diff --git "a/checkpoint-15000/trainer_state.json" "b/checkpoint-15000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-15000/trainer_state.json"
@@ -0,0 +1,105153 @@
+{
+  "best_metric": 0.86865234375,
+  "best_model_checkpoint": "outputs/models/tiny-gpt2_15000_steps_vintage-photography-450k-high-quality-captions/checkpoint-15000",
+  "epoch": 13.84721901684745,
+  "eval_steps": 1000,
+  "global_step": 15000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009231479344564967,
+      "grad_norm": 52.03125,
+      "learning_rate": 1e-05,
+      "loss": 7.1699,
+      "step": 1
+    },
+    {
+      "epoch": 0.0018462958689129934,
+      "grad_norm": 42.8125,
+      "learning_rate": 2e-05,
+      "loss": 6.9014,
+      "step": 2
+    },
+    {
+      "epoch": 0.0027694438033694898,
+      "grad_norm": 40.09375,
+      "learning_rate": 3e-05,
+      "loss": 6.5547,
+      "step": 3
+    },
+    {
+      "epoch": 0.0036925917378259868,
+      "grad_norm": 35.25,
+      "learning_rate": 4e-05,
+      "loss": 5.5684,
+      "step": 4
+    },
+    {
+      "epoch": 0.004615739672282483,
+      "grad_norm": 34.625,
+      "learning_rate": 5e-05,
+      "loss": 4.3682,
+      "step": 5
+    },
+    {
+      "epoch": 0.0055388876067389795,
+      "grad_norm": 43.59375,
+      "learning_rate": 6e-05,
+      "loss": 5.7031,
+      "step": 6
+    },
+    {
+      "epoch": 0.006462035541195477,
+      "grad_norm": 55.71875,
+      "learning_rate": 7e-05,
+      "loss": 5.252,
+      "step": 7
+    },
+    {
+      "epoch": 0.0073851834756519736,
+      "grad_norm": 28.953125,
+      "learning_rate": 8e-05,
+      "loss": 5.9707,
+      "step": 8
+    },
+    {
+      "epoch": 0.00830833141010847,
+      "grad_norm": 27.21875,
+      "learning_rate": 9e-05,
+      "loss": 4.71,
+      "step": 9
+    },
+    {
+      "epoch": 0.009231479344564966,
+      "grad_norm": 33.71875,
+      "learning_rate": 0.0001,
+      "loss": 4.2095,
+      "step": 10
+    },
+    {
+      "epoch": 0.010154627279021463,
+      "grad_norm": 47.1875,
+      "learning_rate": 9.999332888592395e-05,
+      "loss": 5.5498,
+      "step": 11
+    },
+    {
+      "epoch": 0.011077775213477959,
+      "grad_norm": 76.0625,
+      "learning_rate": 9.998665777184791e-05,
+      "loss": 11.9775,
+      "step": 12
+    },
+    {
+      "epoch": 0.012000923147934456,
+      "grad_norm": 19.734375,
+      "learning_rate": 9.997998665777185e-05,
+      "loss": 4.3613,
+      "step": 13
+    },
+    {
+      "epoch": 0.012924071082390954,
+      "grad_norm": 26.140625,
+      "learning_rate": 9.99733155436958e-05,
+      "loss": 3.915,
+      "step": 14
+    },
+    {
+      "epoch": 0.01384721901684745,
+      "grad_norm": 85.0,
+      "learning_rate": 9.996664442961975e-05,
+      "loss": 7.5703,
+      "step": 15
+    },
+    {
+      "epoch": 0.014770366951303947,
+      "grad_norm": 13.84375,
+      "learning_rate": 9.99599733155437e-05,
+      "loss": 3.7168,
+      "step": 16
+    },
+    {
+      "epoch": 0.015693514885760443,
+      "grad_norm": 31.78125,
+      "learning_rate": 9.995330220146765e-05,
+      "loss": 4.3042,
+      "step": 17
+    },
+    {
+      "epoch": 0.01661666282021694,
+      "grad_norm": 15.671875,
+      "learning_rate": 9.99466310873916e-05,
+      "loss": 3.8384,
+      "step": 18
+    },
+    {
+      "epoch": 0.017539810754673438,
+      "grad_norm": 12.4765625,
+      "learning_rate": 9.993995997331554e-05,
+      "loss": 3.1692,
+      "step": 19
+    },
+    {
+      "epoch": 0.018462958689129932,
+      "grad_norm": 12.890625,
+      "learning_rate": 9.99332888592395e-05,
+      "loss": 2.9233,
+      "step": 20
+    },
+    {
+      "epoch": 0.01938610662358643,
+      "grad_norm": 21.515625,
+      "learning_rate": 9.992661774516345e-05,
+      "loss": 4.0273,
+      "step": 21
+    },
+    {
+      "epoch": 0.020309254558042927,
+      "grad_norm": 11.15625,
+      "learning_rate": 9.991994663108739e-05,
+      "loss": 3.0967,
+      "step": 22
+    },
+    {
+      "epoch": 0.021232402492499424,
+      "grad_norm": 13.4609375,
+      "learning_rate": 9.991327551701135e-05,
+      "loss": 3.4238,
+      "step": 23
+    },
+    {
+      "epoch": 0.022155550426955918,
+      "grad_norm": 11.9921875,
+      "learning_rate": 9.99066044029353e-05,
+      "loss": 3.4771,
+      "step": 24
+    },
+    {
+      "epoch": 0.023078698361412416,
+      "grad_norm": 21.0,
+      "learning_rate": 9.989993328885925e-05,
+      "loss": 3.6611,
+      "step": 25
+    },
+    {
+      "epoch": 0.024001846295868913,
+      "grad_norm": 11.3515625,
+      "learning_rate": 9.98932621747832e-05,
+      "loss": 3.064,
+      "step": 26
+    },
+    {
+      "epoch": 0.02492499423032541,
+      "grad_norm": 19.0,
+      "learning_rate": 9.988659106070714e-05,
+      "loss": 3.7188,
+      "step": 27
+    },
+    {
+      "epoch": 0.025848142164781908,
+      "grad_norm": 15.9296875,
+      "learning_rate": 9.987991994663109e-05,
+      "loss": 3.2114,
+      "step": 28
+    },
+    {
+      "epoch": 0.026771290099238402,
+      "grad_norm": 10.703125,
+      "learning_rate": 9.987324883255505e-05,
+      "loss": 3.167,
+      "step": 29
+    },
+    {
+      "epoch": 0.0276944380336949,
+      "grad_norm": 18.984375,
+      "learning_rate": 9.9866577718479e-05,
+      "loss": 3.4575,
+      "step": 30
+    },
+    {
+      "epoch": 0.028617585968151397,
+      "grad_norm": 13.34375,
+      "learning_rate": 9.985990660440294e-05,
+      "loss": 3.5112,
+      "step": 31
+    },
+    {
+      "epoch": 0.029540733902607894,
+      "grad_norm": 19.359375,
+      "learning_rate": 9.985323549032689e-05,
+      "loss": 2.9614,
+      "step": 32
+    },
+    {
+      "epoch": 0.030463881837064388,
+      "grad_norm": 9.2265625,
+      "learning_rate": 9.984656437625084e-05,
+      "loss": 2.6616,
+      "step": 33
+    },
+    {
+      "epoch": 0.031387029771520886,
+      "grad_norm": 12.0234375,
+      "learning_rate": 9.983989326217479e-05,
+      "loss": 3.1118,
+      "step": 34
+    },
+    {
+      "epoch": 0.03231017770597738,
+      "grad_norm": 14.5234375,
+      "learning_rate": 9.983322214809874e-05,
+      "loss": 2.6875,
+      "step": 35
+    },
+    {
+      "epoch": 0.03323332564043388,
+      "grad_norm": 8.4453125,
+      "learning_rate": 9.982655103402268e-05,
+      "loss": 2.7207,
+      "step": 36
+    },
+    {
+      "epoch": 0.03415647357489038,
+      "grad_norm": 16.453125,
+      "learning_rate": 9.981987991994664e-05,
+      "loss": 3.0981,
+      "step": 37
+    },
+    {
+      "epoch": 0.035079621509346876,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.981320880587059e-05,
+      "loss": 2.7559,
+      "step": 38
+    },
+    {
+      "epoch": 0.036002769443803366,
+      "grad_norm": 12.3984375,
+      "learning_rate": 9.980653769179453e-05,
+      "loss": 2.9282,
+      "step": 39
+    },
+    {
+      "epoch": 0.036925917378259863,
+      "grad_norm": 13.125,
+      "learning_rate": 9.979986657771848e-05,
+      "loss": 2.7461,
+      "step": 40
+    },
+    {
+      "epoch": 0.03784906531271636,
+      "grad_norm": 11.203125,
+      "learning_rate": 9.979319546364244e-05,
+      "loss": 2.894,
+      "step": 41
+    },
+    {
+      "epoch": 0.03877221324717286,
+      "grad_norm": 14.3984375,
+      "learning_rate": 9.978652434956638e-05,
+      "loss": 2.6211,
+      "step": 42
+    },
+    {
+      "epoch": 0.039695361181629356,
+      "grad_norm": 9.4921875,
+      "learning_rate": 9.977985323549033e-05,
+      "loss": 2.6655,
+      "step": 43
+    },
+    {
+      "epoch": 0.04061850911608585,
+      "grad_norm": 18.6875,
+      "learning_rate": 9.977318212141427e-05,
+      "loss": 2.9409,
+      "step": 44
+    },
+    {
+      "epoch": 0.04154165705054235,
+      "grad_norm": 8.078125,
+      "learning_rate": 9.976651100733823e-05,
+      "loss": 2.4751,
+      "step": 45
+    },
+    {
+      "epoch": 0.04246480498499885,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.975983989326218e-05,
+      "loss": 2.7466,
+      "step": 46
+    },
+    {
+      "epoch": 0.043387952919455346,
+      "grad_norm": 11.03125,
+      "learning_rate": 9.975316877918612e-05,
+      "loss": 2.9683,
+      "step": 47
+    },
+    {
+      "epoch": 0.044311100853911836,
+      "grad_norm": 132.625,
+      "learning_rate": 9.974649766511008e-05,
+      "loss": 13.0215,
+      "step": 48
+    },
+    {
+      "epoch": 0.045234248788368334,
+      "grad_norm": 15.4921875,
+      "learning_rate": 9.973982655103403e-05,
+      "loss": 2.7544,
+      "step": 49
+    },
+    {
+      "epoch": 0.04615739672282483,
+      "grad_norm": 8.84375,
+      "learning_rate": 9.973315543695798e-05,
+      "loss": 2.9058,
+      "step": 50
+    },
+    {
+      "epoch": 0.04708054465728133,
+      "grad_norm": 26.921875,
+      "learning_rate": 9.972648432288193e-05,
+      "loss": 3.0752,
+      "step": 51
+    },
+    {
+      "epoch": 0.048003692591737826,
+      "grad_norm": 8.9921875,
+      "learning_rate": 9.971981320880588e-05,
+      "loss": 2.4399,
+      "step": 52
+    },
+    {
+      "epoch": 0.04892684052619432,
+      "grad_norm": 12.4140625,
+      "learning_rate": 9.971314209472983e-05,
+      "loss": 2.4858,
+      "step": 53
+    },
+    {
+      "epoch": 0.04984998846065082,
+      "grad_norm": 8.8046875,
+      "learning_rate": 9.970647098065378e-05,
+      "loss": 2.4041,
+      "step": 54
+    },
+    {
+      "epoch": 0.05077313639510732,
+      "grad_norm": 12.6015625,
+      "learning_rate": 9.969979986657773e-05,
+      "loss": 2.3542,
+      "step": 55
+    },
+    {
+      "epoch": 0.051696284329563816,
+      "grad_norm": 13.234375,
+      "learning_rate": 9.969312875250167e-05,
+      "loss": 2.8442,
+      "step": 56
+    },
+    {
+      "epoch": 0.052619432264020306,
+      "grad_norm": 10.890625,
+      "learning_rate": 9.968645763842563e-05,
+      "loss": 2.4028,
+      "step": 57
+    },
+    {
+      "epoch": 0.053542580198476804,
+      "grad_norm": 11.5859375,
+      "learning_rate": 9.967978652434958e-05,
+      "loss": 2.6738,
+      "step": 58
+    },
+    {
+      "epoch": 0.0544657281329333,
+      "grad_norm": 19.046875,
+      "learning_rate": 9.967311541027352e-05,
+      "loss": 2.9263,
+      "step": 59
+    },
+    {
+      "epoch": 0.0553888760673898,
+      "grad_norm": 11.5,
+      "learning_rate": 9.966644429619747e-05,
+      "loss": 2.8789,
+      "step": 60
+    },
+    {
+      "epoch": 0.056312024001846296,
+      "grad_norm": 22.359375,
+      "learning_rate": 9.965977318212143e-05,
+      "loss": 2.9526,
+      "step": 61
+    },
+    {
+      "epoch": 0.057235171936302794,
+      "grad_norm": 8.8203125,
+      "learning_rate": 9.965310206804537e-05,
+      "loss": 2.2493,
+      "step": 62
+    },
+    {
+      "epoch": 0.05815831987075929,
+      "grad_norm": 19.4375,
+      "learning_rate": 9.964643095396932e-05,
+      "loss": 2.7349,
+      "step": 63
+    },
+    {
+      "epoch": 0.05908146780521579,
+      "grad_norm": 11.46875,
+      "learning_rate": 9.963975983989326e-05,
+      "loss": 2.5508,
+      "step": 64
+    },
+    {
+      "epoch": 0.06000461573967228,
+      "grad_norm": 14.9375,
+      "learning_rate": 9.963308872581722e-05,
+      "loss": 2.7378,
+      "step": 65
+    },
+    {
+      "epoch": 0.060927763674128776,
+      "grad_norm": 11.6328125,
+      "learning_rate": 9.962641761174117e-05,
+      "loss": 2.5464,
+      "step": 66
+    },
+    {
+      "epoch": 0.061850911608585274,
+      "grad_norm": 17.359375,
+      "learning_rate": 9.961974649766511e-05,
+      "loss": 2.4465,
+      "step": 67
+    },
+    {
+      "epoch": 0.06277405954304177,
+      "grad_norm": 8.609375,
+      "learning_rate": 9.961307538358906e-05,
+      "loss": 2.5527,
+      "step": 68
+    },
+    {
+      "epoch": 0.06369720747749827,
+      "grad_norm": 26.96875,
+      "learning_rate": 9.960640426951302e-05,
+      "loss": 3.0923,
+      "step": 69
+    },
+    {
+      "epoch": 0.06462035541195477,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.959973315543696e-05,
+      "loss": 2.095,
+      "step": 70
+    },
+    {
+      "epoch": 0.06554350334641126,
+      "grad_norm": 14.2890625,
+      "learning_rate": 9.95930620413609e-05,
+      "loss": 2.7261,
+      "step": 71
+    },
+    {
+      "epoch": 0.06646665128086776,
+      "grad_norm": 9.3046875,
+      "learning_rate": 9.958639092728485e-05,
+      "loss": 2.5474,
+      "step": 72
+    },
+    {
+      "epoch": 0.06738979921532426,
+      "grad_norm": 55.71875,
+      "learning_rate": 9.957971981320881e-05,
+      "loss": 3.0283,
+      "step": 73
+    },
+    {
+      "epoch": 0.06831294714978076,
+      "grad_norm": 85.5,
+      "learning_rate": 9.957304869913277e-05,
+      "loss": 4.8975,
+      "step": 74
+    },
+    {
+      "epoch": 0.06923609508423725,
+      "grad_norm": 11.765625,
+      "learning_rate": 9.956637758505672e-05,
+      "loss": 2.3359,
+      "step": 75
+    },
+    {
+      "epoch": 0.07015924301869375,
+      "grad_norm": 10.5078125,
+      "learning_rate": 9.955970647098066e-05,
+      "loss": 2.6694,
+      "step": 76
+    },
+    {
+      "epoch": 0.07108239095315025,
+      "grad_norm": 9.625,
+      "learning_rate": 9.95530353569046e-05,
+      "loss": 2.0156,
+      "step": 77
+    },
+    {
+      "epoch": 0.07200553888760673,
+      "grad_norm": 10.40625,
+      "learning_rate": 9.954636424282857e-05,
+      "loss": 2.0667,
+      "step": 78
+    },
+    {
+      "epoch": 0.07292868682206323,
+      "grad_norm": 10.9453125,
+      "learning_rate": 9.953969312875251e-05,
+      "loss": 2.519,
+      "step": 79
+    },
+    {
+      "epoch": 0.07385183475651973,
+      "grad_norm": 10.7890625,
+      "learning_rate": 9.953302201467646e-05,
+      "loss": 2.7056,
+      "step": 80
+    },
+    {
+      "epoch": 0.07477498269097622,
+      "grad_norm": 12.71875,
+      "learning_rate": 9.95263509006004e-05,
+      "loss": 1.5345,
+      "step": 81
+    },
+    {
+      "epoch": 0.07569813062543272,
+      "grad_norm": 8.578125,
+      "learning_rate": 9.951967978652436e-05,
+      "loss": 2.5225,
+      "step": 82
+    },
+    {
+      "epoch": 0.07662127855988922,
+      "grad_norm": 15.953125,
+      "learning_rate": 9.95130086724483e-05,
+      "loss": 2.6006,
+      "step": 83
+    },
+    {
+      "epoch": 0.07754442649434572,
+      "grad_norm": 8.9375,
+      "learning_rate": 9.950633755837225e-05,
+      "loss": 2.5767,
+      "step": 84
+    },
+    {
+      "epoch": 0.07846757442880221,
+      "grad_norm": 26.328125,
+      "learning_rate": 9.94996664442962e-05,
+      "loss": 2.8804,
+      "step": 85
+    },
+    {
+      "epoch": 0.07939072236325871,
+      "grad_norm": 6.75,
+      "learning_rate": 9.949299533022016e-05,
+      "loss": 2.2729,
+      "step": 86
+    },
+    {
+      "epoch": 0.08031387029771521,
+      "grad_norm": 25.953125,
+      "learning_rate": 9.94863242161441e-05,
+      "loss": 2.9995,
+      "step": 87
+    },
+    {
+      "epoch": 0.0812370182321717,
+      "grad_norm": 8.0078125,
+      "learning_rate": 9.947965310206805e-05,
+      "loss": 2.4194,
+      "step": 88
+    },
+    {
+      "epoch": 0.0821601661666282,
+      "grad_norm": 14.0390625,
+      "learning_rate": 9.947298198799199e-05,
+      "loss": 2.3035,
+      "step": 89
+    },
+    {
+      "epoch": 0.0830833141010847,
+      "grad_norm": 8.171875,
+      "learning_rate": 9.946631087391595e-05,
+      "loss": 2.2556,
+      "step": 90
+    },
+    {
+      "epoch": 0.0840064620355412,
+      "grad_norm": 21.640625,
+      "learning_rate": 9.94596397598399e-05,
+      "loss": 2.5225,
+      "step": 91
+    },
+    {
+      "epoch": 0.0849296099699977,
+      "grad_norm": 5.3515625,
+      "learning_rate": 9.945296864576384e-05,
+      "loss": 1.4491,
+      "step": 92
+    },
+    {
+      "epoch": 0.0858527579044542,
+      "grad_norm": 14.1015625,
+      "learning_rate": 9.944629753168779e-05,
+      "loss": 2.4536,
+      "step": 93
+    },
+    {
+      "epoch": 0.08677590583891069,
+      "grad_norm": 8.6171875,
+      "learning_rate": 9.943962641761175e-05,
+      "loss": 2.2156,
+      "step": 94
+    },
+    {
+      "epoch": 0.08769905377336719,
+      "grad_norm": 21.640625,
+      "learning_rate": 9.943295530353569e-05,
+      "loss": 2.5742,
+      "step": 95
+    },
+    {
+      "epoch": 0.08862220170782367,
+      "grad_norm": 7.06640625,
+      "learning_rate": 9.942628418945964e-05,
+      "loss": 2.2368,
+      "step": 96
+    },
+    {
+      "epoch": 0.08954534964228017,
+      "grad_norm": 16.9375,
+      "learning_rate": 9.941961307538358e-05,
+      "loss": 2.7134,
+      "step": 97
+    },
+    {
+      "epoch": 0.09046849757673667,
+      "grad_norm": 7.69140625,
+      "learning_rate": 9.941294196130756e-05,
+      "loss": 2.0747,
+      "step": 98
+    },
+    {
+      "epoch": 0.09139164551119316,
+      "grad_norm": 10.3671875,
+      "learning_rate": 9.94062708472315e-05,
+      "loss": 2.1924,
+      "step": 99
+    },
+    {
+      "epoch": 0.09231479344564966,
+      "grad_norm": 8.5859375,
+      "learning_rate": 9.939959973315545e-05,
+      "loss": 2.3035,
+      "step": 100
+    },
+    {
+      "epoch": 0.09323794138010616,
+      "grad_norm": 15.921875,
+      "learning_rate": 9.939292861907939e-05,
+      "loss": 2.4563,
+      "step": 101
+    },
+    {
+      "epoch": 0.09416108931456266,
+      "grad_norm": 8.828125,
+      "learning_rate": 9.938625750500335e-05,
+      "loss": 2.3079,
+      "step": 102
+    },
+    {
+      "epoch": 0.09508423724901915,
+      "grad_norm": 8.59375,
+      "learning_rate": 9.93795863909273e-05,
+      "loss": 2.0386,
+      "step": 103
+    },
+    {
+      "epoch": 0.09600738518347565,
+      "grad_norm": 14.5078125,
+      "learning_rate": 9.937291527685124e-05,
+      "loss": 2.4775,
+      "step": 104
+    },
+    {
+      "epoch": 0.09693053311793215,
+      "grad_norm": 6.2890625,
+      "learning_rate": 9.936624416277519e-05,
+      "loss": 1.8246,
+      "step": 105
+    },
+    {
+      "epoch": 0.09785368105238865,
+      "grad_norm": 10.3828125,
+      "learning_rate": 9.935957304869915e-05,
+      "loss": 2.2129,
+      "step": 106
+    },
+    {
+      "epoch": 0.09877682898684514,
+      "grad_norm": 5.89453125,
+      "learning_rate": 9.935290193462309e-05,
+      "loss": 2.0571,
+      "step": 107
+    },
+    {
+      "epoch": 0.09969997692130164,
+      "grad_norm": 27.109375,
+      "learning_rate": 9.934623082054704e-05,
+      "loss": 2.7046,
+      "step": 108
+    },
+    {
+      "epoch": 0.10062312485575814,
+      "grad_norm": 5.27734375,
+      "learning_rate": 9.933955970647098e-05,
+      "loss": 2.5386,
+      "step": 109
+    },
+    {
+      "epoch": 0.10154627279021464,
+      "grad_norm": 33.53125,
+      "learning_rate": 9.933288859239494e-05,
+      "loss": 2.5601,
+      "step": 110
+    },
+    {
+      "epoch": 0.10246942072467113,
+      "grad_norm": 6.64453125,
+      "learning_rate": 9.932621747831889e-05,
+      "loss": 2.3281,
+      "step": 111
+    },
+    {
+      "epoch": 0.10339256865912763,
+      "grad_norm": 14.4609375,
+      "learning_rate": 9.931954636424283e-05,
+      "loss": 2.6001,
+      "step": 112
+    },
+    {
+      "epoch": 0.10431571659358412,
+      "grad_norm": 7.484375,
+      "learning_rate": 9.931287525016678e-05,
+      "loss": 2.2861,
+      "step": 113
+    },
+    {
+      "epoch": 0.10523886452804061,
+      "grad_norm": 12.6328125,
+      "learning_rate": 9.930620413609074e-05,
+      "loss": 2.1816,
+      "step": 114
+    },
+    {
+      "epoch": 0.10616201246249711,
+      "grad_norm": 6.88671875,
+      "learning_rate": 9.929953302201468e-05,
+      "loss": 2.4395,
+      "step": 115
+    },
+    {
+      "epoch": 0.10708516039695361,
+      "grad_norm": 24.171875,
+      "learning_rate": 9.929286190793863e-05,
+      "loss": 2.5938,
+      "step": 116
+    },
+    {
+      "epoch": 0.1080083083314101,
+      "grad_norm": 6.58203125,
+      "learning_rate": 9.928619079386257e-05,
+      "loss": 2.1472,
+      "step": 117
+    },
+    {
+      "epoch": 0.1089314562658666,
+      "grad_norm": 20.203125,
+      "learning_rate": 9.927951967978653e-05,
+      "loss": 2.5078,
+      "step": 118
+    },
+    {
+      "epoch": 0.1098546042003231,
+      "grad_norm": 5.81640625,
+      "learning_rate": 9.927284856571048e-05,
+      "loss": 1.7283,
+      "step": 119
+    },
+    {
+      "epoch": 0.1107777521347796,
+      "grad_norm": 15.09375,
+      "learning_rate": 9.926617745163442e-05,
+      "loss": 1.8838,
+      "step": 120
+    },
+    {
+      "epoch": 0.1117009000692361,
+      "grad_norm": 5.95703125,
+      "learning_rate": 9.925950633755837e-05,
+      "loss": 2.2444,
+      "step": 121
+    },
+    {
+      "epoch": 0.11262404800369259,
+      "grad_norm": 20.125,
+      "learning_rate": 9.925283522348233e-05,
+      "loss": 2.3965,
+      "step": 122
+    },
+    {
+      "epoch": 0.11354719593814909,
+      "grad_norm": 6.2890625,
+      "learning_rate": 9.924616410940629e-05,
+      "loss": 2.3726,
+      "step": 123
+    },
+    {
+      "epoch": 0.11447034387260559,
+      "grad_norm": 17.140625,
+      "learning_rate": 9.923949299533023e-05,
+      "loss": 2.2402,
+      "step": 124
+    },
+    {
+      "epoch": 0.11539349180706208,
+      "grad_norm": 6.5625,
+      "learning_rate": 9.923282188125418e-05,
+      "loss": 2.1895,
+      "step": 125
+    },
+    {
+      "epoch": 0.11631663974151858,
+      "grad_norm": 10.0234375,
+      "learning_rate": 9.922615076717812e-05,
+      "loss": 1.6777,
+      "step": 126
+    },
+    {
+      "epoch": 0.11723978767597508,
+      "grad_norm": 8.0703125,
+      "learning_rate": 9.921947965310208e-05,
+      "loss": 2.0073,
+      "step": 127
+    },
+    {
+      "epoch": 0.11816293561043158,
+      "grad_norm": 12.3984375,
+      "learning_rate": 9.921280853902603e-05,
+      "loss": 2.1926,
+      "step": 128
+    },
+    {
+      "epoch": 0.11908608354488807,
+      "grad_norm": 7.1953125,
+      "learning_rate": 9.920613742494997e-05,
+      "loss": 1.8434,
+      "step": 129
+    },
+    {
+      "epoch": 0.12000923147934456,
+      "grad_norm": 9.21875,
+      "learning_rate": 9.919946631087392e-05,
+      "loss": 2.105,
+      "step": 130
+    },
+    {
+      "epoch": 0.12093237941380106,
+      "grad_norm": 7.85546875,
+      "learning_rate": 9.919279519679788e-05,
+      "loss": 2.0217,
+      "step": 131
+    },
+    {
+      "epoch": 0.12185552734825755,
+      "grad_norm": 13.6953125,
+      "learning_rate": 9.918612408272182e-05,
+      "loss": 2.377,
+      "step": 132
+    },
+    {
+      "epoch": 0.12277867528271405,
+      "grad_norm": 6.48828125,
+      "learning_rate": 9.917945296864577e-05,
+      "loss": 2.1794,
+      "step": 133
+    },
+    {
+      "epoch": 0.12370182321717055,
+      "grad_norm": 14.8046875,
+      "learning_rate": 9.917278185456971e-05,
+      "loss": 2.373,
+      "step": 134
+    },
+    {
+      "epoch": 0.12462497115162705,
+      "grad_norm": 6.1640625,
+      "learning_rate": 9.916611074049367e-05,
+      "loss": 2.0894,
+      "step": 135
+    },
+    {
+      "epoch": 0.12554811908608354,
+      "grad_norm": 12.0234375,
+      "learning_rate": 9.915943962641762e-05,
+      "loss": 2.2012,
+      "step": 136
+    },
+    {
+      "epoch": 0.12647126702054004,
+      "grad_norm": 9.2109375,
+      "learning_rate": 9.915276851234156e-05,
+      "loss": 2.3765,
+      "step": 137
+    },
+    {
+      "epoch": 0.12739441495499654,
+      "grad_norm": 11.7734375,
+      "learning_rate": 9.914609739826551e-05,
+      "loss": 1.9963,
+      "step": 138
+    },
+    {
+      "epoch": 0.12831756288945304,
+      "grad_norm": 6.06640625,
+      "learning_rate": 9.913942628418947e-05,
+      "loss": 2.3633,
+      "step": 139
+    },
+    {
+      "epoch": 0.12924071082390953,
+      "grad_norm": 16.75,
+      "learning_rate": 9.913275517011341e-05,
+      "loss": 2.4951,
+      "step": 140
+    },
+    {
+      "epoch": 0.13016385875836603,
+      "grad_norm": 4.94140625,
+      "learning_rate": 9.912608405603736e-05,
+      "loss": 2.2407,
+      "step": 141
+    },
+    {
+      "epoch": 0.13108700669282253,
+      "grad_norm": 16.5,
+      "learning_rate": 9.91194129419613e-05,
+      "loss": 2.4033,
+      "step": 142
+    },
+    {
+      "epoch": 0.13201015462727902,
+      "grad_norm": 5.66015625,
+      "learning_rate": 9.911274182788526e-05,
+      "loss": 2.397,
+      "step": 143
+    },
+    {
+      "epoch": 0.13293330256173552,
+      "grad_norm": 21.796875,
+      "learning_rate": 9.910607071380921e-05,
+      "loss": 2.4878,
+      "step": 144
+    },
+    {
+      "epoch": 0.13385645049619202,
+      "grad_norm": 5.15234375,
+      "learning_rate": 9.909939959973315e-05,
+      "loss": 1.854,
+      "step": 145
+    },
+    {
+      "epoch": 0.13477959843064852,
+      "grad_norm": 10.2890625,
+      "learning_rate": 9.90927284856571e-05,
+      "loss": 2.5552,
+      "step": 146
+    },
+    {
+      "epoch": 0.13570274636510501,
+      "grad_norm": 5.109375,
+      "learning_rate": 9.908605737158106e-05,
+      "loss": 1.6755,
+      "step": 147
+    },
+    {
+      "epoch": 0.1366258942995615,
+      "grad_norm": 10.1640625,
+      "learning_rate": 9.907938625750502e-05,
+      "loss": 2.0627,
+      "step": 148
+    },
+    {
+      "epoch": 0.137549042234018,
+      "grad_norm": 7.44921875,
+      "learning_rate": 9.907271514342896e-05,
+      "loss": 2.2327,
+      "step": 149
+    },
+    {
+      "epoch": 0.1384721901684745,
+      "grad_norm": 10.921875,
+      "learning_rate": 9.906604402935291e-05,
+      "loss": 2.5244,
+      "step": 150
+    },
+    {
+      "epoch": 0.139395338102931,
+      "grad_norm": 7.48046875,
+      "learning_rate": 9.905937291527687e-05,
+      "loss": 2.2051,
+      "step": 151
+    },
+    {
+      "epoch": 0.1403184860373875,
+      "grad_norm": 8.828125,
+      "learning_rate": 9.905270180120081e-05,
+      "loss": 1.7011,
+      "step": 152
+    },
+    {
+      "epoch": 0.141241633971844,
+      "grad_norm": 9.6484375,
+      "learning_rate": 9.904603068712476e-05,
+      "loss": 2.2942,
+      "step": 153
+    },
+    {
+      "epoch": 0.1421647819063005,
+      "grad_norm": 8.3046875,
+      "learning_rate": 9.90393595730487e-05,
+      "loss": 1.9937,
+      "step": 154
+    },
+    {
+      "epoch": 0.143087929840757,
+      "grad_norm": 11.234375,
+      "learning_rate": 9.903268845897266e-05,
+      "loss": 1.9941,
+      "step": 155
+    },
+    {
+      "epoch": 0.14401107777521346,
+      "grad_norm": 7.37109375,
+      "learning_rate": 9.902601734489661e-05,
+      "loss": 2.0952,
+      "step": 156
+    },
+    {
+      "epoch": 0.14493422570966996,
+      "grad_norm": 9.9609375,
+      "learning_rate": 9.901934623082055e-05,
+      "loss": 2.1399,
+      "step": 157
+    },
+    {
+      "epoch": 0.14585737364412646,
+      "grad_norm": 8.5390625,
+      "learning_rate": 9.90126751167445e-05,
+      "loss": 1.9287,
+      "step": 158
+    },
+    {
+      "epoch": 0.14678052157858296,
+      "grad_norm": 10.2578125,
+      "learning_rate": 9.900600400266846e-05,
+      "loss": 2.0864,
+      "step": 159
+    },
+    {
+      "epoch": 0.14770366951303945,
+      "grad_norm": 7.52734375,
+      "learning_rate": 9.89993328885924e-05,
+      "loss": 2.0796,
+      "step": 160
+    },
+    {
+      "epoch": 0.14862681744749595,
+      "grad_norm": 9.296875,
+      "learning_rate": 9.899266177451635e-05,
+      "loss": 1.743,
+      "step": 161
+    },
+    {
+      "epoch": 0.14954996538195245,
+      "grad_norm": 7.796875,
+      "learning_rate": 9.898599066044029e-05,
+      "loss": 2.0293,
+      "step": 162
+    },
+    {
+      "epoch": 0.15047311331640895,
+      "grad_norm": 12.859375,
+      "learning_rate": 9.897931954636425e-05,
+      "loss": 2.4751,
+      "step": 163
+    },
+    {
+      "epoch": 0.15139626125086544,
+      "grad_norm": 8.6796875,
+      "learning_rate": 9.89726484322882e-05,
+      "loss": 1.9666,
+      "step": 164
+    },
+    {
+      "epoch": 0.15231940918532194,
+      "grad_norm": 7.921875,
+      "learning_rate": 9.896597731821214e-05,
+      "loss": 1.8916,
+      "step": 165
+    },
+    {
+      "epoch": 0.15324255711977844,
+      "grad_norm": 10.328125,
+      "learning_rate": 9.895930620413609e-05,
+      "loss": 2.2808,
+      "step": 166
+    },
+    {
+      "epoch": 0.15416570505423494,
+      "grad_norm": 8.2890625,
+      "learning_rate": 9.895263509006005e-05,
+      "loss": 2.2419,
+      "step": 167
+    },
+    {
+      "epoch": 0.15508885298869143,
+      "grad_norm": 10.265625,
+      "learning_rate": 9.894596397598399e-05,
+      "loss": 1.9714,
+      "step": 168
+    },
+    {
+      "epoch": 0.15601200092314793,
+      "grad_norm": 7.5,
+      "learning_rate": 9.893929286190794e-05,
+      "loss": 2.2368,
+      "step": 169
+    },
+    {
+      "epoch": 0.15693514885760443,
+      "grad_norm": 9.9296875,
+      "learning_rate": 9.893262174783188e-05,
+      "loss": 2.0728,
+      "step": 170
+    },
+    {
+      "epoch": 0.15785829679206093,
+      "grad_norm": 6.91796875,
+      "learning_rate": 9.892595063375584e-05,
+      "loss": 1.9675,
+      "step": 171
+    },
+    {
+      "epoch": 0.15878144472651742,
+      "grad_norm": 8.5,
+      "learning_rate": 9.891927951967979e-05,
+      "loss": 1.7405,
+      "step": 172
+    },
+    {
+      "epoch": 0.15970459266097392,
+      "grad_norm": 7.671875,
+      "learning_rate": 9.891260840560375e-05,
+      "loss": 2.1221,
+      "step": 173
+    },
+    {
+      "epoch": 0.16062774059543042,
+      "grad_norm": 10.90625,
+      "learning_rate": 9.890593729152769e-05,
+      "loss": 2.2566,
+      "step": 174
+    },
+    {
+      "epoch": 0.16155088852988692,
+      "grad_norm": 7.55859375,
+      "learning_rate": 9.889926617745164e-05,
+      "loss": 2.4946,
+      "step": 175
+    },
+    {
+      "epoch": 0.1624740364643434,
+      "grad_norm": 17.125,
+      "learning_rate": 9.88925950633756e-05,
+      "loss": 2.3101,
+      "step": 176
+    },
+    {
+      "epoch": 0.1633971843987999,
+      "grad_norm": 6.4375,
+      "learning_rate": 9.888592394929954e-05,
+      "loss": 2.2021,
+      "step": 177
+    },
+    {
+      "epoch": 0.1643203323332564,
+      "grad_norm": 17.078125,
+      "learning_rate": 9.887925283522349e-05,
+      "loss": 2.1438,
+      "step": 178
+    },
+    {
+      "epoch": 0.1652434802677129,
+      "grad_norm": 6.91796875,
+      "learning_rate": 9.887258172114743e-05,
+      "loss": 2.0947,
+      "step": 179
+    },
+    {
+      "epoch": 0.1661666282021694,
+      "grad_norm": 12.6796875,
+      "learning_rate": 9.886591060707139e-05,
+      "loss": 1.5601,
+      "step": 180
+    },
+    {
+      "epoch": 0.1670897761366259,
+      "grad_norm": 7.99609375,
+      "learning_rate": 9.885923949299534e-05,
+      "loss": 2.0188,
+      "step": 181
+    },
+    {
+      "epoch": 0.1680129240710824,
+      "grad_norm": 9.1328125,
+      "learning_rate": 9.885256837891928e-05,
+      "loss": 1.7489,
+      "step": 182
+    },
+    {
+      "epoch": 0.1689360720055389,
+      "grad_norm": 5.0,
+      "learning_rate": 9.884589726484323e-05,
+      "loss": 1.9922,
+      "step": 183
+    },
+    {
+      "epoch": 0.1698592199399954,
+      "grad_norm": 12.9296875,
+      "learning_rate": 9.883922615076719e-05,
+      "loss": 2.2686,
+      "step": 184
+    },
+    {
+      "epoch": 0.1707823678744519,
+      "grad_norm": 6.53515625,
+      "learning_rate": 9.883255503669113e-05,
+      "loss": 2.0645,
+      "step": 185
+    },
+    {
+      "epoch": 0.1717055158089084,
+      "grad_norm": 13.171875,
+      "learning_rate": 9.882588392261508e-05,
+      "loss": 2.2109,
+      "step": 186
+    },
+    {
+      "epoch": 0.17262866374336489,
+      "grad_norm": 6.90625,
+      "learning_rate": 9.881921280853902e-05,
+      "loss": 1.7542,
+      "step": 187
+    },
+    {
+      "epoch": 0.17355181167782138,
+      "grad_norm": 10.984375,
+      "learning_rate": 9.881254169446298e-05,
+      "loss": 2.1797,
+      "step": 188
+    },
+    {
+      "epoch": 0.17447495961227788,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.880587058038693e-05,
+      "loss": 1.7866,
+      "step": 189
+    },
+    {
+      "epoch": 0.17539810754673438,
+      "grad_norm": 14.96875,
+      "learning_rate": 9.879919946631087e-05,
+      "loss": 2.2793,
+      "step": 190
+    },
+    {
+      "epoch": 0.17632125548119085,
+      "grad_norm": 5.41796875,
+      "learning_rate": 9.879252835223482e-05,
+      "loss": 1.6826,
+      "step": 191
+    },
+    {
+      "epoch": 0.17724440341564734,
+      "grad_norm": 13.609375,
+      "learning_rate": 9.878585723815878e-05,
+      "loss": 2.3306,
+      "step": 192
+    },
+    {
+      "epoch": 0.17816755135010384,
+      "grad_norm": 8.8828125,
+      "learning_rate": 9.877918612408272e-05,
+      "loss": 2.0718,
+      "step": 193
+    },
+    {
+      "epoch": 0.17909069928456034,
+      "grad_norm": 10.5859375,
+      "learning_rate": 9.877251501000667e-05,
+      "loss": 2.0239,
+      "step": 194
+    },
+    {
+      "epoch": 0.18001384721901684,
+      "grad_norm": 6.79296875,
+      "learning_rate": 9.876584389593061e-05,
+      "loss": 1.8589,
+      "step": 195
+    },
+    {
+      "epoch": 0.18093699515347333,
+      "grad_norm": 9.6328125,
+      "learning_rate": 9.875917278185457e-05,
+      "loss": 1.9973,
+      "step": 196
+    },
+    {
+      "epoch": 0.18186014308792983,
+      "grad_norm": 7.77734375,
+      "learning_rate": 9.875250166777852e-05,
+      "loss": 2.2158,
+      "step": 197
+    },
+    {
+      "epoch": 0.18278329102238633,
+      "grad_norm": 10.2421875,
+      "learning_rate": 9.874583055370246e-05,
+      "loss": 2.5435,
+      "step": 198
+    },
+    {
+      "epoch": 0.18370643895684283,
+      "grad_norm": 9.4375,
+      "learning_rate": 9.873915943962642e-05,
+      "loss": 2.1436,
+      "step": 199
+    },
+    {
+      "epoch": 0.18462958689129932,
+      "grad_norm": 5.36328125,
+      "learning_rate": 9.873248832555038e-05,
+      "loss": 2.1038,
+      "step": 200
+    },
+    {
+      "epoch": 0.18555273482575582,
+      "grad_norm": 8.390625,
+      "learning_rate": 9.872581721147433e-05,
+      "loss": 2.0076,
+      "step": 201
+    },
+    {
+      "epoch": 0.18647588276021232,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.871914609739827e-05,
+      "loss": 1.896,
+      "step": 202
+    },
+    {
+      "epoch": 0.18739903069466882,
+      "grad_norm": 10.3046875,
+      "learning_rate": 9.871247498332222e-05,
+      "loss": 2.1335,
+      "step": 203
+    },
+    {
+      "epoch": 0.18832217862912531,
+      "grad_norm": 8.8671875,
+      "learning_rate": 9.870580386924618e-05,
+      "loss": 2.1733,
+      "step": 204
+    },
+    {
+      "epoch": 0.1892453265635818,
+      "grad_norm": 9.6875,
+      "learning_rate": 9.869913275517012e-05,
+      "loss": 2.0938,
+      "step": 205
+    },
+    {
+      "epoch": 0.1901684744980383,
+      "grad_norm": 7.328125,
+      "learning_rate": 9.869246164109407e-05,
+      "loss": 2.0322,
+      "step": 206
+    },
+    {
+      "epoch": 0.1910916224324948,
+      "grad_norm": 10.953125,
+      "learning_rate": 9.868579052701801e-05,
+      "loss": 2.2046,
+      "step": 207
+    },
+    {
+      "epoch": 0.1920147703669513,
+      "grad_norm": 6.015625,
+      "learning_rate": 9.867911941294197e-05,
+      "loss": 1.866,
+      "step": 208
+    },
+    {
+      "epoch": 0.1929379183014078,
+      "grad_norm": 9.640625,
+      "learning_rate": 9.867244829886592e-05,
+      "loss": 2.0732,
+      "step": 209
+    },
+    {
+      "epoch": 0.1938610662358643,
+      "grad_norm": 6.91796875,
+      "learning_rate": 9.866577718478986e-05,
+      "loss": 1.8025,
+      "step": 210
+    },
+    {
+      "epoch": 0.1947842141703208,
+      "grad_norm": 8.4453125,
+      "learning_rate": 9.865910607071381e-05,
+      "loss": 1.7649,
+      "step": 211
+    },
+    {
+      "epoch": 0.1957073621047773,
+      "grad_norm": 6.015625,
+      "learning_rate": 9.865243495663777e-05,
+      "loss": 2.1985,
+      "step": 212
+    },
+    {
+      "epoch": 0.1966305100392338,
+      "grad_norm": 13.671875,
+      "learning_rate": 9.864576384256171e-05,
+      "loss": 2.1504,
+      "step": 213
+    },
+    {
+      "epoch": 0.1975536579736903,
+      "grad_norm": 6.046875,
+      "learning_rate": 9.863909272848566e-05,
+      "loss": 1.7773,
+      "step": 214
+    },
+    {
+      "epoch": 0.1984768059081468,
+      "grad_norm": 11.078125,
+      "learning_rate": 9.86324216144096e-05,
+      "loss": 2.1885,
+      "step": 215
+    },
+    {
+      "epoch": 0.19939995384260328,
+      "grad_norm": 6.4296875,
+      "learning_rate": 9.862575050033356e-05,
+      "loss": 2.1338,
+      "step": 216
+    },
+    {
+      "epoch": 0.20032310177705978,
+      "grad_norm": 13.828125,
+      "learning_rate": 9.861907938625751e-05,
+      "loss": 1.8284,
+      "step": 217
+    },
+    {
+      "epoch": 0.20124624971151628,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.861240827218145e-05,
+      "loss": 1.9434,
+      "step": 218
+    },
+    {
+      "epoch": 0.20216939764597278,
+      "grad_norm": 11.375,
+      "learning_rate": 9.86057371581054e-05,
+      "loss": 2.1895,
+      "step": 219
+    },
+    {
+      "epoch": 0.20309254558042927,
+      "grad_norm": 6.14453125,
+      "learning_rate": 9.859906604402936e-05,
+      "loss": 1.783,
+      "step": 220
+    },
+    {
+      "epoch": 0.20401569351488577,
+      "grad_norm": 9.21875,
+      "learning_rate": 9.85923949299533e-05,
+      "loss": 1.6179,
+      "step": 221
+    },
+    {
+      "epoch": 0.20493884144934227,
+      "grad_norm": 6.63671875,
+      "learning_rate": 9.858572381587725e-05,
+      "loss": 1.5516,
+      "step": 222
+    },
+    {
+      "epoch": 0.20586198938379877,
+      "grad_norm": 8.6796875,
+      "learning_rate": 9.85790527018012e-05,
+      "loss": 1.812,
+      "step": 223
+    },
+    {
+      "epoch": 0.20678513731825526,
+      "grad_norm": 8.0546875,
+      "learning_rate": 9.857238158772515e-05,
+      "loss": 2.0442,
+      "step": 224
+    },
+    {
+      "epoch": 0.20770828525271176,
+      "grad_norm": 7.828125,
+      "learning_rate": 9.856571047364911e-05,
+      "loss": 1.6648,
+      "step": 225
+    },
+    {
+      "epoch": 0.20863143318716823,
+      "grad_norm": 9.0,
+      "learning_rate": 9.855903935957306e-05,
+      "loss": 1.9956,
+      "step": 226
+    },
+    {
+      "epoch": 0.20955458112162473,
+      "grad_norm": 8.6328125,
+      "learning_rate": 9.8552368245497e-05,
+      "loss": 2.0923,
+      "step": 227
+    },
+    {
+      "epoch": 0.21047772905608123,
+      "grad_norm": 7.171875,
+      "learning_rate": 9.854569713142095e-05,
+      "loss": 1.8752,
+      "step": 228
+    },
+    {
+      "epoch": 0.21140087699053772,
+      "grad_norm": 10.6015625,
+      "learning_rate": 9.853902601734491e-05,
+      "loss": 1.9014,
+      "step": 229
+    },
+    {
+      "epoch": 0.21232402492499422,
+      "grad_norm": 6.56640625,
+      "learning_rate": 9.853235490326885e-05,
+      "loss": 2.0352,
+      "step": 230
+    },
+    {
+      "epoch": 0.21324717285945072,
+      "grad_norm": 8.5703125,
+      "learning_rate": 9.85256837891928e-05,
+      "loss": 1.886,
+      "step": 231
+    },
+    {
+      "epoch": 0.21417032079390722,
+      "grad_norm": 7.7890625,
+      "learning_rate": 9.851901267511674e-05,
+      "loss": 1.8359,
+      "step": 232
+    },
+    {
+      "epoch": 0.2150934687283637,
+      "grad_norm": 9.4453125,
+      "learning_rate": 9.85123415610407e-05,
+      "loss": 2.1206,
+      "step": 233
+    },
+    {
+      "epoch": 0.2160166166628202,
+      "grad_norm": 5.78515625,
+      "learning_rate": 9.850567044696465e-05,
+      "loss": 1.9128,
+      "step": 234
+    },
+    {
+      "epoch": 0.2169397645972767,
+      "grad_norm": 7.55859375,
+      "learning_rate": 9.84989993328886e-05,
+      "loss": 1.8325,
+      "step": 235
+    },
+    {
+      "epoch": 0.2178629125317332,
+      "grad_norm": 7.1328125,
+      "learning_rate": 9.849232821881255e-05,
+      "loss": 2.0356,
+      "step": 236
+    },
+    {
+      "epoch": 0.2187860604661897,
+      "grad_norm": 10.2890625,
+      "learning_rate": 9.84856571047365e-05,
+      "loss": 1.9856,
+      "step": 237
+    },
+    {
+      "epoch": 0.2197092084006462,
+      "grad_norm": 7.74609375,
+      "learning_rate": 9.847898599066044e-05,
+      "loss": 2.2183,
+      "step": 238
+    },
+    {
+      "epoch": 0.2206323563351027,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.847231487658439e-05,
+      "loss": 1.7983,
+      "step": 239
+    },
+    {
+      "epoch": 0.2215555042695592,
+      "grad_norm": 7.97265625,
+      "learning_rate": 9.846564376250835e-05,
+      "loss": 2.0393,
+      "step": 240
+    },
+    {
+      "epoch": 0.2224786522040157,
+      "grad_norm": 10.125,
+      "learning_rate": 9.84589726484323e-05,
+      "loss": 1.8582,
+      "step": 241
+    },
+    {
+      "epoch": 0.2234018001384722,
+      "grad_norm": 8.3515625,
+      "learning_rate": 9.845230153435624e-05,
+      "loss": 2.0391,
+      "step": 242
+    },
+    {
+      "epoch": 0.2243249480729287,
+      "grad_norm": 9.640625,
+      "learning_rate": 9.844563042028018e-05,
+      "loss": 1.8774,
+      "step": 243
+    },
+    {
+      "epoch": 0.22524809600738518,
+      "grad_norm": 7.47265625,
+      "learning_rate": 9.843895930620414e-05,
+      "loss": 1.8223,
+      "step": 244
+    },
+    {
+      "epoch": 0.22617124394184168,
+      "grad_norm": 7.29296875,
+      "learning_rate": 9.843228819212809e-05,
+      "loss": 2.1379,
+      "step": 245
+    },
+    {
+      "epoch": 0.22709439187629818,
+      "grad_norm": 8.3828125,
+      "learning_rate": 9.842561707805203e-05,
+      "loss": 2.2971,
+      "step": 246
+    },
+    {
+      "epoch": 0.22801753981075468,
+      "grad_norm": 8.15625,
+      "learning_rate": 9.841894596397598e-05,
+      "loss": 1.8374,
+      "step": 247
+    },
+    {
+      "epoch": 0.22894068774521117,
+      "grad_norm": 6.7578125,
+      "learning_rate": 9.841227484989994e-05,
+      "loss": 1.6539,
+      "step": 248
+    },
+    {
+      "epoch": 0.22986383567966767,
+      "grad_norm": 6.78125,
+      "learning_rate": 9.84056037358239e-05,
+      "loss": 1.4236,
+      "step": 249
+    },
+    {
+      "epoch": 0.23078698361412417,
+      "grad_norm": 8.3359375,
+      "learning_rate": 9.839893262174784e-05,
+      "loss": 1.9727,
+      "step": 250
+    },
+    {
+      "epoch": 0.23171013154858067,
+      "grad_norm": 8.3671875,
+      "learning_rate": 9.839226150767179e-05,
+      "loss": 2.2769,
+      "step": 251
+    },
+    {
+      "epoch": 0.23263327948303716,
+      "grad_norm": 9.640625,
+      "learning_rate": 9.838559039359573e-05,
+      "loss": 2.116,
+      "step": 252
+    },
+    {
+      "epoch": 0.23355642741749366,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.837891927951969e-05,
+      "loss": 1.8723,
+      "step": 253
+    },
+    {
+      "epoch": 0.23447957535195016,
+      "grad_norm": 9.6640625,
+      "learning_rate": 9.837224816544364e-05,
+      "loss": 1.9448,
+      "step": 254
+    },
+    {
+      "epoch": 0.23540272328640666,
+      "grad_norm": 7.5546875,
+      "learning_rate": 9.836557705136758e-05,
+      "loss": 2.0471,
+      "step": 255
+    },
+    {
+      "epoch": 0.23632587122086315,
+      "grad_norm": 9.9453125,
+      "learning_rate": 9.835890593729153e-05,
+      "loss": 2.3164,
+      "step": 256
+    },
+    {
+      "epoch": 0.23724901915531965,
+      "grad_norm": 7.7421875,
+      "learning_rate": 9.835223482321549e-05,
+      "loss": 2.1377,
+      "step": 257
+    },
+    {
+      "epoch": 0.23817216708977615,
+      "grad_norm": 10.09375,
+      "learning_rate": 9.834556370913943e-05,
+      "loss": 2.1184,
+      "step": 258
+    },
+    {
+      "epoch": 0.23909531502423265,
+      "grad_norm": 5.046875,
+      "learning_rate": 9.833889259506338e-05,
+      "loss": 1.6819,
+      "step": 259
+    },
+    {
+      "epoch": 0.24001846295868912,
+      "grad_norm": 8.015625,
+      "learning_rate": 9.833222148098732e-05,
+      "loss": 2.0598,
+      "step": 260
+    },
+    {
+      "epoch": 0.2409416108931456,
+      "grad_norm": 7.23828125,
+      "learning_rate": 9.832555036691128e-05,
+      "loss": 2.1196,
+      "step": 261
+    },
+    {
+      "epoch": 0.2418647588276021,
+      "grad_norm": 10.4140625,
+      "learning_rate": 9.831887925283523e-05,
+      "loss": 1.8186,
+      "step": 262
+    },
+    {
+      "epoch": 0.2427879067620586,
+      "grad_norm": 5.921875,
+      "learning_rate": 9.831220813875917e-05,
+      "loss": 1.9304,
+      "step": 263
+    },
+    {
+      "epoch": 0.2437110546965151,
+      "grad_norm": 8.5703125,
+      "learning_rate": 9.830553702468312e-05,
+      "loss": 1.7825,
+      "step": 264
+    },
+    {
+      "epoch": 0.2446342026309716,
+      "grad_norm": 7.6328125,
+      "learning_rate": 9.829886591060708e-05,
+      "loss": 2.0845,
+      "step": 265
+    },
+    {
+      "epoch": 0.2455573505654281,
+      "grad_norm": 7.8046875,
+      "learning_rate": 9.829219479653102e-05,
+      "loss": 1.7998,
+      "step": 266
+    },
+    {
+      "epoch": 0.2464804984998846,
+      "grad_norm": 5.50390625,
+      "learning_rate": 9.828552368245497e-05,
+      "loss": 2.0649,
+      "step": 267
+    },
+    {
+      "epoch": 0.2474036464343411,
+      "grad_norm": 10.6796875,
+      "learning_rate": 9.827885256837892e-05,
+      "loss": 1.8015,
+      "step": 268
+    },
+    {
+      "epoch": 0.2483267943687976,
+      "grad_norm": 5.5078125,
+      "learning_rate": 9.827218145430287e-05,
+      "loss": 1.8521,
+      "step": 269
+    },
+    {
+      "epoch": 0.2492499423032541,
+      "grad_norm": 7.36328125,
+      "learning_rate": 9.826551034022682e-05,
+      "loss": 1.9448,
+      "step": 270
+    },
+    {
+      "epoch": 0.2501730902377106,
+      "grad_norm": 5.734375,
+      "learning_rate": 9.825883922615077e-05,
+      "loss": 1.9905,
+      "step": 271
+    },
+    {
+      "epoch": 0.2510962381721671,
+      "grad_norm": 9.1953125,
+      "learning_rate": 9.825216811207471e-05,
+      "loss": 2.0356,
+      "step": 272
+    },
+    {
+      "epoch": 0.2520193861066236,
+      "grad_norm": 7.02734375,
+      "learning_rate": 9.824549699799867e-05,
+      "loss": 1.8069,
+      "step": 273
+    },
+    {
+      "epoch": 0.2529425340410801,
+      "grad_norm": 7.53515625,
+      "learning_rate": 9.823882588392263e-05,
+      "loss": 1.7554,
+      "step": 274
+    },
+    {
+      "epoch": 0.2538656819755366,
+      "grad_norm": 5.09765625,
+      "learning_rate": 9.823215476984657e-05,
+      "loss": 1.7788,
+      "step": 275
+    },
+    {
+      "epoch": 0.2547888299099931,
+      "grad_norm": 6.9140625,
+      "learning_rate": 9.822548365577052e-05,
+      "loss": 1.7656,
+      "step": 276
+    },
+    {
+      "epoch": 0.2557119778444496,
+      "grad_norm": 5.80859375,
+      "learning_rate": 9.821881254169448e-05,
+      "loss": 2.0911,
+      "step": 277
+    },
+    {
+      "epoch": 0.25663512577890607,
+      "grad_norm": 7.72265625,
+      "learning_rate": 9.821214142761842e-05,
+      "loss": 2.1436,
+      "step": 278
+    },
+    {
+      "epoch": 0.25755827371336254,
+      "grad_norm": 8.0390625,
+      "learning_rate": 9.820547031354237e-05,
+      "loss": 2.1826,
+      "step": 279
+    },
+    {
+      "epoch": 0.25848142164781907,
+      "grad_norm": 9.234375,
+      "learning_rate": 9.819879919946631e-05,
+      "loss": 1.709,
+      "step": 280
+    },
+    {
+      "epoch": 0.25940456958227553,
+      "grad_norm": 6.984375,
+      "learning_rate": 9.819212808539027e-05,
+      "loss": 1.8047,
+      "step": 281
+    },
+    {
+      "epoch": 0.26032771751673206,
+      "grad_norm": 7.1953125,
+      "learning_rate": 9.818545697131422e-05,
+      "loss": 1.9277,
+      "step": 282
+    },
+    {
+      "epoch": 0.26125086545118853,
+      "grad_norm": 6.3515625,
+      "learning_rate": 9.817878585723816e-05,
+      "loss": 1.907,
+      "step": 283
+    },
+    {
+      "epoch": 0.26217401338564505,
+      "grad_norm": 9.34375,
+      "learning_rate": 9.817211474316211e-05,
+      "loss": 2.0029,
+      "step": 284
+    },
+    {
+      "epoch": 0.2630971613201015,
+      "grad_norm": 5.8984375,
+      "learning_rate": 9.816544362908607e-05,
+      "loss": 1.9868,
+      "step": 285
+    },
+    {
+      "epoch": 0.26402030925455805,
+      "grad_norm": 8.34375,
+      "learning_rate": 9.815877251501001e-05,
+      "loss": 1.9119,
+      "step": 286
+    },
+    {
+      "epoch": 0.2649434571890145,
+      "grad_norm": 7.07421875,
+      "learning_rate": 9.815210140093396e-05,
+      "loss": 1.9597,
+      "step": 287
+    },
+    {
+      "epoch": 0.26586660512347104,
+      "grad_norm": 9.71875,
+      "learning_rate": 9.81454302868579e-05,
+      "loss": 1.8572,
+      "step": 288
+    },
+    {
+      "epoch": 0.2667897530579275,
+      "grad_norm": 6.24609375,
+      "learning_rate": 9.813875917278186e-05,
+      "loss": 2.1304,
+      "step": 289
+    },
+    {
+      "epoch": 0.26771290099238404,
+      "grad_norm": 11.765625,
+      "learning_rate": 9.813208805870581e-05,
+      "loss": 1.8933,
+      "step": 290
+    },
+    {
+      "epoch": 0.2686360489268405,
+      "grad_norm": 4.8515625,
+      "learning_rate": 9.812541694462975e-05,
+      "loss": 1.9749,
+      "step": 291
+    },
+    {
+      "epoch": 0.26955919686129703,
+      "grad_norm": 11.203125,
+      "learning_rate": 9.81187458305537e-05,
+      "loss": 1.8953,
+      "step": 292
+    },
+    {
+      "epoch": 0.2704823447957535,
+      "grad_norm": 7.21875,
+      "learning_rate": 9.811207471647766e-05,
+      "loss": 2.0579,
+      "step": 293
+    },
+    {
+      "epoch": 0.27140549273021003,
+      "grad_norm": 9.6875,
+      "learning_rate": 9.81054036024016e-05,
+      "loss": 2.0042,
+      "step": 294
+    },
+    {
+      "epoch": 0.2723286406646665,
+      "grad_norm": 6.29296875,
+      "learning_rate": 9.809873248832555e-05,
+      "loss": 2.0029,
+      "step": 295
+    },
+    {
+      "epoch": 0.273251788599123,
+      "grad_norm": 8.5390625,
+      "learning_rate": 9.80920613742495e-05,
+      "loss": 1.9434,
+      "step": 296
+    },
+    {
+      "epoch": 0.2741749365335795,
+      "grad_norm": 7.43359375,
+      "learning_rate": 9.808539026017345e-05,
+      "loss": 1.5764,
+      "step": 297
+    },
+    {
+      "epoch": 0.275098084468036,
+      "grad_norm": 5.984375,
+      "learning_rate": 9.80787191460974e-05,
+      "loss": 1.7959,
+      "step": 298
+    },
+    {
+      "epoch": 0.2760212324024925,
+      "grad_norm": 5.60546875,
+      "learning_rate": 9.807204803202136e-05,
+      "loss": 1.7468,
+      "step": 299
+    },
+    {
+      "epoch": 0.276944380336949,
+      "grad_norm": 7.046875,
+      "learning_rate": 9.80653769179453e-05,
+      "loss": 1.8413,
+      "step": 300
+    },
+    {
+      "epoch": 0.2778675282714055,
+      "grad_norm": 6.80078125,
+      "learning_rate": 9.805870580386925e-05,
+      "loss": 2.3672,
+      "step": 301
+    },
+    {
+      "epoch": 0.278790676205862,
+      "grad_norm": 10.171875,
+      "learning_rate": 9.805203468979321e-05,
+      "loss": 1.8975,
+      "step": 302
+    },
+    {
+      "epoch": 0.2797138241403185,
+      "grad_norm": 7.41015625,
+      "learning_rate": 9.804536357571715e-05,
+      "loss": 1.7681,
+      "step": 303
+    },
+    {
+      "epoch": 0.280636972074775,
+      "grad_norm": 7.19921875,
+      "learning_rate": 9.80386924616411e-05,
+      "loss": 1.9331,
+      "step": 304
+    },
+    {
+      "epoch": 0.2815601200092315,
+      "grad_norm": 7.1328125,
+      "learning_rate": 9.803202134756505e-05,
+      "loss": 1.8254,
+      "step": 305
+    },
+    {
+      "epoch": 0.282483267943688,
+      "grad_norm": 7.89453125,
+      "learning_rate": 9.8025350233489e-05,
+      "loss": 1.9219,
+      "step": 306
+    },
+    {
+      "epoch": 0.28340641587814447,
+      "grad_norm": 6.97265625,
+      "learning_rate": 9.801867911941295e-05,
+      "loss": 1.5212,
+      "step": 307
+    },
+    {
+      "epoch": 0.284329563812601,
+      "grad_norm": 8.3828125,
+      "learning_rate": 9.80120080053369e-05,
+      "loss": 1.7944,
+      "step": 308
+    },
+    {
+      "epoch": 0.28525271174705746,
+      "grad_norm": 6.18359375,
+      "learning_rate": 9.800533689126084e-05,
+      "loss": 1.9309,
+      "step": 309
+    },
+    {
+      "epoch": 0.286175859681514,
+      "grad_norm": 7.734375,
+      "learning_rate": 9.79986657771848e-05,
+      "loss": 1.678,
+      "step": 310
+    },
+    {
+      "epoch": 0.28709900761597046,
+      "grad_norm": 6.33203125,
+      "learning_rate": 9.799199466310874e-05,
+      "loss": 2.0806,
+      "step": 311
+    },
+    {
+      "epoch": 0.28802215555042693,
+      "grad_norm": 10.6640625,
+      "learning_rate": 9.798532354903269e-05,
+      "loss": 1.8906,
+      "step": 312
+    },
+    {
+      "epoch": 0.28894530348488345,
+      "grad_norm": 4.93359375,
+      "learning_rate": 9.797865243495664e-05,
+      "loss": 2.1196,
+      "step": 313
+    },
+    {
+      "epoch": 0.2898684514193399,
+      "grad_norm": 8.109375,
+      "learning_rate": 9.79719813208806e-05,
+      "loss": 1.7913,
+      "step": 314
+    },
+    {
+      "epoch": 0.29079159935379645,
+      "grad_norm": 6.2890625,
+      "learning_rate": 9.796531020680454e-05,
+      "loss": 1.6935,
+      "step": 315
+    },
+    {
+      "epoch": 0.2917147472882529,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.795863909272849e-05,
+      "loss": 1.8784,
+      "step": 316
+    },
+    {
+      "epoch": 0.29263789522270944,
+      "grad_norm": 6.12890625,
+      "learning_rate": 9.795196797865243e-05,
+      "loss": 1.5134,
+      "step": 317
+    },
+    {
+      "epoch": 0.2935610431571659,
+      "grad_norm": 7.2421875,
+      "learning_rate": 9.794529686457639e-05,
+      "loss": 1.8337,
+      "step": 318
+    },
+    {
+      "epoch": 0.29448419109162244,
+      "grad_norm": 6.7109375,
+      "learning_rate": 9.793862575050034e-05,
+      "loss": 1.752,
+      "step": 319
+    },
+    {
+      "epoch": 0.2954073390260789,
+      "grad_norm": 9.84375,
+      "learning_rate": 9.793195463642428e-05,
+      "loss": 1.7925,
+      "step": 320
+    },
+    {
+      "epoch": 0.29633048696053543,
+      "grad_norm": 7.2265625,
+      "learning_rate": 9.792528352234823e-05,
+      "loss": 1.8762,
+      "step": 321
+    },
+    {
+      "epoch": 0.2972536348949919,
+      "grad_norm": 7.6953125,
+      "learning_rate": 9.791861240827219e-05,
+      "loss": 1.8394,
+      "step": 322
+    },
+    {
+      "epoch": 0.29817678282944843,
+      "grad_norm": 7.1640625,
+      "learning_rate": 9.791194129419613e-05,
+      "loss": 2.0505,
+      "step": 323
+    },
+    {
+      "epoch": 0.2990999307639049,
+      "grad_norm": 7.68359375,
+      "learning_rate": 9.790527018012009e-05,
+      "loss": 1.9717,
+      "step": 324
+    },
+    {
+      "epoch": 0.3000230786983614,
+      "grad_norm": 7.50390625,
+      "learning_rate": 9.789859906604404e-05,
+      "loss": 1.8064,
+      "step": 325
+    },
+    {
+      "epoch": 0.3009462266328179,
+      "grad_norm": 6.3125,
+      "learning_rate": 9.7891927951968e-05,
+      "loss": 1.8638,
+      "step": 326
+    },
+    {
+      "epoch": 0.3018693745672744,
+      "grad_norm": 6.42578125,
+      "learning_rate": 9.788525683789194e-05,
+      "loss": 1.7483,
+      "step": 327
+    },
+    {
+      "epoch": 0.3027925225017309,
+      "grad_norm": 6.42578125,
+      "learning_rate": 9.787858572381588e-05,
+      "loss": 1.8079,
+      "step": 328
+    },
+    {
+      "epoch": 0.3037156704361874,
+      "grad_norm": 5.58984375,
+      "learning_rate": 9.787191460973983e-05,
+      "loss": 1.6431,
+      "step": 329
+    },
+    {
+      "epoch": 0.3046388183706439,
+      "grad_norm": 6.25390625,
+      "learning_rate": 9.786524349566379e-05,
+      "loss": 1.4775,
+      "step": 330
+    },
+    {
+      "epoch": 0.3055619663051004,
+      "grad_norm": 6.5859375,
+      "learning_rate": 9.785857238158773e-05,
+      "loss": 1.9299,
+      "step": 331
+    },
+    {
+      "epoch": 0.3064851142395569,
+      "grad_norm": 9.2890625,
+      "learning_rate": 9.785190126751168e-05,
+      "loss": 2.0537,
+      "step": 332
+    },
+    {
+      "epoch": 0.3074082621740134,
+      "grad_norm": 5.5,
+      "learning_rate": 9.784523015343563e-05,
+      "loss": 1.9915,
+      "step": 333
+    },
+    {
+      "epoch": 0.30833141010846987,
+      "grad_norm": 8.265625,
+      "learning_rate": 9.783855903935958e-05,
+      "loss": 2.0579,
+      "step": 334
+    },
+    {
+      "epoch": 0.3092545580429264,
+      "grad_norm": 5.6953125,
+      "learning_rate": 9.783188792528353e-05,
+      "loss": 1.9753,
+      "step": 335
+    },
+    {
+      "epoch": 0.31017770597738287,
+      "grad_norm": 10.421875,
+      "learning_rate": 9.782521681120748e-05,
+      "loss": 2.1694,
+      "step": 336
+    },
+    {
+      "epoch": 0.3111008539118394,
+      "grad_norm": 5.30859375,
+      "learning_rate": 9.781854569713142e-05,
+      "loss": 1.7417,
+      "step": 337
+    },
+    {
+      "epoch": 0.31202400184629586,
+      "grad_norm": 7.38671875,
+      "learning_rate": 9.781187458305538e-05,
+      "loss": 1.8413,
+      "step": 338
+    },
+    {
+      "epoch": 0.3129471497807524,
+      "grad_norm": 7.4140625,
+      "learning_rate": 9.780520346897933e-05,
+      "loss": 1.9443,
+      "step": 339
+    },
+    {
+      "epoch": 0.31387029771520886,
+      "grad_norm": 8.2265625,
+      "learning_rate": 9.779853235490327e-05,
+      "loss": 1.7446,
+      "step": 340
+    },
+    {
+      "epoch": 0.3147934456496654,
+      "grad_norm": 5.609375,
+      "learning_rate": 9.779186124082722e-05,
+      "loss": 1.6411,
+      "step": 341
+    },
+    {
+      "epoch": 0.31571659358412185,
+      "grad_norm": 7.0859375,
+      "learning_rate": 9.778519012675118e-05,
+      "loss": 1.6261,
+      "step": 342
+    },
+    {
+      "epoch": 0.3166397415185784,
+      "grad_norm": 8.703125,
+      "learning_rate": 9.777851901267512e-05,
+      "loss": 1.6178,
+      "step": 343
+    },
+    {
+      "epoch": 0.31756288945303485,
+      "grad_norm": 6.8828125,
+      "learning_rate": 9.777184789859907e-05,
+      "loss": 1.7908,
+      "step": 344
+    },
+    {
+      "epoch": 0.31848603738749137,
+      "grad_norm": 6.76953125,
+      "learning_rate": 9.776517678452301e-05,
+      "loss": 1.8848,
+      "step": 345
+    },
+    {
+      "epoch": 0.31940918532194784,
+      "grad_norm": 6.203125,
+      "learning_rate": 9.775850567044697e-05,
+      "loss": 1.845,
+      "step": 346
+    },
+    {
+      "epoch": 0.3203323332564043,
+      "grad_norm": 6.33984375,
+      "learning_rate": 9.775183455637092e-05,
+      "loss": 1.4749,
+      "step": 347
+    },
+    {
+      "epoch": 0.32125548119086084,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.774516344229486e-05,
+      "loss": 1.8572,
+      "step": 348
+    },
+    {
+      "epoch": 0.3221786291253173,
+      "grad_norm": 7.66015625,
+      "learning_rate": 9.773849232821882e-05,
+      "loss": 1.9917,
+      "step": 349
+    },
+    {
+      "epoch": 0.32310177705977383,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.773182121414277e-05,
+      "loss": 1.8682,
+      "step": 350
+    },
+    {
+      "epoch": 0.3240249249942303,
+      "grad_norm": 6.94140625,
+      "learning_rate": 9.772515010006672e-05,
+      "loss": 1.9636,
+      "step": 351
+    },
+    {
+      "epoch": 0.3249480729286868,
+      "grad_norm": 9.6640625,
+      "learning_rate": 9.771847898599067e-05,
+      "loss": 2.0718,
+      "step": 352
+    },
+    {
+      "epoch": 0.3258712208631433,
+      "grad_norm": 8.4296875,
+      "learning_rate": 9.771180787191462e-05,
+      "loss": 1.2907,
+      "step": 353
+    },
+    {
+      "epoch": 0.3267943687975998,
+      "grad_norm": 6.58203125,
+      "learning_rate": 9.770513675783856e-05,
+      "loss": 1.9324,
+      "step": 354
+    },
+    {
+      "epoch": 0.3277175167320563,
+      "grad_norm": 7.85546875,
+      "learning_rate": 9.769846564376252e-05,
+      "loss": 1.937,
+      "step": 355
+    },
+    {
+      "epoch": 0.3286406646665128,
+      "grad_norm": 7.234375,
+      "learning_rate": 9.769179452968647e-05,
+      "loss": 1.8916,
+      "step": 356
+    },
+    {
+      "epoch": 0.3295638126009693,
+      "grad_norm": 6.82421875,
+      "learning_rate": 9.768512341561041e-05,
+      "loss": 1.855,
+      "step": 357
+    },
+    {
+      "epoch": 0.3304869605354258,
+      "grad_norm": 5.8203125,
+      "learning_rate": 9.767845230153436e-05,
+      "loss": 1.6425,
+      "step": 358
+    },
+    {
+      "epoch": 0.3314101084698823,
+      "grad_norm": 7.96484375,
+      "learning_rate": 9.767178118745832e-05,
+      "loss": 1.8533,
+      "step": 359
+    },
+    {
+      "epoch": 0.3323332564043388,
+      "grad_norm": 5.5,
+      "learning_rate": 9.766511007338226e-05,
+      "loss": 1.8071,
+      "step": 360
+    },
+    {
+      "epoch": 0.3332564043387953,
+      "grad_norm": 5.8359375,
+      "learning_rate": 9.76584389593062e-05,
+      "loss": 1.991,
+      "step": 361
+    },
+    {
+      "epoch": 0.3341795522732518,
+      "grad_norm": 6.8125,
+      "learning_rate": 9.765176784523015e-05,
+      "loss": 1.761,
+      "step": 362
+    },
+    {
+      "epoch": 0.33510270020770827,
+      "grad_norm": 5.96484375,
+      "learning_rate": 9.764509673115411e-05,
+      "loss": 1.9397,
+      "step": 363
+    },
+    {
+      "epoch": 0.3360258481421648,
+      "grad_norm": 8.921875,
+      "learning_rate": 9.763842561707806e-05,
+      "loss": 1.8225,
+      "step": 364
+    },
+    {
+      "epoch": 0.33694899607662127,
+      "grad_norm": 6.515625,
+      "learning_rate": 9.7631754503002e-05,
+      "loss": 2.022,
+      "step": 365
+    },
+    {
+      "epoch": 0.3378721440110778,
+      "grad_norm": 8.109375,
+      "learning_rate": 9.762508338892595e-05,
+      "loss": 2.0222,
+      "step": 366
+    },
+    {
+      "epoch": 0.33879529194553426,
+      "grad_norm": 6.61328125,
+      "learning_rate": 9.76184122748499e-05,
+      "loss": 1.9377,
+      "step": 367
+    },
+    {
+      "epoch": 0.3397184398799908,
+      "grad_norm": 7.3984375,
+      "learning_rate": 9.761174116077385e-05,
+      "loss": 2.1025,
+      "step": 368
+    },
+    {
+      "epoch": 0.34064158781444726,
+      "grad_norm": 6.8359375,
+      "learning_rate": 9.76050700466978e-05,
+      "loss": 2.0266,
+      "step": 369
+    },
+    {
+      "epoch": 0.3415647357489038,
+      "grad_norm": 8.671875,
+      "learning_rate": 9.759839893262174e-05,
+      "loss": 1.6848,
+      "step": 370
+    },
+    {
+      "epoch": 0.34248788368336025,
+      "grad_norm": 5.85546875,
+      "learning_rate": 9.75917278185457e-05,
+      "loss": 1.6594,
+      "step": 371
+    },
+    {
+      "epoch": 0.3434110316178168,
+      "grad_norm": 7.90234375,
+      "learning_rate": 9.758505670446965e-05,
+      "loss": 1.9941,
+      "step": 372
+    },
+    {
+      "epoch": 0.34433417955227325,
+      "grad_norm": 6.57421875,
+      "learning_rate": 9.757838559039359e-05,
+      "loss": 1.5376,
+      "step": 373
+    },
+    {
+      "epoch": 0.34525732748672977,
+      "grad_norm": 9.65625,
+      "learning_rate": 9.757171447631755e-05,
+      "loss": 2.0342,
+      "step": 374
+    },
+    {
+      "epoch": 0.34618047542118624,
+      "grad_norm": 6.25,
+      "learning_rate": 9.756504336224151e-05,
+      "loss": 1.781,
+      "step": 375
+    },
+    {
+      "epoch": 0.34710362335564277,
+      "grad_norm": 8.296875,
+      "learning_rate": 9.755837224816546e-05,
+      "loss": 1.9001,
+      "step": 376
+    },
+    {
+      "epoch": 0.34802677129009923,
+      "grad_norm": 6.1328125,
+      "learning_rate": 9.75517011340894e-05,
+      "loss": 1.583,
+      "step": 377
+    },
+    {
+      "epoch": 0.34894991922455576,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.754503002001335e-05,
+      "loss": 1.7808,
+      "step": 378
+    },
+    {
+      "epoch": 0.34987306715901223,
+      "grad_norm": 6.73046875,
+      "learning_rate": 9.75383589059373e-05,
+      "loss": 1.8645,
+      "step": 379
+    },
+    {
+      "epoch": 0.35079621509346876,
+      "grad_norm": 8.7890625,
+      "learning_rate": 9.753168779186125e-05,
+      "loss": 1.8445,
+      "step": 380
+    },
+    {
+      "epoch": 0.3517193630279252,
+      "grad_norm": 7.484375,
+      "learning_rate": 9.75250166777852e-05,
+      "loss": 1.9241,
+      "step": 381
+    },
+    {
+      "epoch": 0.3526425109623817,
+      "grad_norm": 7.609375,
+      "learning_rate": 9.751834556370914e-05,
+      "loss": 1.9749,
+      "step": 382
+    },
+    {
+      "epoch": 0.3535656588968382,
+      "grad_norm": 6.4453125,
+      "learning_rate": 9.75116744496331e-05,
+      "loss": 1.6152,
+      "step": 383
+    },
+    {
+      "epoch": 0.3544888068312947,
+      "grad_norm": 6.52734375,
+      "learning_rate": 9.750500333555705e-05,
+      "loss": 1.6895,
+      "step": 384
+    },
+    {
+      "epoch": 0.3554119547657512,
+      "grad_norm": 7.890625,
+      "learning_rate": 9.749833222148099e-05,
+      "loss": 1.9922,
+      "step": 385
+    },
+    {
+      "epoch": 0.3563351027002077,
+      "grad_norm": 8.46875,
+      "learning_rate": 9.749166110740494e-05,
+      "loss": 1.8076,
+      "step": 386
+    },
+    {
+      "epoch": 0.3572582506346642,
+      "grad_norm": 4.90234375,
+      "learning_rate": 9.74849899933289e-05,
+      "loss": 1.8621,
+      "step": 387
+    },
+    {
+      "epoch": 0.3581813985691207,
+      "grad_norm": 10.8203125,
+      "learning_rate": 9.747831887925284e-05,
+      "loss": 2.0715,
+      "step": 388
+    },
+    {
+      "epoch": 0.3591045465035772,
+      "grad_norm": 6.17578125,
+      "learning_rate": 9.747164776517679e-05,
+      "loss": 1.8552,
+      "step": 389
+    },
+    {
+      "epoch": 0.3600276944380337,
+      "grad_norm": 7.31640625,
+      "learning_rate": 9.746497665110073e-05,
+      "loss": 1.7153,
+      "step": 390
+    },
+    {
+      "epoch": 0.3609508423724902,
+      "grad_norm": 6.17578125,
+      "learning_rate": 9.745830553702469e-05,
+      "loss": 1.9153,
+      "step": 391
+    },
+    {
+      "epoch": 0.36187399030694667,
+      "grad_norm": 7.26171875,
+      "learning_rate": 9.745163442294864e-05,
+      "loss": 1.665,
+      "step": 392
+    },
+    {
+      "epoch": 0.3627971382414032,
+      "grad_norm": 6.13671875,
+      "learning_rate": 9.744496330887258e-05,
+      "loss": 1.9526,
+      "step": 393
+    },
+    {
+      "epoch": 0.36372028617585966,
+      "grad_norm": 8.1953125,
+      "learning_rate": 9.743829219479653e-05,
+      "loss": 1.7515,
+      "step": 394
+    },
+    {
+      "epoch": 0.3646434341103162,
+      "grad_norm": 6.08984375,
+      "learning_rate": 9.743162108072049e-05,
+      "loss": 1.7913,
+      "step": 395
+    },
+    {
+      "epoch": 0.36556658204477266,
+      "grad_norm": 8.78125,
+      "learning_rate": 9.742494996664443e-05,
+      "loss": 1.8286,
+      "step": 396
+    },
+    {
+      "epoch": 0.3664897299792292,
+      "grad_norm": 6.83203125,
+      "learning_rate": 9.741827885256838e-05,
+      "loss": 2.1528,
+      "step": 397
+    },
+    {
+      "epoch": 0.36741287791368565,
+      "grad_norm": 9.046875,
+      "learning_rate": 9.741160773849232e-05,
+      "loss": 1.895,
+      "step": 398
+    },
+    {
+      "epoch": 0.3683360258481422,
+      "grad_norm": 7.45703125,
+      "learning_rate": 9.740493662441628e-05,
+      "loss": 2.1372,
+      "step": 399
+    },
+    {
+      "epoch": 0.36925917378259865,
+      "grad_norm": 10.2578125,
+      "learning_rate": 9.739826551034024e-05,
+      "loss": 2.1362,
+      "step": 400
+    },
+    {
+      "epoch": 0.3701823217170552,
+      "grad_norm": 5.0,
+      "learning_rate": 9.739159439626419e-05,
+      "loss": 1.8911,
+      "step": 401
+    },
+    {
+      "epoch": 0.37110546965151164,
+      "grad_norm": 6.4140625,
+      "learning_rate": 9.738492328218813e-05,
+      "loss": 1.9431,
+      "step": 402
+    },
+    {
+      "epoch": 0.37202861758596817,
+      "grad_norm": 4.5546875,
+      "learning_rate": 9.737825216811208e-05,
+      "loss": 1.8711,
+      "step": 403
+    },
+    {
+      "epoch": 0.37295176552042464,
+      "grad_norm": 7.3125,
+      "learning_rate": 9.737158105403604e-05,
+      "loss": 1.981,
+      "step": 404
+    },
+    {
+      "epoch": 0.37387491345488116,
+      "grad_norm": 6.55859375,
+      "learning_rate": 9.736490993995998e-05,
+      "loss": 1.835,
+      "step": 405
+    },
+    {
+      "epoch": 0.37479806138933763,
+      "grad_norm": 8.1171875,
+      "learning_rate": 9.735823882588393e-05,
+      "loss": 1.7805,
+      "step": 406
+    },
+    {
+      "epoch": 0.37572120932379416,
+      "grad_norm": 5.953125,
+      "learning_rate": 9.735156771180787e-05,
+      "loss": 1.5876,
+      "step": 407
+    },
+    {
+      "epoch": 0.37664435725825063,
+      "grad_norm": 6.4375,
+      "learning_rate": 9.734489659773183e-05,
+      "loss": 1.8513,
+      "step": 408
+    },
+    {
+      "epoch": 0.37756750519270715,
+      "grad_norm": 7.25390625,
+      "learning_rate": 9.733822548365578e-05,
+      "loss": 1.635,
+      "step": 409
+    },
+    {
+      "epoch": 0.3784906531271636,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.733155436957972e-05,
+      "loss": 1.8447,
+      "step": 410
+    },
+    {
+      "epoch": 0.37941380106162015,
+      "grad_norm": 6.0,
+      "learning_rate": 9.732488325550367e-05,
+      "loss": 1.6052,
+      "step": 411
+    },
+    {
+      "epoch": 0.3803369489960766,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.731821214142763e-05,
+      "loss": 2.0273,
+      "step": 412
+    },
+    {
+      "epoch": 0.38126009693053314,
+      "grad_norm": 7.71484375,
+      "learning_rate": 9.731154102735157e-05,
+      "loss": 1.5293,
+      "step": 413
+    },
+    {
+      "epoch": 0.3821832448649896,
+      "grad_norm": 6.73046875,
+      "learning_rate": 9.730486991327552e-05,
+      "loss": 1.7256,
+      "step": 414
+    },
+    {
+      "epoch": 0.38310639279944614,
+      "grad_norm": 5.984375,
+      "learning_rate": 9.729819879919946e-05,
+      "loss": 1.9116,
+      "step": 415
+    },
+    {
+      "epoch": 0.3840295407339026,
+      "grad_norm": 8.21875,
+      "learning_rate": 9.729152768512342e-05,
+      "loss": 1.8054,
+      "step": 416
+    },
+    {
+      "epoch": 0.3849526886683591,
+      "grad_norm": 6.453125,
+      "learning_rate": 9.728485657104737e-05,
+      "loss": 1.9309,
+      "step": 417
+    },
+    {
+      "epoch": 0.3858758366028156,
+      "grad_norm": 8.390625,
+      "learning_rate": 9.727818545697131e-05,
+      "loss": 1.9006,
+      "step": 418
+    },
+    {
+      "epoch": 0.3867989845372721,
+      "grad_norm": 6.796875,
+      "learning_rate": 9.727151434289526e-05,
+      "loss": 1.6365,
+      "step": 419
+    },
+    {
+      "epoch": 0.3877221324717286,
+      "grad_norm": 7.890625,
+      "learning_rate": 9.726484322881922e-05,
+      "loss": 1.9709,
+      "step": 420
+    },
+    {
+      "epoch": 0.38864528040618507,
+      "grad_norm": 4.74609375,
+      "learning_rate": 9.725817211474316e-05,
+      "loss": 1.5393,
+      "step": 421
+    },
+    {
+      "epoch": 0.3895684283406416,
+      "grad_norm": 5.375,
+      "learning_rate": 9.725150100066711e-05,
+      "loss": 1.884,
+      "step": 422
+    },
+    {
+      "epoch": 0.39049157627509806,
+      "grad_norm": 5.67578125,
+      "learning_rate": 9.724482988659107e-05,
+      "loss": 1.4019,
+      "step": 423
+    },
+    {
+      "epoch": 0.3914147242095546,
+      "grad_norm": 6.8515625,
+      "learning_rate": 9.723815877251503e-05,
+      "loss": 1.9158,
+      "step": 424
+    },
+    {
+      "epoch": 0.39233787214401106,
+      "grad_norm": 8.3671875,
+      "learning_rate": 9.723148765843897e-05,
+      "loss": 1.822,
+      "step": 425
+    },
+    {
+      "epoch": 0.3932610200784676,
+      "grad_norm": 6.08984375,
+      "learning_rate": 9.722481654436292e-05,
+      "loss": 1.4629,
+      "step": 426
+    },
+    {
+      "epoch": 0.39418416801292405,
+      "grad_norm": 6.7421875,
+      "learning_rate": 9.721814543028686e-05,
+      "loss": 1.7261,
+      "step": 427
+    },
+    {
+      "epoch": 0.3951073159473806,
+      "grad_norm": 7.18359375,
+      "learning_rate": 9.721147431621082e-05,
+      "loss": 1.8552,
+      "step": 428
+    },
+    {
+      "epoch": 0.39603046388183705,
+      "grad_norm": 5.55078125,
+      "learning_rate": 9.720480320213477e-05,
+      "loss": 1.9292,
+      "step": 429
+    },
+    {
+      "epoch": 0.3969536118162936,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.719813208805871e-05,
+      "loss": 1.9666,
+      "step": 430
+    },
+    {
+      "epoch": 0.39787675975075004,
+      "grad_norm": 6.58203125,
+      "learning_rate": 9.719146097398266e-05,
+      "loss": 1.7778,
+      "step": 431
+    },
+    {
+      "epoch": 0.39879990768520657,
+      "grad_norm": 6.24609375,
+      "learning_rate": 9.718478985990662e-05,
+      "loss": 1.6648,
+      "step": 432
+    },
+    {
+      "epoch": 0.39972305561966304,
+      "grad_norm": 5.421875,
+      "learning_rate": 9.717811874583056e-05,
+      "loss": 1.6973,
+      "step": 433
+    },
+    {
+      "epoch": 0.40064620355411956,
+      "grad_norm": 7.39453125,
+      "learning_rate": 9.717144763175451e-05,
+      "loss": 1.8547,
+      "step": 434
+    },
+    {
+      "epoch": 0.40156935148857603,
+      "grad_norm": 9.4140625,
+      "learning_rate": 9.716477651767845e-05,
+      "loss": 2.0364,
+      "step": 435
+    },
+    {
+      "epoch": 0.40249249942303256,
+      "grad_norm": 6.80859375,
+      "learning_rate": 9.715810540360241e-05,
+      "loss": 1.8237,
+      "step": 436
+    },
+    {
+      "epoch": 0.403415647357489,
+      "grad_norm": 7.16015625,
+      "learning_rate": 9.715143428952636e-05,
+      "loss": 1.8242,
+      "step": 437
+    },
+    {
+      "epoch": 0.40433879529194555,
+      "grad_norm": 7.56640625,
+      "learning_rate": 9.71447631754503e-05,
+      "loss": 2.1543,
+      "step": 438
+    },
+    {
+      "epoch": 0.405261943226402,
+      "grad_norm": 8.90625,
+      "learning_rate": 9.713809206137425e-05,
+      "loss": 1.7002,
+      "step": 439
+    },
+    {
+      "epoch": 0.40618509116085855,
+      "grad_norm": 6.20703125,
+      "learning_rate": 9.713142094729821e-05,
+      "loss": 1.9949,
+      "step": 440
+    },
+    {
+      "epoch": 0.407108239095315,
+      "grad_norm": 8.7109375,
+      "learning_rate": 9.712474983322215e-05,
+      "loss": 1.8442,
+      "step": 441
+    },
+    {
+      "epoch": 0.40803138702977154,
+      "grad_norm": 6.515625,
+      "learning_rate": 9.71180787191461e-05,
+      "loss": 1.8411,
+      "step": 442
+    },
+    {
+      "epoch": 0.408954534964228,
+      "grad_norm": 8.03125,
+      "learning_rate": 9.711140760507004e-05,
+      "loss": 1.3862,
+      "step": 443
+    },
+    {
+      "epoch": 0.40987768289868454,
+      "grad_norm": 5.99609375,
+      "learning_rate": 9.7104736490994e-05,
+      "loss": 1.9365,
+      "step": 444
+    },
+    {
+      "epoch": 0.410800830833141,
+      "grad_norm": 6.515625,
+      "learning_rate": 9.709806537691795e-05,
+      "loss": 1.8826,
+      "step": 445
+    },
+    {
+      "epoch": 0.41172397876759753,
+      "grad_norm": 5.87109375,
+      "learning_rate": 9.709139426284189e-05,
+      "loss": 1.6826,
+      "step": 446
+    },
+    {
+      "epoch": 0.412647126702054,
+      "grad_norm": 5.68359375,
+      "learning_rate": 9.708472314876584e-05,
+      "loss": 1.6328,
+      "step": 447
+    },
+    {
+      "epoch": 0.4135702746365105,
+      "grad_norm": 6.11328125,
+      "learning_rate": 9.70780520346898e-05,
+      "loss": 1.7905,
+      "step": 448
+    },
+    {
+      "epoch": 0.414493422570967,
+      "grad_norm": 6.421875,
+      "learning_rate": 9.707138092061376e-05,
+      "loss": 1.5215,
+      "step": 449
+    },
+    {
+      "epoch": 0.4154165705054235,
+      "grad_norm": 5.97265625,
+      "learning_rate": 9.70647098065377e-05,
+      "loss": 1.7698,
+      "step": 450
+    },
+    {
+      "epoch": 0.41633971843988,
+      "grad_norm": 9.6875,
+      "learning_rate": 9.705803869246165e-05,
+      "loss": 1.9119,
+      "step": 451
+    },
+    {
+      "epoch": 0.41726286637433646,
+      "grad_norm": 7.3671875,
+      "learning_rate": 9.705136757838559e-05,
+      "loss": 1.9167,
+      "step": 452
+    },
+    {
+      "epoch": 0.418186014308793,
+      "grad_norm": 7.15625,
+      "learning_rate": 9.704469646430955e-05,
+      "loss": 1.4827,
+      "step": 453
+    },
+    {
+      "epoch": 0.41910916224324946,
+      "grad_norm": 5.66796875,
+      "learning_rate": 9.70380253502335e-05,
+      "loss": 1.6245,
+      "step": 454
+    },
+    {
+      "epoch": 0.420032310177706,
+      "grad_norm": 9.53125,
+      "learning_rate": 9.703135423615744e-05,
+      "loss": 1.7849,
+      "step": 455
+    },
+    {
+      "epoch": 0.42095545811216245,
+      "grad_norm": 7.29296875,
+      "learning_rate": 9.702468312208139e-05,
+      "loss": 1.6733,
+      "step": 456
+    },
+    {
+      "epoch": 0.421878606046619,
+      "grad_norm": 8.7421875,
+      "learning_rate": 9.701801200800535e-05,
+      "loss": 1.9556,
+      "step": 457
+    },
+    {
+      "epoch": 0.42280175398107545,
+      "grad_norm": 6.328125,
+      "learning_rate": 9.701134089392929e-05,
+      "loss": 1.7693,
+      "step": 458
+    },
+    {
+      "epoch": 0.42372490191553197,
+      "grad_norm": 7.45703125,
+      "learning_rate": 9.700466977985324e-05,
+      "loss": 1.6722,
+      "step": 459
+    },
+    {
+      "epoch": 0.42464804984998844,
+      "grad_norm": 6.8515625,
+      "learning_rate": 9.699799866577718e-05,
+      "loss": 1.6559,
+      "step": 460
+    },
+    {
+      "epoch": 0.42557119778444497,
+      "grad_norm": 7.47265625,
+      "learning_rate": 9.699132755170114e-05,
+      "loss": 1.9612,
+      "step": 461
+    },
+    {
+      "epoch": 0.42649434571890144,
+      "grad_norm": 5.859375,
+      "learning_rate": 9.698465643762509e-05,
+      "loss": 1.9521,
+      "step": 462
+    },
+    {
+      "epoch": 0.42741749365335796,
+      "grad_norm": 7.72265625,
+      "learning_rate": 9.697798532354903e-05,
+      "loss": 1.8479,
+      "step": 463
+    },
+    {
+      "epoch": 0.42834064158781443,
+      "grad_norm": 5.72265625,
+      "learning_rate": 9.697131420947299e-05,
+      "loss": 1.6024,
+      "step": 464
+    },
+    {
+      "epoch": 0.42926378952227096,
+      "grad_norm": 7.58203125,
+      "learning_rate": 9.696464309539694e-05,
+      "loss": 1.6477,
+      "step": 465
+    },
+    {
+      "epoch": 0.4301869374567274,
+      "grad_norm": 8.546875,
+      "learning_rate": 9.695797198132088e-05,
+      "loss": 1.8975,
+      "step": 466
+    },
+    {
+      "epoch": 0.43111008539118395,
+      "grad_norm": 7.484375,
+      "learning_rate": 9.695130086724483e-05,
+      "loss": 1.8708,
+      "step": 467
+    },
+    {
+      "epoch": 0.4320332333256404,
+      "grad_norm": 7.29296875,
+      "learning_rate": 9.694462975316879e-05,
+      "loss": 1.8948,
+      "step": 468
+    },
+    {
+      "epoch": 0.43295638126009695,
+      "grad_norm": 6.64453125,
+      "learning_rate": 9.693795863909273e-05,
+      "loss": 1.834,
+      "step": 469
+    },
+    {
+      "epoch": 0.4338795291945534,
+      "grad_norm": 8.2421875,
+      "learning_rate": 9.693128752501668e-05,
+      "loss": 1.9355,
+      "step": 470
+    },
+    {
+      "epoch": 0.43480267712900994,
+      "grad_norm": 6.25,
+      "learning_rate": 9.692461641094062e-05,
+      "loss": 1.7043,
+      "step": 471
+    },
+    {
+      "epoch": 0.4357258250634664,
+      "grad_norm": 7.87109375,
+      "learning_rate": 9.691794529686458e-05,
+      "loss": 1.8823,
+      "step": 472
+    },
+    {
+      "epoch": 0.43664897299792294,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.691127418278853e-05,
+      "loss": 1.2206,
+      "step": 473
+    },
+    {
+      "epoch": 0.4375721209323794,
+      "grad_norm": 7.4140625,
+      "learning_rate": 9.690460306871249e-05,
+      "loss": 1.9514,
+      "step": 474
+    },
+    {
+      "epoch": 0.43849526886683593,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.689793195463643e-05,
+      "loss": 1.7356,
+      "step": 475
+    },
+    {
+      "epoch": 0.4394184168012924,
+      "grad_norm": 7.6171875,
+      "learning_rate": 9.689126084056038e-05,
+      "loss": 1.6897,
+      "step": 476
+    },
+    {
+      "epoch": 0.4403415647357489,
+      "grad_norm": 5.48828125,
+      "learning_rate": 9.688458972648434e-05,
+      "loss": 1.6875,
+      "step": 477
+    },
+    {
+      "epoch": 0.4412647126702054,
+      "grad_norm": 6.37109375,
+      "learning_rate": 9.687791861240828e-05,
+      "loss": 1.8765,
+      "step": 478
+    },
+    {
+      "epoch": 0.4421878606046619,
+      "grad_norm": 8.140625,
+      "learning_rate": 9.687124749833223e-05,
+      "loss": 1.9128,
+      "step": 479
+    },
+    {
+      "epoch": 0.4431110085391184,
+      "grad_norm": 6.234375,
+      "learning_rate": 9.686457638425617e-05,
+      "loss": 1.6284,
+      "step": 480
+    },
+    {
+      "epoch": 0.4440341564735749,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.685790527018013e-05,
+      "loss": 1.3005,
+      "step": 481
+    },
+    {
+      "epoch": 0.4449573044080314,
+      "grad_norm": 4.6796875,
+      "learning_rate": 9.685123415610408e-05,
+      "loss": 1.6204,
+      "step": 482
+    },
+    {
+      "epoch": 0.4458804523424879,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.684456304202802e-05,
+      "loss": 1.8169,
+      "step": 483
+    },
+    {
+      "epoch": 0.4468036002769444,
+      "grad_norm": 5.30859375,
+      "learning_rate": 9.683789192795197e-05,
+      "loss": 1.4808,
+      "step": 484
+    },
+    {
+      "epoch": 0.4477267482114009,
+      "grad_norm": 8.140625,
+      "learning_rate": 9.683122081387593e-05,
+      "loss": 1.7424,
+      "step": 485
+    },
+    {
+      "epoch": 0.4486498961458574,
+      "grad_norm": 5.1484375,
+      "learning_rate": 9.682454969979987e-05,
+      "loss": 1.7678,
+      "step": 486
+    },
+    {
+      "epoch": 0.44957304408031384,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.681787858572382e-05,
+      "loss": 1.6692,
+      "step": 487
+    },
+    {
+      "epoch": 0.45049619201477037,
+      "grad_norm": 5.8203125,
+      "learning_rate": 9.681120747164776e-05,
+      "loss": 1.9548,
+      "step": 488
+    },
+    {
+      "epoch": 0.45141933994922684,
+      "grad_norm": 8.2421875,
+      "learning_rate": 9.680453635757172e-05,
+      "loss": 1.6488,
+      "step": 489
+    },
+    {
+      "epoch": 0.45234248788368336,
+      "grad_norm": 6.671875,
+      "learning_rate": 9.679786524349567e-05,
+      "loss": 1.7976,
+      "step": 490
+    },
+    {
+      "epoch": 0.45326563581813983,
+      "grad_norm": 7.57421875,
+      "learning_rate": 9.679119412941961e-05,
+      "loss": 1.8059,
+      "step": 491
+    },
+    {
+      "epoch": 0.45418878375259636,
+      "grad_norm": 5.5234375,
+      "learning_rate": 9.678452301534356e-05,
+      "loss": 1.4451,
+      "step": 492
+    },
+    {
+      "epoch": 0.45511193168705283,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.677785190126752e-05,
+      "loss": 1.9504,
+      "step": 493
+    },
+    {
+      "epoch": 0.45603507962150935,
+      "grad_norm": 5.3203125,
+      "learning_rate": 9.677118078719146e-05,
+      "loss": 1.5417,
+      "step": 494
+    },
+    {
+      "epoch": 0.4569582275559658,
+      "grad_norm": 5.8515625,
+      "learning_rate": 9.676450967311541e-05,
+      "loss": 1.8123,
+      "step": 495
+    },
+    {
+      "epoch": 0.45788137549042235,
+      "grad_norm": 5.43359375,
+      "learning_rate": 9.675783855903935e-05,
+      "loss": 1.6172,
+      "step": 496
+    },
+    {
+      "epoch": 0.4588045234248788,
+      "grad_norm": 73.0,
+      "learning_rate": 9.675116744496331e-05,
+      "loss": 2.8965,
+      "step": 497
+    },
+    {
+      "epoch": 0.45972767135933534,
+      "grad_norm": 95.1875,
+      "learning_rate": 9.674449633088726e-05,
+      "loss": 3.8291,
+      "step": 498
+    },
+    {
+      "epoch": 0.4606508192937918,
+      "grad_norm": 6.1328125,
+      "learning_rate": 9.673782521681122e-05,
+      "loss": 1.7273,
+      "step": 499
+    },
+    {
+      "epoch": 0.46157396722824834,
+      "grad_norm": 6.2265625,
+      "learning_rate": 9.673115410273516e-05,
+      "loss": 2.0942,
+      "step": 500
+    },
+    {
+      "epoch": 0.4624971151627048,
+      "grad_norm": 7.61328125,
+      "learning_rate": 9.672448298865911e-05,
+      "loss": 2.0205,
+      "step": 501
+    },
+    {
+      "epoch": 0.46342026309716133,
+      "grad_norm": 6.74609375,
+      "learning_rate": 9.671781187458307e-05,
+      "loss": 1.9387,
+      "step": 502
+    },
+    {
+      "epoch": 0.4643434110316178,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.671114076050701e-05,
+      "loss": 1.7751,
+      "step": 503
+    },
+    {
+      "epoch": 0.46526655896607433,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.670446964643096e-05,
+      "loss": 1.8804,
+      "step": 504
+    },
+    {
+      "epoch": 0.4661897069005308,
+      "grad_norm": 5.3046875,
+      "learning_rate": 9.669779853235492e-05,
+      "loss": 1.7546,
+      "step": 505
+    },
+    {
+      "epoch": 0.4671128548349873,
+      "grad_norm": 7.39453125,
+      "learning_rate": 9.669112741827886e-05,
+      "loss": 1.9646,
+      "step": 506
+    },
+    {
+      "epoch": 0.4680360027694438,
+      "grad_norm": 6.828125,
+      "learning_rate": 9.668445630420281e-05,
+      "loss": 1.5659,
+      "step": 507
+    },
+    {
+      "epoch": 0.4689591507039003,
+      "grad_norm": 6.4765625,
+      "learning_rate": 9.667778519012675e-05,
+      "loss": 1.7634,
+      "step": 508
+    },
+    {
+      "epoch": 0.4698822986383568,
+      "grad_norm": 6.31640625,
+      "learning_rate": 9.667111407605071e-05,
+      "loss": 1.6632,
+      "step": 509
+    },
+    {
+      "epoch": 0.4708054465728133,
+      "grad_norm": 6.6953125,
+      "learning_rate": 9.666444296197466e-05,
+      "loss": 1.7966,
+      "step": 510
+    },
+    {
+      "epoch": 0.4717285945072698,
+      "grad_norm": 6.1640625,
+      "learning_rate": 9.66577718478986e-05,
+      "loss": 1.6934,
+      "step": 511
+    },
+    {
+      "epoch": 0.4726517424417263,
+      "grad_norm": 6.6171875,
+      "learning_rate": 9.665110073382255e-05,
+      "loss": 1.595,
+      "step": 512
+    },
+    {
+      "epoch": 0.4735748903761828,
+      "grad_norm": 4.90234375,
+      "learning_rate": 9.664442961974651e-05,
+      "loss": 1.7456,
+      "step": 513
+    },
+    {
+      "epoch": 0.4744980383106393,
+      "grad_norm": 5.484375,
+      "learning_rate": 9.663775850567045e-05,
+      "loss": 1.9524,
+      "step": 514
+    },
+    {
+      "epoch": 0.4754211862450958,
+      "grad_norm": 5.65234375,
+      "learning_rate": 9.66310873915944e-05,
+      "loss": 1.9097,
+      "step": 515
+    },
+    {
+      "epoch": 0.4763443341795523,
+      "grad_norm": 6.34765625,
+      "learning_rate": 9.662441627751834e-05,
+      "loss": 1.7061,
+      "step": 516
+    },
+    {
+      "epoch": 0.47726748211400877,
+      "grad_norm": 6.80078125,
+      "learning_rate": 9.66177451634423e-05,
+      "loss": 1.8489,
+      "step": 517
+    },
+    {
+      "epoch": 0.4781906300484653,
+      "grad_norm": 7.96484375,
+      "learning_rate": 9.661107404936625e-05,
+      "loss": 1.8684,
+      "step": 518
+    },
+    {
+      "epoch": 0.47911377798292176,
+      "grad_norm": 6.7890625,
+      "learning_rate": 9.66044029352902e-05,
+      "loss": 1.7974,
+      "step": 519
+    },
+    {
+      "epoch": 0.48003692591737823,
+      "grad_norm": 8.859375,
+      "learning_rate": 9.659773182121414e-05,
+      "loss": 1.8508,
+      "step": 520
+    },
+    {
+      "epoch": 0.48096007385183476,
+      "grad_norm": 5.4765625,
+      "learning_rate": 9.65910607071381e-05,
+      "loss": 2.0354,
+      "step": 521
+    },
+    {
+      "epoch": 0.4818832217862912,
+      "grad_norm": 7.5390625,
+      "learning_rate": 9.658438959306204e-05,
+      "loss": 1.6227,
+      "step": 522
+    },
+    {
+      "epoch": 0.48280636972074775,
+      "grad_norm": 6.375,
+      "learning_rate": 9.657771847898599e-05,
+      "loss": 1.6504,
+      "step": 523
+    },
+    {
+      "epoch": 0.4837295176552042,
+      "grad_norm": 6.3125,
+      "learning_rate": 9.657104736490995e-05,
+      "loss": 1.7908,
+      "step": 524
+    },
+    {
+      "epoch": 0.48465266558966075,
+      "grad_norm": 6.44921875,
+      "learning_rate": 9.65643762508339e-05,
+      "loss": 1.9624,
+      "step": 525
+    },
+    {
+      "epoch": 0.4855758135241172,
+      "grad_norm": 4.9296875,
+      "learning_rate": 9.655770513675785e-05,
+      "loss": 1.675,
+      "step": 526
+    },
+    {
+      "epoch": 0.48649896145857374,
+      "grad_norm": 5.19921875,
+      "learning_rate": 9.65510340226818e-05,
+      "loss": 1.8286,
+      "step": 527
+    },
+    {
+      "epoch": 0.4874221093930302,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.654436290860574e-05,
+      "loss": 1.8921,
+      "step": 528
+    },
+    {
+      "epoch": 0.48834525732748674,
+      "grad_norm": 5.953125,
+      "learning_rate": 9.653769179452969e-05,
+      "loss": 1.8113,
+      "step": 529
+    },
+    {
+      "epoch": 0.4892684052619432,
+      "grad_norm": 6.6640625,
+      "learning_rate": 9.653102068045365e-05,
+      "loss": 1.5984,
+      "step": 530
+    },
+    {
+      "epoch": 0.49019155319639973,
+      "grad_norm": 6.32421875,
+      "learning_rate": 9.65243495663776e-05,
+      "loss": 1.7388,
+      "step": 531
+    },
+    {
+      "epoch": 0.4911147011308562,
+      "grad_norm": 8.296875,
+      "learning_rate": 9.651767845230154e-05,
+      "loss": 1.7832,
+      "step": 532
+    },
+    {
+      "epoch": 0.4920378490653127,
+      "grad_norm": 5.96484375,
+      "learning_rate": 9.651100733822548e-05,
+      "loss": 1.7662,
+      "step": 533
+    },
+    {
+      "epoch": 0.4929609969997692,
+      "grad_norm": 6.796875,
+      "learning_rate": 9.650433622414944e-05,
+      "loss": 1.5933,
+      "step": 534
+    },
+    {
+      "epoch": 0.4938841449342257,
+      "grad_norm": 4.88671875,
+      "learning_rate": 9.649766511007339e-05,
+      "loss": 1.8276,
+      "step": 535
+    },
+    {
+      "epoch": 0.4948072928686822,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.649099399599733e-05,
+      "loss": 1.7932,
+      "step": 536
+    },
+    {
+      "epoch": 0.4957304408031387,
+      "grad_norm": 6.3359375,
+      "learning_rate": 9.648432288192128e-05,
+      "loss": 1.4583,
+      "step": 537
+    },
+    {
+      "epoch": 0.4966535887375952,
+      "grad_norm": 6.828125,
+      "learning_rate": 9.647765176784524e-05,
+      "loss": 1.8784,
+      "step": 538
+    },
+    {
+      "epoch": 0.4975767366720517,
+      "grad_norm": 6.05078125,
+      "learning_rate": 9.647098065376918e-05,
+      "loss": 1.6936,
+      "step": 539
+    },
+    {
+      "epoch": 0.4984998846065082,
+      "grad_norm": 8.625,
+      "learning_rate": 9.646430953969313e-05,
+      "loss": 1.5133,
+      "step": 540
+    },
+    {
+      "epoch": 0.4994230325409647,
+      "grad_norm": 5.23828125,
+      "learning_rate": 9.645763842561707e-05,
+      "loss": 1.5984,
+      "step": 541
+    },
+    {
+      "epoch": 0.5003461804754212,
+      "grad_norm": 5.875,
+      "learning_rate": 9.645096731154103e-05,
+      "loss": 1.8914,
+      "step": 542
+    },
+    {
+      "epoch": 0.5012693284098777,
+      "grad_norm": 4.83984375,
+      "learning_rate": 9.644429619746498e-05,
+      "loss": 1.7124,
+      "step": 543
+    },
+    {
+      "epoch": 0.5021924763443342,
+      "grad_norm": 7.6171875,
+      "learning_rate": 9.643762508338892e-05,
+      "loss": 1.9399,
+      "step": 544
+    },
+    {
+      "epoch": 0.5031156242787906,
+      "grad_norm": 5.10546875,
+      "learning_rate": 9.643095396931287e-05,
+      "loss": 1.8704,
+      "step": 545
+    },
+    {
+      "epoch": 0.5040387722132472,
+      "grad_norm": 7.140625,
+      "learning_rate": 9.642428285523683e-05,
+      "loss": 1.7798,
+      "step": 546
+    },
+    {
+      "epoch": 0.5049619201477037,
+      "grad_norm": 5.6484375,
+      "learning_rate": 9.641761174116077e-05,
+      "loss": 1.7075,
+      "step": 547
+    },
+    {
+      "epoch": 0.5058850680821602,
+      "grad_norm": 8.609375,
+      "learning_rate": 9.641094062708472e-05,
+      "loss": 1.7932,
+      "step": 548
+    },
+    {
+      "epoch": 0.5068082160166166,
+      "grad_norm": 5.70703125,
+      "learning_rate": 9.640426951300868e-05,
+      "loss": 1.2843,
+      "step": 549
+    },
+    {
+      "epoch": 0.5077313639510732,
+      "grad_norm": 7.9609375,
+      "learning_rate": 9.639759839893264e-05,
+      "loss": 1.9604,
+      "step": 550
+    },
+    {
+      "epoch": 0.5086545118855297,
+      "grad_norm": 7.8046875,
+      "learning_rate": 9.639092728485658e-05,
+      "loss": 1.8276,
+      "step": 551
+    },
+    {
+      "epoch": 0.5095776598199862,
+      "grad_norm": 7.21484375,
+      "learning_rate": 9.638425617078053e-05,
+      "loss": 1.948,
+      "step": 552
+    },
+    {
+      "epoch": 0.5105008077544426,
+      "grad_norm": 5.921875,
+      "learning_rate": 9.637758505670447e-05,
+      "loss": 1.7626,
+      "step": 553
+    },
+    {
+      "epoch": 0.5114239556888992,
+      "grad_norm": 6.453125,
+      "learning_rate": 9.637091394262843e-05,
+      "loss": 1.3768,
+      "step": 554
+    },
+    {
+      "epoch": 0.5123471036233557,
+      "grad_norm": 5.88671875,
+      "learning_rate": 9.636424282855238e-05,
+      "loss": 1.603,
+      "step": 555
+    },
+    {
+      "epoch": 0.5132702515578121,
+      "grad_norm": 6.5078125,
+      "learning_rate": 9.635757171447632e-05,
+      "loss": 1.7952,
+      "step": 556
+    },
+    {
+      "epoch": 0.5141933994922686,
+      "grad_norm": 8.484375,
+      "learning_rate": 9.635090060040027e-05,
+      "loss": 1.7161,
+      "step": 557
+    },
+    {
+      "epoch": 0.5151165474267251,
+      "grad_norm": 5.98046875,
+      "learning_rate": 9.634422948632423e-05,
+      "loss": 1.6995,
+      "step": 558
+    },
+    {
+      "epoch": 0.5160396953611817,
+      "grad_norm": 7.5390625,
+      "learning_rate": 9.633755837224817e-05,
+      "loss": 1.8037,
+      "step": 559
+    },
+    {
+      "epoch": 0.5169628432956381,
+      "grad_norm": 8.1171875,
+      "learning_rate": 9.633088725817212e-05,
+      "loss": 1.4825,
+      "step": 560
+    },
+    {
+      "epoch": 0.5178859912300946,
+      "grad_norm": 5.81640625,
+      "learning_rate": 9.632421614409606e-05,
+      "loss": 1.5466,
+      "step": 561
+    },
+    {
+      "epoch": 0.5188091391645511,
+      "grad_norm": 5.7109375,
+      "learning_rate": 9.631754503002002e-05,
+      "loss": 1.8303,
+      "step": 562
+    },
+    {
+      "epoch": 0.5197322870990077,
+      "grad_norm": 8.7734375,
+      "learning_rate": 9.631087391594397e-05,
+      "loss": 1.8416,
+      "step": 563
+    },
+    {
+      "epoch": 0.5206554350334641,
+      "grad_norm": 5.078125,
+      "learning_rate": 9.630420280186791e-05,
+      "loss": 1.679,
+      "step": 564
+    },
+    {
+      "epoch": 0.5215785829679206,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.629753168779186e-05,
+      "loss": 1.9231,
+      "step": 565
+    },
+    {
+      "epoch": 0.5225017309023771,
+      "grad_norm": 5.88671875,
+      "learning_rate": 9.629086057371582e-05,
+      "loss": 1.5483,
+      "step": 566
+    },
+    {
+      "epoch": 0.5234248788368336,
+      "grad_norm": 7.24609375,
+      "learning_rate": 9.628418945963976e-05,
+      "loss": 1.3227,
+      "step": 567
+    },
+    {
+      "epoch": 0.5243480267712901,
+      "grad_norm": 5.48828125,
+      "learning_rate": 9.627751834556371e-05,
+      "loss": 1.8123,
+      "step": 568
+    },
+    {
+      "epoch": 0.5252711747057466,
+      "grad_norm": 6.14453125,
+      "learning_rate": 9.627084723148766e-05,
+      "loss": 1.97,
+      "step": 569
+    },
+    {
+      "epoch": 0.526194322640203,
+      "grad_norm": 5.64453125,
+      "learning_rate": 9.626417611741161e-05,
+      "loss": 1.7124,
+      "step": 570
+    },
+    {
+      "epoch": 0.5271174705746596,
+      "grad_norm": 5.75,
+      "learning_rate": 9.625750500333556e-05,
+      "loss": 1.6763,
+      "step": 571
+    },
+    {
+      "epoch": 0.5280406185091161,
+      "grad_norm": 6.3984375,
+      "learning_rate": 9.62508338892595e-05,
+      "loss": 1.6902,
+      "step": 572
+    },
+    {
+      "epoch": 0.5289637664435726,
+      "grad_norm": 6.14453125,
+      "learning_rate": 9.624416277518345e-05,
+      "loss": 1.9634,
+      "step": 573
+    },
+    {
+      "epoch": 0.529886914378029,
+      "grad_norm": 6.1484375,
+      "learning_rate": 9.623749166110741e-05,
+      "loss": 1.8433,
+      "step": 574
+    },
+    {
+      "epoch": 0.5308100623124856,
+      "grad_norm": 8.0234375,
+      "learning_rate": 9.623082054703137e-05,
+      "loss": 1.9392,
+      "step": 575
+    },
+    {
+      "epoch": 0.5317332102469421,
+      "grad_norm": 5.74609375,
+      "learning_rate": 9.622414943295531e-05,
+      "loss": 1.8005,
+      "step": 576
+    },
+    {
+      "epoch": 0.5326563581813986,
+      "grad_norm": 7.8515625,
+      "learning_rate": 9.621747831887926e-05,
+      "loss": 1.4319,
+      "step": 577
+    },
+    {
+      "epoch": 0.533579506115855,
+      "grad_norm": 4.890625,
+      "learning_rate": 9.62108072048032e-05,
+      "loss": 1.7603,
+      "step": 578
+    },
+    {
+      "epoch": 0.5345026540503116,
+      "grad_norm": 5.35546875,
+      "learning_rate": 9.620413609072716e-05,
+      "loss": 1.7097,
+      "step": 579
+    },
+    {
+      "epoch": 0.5354258019847681,
+      "grad_norm": 4.83203125,
+      "learning_rate": 9.619746497665111e-05,
+      "loss": 1.8438,
+      "step": 580
+    },
+    {
+      "epoch": 0.5363489499192245,
+      "grad_norm": 5.171875,
+      "learning_rate": 9.619079386257505e-05,
+      "loss": 1.4211,
+      "step": 581
+    },
+    {
+      "epoch": 0.537272097853681,
+      "grad_norm": 4.91015625,
+      "learning_rate": 9.6184122748499e-05,
+      "loss": 1.5386,
+      "step": 582
+    },
+    {
+      "epoch": 0.5381952457881376,
+      "grad_norm": 5.69921875,
+      "learning_rate": 9.617745163442296e-05,
+      "loss": 1.7856,
+      "step": 583
+    },
+    {
+      "epoch": 0.5391183937225941,
+      "grad_norm": 5.93359375,
+      "learning_rate": 9.61707805203469e-05,
+      "loss": 1.9592,
+      "step": 584
+    },
+    {
+      "epoch": 0.5400415416570505,
+      "grad_norm": 8.375,
+      "learning_rate": 9.616410940627085e-05,
+      "loss": 1.7913,
+      "step": 585
+    },
+    {
+      "epoch": 0.540964689591507,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.61574382921948e-05,
+      "loss": 1.6965,
+      "step": 586
+    },
+    {
+      "epoch": 0.5418878375259636,
+      "grad_norm": 6.7421875,
+      "learning_rate": 9.615076717811875e-05,
+      "loss": 1.9368,
+      "step": 587
+    },
+    {
+      "epoch": 0.5428109854604201,
+      "grad_norm": 5.625,
+      "learning_rate": 9.61440960640427e-05,
+      "loss": 1.6428,
+      "step": 588
+    },
+    {
+      "epoch": 0.5437341333948765,
+      "grad_norm": 4.81640625,
+      "learning_rate": 9.613742494996665e-05,
+      "loss": 1.4694,
+      "step": 589
+    },
+    {
+      "epoch": 0.544657281329333,
+      "grad_norm": 5.72265625,
+      "learning_rate": 9.613075383589059e-05,
+      "loss": 1.8787,
+      "step": 590
+    },
+    {
+      "epoch": 0.5455804292637895,
+      "grad_norm": 4.57421875,
+      "learning_rate": 9.612408272181455e-05,
+      "loss": 1.3748,
+      "step": 591
+    },
+    {
+      "epoch": 0.546503577198246,
+      "grad_norm": 5.63671875,
+      "learning_rate": 9.61174116077385e-05,
+      "loss": 1.7695,
+      "step": 592
+    },
+    {
+      "epoch": 0.5474267251327025,
+      "grad_norm": 6.31640625,
+      "learning_rate": 9.611074049366244e-05,
+      "loss": 1.6885,
+      "step": 593
+    },
+    {
+      "epoch": 0.548349873067159,
+      "grad_norm": 7.1484375,
+      "learning_rate": 9.610406937958639e-05,
+      "loss": 2.053,
+      "step": 594
+    },
+    {
+      "epoch": 0.5492730210016155,
+      "grad_norm": 5.609375,
+      "learning_rate": 9.609739826551034e-05,
+      "loss": 1.6252,
+      "step": 595
+    },
+    {
+      "epoch": 0.550196168936072,
+      "grad_norm": 6.0859375,
+      "learning_rate": 9.609072715143429e-05,
+      "loss": 1.752,
+      "step": 596
+    },
+    {
+      "epoch": 0.5511193168705285,
+      "grad_norm": 7.35546875,
+      "learning_rate": 9.608405603735824e-05,
+      "loss": 1.5771,
+      "step": 597
+    },
+    {
+      "epoch": 0.552042464804985,
+      "grad_norm": 5.4609375,
+      "learning_rate": 9.607738492328218e-05,
+      "loss": 1.3693,
+      "step": 598
+    },
+    {
+      "epoch": 0.5529656127394414,
+      "grad_norm": 6.55078125,
+      "learning_rate": 9.607071380920614e-05,
+      "loss": 2.0869,
+      "step": 599
+    },
+    {
+      "epoch": 0.553888760673898,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.60640426951301e-05,
+      "loss": 1.4077,
+      "step": 600
+    },
+    {
+      "epoch": 0.5548119086083545,
+      "grad_norm": 4.9453125,
+      "learning_rate": 9.605737158105404e-05,
+      "loss": 1.3591,
+      "step": 601
+    },
+    {
+      "epoch": 0.555735056542811,
+      "grad_norm": 5.45703125,
+      "learning_rate": 9.605070046697799e-05,
+      "loss": 1.6968,
+      "step": 602
+    },
+    {
+      "epoch": 0.5566582044772674,
+      "grad_norm": 5.58203125,
+      "learning_rate": 9.604402935290195e-05,
+      "loss": 1.1097,
+      "step": 603
+    },
+    {
+      "epoch": 0.557581352411724,
+      "grad_norm": 4.79296875,
+      "learning_rate": 9.60373582388259e-05,
+      "loss": 1.476,
+      "step": 604
+    },
+    {
+      "epoch": 0.5585045003461805,
+      "grad_norm": 6.58203125,
+      "learning_rate": 9.603068712474984e-05,
+      "loss": 1.9341,
+      "step": 605
+    },
+    {
+      "epoch": 0.559427648280637,
+      "grad_norm": 7.078125,
+      "learning_rate": 9.602401601067379e-05,
+      "loss": 1.8635,
+      "step": 606
+    },
+    {
+      "epoch": 0.5603507962150934,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.601734489659774e-05,
+      "loss": 1.6674,
+      "step": 607
+    },
+    {
+      "epoch": 0.56127394414955,
+      "grad_norm": 5.78515625,
+      "learning_rate": 9.601067378252169e-05,
+      "loss": 1.6882,
+      "step": 608
+    },
+    {
+      "epoch": 0.5621970920840065,
+      "grad_norm": 8.453125,
+      "learning_rate": 9.600400266844564e-05,
+      "loss": 1.8459,
+      "step": 609
+    },
+    {
+      "epoch": 0.563120240018463,
+      "grad_norm": 5.62890625,
+      "learning_rate": 9.599733155436958e-05,
+      "loss": 1.5283,
+      "step": 610
+    },
+    {
+      "epoch": 0.5640433879529194,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.599066044029354e-05,
+      "loss": 1.6772,
+      "step": 611
+    },
+    {
+      "epoch": 0.564966535887376,
+      "grad_norm": 5.23046875,
+      "learning_rate": 9.598398932621749e-05,
+      "loss": 1.4495,
+      "step": 612
+    },
+    {
+      "epoch": 0.5658896838218325,
+      "grad_norm": 7.49609375,
+      "learning_rate": 9.597731821214143e-05,
+      "loss": 1.885,
+      "step": 613
+    },
+    {
+      "epoch": 0.5668128317562889,
+      "grad_norm": 6.80859375,
+      "learning_rate": 9.597064709806538e-05,
+      "loss": 1.9661,
+      "step": 614
+    },
+    {
+      "epoch": 0.5677359796907454,
+      "grad_norm": 8.2421875,
+      "learning_rate": 9.596397598398933e-05,
+      "loss": 1.9792,
+      "step": 615
+    },
+    {
+      "epoch": 0.568659127625202,
+      "grad_norm": 7.31640625,
+      "learning_rate": 9.595730486991328e-05,
+      "loss": 1.8926,
+      "step": 616
+    },
+    {
+      "epoch": 0.5695822755596585,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.595063375583723e-05,
+      "loss": 1.7109,
+      "step": 617
+    },
+    {
+      "epoch": 0.5705054234941149,
+      "grad_norm": 7.7265625,
+      "learning_rate": 9.594396264176117e-05,
+      "loss": 1.8887,
+      "step": 618
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 7.015625,
+      "learning_rate": 9.593729152768513e-05,
+      "loss": 1.5822,
+      "step": 619
+    },
+    {
+      "epoch": 0.572351719363028,
+      "grad_norm": 4.16796875,
+      "learning_rate": 9.593062041360908e-05,
+      "loss": 1.9116,
+      "step": 620
+    },
+    {
+      "epoch": 0.5732748672974844,
+      "grad_norm": 5.5,
+      "learning_rate": 9.592394929953302e-05,
+      "loss": 1.6375,
+      "step": 621
+    },
+    {
+      "epoch": 0.5741980152319409,
+      "grad_norm": 6.30078125,
+      "learning_rate": 9.591727818545697e-05,
+      "loss": 1.634,
+      "step": 622
+    },
+    {
+      "epoch": 0.5751211631663974,
+      "grad_norm": 7.296875,
+      "learning_rate": 9.591060707138093e-05,
+      "loss": 1.655,
+      "step": 623
+    },
+    {
+      "epoch": 0.5760443111008539,
+      "grad_norm": 6.33203125,
+      "learning_rate": 9.590393595730487e-05,
+      "loss": 1.9722,
+      "step": 624
+    },
+    {
+      "epoch": 0.5769674590353104,
+      "grad_norm": 7.38671875,
+      "learning_rate": 9.589726484322883e-05,
+      "loss": 1.937,
+      "step": 625
+    },
+    {
+      "epoch": 0.5778906069697669,
+      "grad_norm": 4.765625,
+      "learning_rate": 9.589059372915278e-05,
+      "loss": 1.9639,
+      "step": 626
+    },
+    {
+      "epoch": 0.5788137549042234,
+      "grad_norm": 6.9296875,
+      "learning_rate": 9.588392261507672e-05,
+      "loss": 1.6255,
+      "step": 627
+    },
+    {
+      "epoch": 0.5797369028386798,
+      "grad_norm": 5.94140625,
+      "learning_rate": 9.587725150100068e-05,
+      "loss": 1.6733,
+      "step": 628
+    },
+    {
+      "epoch": 0.5806600507731364,
+      "grad_norm": 7.5078125,
+      "learning_rate": 9.587058038692463e-05,
+      "loss": 1.8135,
+      "step": 629
+    },
+    {
+      "epoch": 0.5815831987075929,
+      "grad_norm": 6.515625,
+      "learning_rate": 9.586390927284857e-05,
+      "loss": 1.5801,
+      "step": 630
+    },
+    {
+      "epoch": 0.5825063466420494,
+      "grad_norm": 6.92578125,
+      "learning_rate": 9.585723815877252e-05,
+      "loss": 1.7769,
+      "step": 631
+    },
+    {
+      "epoch": 0.5834294945765058,
+      "grad_norm": 6.37109375,
+      "learning_rate": 9.585056704469647e-05,
+      "loss": 1.6985,
+      "step": 632
+    },
+    {
+      "epoch": 0.5843526425109624,
+      "grad_norm": 6.82421875,
+      "learning_rate": 9.584389593062042e-05,
+      "loss": 2.0203,
+      "step": 633
+    },
+    {
+      "epoch": 0.5852757904454189,
+      "grad_norm": 5.39453125,
+      "learning_rate": 9.583722481654437e-05,
+      "loss": 1.562,
+      "step": 634
+    },
+    {
+      "epoch": 0.5861989383798754,
+      "grad_norm": 5.41796875,
+      "learning_rate": 9.583055370246831e-05,
+      "loss": 1.5032,
+      "step": 635
+    },
+    {
+      "epoch": 0.5871220863143318,
+      "grad_norm": 7.88671875,
+      "learning_rate": 9.582388258839227e-05,
+      "loss": 1.9116,
+      "step": 636
+    },
+    {
+      "epoch": 0.5880452342487884,
+      "grad_norm": 6.26171875,
+      "learning_rate": 9.581721147431622e-05,
+      "loss": 1.6079,
+      "step": 637
+    },
+    {
+      "epoch": 0.5889683821832449,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.581054036024016e-05,
+      "loss": 1.8486,
+      "step": 638
+    },
+    {
+      "epoch": 0.5898915301177013,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.58038692461641e-05,
+      "loss": 1.8347,
+      "step": 639
+    },
+    {
+      "epoch": 0.5908146780521578,
+      "grad_norm": 7.3046875,
+      "learning_rate": 9.579719813208807e-05,
+      "loss": 1.9507,
+      "step": 640
+    },
+    {
+      "epoch": 0.5917378259866144,
+      "grad_norm": 7.0078125,
+      "learning_rate": 9.579052701801201e-05,
+      "loss": 1.5554,
+      "step": 641
+    },
+    {
+      "epoch": 0.5926609739210709,
+      "grad_norm": 8.4765625,
+      "learning_rate": 9.578385590393596e-05,
+      "loss": 1.7551,
+      "step": 642
+    },
+    {
+      "epoch": 0.5935841218555273,
+      "grad_norm": 5.98828125,
+      "learning_rate": 9.57771847898599e-05,
+      "loss": 1.7261,
+      "step": 643
+    },
+    {
+      "epoch": 0.5945072697899838,
+      "grad_norm": 7.76953125,
+      "learning_rate": 9.577051367578386e-05,
+      "loss": 1.7295,
+      "step": 644
+    },
+    {
+      "epoch": 0.5954304177244404,
+      "grad_norm": 5.25,
+      "learning_rate": 9.57638425617078e-05,
+      "loss": 1.0792,
+      "step": 645
+    },
+    {
+      "epoch": 0.5963535656588969,
+      "grad_norm": 5.73046875,
+      "learning_rate": 9.575717144763175e-05,
+      "loss": 1.6802,
+      "step": 646
+    },
+    {
+      "epoch": 0.5972767135933533,
+      "grad_norm": 5.57421875,
+      "learning_rate": 9.57505003335557e-05,
+      "loss": 1.7537,
+      "step": 647
+    },
+    {
+      "epoch": 0.5981998615278098,
+      "grad_norm": 7.671875,
+      "learning_rate": 9.574382921947966e-05,
+      "loss": 1.5603,
+      "step": 648
+    },
+    {
+      "epoch": 0.5991230094622664,
+      "grad_norm": 6.875,
+      "learning_rate": 9.57371581054036e-05,
+      "loss": 1.467,
+      "step": 649
+    },
+    {
+      "epoch": 0.6000461573967228,
+      "grad_norm": 7.75390625,
+      "learning_rate": 9.573048699132756e-05,
+      "loss": 1.6992,
+      "step": 650
+    },
+    {
+      "epoch": 0.6009693053311793,
+      "grad_norm": 4.68359375,
+      "learning_rate": 9.57238158772515e-05,
+      "loss": 1.563,
+      "step": 651
+    },
+    {
+      "epoch": 0.6018924532656358,
+      "grad_norm": 5.890625,
+      "learning_rate": 9.571714476317546e-05,
+      "loss": 1.4746,
+      "step": 652
+    },
+    {
+      "epoch": 0.6028156012000924,
+      "grad_norm": 5.4765625,
+      "learning_rate": 9.571047364909941e-05,
+      "loss": 1.4298,
+      "step": 653
+    },
+    {
+      "epoch": 0.6037387491345488,
+      "grad_norm": 5.4609375,
+      "learning_rate": 9.570380253502336e-05,
+      "loss": 1.968,
+      "step": 654
+    },
+    {
+      "epoch": 0.6046618970690053,
+      "grad_norm": 7.06640625,
+      "learning_rate": 9.56971314209473e-05,
+      "loss": 1.7017,
+      "step": 655
+    },
+    {
+      "epoch": 0.6055850450034618,
+      "grad_norm": 7.296875,
+      "learning_rate": 9.569046030687126e-05,
+      "loss": 1.7031,
+      "step": 656
+    },
+    {
+      "epoch": 0.6065081929379184,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.56837891927952e-05,
+      "loss": 1.4336,
+      "step": 657
+    },
+    {
+      "epoch": 0.6074313408723748,
+      "grad_norm": 6.71484375,
+      "learning_rate": 9.567711807871915e-05,
+      "loss": 1.585,
+      "step": 658
+    },
+    {
+      "epoch": 0.6083544888068313,
+      "grad_norm": 4.5,
+      "learning_rate": 9.56704469646431e-05,
+      "loss": 1.6884,
+      "step": 659
+    },
+    {
+      "epoch": 0.6092776367412878,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.566377585056706e-05,
+      "loss": 1.769,
+      "step": 660
+    },
+    {
+      "epoch": 0.6102007846757442,
+      "grad_norm": 6.95703125,
+      "learning_rate": 9.5657104736491e-05,
+      "loss": 1.761,
+      "step": 661
+    },
+    {
+      "epoch": 0.6111239326102008,
+      "grad_norm": 6.9296875,
+      "learning_rate": 9.565043362241495e-05,
+      "loss": 1.6655,
+      "step": 662
+    },
+    {
+      "epoch": 0.6120470805446573,
+      "grad_norm": 5.51953125,
+      "learning_rate": 9.564376250833889e-05,
+      "loss": 1.6699,
+      "step": 663
+    },
+    {
+      "epoch": 0.6129702284791138,
+      "grad_norm": 4.58203125,
+      "learning_rate": 9.563709139426285e-05,
+      "loss": 1.2987,
+      "step": 664
+    },
+    {
+      "epoch": 0.6138933764135702,
+      "grad_norm": 5.234375,
+      "learning_rate": 9.56304202801868e-05,
+      "loss": 1.5502,
+      "step": 665
+    },
+    {
+      "epoch": 0.6148165243480268,
+      "grad_norm": 5.8671875,
+      "learning_rate": 9.562374916611074e-05,
+      "loss": 1.979,
+      "step": 666
+    },
+    {
+      "epoch": 0.6157396722824833,
+      "grad_norm": 6.76953125,
+      "learning_rate": 9.561707805203469e-05,
+      "loss": 1.6648,
+      "step": 667
+    },
+    {
+      "epoch": 0.6166628202169397,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.561040693795865e-05,
+      "loss": 1.9653,
+      "step": 668
+    },
+    {
+      "epoch": 0.6175859681513962,
+      "grad_norm": 5.16015625,
+      "learning_rate": 9.560373582388259e-05,
+      "loss": 1.8142,
+      "step": 669
+    },
+    {
+      "epoch": 0.6185091160858528,
+      "grad_norm": 4.61328125,
+      "learning_rate": 9.559706470980654e-05,
+      "loss": 1.9058,
+      "step": 670
+    },
+    {
+      "epoch": 0.6194322640203093,
+      "grad_norm": 7.5859375,
+      "learning_rate": 9.559039359573048e-05,
+      "loss": 1.5969,
+      "step": 671
+    },
+    {
+      "epoch": 0.6203554119547657,
+      "grad_norm": 5.8515625,
+      "learning_rate": 9.558372248165444e-05,
+      "loss": 1.6433,
+      "step": 672
+    },
+    {
+      "epoch": 0.6212785598892222,
+      "grad_norm": 6.06640625,
+      "learning_rate": 9.557705136757839e-05,
+      "loss": 1.8311,
+      "step": 673
+    },
+    {
+      "epoch": 0.6222017078236788,
+      "grad_norm": 5.0390625,
+      "learning_rate": 9.557038025350233e-05,
+      "loss": 1.5918,
+      "step": 674
+    },
+    {
+      "epoch": 0.6231248557581353,
+      "grad_norm": 6.52734375,
+      "learning_rate": 9.556370913942629e-05,
+      "loss": 1.7715,
+      "step": 675
+    },
+    {
+      "epoch": 0.6240480036925917,
+      "grad_norm": 5.25390625,
+      "learning_rate": 9.555703802535024e-05,
+      "loss": 1.7236,
+      "step": 676
+    },
+    {
+      "epoch": 0.6249711516270482,
+      "grad_norm": 5.609375,
+      "learning_rate": 9.55503669112742e-05,
+      "loss": 1.6426,
+      "step": 677
+    },
+    {
+      "epoch": 0.6258942995615048,
+      "grad_norm": 5.87890625,
+      "learning_rate": 9.554369579719814e-05,
+      "loss": 1.7766,
+      "step": 678
+    },
+    {
+      "epoch": 0.6268174474959612,
+      "grad_norm": 6.9296875,
+      "learning_rate": 9.553702468312209e-05,
+      "loss": 1.7314,
+      "step": 679
+    },
+    {
+      "epoch": 0.6277405954304177,
+      "grad_norm": 6.390625,
+      "learning_rate": 9.553035356904603e-05,
+      "loss": 1.8062,
+      "step": 680
+    },
+    {
+      "epoch": 0.6286637433648742,
+      "grad_norm": 6.3515625,
+      "learning_rate": 9.552368245496999e-05,
+      "loss": 1.7476,
+      "step": 681
+    },
+    {
+      "epoch": 0.6295868912993308,
+      "grad_norm": 6.3203125,
+      "learning_rate": 9.551701134089394e-05,
+      "loss": 1.5786,
+      "step": 682
+    },
+    {
+      "epoch": 0.6305100392337872,
+      "grad_norm": 7.78515625,
+      "learning_rate": 9.551034022681788e-05,
+      "loss": 1.5842,
+      "step": 683
+    },
+    {
+      "epoch": 0.6314331871682437,
+      "grad_norm": 6.0234375,
+      "learning_rate": 9.550366911274183e-05,
+      "loss": 1.8059,
+      "step": 684
+    },
+    {
+      "epoch": 0.6323563351027002,
+      "grad_norm": 8.0546875,
+      "learning_rate": 9.549699799866579e-05,
+      "loss": 1.7307,
+      "step": 685
+    },
+    {
+      "epoch": 0.6332794830371568,
+      "grad_norm": 5.73828125,
+      "learning_rate": 9.549032688458973e-05,
+      "loss": 1.6541,
+      "step": 686
+    },
+    {
+      "epoch": 0.6342026309716132,
+      "grad_norm": 7.92578125,
+      "learning_rate": 9.548365577051368e-05,
+      "loss": 1.678,
+      "step": 687
+    },
+    {
+      "epoch": 0.6351257789060697,
+      "grad_norm": 6.203125,
+      "learning_rate": 9.547698465643764e-05,
+      "loss": 1.7688,
+      "step": 688
+    },
+    {
+      "epoch": 0.6360489268405262,
+      "grad_norm": 6.140625,
+      "learning_rate": 9.547031354236158e-05,
+      "loss": 1.9966,
+      "step": 689
+    },
+    {
+      "epoch": 0.6369720747749827,
+      "grad_norm": 6.79296875,
+      "learning_rate": 9.546364242828553e-05,
+      "loss": 1.6733,
+      "step": 690
+    },
+    {
+      "epoch": 0.6378952227094392,
+      "grad_norm": 6.22265625,
+      "learning_rate": 9.545697131420947e-05,
+      "loss": 1.8062,
+      "step": 691
+    },
+    {
+      "epoch": 0.6388183706438957,
+      "grad_norm": 4.91015625,
+      "learning_rate": 9.545030020013343e-05,
+      "loss": 1.9199,
+      "step": 692
+    },
+    {
+      "epoch": 0.6397415185783522,
+      "grad_norm": 7.04296875,
+      "learning_rate": 9.544362908605738e-05,
+      "loss": 1.7607,
+      "step": 693
+    },
+    {
+      "epoch": 0.6406646665128086,
+      "grad_norm": 4.8828125,
+      "learning_rate": 9.543695797198132e-05,
+      "loss": 1.9644,
+      "step": 694
+    },
+    {
+      "epoch": 0.6415878144472652,
+      "grad_norm": 8.015625,
+      "learning_rate": 9.543028685790527e-05,
+      "loss": 1.2926,
+      "step": 695
+    },
+    {
+      "epoch": 0.6425109623817217,
+      "grad_norm": 5.20703125,
+      "learning_rate": 9.542361574382923e-05,
+      "loss": 1.8853,
+      "step": 696
+    },
+    {
+      "epoch": 0.6434341103161781,
+      "grad_norm": 5.71484375,
+      "learning_rate": 9.541694462975317e-05,
+      "loss": 1.7791,
+      "step": 697
+    },
+    {
+      "epoch": 0.6443572582506346,
+      "grad_norm": 5.83984375,
+      "learning_rate": 9.541027351567712e-05,
+      "loss": 1.3151,
+      "step": 698
+    },
+    {
+      "epoch": 0.6452804061850912,
+      "grad_norm": 6.1640625,
+      "learning_rate": 9.540360240160106e-05,
+      "loss": 1.9634,
+      "step": 699
+    },
+    {
+      "epoch": 0.6462035541195477,
+      "grad_norm": 5.91796875,
+      "learning_rate": 9.539693128752502e-05,
+      "loss": 1.5525,
+      "step": 700
+    },
+    {
+      "epoch": 0.6471267020540041,
+      "grad_norm": 5.23046875,
+      "learning_rate": 9.539026017344898e-05,
+      "loss": 1.7407,
+      "step": 701
+    },
+    {
+      "epoch": 0.6480498499884606,
+      "grad_norm": 6.9296875,
+      "learning_rate": 9.538358905937293e-05,
+      "loss": 1.7043,
+      "step": 702
+    },
+    {
+      "epoch": 0.6489729979229172,
+      "grad_norm": 7.39453125,
+      "learning_rate": 9.537691794529687e-05,
+      "loss": 1.6333,
+      "step": 703
+    },
+    {
+      "epoch": 0.6498961458573737,
+      "grad_norm": 5.54296875,
+      "learning_rate": 9.537024683122082e-05,
+      "loss": 1.5588,
+      "step": 704
+    },
+    {
+      "epoch": 0.6508192937918301,
+      "grad_norm": 7.22265625,
+      "learning_rate": 9.536357571714478e-05,
+      "loss": 1.78,
+      "step": 705
+    },
+    {
+      "epoch": 0.6517424417262866,
+      "grad_norm": 7.015625,
+      "learning_rate": 9.535690460306872e-05,
+      "loss": 1.5254,
+      "step": 706
+    },
+    {
+      "epoch": 0.6526655896607432,
+      "grad_norm": 5.04296875,
+      "learning_rate": 9.535023348899267e-05,
+      "loss": 1.6833,
+      "step": 707
+    },
+    {
+      "epoch": 0.6535887375951996,
+      "grad_norm": 6.90625,
+      "learning_rate": 9.534356237491661e-05,
+      "loss": 1.6902,
+      "step": 708
+    },
+    {
+      "epoch": 0.6545118855296561,
+      "grad_norm": 6.35546875,
+      "learning_rate": 9.533689126084057e-05,
+      "loss": 1.7273,
+      "step": 709
+    },
+    {
+      "epoch": 0.6554350334641126,
+      "grad_norm": 6.8984375,
+      "learning_rate": 9.533022014676452e-05,
+      "loss": 1.7571,
+      "step": 710
+    },
+    {
+      "epoch": 0.6563581813985692,
+      "grad_norm": 7.12890625,
+      "learning_rate": 9.532354903268846e-05,
+      "loss": 1.2982,
+      "step": 711
+    },
+    {
+      "epoch": 0.6572813293330256,
+      "grad_norm": 6.0,
+      "learning_rate": 9.531687791861241e-05,
+      "loss": 1.636,
+      "step": 712
+    },
+    {
+      "epoch": 0.6582044772674821,
+      "grad_norm": 5.47265625,
+      "learning_rate": 9.531020680453637e-05,
+      "loss": 1.314,
+      "step": 713
+    },
+    {
+      "epoch": 0.6591276252019386,
+      "grad_norm": 3.720703125,
+      "learning_rate": 9.530353569046031e-05,
+      "loss": 1.2672,
+      "step": 714
+    },
+    {
+      "epoch": 0.6600507731363952,
+      "grad_norm": 4.48828125,
+      "learning_rate": 9.529686457638426e-05,
+      "loss": 1.6804,
+      "step": 715
+    },
+    {
+      "epoch": 0.6609739210708516,
+      "grad_norm": 4.41796875,
+      "learning_rate": 9.52901934623082e-05,
+      "loss": 1.8533,
+      "step": 716
+    },
+    {
+      "epoch": 0.6618970690053081,
+      "grad_norm": 5.10546875,
+      "learning_rate": 9.528352234823216e-05,
+      "loss": 1.6057,
+      "step": 717
+    },
+    {
+      "epoch": 0.6628202169397646,
+      "grad_norm": 5.55078125,
+      "learning_rate": 9.527685123415611e-05,
+      "loss": 1.6587,
+      "step": 718
+    },
+    {
+      "epoch": 0.6637433648742211,
+      "grad_norm": 7.9140625,
+      "learning_rate": 9.527018012008005e-05,
+      "loss": 1.8486,
+      "step": 719
+    },
+    {
+      "epoch": 0.6646665128086776,
+      "grad_norm": 7.05078125,
+      "learning_rate": 9.5263509006004e-05,
+      "loss": 1.8313,
+      "step": 720
+    },
+    {
+      "epoch": 0.6655896607431341,
+      "grad_norm": 7.2265625,
+      "learning_rate": 9.525683789192796e-05,
+      "loss": 1.6917,
+      "step": 721
+    },
+    {
+      "epoch": 0.6665128086775906,
+      "grad_norm": 6.77734375,
+      "learning_rate": 9.52501667778519e-05,
+      "loss": 2.0098,
+      "step": 722
+    },
+    {
+      "epoch": 0.6674359566120471,
+      "grad_norm": 8.859375,
+      "learning_rate": 9.524349566377585e-05,
+      "loss": 1.8965,
+      "step": 723
+    },
+    {
+      "epoch": 0.6683591045465036,
+      "grad_norm": 5.84375,
+      "learning_rate": 9.52368245496998e-05,
+      "loss": 1.4966,
+      "step": 724
+    },
+    {
+      "epoch": 0.6692822524809601,
+      "grad_norm": 6.7734375,
+      "learning_rate": 9.523015343562375e-05,
+      "loss": 1.6836,
+      "step": 725
+    },
+    {
+      "epoch": 0.6702054004154165,
+      "grad_norm": 4.51171875,
+      "learning_rate": 9.522348232154771e-05,
+      "loss": 1.3839,
+      "step": 726
+    },
+    {
+      "epoch": 0.6711285483498731,
+      "grad_norm": 5.18359375,
+      "learning_rate": 9.521681120747166e-05,
+      "loss": 1.5664,
+      "step": 727
+    },
+    {
+      "epoch": 0.6720516962843296,
+      "grad_norm": 6.14453125,
+      "learning_rate": 9.52101400933956e-05,
+      "loss": 1.6797,
+      "step": 728
+    },
+    {
+      "epoch": 0.6729748442187861,
+      "grad_norm": 7.41015625,
+      "learning_rate": 9.520346897931956e-05,
+      "loss": 1.9233,
+      "step": 729
+    },
+    {
+      "epoch": 0.6738979921532425,
+      "grad_norm": 7.1015625,
+      "learning_rate": 9.51967978652435e-05,
+      "loss": 1.5088,
+      "step": 730
+    },
+    {
+      "epoch": 0.674821140087699,
+      "grad_norm": 6.51953125,
+      "learning_rate": 9.519012675116745e-05,
+      "loss": 1.6035,
+      "step": 731
+    },
+    {
+      "epoch": 0.6757442880221556,
+      "grad_norm": 6.5078125,
+      "learning_rate": 9.51834556370914e-05,
+      "loss": 1.4622,
+      "step": 732
+    },
+    {
+      "epoch": 0.676667435956612,
+      "grad_norm": 6.23828125,
+      "learning_rate": 9.517678452301536e-05,
+      "loss": 1.5713,
+      "step": 733
+    },
+    {
+      "epoch": 0.6775905838910685,
+      "grad_norm": 6.296875,
+      "learning_rate": 9.51701134089393e-05,
+      "loss": 1.4585,
+      "step": 734
+    },
+    {
+      "epoch": 0.678513731825525,
+      "grad_norm": 4.62109375,
+      "learning_rate": 9.516344229486325e-05,
+      "loss": 1.5024,
+      "step": 735
+    },
+    {
+      "epoch": 0.6794368797599816,
+      "grad_norm": 7.296875,
+      "learning_rate": 9.515677118078719e-05,
+      "loss": 1.8953,
+      "step": 736
+    },
+    {
+      "epoch": 0.680360027694438,
+      "grad_norm": 5.8515625,
+      "learning_rate": 9.515010006671115e-05,
+      "loss": 1.7466,
+      "step": 737
+    },
+    {
+      "epoch": 0.6812831756288945,
+      "grad_norm": 5.5078125,
+      "learning_rate": 9.51434289526351e-05,
+      "loss": 1.6388,
+      "step": 738
+    },
+    {
+      "epoch": 0.682206323563351,
+      "grad_norm": 5.953125,
+      "learning_rate": 9.513675783855904e-05,
+      "loss": 1.6174,
+      "step": 739
+    },
+    {
+      "epoch": 0.6831294714978076,
+      "grad_norm": 5.89453125,
+      "learning_rate": 9.513008672448299e-05,
+      "loss": 1.9897,
+      "step": 740
+    },
+    {
+      "epoch": 0.684052619432264,
+      "grad_norm": 5.890625,
+      "learning_rate": 9.512341561040695e-05,
+      "loss": 1.7686,
+      "step": 741
+    },
+    {
+      "epoch": 0.6849757673667205,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.511674449633089e-05,
+      "loss": 1.1656,
+      "step": 742
+    },
+    {
+      "epoch": 0.685898915301177,
+      "grad_norm": 4.97265625,
+      "learning_rate": 9.511007338225484e-05,
+      "loss": 1.448,
+      "step": 743
+    },
+    {
+      "epoch": 0.6868220632356336,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.510340226817878e-05,
+      "loss": 1.5461,
+      "step": 744
+    },
+    {
+      "epoch": 0.68774521117009,
+      "grad_norm": 4.63671875,
+      "learning_rate": 9.509673115410274e-05,
+      "loss": 1.8838,
+      "step": 745
+    },
+    {
+      "epoch": 0.6886683591045465,
+      "grad_norm": 4.65234375,
+      "learning_rate": 9.509006004002669e-05,
+      "loss": 1.5725,
+      "step": 746
+    },
+    {
+      "epoch": 0.689591507039003,
+      "grad_norm": 5.70703125,
+      "learning_rate": 9.508338892595063e-05,
+      "loss": 1.6052,
+      "step": 747
+    },
+    {
+      "epoch": 0.6905146549734595,
+      "grad_norm": 6.05078125,
+      "learning_rate": 9.507671781187458e-05,
+      "loss": 1.4915,
+      "step": 748
+    },
+    {
+      "epoch": 0.691437802907916,
+      "grad_norm": 5.34765625,
+      "learning_rate": 9.507004669779854e-05,
+      "loss": 1.5988,
+      "step": 749
+    },
+    {
+      "epoch": 0.6923609508423725,
+      "grad_norm": 5.71484375,
+      "learning_rate": 9.50633755837225e-05,
+      "loss": 1.5317,
+      "step": 750
+    },
+    {
+      "epoch": 0.693284098776829,
+      "grad_norm": 5.60546875,
+      "learning_rate": 9.505670446964644e-05,
+      "loss": 1.4956,
+      "step": 751
+    },
+    {
+      "epoch": 0.6942072467112855,
+      "grad_norm": 6.7578125,
+      "learning_rate": 9.505003335557039e-05,
+      "loss": 1.7122,
+      "step": 752
+    },
+    {
+      "epoch": 0.695130394645742,
+      "grad_norm": 6.37890625,
+      "learning_rate": 9.504336224149433e-05,
+      "loss": 1.582,
+      "step": 753
+    },
+    {
+      "epoch": 0.6960535425801985,
+      "grad_norm": 5.6015625,
+      "learning_rate": 9.503669112741829e-05,
+      "loss": 1.6853,
+      "step": 754
+    },
+    {
+      "epoch": 0.6969766905146549,
+      "grad_norm": 5.78515625,
+      "learning_rate": 9.503002001334224e-05,
+      "loss": 1.5457,
+      "step": 755
+    },
+    {
+      "epoch": 0.6978998384491115,
+      "grad_norm": 7.87890625,
+      "learning_rate": 9.502334889926618e-05,
+      "loss": 1.551,
+      "step": 756
+    },
+    {
+      "epoch": 0.698822986383568,
+      "grad_norm": 5.95703125,
+      "learning_rate": 9.501667778519013e-05,
+      "loss": 1.6221,
+      "step": 757
+    },
+    {
+      "epoch": 0.6997461343180245,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.501000667111409e-05,
+      "loss": 1.6292,
+      "step": 758
+    },
+    {
+      "epoch": 0.7006692822524809,
+      "grad_norm": 7.359375,
+      "learning_rate": 9.500333555703803e-05,
+      "loss": 2.0242,
+      "step": 759
+    },
+    {
+      "epoch": 0.7015924301869375,
+      "grad_norm": 7.48828125,
+      "learning_rate": 9.499666444296198e-05,
+      "loss": 1.8083,
+      "step": 760
+    },
+    {
+      "epoch": 0.702515578121394,
+      "grad_norm": 6.2265625,
+      "learning_rate": 9.498999332888592e-05,
+      "loss": 1.5212,
+      "step": 761
+    },
+    {
+      "epoch": 0.7034387260558504,
+      "grad_norm": 6.5390625,
+      "learning_rate": 9.498332221480988e-05,
+      "loss": 1.4962,
+      "step": 762
+    },
+    {
+      "epoch": 0.7043618739903069,
+      "grad_norm": 6.1875,
+      "learning_rate": 9.497665110073383e-05,
+      "loss": 1.6187,
+      "step": 763
+    },
+    {
+      "epoch": 0.7052850219247634,
+      "grad_norm": 7.19140625,
+      "learning_rate": 9.496997998665777e-05,
+      "loss": 1.7974,
+      "step": 764
+    },
+    {
+      "epoch": 0.70620816985922,
+      "grad_norm": 6.171875,
+      "learning_rate": 9.496330887258172e-05,
+      "loss": 1.6577,
+      "step": 765
+    },
+    {
+      "epoch": 0.7071313177936764,
+      "grad_norm": 6.5546875,
+      "learning_rate": 9.495663775850568e-05,
+      "loss": 1.6064,
+      "step": 766
+    },
+    {
+      "epoch": 0.7080544657281329,
+      "grad_norm": 6.94921875,
+      "learning_rate": 9.494996664442962e-05,
+      "loss": 1.4341,
+      "step": 767
+    },
+    {
+      "epoch": 0.7089776136625894,
+      "grad_norm": 5.9375,
+      "learning_rate": 9.494329553035357e-05,
+      "loss": 1.5447,
+      "step": 768
+    },
+    {
+      "epoch": 0.709900761597046,
+      "grad_norm": 6.58203125,
+      "learning_rate": 9.493662441627751e-05,
+      "loss": 1.9722,
+      "step": 769
+    },
+    {
+      "epoch": 0.7108239095315024,
+      "grad_norm": 7.81640625,
+      "learning_rate": 9.492995330220147e-05,
+      "loss": 1.9924,
+      "step": 770
+    },
+    {
+      "epoch": 0.7117470574659589,
+      "grad_norm": 6.58203125,
+      "learning_rate": 9.492328218812542e-05,
+      "loss": 1.571,
+      "step": 771
+    },
+    {
+      "epoch": 0.7126702054004154,
+      "grad_norm": 6.09765625,
+      "learning_rate": 9.491661107404936e-05,
+      "loss": 1.7566,
+      "step": 772
+    },
+    {
+      "epoch": 0.713593353334872,
+      "grad_norm": 7.19140625,
+      "learning_rate": 9.490993995997331e-05,
+      "loss": 1.8276,
+      "step": 773
+    },
+    {
+      "epoch": 0.7145165012693284,
+      "grad_norm": 6.19921875,
+      "learning_rate": 9.490326884589727e-05,
+      "loss": 1.5541,
+      "step": 774
+    },
+    {
+      "epoch": 0.7154396492037849,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.489659773182123e-05,
+      "loss": 1.4622,
+      "step": 775
+    },
+    {
+      "epoch": 0.7163627971382414,
+      "grad_norm": 5.140625,
+      "learning_rate": 9.488992661774517e-05,
+      "loss": 1.8135,
+      "step": 776
+    },
+    {
+      "epoch": 0.7172859450726979,
+      "grad_norm": 6.26171875,
+      "learning_rate": 9.488325550366912e-05,
+      "loss": 1.7998,
+      "step": 777
+    },
+    {
+      "epoch": 0.7182090930071544,
+      "grad_norm": 6.32421875,
+      "learning_rate": 9.487658438959308e-05,
+      "loss": 1.875,
+      "step": 778
+    },
+    {
+      "epoch": 0.7191322409416109,
+      "grad_norm": 5.36328125,
+      "learning_rate": 9.486991327551702e-05,
+      "loss": 1.49,
+      "step": 779
+    },
+    {
+      "epoch": 0.7200553888760673,
+      "grad_norm": 6.62890625,
+      "learning_rate": 9.486324216144097e-05,
+      "loss": 1.9866,
+      "step": 780
+    },
+    {
+      "epoch": 0.7209785368105239,
+      "grad_norm": 5.4453125,
+      "learning_rate": 9.485657104736491e-05,
+      "loss": 1.6892,
+      "step": 781
+    },
+    {
+      "epoch": 0.7219016847449804,
+      "grad_norm": 5.45703125,
+      "learning_rate": 9.484989993328887e-05,
+      "loss": 1.5481,
+      "step": 782
+    },
+    {
+      "epoch": 0.7228248326794369,
+      "grad_norm": 4.76953125,
+      "learning_rate": 9.484322881921282e-05,
+      "loss": 1.6526,
+      "step": 783
+    },
+    {
+      "epoch": 0.7237479806138933,
+      "grad_norm": 7.2109375,
+      "learning_rate": 9.483655770513676e-05,
+      "loss": 1.8198,
+      "step": 784
+    },
+    {
+      "epoch": 0.7246711285483499,
+      "grad_norm": 5.65234375,
+      "learning_rate": 9.482988659106071e-05,
+      "loss": 1.6543,
+      "step": 785
+    },
+    {
+      "epoch": 0.7255942764828064,
+      "grad_norm": 6.8515625,
+      "learning_rate": 9.482321547698467e-05,
+      "loss": 1.7812,
+      "step": 786
+    },
+    {
+      "epoch": 0.7265174244172629,
+      "grad_norm": 5.9609375,
+      "learning_rate": 9.481654436290861e-05,
+      "loss": 1.7817,
+      "step": 787
+    },
+    {
+      "epoch": 0.7274405723517193,
+      "grad_norm": 5.484375,
+      "learning_rate": 9.480987324883256e-05,
+      "loss": 1.856,
+      "step": 788
+    },
+    {
+      "epoch": 0.7283637202861759,
+      "grad_norm": 5.7734375,
+      "learning_rate": 9.48032021347565e-05,
+      "loss": 1.7292,
+      "step": 789
+    },
+    {
+      "epoch": 0.7292868682206324,
+      "grad_norm": 7.1171875,
+      "learning_rate": 9.479653102068046e-05,
+      "loss": 1.8342,
+      "step": 790
+    },
+    {
+      "epoch": 0.7302100161550888,
+      "grad_norm": 5.21484375,
+      "learning_rate": 9.478985990660441e-05,
+      "loss": 1.8389,
+      "step": 791
+    },
+    {
+      "epoch": 0.7311331640895453,
+      "grad_norm": 5.6484375,
+      "learning_rate": 9.478318879252835e-05,
+      "loss": 1.96,
+      "step": 792
+    },
+    {
+      "epoch": 0.7320563120240019,
+      "grad_norm": 6.4609375,
+      "learning_rate": 9.47765176784523e-05,
+      "loss": 1.8774,
+      "step": 793
+    },
+    {
+      "epoch": 0.7329794599584584,
+      "grad_norm": 7.72265625,
+      "learning_rate": 9.476984656437626e-05,
+      "loss": 1.5662,
+      "step": 794
+    },
+    {
+      "epoch": 0.7339026078929148,
+      "grad_norm": 6.80859375,
+      "learning_rate": 9.47631754503002e-05,
+      "loss": 1.8149,
+      "step": 795
+    },
+    {
+      "epoch": 0.7348257558273713,
+      "grad_norm": 6.015625,
+      "learning_rate": 9.475650433622415e-05,
+      "loss": 1.7097,
+      "step": 796
+    },
+    {
+      "epoch": 0.7357489037618279,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.47498332221481e-05,
+      "loss": 1.6299,
+      "step": 797
+    },
+    {
+      "epoch": 0.7366720516962844,
+      "grad_norm": 6.4765625,
+      "learning_rate": 9.474316210807205e-05,
+      "loss": 1.4478,
+      "step": 798
+    },
+    {
+      "epoch": 0.7375951996307408,
+      "grad_norm": 7.1328125,
+      "learning_rate": 9.4736490993996e-05,
+      "loss": 1.6638,
+      "step": 799
+    },
+    {
+      "epoch": 0.7385183475651973,
+      "grad_norm": 5.8203125,
+      "learning_rate": 9.472981987991996e-05,
+      "loss": 1.8018,
+      "step": 800
+    },
+    {
+      "epoch": 0.7394414954996538,
+      "grad_norm": 7.1484375,
+      "learning_rate": 9.47231487658439e-05,
+      "loss": 1.6443,
+      "step": 801
+    },
+    {
+      "epoch": 0.7403646434341103,
+      "grad_norm": 4.81640625,
+      "learning_rate": 9.471647765176785e-05,
+      "loss": 1.5834,
+      "step": 802
+    },
+    {
+      "epoch": 0.7412877913685668,
+      "grad_norm": 4.703125,
+      "learning_rate": 9.470980653769181e-05,
+      "loss": 1.6411,
+      "step": 803
+    },
+    {
+      "epoch": 0.7422109393030233,
+      "grad_norm": 5.3359375,
+      "learning_rate": 9.470313542361575e-05,
+      "loss": 1.917,
+      "step": 804
+    },
+    {
+      "epoch": 0.7431340872374798,
+      "grad_norm": 5.60546875,
+      "learning_rate": 9.46964643095397e-05,
+      "loss": 1.7124,
+      "step": 805
+    },
+    {
+      "epoch": 0.7440572351719363,
+      "grad_norm": 4.7734375,
+      "learning_rate": 9.468979319546364e-05,
+      "loss": 1.8513,
+      "step": 806
+    },
+    {
+      "epoch": 0.7449803831063928,
+      "grad_norm": 4.3828125,
+      "learning_rate": 9.46831220813876e-05,
+      "loss": 1.4612,
+      "step": 807
+    },
+    {
+      "epoch": 0.7459035310408493,
+      "grad_norm": 5.5078125,
+      "learning_rate": 9.467645096731155e-05,
+      "loss": 1.6665,
+      "step": 808
+    },
+    {
+      "epoch": 0.7468266789753057,
+      "grad_norm": 5.56640625,
+      "learning_rate": 9.46697798532355e-05,
+      "loss": 1.8179,
+      "step": 809
+    },
+    {
+      "epoch": 0.7477498269097623,
+      "grad_norm": 5.20703125,
+      "learning_rate": 9.466310873915944e-05,
+      "loss": 1.5579,
+      "step": 810
+    },
+    {
+      "epoch": 0.7486729748442188,
+      "grad_norm": 5.95703125,
+      "learning_rate": 9.46564376250834e-05,
+      "loss": 1.876,
+      "step": 811
+    },
+    {
+      "epoch": 0.7495961227786753,
+      "grad_norm": 6.88671875,
+      "learning_rate": 9.464976651100734e-05,
+      "loss": 1.9658,
+      "step": 812
+    },
+    {
+      "epoch": 0.7505192707131317,
+      "grad_norm": 9.8828125,
+      "learning_rate": 9.464309539693129e-05,
+      "loss": 1.3588,
+      "step": 813
+    },
+    {
+      "epoch": 0.7514424186475883,
+      "grad_norm": 6.28125,
+      "learning_rate": 9.463642428285523e-05,
+      "loss": 1.5037,
+      "step": 814
+    },
+    {
+      "epoch": 0.7523655665820448,
+      "grad_norm": 5.03515625,
+      "learning_rate": 9.46297531687792e-05,
+      "loss": 1.8259,
+      "step": 815
+    },
+    {
+      "epoch": 0.7532887145165013,
+      "grad_norm": 5.2421875,
+      "learning_rate": 9.462308205470314e-05,
+      "loss": 1.4899,
+      "step": 816
+    },
+    {
+      "epoch": 0.7542118624509577,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.461641094062708e-05,
+      "loss": 1.7937,
+      "step": 817
+    },
+    {
+      "epoch": 0.7551350103854143,
+      "grad_norm": 6.48046875,
+      "learning_rate": 9.460973982655103e-05,
+      "loss": 1.7957,
+      "step": 818
+    },
+    {
+      "epoch": 0.7560581583198708,
+      "grad_norm": 6.3984375,
+      "learning_rate": 9.460306871247499e-05,
+      "loss": 1.9944,
+      "step": 819
+    },
+    {
+      "epoch": 0.7569813062543272,
+      "grad_norm": 7.9921875,
+      "learning_rate": 9.459639759839893e-05,
+      "loss": 1.7229,
+      "step": 820
+    },
+    {
+      "epoch": 0.7579044541887837,
+      "grad_norm": 5.80078125,
+      "learning_rate": 9.458972648432288e-05,
+      "loss": 1.5466,
+      "step": 821
+    },
+    {
+      "epoch": 0.7588276021232403,
+      "grad_norm": 6.35546875,
+      "learning_rate": 9.458305537024683e-05,
+      "loss": 1.5505,
+      "step": 822
+    },
+    {
+      "epoch": 0.7597507500576968,
+      "grad_norm": 5.578125,
+      "learning_rate": 9.457638425617078e-05,
+      "loss": 1.3555,
+      "step": 823
+    },
+    {
+      "epoch": 0.7606738979921532,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.456971314209473e-05,
+      "loss": 1.7251,
+      "step": 824
+    },
+    {
+      "epoch": 0.7615970459266097,
+      "grad_norm": 5.40234375,
+      "learning_rate": 9.456304202801869e-05,
+      "loss": 1.8687,
+      "step": 825
+    },
+    {
+      "epoch": 0.7625201938610663,
+      "grad_norm": 6.68359375,
+      "learning_rate": 9.455637091394263e-05,
+      "loss": 1.6614,
+      "step": 826
+    },
+    {
+      "epoch": 0.7634433417955228,
+      "grad_norm": 6.5078125,
+      "learning_rate": 9.454969979986659e-05,
+      "loss": 1.7637,
+      "step": 827
+    },
+    {
+      "epoch": 0.7643664897299792,
+      "grad_norm": 6.24609375,
+      "learning_rate": 9.454302868579054e-05,
+      "loss": 1.495,
+      "step": 828
+    },
+    {
+      "epoch": 0.7652896376644357,
+      "grad_norm": 6.4921875,
+      "learning_rate": 9.453635757171448e-05,
+      "loss": 1.6189,
+      "step": 829
+    },
+    {
+      "epoch": 0.7662127855988923,
+      "grad_norm": 6.90234375,
+      "learning_rate": 9.452968645763843e-05,
+      "loss": 1.9504,
+      "step": 830
+    },
+    {
+      "epoch": 0.7671359335333487,
+      "grad_norm": 7.0703125,
+      "learning_rate": 9.452301534356239e-05,
+      "loss": 1.7397,
+      "step": 831
+    },
+    {
+      "epoch": 0.7680590814678052,
+      "grad_norm": 8.796875,
+      "learning_rate": 9.451634422948633e-05,
+      "loss": 1.5737,
+      "step": 832
+    },
+    {
+      "epoch": 0.7689822294022617,
+      "grad_norm": 5.60546875,
+      "learning_rate": 9.450967311541028e-05,
+      "loss": 1.8564,
+      "step": 833
+    },
+    {
+      "epoch": 0.7699053773367182,
+      "grad_norm": 5.890625,
+      "learning_rate": 9.450300200133422e-05,
+      "loss": 1.8362,
+      "step": 834
+    },
+    {
+      "epoch": 0.7708285252711747,
+      "grad_norm": 6.64453125,
+      "learning_rate": 9.449633088725818e-05,
+      "loss": 1.8613,
+      "step": 835
+    },
+    {
+      "epoch": 0.7717516732056312,
+      "grad_norm": 6.7421875,
+      "learning_rate": 9.448965977318213e-05,
+      "loss": 1.9258,
+      "step": 836
+    },
+    {
+      "epoch": 0.7726748211400877,
+      "grad_norm": 6.32421875,
+      "learning_rate": 9.448298865910607e-05,
+      "loss": 1.6729,
+      "step": 837
+    },
+    {
+      "epoch": 0.7735979690745441,
+      "grad_norm": 5.2578125,
+      "learning_rate": 9.447631754503002e-05,
+      "loss": 1.738,
+      "step": 838
+    },
+    {
+      "epoch": 0.7745211170090007,
+      "grad_norm": 5.38671875,
+      "learning_rate": 9.446964643095398e-05,
+      "loss": 1.8274,
+      "step": 839
+    },
+    {
+      "epoch": 0.7754442649434572,
+      "grad_norm": 5.3359375,
+      "learning_rate": 9.446297531687792e-05,
+      "loss": 1.7192,
+      "step": 840
+    },
+    {
+      "epoch": 0.7763674128779137,
+      "grad_norm": 7.28515625,
+      "learning_rate": 9.445630420280187e-05,
+      "loss": 1.7424,
+      "step": 841
+    },
+    {
+      "epoch": 0.7772905608123701,
+      "grad_norm": 7.23828125,
+      "learning_rate": 9.444963308872581e-05,
+      "loss": 0.9667,
+      "step": 842
+    },
+    {
+      "epoch": 0.7782137087468267,
+      "grad_norm": 6.12109375,
+      "learning_rate": 9.444296197464977e-05,
+      "loss": 1.6619,
+      "step": 843
+    },
+    {
+      "epoch": 0.7791368566812832,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.443629086057372e-05,
+      "loss": 1.6514,
+      "step": 844
+    },
+    {
+      "epoch": 0.7800600046157397,
+      "grad_norm": 5.39453125,
+      "learning_rate": 9.442961974649766e-05,
+      "loss": 1.6792,
+      "step": 845
+    },
+    {
+      "epoch": 0.7809831525501961,
+      "grad_norm": 5.69140625,
+      "learning_rate": 9.442294863242161e-05,
+      "loss": 1.7346,
+      "step": 846
+    },
+    {
+      "epoch": 0.7819063004846527,
+      "grad_norm": 7.56640625,
+      "learning_rate": 9.441627751834557e-05,
+      "loss": 1.593,
+      "step": 847
+    },
+    {
+      "epoch": 0.7828294484191092,
+      "grad_norm": 5.53515625,
+      "learning_rate": 9.440960640426951e-05,
+      "loss": 1.3112,
+      "step": 848
+    },
+    {
+      "epoch": 0.7837525963535656,
+      "grad_norm": 5.80078125,
+      "learning_rate": 9.440293529019346e-05,
+      "loss": 1.9114,
+      "step": 849
+    },
+    {
+      "epoch": 0.7846757442880221,
+      "grad_norm": 4.97265625,
+      "learning_rate": 9.439626417611742e-05,
+      "loss": 1.7561,
+      "step": 850
+    },
+    {
+      "epoch": 0.7855988922224787,
+      "grad_norm": 5.42578125,
+      "learning_rate": 9.438959306204136e-05,
+      "loss": 1.3966,
+      "step": 851
+    },
+    {
+      "epoch": 0.7865220401569352,
+      "grad_norm": 7.61328125,
+      "learning_rate": 9.438292194796532e-05,
+      "loss": 1.7109,
+      "step": 852
+    },
+    {
+      "epoch": 0.7874451880913916,
+      "grad_norm": 5.38671875,
+      "learning_rate": 9.437625083388927e-05,
+      "loss": 1.5596,
+      "step": 853
+    },
+    {
+      "epoch": 0.7883683360258481,
+      "grad_norm": 6.390625,
+      "learning_rate": 9.436957971981321e-05,
+      "loss": 1.5383,
+      "step": 854
+    },
+    {
+      "epoch": 0.7892914839603047,
+      "grad_norm": 6.87890625,
+      "learning_rate": 9.436290860573716e-05,
+      "loss": 1.5833,
+      "step": 855
+    },
+    {
+      "epoch": 0.7902146318947612,
+      "grad_norm": 7.125,
+      "learning_rate": 9.435623749166112e-05,
+      "loss": 1.7986,
+      "step": 856
+    },
+    {
+      "epoch": 0.7911377798292176,
+      "grad_norm": 6.8125,
+      "learning_rate": 9.434956637758506e-05,
+      "loss": 1.5225,
+      "step": 857
+    },
+    {
+      "epoch": 0.7920609277636741,
+      "grad_norm": 6.1484375,
+      "learning_rate": 9.434289526350901e-05,
+      "loss": 1.4692,
+      "step": 858
+    },
+    {
+      "epoch": 0.7929840756981307,
+      "grad_norm": 5.109375,
+      "learning_rate": 9.433622414943296e-05,
+      "loss": 1.426,
+      "step": 859
+    },
+    {
+      "epoch": 0.7939072236325871,
+      "grad_norm": 4.6328125,
+      "learning_rate": 9.432955303535691e-05,
+      "loss": 1.7302,
+      "step": 860
+    },
+    {
+      "epoch": 0.7948303715670436,
+      "grad_norm": 5.08203125,
+      "learning_rate": 9.432288192128086e-05,
+      "loss": 1.6643,
+      "step": 861
+    },
+    {
+      "epoch": 0.7957535195015001,
+      "grad_norm": 6.43359375,
+      "learning_rate": 9.43162108072048e-05,
+      "loss": 1.8638,
+      "step": 862
+    },
+    {
+      "epoch": 0.7966766674359567,
+      "grad_norm": 7.09765625,
+      "learning_rate": 9.430953969312875e-05,
+      "loss": 1.8235,
+      "step": 863
+    },
+    {
+      "epoch": 0.7975998153704131,
+      "grad_norm": 7.140625,
+      "learning_rate": 9.430286857905271e-05,
+      "loss": 1.9155,
+      "step": 864
+    },
+    {
+      "epoch": 0.7985229633048696,
+      "grad_norm": 7.4453125,
+      "learning_rate": 9.429619746497665e-05,
+      "loss": 1.5885,
+      "step": 865
+    },
+    {
+      "epoch": 0.7994461112393261,
+      "grad_norm": 5.171875,
+      "learning_rate": 9.42895263509006e-05,
+      "loss": 1.5144,
+      "step": 866
+    },
+    {
+      "epoch": 0.8003692591737825,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.428285523682455e-05,
+      "loss": 1.6016,
+      "step": 867
+    },
+    {
+      "epoch": 0.8012924071082391,
+      "grad_norm": 4.69921875,
+      "learning_rate": 9.42761841227485e-05,
+      "loss": 1.3997,
+      "step": 868
+    },
+    {
+      "epoch": 0.8022155550426956,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.426951300867245e-05,
+      "loss": 1.6943,
+      "step": 869
+    },
+    {
+      "epoch": 0.8031387029771521,
+      "grad_norm": 5.65625,
+      "learning_rate": 9.42628418945964e-05,
+      "loss": 1.7451,
+      "step": 870
+    },
+    {
+      "epoch": 0.8040618509116085,
+      "grad_norm": 5.125,
+      "learning_rate": 9.425617078052034e-05,
+      "loss": 1.772,
+      "step": 871
+    },
+    {
+      "epoch": 0.8049849988460651,
+      "grad_norm": 4.93359375,
+      "learning_rate": 9.42494996664443e-05,
+      "loss": 1.9177,
+      "step": 872
+    },
+    {
+      "epoch": 0.8059081467805216,
+      "grad_norm": 5.0,
+      "learning_rate": 9.424282855236825e-05,
+      "loss": 1.7983,
+      "step": 873
+    },
+    {
+      "epoch": 0.806831294714978,
+      "grad_norm": 5.70703125,
+      "learning_rate": 9.423615743829219e-05,
+      "loss": 1.5747,
+      "step": 874
+    },
+    {
+      "epoch": 0.8077544426494345,
+      "grad_norm": 5.09375,
+      "learning_rate": 9.422948632421615e-05,
+      "loss": 1.448,
+      "step": 875
+    },
+    {
+      "epoch": 0.8086775905838911,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.422281521014011e-05,
+      "loss": 1.4792,
+      "step": 876
+    },
+    {
+      "epoch": 0.8096007385183476,
+      "grad_norm": 5.44921875,
+      "learning_rate": 9.421614409606405e-05,
+      "loss": 1.6697,
+      "step": 877
+    },
+    {
+      "epoch": 0.810523886452804,
+      "grad_norm": 6.1328125,
+      "learning_rate": 9.4209472981988e-05,
+      "loss": 1.7341,
+      "step": 878
+    },
+    {
+      "epoch": 0.8114470343872605,
+      "grad_norm": 5.91015625,
+      "learning_rate": 9.420280186791194e-05,
+      "loss": 1.6316,
+      "step": 879
+    },
+    {
+      "epoch": 0.8123701823217171,
+      "grad_norm": 7.515625,
+      "learning_rate": 9.41961307538359e-05,
+      "loss": 1.7244,
+      "step": 880
+    },
+    {
+      "epoch": 0.8132933302561736,
+      "grad_norm": 4.92578125,
+      "learning_rate": 9.418945963975985e-05,
+      "loss": 1.8018,
+      "step": 881
+    },
+    {
+      "epoch": 0.81421647819063,
+      "grad_norm": 5.84375,
+      "learning_rate": 9.41827885256838e-05,
+      "loss": 1.6414,
+      "step": 882
+    },
+    {
+      "epoch": 0.8151396261250865,
+      "grad_norm": 7.70703125,
+      "learning_rate": 9.417611741160774e-05,
+      "loss": 1.7476,
+      "step": 883
+    },
+    {
+      "epoch": 0.8160627740595431,
+      "grad_norm": 6.8359375,
+      "learning_rate": 9.41694462975317e-05,
+      "loss": 1.5718,
+      "step": 884
+    },
+    {
+      "epoch": 0.8169859219939996,
+      "grad_norm": 5.07421875,
+      "learning_rate": 9.416277518345564e-05,
+      "loss": 1.8123,
+      "step": 885
+    },
+    {
+      "epoch": 0.817909069928456,
+      "grad_norm": 6.96875,
+      "learning_rate": 9.415610406937959e-05,
+      "loss": 1.6936,
+      "step": 886
+    },
+    {
+      "epoch": 0.8188322178629125,
+      "grad_norm": 5.51953125,
+      "learning_rate": 9.414943295530354e-05,
+      "loss": 1.3618,
+      "step": 887
+    },
+    {
+      "epoch": 0.8197553657973691,
+      "grad_norm": 6.96875,
+      "learning_rate": 9.41427618412275e-05,
+      "loss": 1.6526,
+      "step": 888
+    },
+    {
+      "epoch": 0.8206785137318255,
+      "grad_norm": 7.453125,
+      "learning_rate": 9.413609072715144e-05,
+      "loss": 1.9917,
+      "step": 889
+    },
+    {
+      "epoch": 0.821601661666282,
+      "grad_norm": 7.7265625,
+      "learning_rate": 9.412941961307539e-05,
+      "loss": 1.8962,
+      "step": 890
+    },
+    {
+      "epoch": 0.8225248096007385,
+      "grad_norm": 5.5703125,
+      "learning_rate": 9.412274849899933e-05,
+      "loss": 1.6902,
+      "step": 891
+    },
+    {
+      "epoch": 0.8234479575351951,
+      "grad_norm": 6.3359375,
+      "learning_rate": 9.411607738492329e-05,
+      "loss": 1.6416,
+      "step": 892
+    },
+    {
+      "epoch": 0.8243711054696515,
+      "grad_norm": 6.2421875,
+      "learning_rate": 9.410940627084724e-05,
+      "loss": 1.6912,
+      "step": 893
+    },
+    {
+      "epoch": 0.825294253404108,
+      "grad_norm": 6.03125,
+      "learning_rate": 9.410273515677118e-05,
+      "loss": 1.5999,
+      "step": 894
+    },
+    {
+      "epoch": 0.8262174013385645,
+      "grad_norm": 6.42578125,
+      "learning_rate": 9.409606404269513e-05,
+      "loss": 1.7998,
+      "step": 895
+    },
+    {
+      "epoch": 0.827140549273021,
+      "grad_norm": 6.06640625,
+      "learning_rate": 9.408939292861909e-05,
+      "loss": 1.5999,
+      "step": 896
+    },
+    {
+      "epoch": 0.8280636972074775,
+      "grad_norm": 5.0390625,
+      "learning_rate": 9.408272181454303e-05,
+      "loss": 1.6477,
+      "step": 897
+    },
+    {
+      "epoch": 0.828986845141934,
+      "grad_norm": 8.3671875,
+      "learning_rate": 9.407605070046698e-05,
+      "loss": 1.8042,
+      "step": 898
+    },
+    {
+      "epoch": 0.8299099930763905,
+      "grad_norm": 6.5,
+      "learning_rate": 9.406937958639092e-05,
+      "loss": 1.5618,
+      "step": 899
+    },
+    {
+      "epoch": 0.830833141010847,
+      "grad_norm": 6.92578125,
+      "learning_rate": 9.406270847231488e-05,
+      "loss": 1.5723,
+      "step": 900
+    },
+    {
+      "epoch": 0.8317562889453035,
+      "grad_norm": 6.1953125,
+      "learning_rate": 9.405603735823884e-05,
+      "loss": 1.6638,
+      "step": 901
+    },
+    {
+      "epoch": 0.83267943687976,
+      "grad_norm": 8.6015625,
+      "learning_rate": 9.404936624416278e-05,
+      "loss": 1.5999,
+      "step": 902
+    },
+    {
+      "epoch": 0.8336025848142165,
+      "grad_norm": 5.91015625,
+      "learning_rate": 9.404269513008673e-05,
+      "loss": 1.4272,
+      "step": 903
+    },
+    {
+      "epoch": 0.8345257327486729,
+      "grad_norm": 7.18359375,
+      "learning_rate": 9.403602401601068e-05,
+      "loss": 1.7874,
+      "step": 904
+    },
+    {
+      "epoch": 0.8354488806831295,
+      "grad_norm": 5.56640625,
+      "learning_rate": 9.402935290193463e-05,
+      "loss": 1.437,
+      "step": 905
+    },
+    {
+      "epoch": 0.836372028617586,
+      "grad_norm": 6.265625,
+      "learning_rate": 9.402268178785858e-05,
+      "loss": 1.3362,
+      "step": 906
+    },
+    {
+      "epoch": 0.8372951765520424,
+      "grad_norm": 5.5078125,
+      "learning_rate": 9.401601067378253e-05,
+      "loss": 1.2535,
+      "step": 907
+    },
+    {
+      "epoch": 0.8382183244864989,
+      "grad_norm": 7.671875,
+      "learning_rate": 9.400933955970647e-05,
+      "loss": 1.8054,
+      "step": 908
+    },
+    {
+      "epoch": 0.8391414724209555,
+      "grad_norm": 5.25390625,
+      "learning_rate": 9.400266844563043e-05,
+      "loss": 1.7761,
+      "step": 909
+    },
+    {
+      "epoch": 0.840064620355412,
+      "grad_norm": 4.859375,
+      "learning_rate": 9.399599733155438e-05,
+      "loss": 1.4781,
+      "step": 910
+    },
+    {
+      "epoch": 0.8409877682898684,
+      "grad_norm": 5.24609375,
+      "learning_rate": 9.398932621747832e-05,
+      "loss": 1.7983,
+      "step": 911
+    },
+    {
+      "epoch": 0.8419109162243249,
+      "grad_norm": 6.40234375,
+      "learning_rate": 9.398265510340227e-05,
+      "loss": 1.6172,
+      "step": 912
+    },
+    {
+      "epoch": 0.8428340641587815,
+      "grad_norm": 5.453125,
+      "learning_rate": 9.397598398932623e-05,
+      "loss": 1.5691,
+      "step": 913
+    },
+    {
+      "epoch": 0.843757212093238,
+      "grad_norm": 5.1328125,
+      "learning_rate": 9.396931287525017e-05,
+      "loss": 1.1881,
+      "step": 914
+    },
+    {
+      "epoch": 0.8446803600276944,
+      "grad_norm": 5.30859375,
+      "learning_rate": 9.396264176117412e-05,
+      "loss": 1.5906,
+      "step": 915
+    },
+    {
+      "epoch": 0.8456035079621509,
+      "grad_norm": 6.39453125,
+      "learning_rate": 9.395597064709807e-05,
+      "loss": 1.731,
+      "step": 916
+    },
+    {
+      "epoch": 0.8465266558966075,
+      "grad_norm": 7.0546875,
+      "learning_rate": 9.394929953302202e-05,
+      "loss": 1.8513,
+      "step": 917
+    },
+    {
+      "epoch": 0.8474498038310639,
+      "grad_norm": 5.34765625,
+      "learning_rate": 9.394262841894597e-05,
+      "loss": 1.5609,
+      "step": 918
+    },
+    {
+      "epoch": 0.8483729517655204,
+      "grad_norm": 6.08203125,
+      "learning_rate": 9.393595730486991e-05,
+      "loss": 1.6909,
+      "step": 919
+    },
+    {
+      "epoch": 0.8492960996999769,
+      "grad_norm": 5.27734375,
+      "learning_rate": 9.392928619079387e-05,
+      "loss": 1.7,
+      "step": 920
+    },
+    {
+      "epoch": 0.8502192476344335,
+      "grad_norm": 5.7109375,
+      "learning_rate": 9.392261507671782e-05,
+      "loss": 1.4337,
+      "step": 921
+    },
+    {
+      "epoch": 0.8511423955688899,
+      "grad_norm": 5.29296875,
+      "learning_rate": 9.391594396264176e-05,
+      "loss": 1.739,
+      "step": 922
+    },
+    {
+      "epoch": 0.8520655435033464,
+      "grad_norm": 7.55859375,
+      "learning_rate": 9.39092728485657e-05,
+      "loss": 1.5928,
+      "step": 923
+    },
+    {
+      "epoch": 0.8529886914378029,
+      "grad_norm": 5.37890625,
+      "learning_rate": 9.390260173448967e-05,
+      "loss": 1.7253,
+      "step": 924
+    },
+    {
+      "epoch": 0.8539118393722595,
+      "grad_norm": 6.97265625,
+      "learning_rate": 9.389593062041362e-05,
+      "loss": 1.6184,
+      "step": 925
+    },
+    {
+      "epoch": 0.8548349873067159,
+      "grad_norm": 5.61328125,
+      "learning_rate": 9.388925950633757e-05,
+      "loss": 1.7786,
+      "step": 926
+    },
+    {
+      "epoch": 0.8557581352411724,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.388258839226152e-05,
+      "loss": 1.5327,
+      "step": 927
+    },
+    {
+      "epoch": 0.8566812831756289,
+      "grad_norm": 5.8671875,
+      "learning_rate": 9.387591727818546e-05,
+      "loss": 1.9333,
+      "step": 928
+    },
+    {
+      "epoch": 0.8576044311100854,
+      "grad_norm": 6.46484375,
+      "learning_rate": 9.386924616410942e-05,
+      "loss": 1.7732,
+      "step": 929
+    },
+    {
+      "epoch": 0.8585275790445419,
+      "grad_norm": 5.4609375,
+      "learning_rate": 9.386257505003337e-05,
+      "loss": 1.613,
+      "step": 930
+    },
+    {
+      "epoch": 0.8594507269789984,
+      "grad_norm": 4.390625,
+      "learning_rate": 9.385590393595731e-05,
+      "loss": 1.6768,
+      "step": 931
+    },
+    {
+      "epoch": 0.8603738749134549,
+      "grad_norm": 5.3359375,
+      "learning_rate": 9.384923282188126e-05,
+      "loss": 1.0653,
+      "step": 932
+    },
+    {
+      "epoch": 0.8612970228479114,
+      "grad_norm": 6.42578125,
+      "learning_rate": 9.384256170780522e-05,
+      "loss": 1.8276,
+      "step": 933
+    },
+    {
+      "epoch": 0.8622201707823679,
+      "grad_norm": 7.55078125,
+      "learning_rate": 9.383589059372916e-05,
+      "loss": 1.5999,
+      "step": 934
+    },
+    {
+      "epoch": 0.8631433187168244,
+      "grad_norm": 6.43359375,
+      "learning_rate": 9.38292194796531e-05,
+      "loss": 1.626,
+      "step": 935
+    },
+    {
+      "epoch": 0.8640664666512808,
+      "grad_norm": 5.08203125,
+      "learning_rate": 9.382254836557705e-05,
+      "loss": 1.7175,
+      "step": 936
+    },
+    {
+      "epoch": 0.8649896145857373,
+      "grad_norm": 6.93359375,
+      "learning_rate": 9.381587725150101e-05,
+      "loss": 1.6082,
+      "step": 937
+    },
+    {
+      "epoch": 0.8659127625201939,
+      "grad_norm": 5.4921875,
+      "learning_rate": 9.380920613742496e-05,
+      "loss": 1.3915,
+      "step": 938
+    },
+    {
+      "epoch": 0.8668359104546504,
+      "grad_norm": 5.94140625,
+      "learning_rate": 9.38025350233489e-05,
+      "loss": 1.9192,
+      "step": 939
+    },
+    {
+      "epoch": 0.8677590583891068,
+      "grad_norm": 4.4296875,
+      "learning_rate": 9.379586390927285e-05,
+      "loss": 1.9048,
+      "step": 940
+    },
+    {
+      "epoch": 0.8686822063235633,
+      "grad_norm": 6.078125,
+      "learning_rate": 9.37891927951968e-05,
+      "loss": 1.5482,
+      "step": 941
+    },
+    {
+      "epoch": 0.8696053542580199,
+      "grad_norm": 4.30859375,
+      "learning_rate": 9.378252168112075e-05,
+      "loss": 1.2573,
+      "step": 942
+    },
+    {
+      "epoch": 0.8705285021924763,
+      "grad_norm": 5.79296875,
+      "learning_rate": 9.37758505670447e-05,
+      "loss": 1.623,
+      "step": 943
+    },
+    {
+      "epoch": 0.8714516501269328,
+      "grad_norm": 6.828125,
+      "learning_rate": 9.376917945296864e-05,
+      "loss": 1.8745,
+      "step": 944
+    },
+    {
+      "epoch": 0.8723747980613893,
+      "grad_norm": 7.375,
+      "learning_rate": 9.37625083388926e-05,
+      "loss": 1.6946,
+      "step": 945
+    },
+    {
+      "epoch": 0.8732979459958459,
+      "grad_norm": 7.87109375,
+      "learning_rate": 9.375583722481655e-05,
+      "loss": 1.7871,
+      "step": 946
+    },
+    {
+      "epoch": 0.8742210939303023,
+      "grad_norm": 5.7421875,
+      "learning_rate": 9.374916611074049e-05,
+      "loss": 1.5681,
+      "step": 947
+    },
+    {
+      "epoch": 0.8751442418647588,
+      "grad_norm": 6.4453125,
+      "learning_rate": 9.374249499666444e-05,
+      "loss": 1.5042,
+      "step": 948
+    },
+    {
+      "epoch": 0.8760673897992153,
+      "grad_norm": 6.16015625,
+      "learning_rate": 9.37358238825884e-05,
+      "loss": 1.7476,
+      "step": 949
+    },
+    {
+      "epoch": 0.8769905377336719,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.372915276851234e-05,
+      "loss": 1.8833,
+      "step": 950
+    },
+    {
+      "epoch": 0.8779136856681283,
+      "grad_norm": 6.6328125,
+      "learning_rate": 9.37224816544363e-05,
+      "loss": 1.6216,
+      "step": 951
+    },
+    {
+      "epoch": 0.8788368336025848,
+      "grad_norm": 6.45703125,
+      "learning_rate": 9.371581054036025e-05,
+      "loss": 1.9033,
+      "step": 952
+    },
+    {
+      "epoch": 0.8797599815370413,
+      "grad_norm": 5.0234375,
+      "learning_rate": 9.370913942628419e-05,
+      "loss": 1.6699,
+      "step": 953
+    },
+    {
+      "epoch": 0.8806831294714978,
+      "grad_norm": 8.3203125,
+      "learning_rate": 9.370246831220815e-05,
+      "loss": 1.8667,
+      "step": 954
+    },
+    {
+      "epoch": 0.8816062774059543,
+      "grad_norm": 6.29296875,
+      "learning_rate": 9.36957971981321e-05,
+      "loss": 1.7085,
+      "step": 955
+    },
+    {
+      "epoch": 0.8825294253404108,
+      "grad_norm": 6.875,
+      "learning_rate": 9.368912608405604e-05,
+      "loss": 1.4915,
+      "step": 956
+    },
+    {
+      "epoch": 0.8834525732748673,
+      "grad_norm": 5.9765625,
+      "learning_rate": 9.368245496998e-05,
+      "loss": 1.7275,
+      "step": 957
+    },
+    {
+      "epoch": 0.8843757212093238,
+      "grad_norm": 5.45703125,
+      "learning_rate": 9.367578385590395e-05,
+      "loss": 1.4167,
+      "step": 958
+    },
+    {
+      "epoch": 0.8852988691437803,
+      "grad_norm": 4.609375,
+      "learning_rate": 9.366911274182789e-05,
+      "loss": 1.4982,
+      "step": 959
+    },
+    {
+      "epoch": 0.8862220170782368,
+      "grad_norm": 6.32421875,
+      "learning_rate": 9.366244162775184e-05,
+      "loss": 1.7322,
+      "step": 960
+    },
+    {
+      "epoch": 0.8871451650126932,
+      "grad_norm": 5.734375,
+      "learning_rate": 9.36557705136758e-05,
+      "loss": 1.7642,
+      "step": 961
+    },
+    {
+      "epoch": 0.8880683129471498,
+      "grad_norm": 7.21875,
+      "learning_rate": 9.364909939959974e-05,
+      "loss": 1.7068,
+      "step": 962
+    },
+    {
+      "epoch": 0.8889914608816063,
+      "grad_norm": 6.2421875,
+      "learning_rate": 9.364242828552369e-05,
+      "loss": 1.8381,
+      "step": 963
+    },
+    {
+      "epoch": 0.8899146088160628,
+      "grad_norm": 6.08984375,
+      "learning_rate": 9.363575717144763e-05,
+      "loss": 1.7646,
+      "step": 964
+    },
+    {
+      "epoch": 0.8908377567505192,
+      "grad_norm": 5.796875,
+      "learning_rate": 9.362908605737159e-05,
+      "loss": 1.7698,
+      "step": 965
+    },
+    {
+      "epoch": 0.8917609046849758,
+      "grad_norm": 6.11328125,
+      "learning_rate": 9.362241494329554e-05,
+      "loss": 1.7478,
+      "step": 966
+    },
+    {
+      "epoch": 0.8926840526194323,
+      "grad_norm": 5.8828125,
+      "learning_rate": 9.361574382921948e-05,
+      "loss": 1.8286,
+      "step": 967
+    },
+    {
+      "epoch": 0.8936072005538888,
+      "grad_norm": 7.08203125,
+      "learning_rate": 9.360907271514343e-05,
+      "loss": 1.6328,
+      "step": 968
+    },
+    {
+      "epoch": 0.8945303484883452,
+      "grad_norm": 5.453125,
+      "learning_rate": 9.360240160106739e-05,
+      "loss": 1.5769,
+      "step": 969
+    },
+    {
+      "epoch": 0.8954534964228018,
+      "grad_norm": 5.5078125,
+      "learning_rate": 9.359573048699133e-05,
+      "loss": 1.8782,
+      "step": 970
+    },
+    {
+      "epoch": 0.8963766443572583,
+      "grad_norm": 4.64453125,
+      "learning_rate": 9.358905937291528e-05,
+      "loss": 1.7219,
+      "step": 971
+    },
+    {
+      "epoch": 0.8972997922917147,
+      "grad_norm": 5.14453125,
+      "learning_rate": 9.358238825883922e-05,
+      "loss": 1.3992,
+      "step": 972
+    },
+    {
+      "epoch": 0.8982229402261712,
+      "grad_norm": 5.6640625,
+      "learning_rate": 9.357571714476318e-05,
+      "loss": 1.6097,
+      "step": 973
+    },
+    {
+      "epoch": 0.8991460881606277,
+      "grad_norm": 4.8671875,
+      "learning_rate": 9.356904603068713e-05,
+      "loss": 1.8716,
+      "step": 974
+    },
+    {
+      "epoch": 0.9000692360950843,
+      "grad_norm": 6.25,
+      "learning_rate": 9.356237491661107e-05,
+      "loss": 1.9075,
+      "step": 975
+    },
+    {
+      "epoch": 0.9009923840295407,
+      "grad_norm": 6.24609375,
+      "learning_rate": 9.355570380253503e-05,
+      "loss": 1.6992,
+      "step": 976
+    },
+    {
+      "epoch": 0.9019155319639972,
+      "grad_norm": 5.5859375,
+      "learning_rate": 9.354903268845898e-05,
+      "loss": 1.656,
+      "step": 977
+    },
+    {
+      "epoch": 0.9028386798984537,
+      "grad_norm": 5.55078125,
+      "learning_rate": 9.354236157438294e-05,
+      "loss": 1.7817,
+      "step": 978
+    },
+    {
+      "epoch": 0.9037618278329103,
+      "grad_norm": 6.55078125,
+      "learning_rate": 9.353569046030688e-05,
+      "loss": 1.8042,
+      "step": 979
+    },
+    {
+      "epoch": 0.9046849757673667,
+      "grad_norm": 5.57421875,
+      "learning_rate": 9.352901934623083e-05,
+      "loss": 1.603,
+      "step": 980
+    },
+    {
+      "epoch": 0.9056081237018232,
+      "grad_norm": 6.70703125,
+      "learning_rate": 9.352234823215477e-05,
+      "loss": 1.9219,
+      "step": 981
+    },
+    {
+      "epoch": 0.9065312716362797,
+      "grad_norm": 4.734375,
+      "learning_rate": 9.351567711807873e-05,
+      "loss": 1.7161,
+      "step": 982
+    },
+    {
+      "epoch": 0.9074544195707362,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.350900600400268e-05,
+      "loss": 1.7185,
+      "step": 983
+    },
+    {
+      "epoch": 0.9083775675051927,
+      "grad_norm": 7.08984375,
+      "learning_rate": 9.350233488992662e-05,
+      "loss": 1.2151,
+      "step": 984
+    },
+    {
+      "epoch": 0.9093007154396492,
+      "grad_norm": 6.078125,
+      "learning_rate": 9.349566377585057e-05,
+      "loss": 1.6221,
+      "step": 985
+    },
+    {
+      "epoch": 0.9102238633741057,
+      "grad_norm": 4.859375,
+      "learning_rate": 9.348899266177453e-05,
+      "loss": 1.2326,
+      "step": 986
+    },
+    {
+      "epoch": 0.9111470113085622,
+      "grad_norm": 4.83984375,
+      "learning_rate": 9.348232154769847e-05,
+      "loss": 1.5092,
+      "step": 987
+    },
+    {
+      "epoch": 0.9120701592430187,
+      "grad_norm": 6.046875,
+      "learning_rate": 9.347565043362242e-05,
+      "loss": 1.5823,
+      "step": 988
+    },
+    {
+      "epoch": 0.9129933071774752,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.346897931954636e-05,
+      "loss": 1.7266,
+      "step": 989
+    },
+    {
+      "epoch": 0.9139164551119316,
+      "grad_norm": 6.41796875,
+      "learning_rate": 9.346230820547032e-05,
+      "loss": 1.5753,
+      "step": 990
+    },
+    {
+      "epoch": 0.9148396030463882,
+      "grad_norm": 8.4453125,
+      "learning_rate": 9.345563709139427e-05,
+      "loss": 1.5088,
+      "step": 991
+    },
+    {
+      "epoch": 0.9157627509808447,
+      "grad_norm": 4.51953125,
+      "learning_rate": 9.344896597731821e-05,
+      "loss": 1.8713,
+      "step": 992
+    },
+    {
+      "epoch": 0.9166858989153012,
+      "grad_norm": 5.48046875,
+      "learning_rate": 9.344229486324216e-05,
+      "loss": 1.6277,
+      "step": 993
+    },
+    {
+      "epoch": 0.9176090468497576,
+      "grad_norm": 4.234375,
+      "learning_rate": 9.343562374916612e-05,
+      "loss": 1.5293,
+      "step": 994
+    },
+    {
+      "epoch": 0.9185321947842142,
+      "grad_norm": 6.546875,
+      "learning_rate": 9.342895263509006e-05,
+      "loss": 1.8459,
+      "step": 995
+    },
+    {
+      "epoch": 0.9194553427186707,
+      "grad_norm": 7.2265625,
+      "learning_rate": 9.342228152101401e-05,
+      "loss": 1.8059,
+      "step": 996
+    },
+    {
+      "epoch": 0.9203784906531272,
+      "grad_norm": 5.828125,
+      "learning_rate": 9.341561040693795e-05,
+      "loss": 1.2915,
+      "step": 997
+    },
+    {
+      "epoch": 0.9213016385875836,
+      "grad_norm": 5.30859375,
+      "learning_rate": 9.340893929286191e-05,
+      "loss": 1.4839,
+      "step": 998
+    },
+    {
+      "epoch": 0.9222247865220402,
+      "grad_norm": 5.48046875,
+      "learning_rate": 9.340226817878586e-05,
+      "loss": 1.4783,
+      "step": 999
+    },
+    {
+      "epoch": 0.9231479344564967,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.33955970647098e-05,
+      "loss": 1.5859,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9231479344564967,
+      "eval_loss": 0.9296875,
+      "eval_runtime": 23.3057,
+      "eval_samples_per_second": 978.345,
+      "eval_steps_per_second": 61.187,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9240710823909531,
+      "grad_norm": 4.203125,
+      "learning_rate": 9.338892595063376e-05,
+      "loss": 1.6025,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9249942303254096,
+      "grad_norm": 4.12890625,
+      "learning_rate": 9.338225483655772e-05,
+      "loss": 1.6182,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9259173782598662,
+      "grad_norm": 5.75,
+      "learning_rate": 9.337558372248167e-05,
+      "loss": 1.8306,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9268405261943227,
+      "grad_norm": 6.140625,
+      "learning_rate": 9.336891260840561e-05,
+      "loss": 1.5002,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9277636741287791,
+      "grad_norm": 5.890625,
+      "learning_rate": 9.336224149432956e-05,
+      "loss": 1.71,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9286868220632356,
+      "grad_norm": 6.51953125,
+      "learning_rate": 9.335557038025352e-05,
+      "loss": 1.6074,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9296099699976921,
+      "grad_norm": 6.28515625,
+      "learning_rate": 9.334889926617746e-05,
+      "loss": 1.5688,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9305331179321487,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.334222815210141e-05,
+      "loss": 1.5261,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9314562658666051,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.333555703802535e-05,
+      "loss": 1.6099,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9323794138010616,
+      "grad_norm": 5.08203125,
+      "learning_rate": 9.332888592394931e-05,
+      "loss": 1.5815,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9333025617355181,
+      "grad_norm": 5.09765625,
+      "learning_rate": 9.332221480987326e-05,
+      "loss": 1.6262,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9342257096699746,
+      "grad_norm": 5.046875,
+      "learning_rate": 9.33155436957972e-05,
+      "loss": 1.6848,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9351488576044311,
+      "grad_norm": 6.40234375,
+      "learning_rate": 9.330887258172115e-05,
+      "loss": 1.6758,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9360720055388876,
+      "grad_norm": 6.3828125,
+      "learning_rate": 9.33022014676451e-05,
+      "loss": 1.6411,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9369951534733441,
+      "grad_norm": 6.734375,
+      "learning_rate": 9.329553035356905e-05,
+      "loss": 1.5249,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9379183014078006,
+      "grad_norm": 6.6796875,
+      "learning_rate": 9.3288859239493e-05,
+      "loss": 1.5647,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9388414493422571,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.328218812541694e-05,
+      "loss": 1.6465,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9397645972767136,
+      "grad_norm": 6.27734375,
+      "learning_rate": 9.32755170113409e-05,
+      "loss": 1.6035,
+      "step": 1018
+    },
+    {
+      "epoch": 0.94068774521117,
+      "grad_norm": 7.6796875,
+      "learning_rate": 9.326884589726485e-05,
+      "loss": 1.7441,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9416108931456266,
+      "grad_norm": 5.765625,
+      "learning_rate": 9.326217478318879e-05,
+      "loss": 1.6499,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9425340410800831,
+      "grad_norm": 4.828125,
+      "learning_rate": 9.325550366911274e-05,
+      "loss": 1.0149,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9434571890145396,
+      "grad_norm": 5.58203125,
+      "learning_rate": 9.32488325550367e-05,
+      "loss": 1.5903,
+      "step": 1022
+    },
+    {
+      "epoch": 0.944380336948996,
+      "grad_norm": 4.2734375,
+      "learning_rate": 9.324216144096064e-05,
+      "loss": 1.7229,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9453034848834526,
+      "grad_norm": 4.84765625,
+      "learning_rate": 9.323549032688459e-05,
+      "loss": 1.7375,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9462266328179091,
+      "grad_norm": 4.83203125,
+      "learning_rate": 9.322881921280853e-05,
+      "loss": 2.0337,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9471497807523656,
+      "grad_norm": 4.4453125,
+      "learning_rate": 9.322214809873249e-05,
+      "loss": 1.4819,
+      "step": 1026
+    },
+    {
+      "epoch": 0.948072928686822,
+      "grad_norm": 4.42578125,
+      "learning_rate": 9.321547698465645e-05,
+      "loss": 1.4926,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9489960766212786,
+      "grad_norm": 4.57421875,
+      "learning_rate": 9.32088058705804e-05,
+      "loss": 1.614,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9499192245557351,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.320213475650434e-05,
+      "loss": 1.7271,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9508423724901915,
+      "grad_norm": 5.27734375,
+      "learning_rate": 9.319546364242829e-05,
+      "loss": 1.5417,
+      "step": 1030
+    },
+    {
+      "epoch": 0.951765520424648,
+      "grad_norm": 6.6328125,
+      "learning_rate": 9.318879252835225e-05,
+      "loss": 1.8098,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9526886683591046,
+      "grad_norm": 5.75390625,
+      "learning_rate": 9.318212141427619e-05,
+      "loss": 1.6641,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9536118162935611,
+      "grad_norm": 5.5546875,
+      "learning_rate": 9.317545030020014e-05,
+      "loss": 1.7668,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9545349642280175,
+      "grad_norm": 6.28515625,
+      "learning_rate": 9.316877918612408e-05,
+      "loss": 1.1262,
+      "step": 1034
+    },
+    {
+      "epoch": 0.955458112162474,
+      "grad_norm": 5.73046875,
+      "learning_rate": 9.316210807204804e-05,
+      "loss": 1.6301,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9563812600969306,
+      "grad_norm": 4.97265625,
+      "learning_rate": 9.315543695797199e-05,
+      "loss": 1.6736,
+      "step": 1036
+    },
+    {
+      "epoch": 0.957304408031387,
+      "grad_norm": 4.0078125,
+      "learning_rate": 9.314876584389593e-05,
+      "loss": 1.1693,
+      "step": 1037
+    },
+    {
+      "epoch": 0.9582275559658435,
+      "grad_norm": 5.66796875,
+      "learning_rate": 9.314209472981988e-05,
+      "loss": 1.7546,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9591507039003,
+      "grad_norm": 6.15234375,
+      "learning_rate": 9.313542361574384e-05,
+      "loss": 1.7239,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9600738518347565,
+      "grad_norm": 5.87109375,
+      "learning_rate": 9.312875250166778e-05,
+      "loss": 1.5864,
+      "step": 1040
+    },
+    {
+      "epoch": 0.960996999769213,
+      "grad_norm": 5.36328125,
+      "learning_rate": 9.312208138759173e-05,
+      "loss": 1.7222,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9619201477036695,
+      "grad_norm": 6.50390625,
+      "learning_rate": 9.311541027351567e-05,
+      "loss": 1.7256,
+      "step": 1042
+    },
+    {
+      "epoch": 0.962843295638126,
+      "grad_norm": 6.34765625,
+      "learning_rate": 9.310873915943963e-05,
+      "loss": 1.9543,
+      "step": 1043
+    },
+    {
+      "epoch": 0.9637664435725825,
+      "grad_norm": 7.0,
+      "learning_rate": 9.310206804536358e-05,
+      "loss": 1.5813,
+      "step": 1044
+    },
+    {
+      "epoch": 0.964689591507039,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.309539693128752e-05,
+      "loss": 1.4851,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9656127394414955,
+      "grad_norm": 5.95703125,
+      "learning_rate": 9.308872581721147e-05,
+      "loss": 1.5093,
+      "step": 1046
+    },
+    {
+      "epoch": 0.966535887375952,
+      "grad_norm": 5.7734375,
+      "learning_rate": 9.308205470313543e-05,
+      "loss": 1.1499,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9674590353104084,
+      "grad_norm": 5.3203125,
+      "learning_rate": 9.307538358905937e-05,
+      "loss": 1.1495,
+      "step": 1048
+    },
+    {
+      "epoch": 0.968382183244865,
+      "grad_norm": 4.875,
+      "learning_rate": 9.306871247498332e-05,
+      "loss": 1.7661,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9693053311793215,
+      "grad_norm": 6.125,
+      "learning_rate": 9.306204136090726e-05,
+      "loss": 1.6672,
+      "step": 1050
+    },
+    {
+      "epoch": 0.970228479113778,
+      "grad_norm": 7.4609375,
+      "learning_rate": 9.305537024683124e-05,
+      "loss": 1.6372,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9711516270482344,
+      "grad_norm": 5.51953125,
+      "learning_rate": 9.304869913275518e-05,
+      "loss": 1.3215,
+      "step": 1052
+    },
+    {
+      "epoch": 0.972074774982691,
+      "grad_norm": 6.7890625,
+      "learning_rate": 9.304202801867913e-05,
+      "loss": 1.8623,
+      "step": 1053
+    },
+    {
+      "epoch": 0.9729979229171475,
+      "grad_norm": 5.20703125,
+      "learning_rate": 9.303535690460307e-05,
+      "loss": 1.4141,
+      "step": 1054
+    },
+    {
+      "epoch": 0.973921070851604,
+      "grad_norm": 4.6953125,
+      "learning_rate": 9.302868579052703e-05,
+      "loss": 1.5728,
+      "step": 1055
+    },
+    {
+      "epoch": 0.9748442187860604,
+      "grad_norm": 5.4296875,
+      "learning_rate": 9.302201467645098e-05,
+      "loss": 1.6165,
+      "step": 1056
+    },
+    {
+      "epoch": 0.975767366720517,
+      "grad_norm": 4.95703125,
+      "learning_rate": 9.301534356237492e-05,
+      "loss": 1.2343,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9766905146549735,
+      "grad_norm": 6.265625,
+      "learning_rate": 9.300867244829887e-05,
+      "loss": 1.772,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9776136625894299,
+      "grad_norm": 5.47265625,
+      "learning_rate": 9.300200133422283e-05,
+      "loss": 1.5791,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9785368105238864,
+      "grad_norm": 4.3203125,
+      "learning_rate": 9.299533022014677e-05,
+      "loss": 1.4755,
+      "step": 1060
+    },
+    {
+      "epoch": 0.979459958458343,
+      "grad_norm": 4.56640625,
+      "learning_rate": 9.298865910607072e-05,
+      "loss": 1.7073,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9803831063927995,
+      "grad_norm": 4.953125,
+      "learning_rate": 9.298198799199466e-05,
+      "loss": 1.8699,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9813062543272559,
+      "grad_norm": 5.9140625,
+      "learning_rate": 9.297531687791862e-05,
+      "loss": 1.8987,
+      "step": 1063
+    },
+    {
+      "epoch": 0.9822294022617124,
+      "grad_norm": 7.07421875,
+      "learning_rate": 9.296864576384257e-05,
+      "loss": 1.3163,
+      "step": 1064
+    },
+    {
+      "epoch": 0.983152550196169,
+      "grad_norm": 6.16015625,
+      "learning_rate": 9.296197464976651e-05,
+      "loss": 1.4651,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9840756981306255,
+      "grad_norm": 3.82421875,
+      "learning_rate": 9.295530353569046e-05,
+      "loss": 1.6228,
+      "step": 1066
+    },
+    {
+      "epoch": 0.9849988460650819,
+      "grad_norm": 4.22265625,
+      "learning_rate": 9.294863242161442e-05,
+      "loss": 1.2941,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9859219939995384,
+      "grad_norm": 5.234375,
+      "learning_rate": 9.294196130753836e-05,
+      "loss": 1.708,
+      "step": 1068
+    },
+    {
+      "epoch": 0.986845141933995,
+      "grad_norm": 5.86328125,
+      "learning_rate": 9.293529019346231e-05,
+      "loss": 1.6572,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9877682898684514,
+      "grad_norm": 4.94140625,
+      "learning_rate": 9.292861907938625e-05,
+      "loss": 1.5369,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9886914378029079,
+      "grad_norm": 6.06640625,
+      "learning_rate": 9.292194796531021e-05,
+      "loss": 1.4672,
+      "step": 1071
+    },
+    {
+      "epoch": 0.9896145857373644,
+      "grad_norm": 6.5859375,
+      "learning_rate": 9.291527685123416e-05,
+      "loss": 1.8726,
+      "step": 1072
+    },
+    {
+      "epoch": 0.990537733671821,
+      "grad_norm": 7.08203125,
+      "learning_rate": 9.29086057371581e-05,
+      "loss": 1.8376,
+      "step": 1073
+    },
+    {
+      "epoch": 0.9914608816062774,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.290193462308205e-05,
+      "loss": 1.3589,
+      "step": 1074
+    },
+    {
+      "epoch": 0.9923840295407339,
+      "grad_norm": 4.63671875,
+      "learning_rate": 9.289526350900601e-05,
+      "loss": 1.6405,
+      "step": 1075
+    },
+    {
+      "epoch": 0.9933071774751904,
+      "grad_norm": 4.390625,
+      "learning_rate": 9.288859239492997e-05,
+      "loss": 1.7305,
+      "step": 1076
+    },
+    {
+      "epoch": 0.9942303254096468,
+      "grad_norm": 5.4296875,
+      "learning_rate": 9.288192128085391e-05,
+      "loss": 1.7634,
+      "step": 1077
+    },
+    {
+      "epoch": 0.9951534733441034,
+      "grad_norm": 7.15625,
+      "learning_rate": 9.287525016677786e-05,
+      "loss": 1.7163,
+      "step": 1078
+    },
+    {
+      "epoch": 0.9960766212785599,
+      "grad_norm": 7.3359375,
+      "learning_rate": 9.28685790527018e-05,
+      "loss": 1.7214,
+      "step": 1079
+    },
+    {
+      "epoch": 0.9969997692130164,
+      "grad_norm": 6.07421875,
+      "learning_rate": 9.286190793862576e-05,
+      "loss": 1.6061,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9979229171474728,
+      "grad_norm": 6.9609375,
+      "learning_rate": 9.285523682454971e-05,
+      "loss": 1.8003,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9988460650819294,
+      "grad_norm": 7.66796875,
+      "learning_rate": 9.284856571047365e-05,
+      "loss": 1.3994,
+      "step": 1082
+    },
+    {
+      "epoch": 0.9997692130163859,
+      "grad_norm": 5.6328125,
+      "learning_rate": 9.28418945963976e-05,
+      "loss": 1.637,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0006923609508425,
+      "grad_norm": 3.35546875,
+      "learning_rate": 9.283522348232156e-05,
+      "loss": 1.1548,
+      "step": 1084
+    },
+    {
+      "epoch": 1.001615508885299,
+      "grad_norm": 3.412109375,
+      "learning_rate": 9.28285523682455e-05,
+      "loss": 0.9741,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0025386568197554,
+      "grad_norm": 4.2109375,
+      "learning_rate": 9.282188125416945e-05,
+      "loss": 1.2368,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0034618047542119,
+      "grad_norm": 5.0546875,
+      "learning_rate": 9.28152101400934e-05,
+      "loss": 1.6316,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0043849526886683,
+      "grad_norm": 7.09765625,
+      "learning_rate": 9.280853902601735e-05,
+      "loss": 1.7778,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0053081006231248,
+      "grad_norm": 3.935546875,
+      "learning_rate": 9.28018679119413e-05,
+      "loss": 1.4246,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0062312485575813,
+      "grad_norm": 4.39453125,
+      "learning_rate": 9.279519679786524e-05,
+      "loss": 1.481,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0071543964920378,
+      "grad_norm": 4.859375,
+      "learning_rate": 9.278852568378919e-05,
+      "loss": 1.291,
+      "step": 1091
+    },
+    {
+      "epoch": 1.0080775444264944,
+      "grad_norm": 5.52734375,
+      "learning_rate": 9.278185456971315e-05,
+      "loss": 1.6538,
+      "step": 1092
+    },
+    {
+      "epoch": 1.009000692360951,
+      "grad_norm": 4.65234375,
+      "learning_rate": 9.27751834556371e-05,
+      "loss": 1.4314,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0099238402954074,
+      "grad_norm": 5.5234375,
+      "learning_rate": 9.276851234156104e-05,
+      "loss": 1.5679,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0108469882298639,
+      "grad_norm": 5.24609375,
+      "learning_rate": 9.276184122748498e-05,
+      "loss": 1.4297,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0117701361643203,
+      "grad_norm": 5.578125,
+      "learning_rate": 9.275517011340894e-05,
+      "loss": 1.719,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0126932840987768,
+      "grad_norm": 5.98828125,
+      "learning_rate": 9.274849899933289e-05,
+      "loss": 1.7236,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0136164320332333,
+      "grad_norm": 5.20703125,
+      "learning_rate": 9.274182788525683e-05,
+      "loss": 1.5188,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0145395799676897,
+      "grad_norm": 5.91796875,
+      "learning_rate": 9.273515677118078e-05,
+      "loss": 1.4684,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0154627279021464,
+      "grad_norm": 4.6171875,
+      "learning_rate": 9.272848565710474e-05,
+      "loss": 1.5646,
+      "step": 1100
+    },
+    {
+      "epoch": 1.016385875836603,
+      "grad_norm": 4.75390625,
+      "learning_rate": 9.27218145430287e-05,
+      "loss": 1.4172,
+      "step": 1101
+    },
+    {
+      "epoch": 1.0173090237710594,
+      "grad_norm": 5.74609375,
+      "learning_rate": 9.271514342895264e-05,
+      "loss": 1.8022,
+      "step": 1102
+    },
+    {
+      "epoch": 1.0182321717055158,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.270847231487659e-05,
+      "loss": 1.2927,
+      "step": 1103
+    },
+    {
+      "epoch": 1.0191553196399723,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.270180120080055e-05,
+      "loss": 1.5271,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0200784675744288,
+      "grad_norm": 6.54296875,
+      "learning_rate": 9.269513008672449e-05,
+      "loss": 1.6819,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0210016155088852,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.268845897264844e-05,
+      "loss": 1.5149,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0219247634433417,
+      "grad_norm": 6.11328125,
+      "learning_rate": 9.268178785857238e-05,
+      "loss": 1.6802,
+      "step": 1107
+    },
+    {
+      "epoch": 1.0228479113777982,
+      "grad_norm": 5.2265625,
+      "learning_rate": 9.267511674449634e-05,
+      "loss": 1.8127,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0237710593122549,
+      "grad_norm": 5.640625,
+      "learning_rate": 9.266844563042029e-05,
+      "loss": 1.4617,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0246942072467113,
+      "grad_norm": 5.08203125,
+      "learning_rate": 9.266177451634423e-05,
+      "loss": 1.2969,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0256173551811678,
+      "grad_norm": 6.98046875,
+      "learning_rate": 9.265510340226818e-05,
+      "loss": 1.7305,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0265405031156243,
+      "grad_norm": 3.966796875,
+      "learning_rate": 9.264843228819214e-05,
+      "loss": 1.8816,
+      "step": 1112
+    },
+    {
+      "epoch": 1.0274636510500808,
+      "grad_norm": 4.86328125,
+      "learning_rate": 9.264176117411608e-05,
+      "loss": 1.6683,
+      "step": 1113
+    },
+    {
+      "epoch": 1.0283867989845372,
+      "grad_norm": 5.19140625,
+      "learning_rate": 9.263509006004003e-05,
+      "loss": 1.5781,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0293099469189937,
+      "grad_norm": 5.66796875,
+      "learning_rate": 9.262841894596397e-05,
+      "loss": 1.6003,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0302330948534502,
+      "grad_norm": 5.81640625,
+      "learning_rate": 9.262174783188793e-05,
+      "loss": 1.5408,
+      "step": 1116
+    },
+    {
+      "epoch": 1.0311562427879069,
+      "grad_norm": 5.4140625,
+      "learning_rate": 9.261507671781188e-05,
+      "loss": 1.5732,
+      "step": 1117
+    },
+    {
+      "epoch": 1.0320793907223633,
+      "grad_norm": 4.5859375,
+      "learning_rate": 9.260840560373582e-05,
+      "loss": 1.7153,
+      "step": 1118
+    },
+    {
+      "epoch": 1.0330025386568198,
+      "grad_norm": 5.19921875,
+      "learning_rate": 9.260173448965977e-05,
+      "loss": 1.5974,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0339256865912763,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.259506337558373e-05,
+      "loss": 1.6589,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0348488345257327,
+      "grad_norm": 5.55078125,
+      "learning_rate": 9.258839226150767e-05,
+      "loss": 1.4924,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0357719824601892,
+      "grad_norm": 4.83984375,
+      "learning_rate": 9.258172114743162e-05,
+      "loss": 1.5154,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0366951303946457,
+      "grad_norm": 5.40234375,
+      "learning_rate": 9.257505003335557e-05,
+      "loss": 1.7073,
+      "step": 1123
+    },
+    {
+      "epoch": 1.0376182783291021,
+      "grad_norm": 4.734375,
+      "learning_rate": 9.256837891927952e-05,
+      "loss": 1.5798,
+      "step": 1124
+    },
+    {
+      "epoch": 1.0385414262635588,
+      "grad_norm": 3.89453125,
+      "learning_rate": 9.256170780520347e-05,
+      "loss": 1.5762,
+      "step": 1125
+    },
+    {
+      "epoch": 1.0394645741980153,
+      "grad_norm": 3.912109375,
+      "learning_rate": 9.255503669112743e-05,
+      "loss": 1.4795,
+      "step": 1126
+    },
+    {
+      "epoch": 1.0403877221324718,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.254836557705137e-05,
+      "loss": 1.4421,
+      "step": 1127
+    },
+    {
+      "epoch": 1.0413108700669282,
+      "grad_norm": 5.3359375,
+      "learning_rate": 9.254169446297532e-05,
+      "loss": 1.7573,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0422340180013847,
+      "grad_norm": 6.30078125,
+      "learning_rate": 9.253502334889928e-05,
+      "loss": 1.5725,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0431571659358412,
+      "grad_norm": 6.13671875,
+      "learning_rate": 9.252835223482322e-05,
+      "loss": 1.5358,
+      "step": 1130
+    },
+    {
+      "epoch": 1.0440803138702976,
+      "grad_norm": 6.38671875,
+      "learning_rate": 9.252168112074717e-05,
+      "loss": 1.5845,
+      "step": 1131
+    },
+    {
+      "epoch": 1.0450034618047541,
+      "grad_norm": 6.30859375,
+      "learning_rate": 9.251501000667111e-05,
+      "loss": 1.8074,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0459266097392108,
+      "grad_norm": 6.81640625,
+      "learning_rate": 9.250833889259507e-05,
+      "loss": 1.6895,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0468497576736673,
+      "grad_norm": 5.4296875,
+      "learning_rate": 9.250166777851902e-05,
+      "loss": 1.5059,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0477729056081238,
+      "grad_norm": 4.859375,
+      "learning_rate": 9.249499666444296e-05,
+      "loss": 1.772,
+      "step": 1135
+    },
+    {
+      "epoch": 1.0486960535425802,
+      "grad_norm": 6.14453125,
+      "learning_rate": 9.248832555036691e-05,
+      "loss": 1.5432,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0496192014770367,
+      "grad_norm": 5.828125,
+      "learning_rate": 9.248165443629087e-05,
+      "loss": 1.4747,
+      "step": 1137
+    },
+    {
+      "epoch": 1.0505423494114932,
+      "grad_norm": 5.578125,
+      "learning_rate": 9.247498332221481e-05,
+      "loss": 1.6279,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0514654973459496,
+      "grad_norm": 5.4296875,
+      "learning_rate": 9.246831220813876e-05,
+      "loss": 1.4764,
+      "step": 1139
+    },
+    {
+      "epoch": 1.052388645280406,
+      "grad_norm": 5.01953125,
+      "learning_rate": 9.246164109406272e-05,
+      "loss": 1.3867,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0533117932148626,
+      "grad_norm": 5.26953125,
+      "learning_rate": 9.245496997998666e-05,
+      "loss": 1.2252,
+      "step": 1141
+    },
+    {
+      "epoch": 1.0542349411493193,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.244829886591061e-05,
+      "loss": 1.4495,
+      "step": 1142
+    },
+    {
+      "epoch": 1.0551580890837757,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.244162775183456e-05,
+      "loss": 1.2444,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0560812370182322,
+      "grad_norm": 5.06640625,
+      "learning_rate": 9.243495663775851e-05,
+      "loss": 1.594,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0570043849526887,
+      "grad_norm": 6.75390625,
+      "learning_rate": 9.242828552368246e-05,
+      "loss": 1.5884,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0579275328871451,
+      "grad_norm": 7.00390625,
+      "learning_rate": 9.24216144096064e-05,
+      "loss": 1.7502,
+      "step": 1146
+    },
+    {
+      "epoch": 1.0588506808216016,
+      "grad_norm": 6.94140625,
+      "learning_rate": 9.241494329553035e-05,
+      "loss": 1.6643,
+      "step": 1147
+    },
+    {
+      "epoch": 1.059773828756058,
+      "grad_norm": 5.12890625,
+      "learning_rate": 9.240827218145431e-05,
+      "loss": 1.6604,
+      "step": 1148
+    },
+    {
+      "epoch": 1.0606969766905145,
+      "grad_norm": 4.48046875,
+      "learning_rate": 9.240160106737825e-05,
+      "loss": 1.6609,
+      "step": 1149
+    },
+    {
+      "epoch": 1.0616201246249712,
+      "grad_norm": 4.296875,
+      "learning_rate": 9.23949299533022e-05,
+      "loss": 1.5054,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0625432725594277,
+      "grad_norm": 4.2890625,
+      "learning_rate": 9.238825883922616e-05,
+      "loss": 1.7817,
+      "step": 1151
+    },
+    {
+      "epoch": 1.0634664204938842,
+      "grad_norm": 5.88671875,
+      "learning_rate": 9.23815877251501e-05,
+      "loss": 1.6987,
+      "step": 1152
+    },
+    {
+      "epoch": 1.0643895684283406,
+      "grad_norm": 7.52734375,
+      "learning_rate": 9.237491661107406e-05,
+      "loss": 1.5134,
+      "step": 1153
+    },
+    {
+      "epoch": 1.0653127163627971,
+      "grad_norm": 5.6953125,
+      "learning_rate": 9.236824549699801e-05,
+      "loss": 1.4458,
+      "step": 1154
+    },
+    {
+      "epoch": 1.0662358642972536,
+      "grad_norm": 5.5234375,
+      "learning_rate": 9.236157438292195e-05,
+      "loss": 1.7146,
+      "step": 1155
+    },
+    {
+      "epoch": 1.06715901223171,
+      "grad_norm": 6.68359375,
+      "learning_rate": 9.23549032688459e-05,
+      "loss": 1.5564,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0680821601661665,
+      "grad_norm": 6.11328125,
+      "learning_rate": 9.234823215476986e-05,
+      "loss": 1.8232,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0690053081006232,
+      "grad_norm": 6.37890625,
+      "learning_rate": 9.23415610406938e-05,
+      "loss": 1.6282,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0699284560350797,
+      "grad_norm": 4.50390625,
+      "learning_rate": 9.233488992661775e-05,
+      "loss": 1.1858,
+      "step": 1159
+    },
+    {
+      "epoch": 1.0708516039695362,
+      "grad_norm": 5.7890625,
+      "learning_rate": 9.23282188125417e-05,
+      "loss": 1.6021,
+      "step": 1160
+    },
+    {
+      "epoch": 1.0717747519039926,
+      "grad_norm": 6.64453125,
+      "learning_rate": 9.232154769846565e-05,
+      "loss": 1.6257,
+      "step": 1161
+    },
+    {
+      "epoch": 1.072697899838449,
+      "grad_norm": 6.8515625,
+      "learning_rate": 9.23148765843896e-05,
+      "loss": 1.699,
+      "step": 1162
+    },
+    {
+      "epoch": 1.0736210477729056,
+      "grad_norm": 5.71484375,
+      "learning_rate": 9.230820547031354e-05,
+      "loss": 1.4126,
+      "step": 1163
+    },
+    {
+      "epoch": 1.074544195707362,
+      "grad_norm": 5.54296875,
+      "learning_rate": 9.230153435623749e-05,
+      "loss": 1.5669,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0754673436418185,
+      "grad_norm": 5.296875,
+      "learning_rate": 9.229486324216145e-05,
+      "loss": 1.5073,
+      "step": 1165
+    },
+    {
+      "epoch": 1.0763904915762752,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.22881921280854e-05,
+      "loss": 1.4009,
+      "step": 1166
+    },
+    {
+      "epoch": 1.0773136395107317,
+      "grad_norm": 5.09375,
+      "learning_rate": 9.228152101400934e-05,
+      "loss": 1.4463,
+      "step": 1167
+    },
+    {
+      "epoch": 1.0782367874451881,
+      "grad_norm": 6.1640625,
+      "learning_rate": 9.227484989993329e-05,
+      "loss": 1.418,
+      "step": 1168
+    },
+    {
+      "epoch": 1.0791599353796446,
+      "grad_norm": 5.09765625,
+      "learning_rate": 9.226817878585724e-05,
+      "loss": 1.6909,
+      "step": 1169
+    },
+    {
+      "epoch": 1.080083083314101,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.226150767178119e-05,
+      "loss": 1.5823,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0810062312485575,
+      "grad_norm": 4.51953125,
+      "learning_rate": 9.225483655770514e-05,
+      "loss": 1.7126,
+      "step": 1171
+    },
+    {
+      "epoch": 1.081929379183014,
+      "grad_norm": 5.2578125,
+      "learning_rate": 9.224816544362908e-05,
+      "loss": 1.6545,
+      "step": 1172
+    },
+    {
+      "epoch": 1.0828525271174705,
+      "grad_norm": 5.37109375,
+      "learning_rate": 9.224149432955304e-05,
+      "loss": 1.5762,
+      "step": 1173
+    },
+    {
+      "epoch": 1.083775675051927,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.223482321547699e-05,
+      "loss": 1.325,
+      "step": 1174
+    },
+    {
+      "epoch": 1.0846988229863836,
+      "grad_norm": 6.484375,
+      "learning_rate": 9.222815210140093e-05,
+      "loss": 1.6548,
+      "step": 1175
+    },
+    {
+      "epoch": 1.0856219709208401,
+      "grad_norm": 5.6640625,
+      "learning_rate": 9.222148098732489e-05,
+      "loss": 1.7952,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0865451188552966,
+      "grad_norm": 6.37890625,
+      "learning_rate": 9.221480987324884e-05,
+      "loss": 1.4452,
+      "step": 1177
+    },
+    {
+      "epoch": 1.087468266789753,
+      "grad_norm": 5.6171875,
+      "learning_rate": 9.22081387591728e-05,
+      "loss": 1.4934,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0883914147242095,
+      "grad_norm": 4.43359375,
+      "learning_rate": 9.220146764509674e-05,
+      "loss": 1.4338,
+      "step": 1179
+    },
+    {
+      "epoch": 1.089314562658666,
+      "grad_norm": 5.52734375,
+      "learning_rate": 9.219479653102069e-05,
+      "loss": 1.3093,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0902377105931225,
+      "grad_norm": 6.22265625,
+      "learning_rate": 9.218812541694464e-05,
+      "loss": 1.7593,
+      "step": 1181
+    },
+    {
+      "epoch": 1.0911608585275792,
+      "grad_norm": 6.50390625,
+      "learning_rate": 9.218145430286859e-05,
+      "loss": 1.323,
+      "step": 1182
+    },
+    {
+      "epoch": 1.0920840064620356,
+      "grad_norm": 6.02734375,
+      "learning_rate": 9.217478318879253e-05,
+      "loss": 1.5483,
+      "step": 1183
+    },
+    {
+      "epoch": 1.093007154396492,
+      "grad_norm": 5.734375,
+      "learning_rate": 9.216811207471648e-05,
+      "loss": 1.5703,
+      "step": 1184
+    },
+    {
+      "epoch": 1.0939303023309486,
+      "grad_norm": 4.69921875,
+      "learning_rate": 9.216144096064044e-05,
+      "loss": 1.4093,
+      "step": 1185
+    },
+    {
+      "epoch": 1.094853450265405,
+      "grad_norm": 7.2578125,
+      "learning_rate": 9.215476984656438e-05,
+      "loss": 1.583,
+      "step": 1186
+    },
+    {
+      "epoch": 1.0957765981998615,
+      "grad_norm": 5.2890625,
+      "learning_rate": 9.214809873248833e-05,
+      "loss": 1.2965,
+      "step": 1187
+    },
+    {
+      "epoch": 1.096699746134318,
+      "grad_norm": 4.87109375,
+      "learning_rate": 9.214142761841228e-05,
+      "loss": 1.5645,
+      "step": 1188
+    },
+    {
+      "epoch": 1.0976228940687744,
+      "grad_norm": 5.10546875,
+      "learning_rate": 9.213475650433623e-05,
+      "loss": 1.5764,
+      "step": 1189
+    },
+    {
+      "epoch": 1.098546042003231,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.212808539026018e-05,
+      "loss": 1.469,
+      "step": 1190
+    },
+    {
+      "epoch": 1.0994691899376876,
+      "grad_norm": 4.95703125,
+      "learning_rate": 9.212141427618413e-05,
+      "loss": 1.7725,
+      "step": 1191
+    },
+    {
+      "epoch": 1.100392337872144,
+      "grad_norm": 5.47265625,
+      "learning_rate": 9.211474316210807e-05,
+      "loss": 1.459,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1013154858066005,
+      "grad_norm": 4.44140625,
+      "learning_rate": 9.210807204803203e-05,
+      "loss": 1.6628,
+      "step": 1193
+    },
+    {
+      "epoch": 1.102238633741057,
+      "grad_norm": 4.42578125,
+      "learning_rate": 9.210140093395598e-05,
+      "loss": 1.4109,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1031617816755135,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.209472981987992e-05,
+      "loss": 1.446,
+      "step": 1195
+    },
+    {
+      "epoch": 1.10408492960997,
+      "grad_norm": 5.34765625,
+      "learning_rate": 9.208805870580387e-05,
+      "loss": 1.582,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1050080775444264,
+      "grad_norm": 4.62890625,
+      "learning_rate": 9.208138759172783e-05,
+      "loss": 1.6077,
+      "step": 1197
+    },
+    {
+      "epoch": 1.105931225478883,
+      "grad_norm": 5.39453125,
+      "learning_rate": 9.207471647765177e-05,
+      "loss": 1.6626,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1068543734133396,
+      "grad_norm": 6.66796875,
+      "learning_rate": 9.206804536357572e-05,
+      "loss": 1.5491,
+      "step": 1199
+    },
+    {
+      "epoch": 1.107777521347796,
+      "grad_norm": 5.80859375,
+      "learning_rate": 9.206137424949966e-05,
+      "loss": 1.541,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1087006692822525,
+      "grad_norm": 5.125,
+      "learning_rate": 9.205470313542362e-05,
+      "loss": 1.3374,
+      "step": 1201
+    },
+    {
+      "epoch": 1.109623817216709,
+      "grad_norm": 4.94140625,
+      "learning_rate": 9.204803202134758e-05,
+      "loss": 1.6782,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1105469651511655,
+      "grad_norm": 4.3828125,
+      "learning_rate": 9.204136090727152e-05,
+      "loss": 1.8843,
+      "step": 1203
+    },
+    {
+      "epoch": 1.111470113085622,
+      "grad_norm": 5.09765625,
+      "learning_rate": 9.203468979319547e-05,
+      "loss": 1.6956,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1123932610200784,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.202801867911942e-05,
+      "loss": 1.5592,
+      "step": 1205
+    },
+    {
+      "epoch": 1.1133164089545349,
+      "grad_norm": 5.01171875,
+      "learning_rate": 9.202134756504337e-05,
+      "loss": 1.564,
+      "step": 1206
+    },
+    {
+      "epoch": 1.1142395568889913,
+      "grad_norm": 4.12890625,
+      "learning_rate": 9.201467645096732e-05,
+      "loss": 1.2706,
+      "step": 1207
+    },
+    {
+      "epoch": 1.115162704823448,
+      "grad_norm": 4.91015625,
+      "learning_rate": 9.200800533689127e-05,
+      "loss": 1.479,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1160858527579045,
+      "grad_norm": 4.609375,
+      "learning_rate": 9.200133422281521e-05,
+      "loss": 2.0493,
+      "step": 1209
+    },
+    {
+      "epoch": 1.117009000692361,
+      "grad_norm": 5.0546875,
+      "learning_rate": 9.199466310873917e-05,
+      "loss": 1.6399,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1179321486268174,
+      "grad_norm": 5.01171875,
+      "learning_rate": 9.198799199466312e-05,
+      "loss": 1.2255,
+      "step": 1211
+    },
+    {
+      "epoch": 1.118855296561274,
+      "grad_norm": 5.625,
+      "learning_rate": 9.198132088058706e-05,
+      "loss": 1.3223,
+      "step": 1212
+    },
+    {
+      "epoch": 1.1197784444957304,
+      "grad_norm": 6.73828125,
+      "learning_rate": 9.1974649766511e-05,
+      "loss": 1.5007,
+      "step": 1213
+    },
+    {
+      "epoch": 1.1207015924301869,
+      "grad_norm": 6.8203125,
+      "learning_rate": 9.196797865243497e-05,
+      "loss": 1.6023,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1216247403646435,
+      "grad_norm": 6.08203125,
+      "learning_rate": 9.196130753835891e-05,
+      "loss": 1.6482,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1225478882991,
+      "grad_norm": 6.078125,
+      "learning_rate": 9.195463642428286e-05,
+      "loss": 1.2251,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1234710362335565,
+      "grad_norm": 6.1328125,
+      "learning_rate": 9.19479653102068e-05,
+      "loss": 1.6067,
+      "step": 1217
+    },
+    {
+      "epoch": 1.124394184168013,
+      "grad_norm": 5.1171875,
+      "learning_rate": 9.194129419613076e-05,
+      "loss": 1.4653,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1253173321024694,
+      "grad_norm": 6.35546875,
+      "learning_rate": 9.19346230820547e-05,
+      "loss": 1.6697,
+      "step": 1219
+    },
+    {
+      "epoch": 1.126240480036926,
+      "grad_norm": 5.8359375,
+      "learning_rate": 9.192795196797865e-05,
+      "loss": 1.6567,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1271636279713824,
+      "grad_norm": 5.1640625,
+      "learning_rate": 9.19212808539026e-05,
+      "loss": 1.6921,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1280867759058388,
+      "grad_norm": 5.6796875,
+      "learning_rate": 9.191460973982656e-05,
+      "loss": 1.4022,
+      "step": 1222
+    },
+    {
+      "epoch": 1.1290099238402953,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.19079386257505e-05,
+      "loss": 1.4941,
+      "step": 1223
+    },
+    {
+      "epoch": 1.129933071774752,
+      "grad_norm": 4.640625,
+      "learning_rate": 9.190126751167445e-05,
+      "loss": 1.6147,
+      "step": 1224
+    },
+    {
+      "epoch": 1.1308562197092085,
+      "grad_norm": 5.31640625,
+      "learning_rate": 9.189459639759839e-05,
+      "loss": 1.824,
+      "step": 1225
+    },
+    {
+      "epoch": 1.131779367643665,
+      "grad_norm": 5.56640625,
+      "learning_rate": 9.188792528352236e-05,
+      "loss": 1.6111,
+      "step": 1226
+    },
+    {
+      "epoch": 1.1327025155781214,
+      "grad_norm": 5.546875,
+      "learning_rate": 9.188125416944631e-05,
+      "loss": 1.3263,
+      "step": 1227
+    },
+    {
+      "epoch": 1.1336256635125779,
+      "grad_norm": 4.7265625,
+      "learning_rate": 9.187458305537026e-05,
+      "loss": 1.1575,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1345488114470343,
+      "grad_norm": 5.859375,
+      "learning_rate": 9.18679119412942e-05,
+      "loss": 1.5986,
+      "step": 1229
+    },
+    {
+      "epoch": 1.1354719593814908,
+      "grad_norm": 5.6640625,
+      "learning_rate": 9.186124082721816e-05,
+      "loss": 1.7119,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1363951073159473,
+      "grad_norm": 4.76953125,
+      "learning_rate": 9.18545697131421e-05,
+      "loss": 1.5081,
+      "step": 1231
+    },
+    {
+      "epoch": 1.137318255250404,
+      "grad_norm": 5.35546875,
+      "learning_rate": 9.184789859906605e-05,
+      "loss": 1.6157,
+      "step": 1232
+    },
+    {
+      "epoch": 1.1382414031848604,
+      "grad_norm": 5.94921875,
+      "learning_rate": 9.184122748499e-05,
+      "loss": 1.8955,
+      "step": 1233
+    },
+    {
+      "epoch": 1.139164551119317,
+      "grad_norm": 5.90234375,
+      "learning_rate": 9.183455637091396e-05,
+      "loss": 1.5942,
+      "step": 1234
+    },
+    {
+      "epoch": 1.1400876990537734,
+      "grad_norm": 6.8203125,
+      "learning_rate": 9.18278852568379e-05,
+      "loss": 1.5552,
+      "step": 1235
+    },
+    {
+      "epoch": 1.1410108469882299,
+      "grad_norm": 6.51171875,
+      "learning_rate": 9.182121414276185e-05,
+      "loss": 1.6123,
+      "step": 1236
+    },
+    {
+      "epoch": 1.1419339949226863,
+      "grad_norm": 5.84375,
+      "learning_rate": 9.181454302868579e-05,
+      "loss": 1.3,
+      "step": 1237
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 6.3984375,
+      "learning_rate": 9.180787191460975e-05,
+      "loss": 1.6846,
+      "step": 1238
+    },
+    {
+      "epoch": 1.1437802907915993,
+      "grad_norm": 6.875,
+      "learning_rate": 9.18012008005337e-05,
+      "loss": 1.4868,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1447034387260557,
+      "grad_norm": 7.26953125,
+      "learning_rate": 9.179452968645764e-05,
+      "loss": 2.0217,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1456265866605124,
+      "grad_norm": 5.09765625,
+      "learning_rate": 9.178785857238159e-05,
+      "loss": 1.4961,
+      "step": 1241
+    },
+    {
+      "epoch": 1.146549734594969,
+      "grad_norm": 5.48046875,
+      "learning_rate": 9.178118745830555e-05,
+      "loss": 1.707,
+      "step": 1242
+    },
+    {
+      "epoch": 1.1474728825294254,
+      "grad_norm": 6.08203125,
+      "learning_rate": 9.177451634422949e-05,
+      "loss": 1.2684,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1483960304638818,
+      "grad_norm": 5.74609375,
+      "learning_rate": 9.176784523015344e-05,
+      "loss": 1.6558,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1493191783983383,
+      "grad_norm": 6.08203125,
+      "learning_rate": 9.176117411607738e-05,
+      "loss": 1.6282,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1502423263327948,
+      "grad_norm": 5.83984375,
+      "learning_rate": 9.175450300200134e-05,
+      "loss": 1.6628,
+      "step": 1246
+    },
+    {
+      "epoch": 1.1511654742672512,
+      "grad_norm": 7.0859375,
+      "learning_rate": 9.174783188792529e-05,
+      "loss": 1.6221,
+      "step": 1247
+    },
+    {
+      "epoch": 1.152088622201708,
+      "grad_norm": 5.6796875,
+      "learning_rate": 9.174116077384923e-05,
+      "loss": 1.6521,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1530117701361644,
+      "grad_norm": 5.83984375,
+      "learning_rate": 9.173448965977318e-05,
+      "loss": 1.4155,
+      "step": 1249
+    },
+    {
+      "epoch": 1.1539349180706209,
+      "grad_norm": 5.4140625,
+      "learning_rate": 9.172781854569714e-05,
+      "loss": 1.606,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1548580660050773,
+      "grad_norm": 5.26953125,
+      "learning_rate": 9.17211474316211e-05,
+      "loss": 1.6643,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1557812139395338,
+      "grad_norm": 4.25,
+      "learning_rate": 9.171447631754504e-05,
+      "loss": 1.3879,
+      "step": 1252
+    },
+    {
+      "epoch": 1.1567043618739903,
+      "grad_norm": 4.9609375,
+      "learning_rate": 9.170780520346899e-05,
+      "loss": 1.5092,
+      "step": 1253
+    },
+    {
+      "epoch": 1.1576275098084468,
+      "grad_norm": 5.3203125,
+      "learning_rate": 9.170113408939293e-05,
+      "loss": 1.6931,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1585506577429032,
+      "grad_norm": 6.96484375,
+      "learning_rate": 9.169446297531689e-05,
+      "loss": 1.4854,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1594738056773597,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.168779186124084e-05,
+      "loss": 1.6082,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1603969536118164,
+      "grad_norm": 6.30078125,
+      "learning_rate": 9.168112074716478e-05,
+      "loss": 1.6997,
+      "step": 1257
+    },
+    {
+      "epoch": 1.1613201015462729,
+      "grad_norm": 5.3046875,
+      "learning_rate": 9.167444963308873e-05,
+      "loss": 1.3036,
+      "step": 1258
+    },
+    {
+      "epoch": 1.1622432494807293,
+      "grad_norm": 5.48046875,
+      "learning_rate": 9.166777851901269e-05,
+      "loss": 1.4553,
+      "step": 1259
+    },
+    {
+      "epoch": 1.1631663974151858,
+      "grad_norm": 4.78515625,
+      "learning_rate": 9.166110740493663e-05,
+      "loss": 1.5256,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1640895453496423,
+      "grad_norm": 6.55859375,
+      "learning_rate": 9.165443629086058e-05,
+      "loss": 1.6111,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1650126932840987,
+      "grad_norm": 5.671875,
+      "learning_rate": 9.164776517678452e-05,
+      "loss": 1.5928,
+      "step": 1262
+    },
+    {
+      "epoch": 1.1659358412185552,
+      "grad_norm": 6.33203125,
+      "learning_rate": 9.164109406270848e-05,
+      "loss": 1.7388,
+      "step": 1263
+    },
+    {
+      "epoch": 1.166858989153012,
+      "grad_norm": 5.0390625,
+      "learning_rate": 9.163442294863243e-05,
+      "loss": 1.8547,
+      "step": 1264
+    },
+    {
+      "epoch": 1.1677821370874684,
+      "grad_norm": 6.16015625,
+      "learning_rate": 9.162775183455637e-05,
+      "loss": 1.6155,
+      "step": 1265
+    },
+    {
+      "epoch": 1.1687052850219248,
+      "grad_norm": 5.38671875,
+      "learning_rate": 9.162108072048032e-05,
+      "loss": 1.6431,
+      "step": 1266
+    },
+    {
+      "epoch": 1.1696284329563813,
+      "grad_norm": 5.7734375,
+      "learning_rate": 9.161440960640428e-05,
+      "loss": 1.6255,
+      "step": 1267
+    },
+    {
+      "epoch": 1.1705515808908378,
+      "grad_norm": 5.2578125,
+      "learning_rate": 9.160773849232822e-05,
+      "loss": 1.4705,
+      "step": 1268
+    },
+    {
+      "epoch": 1.1714747288252942,
+      "grad_norm": 6.02734375,
+      "learning_rate": 9.160106737825217e-05,
+      "loss": 1.4907,
+      "step": 1269
+    },
+    {
+      "epoch": 1.1723978767597507,
+      "grad_norm": 5.9296875,
+      "learning_rate": 9.159439626417611e-05,
+      "loss": 1.4546,
+      "step": 1270
+    },
+    {
+      "epoch": 1.1733210246942072,
+      "grad_norm": 5.578125,
+      "learning_rate": 9.158772515010007e-05,
+      "loss": 1.7136,
+      "step": 1271
+    },
+    {
+      "epoch": 1.1742441726286637,
+      "grad_norm": 5.59765625,
+      "learning_rate": 9.158105403602402e-05,
+      "loss": 1.5771,
+      "step": 1272
+    },
+    {
+      "epoch": 1.1751673205631201,
+      "grad_norm": 4.16015625,
+      "learning_rate": 9.157438292194796e-05,
+      "loss": 1.7046,
+      "step": 1273
+    },
+    {
+      "epoch": 1.1760904684975768,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.156771180787191e-05,
+      "loss": 1.8916,
+      "step": 1274
+    },
+    {
+      "epoch": 1.1770136164320333,
+      "grad_norm": 6.41796875,
+      "learning_rate": 9.156104069379587e-05,
+      "loss": 1.6577,
+      "step": 1275
+    },
+    {
+      "epoch": 1.1779367643664898,
+      "grad_norm": 6.734375,
+      "learning_rate": 9.155436957971983e-05,
+      "loss": 1.5151,
+      "step": 1276
+    },
+    {
+      "epoch": 1.1788599123009462,
+      "grad_norm": 4.76953125,
+      "learning_rate": 9.154769846564377e-05,
+      "loss": 1.5989,
+      "step": 1277
+    },
+    {
+      "epoch": 1.1797830602354027,
+      "grad_norm": 5.35546875,
+      "learning_rate": 9.154102735156772e-05,
+      "loss": 1.5959,
+      "step": 1278
+    },
+    {
+      "epoch": 1.1807062081698592,
+      "grad_norm": 4.58984375,
+      "learning_rate": 9.153435623749168e-05,
+      "loss": 1.4647,
+      "step": 1279
+    },
+    {
+      "epoch": 1.1816293561043156,
+      "grad_norm": 5.62109375,
+      "learning_rate": 9.152768512341562e-05,
+      "loss": 1.3519,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1825525040387723,
+      "grad_norm": 6.109375,
+      "learning_rate": 9.152101400933957e-05,
+      "loss": 1.5879,
+      "step": 1281
+    },
+    {
+      "epoch": 1.1834756519732288,
+      "grad_norm": 6.25390625,
+      "learning_rate": 9.151434289526351e-05,
+      "loss": 1.4878,
+      "step": 1282
+    },
+    {
+      "epoch": 1.1843987999076853,
+      "grad_norm": 5.4296875,
+      "learning_rate": 9.150767178118747e-05,
+      "loss": 1.623,
+      "step": 1283
+    },
+    {
+      "epoch": 1.1853219478421417,
+      "grad_norm": 6.28515625,
+      "learning_rate": 9.150100066711142e-05,
+      "loss": 1.3247,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1862450957765982,
+      "grad_norm": 6.109375,
+      "learning_rate": 9.149432955303536e-05,
+      "loss": 1.4988,
+      "step": 1285
+    },
+    {
+      "epoch": 1.1871682437110547,
+      "grad_norm": 4.1484375,
+      "learning_rate": 9.148765843895931e-05,
+      "loss": 0.7976,
+      "step": 1286
+    },
+    {
+      "epoch": 1.1880913916455111,
+      "grad_norm": 3.970703125,
+      "learning_rate": 9.148098732488327e-05,
+      "loss": 1.3647,
+      "step": 1287
+    },
+    {
+      "epoch": 1.1890145395799676,
+      "grad_norm": 5.39453125,
+      "learning_rate": 9.147431621080721e-05,
+      "loss": 1.6846,
+      "step": 1288
+    },
+    {
+      "epoch": 1.189937687514424,
+      "grad_norm": 5.68359375,
+      "learning_rate": 9.146764509673116e-05,
+      "loss": 1.5327,
+      "step": 1289
+    },
+    {
+      "epoch": 1.1908608354488808,
+      "grad_norm": 4.703125,
+      "learning_rate": 9.14609739826551e-05,
+      "loss": 1.5647,
+      "step": 1290
+    },
+    {
+      "epoch": 1.1917839833833372,
+      "grad_norm": 5.73828125,
+      "learning_rate": 9.145430286857906e-05,
+      "loss": 2.0581,
+      "step": 1291
+    },
+    {
+      "epoch": 1.1927071313177937,
+      "grad_norm": 6.04296875,
+      "learning_rate": 9.144763175450301e-05,
+      "loss": 1.4941,
+      "step": 1292
+    },
+    {
+      "epoch": 1.1936302792522502,
+      "grad_norm": 5.31640625,
+      "learning_rate": 9.144096064042695e-05,
+      "loss": 1.647,
+      "step": 1293
+    },
+    {
+      "epoch": 1.1945534271867067,
+      "grad_norm": 5.8515625,
+      "learning_rate": 9.14342895263509e-05,
+      "loss": 1.7988,
+      "step": 1294
+    },
+    {
+      "epoch": 1.1954765751211631,
+      "grad_norm": 5.46484375,
+      "learning_rate": 9.142761841227486e-05,
+      "loss": 1.8242,
+      "step": 1295
+    },
+    {
+      "epoch": 1.1963997230556196,
+      "grad_norm": 7.41796875,
+      "learning_rate": 9.14209472981988e-05,
+      "loss": 1.3851,
+      "step": 1296
+    },
+    {
+      "epoch": 1.1973228709900763,
+      "grad_norm": 5.14453125,
+      "learning_rate": 9.141427618412275e-05,
+      "loss": 1.7537,
+      "step": 1297
+    },
+    {
+      "epoch": 1.1982460189245328,
+      "grad_norm": 5.10546875,
+      "learning_rate": 9.140760507004669e-05,
+      "loss": 1.689,
+      "step": 1298
+    },
+    {
+      "epoch": 1.1991691668589892,
+      "grad_norm": 7.1328125,
+      "learning_rate": 9.140093395597065e-05,
+      "loss": 1.4878,
+      "step": 1299
+    },
+    {
+      "epoch": 1.2000923147934457,
+      "grad_norm": 4.3984375,
+      "learning_rate": 9.13942628418946e-05,
+      "loss": 1.2218,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2010154627279022,
+      "grad_norm": 4.3984375,
+      "learning_rate": 9.138759172781856e-05,
+      "loss": 1.4351,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2019386106623586,
+      "grad_norm": 4.75,
+      "learning_rate": 9.13809206137425e-05,
+      "loss": 1.4076,
+      "step": 1302
+    },
+    {
+      "epoch": 1.202861758596815,
+      "grad_norm": 4.0234375,
+      "learning_rate": 9.137424949966645e-05,
+      "loss": 1.5237,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2037849065312716,
+      "grad_norm": 5.76953125,
+      "learning_rate": 9.13675783855904e-05,
+      "loss": 1.5442,
+      "step": 1304
+    },
+    {
+      "epoch": 1.204708054465728,
+      "grad_norm": 6.1953125,
+      "learning_rate": 9.136090727151435e-05,
+      "loss": 1.7207,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2056312024001845,
+      "grad_norm": 5.4140625,
+      "learning_rate": 9.13542361574383e-05,
+      "loss": 1.5918,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2065543503346412,
+      "grad_norm": 4.38671875,
+      "learning_rate": 9.134756504336224e-05,
+      "loss": 1.7822,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2074774982690977,
+      "grad_norm": 6.171875,
+      "learning_rate": 9.13408939292862e-05,
+      "loss": 1.6279,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2084006462035541,
+      "grad_norm": 6.609375,
+      "learning_rate": 9.133422281521015e-05,
+      "loss": 1.7473,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2093237941380106,
+      "grad_norm": 6.70703125,
+      "learning_rate": 9.132755170113409e-05,
+      "loss": 1.8313,
+      "step": 1310
+    },
+    {
+      "epoch": 1.210246942072467,
+      "grad_norm": 5.00390625,
+      "learning_rate": 9.132088058705804e-05,
+      "loss": 1.5183,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2111700900069235,
+      "grad_norm": 5.765625,
+      "learning_rate": 9.1314209472982e-05,
+      "loss": 1.4282,
+      "step": 1312
+    },
+    {
+      "epoch": 1.21209323794138,
+      "grad_norm": 5.5859375,
+      "learning_rate": 9.130753835890594e-05,
+      "loss": 1.6082,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2130163858758367,
+      "grad_norm": 6.42578125,
+      "learning_rate": 9.130086724482989e-05,
+      "loss": 1.4431,
+      "step": 1314
+    },
+    {
+      "epoch": 1.2139395338102932,
+      "grad_norm": 5.64453125,
+      "learning_rate": 9.129419613075383e-05,
+      "loss": 1.2494,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2148626817447497,
+      "grad_norm": 5.671875,
+      "learning_rate": 9.128752501667779e-05,
+      "loss": 1.6897,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2157858296792061,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.128085390260174e-05,
+      "loss": 1.5679,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2167089776136626,
+      "grad_norm": 5.1015625,
+      "learning_rate": 9.127418278852568e-05,
+      "loss": 1.4526,
+      "step": 1318
+    },
+    {
+      "epoch": 1.217632125548119,
+      "grad_norm": 5.2578125,
+      "learning_rate": 9.126751167444963e-05,
+      "loss": 1.5183,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2185552734825755,
+      "grad_norm": 5.3203125,
+      "learning_rate": 9.126084056037359e-05,
+      "loss": 1.1979,
+      "step": 1320
+    },
+    {
+      "epoch": 1.219478421417032,
+      "grad_norm": 6.18359375,
+      "learning_rate": 9.125416944629753e-05,
+      "loss": 1.6084,
+      "step": 1321
+    },
+    {
+      "epoch": 1.2204015693514885,
+      "grad_norm": 5.984375,
+      "learning_rate": 9.124749833222148e-05,
+      "loss": 1.4037,
+      "step": 1322
+    },
+    {
+      "epoch": 1.2213247172859452,
+      "grad_norm": 6.39453125,
+      "learning_rate": 9.124082721814542e-05,
+      "loss": 1.6038,
+      "step": 1323
+    },
+    {
+      "epoch": 1.2222478652204016,
+      "grad_norm": 5.6484375,
+      "learning_rate": 9.123415610406938e-05,
+      "loss": 1.645,
+      "step": 1324
+    },
+    {
+      "epoch": 1.223171013154858,
+      "grad_norm": 5.65625,
+      "learning_rate": 9.122748498999333e-05,
+      "loss": 1.2991,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2240941610893146,
+      "grad_norm": 6.3671875,
+      "learning_rate": 9.122081387591727e-05,
+      "loss": 1.5696,
+      "step": 1326
+    },
+    {
+      "epoch": 1.225017309023771,
+      "grad_norm": 5.03515625,
+      "learning_rate": 9.121414276184123e-05,
+      "loss": 1.7598,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2259404569582275,
+      "grad_norm": 5.18359375,
+      "learning_rate": 9.120747164776519e-05,
+      "loss": 1.4146,
+      "step": 1328
+    },
+    {
+      "epoch": 1.226863604892684,
+      "grad_norm": 6.04296875,
+      "learning_rate": 9.120080053368914e-05,
+      "loss": 1.51,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2277867528271407,
+      "grad_norm": 5.8515625,
+      "learning_rate": 9.119412941961308e-05,
+      "loss": 1.8059,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2287099007615971,
+      "grad_norm": 6.41015625,
+      "learning_rate": 9.118745830553703e-05,
+      "loss": 1.7656,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2296330486960536,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.118078719146099e-05,
+      "loss": 1.4475,
+      "step": 1332
+    },
+    {
+      "epoch": 1.23055619663051,
+      "grad_norm": 5.1328125,
+      "learning_rate": 9.117411607738493e-05,
+      "loss": 1.6421,
+      "step": 1333
+    },
+    {
+      "epoch": 1.2314793445649665,
+      "grad_norm": 3.623046875,
+      "learning_rate": 9.116744496330888e-05,
+      "loss": 1.7266,
+      "step": 1334
+    },
+    {
+      "epoch": 1.232402492499423,
+      "grad_norm": 5.22265625,
+      "learning_rate": 9.116077384923282e-05,
+      "loss": 1.4076,
+      "step": 1335
+    },
+    {
+      "epoch": 1.2333256404338795,
+      "grad_norm": 5.765625,
+      "learning_rate": 9.115410273515678e-05,
+      "loss": 1.5577,
+      "step": 1336
+    },
+    {
+      "epoch": 1.234248788368336,
+      "grad_norm": 6.015625,
+      "learning_rate": 9.114743162108073e-05,
+      "loss": 1.6494,
+      "step": 1337
+    },
+    {
+      "epoch": 1.2351719363027924,
+      "grad_norm": 6.0390625,
+      "learning_rate": 9.114076050700467e-05,
+      "loss": 1.5386,
+      "step": 1338
+    },
+    {
+      "epoch": 1.236095084237249,
+      "grad_norm": 6.3359375,
+      "learning_rate": 9.113408939292862e-05,
+      "loss": 1.5239,
+      "step": 1339
+    },
+    {
+      "epoch": 1.2370182321717056,
+      "grad_norm": 5.3828125,
+      "learning_rate": 9.112741827885258e-05,
+      "loss": 1.4299,
+      "step": 1340
+    },
+    {
+      "epoch": 1.237941380106162,
+      "grad_norm": 6.140625,
+      "learning_rate": 9.112074716477652e-05,
+      "loss": 1.7158,
+      "step": 1341
+    },
+    {
+      "epoch": 1.2388645280406185,
+      "grad_norm": 4.03515625,
+      "learning_rate": 9.111407605070047e-05,
+      "loss": 1.7,
+      "step": 1342
+    },
+    {
+      "epoch": 1.239787675975075,
+      "grad_norm": 4.56640625,
+      "learning_rate": 9.110740493662441e-05,
+      "loss": 1.5884,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2407108239095315,
+      "grad_norm": 4.921875,
+      "learning_rate": 9.110073382254837e-05,
+      "loss": 1.718,
+      "step": 1344
+    },
+    {
+      "epoch": 1.241633971843988,
+      "grad_norm": 5.7890625,
+      "learning_rate": 9.109406270847232e-05,
+      "loss": 1.6257,
+      "step": 1345
+    },
+    {
+      "epoch": 1.2425571197784444,
+      "grad_norm": 5.15234375,
+      "learning_rate": 9.108739159439626e-05,
+      "loss": 1.7332,
+      "step": 1346
+    },
+    {
+      "epoch": 1.243480267712901,
+      "grad_norm": 4.9140625,
+      "learning_rate": 9.108072048032021e-05,
+      "loss": 1.7925,
+      "step": 1347
+    },
+    {
+      "epoch": 1.2444034156473576,
+      "grad_norm": 4.7421875,
+      "learning_rate": 9.107404936624417e-05,
+      "loss": 1.436,
+      "step": 1348
+    },
+    {
+      "epoch": 1.245326563581814,
+      "grad_norm": 5.95703125,
+      "learning_rate": 9.106737825216811e-05,
+      "loss": 1.5464,
+      "step": 1349
+    },
+    {
+      "epoch": 1.2462497115162705,
+      "grad_norm": 5.9921875,
+      "learning_rate": 9.106070713809206e-05,
+      "loss": 1.646,
+      "step": 1350
+    },
+    {
+      "epoch": 1.247172859450727,
+      "grad_norm": 5.84765625,
+      "learning_rate": 9.1054036024016e-05,
+      "loss": 1.7029,
+      "step": 1351
+    },
+    {
+      "epoch": 1.2480960073851834,
+      "grad_norm": 4.66015625,
+      "learning_rate": 9.104736490993996e-05,
+      "loss": 1.7036,
+      "step": 1352
+    },
+    {
+      "epoch": 1.24901915531964,
+      "grad_norm": 6.125,
+      "learning_rate": 9.104069379586392e-05,
+      "loss": 1.8037,
+      "step": 1353
+    },
+    {
+      "epoch": 1.2499423032540964,
+      "grad_norm": 6.5546875,
+      "learning_rate": 9.103402268178787e-05,
+      "loss": 1.5835,
+      "step": 1354
+    },
+    {
+      "epoch": 1.2508654511885529,
+      "grad_norm": 7.140625,
+      "learning_rate": 9.102735156771181e-05,
+      "loss": 1.6707,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2517885991230093,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.102068045363576e-05,
+      "loss": 1.4857,
+      "step": 1356
+    },
+    {
+      "epoch": 1.252711747057466,
+      "grad_norm": 5.83203125,
+      "learning_rate": 9.101400933955972e-05,
+      "loss": 1.7847,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2536348949919225,
+      "grad_norm": 4.5859375,
+      "learning_rate": 9.100733822548366e-05,
+      "loss": 1.887,
+      "step": 1358
+    },
+    {
+      "epoch": 1.254558042926379,
+      "grad_norm": 4.734375,
+      "learning_rate": 9.100066711140761e-05,
+      "loss": 1.4753,
+      "step": 1359
+    },
+    {
+      "epoch": 1.2554811908608354,
+      "grad_norm": 4.32421875,
+      "learning_rate": 9.099399599733155e-05,
+      "loss": 1.7917,
+      "step": 1360
+    },
+    {
+      "epoch": 1.256404338795292,
+      "grad_norm": 4.375,
+      "learning_rate": 9.098732488325551e-05,
+      "loss": 1.6304,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2573274867297484,
+      "grad_norm": 4.640625,
+      "learning_rate": 9.098065376917946e-05,
+      "loss": 1.6042,
+      "step": 1362
+    },
+    {
+      "epoch": 1.258250634664205,
+      "grad_norm": 4.41015625,
+      "learning_rate": 9.09739826551034e-05,
+      "loss": 1.4056,
+      "step": 1363
+    },
+    {
+      "epoch": 1.2591737825986615,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.096731154102735e-05,
+      "loss": 1.5754,
+      "step": 1364
+    },
+    {
+      "epoch": 1.260096930533118,
+      "grad_norm": 5.453125,
+      "learning_rate": 9.096064042695131e-05,
+      "loss": 1.5779,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2610200784675745,
+      "grad_norm": 4.875,
+      "learning_rate": 9.095396931287525e-05,
+      "loss": 1.6565,
+      "step": 1366
+    },
+    {
+      "epoch": 1.261943226402031,
+      "grad_norm": 5.21484375,
+      "learning_rate": 9.09472981987992e-05,
+      "loss": 1.5649,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2628663743364874,
+      "grad_norm": 5.41796875,
+      "learning_rate": 9.094062708472316e-05,
+      "loss": 1.593,
+      "step": 1368
+    },
+    {
+      "epoch": 1.2637895222709439,
+      "grad_norm": 5.265625,
+      "learning_rate": 9.09339559706471e-05,
+      "loss": 1.5034,
+      "step": 1369
+    },
+    {
+      "epoch": 1.2647126702054003,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.092728485657105e-05,
+      "loss": 1.6489,
+      "step": 1370
+    },
+    {
+      "epoch": 1.2656358181398568,
+      "grad_norm": 6.34765625,
+      "learning_rate": 9.0920613742495e-05,
+      "loss": 1.3091,
+      "step": 1371
+    },
+    {
+      "epoch": 1.2665589660743133,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.091394262841895e-05,
+      "loss": 1.5825,
+      "step": 1372
+    },
+    {
+      "epoch": 1.26748211400877,
+      "grad_norm": 4.24609375,
+      "learning_rate": 9.09072715143429e-05,
+      "loss": 1.5398,
+      "step": 1373
+    },
+    {
+      "epoch": 1.2684052619432264,
+      "grad_norm": 3.947265625,
+      "learning_rate": 9.090060040026684e-05,
+      "loss": 1.8296,
+      "step": 1374
+    },
+    {
+      "epoch": 1.269328409877683,
+      "grad_norm": 4.921875,
+      "learning_rate": 9.089392928619079e-05,
+      "loss": 1.7175,
+      "step": 1375
+    },
+    {
+      "epoch": 1.2702515578121394,
+      "grad_norm": 7.43359375,
+      "learning_rate": 9.088725817211475e-05,
+      "loss": 1.6826,
+      "step": 1376
+    },
+    {
+      "epoch": 1.2711747057465959,
+      "grad_norm": 7.2578125,
+      "learning_rate": 9.088058705803871e-05,
+      "loss": 1.7949,
+      "step": 1377
+    },
+    {
+      "epoch": 1.2720978536810523,
+      "grad_norm": 5.21484375,
+      "learning_rate": 9.087391594396265e-05,
+      "loss": 1.4341,
+      "step": 1378
+    },
+    {
+      "epoch": 1.273021001615509,
+      "grad_norm": 5.83984375,
+      "learning_rate": 9.08672448298866e-05,
+      "loss": 1.7351,
+      "step": 1379
+    },
+    {
+      "epoch": 1.2739441495499655,
+      "grad_norm": 4.0546875,
+      "learning_rate": 9.086057371581054e-05,
+      "loss": 1.7327,
+      "step": 1380
+    },
+    {
+      "epoch": 1.274867297484422,
+      "grad_norm": 5.64453125,
+      "learning_rate": 9.08539026017345e-05,
+      "loss": 1.8618,
+      "step": 1381
+    },
+    {
+      "epoch": 1.2757904454188784,
+      "grad_norm": 5.86328125,
+      "learning_rate": 9.084723148765845e-05,
+      "loss": 1.3447,
+      "step": 1382
+    },
+    {
+      "epoch": 1.276713593353335,
+      "grad_norm": 6.62890625,
+      "learning_rate": 9.08405603735824e-05,
+      "loss": 1.4868,
+      "step": 1383
+    },
+    {
+      "epoch": 1.2776367412877914,
+      "grad_norm": 5.2578125,
+      "learning_rate": 9.083388925950634e-05,
+      "loss": 1.6182,
+      "step": 1384
+    },
+    {
+      "epoch": 1.2785598892222478,
+      "grad_norm": 6.125,
+      "learning_rate": 9.08272181454303e-05,
+      "loss": 1.6167,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2794830371567043,
+      "grad_norm": 6.37890625,
+      "learning_rate": 9.082054703135424e-05,
+      "loss": 1.6223,
+      "step": 1386
+    },
+    {
+      "epoch": 1.2804061850911608,
+      "grad_norm": 6.27734375,
+      "learning_rate": 9.081387591727819e-05,
+      "loss": 1.4058,
+      "step": 1387
+    },
+    {
+      "epoch": 1.2813293330256172,
+      "grad_norm": 5.3203125,
+      "learning_rate": 9.080720480320213e-05,
+      "loss": 1.6904,
+      "step": 1388
+    },
+    {
+      "epoch": 1.2822524809600737,
+      "grad_norm": 6.9921875,
+      "learning_rate": 9.080053368912609e-05,
+      "loss": 1.6902,
+      "step": 1389
+    },
+    {
+      "epoch": 1.2831756288945304,
+      "grad_norm": 5.1640625,
+      "learning_rate": 9.079386257505004e-05,
+      "loss": 1.6165,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2840987768289869,
+      "grad_norm": 6.8984375,
+      "learning_rate": 9.078719146097398e-05,
+      "loss": 1.2557,
+      "step": 1391
+    },
+    {
+      "epoch": 1.2850219247634433,
+      "grad_norm": 6.734375,
+      "learning_rate": 9.078052034689793e-05,
+      "loss": 1.5071,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2859450726978998,
+      "grad_norm": 6.47265625,
+      "learning_rate": 9.077384923282189e-05,
+      "loss": 1.4807,
+      "step": 1393
+    },
+    {
+      "epoch": 1.2868682206323563,
+      "grad_norm": 6.08984375,
+      "learning_rate": 9.076717811874583e-05,
+      "loss": 1.7539,
+      "step": 1394
+    },
+    {
+      "epoch": 1.2877913685668128,
+      "grad_norm": 5.59765625,
+      "learning_rate": 9.076050700466978e-05,
+      "loss": 1.5566,
+      "step": 1395
+    },
+    {
+      "epoch": 1.2887145165012694,
+      "grad_norm": 5.3828125,
+      "learning_rate": 9.075383589059372e-05,
+      "loss": 1.5596,
+      "step": 1396
+    },
+    {
+      "epoch": 1.289637664435726,
+      "grad_norm": 7.75390625,
+      "learning_rate": 9.074716477651768e-05,
+      "loss": 1.5889,
+      "step": 1397
+    },
+    {
+      "epoch": 1.2905608123701824,
+      "grad_norm": 5.828125,
+      "learning_rate": 9.074049366244163e-05,
+      "loss": 1.4817,
+      "step": 1398
+    },
+    {
+      "epoch": 1.2914839603046389,
+      "grad_norm": 7.1171875,
+      "learning_rate": 9.073382254836557e-05,
+      "loss": 1.9001,
+      "step": 1399
+    },
+    {
+      "epoch": 1.2924071082390953,
+      "grad_norm": 6.3125,
+      "learning_rate": 9.072715143428952e-05,
+      "loss": 1.7561,
+      "step": 1400
+    },
+    {
+      "epoch": 1.2933302561735518,
+      "grad_norm": 5.51953125,
+      "learning_rate": 9.072048032021348e-05,
+      "loss": 1.3307,
+      "step": 1401
+    },
+    {
+      "epoch": 1.2942534041080083,
+      "grad_norm": 4.953125,
+      "learning_rate": 9.071380920613744e-05,
+      "loss": 1.5674,
+      "step": 1402
+    },
+    {
+      "epoch": 1.2951765520424647,
+      "grad_norm": 4.265625,
+      "learning_rate": 9.070713809206138e-05,
+      "loss": 1.7214,
+      "step": 1403
+    },
+    {
+      "epoch": 1.2960996999769212,
+      "grad_norm": 4.484375,
+      "learning_rate": 9.070046697798533e-05,
+      "loss": 1.8926,
+      "step": 1404
+    },
+    {
+      "epoch": 1.2970228479113777,
+      "grad_norm": 3.892578125,
+      "learning_rate": 9.069379586390927e-05,
+      "loss": 1.6005,
+      "step": 1405
+    },
+    {
+      "epoch": 1.2979459958458344,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.068712474983323e-05,
+      "loss": 1.1433,
+      "step": 1406
+    },
+    {
+      "epoch": 1.2988691437802908,
+      "grad_norm": 4.97265625,
+      "learning_rate": 9.068045363575718e-05,
+      "loss": 1.5225,
+      "step": 1407
+    },
+    {
+      "epoch": 1.2997922917147473,
+      "grad_norm": 5.4609375,
+      "learning_rate": 9.067378252168112e-05,
+      "loss": 1.6577,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3007154396492038,
+      "grad_norm": 6.60546875,
+      "learning_rate": 9.066711140760508e-05,
+      "loss": 1.6738,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3016385875836602,
+      "grad_norm": 6.16015625,
+      "learning_rate": 9.066044029352903e-05,
+      "loss": 1.6895,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3025617355181167,
+      "grad_norm": 5.3984375,
+      "learning_rate": 9.065376917945297e-05,
+      "loss": 1.606,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3034848834525734,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.064709806537692e-05,
+      "loss": 1.6643,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3044080313870299,
+      "grad_norm": 4.98046875,
+      "learning_rate": 9.064042695130088e-05,
+      "loss": 1.7334,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3053311793214863,
+      "grad_norm": 4.98828125,
+      "learning_rate": 9.063375583722482e-05,
+      "loss": 1.6924,
+      "step": 1414
+    },
+    {
+      "epoch": 1.3062543272559428,
+      "grad_norm": 6.59375,
+      "learning_rate": 9.062708472314877e-05,
+      "loss": 1.5049,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3071774751903993,
+      "grad_norm": 5.41015625,
+      "learning_rate": 9.062041360907271e-05,
+      "loss": 1.3574,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3081006231248558,
+      "grad_norm": 4.03515625,
+      "learning_rate": 9.061374249499667e-05,
+      "loss": 1.436,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3090237710593122,
+      "grad_norm": 4.171875,
+      "learning_rate": 9.060707138092062e-05,
+      "loss": 1.7229,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3099469189937687,
+      "grad_norm": 5.27734375,
+      "learning_rate": 9.060040026684456e-05,
+      "loss": 0.9309,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3108700669282252,
+      "grad_norm": 5.13671875,
+      "learning_rate": 9.059372915276851e-05,
+      "loss": 1.6145,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3117932148626816,
+      "grad_norm": 3.875,
+      "learning_rate": 9.058705803869247e-05,
+      "loss": 1.4683,
+      "step": 1421
+    },
+    {
+      "epoch": 1.312716362797138,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.058038692461641e-05,
+      "loss": 1.4851,
+      "step": 1422
+    },
+    {
+      "epoch": 1.3136395107315948,
+      "grad_norm": 5.05078125,
+      "learning_rate": 9.057371581054036e-05,
+      "loss": 1.4282,
+      "step": 1423
+    },
+    {
+      "epoch": 1.3145626586660513,
+      "grad_norm": 5.7109375,
+      "learning_rate": 9.05670446964643e-05,
+      "loss": 1.5483,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3154858066005077,
+      "grad_norm": 4.390625,
+      "learning_rate": 9.056037358238826e-05,
+      "loss": 1.1928,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3164089545349642,
+      "grad_norm": 7.0703125,
+      "learning_rate": 9.055370246831221e-05,
+      "loss": 1.8879,
+      "step": 1426
+    },
+    {
+      "epoch": 1.3173321024694207,
+      "grad_norm": 5.73046875,
+      "learning_rate": 9.054703135423617e-05,
+      "loss": 1.5925,
+      "step": 1427
+    },
+    {
+      "epoch": 1.3182552504038771,
+      "grad_norm": 5.6796875,
+      "learning_rate": 9.054036024016011e-05,
+      "loss": 1.2795,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3191783983383338,
+      "grad_norm": 5.046875,
+      "learning_rate": 9.053368912608406e-05,
+      "loss": 1.8608,
+      "step": 1429
+    },
+    {
+      "epoch": 1.3201015462727903,
+      "grad_norm": 5.5859375,
+      "learning_rate": 9.052701801200802e-05,
+      "loss": 1.5701,
+      "step": 1430
+    },
+    {
+      "epoch": 1.3210246942072468,
+      "grad_norm": 4.51171875,
+      "learning_rate": 9.052034689793196e-05,
+      "loss": 1.7183,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3219478421417032,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.051367578385591e-05,
+      "loss": 1.8972,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3228709900761597,
+      "grad_norm": 4.171875,
+      "learning_rate": 9.050700466977985e-05,
+      "loss": 1.5472,
+      "step": 1433
+    },
+    {
+      "epoch": 1.3237941380106162,
+      "grad_norm": 4.70703125,
+      "learning_rate": 9.050033355570381e-05,
+      "loss": 1.5842,
+      "step": 1434
+    },
+    {
+      "epoch": 1.3247172859450727,
+      "grad_norm": 5.44921875,
+      "learning_rate": 9.049366244162776e-05,
+      "loss": 1.844,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3256404338795291,
+      "grad_norm": 6.015625,
+      "learning_rate": 9.04869913275517e-05,
+      "loss": 1.3812,
+      "step": 1436
+    },
+    {
+      "epoch": 1.3265635818139856,
+      "grad_norm": 5.06640625,
+      "learning_rate": 9.048032021347565e-05,
+      "loss": 1.2316,
+      "step": 1437
+    },
+    {
+      "epoch": 1.327486729748442,
+      "grad_norm": 4.36328125,
+      "learning_rate": 9.047364909939961e-05,
+      "loss": 1.5554,
+      "step": 1438
+    },
+    {
+      "epoch": 1.3284098776828988,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.046697798532355e-05,
+      "loss": 1.5869,
+      "step": 1439
+    },
+    {
+      "epoch": 1.3293330256173552,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.04603068712475e-05,
+      "loss": 1.6357,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3302561735518117,
+      "grad_norm": 4.9921875,
+      "learning_rate": 9.045363575717145e-05,
+      "loss": 1.6858,
+      "step": 1441
+    },
+    {
+      "epoch": 1.3311793214862682,
+      "grad_norm": 5.70703125,
+      "learning_rate": 9.04469646430954e-05,
+      "loss": 1.4003,
+      "step": 1442
+    },
+    {
+      "epoch": 1.3321024694207246,
+      "grad_norm": 5.5,
+      "learning_rate": 9.044029352901935e-05,
+      "loss": 1.7051,
+      "step": 1443
+    },
+    {
+      "epoch": 1.333025617355181,
+      "grad_norm": 6.0,
+      "learning_rate": 9.04336224149433e-05,
+      "loss": 1.6504,
+      "step": 1444
+    },
+    {
+      "epoch": 1.3339487652896378,
+      "grad_norm": 6.49609375,
+      "learning_rate": 9.042695130086724e-05,
+      "loss": 1.511,
+      "step": 1445
+    },
+    {
+      "epoch": 1.3348719132240943,
+      "grad_norm": 5.58984375,
+      "learning_rate": 9.04202801867912e-05,
+      "loss": 1.2622,
+      "step": 1446
+    },
+    {
+      "epoch": 1.3357950611585507,
+      "grad_norm": 4.66015625,
+      "learning_rate": 9.041360907271515e-05,
+      "loss": 1.4618,
+      "step": 1447
+    },
+    {
+      "epoch": 1.3367182090930072,
+      "grad_norm": 5.671875,
+      "learning_rate": 9.040693795863909e-05,
+      "loss": 1.5237,
+      "step": 1448
+    },
+    {
+      "epoch": 1.3376413570274637,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.040026684456304e-05,
+      "loss": 1.3364,
+      "step": 1449
+    },
+    {
+      "epoch": 1.3385645049619201,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.0393595730487e-05,
+      "loss": 1.5125,
+      "step": 1450
+    },
+    {
+      "epoch": 1.3394876528963766,
+      "grad_norm": 4.32421875,
+      "learning_rate": 9.038692461641094e-05,
+      "loss": 1.7925,
+      "step": 1451
+    },
+    {
+      "epoch": 1.340410800830833,
+      "grad_norm": 5.11328125,
+      "learning_rate": 9.03802535023349e-05,
+      "loss": 1.7773,
+      "step": 1452
+    },
+    {
+      "epoch": 1.3413339487652896,
+      "grad_norm": 4.9609375,
+      "learning_rate": 9.037358238825884e-05,
+      "loss": 1.7959,
+      "step": 1453
+    },
+    {
+      "epoch": 1.342257096699746,
+      "grad_norm": 4.7734375,
+      "learning_rate": 9.03669112741828e-05,
+      "loss": 1.7175,
+      "step": 1454
+    },
+    {
+      "epoch": 1.3431802446342027,
+      "grad_norm": 5.22265625,
+      "learning_rate": 9.036024016010675e-05,
+      "loss": 1.4958,
+      "step": 1455
+    },
+    {
+      "epoch": 1.3441033925686592,
+      "grad_norm": 5.9375,
+      "learning_rate": 9.03535690460307e-05,
+      "loss": 1.54,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3450265405031157,
+      "grad_norm": 3.896484375,
+      "learning_rate": 9.034689793195464e-05,
+      "loss": 1.5127,
+      "step": 1457
+    },
+    {
+      "epoch": 1.3459496884375721,
+      "grad_norm": 4.0546875,
+      "learning_rate": 9.03402268178786e-05,
+      "loss": 1.217,
+      "step": 1458
+    },
+    {
+      "epoch": 1.3468728363720286,
+      "grad_norm": 4.98828125,
+      "learning_rate": 9.033355570380254e-05,
+      "loss": 1.6055,
+      "step": 1459
+    },
+    {
+      "epoch": 1.347795984306485,
+      "grad_norm": 5.96875,
+      "learning_rate": 9.032688458972649e-05,
+      "loss": 1.6304,
+      "step": 1460
+    },
+    {
+      "epoch": 1.3487191322409415,
+      "grad_norm": 4.88671875,
+      "learning_rate": 9.032021347565044e-05,
+      "loss": 1.4519,
+      "step": 1461
+    },
+    {
+      "epoch": 1.3496422801753982,
+      "grad_norm": 6.54296875,
+      "learning_rate": 9.03135423615744e-05,
+      "loss": 1.8608,
+      "step": 1462
+    },
+    {
+      "epoch": 1.3505654281098547,
+      "grad_norm": 6.01953125,
+      "learning_rate": 9.030687124749834e-05,
+      "loss": 1.5791,
+      "step": 1463
+    },
+    {
+      "epoch": 1.3514885760443112,
+      "grad_norm": 5.34765625,
+      "learning_rate": 9.030020013342229e-05,
+      "loss": 1.4628,
+      "step": 1464
+    },
+    {
+      "epoch": 1.3524117239787676,
+      "grad_norm": 4.9765625,
+      "learning_rate": 9.029352901934623e-05,
+      "loss": 1.5769,
+      "step": 1465
+    },
+    {
+      "epoch": 1.353334871913224,
+      "grad_norm": 6.29296875,
+      "learning_rate": 9.028685790527019e-05,
+      "loss": 1.5989,
+      "step": 1466
+    },
+    {
+      "epoch": 1.3542580198476806,
+      "grad_norm": 5.64453125,
+      "learning_rate": 9.028018679119413e-05,
+      "loss": 1.7554,
+      "step": 1467
+    },
+    {
+      "epoch": 1.355181167782137,
+      "grad_norm": 6.1875,
+      "learning_rate": 9.027351567711808e-05,
+      "loss": 1.6021,
+      "step": 1468
+    },
+    {
+      "epoch": 1.3561043157165935,
+      "grad_norm": 4.43359375,
+      "learning_rate": 9.026684456304203e-05,
+      "loss": 1.6875,
+      "step": 1469
+    },
+    {
+      "epoch": 1.35702746365105,
+      "grad_norm": 5.671875,
+      "learning_rate": 9.026017344896598e-05,
+      "loss": 1.5945,
+      "step": 1470
+    },
+    {
+      "epoch": 1.3579506115855065,
+      "grad_norm": 6.4921875,
+      "learning_rate": 9.025350233488993e-05,
+      "loss": 1.7417,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3588737595199631,
+      "grad_norm": 6.3046875,
+      "learning_rate": 9.024683122081388e-05,
+      "loss": 1.6919,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3597969074544196,
+      "grad_norm": 5.35546875,
+      "learning_rate": 9.024016010673782e-05,
+      "loss": 1.8782,
+      "step": 1473
+    },
+    {
+      "epoch": 1.360720055388876,
+      "grad_norm": 5.4609375,
+      "learning_rate": 9.023348899266178e-05,
+      "loss": 1.5737,
+      "step": 1474
+    },
+    {
+      "epoch": 1.3616432033233326,
+      "grad_norm": 4.83984375,
+      "learning_rate": 9.022681787858573e-05,
+      "loss": 1.7549,
+      "step": 1475
+    },
+    {
+      "epoch": 1.362566351257789,
+      "grad_norm": 5.41796875,
+      "learning_rate": 9.022014676450967e-05,
+      "loss": 1.614,
+      "step": 1476
+    },
+    {
+      "epoch": 1.3634894991922455,
+      "grad_norm": 5.21484375,
+      "learning_rate": 9.021347565043363e-05,
+      "loss": 1.6299,
+      "step": 1477
+    },
+    {
+      "epoch": 1.3644126471267022,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.020680453635758e-05,
+      "loss": 1.4678,
+      "step": 1478
+    },
+    {
+      "epoch": 1.3653357950611587,
+      "grad_norm": 4.765625,
+      "learning_rate": 9.020013342228153e-05,
+      "loss": 1.4976,
+      "step": 1479
+    },
+    {
+      "epoch": 1.3662589429956151,
+      "grad_norm": 6.29296875,
+      "learning_rate": 9.019346230820548e-05,
+      "loss": 1.7234,
+      "step": 1480
+    },
+    {
+      "epoch": 1.3671820909300716,
+      "grad_norm": 3.86328125,
+      "learning_rate": 9.018679119412943e-05,
+      "loss": 1.2487,
+      "step": 1481
+    },
+    {
+      "epoch": 1.368105238864528,
+      "grad_norm": 4.66796875,
+      "learning_rate": 9.018012008005337e-05,
+      "loss": 1.5735,
+      "step": 1482
+    },
+    {
+      "epoch": 1.3690283867989845,
+      "grad_norm": 5.2109375,
+      "learning_rate": 9.017344896597733e-05,
+      "loss": 1.5115,
+      "step": 1483
+    },
+    {
+      "epoch": 1.369951534733441,
+      "grad_norm": 5.51171875,
+      "learning_rate": 9.016677785190128e-05,
+      "loss": 1.5674,
+      "step": 1484
+    },
+    {
+      "epoch": 1.3708746826678975,
+      "grad_norm": 4.3359375,
+      "learning_rate": 9.016010673782522e-05,
+      "loss": 1.7947,
+      "step": 1485
+    },
+    {
+      "epoch": 1.371797830602354,
+      "grad_norm": 5.01953125,
+      "learning_rate": 9.015343562374917e-05,
+      "loss": 1.582,
+      "step": 1486
+    },
+    {
+      "epoch": 1.3727209785368104,
+      "grad_norm": 4.80859375,
+      "learning_rate": 9.014676450967312e-05,
+      "loss": 1.2512,
+      "step": 1487
+    },
+    {
+      "epoch": 1.373644126471267,
+      "grad_norm": 4.95703125,
+      "learning_rate": 9.014009339559707e-05,
+      "loss": 1.2954,
+      "step": 1488
+    },
+    {
+      "epoch": 1.3745672744057236,
+      "grad_norm": 6.59375,
+      "learning_rate": 9.013342228152102e-05,
+      "loss": 1.6591,
+      "step": 1489
+    },
+    {
+      "epoch": 1.37549042234018,
+      "grad_norm": 5.5,
+      "learning_rate": 9.012675116744496e-05,
+      "loss": 1.5652,
+      "step": 1490
+    },
+    {
+      "epoch": 1.3764135702746365,
+      "grad_norm": 5.1796875,
+      "learning_rate": 9.012008005336892e-05,
+      "loss": 1.5172,
+      "step": 1491
+    },
+    {
+      "epoch": 1.377336718209093,
+      "grad_norm": 5.36328125,
+      "learning_rate": 9.011340893929287e-05,
+      "loss": 1.5947,
+      "step": 1492
+    },
+    {
+      "epoch": 1.3782598661435495,
+      "grad_norm": 4.90234375,
+      "learning_rate": 9.010673782521681e-05,
+      "loss": 1.8462,
+      "step": 1493
+    },
+    {
+      "epoch": 1.379183014078006,
+      "grad_norm": 4.07421875,
+      "learning_rate": 9.010006671114076e-05,
+      "loss": 1.6526,
+      "step": 1494
+    },
+    {
+      "epoch": 1.3801061620124626,
+      "grad_norm": 3.814453125,
+      "learning_rate": 9.009339559706472e-05,
+      "loss": 1.1341,
+      "step": 1495
+    },
+    {
+      "epoch": 1.381029309946919,
+      "grad_norm": 4.77734375,
+      "learning_rate": 9.008672448298866e-05,
+      "loss": 1.5581,
+      "step": 1496
+    },
+    {
+      "epoch": 1.3819524578813756,
+      "grad_norm": 4.59765625,
+      "learning_rate": 9.00800533689126e-05,
+      "loss": 1.5723,
+      "step": 1497
+    },
+    {
+      "epoch": 1.382875605815832,
+      "grad_norm": 4.4296875,
+      "learning_rate": 9.007338225483655e-05,
+      "loss": 1.6191,
+      "step": 1498
+    },
+    {
+      "epoch": 1.3837987537502885,
+      "grad_norm": 4.671875,
+      "learning_rate": 9.006671114076051e-05,
+      "loss": 1.687,
+      "step": 1499
+    },
+    {
+      "epoch": 1.384721901684745,
+      "grad_norm": 5.5078125,
+      "learning_rate": 9.006004002668446e-05,
+      "loss": 1.7947,
+      "step": 1500
+    },
+    {
+      "epoch": 1.3856450496192014,
+      "grad_norm": 7.51953125,
+      "learning_rate": 9.00533689126084e-05,
+      "loss": 0.9882,
+      "step": 1501
+    },
+    {
+      "epoch": 1.386568197553658,
+      "grad_norm": 5.41015625,
+      "learning_rate": 9.004669779853236e-05,
+      "loss": 1.5234,
+      "step": 1502
+    },
+    {
+      "epoch": 1.3874913454881144,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.004002668445632e-05,
+      "loss": 1.1596,
+      "step": 1503
+    },
+    {
+      "epoch": 1.3884144934225708,
+      "grad_norm": 5.3203125,
+      "learning_rate": 9.003335557038026e-05,
+      "loss": 1.5073,
+      "step": 1504
+    },
+    {
+      "epoch": 1.3893376413570275,
+      "grad_norm": 5.33203125,
+      "learning_rate": 9.002668445630421e-05,
+      "loss": 1.5471,
+      "step": 1505
+    },
+    {
+      "epoch": 1.390260789291484,
+      "grad_norm": 5.8671875,
+      "learning_rate": 9.002001334222816e-05,
+      "loss": 1.7251,
+      "step": 1506
+    },
+    {
+      "epoch": 1.3911839372259405,
+      "grad_norm": 5.5078125,
+      "learning_rate": 9.001334222815211e-05,
+      "loss": 1.6194,
+      "step": 1507
+    },
+    {
+      "epoch": 1.392107085160397,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.000667111407606e-05,
+      "loss": 1.5975,
+      "step": 1508
+    },
+    {
+      "epoch": 1.3930302330948534,
+      "grad_norm": 4.1875,
+      "learning_rate": 9e-05,
+      "loss": 1.5146,
+      "step": 1509
+    },
+    {
+      "epoch": 1.3939533810293099,
+      "grad_norm": 5.46484375,
+      "learning_rate": 8.999332888592395e-05,
+      "loss": 1.5341,
+      "step": 1510
+    },
+    {
+      "epoch": 1.3948765289637666,
+      "grad_norm": 5.05859375,
+      "learning_rate": 8.998665777184791e-05,
+      "loss": 1.7688,
+      "step": 1511
+    },
+    {
+      "epoch": 1.395799676898223,
+      "grad_norm": 3.62890625,
+      "learning_rate": 8.997998665777186e-05,
+      "loss": 1.1881,
+      "step": 1512
+    },
+    {
+      "epoch": 1.3967228248326795,
+      "grad_norm": 4.61328125,
+      "learning_rate": 8.99733155436958e-05,
+      "loss": 1.4502,
+      "step": 1513
+    },
+    {
+      "epoch": 1.397645972767136,
+      "grad_norm": 6.20703125,
+      "learning_rate": 8.996664442961975e-05,
+      "loss": 1.6191,
+      "step": 1514
+    },
+    {
+      "epoch": 1.3985691207015924,
+      "grad_norm": 6.23046875,
+      "learning_rate": 8.99599733155437e-05,
+      "loss": 1.7415,
+      "step": 1515
+    },
+    {
+      "epoch": 1.399492268636049,
+      "grad_norm": 6.3359375,
+      "learning_rate": 8.995330220146765e-05,
+      "loss": 1.4988,
+      "step": 1516
+    },
+    {
+      "epoch": 1.4004154165705054,
+      "grad_norm": 5.875,
+      "learning_rate": 8.99466310873916e-05,
+      "loss": 1.4492,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4013385645049619,
+      "grad_norm": 7.28515625,
+      "learning_rate": 8.993995997331554e-05,
+      "loss": 1.5886,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4022617124394183,
+      "grad_norm": 5.265625,
+      "learning_rate": 8.99332888592395e-05,
+      "loss": 1.2694,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4031848603738748,
+      "grad_norm": 6.76953125,
+      "learning_rate": 8.992661774516345e-05,
+      "loss": 1.7893,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4041080083083315,
+      "grad_norm": 5.4375,
+      "learning_rate": 8.991994663108739e-05,
+      "loss": 1.5349,
+      "step": 1521
+    },
+    {
+      "epoch": 1.405031156242788,
+      "grad_norm": 4.57421875,
+      "learning_rate": 8.991327551701134e-05,
+      "loss": 1.3717,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4059543041772444,
+      "grad_norm": 4.7890625,
+      "learning_rate": 8.99066044029353e-05,
+      "loss": 1.323,
+      "step": 1523
+    },
+    {
+      "epoch": 1.406877452111701,
+      "grad_norm": 4.359375,
+      "learning_rate": 8.989993328885924e-05,
+      "loss": 1.3378,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4078006000461574,
+      "grad_norm": 4.75,
+      "learning_rate": 8.989326217478319e-05,
+      "loss": 1.5898,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4087237479806138,
+      "grad_norm": 4.86328125,
+      "learning_rate": 8.988659106070713e-05,
+      "loss": 1.51,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4096468959150703,
+      "grad_norm": 4.16796875,
+      "learning_rate": 8.987991994663109e-05,
+      "loss": 1.7393,
+      "step": 1527
+    },
+    {
+      "epoch": 1.410570043849527,
+      "grad_norm": 4.50390625,
+      "learning_rate": 8.987324883255505e-05,
+      "loss": 1.7595,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4114931917839835,
+      "grad_norm": 3.623046875,
+      "learning_rate": 8.9866577718479e-05,
+      "loss": 1.7317,
+      "step": 1529
+    },
+    {
+      "epoch": 1.41241633971844,
+      "grad_norm": 4.94140625,
+      "learning_rate": 8.985990660440294e-05,
+      "loss": 1.6565,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4133394876528964,
+      "grad_norm": 4.83203125,
+      "learning_rate": 8.985323549032689e-05,
+      "loss": 1.6052,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4142626355873529,
+      "grad_norm": 5.25390625,
+      "learning_rate": 8.984656437625085e-05,
+      "loss": 1.6887,
+      "step": 1532
+    },
+    {
+      "epoch": 1.4151857835218093,
+      "grad_norm": 4.546875,
+      "learning_rate": 8.983989326217479e-05,
+      "loss": 1.4277,
+      "step": 1533
+    },
+    {
+      "epoch": 1.4161089314562658,
+      "grad_norm": 9.671875,
+      "learning_rate": 8.983322214809874e-05,
+      "loss": 1.2357,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4170320793907223,
+      "grad_norm": 4.42578125,
+      "learning_rate": 8.982655103402268e-05,
+      "loss": 1.4636,
+      "step": 1535
+    },
+    {
+      "epoch": 1.4179552273251788,
+      "grad_norm": 3.66015625,
+      "learning_rate": 8.981987991994664e-05,
+      "loss": 1.4434,
+      "step": 1536
+    },
+    {
+      "epoch": 1.4188783752596352,
+      "grad_norm": 4.2109375,
+      "learning_rate": 8.981320880587059e-05,
+      "loss": 1.5515,
+      "step": 1537
+    },
+    {
+      "epoch": 1.419801523194092,
+      "grad_norm": 4.83984375,
+      "learning_rate": 8.980653769179453e-05,
+      "loss": 1.6262,
+      "step": 1538
+    },
+    {
+      "epoch": 1.4207246711285484,
+      "grad_norm": 5.80078125,
+      "learning_rate": 8.979986657771848e-05,
+      "loss": 1.5479,
+      "step": 1539
+    },
+    {
+      "epoch": 1.4216478190630049,
+      "grad_norm": 6.359375,
+      "learning_rate": 8.979319546364244e-05,
+      "loss": 1.8555,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4225709669974613,
+      "grad_norm": 4.3046875,
+      "learning_rate": 8.978652434956638e-05,
+      "loss": 1.4133,
+      "step": 1541
+    },
+    {
+      "epoch": 1.4234941149319178,
+      "grad_norm": 4.4296875,
+      "learning_rate": 8.977985323549033e-05,
+      "loss": 1.3116,
+      "step": 1542
+    },
+    {
+      "epoch": 1.4244172628663743,
+      "grad_norm": 3.82421875,
+      "learning_rate": 8.977318212141427e-05,
+      "loss": 1.2089,
+      "step": 1543
+    },
+    {
+      "epoch": 1.425340410800831,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.976651100733823e-05,
+      "loss": 1.3937,
+      "step": 1544
+    },
+    {
+      "epoch": 1.4262635587352874,
+      "grad_norm": 5.2890625,
+      "learning_rate": 8.975983989326218e-05,
+      "loss": 1.4387,
+      "step": 1545
+    },
+    {
+      "epoch": 1.427186706669744,
+      "grad_norm": 4.50390625,
+      "learning_rate": 8.975316877918612e-05,
+      "loss": 1.4497,
+      "step": 1546
+    },
+    {
+      "epoch": 1.4281098546042004,
+      "grad_norm": 6.43359375,
+      "learning_rate": 8.974649766511007e-05,
+      "loss": 1.7957,
+      "step": 1547
+    },
+    {
+      "epoch": 1.4290330025386568,
+      "grad_norm": 7.10546875,
+      "learning_rate": 8.973982655103403e-05,
+      "loss": 1.6055,
+      "step": 1548
+    },
+    {
+      "epoch": 1.4299561504731133,
+      "grad_norm": 5.48828125,
+      "learning_rate": 8.973315543695797e-05,
+      "loss": 1.6433,
+      "step": 1549
+    },
+    {
+      "epoch": 1.4308792984075698,
+      "grad_norm": 5.21484375,
+      "learning_rate": 8.972648432288192e-05,
+      "loss": 1.4299,
+      "step": 1550
+    },
+    {
+      "epoch": 1.4318024463420262,
+      "grad_norm": 4.80859375,
+      "learning_rate": 8.971981320880586e-05,
+      "loss": 1.7317,
+      "step": 1551
+    },
+    {
+      "epoch": 1.4327255942764827,
+      "grad_norm": 4.08984375,
+      "learning_rate": 8.971314209472984e-05,
+      "loss": 1.4468,
+      "step": 1552
+    },
+    {
+      "epoch": 1.4336487422109392,
+      "grad_norm": 5.26171875,
+      "learning_rate": 8.970647098065378e-05,
+      "loss": 1.6589,
+      "step": 1553
+    },
+    {
+      "epoch": 1.4345718901453959,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.969979986657773e-05,
+      "loss": 1.3787,
+      "step": 1554
+    },
+    {
+      "epoch": 1.4354950380798523,
+      "grad_norm": 4.40625,
+      "learning_rate": 8.969312875250167e-05,
+      "loss": 1.5503,
+      "step": 1555
+    },
+    {
+      "epoch": 1.4364181860143088,
+      "grad_norm": 6.59765625,
+      "learning_rate": 8.968645763842563e-05,
+      "loss": 1.5969,
+      "step": 1556
+    },
+    {
+      "epoch": 1.4373413339487653,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.967978652434958e-05,
+      "loss": 1.5455,
+      "step": 1557
+    },
+    {
+      "epoch": 1.4382644818832218,
+      "grad_norm": 5.4609375,
+      "learning_rate": 8.967311541027352e-05,
+      "loss": 1.6235,
+      "step": 1558
+    },
+    {
+      "epoch": 1.4391876298176782,
+      "grad_norm": 4.734375,
+      "learning_rate": 8.966644429619747e-05,
+      "loss": 1.5605,
+      "step": 1559
+    },
+    {
+      "epoch": 1.4401107777521347,
+      "grad_norm": 4.25390625,
+      "learning_rate": 8.965977318212143e-05,
+      "loss": 1.5615,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4410339256865914,
+      "grad_norm": 4.234375,
+      "learning_rate": 8.965310206804537e-05,
+      "loss": 1.541,
+      "step": 1561
+    },
+    {
+      "epoch": 1.4419570736210479,
+      "grad_norm": 3.779296875,
+      "learning_rate": 8.964643095396932e-05,
+      "loss": 1.3233,
+      "step": 1562
+    },
+    {
+      "epoch": 1.4428802215555043,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.963975983989326e-05,
+      "loss": 1.5137,
+      "step": 1563
+    },
+    {
+      "epoch": 1.4438033694899608,
+      "grad_norm": 5.68359375,
+      "learning_rate": 8.963308872581722e-05,
+      "loss": 1.6584,
+      "step": 1564
+    },
+    {
+      "epoch": 1.4447265174244173,
+      "grad_norm": 4.41015625,
+      "learning_rate": 8.962641761174117e-05,
+      "loss": 1.7148,
+      "step": 1565
+    },
+    {
+      "epoch": 1.4456496653588737,
+      "grad_norm": 4.42578125,
+      "learning_rate": 8.961974649766511e-05,
+      "loss": 1.4131,
+      "step": 1566
+    },
+    {
+      "epoch": 1.4465728132933302,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.961307538358906e-05,
+      "loss": 1.6167,
+      "step": 1567
+    },
+    {
+      "epoch": 1.4474959612277867,
+      "grad_norm": 5.65234375,
+      "learning_rate": 8.960640426951302e-05,
+      "loss": 1.5378,
+      "step": 1568
+    },
+    {
+      "epoch": 1.4484191091622431,
+      "grad_norm": 4.82421875,
+      "learning_rate": 8.959973315543696e-05,
+      "loss": 1.4897,
+      "step": 1569
+    },
+    {
+      "epoch": 1.4493422570966996,
+      "grad_norm": 6.0859375,
+      "learning_rate": 8.959306204136091e-05,
+      "loss": 1.5618,
+      "step": 1570
+    },
+    {
+      "epoch": 1.4502654050311563,
+      "grad_norm": 5.1875,
+      "learning_rate": 8.958639092728485e-05,
+      "loss": 1.7239,
+      "step": 1571
+    },
+    {
+      "epoch": 1.4511885529656128,
+      "grad_norm": 5.48828125,
+      "learning_rate": 8.957971981320881e-05,
+      "loss": 1.877,
+      "step": 1572
+    },
+    {
+      "epoch": 1.4521117009000692,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.957304869913276e-05,
+      "loss": 1.2004,
+      "step": 1573
+    },
+    {
+      "epoch": 1.4530348488345257,
+      "grad_norm": 6.21875,
+      "learning_rate": 8.95663775850567e-05,
+      "loss": 1.3977,
+      "step": 1574
+    },
+    {
+      "epoch": 1.4539579967689822,
+      "grad_norm": 4.3203125,
+      "learning_rate": 8.955970647098065e-05,
+      "loss": 1.6887,
+      "step": 1575
+    },
+    {
+      "epoch": 1.4548811447034387,
+      "grad_norm": 5.27734375,
+      "learning_rate": 8.955303535690461e-05,
+      "loss": 1.6743,
+      "step": 1576
+    },
+    {
+      "epoch": 1.4558042926378953,
+      "grad_norm": 5.06640625,
+      "learning_rate": 8.954636424282857e-05,
+      "loss": 1.2323,
+      "step": 1577
+    },
+    {
+      "epoch": 1.4567274405723518,
+      "grad_norm": 5.01953125,
+      "learning_rate": 8.953969312875251e-05,
+      "loss": 1.7185,
+      "step": 1578
+    },
+    {
+      "epoch": 1.4576505885068083,
+      "grad_norm": 5.8828125,
+      "learning_rate": 8.953302201467646e-05,
+      "loss": 1.5806,
+      "step": 1579
+    },
+    {
+      "epoch": 1.4585737364412648,
+      "grad_norm": 5.12890625,
+      "learning_rate": 8.95263509006004e-05,
+      "loss": 1.527,
+      "step": 1580
+    },
+    {
+      "epoch": 1.4594968843757212,
+      "grad_norm": 5.60546875,
+      "learning_rate": 8.951967978652436e-05,
+      "loss": 1.3378,
+      "step": 1581
+    },
+    {
+      "epoch": 1.4604200323101777,
+      "grad_norm": 4.90234375,
+      "learning_rate": 8.95130086724483e-05,
+      "loss": 1.481,
+      "step": 1582
+    },
+    {
+      "epoch": 1.4613431802446342,
+      "grad_norm": 4.390625,
+      "learning_rate": 8.950633755837225e-05,
+      "loss": 1.3628,
+      "step": 1583
+    },
+    {
+      "epoch": 1.4622663281790906,
+      "grad_norm": 4.7890625,
+      "learning_rate": 8.94996664442962e-05,
+      "loss": 1.6199,
+      "step": 1584
+    },
+    {
+      "epoch": 1.463189476113547,
+      "grad_norm": 5.61328125,
+      "learning_rate": 8.949299533022016e-05,
+      "loss": 1.6536,
+      "step": 1585
+    },
+    {
+      "epoch": 1.4641126240480036,
+      "grad_norm": 5.80859375,
+      "learning_rate": 8.94863242161441e-05,
+      "loss": 1.5669,
+      "step": 1586
+    },
+    {
+      "epoch": 1.4650357719824603,
+      "grad_norm": 4.4921875,
+      "learning_rate": 8.947965310206805e-05,
+      "loss": 1.5955,
+      "step": 1587
+    },
+    {
+      "epoch": 1.4659589199169167,
+      "grad_norm": 5.3359375,
+      "learning_rate": 8.947298198799199e-05,
+      "loss": 1.7417,
+      "step": 1588
+    },
+    {
+      "epoch": 1.4668820678513732,
+      "grad_norm": 4.74609375,
+      "learning_rate": 8.946631087391595e-05,
+      "loss": 1.2981,
+      "step": 1589
+    },
+    {
+      "epoch": 1.4678052157858297,
+      "grad_norm": 6.05859375,
+      "learning_rate": 8.94596397598399e-05,
+      "loss": 1.6062,
+      "step": 1590
+    },
+    {
+      "epoch": 1.4687283637202861,
+      "grad_norm": 3.830078125,
+      "learning_rate": 8.945296864576384e-05,
+      "loss": 1.2984,
+      "step": 1591
+    },
+    {
+      "epoch": 1.4696515116547426,
+      "grad_norm": 4.01953125,
+      "learning_rate": 8.94462975316878e-05,
+      "loss": 1.5767,
+      "step": 1592
+    },
+    {
+      "epoch": 1.4705746595891993,
+      "grad_norm": 4.37890625,
+      "learning_rate": 8.943962641761175e-05,
+      "loss": 1.6592,
+      "step": 1593
+    },
+    {
+      "epoch": 1.4714978075236558,
+      "grad_norm": 4.83984375,
+      "learning_rate": 8.943295530353569e-05,
+      "loss": 1.6765,
+      "step": 1594
+    },
+    {
+      "epoch": 1.4724209554581122,
+      "grad_norm": 5.19921875,
+      "learning_rate": 8.942628418945964e-05,
+      "loss": 1.6077,
+      "step": 1595
+    },
+    {
+      "epoch": 1.4733441033925687,
+      "grad_norm": 5.26171875,
+      "learning_rate": 8.94196130753836e-05,
+      "loss": 1.6265,
+      "step": 1596
+    },
+    {
+      "epoch": 1.4742672513270252,
+      "grad_norm": 5.67578125,
+      "learning_rate": 8.941294196130754e-05,
+      "loss": 1.5464,
+      "step": 1597
+    },
+    {
+      "epoch": 1.4751903992614817,
+      "grad_norm": 4.6015625,
+      "learning_rate": 8.940627084723149e-05,
+      "loss": 1.6479,
+      "step": 1598
+    },
+    {
+      "epoch": 1.4761135471959381,
+      "grad_norm": 5.42578125,
+      "learning_rate": 8.939959973315543e-05,
+      "loss": 1.543,
+      "step": 1599
+    },
+    {
+      "epoch": 1.4770366951303946,
+      "grad_norm": 4.97265625,
+      "learning_rate": 8.939292861907939e-05,
+      "loss": 1.4207,
+      "step": 1600
+    },
+    {
+      "epoch": 1.477959843064851,
+      "grad_norm": 4.3359375,
+      "learning_rate": 8.938625750500334e-05,
+      "loss": 1.2529,
+      "step": 1601
+    },
+    {
+      "epoch": 1.4788829909993075,
+      "grad_norm": 4.62109375,
+      "learning_rate": 8.93795863909273e-05,
+      "loss": 1.4501,
+      "step": 1602
+    },
+    {
+      "epoch": 1.479806138933764,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.937291527685124e-05,
+      "loss": 1.6589,
+      "step": 1603
+    },
+    {
+      "epoch": 1.4807292868682207,
+      "grad_norm": 4.58203125,
+      "learning_rate": 8.936624416277519e-05,
+      "loss": 1.2875,
+      "step": 1604
+    },
+    {
+      "epoch": 1.4816524348026772,
+      "grad_norm": 4.359375,
+      "learning_rate": 8.935957304869915e-05,
+      "loss": 1.7136,
+      "step": 1605
+    },
+    {
+      "epoch": 1.4825755827371336,
+      "grad_norm": 5.1796875,
+      "learning_rate": 8.935290193462309e-05,
+      "loss": 1.6973,
+      "step": 1606
+    },
+    {
+      "epoch": 1.48349873067159,
+      "grad_norm": 4.3828125,
+      "learning_rate": 8.934623082054704e-05,
+      "loss": 1.5884,
+      "step": 1607
+    },
+    {
+      "epoch": 1.4844218786060466,
+      "grad_norm": 5.078125,
+      "learning_rate": 8.933955970647098e-05,
+      "loss": 1.5854,
+      "step": 1608
+    },
+    {
+      "epoch": 1.485345026540503,
+      "grad_norm": 5.328125,
+      "learning_rate": 8.933288859239494e-05,
+      "loss": 1.3263,
+      "step": 1609
+    },
+    {
+      "epoch": 1.4862681744749597,
+      "grad_norm": 5.55078125,
+      "learning_rate": 8.932621747831889e-05,
+      "loss": 1.6199,
+      "step": 1610
+    },
+    {
+      "epoch": 1.4871913224094162,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.931954636424283e-05,
+      "loss": 1.5679,
+      "step": 1611
+    },
+    {
+      "epoch": 1.4881144703438727,
+      "grad_norm": 5.67578125,
+      "learning_rate": 8.931287525016678e-05,
+      "loss": 1.5709,
+      "step": 1612
+    },
+    {
+      "epoch": 1.4890376182783291,
+      "grad_norm": 6.3203125,
+      "learning_rate": 8.930620413609074e-05,
+      "loss": 1.7849,
+      "step": 1613
+    },
+    {
+      "epoch": 1.4899607662127856,
+      "grad_norm": 5.31640625,
+      "learning_rate": 8.929953302201468e-05,
+      "loss": 1.4177,
+      "step": 1614
+    },
+    {
+      "epoch": 1.490883914147242,
+      "grad_norm": 5.73828125,
+      "learning_rate": 8.929286190793863e-05,
+      "loss": 1.6208,
+      "step": 1615
+    },
+    {
+      "epoch": 1.4918070620816986,
+      "grad_norm": 5.44921875,
+      "learning_rate": 8.928619079386257e-05,
+      "loss": 1.7869,
+      "step": 1616
+    },
+    {
+      "epoch": 1.492730210016155,
+      "grad_norm": 6.58984375,
+      "learning_rate": 8.927951967978653e-05,
+      "loss": 1.5776,
+      "step": 1617
+    },
+    {
+      "epoch": 1.4936533579506115,
+      "grad_norm": 6.61328125,
+      "learning_rate": 8.927284856571048e-05,
+      "loss": 1.7261,
+      "step": 1618
+    },
+    {
+      "epoch": 1.494576505885068,
+      "grad_norm": 4.75,
+      "learning_rate": 8.926617745163442e-05,
+      "loss": 1.3481,
+      "step": 1619
+    },
+    {
+      "epoch": 1.4954996538195247,
+      "grad_norm": 6.58984375,
+      "learning_rate": 8.925950633755837e-05,
+      "loss": 1.4353,
+      "step": 1620
+    },
+    {
+      "epoch": 1.4964228017539811,
+      "grad_norm": 4.3984375,
+      "learning_rate": 8.925283522348233e-05,
+      "loss": 1.5076,
+      "step": 1621
+    },
+    {
+      "epoch": 1.4973459496884376,
+      "grad_norm": 6.65234375,
+      "learning_rate": 8.924616410940627e-05,
+      "loss": 1.6372,
+      "step": 1622
+    },
+    {
+      "epoch": 1.498269097622894,
+      "grad_norm": 5.22265625,
+      "learning_rate": 8.923949299533022e-05,
+      "loss": 1.4663,
+      "step": 1623
+    },
+    {
+      "epoch": 1.4991922455573505,
+      "grad_norm": 6.0234375,
+      "learning_rate": 8.923282188125416e-05,
+      "loss": 1.5872,
+      "step": 1624
+    },
+    {
+      "epoch": 1.5001153934918072,
+      "grad_norm": 6.28125,
+      "learning_rate": 8.922615076717812e-05,
+      "loss": 1.1062,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5010385414262637,
+      "grad_norm": 5.37109375,
+      "learning_rate": 8.921947965310207e-05,
+      "loss": 1.4434,
+      "step": 1626
+    },
+    {
+      "epoch": 1.5019616893607202,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.921280853902603e-05,
+      "loss": 1.5808,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5028848372951766,
+      "grad_norm": 4.54296875,
+      "learning_rate": 8.920613742494997e-05,
+      "loss": 1.6736,
+      "step": 1628
+    },
+    {
+      "epoch": 1.503807985229633,
+      "grad_norm": 4.2734375,
+      "learning_rate": 8.919946631087392e-05,
+      "loss": 1.4097,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5047311331640896,
+      "grad_norm": 4.19140625,
+      "learning_rate": 8.919279519679788e-05,
+      "loss": 1.5798,
+      "step": 1630
+    },
+    {
+      "epoch": 1.505654281098546,
+      "grad_norm": 4.55078125,
+      "learning_rate": 8.918612408272182e-05,
+      "loss": 1.5703,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5065774290330025,
+      "grad_norm": 5.23046875,
+      "learning_rate": 8.917945296864577e-05,
+      "loss": 1.408,
+      "step": 1632
+    },
+    {
+      "epoch": 1.507500576967459,
+      "grad_norm": 5.01953125,
+      "learning_rate": 8.917278185456973e-05,
+      "loss": 1.52,
+      "step": 1633
+    },
+    {
+      "epoch": 1.5084237249019155,
+      "grad_norm": 6.5234375,
+      "learning_rate": 8.916611074049367e-05,
+      "loss": 1.6458,
+      "step": 1634
+    },
+    {
+      "epoch": 1.509346872836372,
+      "grad_norm": 5.00390625,
+      "learning_rate": 8.915943962641762e-05,
+      "loss": 1.6729,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5102700207708284,
+      "grad_norm": 6.3984375,
+      "learning_rate": 8.915276851234156e-05,
+      "loss": 1.4227,
+      "step": 1636
+    },
+    {
+      "epoch": 1.5111931687052849,
+      "grad_norm": 5.37890625,
+      "learning_rate": 8.914609739826552e-05,
+      "loss": 1.4182,
+      "step": 1637
+    },
+    {
+      "epoch": 1.5121163166397416,
+      "grad_norm": 6.6796875,
+      "learning_rate": 8.913942628418947e-05,
+      "loss": 1.5012,
+      "step": 1638
+    },
+    {
+      "epoch": 1.513039464574198,
+      "grad_norm": 5.45703125,
+      "learning_rate": 8.913275517011341e-05,
+      "loss": 1.4641,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5139626125086545,
+      "grad_norm": 4.31640625,
+      "learning_rate": 8.912608405603736e-05,
+      "loss": 1.4546,
+      "step": 1640
+    },
+    {
+      "epoch": 1.514885760443111,
+      "grad_norm": 3.7578125,
+      "learning_rate": 8.911941294196132e-05,
+      "loss": 1.5461,
+      "step": 1641
+    },
+    {
+      "epoch": 1.5158089083775677,
+      "grad_norm": 4.8828125,
+      "learning_rate": 8.911274182788526e-05,
+      "loss": 1.8225,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5167320563120241,
+      "grad_norm": 4.41796875,
+      "learning_rate": 8.910607071380921e-05,
+      "loss": 1.2091,
+      "step": 1643
+    },
+    {
+      "epoch": 1.5176552042464806,
+      "grad_norm": 6.13671875,
+      "learning_rate": 8.909939959973315e-05,
+      "loss": 1.4376,
+      "step": 1644
+    },
+    {
+      "epoch": 1.518578352180937,
+      "grad_norm": 4.1484375,
+      "learning_rate": 8.909272848565711e-05,
+      "loss": 1.4823,
+      "step": 1645
+    },
+    {
+      "epoch": 1.5195015001153935,
+      "grad_norm": 5.1640625,
+      "learning_rate": 8.908605737158106e-05,
+      "loss": 1.5842,
+      "step": 1646
+    },
+    {
+      "epoch": 1.52042464804985,
+      "grad_norm": 4.671875,
+      "learning_rate": 8.9079386257505e-05,
+      "loss": 1.3035,
+      "step": 1647
+    },
+    {
+      "epoch": 1.5213477959843065,
+      "grad_norm": 5.34765625,
+      "learning_rate": 8.907271514342895e-05,
+      "loss": 1.3528,
+      "step": 1648
+    },
+    {
+      "epoch": 1.522270943918763,
+      "grad_norm": 5.5703125,
+      "learning_rate": 8.906604402935291e-05,
+      "loss": 1.5171,
+      "step": 1649
+    },
+    {
+      "epoch": 1.5231940918532194,
+      "grad_norm": 4.78515625,
+      "learning_rate": 8.905937291527685e-05,
+      "loss": 1.0854,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5241172397876759,
+      "grad_norm": 4.5234375,
+      "learning_rate": 8.90527018012008e-05,
+      "loss": 1.3817,
+      "step": 1651
+    },
+    {
+      "epoch": 1.5250403877221324,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.904603068712476e-05,
+      "loss": 1.5398,
+      "step": 1652
+    },
+    {
+      "epoch": 1.5259635356565888,
+      "grad_norm": 4.55078125,
+      "learning_rate": 8.90393595730487e-05,
+      "loss": 1.4724,
+      "step": 1653
+    },
+    {
+      "epoch": 1.5268866835910455,
+      "grad_norm": 6.0390625,
+      "learning_rate": 8.903268845897266e-05,
+      "loss": 1.5979,
+      "step": 1654
+    },
+    {
+      "epoch": 1.527809831525502,
+      "grad_norm": 4.63671875,
+      "learning_rate": 8.902601734489661e-05,
+      "loss": 1.3206,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5287329794599585,
+      "grad_norm": 5.296875,
+      "learning_rate": 8.901934623082055e-05,
+      "loss": 1.7578,
+      "step": 1656
+    },
+    {
+      "epoch": 1.529656127394415,
+      "grad_norm": 5.30859375,
+      "learning_rate": 8.90126751167445e-05,
+      "loss": 1.4369,
+      "step": 1657
+    },
+    {
+      "epoch": 1.5305792753288716,
+      "grad_norm": 5.55078125,
+      "learning_rate": 8.900600400266846e-05,
+      "loss": 1.5066,
+      "step": 1658
+    },
+    {
+      "epoch": 1.531502423263328,
+      "grad_norm": 5.23828125,
+      "learning_rate": 8.89993328885924e-05,
+      "loss": 1.4478,
+      "step": 1659
+    },
+    {
+      "epoch": 1.5324255711977846,
+      "grad_norm": 5.30078125,
+      "learning_rate": 8.899266177451635e-05,
+      "loss": 1.5691,
+      "step": 1660
+    },
+    {
+      "epoch": 1.533348719132241,
+      "grad_norm": 3.857421875,
+      "learning_rate": 8.89859906604403e-05,
+      "loss": 1.5522,
+      "step": 1661
+    },
+    {
+      "epoch": 1.5342718670666975,
+      "grad_norm": 3.548828125,
+      "learning_rate": 8.897931954636425e-05,
+      "loss": 1.5078,
+      "step": 1662
+    },
+    {
+      "epoch": 1.535195015001154,
+      "grad_norm": 4.20703125,
+      "learning_rate": 8.89726484322882e-05,
+      "loss": 1.3296,
+      "step": 1663
+    },
+    {
+      "epoch": 1.5361181629356104,
+      "grad_norm": 4.56640625,
+      "learning_rate": 8.896597731821214e-05,
+      "loss": 1.741,
+      "step": 1664
+    },
+    {
+      "epoch": 1.537041310870067,
+      "grad_norm": 6.4609375,
+      "learning_rate": 8.895930620413609e-05,
+      "loss": 1.3635,
+      "step": 1665
+    },
+    {
+      "epoch": 1.5379644588045234,
+      "grad_norm": 6.48828125,
+      "learning_rate": 8.895263509006005e-05,
+      "loss": 1.4268,
+      "step": 1666
+    },
+    {
+      "epoch": 1.5388876067389798,
+      "grad_norm": 4.19921875,
+      "learning_rate": 8.8945963975984e-05,
+      "loss": 1.6067,
+      "step": 1667
+    },
+    {
+      "epoch": 1.5398107546734363,
+      "grad_norm": 5.49609375,
+      "learning_rate": 8.893929286190794e-05,
+      "loss": 1.3923,
+      "step": 1668
+    },
+    {
+      "epoch": 1.5407339026078928,
+      "grad_norm": 5.015625,
+      "learning_rate": 8.893262174783188e-05,
+      "loss": 1.4585,
+      "step": 1669
+    },
+    {
+      "epoch": 1.5416570505423492,
+      "grad_norm": 4.9609375,
+      "learning_rate": 8.892595063375584e-05,
+      "loss": 1.6057,
+      "step": 1670
+    },
+    {
+      "epoch": 1.542580198476806,
+      "grad_norm": 4.50390625,
+      "learning_rate": 8.891927951967979e-05,
+      "loss": 1.47,
+      "step": 1671
+    },
+    {
+      "epoch": 1.5435033464112624,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.891260840560373e-05,
+      "loss": 1.1947,
+      "step": 1672
+    },
+    {
+      "epoch": 1.5444264943457189,
+      "grad_norm": 4.82421875,
+      "learning_rate": 8.890593729152768e-05,
+      "loss": 1.6082,
+      "step": 1673
+    },
+    {
+      "epoch": 1.5453496422801754,
+      "grad_norm": 4.73828125,
+      "learning_rate": 8.889926617745164e-05,
+      "loss": 1.5492,
+      "step": 1674
+    },
+    {
+      "epoch": 1.546272790214632,
+      "grad_norm": 5.61328125,
+      "learning_rate": 8.889259506337558e-05,
+      "loss": 1.813,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5471959381490885,
+      "grad_norm": 7.26171875,
+      "learning_rate": 8.888592394929953e-05,
+      "loss": 1.5945,
+      "step": 1676
+    },
+    {
+      "epoch": 1.548119086083545,
+      "grad_norm": 5.3515625,
+      "learning_rate": 8.887925283522349e-05,
+      "loss": 1.5546,
+      "step": 1677
+    },
+    {
+      "epoch": 1.5490422340180015,
+      "grad_norm": 5.0234375,
+      "learning_rate": 8.887258172114745e-05,
+      "loss": 1.572,
+      "step": 1678
+    },
+    {
+      "epoch": 1.549965381952458,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.886591060707139e-05,
+      "loss": 1.0556,
+      "step": 1679
+    },
+    {
+      "epoch": 1.5508885298869144,
+      "grad_norm": 4.765625,
+      "learning_rate": 8.885923949299534e-05,
+      "loss": 1.6829,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5518116778213709,
+      "grad_norm": 4.2421875,
+      "learning_rate": 8.885256837891928e-05,
+      "loss": 1.3682,
+      "step": 1681
+    },
+    {
+      "epoch": 1.5527348257558273,
+      "grad_norm": 6.47265625,
+      "learning_rate": 8.884589726484324e-05,
+      "loss": 1.5249,
+      "step": 1682
+    },
+    {
+      "epoch": 1.5536579736902838,
+      "grad_norm": 4.2578125,
+      "learning_rate": 8.883922615076719e-05,
+      "loss": 1.5239,
+      "step": 1683
+    },
+    {
+      "epoch": 1.5545811216247403,
+      "grad_norm": 4.45703125,
+      "learning_rate": 8.883255503669113e-05,
+      "loss": 1.7612,
+      "step": 1684
+    },
+    {
+      "epoch": 1.5555042695591967,
+      "grad_norm": 4.7578125,
+      "learning_rate": 8.882588392261508e-05,
+      "loss": 1.5447,
+      "step": 1685
+    },
+    {
+      "epoch": 1.5564274174936532,
+      "grad_norm": 6.31640625,
+      "learning_rate": 8.881921280853904e-05,
+      "loss": 1.6035,
+      "step": 1686
+    },
+    {
+      "epoch": 1.55735056542811,
+      "grad_norm": 4.50390625,
+      "learning_rate": 8.881254169446298e-05,
+      "loss": 1.6565,
+      "step": 1687
+    },
+    {
+      "epoch": 1.5582737133625664,
+      "grad_norm": 6.64453125,
+      "learning_rate": 8.880587058038693e-05,
+      "loss": 1.7493,
+      "step": 1688
+    },
+    {
+      "epoch": 1.5591968612970228,
+      "grad_norm": 5.5546875,
+      "learning_rate": 8.879919946631087e-05,
+      "loss": 1.6655,
+      "step": 1689
+    },
+    {
+      "epoch": 1.5601200092314793,
+      "grad_norm": 6.25,
+      "learning_rate": 8.879252835223483e-05,
+      "loss": 1.4026,
+      "step": 1690
+    },
+    {
+      "epoch": 1.561043157165936,
+      "grad_norm": 4.859375,
+      "learning_rate": 8.878585723815878e-05,
+      "loss": 1.5055,
+      "step": 1691
+    },
+    {
+      "epoch": 1.5619663051003925,
+      "grad_norm": 4.1484375,
+      "learning_rate": 8.877918612408272e-05,
+      "loss": 1.501,
+      "step": 1692
+    },
+    {
+      "epoch": 1.562889453034849,
+      "grad_norm": 3.6953125,
+      "learning_rate": 8.877251501000667e-05,
+      "loss": 1.5161,
+      "step": 1693
+    },
+    {
+      "epoch": 1.5638126009693054,
+      "grad_norm": 4.88671875,
+      "learning_rate": 8.876584389593063e-05,
+      "loss": 1.7153,
+      "step": 1694
+    },
+    {
+      "epoch": 1.5647357489037619,
+      "grad_norm": 5.76171875,
+      "learning_rate": 8.875917278185457e-05,
+      "loss": 1.3585,
+      "step": 1695
+    },
+    {
+      "epoch": 1.5656588968382184,
+      "grad_norm": 5.828125,
+      "learning_rate": 8.875250166777852e-05,
+      "loss": 1.5703,
+      "step": 1696
+    },
+    {
+      "epoch": 1.5665820447726748,
+      "grad_norm": 5.80078125,
+      "learning_rate": 8.874583055370246e-05,
+      "loss": 1.6404,
+      "step": 1697
+    },
+    {
+      "epoch": 1.5675051927071313,
+      "grad_norm": 6.04296875,
+      "learning_rate": 8.873915943962642e-05,
+      "loss": 1.5637,
+      "step": 1698
+    },
+    {
+      "epoch": 1.5684283406415878,
+      "grad_norm": 5.94921875,
+      "learning_rate": 8.873248832555037e-05,
+      "loss": 1.4105,
+      "step": 1699
+    },
+    {
+      "epoch": 1.5693514885760442,
+      "grad_norm": 5.22265625,
+      "learning_rate": 8.872581721147431e-05,
+      "loss": 1.7124,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5702746365105007,
+      "grad_norm": 3.9609375,
+      "learning_rate": 8.871914609739826e-05,
+      "loss": 1.5076,
+      "step": 1701
+    },
+    {
+      "epoch": 1.5711977844449572,
+      "grad_norm": 4.20703125,
+      "learning_rate": 8.871247498332222e-05,
+      "loss": 1.2972,
+      "step": 1702
+    },
+    {
+      "epoch": 1.5721209323794136,
+      "grad_norm": 5.44921875,
+      "learning_rate": 8.870580386924618e-05,
+      "loss": 1.8735,
+      "step": 1703
+    },
+    {
+      "epoch": 1.5730440803138703,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.869913275517012e-05,
+      "loss": 1.4905,
+      "step": 1704
+    },
+    {
+      "epoch": 1.5739672282483268,
+      "grad_norm": 4.51953125,
+      "learning_rate": 8.869246164109407e-05,
+      "loss": 1.4404,
+      "step": 1705
+    },
+    {
+      "epoch": 1.5748903761827833,
+      "grad_norm": 5.703125,
+      "learning_rate": 8.868579052701801e-05,
+      "loss": 1.5686,
+      "step": 1706
+    },
+    {
+      "epoch": 1.5758135241172397,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.867911941294197e-05,
+      "loss": 1.6921,
+      "step": 1707
+    },
+    {
+      "epoch": 1.5767366720516964,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.867244829886592e-05,
+      "loss": 1.4579,
+      "step": 1708
+    },
+    {
+      "epoch": 1.577659819986153,
+      "grad_norm": 5.22265625,
+      "learning_rate": 8.866577718478986e-05,
+      "loss": 1.894,
+      "step": 1709
+    },
+    {
+      "epoch": 1.5785829679206094,
+      "grad_norm": 5.01953125,
+      "learning_rate": 8.865910607071381e-05,
+      "loss": 1.6326,
+      "step": 1710
+    },
+    {
+      "epoch": 1.5795061158550658,
+      "grad_norm": 5.0234375,
+      "learning_rate": 8.865243495663777e-05,
+      "loss": 1.6213,
+      "step": 1711
+    },
+    {
+      "epoch": 1.5804292637895223,
+      "grad_norm": 3.931640625,
+      "learning_rate": 8.864576384256171e-05,
+      "loss": 1.1993,
+      "step": 1712
+    },
+    {
+      "epoch": 1.5813524117239788,
+      "grad_norm": 4.796875,
+      "learning_rate": 8.863909272848566e-05,
+      "loss": 1.5955,
+      "step": 1713
+    },
+    {
+      "epoch": 1.5822755596584352,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.86324216144096e-05,
+      "loss": 1.896,
+      "step": 1714
+    },
+    {
+      "epoch": 1.5831987075928917,
+      "grad_norm": 6.33203125,
+      "learning_rate": 8.862575050033356e-05,
+      "loss": 1.4543,
+      "step": 1715
+    },
+    {
+      "epoch": 1.5841218555273482,
+      "grad_norm": 6.21875,
+      "learning_rate": 8.861907938625751e-05,
+      "loss": 1.3813,
+      "step": 1716
+    },
+    {
+      "epoch": 1.5850450034618047,
+      "grad_norm": 5.42578125,
+      "learning_rate": 8.861240827218145e-05,
+      "loss": 1.5627,
+      "step": 1717
+    },
+    {
+      "epoch": 1.5859681513962611,
+      "grad_norm": 5.08203125,
+      "learning_rate": 8.86057371581054e-05,
+      "loss": 1.3127,
+      "step": 1718
+    },
+    {
+      "epoch": 1.5868912993307176,
+      "grad_norm": 5.1953125,
+      "learning_rate": 8.859906604402936e-05,
+      "loss": 1.5449,
+      "step": 1719
+    },
+    {
+      "epoch": 1.5878144472651743,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.85923949299533e-05,
+      "loss": 1.2366,
+      "step": 1720
+    },
+    {
+      "epoch": 1.5887375951996308,
+      "grad_norm": 4.46484375,
+      "learning_rate": 8.858572381587725e-05,
+      "loss": 1.0721,
+      "step": 1721
+    },
+    {
+      "epoch": 1.5896607431340872,
+      "grad_norm": 4.921875,
+      "learning_rate": 8.85790527018012e-05,
+      "loss": 1.5596,
+      "step": 1722
+    },
+    {
+      "epoch": 1.5905838910685437,
+      "grad_norm": 3.740234375,
+      "learning_rate": 8.857238158772515e-05,
+      "loss": 1.0237,
+      "step": 1723
+    },
+    {
+      "epoch": 1.5915070390030004,
+      "grad_norm": 4.515625,
+      "learning_rate": 8.85657104736491e-05,
+      "loss": 1.6675,
+      "step": 1724
+    },
+    {
+      "epoch": 1.5924301869374569,
+      "grad_norm": 4.6015625,
+      "learning_rate": 8.855903935957305e-05,
+      "loss": 1.3711,
+      "step": 1725
+    },
+    {
+      "epoch": 1.5933533348719133,
+      "grad_norm": 4.7421875,
+      "learning_rate": 8.855236824549699e-05,
+      "loss": 1.3669,
+      "step": 1726
+    },
+    {
+      "epoch": 1.5942764828063698,
+      "grad_norm": 6.3203125,
+      "learning_rate": 8.854569713142095e-05,
+      "loss": 1.4453,
+      "step": 1727
+    },
+    {
+      "epoch": 1.5951996307408263,
+      "grad_norm": 6.4765625,
+      "learning_rate": 8.853902601734491e-05,
+      "loss": 1.509,
+      "step": 1728
+    },
+    {
+      "epoch": 1.5961227786752827,
+      "grad_norm": 5.91015625,
+      "learning_rate": 8.853235490326885e-05,
+      "loss": 1.8184,
+      "step": 1729
+    },
+    {
+      "epoch": 1.5970459266097392,
+      "grad_norm": 4.453125,
+      "learning_rate": 8.85256837891928e-05,
+      "loss": 1.3717,
+      "step": 1730
+    },
+    {
+      "epoch": 1.5979690745441957,
+      "grad_norm": 3.904296875,
+      "learning_rate": 8.851901267511676e-05,
+      "loss": 1.4727,
+      "step": 1731
+    },
+    {
+      "epoch": 1.5988922224786521,
+      "grad_norm": 4.7578125,
+      "learning_rate": 8.85123415610407e-05,
+      "loss": 1.6992,
+      "step": 1732
+    },
+    {
+      "epoch": 1.5998153704131086,
+      "grad_norm": 5.40234375,
+      "learning_rate": 8.850567044696465e-05,
+      "loss": 1.4802,
+      "step": 1733
+    },
+    {
+      "epoch": 1.600738518347565,
+      "grad_norm": 3.900390625,
+      "learning_rate": 8.84989993328886e-05,
+      "loss": 1.4138,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6016616662820216,
+      "grad_norm": 4.6640625,
+      "learning_rate": 8.849232821881255e-05,
+      "loss": 1.293,
+      "step": 1735
+    },
+    {
+      "epoch": 1.602584814216478,
+      "grad_norm": 4.5078125,
+      "learning_rate": 8.84856571047365e-05,
+      "loss": 1.4412,
+      "step": 1736
+    },
+    {
+      "epoch": 1.6035079621509347,
+      "grad_norm": 5.18359375,
+      "learning_rate": 8.847898599066044e-05,
+      "loss": 1.5945,
+      "step": 1737
+    },
+    {
+      "epoch": 1.6044311100853912,
+      "grad_norm": 3.87109375,
+      "learning_rate": 8.847231487658439e-05,
+      "loss": 1.5408,
+      "step": 1738
+    },
+    {
+      "epoch": 1.6053542580198477,
+      "grad_norm": 4.16015625,
+      "learning_rate": 8.846564376250835e-05,
+      "loss": 1.5195,
+      "step": 1739
+    },
+    {
+      "epoch": 1.6062774059543041,
+      "grad_norm": 5.39453125,
+      "learning_rate": 8.84589726484323e-05,
+      "loss": 1.2479,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6072005538887608,
+      "grad_norm": 5.64453125,
+      "learning_rate": 8.845230153435624e-05,
+      "loss": 1.4521,
+      "step": 1741
+    },
+    {
+      "epoch": 1.6081237018232173,
+      "grad_norm": 4.98046875,
+      "learning_rate": 8.844563042028019e-05,
+      "loss": 1.4919,
+      "step": 1742
+    },
+    {
+      "epoch": 1.6090468497576738,
+      "grad_norm": 3.65234375,
+      "learning_rate": 8.843895930620414e-05,
+      "loss": 1.3817,
+      "step": 1743
+    },
+    {
+      "epoch": 1.6099699976921302,
+      "grad_norm": 4.36328125,
+      "learning_rate": 8.843228819212809e-05,
+      "loss": 1.3457,
+      "step": 1744
+    },
+    {
+      "epoch": 1.6108931456265867,
+      "grad_norm": 4.47265625,
+      "learning_rate": 8.842561707805204e-05,
+      "loss": 1.3671,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6118162935610432,
+      "grad_norm": 5.23828125,
+      "learning_rate": 8.841894596397598e-05,
+      "loss": 1.687,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6127394414954996,
+      "grad_norm": 4.67578125,
+      "learning_rate": 8.841227484989994e-05,
+      "loss": 1.6365,
+      "step": 1747
+    },
+    {
+      "epoch": 1.613662589429956,
+      "grad_norm": 6.55078125,
+      "learning_rate": 8.840560373582389e-05,
+      "loss": 1.668,
+      "step": 1748
+    },
+    {
+      "epoch": 1.6145857373644126,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.839893262174783e-05,
+      "loss": 1.5486,
+      "step": 1749
+    },
+    {
+      "epoch": 1.615508885298869,
+      "grad_norm": 5.58203125,
+      "learning_rate": 8.839226150767178e-05,
+      "loss": 1.2413,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6164320332333255,
+      "grad_norm": 5.60546875,
+      "learning_rate": 8.838559039359573e-05,
+      "loss": 1.6616,
+      "step": 1751
+    },
+    {
+      "epoch": 1.617355181167782,
+      "grad_norm": 5.72265625,
+      "learning_rate": 8.837891927951968e-05,
+      "loss": 1.2502,
+      "step": 1752
+    },
+    {
+      "epoch": 1.6182783291022387,
+      "grad_norm": 5.6015625,
+      "learning_rate": 8.837224816544364e-05,
+      "loss": 1.3336,
+      "step": 1753
+    },
+    {
+      "epoch": 1.6192014770366951,
+      "grad_norm": 4.6328125,
+      "learning_rate": 8.836557705136758e-05,
+      "loss": 1.5073,
+      "step": 1754
+    },
+    {
+      "epoch": 1.6201246249711516,
+      "grad_norm": 5.36328125,
+      "learning_rate": 8.835890593729153e-05,
+      "loss": 1.5522,
+      "step": 1755
+    },
+    {
+      "epoch": 1.621047772905608,
+      "grad_norm": 5.0859375,
+      "learning_rate": 8.835223482321549e-05,
+      "loss": 1.4749,
+      "step": 1756
+    },
+    {
+      "epoch": 1.6219709208400648,
+      "grad_norm": 6.140625,
+      "learning_rate": 8.834556370913943e-05,
+      "loss": 1.4082,
+      "step": 1757
+    },
+    {
+      "epoch": 1.6228940687745212,
+      "grad_norm": 5.98828125,
+      "learning_rate": 8.833889259506338e-05,
+      "loss": 1.5901,
+      "step": 1758
+    },
+    {
+      "epoch": 1.6238172167089777,
+      "grad_norm": 5.2578125,
+      "learning_rate": 8.833222148098733e-05,
+      "loss": 1.6968,
+      "step": 1759
+    },
+    {
+      "epoch": 1.6247403646434342,
+      "grad_norm": 5.171875,
+      "learning_rate": 8.832555036691128e-05,
+      "loss": 1.6152,
+      "step": 1760
+    },
+    {
+      "epoch": 1.6256635125778907,
+      "grad_norm": 6.25,
+      "learning_rate": 8.831887925283523e-05,
+      "loss": 1.5044,
+      "step": 1761
+    },
+    {
+      "epoch": 1.6265866605123471,
+      "grad_norm": 4.89453125,
+      "learning_rate": 8.831220813875918e-05,
+      "loss": 1.2483,
+      "step": 1762
+    },
+    {
+      "epoch": 1.6275098084468036,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.830553702468312e-05,
+      "loss": 1.415,
+      "step": 1763
+    },
+    {
+      "epoch": 1.62843295638126,
+      "grad_norm": 3.625,
+      "learning_rate": 8.829886591060708e-05,
+      "loss": 1.624,
+      "step": 1764
+    },
+    {
+      "epoch": 1.6293561043157165,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.829219479653103e-05,
+      "loss": 1.2507,
+      "step": 1765
+    },
+    {
+      "epoch": 1.630279252250173,
+      "grad_norm": 5.625,
+      "learning_rate": 8.828552368245497e-05,
+      "loss": 1.1002,
+      "step": 1766
+    },
+    {
+      "epoch": 1.6312024001846295,
+      "grad_norm": 4.53515625,
+      "learning_rate": 8.827885256837892e-05,
+      "loss": 1.6143,
+      "step": 1767
+    },
+    {
+      "epoch": 1.632125548119086,
+      "grad_norm": 4.890625,
+      "learning_rate": 8.827218145430288e-05,
+      "loss": 1.2435,
+      "step": 1768
+    },
+    {
+      "epoch": 1.6330486960535424,
+      "grad_norm": 4.30078125,
+      "learning_rate": 8.826551034022682e-05,
+      "loss": 1.7815,
+      "step": 1769
+    },
+    {
+      "epoch": 1.633971843987999,
+      "grad_norm": 4.8046875,
+      "learning_rate": 8.825883922615077e-05,
+      "loss": 1.4517,
+      "step": 1770
+    },
+    {
+      "epoch": 1.6348949919224556,
+      "grad_norm": 6.14453125,
+      "learning_rate": 8.825216811207471e-05,
+      "loss": 1.8604,
+      "step": 1771
+    },
+    {
+      "epoch": 1.635818139856912,
+      "grad_norm": 5.40234375,
+      "learning_rate": 8.824549699799867e-05,
+      "loss": 1.2853,
+      "step": 1772
+    },
+    {
+      "epoch": 1.6367412877913685,
+      "grad_norm": 5.18359375,
+      "learning_rate": 8.823882588392262e-05,
+      "loss": 1.4856,
+      "step": 1773
+    },
+    {
+      "epoch": 1.6376644357258252,
+      "grad_norm": 4.80078125,
+      "learning_rate": 8.823215476984656e-05,
+      "loss": 1.182,
+      "step": 1774
+    },
+    {
+      "epoch": 1.6385875836602817,
+      "grad_norm": 5.2890625,
+      "learning_rate": 8.82254836557705e-05,
+      "loss": 1.5088,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6395107315947381,
+      "grad_norm": 4.51953125,
+      "learning_rate": 8.821881254169447e-05,
+      "loss": 1.4958,
+      "step": 1776
+    },
+    {
+      "epoch": 1.6404338795291946,
+      "grad_norm": 5.2265625,
+      "learning_rate": 8.821214142761841e-05,
+      "loss": 1.5508,
+      "step": 1777
+    },
+    {
+      "epoch": 1.641357027463651,
+      "grad_norm": 5.140625,
+      "learning_rate": 8.820547031354237e-05,
+      "loss": 1.4741,
+      "step": 1778
+    },
+    {
+      "epoch": 1.6422801753981076,
+      "grad_norm": 4.640625,
+      "learning_rate": 8.819879919946632e-05,
+      "loss": 1.5364,
+      "step": 1779
+    },
+    {
+      "epoch": 1.643203323332564,
+      "grad_norm": 4.99609375,
+      "learning_rate": 8.819212808539027e-05,
+      "loss": 1.6765,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6441264712670205,
+      "grad_norm": 4.3671875,
+      "learning_rate": 8.818545697131422e-05,
+      "loss": 1.4695,
+      "step": 1781
+    },
+    {
+      "epoch": 1.645049619201477,
+      "grad_norm": 4.17578125,
+      "learning_rate": 8.817878585723817e-05,
+      "loss": 1.5564,
+      "step": 1782
+    },
+    {
+      "epoch": 1.6459727671359334,
+      "grad_norm": 4.875,
+      "learning_rate": 8.817211474316211e-05,
+      "loss": 1.708,
+      "step": 1783
+    },
+    {
+      "epoch": 1.64689591507039,
+      "grad_norm": 4.53515625,
+      "learning_rate": 8.816544362908607e-05,
+      "loss": 1.384,
+      "step": 1784
+    },
+    {
+      "epoch": 1.6478190630048464,
+      "grad_norm": 5.6171875,
+      "learning_rate": 8.815877251501002e-05,
+      "loss": 1.5508,
+      "step": 1785
+    },
+    {
+      "epoch": 1.648742210939303,
+      "grad_norm": 5.9765625,
+      "learning_rate": 8.815210140093396e-05,
+      "loss": 1.3604,
+      "step": 1786
+    },
+    {
+      "epoch": 1.6496653588737595,
+      "grad_norm": 5.38671875,
+      "learning_rate": 8.81454302868579e-05,
+      "loss": 1.3442,
+      "step": 1787
+    },
+    {
+      "epoch": 1.650588506808216,
+      "grad_norm": 5.625,
+      "learning_rate": 8.813875917278186e-05,
+      "loss": 1.3782,
+      "step": 1788
+    },
+    {
+      "epoch": 1.6515116547426725,
+      "grad_norm": 5.38671875,
+      "learning_rate": 8.813208805870581e-05,
+      "loss": 1.5251,
+      "step": 1789
+    },
+    {
+      "epoch": 1.6524348026771292,
+      "grad_norm": 5.234375,
+      "learning_rate": 8.812541694462976e-05,
+      "loss": 1.5117,
+      "step": 1790
+    },
+    {
+      "epoch": 1.6533579506115856,
+      "grad_norm": 5.1015625,
+      "learning_rate": 8.81187458305537e-05,
+      "loss": 1.624,
+      "step": 1791
+    },
+    {
+      "epoch": 1.654281098546042,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.811207471647766e-05,
+      "loss": 1.2148,
+      "step": 1792
+    },
+    {
+      "epoch": 1.6552042464804986,
+      "grad_norm": 4.46484375,
+      "learning_rate": 8.81054036024016e-05,
+      "loss": 1.6946,
+      "step": 1793
+    },
+    {
+      "epoch": 1.656127394414955,
+      "grad_norm": 4.359375,
+      "learning_rate": 8.809873248832555e-05,
+      "loss": 1.145,
+      "step": 1794
+    },
+    {
+      "epoch": 1.6570505423494115,
+      "grad_norm": 6.59765625,
+      "learning_rate": 8.80920613742495e-05,
+      "loss": 1.455,
+      "step": 1795
+    },
+    {
+      "epoch": 1.657973690283868,
+      "grad_norm": 5.1328125,
+      "learning_rate": 8.808539026017346e-05,
+      "loss": 1.4851,
+      "step": 1796
+    },
+    {
+      "epoch": 1.6588968382183245,
+      "grad_norm": 4.2734375,
+      "learning_rate": 8.80787191460974e-05,
+      "loss": 1.6389,
+      "step": 1797
+    },
+    {
+      "epoch": 1.659819986152781,
+      "grad_norm": 4.3515625,
+      "learning_rate": 8.807204803202135e-05,
+      "loss": 1.5994,
+      "step": 1798
+    },
+    {
+      "epoch": 1.6607431340872374,
+      "grad_norm": 3.916015625,
+      "learning_rate": 8.806537691794529e-05,
+      "loss": 1.6079,
+      "step": 1799
+    },
+    {
+      "epoch": 1.6616662820216939,
+      "grad_norm": 5.5546875,
+      "learning_rate": 8.805870580386925e-05,
+      "loss": 1.116,
+      "step": 1800
+    },
+    {
+      "epoch": 1.6625894299561503,
+      "grad_norm": 6.6328125,
+      "learning_rate": 8.80520346897932e-05,
+      "loss": 1.5334,
+      "step": 1801
+    },
+    {
+      "epoch": 1.663512577890607,
+      "grad_norm": 5.203125,
+      "learning_rate": 8.804536357571714e-05,
+      "loss": 1.0701,
+      "step": 1802
+    },
+    {
+      "epoch": 1.6644357258250635,
+      "grad_norm": 5.97265625,
+      "learning_rate": 8.80386924616411e-05,
+      "loss": 1.5916,
+      "step": 1803
+    },
+    {
+      "epoch": 1.66535887375952,
+      "grad_norm": 7.33984375,
+      "learning_rate": 8.803202134756505e-05,
+      "loss": 1.752,
+      "step": 1804
+    },
+    {
+      "epoch": 1.6662820216939764,
+      "grad_norm": 5.90234375,
+      "learning_rate": 8.8025350233489e-05,
+      "loss": 1.1888,
+      "step": 1805
+    },
+    {
+      "epoch": 1.667205169628433,
+      "grad_norm": 6.16015625,
+      "learning_rate": 8.801867911941295e-05,
+      "loss": 1.3972,
+      "step": 1806
+    },
+    {
+      "epoch": 1.6681283175628896,
+      "grad_norm": 4.73046875,
+      "learning_rate": 8.80120080053369e-05,
+      "loss": 1.5356,
+      "step": 1807
+    },
+    {
+      "epoch": 1.669051465497346,
+      "grad_norm": 5.13671875,
+      "learning_rate": 8.800533689126084e-05,
+      "loss": 1.418,
+      "step": 1808
+    },
+    {
+      "epoch": 1.6699746134318025,
+      "grad_norm": 4.671875,
+      "learning_rate": 8.79986657771848e-05,
+      "loss": 1.2343,
+      "step": 1809
+    },
+    {
+      "epoch": 1.670897761366259,
+      "grad_norm": 6.39453125,
+      "learning_rate": 8.799199466310875e-05,
+      "loss": 1.384,
+      "step": 1810
+    },
+    {
+      "epoch": 1.6718209093007155,
+      "grad_norm": 5.4609375,
+      "learning_rate": 8.798532354903269e-05,
+      "loss": 1.5928,
+      "step": 1811
+    },
+    {
+      "epoch": 1.672744057235172,
+      "grad_norm": 4.74609375,
+      "learning_rate": 8.797865243495664e-05,
+      "loss": 1.5227,
+      "step": 1812
+    },
+    {
+      "epoch": 1.6736672051696284,
+      "grad_norm": 4.99609375,
+      "learning_rate": 8.79719813208806e-05,
+      "loss": 1.6514,
+      "step": 1813
+    },
+    {
+      "epoch": 1.6745903531040849,
+      "grad_norm": 5.4296875,
+      "learning_rate": 8.796531020680454e-05,
+      "loss": 1.6687,
+      "step": 1814
+    },
+    {
+      "epoch": 1.6755135010385414,
+      "grad_norm": 6.3125,
+      "learning_rate": 8.795863909272849e-05,
+      "loss": 1.6331,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6764366489729978,
+      "grad_norm": 4.9921875,
+      "learning_rate": 8.795196797865243e-05,
+      "loss": 1.6226,
+      "step": 1816
+    },
+    {
+      "epoch": 1.6773597969074543,
+      "grad_norm": 5.59765625,
+      "learning_rate": 8.794529686457639e-05,
+      "loss": 1.9407,
+      "step": 1817
+    },
+    {
+      "epoch": 1.6782829448419108,
+      "grad_norm": 5.66796875,
+      "learning_rate": 8.793862575050034e-05,
+      "loss": 1.4258,
+      "step": 1818
+    },
+    {
+      "epoch": 1.6792060927763675,
+      "grad_norm": 4.203125,
+      "learning_rate": 8.793195463642428e-05,
+      "loss": 1.4072,
+      "step": 1819
+    },
+    {
+      "epoch": 1.680129240710824,
+      "grad_norm": 4.84765625,
+      "learning_rate": 8.792528352234824e-05,
+      "loss": 1.7031,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6810523886452804,
+      "grad_norm": 5.765625,
+      "learning_rate": 8.791861240827219e-05,
+      "loss": 1.469,
+      "step": 1821
+    },
+    {
+      "epoch": 1.6819755365797369,
+      "grad_norm": 5.2734375,
+      "learning_rate": 8.791194129419613e-05,
+      "loss": 1.448,
+      "step": 1822
+    },
+    {
+      "epoch": 1.6828986845141936,
+      "grad_norm": 4.23046875,
+      "learning_rate": 8.790527018012008e-05,
+      "loss": 1.4127,
+      "step": 1823
+    },
+    {
+      "epoch": 1.68382183244865,
+      "grad_norm": 4.75390625,
+      "learning_rate": 8.789859906604404e-05,
+      "loss": 1.429,
+      "step": 1824
+    },
+    {
+      "epoch": 1.6847449803831065,
+      "grad_norm": 4.2578125,
+      "learning_rate": 8.789192795196798e-05,
+      "loss": 1.6965,
+      "step": 1825
+    },
+    {
+      "epoch": 1.685668128317563,
+      "grad_norm": 4.671875,
+      "learning_rate": 8.788525683789193e-05,
+      "loss": 1.5496,
+      "step": 1826
+    },
+    {
+      "epoch": 1.6865912762520194,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.787858572381587e-05,
+      "loss": 1.5869,
+      "step": 1827
+    },
+    {
+      "epoch": 1.687514424186476,
+      "grad_norm": 5.359375,
+      "learning_rate": 8.787191460973983e-05,
+      "loss": 1.6008,
+      "step": 1828
+    },
+    {
+      "epoch": 1.6884375721209324,
+      "grad_norm": 5.20703125,
+      "learning_rate": 8.786524349566379e-05,
+      "loss": 1.4297,
+      "step": 1829
+    },
+    {
+      "epoch": 1.6893607200553888,
+      "grad_norm": 5.1875,
+      "learning_rate": 8.785857238158774e-05,
+      "loss": 1.6265,
+      "step": 1830
+    },
+    {
+      "epoch": 1.6902838679898453,
+      "grad_norm": 5.42578125,
+      "learning_rate": 8.785190126751168e-05,
+      "loss": 1.5959,
+      "step": 1831
+    },
+    {
+      "epoch": 1.6912070159243018,
+      "grad_norm": 6.17578125,
+      "learning_rate": 8.784523015343563e-05,
+      "loss": 1.7766,
+      "step": 1832
+    },
+    {
+      "epoch": 1.6921301638587583,
+      "grad_norm": 5.328125,
+      "learning_rate": 8.783855903935959e-05,
+      "loss": 1.4441,
+      "step": 1833
+    },
+    {
+      "epoch": 1.6930533117932147,
+      "grad_norm": 5.37890625,
+      "learning_rate": 8.783188792528353e-05,
+      "loss": 1.3608,
+      "step": 1834
+    },
+    {
+      "epoch": 1.6939764597276714,
+      "grad_norm": 4.34765625,
+      "learning_rate": 8.782521681120748e-05,
+      "loss": 1.5117,
+      "step": 1835
+    },
+    {
+      "epoch": 1.6948996076621279,
+      "grad_norm": 4.62890625,
+      "learning_rate": 8.781854569713142e-05,
+      "loss": 1.5408,
+      "step": 1836
+    },
+    {
+      "epoch": 1.6958227555965844,
+      "grad_norm": 4.83203125,
+      "learning_rate": 8.781187458305538e-05,
+      "loss": 1.1581,
+      "step": 1837
+    },
+    {
+      "epoch": 1.6967459035310408,
+      "grad_norm": 5.1484375,
+      "learning_rate": 8.780520346897933e-05,
+      "loss": 1.5076,
+      "step": 1838
+    },
+    {
+      "epoch": 1.6976690514654973,
+      "grad_norm": 3.84765625,
+      "learning_rate": 8.779853235490327e-05,
+      "loss": 1.6558,
+      "step": 1839
+    },
+    {
+      "epoch": 1.698592199399954,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.779186124082722e-05,
+      "loss": 1.6414,
+      "step": 1840
+    },
+    {
+      "epoch": 1.6995153473344105,
+      "grad_norm": 4.9140625,
+      "learning_rate": 8.778519012675118e-05,
+      "loss": 1.6025,
+      "step": 1841
+    },
+    {
+      "epoch": 1.700438495268867,
+      "grad_norm": 5.796875,
+      "learning_rate": 8.777851901267512e-05,
+      "loss": 1.453,
+      "step": 1842
+    },
+    {
+      "epoch": 1.7013616432033234,
+      "grad_norm": 5.33203125,
+      "learning_rate": 8.777184789859907e-05,
+      "loss": 1.6226,
+      "step": 1843
+    },
+    {
+      "epoch": 1.7022847911377799,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.776517678452301e-05,
+      "loss": 1.7344,
+      "step": 1844
+    },
+    {
+      "epoch": 1.7032079390722363,
+      "grad_norm": 4.2734375,
+      "learning_rate": 8.775850567044697e-05,
+      "loss": 1.6055,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7041310870066928,
+      "grad_norm": 5.4453125,
+      "learning_rate": 8.775183455637092e-05,
+      "loss": 1.2634,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7050542349411493,
+      "grad_norm": 4.48046875,
+      "learning_rate": 8.774516344229486e-05,
+      "loss": 1.1705,
+      "step": 1847
+    },
+    {
+      "epoch": 1.7059773828756057,
+      "grad_norm": 5.21484375,
+      "learning_rate": 8.773849232821881e-05,
+      "loss": 1.4675,
+      "step": 1848
+    },
+    {
+      "epoch": 1.7069005308100622,
+      "grad_norm": 4.91015625,
+      "learning_rate": 8.773182121414277e-05,
+      "loss": 1.4817,
+      "step": 1849
+    },
+    {
+      "epoch": 1.7078236787445187,
+      "grad_norm": 4.7421875,
+      "learning_rate": 8.772515010006671e-05,
+      "loss": 1.5737,
+      "step": 1850
+    },
+    {
+      "epoch": 1.7087468266789752,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.771847898599066e-05,
+      "loss": 1.5715,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7096699746134318,
+      "grad_norm": 4.33984375,
+      "learning_rate": 8.77118078719146e-05,
+      "loss": 1.3645,
+      "step": 1852
+    },
+    {
+      "epoch": 1.7105931225478883,
+      "grad_norm": 5.84375,
+      "learning_rate": 8.770513675783856e-05,
+      "loss": 1.6523,
+      "step": 1853
+    },
+    {
+      "epoch": 1.7115162704823448,
+      "grad_norm": 5.94140625,
+      "learning_rate": 8.769846564376252e-05,
+      "loss": 1.5122,
+      "step": 1854
+    },
+    {
+      "epoch": 1.7124394184168013,
+      "grad_norm": 5.0546875,
+      "learning_rate": 8.769179452968647e-05,
+      "loss": 1.6467,
+      "step": 1855
+    },
+    {
+      "epoch": 1.713362566351258,
+      "grad_norm": 4.875,
+      "learning_rate": 8.768512341561041e-05,
+      "loss": 1.0255,
+      "step": 1856
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 4.73828125,
+      "learning_rate": 8.767845230153436e-05,
+      "loss": 1.1176,
+      "step": 1857
+    },
+    {
+      "epoch": 1.7152088622201709,
+      "grad_norm": 4.4296875,
+      "learning_rate": 8.767178118745832e-05,
+      "loss": 1.3953,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7161320101546274,
+      "grad_norm": 4.93359375,
+      "learning_rate": 8.766511007338226e-05,
+      "loss": 1.6001,
+      "step": 1859
+    },
+    {
+      "epoch": 1.7170551580890838,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.765843895930621e-05,
+      "loss": 1.5591,
+      "step": 1860
+    },
+    {
+      "epoch": 1.7179783060235403,
+      "grad_norm": 3.693359375,
+      "learning_rate": 8.765176784523017e-05,
+      "loss": 1.7979,
+      "step": 1861
+    },
+    {
+      "epoch": 1.7189014539579968,
+      "grad_norm": 4.16796875,
+      "learning_rate": 8.764509673115411e-05,
+      "loss": 1.793,
+      "step": 1862
+    },
+    {
+      "epoch": 1.7198246018924532,
+      "grad_norm": 4.828125,
+      "learning_rate": 8.763842561707806e-05,
+      "loss": 1.4348,
+      "step": 1863
+    },
+    {
+      "epoch": 1.7207477498269097,
+      "grad_norm": 5.2578125,
+      "learning_rate": 8.7631754503002e-05,
+      "loss": 1.2255,
+      "step": 1864
+    },
+    {
+      "epoch": 1.7216708977613662,
+      "grad_norm": 4.6484375,
+      "learning_rate": 8.762508338892596e-05,
+      "loss": 1.6372,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7225940456958226,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.76184122748499e-05,
+      "loss": 1.353,
+      "step": 1866
+    },
+    {
+      "epoch": 1.723517193630279,
+      "grad_norm": 6.0859375,
+      "learning_rate": 8.761174116077385e-05,
+      "loss": 1.5227,
+      "step": 1867
+    },
+    {
+      "epoch": 1.7244403415647358,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.76050700466978e-05,
+      "loss": 1.4839,
+      "step": 1868
+    },
+    {
+      "epoch": 1.7253634894991923,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.759839893262176e-05,
+      "loss": 1.5215,
+      "step": 1869
+    },
+    {
+      "epoch": 1.7262866374336487,
+      "grad_norm": 4.8515625,
+      "learning_rate": 8.75917278185457e-05,
+      "loss": 1.6687,
+      "step": 1870
+    },
+    {
+      "epoch": 1.7272097853681052,
+      "grad_norm": 5.17578125,
+      "learning_rate": 8.758505670446965e-05,
+      "loss": 1.4155,
+      "step": 1871
+    },
+    {
+      "epoch": 1.7281329333025617,
+      "grad_norm": 5.3203125,
+      "learning_rate": 8.757838559039359e-05,
+      "loss": 1.6382,
+      "step": 1872
+    },
+    {
+      "epoch": 1.7290560812370184,
+      "grad_norm": 4.01953125,
+      "learning_rate": 8.757171447631755e-05,
+      "loss": 1.6692,
+      "step": 1873
+    },
+    {
+      "epoch": 1.7299792291714748,
+      "grad_norm": 3.791015625,
+      "learning_rate": 8.75650433622415e-05,
+      "loss": 1.5024,
+      "step": 1874
+    },
+    {
+      "epoch": 1.7309023771059313,
+      "grad_norm": 4.453125,
+      "learning_rate": 8.755837224816544e-05,
+      "loss": 1.7522,
+      "step": 1875
+    },
+    {
+      "epoch": 1.7318255250403878,
+      "grad_norm": 6.0625,
+      "learning_rate": 8.755170113408939e-05,
+      "loss": 1.6404,
+      "step": 1876
+    },
+    {
+      "epoch": 1.7327486729748443,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.754503002001335e-05,
+      "loss": 1.4822,
+      "step": 1877
+    },
+    {
+      "epoch": 1.7336718209093007,
+      "grad_norm": 4.796875,
+      "learning_rate": 8.75383589059373e-05,
+      "loss": 1.5115,
+      "step": 1878
+    },
+    {
+      "epoch": 1.7345949688437572,
+      "grad_norm": 5.4609375,
+      "learning_rate": 8.753168779186125e-05,
+      "loss": 1.5159,
+      "step": 1879
+    },
+    {
+      "epoch": 1.7355181167782137,
+      "grad_norm": 5.96484375,
+      "learning_rate": 8.75250166777852e-05,
+      "loss": 1.4814,
+      "step": 1880
+    },
+    {
+      "epoch": 1.7364412647126701,
+      "grad_norm": 4.65625,
+      "learning_rate": 8.751834556370914e-05,
+      "loss": 1.016,
+      "step": 1881
+    },
+    {
+      "epoch": 1.7373644126471266,
+      "grad_norm": 5.59765625,
+      "learning_rate": 8.75116744496331e-05,
+      "loss": 1.3928,
+      "step": 1882
+    },
+    {
+      "epoch": 1.738287560581583,
+      "grad_norm": 6.6328125,
+      "learning_rate": 8.750500333555705e-05,
+      "loss": 1.4678,
+      "step": 1883
+    },
+    {
+      "epoch": 1.7392107085160395,
+      "grad_norm": 5.73828125,
+      "learning_rate": 8.749833222148099e-05,
+      "loss": 1.4932,
+      "step": 1884
+    },
+    {
+      "epoch": 1.7401338564504962,
+      "grad_norm": 5.125,
+      "learning_rate": 8.749166110740494e-05,
+      "loss": 1.6182,
+      "step": 1885
+    },
+    {
+      "epoch": 1.7410570043849527,
+      "grad_norm": 4.75,
+      "learning_rate": 8.74849899933289e-05,
+      "loss": 1.8672,
+      "step": 1886
+    },
+    {
+      "epoch": 1.7419801523194092,
+      "grad_norm": 5.37109375,
+      "learning_rate": 8.747831887925284e-05,
+      "loss": 1.4521,
+      "step": 1887
+    },
+    {
+      "epoch": 1.7429033002538656,
+      "grad_norm": 4.76171875,
+      "learning_rate": 8.747164776517679e-05,
+      "loss": 1.6028,
+      "step": 1888
+    },
+    {
+      "epoch": 1.7438264481883223,
+      "grad_norm": 4.75390625,
+      "learning_rate": 8.746497665110073e-05,
+      "loss": 1.5837,
+      "step": 1889
+    },
+    {
+      "epoch": 1.7447495961227788,
+      "grad_norm": 4.77734375,
+      "learning_rate": 8.745830553702469e-05,
+      "loss": 1.3542,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7456727440572353,
+      "grad_norm": 5.4453125,
+      "learning_rate": 8.745163442294864e-05,
+      "loss": 1.7068,
+      "step": 1891
+    },
+    {
+      "epoch": 1.7465958919916917,
+      "grad_norm": 4.3515625,
+      "learning_rate": 8.744496330887258e-05,
+      "loss": 1.6299,
+      "step": 1892
+    },
+    {
+      "epoch": 1.7475190399261482,
+      "grad_norm": 5.125,
+      "learning_rate": 8.743829219479653e-05,
+      "loss": 1.5874,
+      "step": 1893
+    },
+    {
+      "epoch": 1.7484421878606047,
+      "grad_norm": 4.015625,
+      "learning_rate": 8.743162108072049e-05,
+      "loss": 1.6335,
+      "step": 1894
+    },
+    {
+      "epoch": 1.7493653357950611,
+      "grad_norm": 4.6171875,
+      "learning_rate": 8.742494996664443e-05,
+      "loss": 1.6523,
+      "step": 1895
+    },
+    {
+      "epoch": 1.7502884837295176,
+      "grad_norm": 4.71484375,
+      "learning_rate": 8.741827885256838e-05,
+      "loss": 1.6572,
+      "step": 1896
+    },
+    {
+      "epoch": 1.751211631663974,
+      "grad_norm": 5.1796875,
+      "learning_rate": 8.741160773849232e-05,
+      "loss": 1.2283,
+      "step": 1897
+    },
+    {
+      "epoch": 1.7521347795984306,
+      "grad_norm": 5.41015625,
+      "learning_rate": 8.740493662441628e-05,
+      "loss": 1.3842,
+      "step": 1898
+    },
+    {
+      "epoch": 1.753057927532887,
+      "grad_norm": 4.4609375,
+      "learning_rate": 8.739826551034023e-05,
+      "loss": 1.2003,
+      "step": 1899
+    },
+    {
+      "epoch": 1.7539810754673435,
+      "grad_norm": 4.546875,
+      "learning_rate": 8.739159439626417e-05,
+      "loss": 1.6355,
+      "step": 1900
+    },
+    {
+      "epoch": 1.7549042234018002,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.738492328218812e-05,
+      "loss": 1.2771,
+      "step": 1901
+    },
+    {
+      "epoch": 1.7558273713362567,
+      "grad_norm": 4.7578125,
+      "learning_rate": 8.737825216811208e-05,
+      "loss": 1.4995,
+      "step": 1902
+    },
+    {
+      "epoch": 1.7567505192707131,
+      "grad_norm": 3.779296875,
+      "learning_rate": 8.737158105403604e-05,
+      "loss": 1.2032,
+      "step": 1903
+    },
+    {
+      "epoch": 1.7576736672051696,
+      "grad_norm": 4.3515625,
+      "learning_rate": 8.736490993995998e-05,
+      "loss": 1.6804,
+      "step": 1904
+    },
+    {
+      "epoch": 1.7585968151396263,
+      "grad_norm": 4.8671875,
+      "learning_rate": 8.735823882588393e-05,
+      "loss": 1.4873,
+      "step": 1905
+    },
+    {
+      "epoch": 1.7595199630740828,
+      "grad_norm": 4.81640625,
+      "learning_rate": 8.735156771180789e-05,
+      "loss": 1.34,
+      "step": 1906
+    },
+    {
+      "epoch": 1.7604431110085392,
+      "grad_norm": 6.14453125,
+      "learning_rate": 8.734489659773183e-05,
+      "loss": 1.7595,
+      "step": 1907
+    },
+    {
+      "epoch": 1.7613662589429957,
+      "grad_norm": 5.76953125,
+      "learning_rate": 8.733822548365578e-05,
+      "loss": 1.2703,
+      "step": 1908
+    },
+    {
+      "epoch": 1.7622894068774522,
+      "grad_norm": 4.94921875,
+      "learning_rate": 8.733155436957972e-05,
+      "loss": 1.2362,
+      "step": 1909
+    },
+    {
+      "epoch": 1.7632125548119086,
+      "grad_norm": 4.39453125,
+      "learning_rate": 8.732488325550368e-05,
+      "loss": 1.5435,
+      "step": 1910
+    },
+    {
+      "epoch": 1.764135702746365,
+      "grad_norm": 4.5,
+      "learning_rate": 8.731821214142763e-05,
+      "loss": 1.4938,
+      "step": 1911
+    },
+    {
+      "epoch": 1.7650588506808216,
+      "grad_norm": 4.1796875,
+      "learning_rate": 8.731154102735157e-05,
+      "loss": 1.675,
+      "step": 1912
+    },
+    {
+      "epoch": 1.765981998615278,
+      "grad_norm": 5.390625,
+      "learning_rate": 8.730486991327552e-05,
+      "loss": 1.5522,
+      "step": 1913
+    },
+    {
+      "epoch": 1.7669051465497345,
+      "grad_norm": 4.984375,
+      "learning_rate": 8.729819879919948e-05,
+      "loss": 1.5786,
+      "step": 1914
+    },
+    {
+      "epoch": 1.767828294484191,
+      "grad_norm": 6.1796875,
+      "learning_rate": 8.729152768512342e-05,
+      "loss": 1.6709,
+      "step": 1915
+    },
+    {
+      "epoch": 1.7687514424186475,
+      "grad_norm": 6.89453125,
+      "learning_rate": 8.728485657104737e-05,
+      "loss": 1.6692,
+      "step": 1916
+    },
+    {
+      "epoch": 1.769674590353104,
+      "grad_norm": 5.0078125,
+      "learning_rate": 8.727818545697131e-05,
+      "loss": 1.5549,
+      "step": 1917
+    },
+    {
+      "epoch": 1.7705977382875606,
+      "grad_norm": 5.87890625,
+      "learning_rate": 8.727151434289527e-05,
+      "loss": 1.7046,
+      "step": 1918
+    },
+    {
+      "epoch": 1.771520886222017,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.726484322881922e-05,
+      "loss": 1.5208,
+      "step": 1919
+    },
+    {
+      "epoch": 1.7724440341564736,
+      "grad_norm": 4.1796875,
+      "learning_rate": 8.725817211474316e-05,
+      "loss": 1.1234,
+      "step": 1920
+    },
+    {
+      "epoch": 1.77336718209093,
+      "grad_norm": 4.66015625,
+      "learning_rate": 8.725150100066711e-05,
+      "loss": 1.6108,
+      "step": 1921
+    },
+    {
+      "epoch": 1.7742903300253867,
+      "grad_norm": 4.93359375,
+      "learning_rate": 8.724482988659107e-05,
+      "loss": 1.4183,
+      "step": 1922
+    },
+    {
+      "epoch": 1.7752134779598432,
+      "grad_norm": 4.43359375,
+      "learning_rate": 8.723815877251501e-05,
+      "loss": 1.3871,
+      "step": 1923
+    },
+    {
+      "epoch": 1.7761366258942997,
+      "grad_norm": 5.3828125,
+      "learning_rate": 8.723148765843896e-05,
+      "loss": 1.4438,
+      "step": 1924
+    },
+    {
+      "epoch": 1.7770597738287561,
+      "grad_norm": 5.609375,
+      "learning_rate": 8.72248165443629e-05,
+      "loss": 1.4749,
+      "step": 1925
+    },
+    {
+      "epoch": 1.7779829217632126,
+      "grad_norm": 6.37109375,
+      "learning_rate": 8.721814543028686e-05,
+      "loss": 1.7646,
+      "step": 1926
+    },
+    {
+      "epoch": 1.778906069697669,
+      "grad_norm": 4.6015625,
+      "learning_rate": 8.721147431621081e-05,
+      "loss": 1.5186,
+      "step": 1927
+    },
+    {
+      "epoch": 1.7798292176321255,
+      "grad_norm": 5.38671875,
+      "learning_rate": 8.720480320213477e-05,
+      "loss": 1.7568,
+      "step": 1928
+    },
+    {
+      "epoch": 1.780752365566582,
+      "grad_norm": 6.06640625,
+      "learning_rate": 8.719813208805871e-05,
+      "loss": 1.6089,
+      "step": 1929
+    },
+    {
+      "epoch": 1.7816755135010385,
+      "grad_norm": 5.29296875,
+      "learning_rate": 8.719146097398266e-05,
+      "loss": 1.7681,
+      "step": 1930
+    },
+    {
+      "epoch": 1.782598661435495,
+      "grad_norm": 5.328125,
+      "learning_rate": 8.718478985990662e-05,
+      "loss": 1.301,
+      "step": 1931
+    },
+    {
+      "epoch": 1.7835218093699514,
+      "grad_norm": 4.7578125,
+      "learning_rate": 8.717811874583056e-05,
+      "loss": 1.6211,
+      "step": 1932
+    },
+    {
+      "epoch": 1.7844449573044079,
+      "grad_norm": 6.45703125,
+      "learning_rate": 8.717144763175451e-05,
+      "loss": 1.6348,
+      "step": 1933
+    },
+    {
+      "epoch": 1.7853681052388646,
+      "grad_norm": 5.41796875,
+      "learning_rate": 8.716477651767845e-05,
+      "loss": 1.8987,
+      "step": 1934
+    },
+    {
+      "epoch": 1.786291253173321,
+      "grad_norm": 5.3828125,
+      "learning_rate": 8.715810540360241e-05,
+      "loss": 1.5059,
+      "step": 1935
+    },
+    {
+      "epoch": 1.7872144011077775,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.715143428952636e-05,
+      "loss": 1.7268,
+      "step": 1936
+    },
+    {
+      "epoch": 1.788137549042234,
+      "grad_norm": 3.701171875,
+      "learning_rate": 8.71447631754503e-05,
+      "loss": 1.6631,
+      "step": 1937
+    },
+    {
+      "epoch": 1.7890606969766907,
+      "grad_norm": 3.78515625,
+      "learning_rate": 8.713809206137425e-05,
+      "loss": 1.448,
+      "step": 1938
+    },
+    {
+      "epoch": 1.7899838449111471,
+      "grad_norm": 4.68359375,
+      "learning_rate": 8.713142094729821e-05,
+      "loss": 1.644,
+      "step": 1939
+    },
+    {
+      "epoch": 1.7909069928456036,
+      "grad_norm": 4.32421875,
+      "learning_rate": 8.712474983322215e-05,
+      "loss": 1.677,
+      "step": 1940
+    },
+    {
+      "epoch": 1.79183014078006,
+      "grad_norm": 5.6796875,
+      "learning_rate": 8.71180787191461e-05,
+      "loss": 1.6843,
+      "step": 1941
+    },
+    {
+      "epoch": 1.7927532887145166,
+      "grad_norm": 6.26171875,
+      "learning_rate": 8.711140760507004e-05,
+      "loss": 1.4949,
+      "step": 1942
+    },
+    {
+      "epoch": 1.793676436648973,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.7104736490994e-05,
+      "loss": 1.467,
+      "step": 1943
+    },
+    {
+      "epoch": 1.7945995845834295,
+      "grad_norm": 4.88671875,
+      "learning_rate": 8.709806537691795e-05,
+      "loss": 0.8904,
+      "step": 1944
+    },
+    {
+      "epoch": 1.795522732517886,
+      "grad_norm": 6.25,
+      "learning_rate": 8.70913942628419e-05,
+      "loss": 1.708,
+      "step": 1945
+    },
+    {
+      "epoch": 1.7964458804523424,
+      "grad_norm": 4.90234375,
+      "learning_rate": 8.708472314876584e-05,
+      "loss": 1.3767,
+      "step": 1946
+    },
+    {
+      "epoch": 1.797369028386799,
+      "grad_norm": 5.3203125,
+      "learning_rate": 8.70780520346898e-05,
+      "loss": 1.6194,
+      "step": 1947
+    },
+    {
+      "epoch": 1.7982921763212554,
+      "grad_norm": 5.125,
+      "learning_rate": 8.707138092061374e-05,
+      "loss": 1.2546,
+      "step": 1948
+    },
+    {
+      "epoch": 1.7992153242557118,
+      "grad_norm": 5.7265625,
+      "learning_rate": 8.706470980653769e-05,
+      "loss": 1.5471,
+      "step": 1949
+    },
+    {
+      "epoch": 1.8001384721901683,
+      "grad_norm": 5.9296875,
+      "learning_rate": 8.705803869246163e-05,
+      "loss": 1.5725,
+      "step": 1950
+    },
+    {
+      "epoch": 1.801061620124625,
+      "grad_norm": 4.7109375,
+      "learning_rate": 8.70513675783856e-05,
+      "loss": 1.5317,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8019847680590815,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.704469646430954e-05,
+      "loss": 1.8169,
+      "step": 1952
+    },
+    {
+      "epoch": 1.802907915993538,
+      "grad_norm": 4.62890625,
+      "learning_rate": 8.70380253502335e-05,
+      "loss": 1.77,
+      "step": 1953
+    },
+    {
+      "epoch": 1.8038310639279944,
+      "grad_norm": 6.50390625,
+      "learning_rate": 8.703135423615744e-05,
+      "loss": 1.551,
+      "step": 1954
+    },
+    {
+      "epoch": 1.804754211862451,
+      "grad_norm": 4.6171875,
+      "learning_rate": 8.70246831220814e-05,
+      "loss": 1.4495,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8056773597969076,
+      "grad_norm": 4.65625,
+      "learning_rate": 8.701801200800535e-05,
+      "loss": 1.3745,
+      "step": 1956
+    },
+    {
+      "epoch": 1.806600507731364,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.70113408939293e-05,
+      "loss": 1.5942,
+      "step": 1957
+    },
+    {
+      "epoch": 1.8075236556658205,
+      "grad_norm": 5.140625,
+      "learning_rate": 8.700466977985324e-05,
+      "loss": 1.6057,
+      "step": 1958
+    },
+    {
+      "epoch": 1.808446803600277,
+      "grad_norm": 4.4765625,
+      "learning_rate": 8.69979986657772e-05,
+      "loss": 1.1013,
+      "step": 1959
+    },
+    {
+      "epoch": 1.8093699515347335,
+      "grad_norm": 5.2734375,
+      "learning_rate": 8.699132755170114e-05,
+      "loss": 1.6633,
+      "step": 1960
+    },
+    {
+      "epoch": 1.81029309946919,
+      "grad_norm": 3.7265625,
+      "learning_rate": 8.698465643762509e-05,
+      "loss": 1.5103,
+      "step": 1961
+    },
+    {
+      "epoch": 1.8112162474036464,
+      "grad_norm": 4.74609375,
+      "learning_rate": 8.697798532354903e-05,
+      "loss": 1.8555,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8121393953381029,
+      "grad_norm": 3.732421875,
+      "learning_rate": 8.697131420947299e-05,
+      "loss": 1.3558,
+      "step": 1963
+    },
+    {
+      "epoch": 1.8130625432725593,
+      "grad_norm": 4.66015625,
+      "learning_rate": 8.696464309539694e-05,
+      "loss": 1.6287,
+      "step": 1964
+    },
+    {
+      "epoch": 1.8139856912070158,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.695797198132088e-05,
+      "loss": 1.623,
+      "step": 1965
+    },
+    {
+      "epoch": 1.8149088391414723,
+      "grad_norm": 4.69921875,
+      "learning_rate": 8.695130086724483e-05,
+      "loss": 1.5369,
+      "step": 1966
+    },
+    {
+      "epoch": 1.815831987075929,
+      "grad_norm": 4.734375,
+      "learning_rate": 8.694462975316879e-05,
+      "loss": 1.478,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8167551350103854,
+      "grad_norm": 5.92578125,
+      "learning_rate": 8.693795863909273e-05,
+      "loss": 1.5637,
+      "step": 1968
+    },
+    {
+      "epoch": 1.817678282944842,
+      "grad_norm": 6.87890625,
+      "learning_rate": 8.693128752501668e-05,
+      "loss": 1.8057,
+      "step": 1969
+    },
+    {
+      "epoch": 1.8186014308792984,
+      "grad_norm": 5.26171875,
+      "learning_rate": 8.692461641094062e-05,
+      "loss": 1.3192,
+      "step": 1970
+    },
+    {
+      "epoch": 1.819524578813755,
+      "grad_norm": 6.20703125,
+      "learning_rate": 8.691794529686458e-05,
+      "loss": 1.4622,
+      "step": 1971
+    },
+    {
+      "epoch": 1.8204477267482115,
+      "grad_norm": 6.8515625,
+      "learning_rate": 8.691127418278853e-05,
+      "loss": 1.4573,
+      "step": 1972
+    },
+    {
+      "epoch": 1.821370874682668,
+      "grad_norm": 5.1875,
+      "learning_rate": 8.690460306871247e-05,
+      "loss": 1.4294,
+      "step": 1973
+    },
+    {
+      "epoch": 1.8222940226171245,
+      "grad_norm": 5.1640625,
+      "learning_rate": 8.689793195463642e-05,
+      "loss": 1.6606,
+      "step": 1974
+    },
+    {
+      "epoch": 1.823217170551581,
+      "grad_norm": 4.62109375,
+      "learning_rate": 8.689126084056038e-05,
+      "loss": 1.553,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8241403184860374,
+      "grad_norm": 4.2109375,
+      "learning_rate": 8.688458972648432e-05,
+      "loss": 1.5842,
+      "step": 1976
+    },
+    {
+      "epoch": 1.8250634664204939,
+      "grad_norm": 5.171875,
+      "learning_rate": 8.687791861240827e-05,
+      "loss": 1.3838,
+      "step": 1977
+    },
+    {
+      "epoch": 1.8259866143549504,
+      "grad_norm": 4.12890625,
+      "learning_rate": 8.687124749833223e-05,
+      "loss": 1.4796,
+      "step": 1978
+    },
+    {
+      "epoch": 1.8269097622894068,
+      "grad_norm": 4.32421875,
+      "learning_rate": 8.686457638425617e-05,
+      "loss": 1.719,
+      "step": 1979
+    },
+    {
+      "epoch": 1.8278329102238633,
+      "grad_norm": 4.26171875,
+      "learning_rate": 8.685790527018013e-05,
+      "loss": 1.6592,
+      "step": 1980
+    },
+    {
+      "epoch": 1.8287560581583198,
+      "grad_norm": 4.27734375,
+      "learning_rate": 8.685123415610408e-05,
+      "loss": 1.4336,
+      "step": 1981
+    },
+    {
+      "epoch": 1.8296792060927762,
+      "grad_norm": 5.05078125,
+      "learning_rate": 8.684456304202802e-05,
+      "loss": 1.5698,
+      "step": 1982
+    },
+    {
+      "epoch": 1.8306023540272327,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.683789192795197e-05,
+      "loss": 1.7744,
+      "step": 1983
+    },
+    {
+      "epoch": 1.8315255019616894,
+      "grad_norm": 3.943359375,
+      "learning_rate": 8.683122081387593e-05,
+      "loss": 1.8877,
+      "step": 1984
+    },
+    {
+      "epoch": 1.8324486498961459,
+      "grad_norm": 3.18359375,
+      "learning_rate": 8.682454969979987e-05,
+      "loss": 1.502,
+      "step": 1985
+    },
+    {
+      "epoch": 1.8333717978306023,
+      "grad_norm": 3.427734375,
+      "learning_rate": 8.681787858572382e-05,
+      "loss": 1.5519,
+      "step": 1986
+    },
+    {
+      "epoch": 1.8342949457650588,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.681120747164776e-05,
+      "loss": 1.6589,
+      "step": 1987
+    },
+    {
+      "epoch": 1.8352180936995155,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.680453635757172e-05,
+      "loss": 1.6599,
+      "step": 1988
+    },
+    {
+      "epoch": 1.836141241633972,
+      "grad_norm": 5.5234375,
+      "learning_rate": 8.679786524349567e-05,
+      "loss": 1.6516,
+      "step": 1989
+    },
+    {
+      "epoch": 1.8370643895684284,
+      "grad_norm": 6.03515625,
+      "learning_rate": 8.679119412941961e-05,
+      "loss": 1.6843,
+      "step": 1990
+    },
+    {
+      "epoch": 1.837987537502885,
+      "grad_norm": 5.953125,
+      "learning_rate": 8.678452301534356e-05,
+      "loss": 2.0122,
+      "step": 1991
+    },
+    {
+      "epoch": 1.8389106854373414,
+      "grad_norm": 4.3203125,
+      "learning_rate": 8.677785190126752e-05,
+      "loss": 1.575,
+      "step": 1992
+    },
+    {
+      "epoch": 1.8398338333717978,
+      "grad_norm": 4.37890625,
+      "learning_rate": 8.677118078719146e-05,
+      "loss": 1.4369,
+      "step": 1993
+    },
+    {
+      "epoch": 1.8407569813062543,
+      "grad_norm": 5.41015625,
+      "learning_rate": 8.676450967311541e-05,
+      "loss": 1.6829,
+      "step": 1994
+    },
+    {
+      "epoch": 1.8416801292407108,
+      "grad_norm": 6.4765625,
+      "learning_rate": 8.675783855903936e-05,
+      "loss": 1.1851,
+      "step": 1995
+    },
+    {
+      "epoch": 1.8426032771751673,
+      "grad_norm": 5.82421875,
+      "learning_rate": 8.675116744496331e-05,
+      "loss": 1.699,
+      "step": 1996
+    },
+    {
+      "epoch": 1.8435264251096237,
+      "grad_norm": 3.61328125,
+      "learning_rate": 8.674449633088726e-05,
+      "loss": 1.7444,
+      "step": 1997
+    },
+    {
+      "epoch": 1.8444495730440802,
+      "grad_norm": 3.654296875,
+      "learning_rate": 8.67378252168112e-05,
+      "loss": 1.3804,
+      "step": 1998
+    },
+    {
+      "epoch": 1.8453727209785367,
+      "grad_norm": 4.9296875,
+      "learning_rate": 8.673115410273515e-05,
+      "loss": 1.4243,
+      "step": 1999
+    },
+    {
+      "epoch": 1.8462958689129934,
+      "grad_norm": 4.76171875,
+      "learning_rate": 8.672448298865911e-05,
+      "loss": 1.5527,
+      "step": 2000
+    },
+    {
+      "epoch": 1.8462958689129934,
+      "eval_loss": 0.8994140625,
+      "eval_runtime": 23.3732,
+      "eval_samples_per_second": 975.517,
+      "eval_steps_per_second": 61.01,
+      "step": 2000
+    },
+    {
+      "epoch": 1.8472190168474498,
+      "grad_norm": 5.20703125,
+      "learning_rate": 8.671781187458305e-05,
+      "loss": 1.4785,
+      "step": 2001
+    },
+    {
+      "epoch": 1.8481421647819063,
+      "grad_norm": 7.12109375,
+      "learning_rate": 8.6711140760507e-05,
+      "loss": 1.5815,
+      "step": 2002
+    },
+    {
+      "epoch": 1.8490653127163628,
+      "grad_norm": 5.109375,
+      "learning_rate": 8.670446964643096e-05,
+      "loss": 1.5876,
+      "step": 2003
+    },
+    {
+      "epoch": 1.8499884606508195,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.669779853235492e-05,
+      "loss": 1.7468,
+      "step": 2004
+    },
+    {
+      "epoch": 1.850911608585276,
+      "grad_norm": 3.646484375,
+      "learning_rate": 8.669112741827886e-05,
+      "loss": 1.2787,
+      "step": 2005
+    },
+    {
+      "epoch": 1.8518347565197324,
+      "grad_norm": 3.880859375,
+      "learning_rate": 8.668445630420281e-05,
+      "loss": 1.1504,
+      "step": 2006
+    },
+    {
+      "epoch": 1.8527579044541889,
+      "grad_norm": 4.3828125,
+      "learning_rate": 8.667778519012675e-05,
+      "loss": 1.343,
+      "step": 2007
+    },
+    {
+      "epoch": 1.8536810523886453,
+      "grad_norm": 4.25390625,
+      "learning_rate": 8.667111407605071e-05,
+      "loss": 1.5304,
+      "step": 2008
+    },
+    {
+      "epoch": 1.8546042003231018,
+      "grad_norm": 4.765625,
+      "learning_rate": 8.666444296197466e-05,
+      "loss": 1.5815,
+      "step": 2009
+    },
+    {
+      "epoch": 1.8555273482575583,
+      "grad_norm": 5.09765625,
+      "learning_rate": 8.66577718478986e-05,
+      "loss": 1.4264,
+      "step": 2010
+    },
+    {
+      "epoch": 1.8564504961920147,
+      "grad_norm": 5.45703125,
+      "learning_rate": 8.665110073382255e-05,
+      "loss": 1.9573,
+      "step": 2011
+    },
+    {
+      "epoch": 1.8573736441264712,
+      "grad_norm": 4.26171875,
+      "learning_rate": 8.664442961974651e-05,
+      "loss": 1.7117,
+      "step": 2012
+    },
+    {
+      "epoch": 1.8582967920609277,
+      "grad_norm": 5.18359375,
+      "learning_rate": 8.663775850567045e-05,
+      "loss": 1.2875,
+      "step": 2013
+    },
+    {
+      "epoch": 1.8592199399953842,
+      "grad_norm": 5.796875,
+      "learning_rate": 8.66310873915944e-05,
+      "loss": 1.5266,
+      "step": 2014
+    },
+    {
+      "epoch": 1.8601430879298406,
+      "grad_norm": 5.12109375,
+      "learning_rate": 8.662441627751835e-05,
+      "loss": 1.7407,
+      "step": 2015
+    },
+    {
+      "epoch": 1.861066235864297,
+      "grad_norm": 3.80078125,
+      "learning_rate": 8.66177451634423e-05,
+      "loss": 1.5903,
+      "step": 2016
+    },
+    {
+      "epoch": 1.8619893837987538,
+      "grad_norm": 4.60546875,
+      "learning_rate": 8.661107404936625e-05,
+      "loss": 1.329,
+      "step": 2017
+    },
+    {
+      "epoch": 1.8629125317332103,
+      "grad_norm": 4.62109375,
+      "learning_rate": 8.66044029352902e-05,
+      "loss": 1.4067,
+      "step": 2018
+    },
+    {
+      "epoch": 1.8638356796676667,
+      "grad_norm": 4.4453125,
+      "learning_rate": 8.659773182121414e-05,
+      "loss": 1.4539,
+      "step": 2019
+    },
+    {
+      "epoch": 1.8647588276021232,
+      "grad_norm": 4.39453125,
+      "learning_rate": 8.65910607071381e-05,
+      "loss": 1.4775,
+      "step": 2020
+    },
+    {
+      "epoch": 1.8656819755365799,
+      "grad_norm": 4.0234375,
+      "learning_rate": 8.658438959306204e-05,
+      "loss": 1.4666,
+      "step": 2021
+    },
+    {
+      "epoch": 1.8666051234710364,
+      "grad_norm": 4.12890625,
+      "learning_rate": 8.657771847898599e-05,
+      "loss": 1.4912,
+      "step": 2022
+    },
+    {
+      "epoch": 1.8675282714054928,
+      "grad_norm": 5.0390625,
+      "learning_rate": 8.657104736490994e-05,
+      "loss": 1.7029,
+      "step": 2023
+    },
+    {
+      "epoch": 1.8684514193399493,
+      "grad_norm": 4.51953125,
+      "learning_rate": 8.65643762508339e-05,
+      "loss": 1.0549,
+      "step": 2024
+    },
+    {
+      "epoch": 1.8693745672744058,
+      "grad_norm": 5.0703125,
+      "learning_rate": 8.655770513675784e-05,
+      "loss": 1.3423,
+      "step": 2025
+    },
+    {
+      "epoch": 1.8702977152088622,
+      "grad_norm": 4.51171875,
+      "learning_rate": 8.655103402268179e-05,
+      "loss": 1.5779,
+      "step": 2026
+    },
+    {
+      "epoch": 1.8712208631433187,
+      "grad_norm": 6.3046875,
+      "learning_rate": 8.654436290860573e-05,
+      "loss": 1.6267,
+      "step": 2027
+    },
+    {
+      "epoch": 1.8721440110777752,
+      "grad_norm": 5.859375,
+      "learning_rate": 8.653769179452969e-05,
+      "loss": 1.4697,
+      "step": 2028
+    },
+    {
+      "epoch": 1.8730671590122316,
+      "grad_norm": 5.90234375,
+      "learning_rate": 8.653102068045365e-05,
+      "loss": 1.7134,
+      "step": 2029
+    },
+    {
+      "epoch": 1.8739903069466881,
+      "grad_norm": 3.837890625,
+      "learning_rate": 8.65243495663776e-05,
+      "loss": 1.7056,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8749134548811446,
+      "grad_norm": 5.3046875,
+      "learning_rate": 8.651767845230154e-05,
+      "loss": 1.7871,
+      "step": 2031
+    },
+    {
+      "epoch": 1.875836602815601,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.651100733822549e-05,
+      "loss": 1.7236,
+      "step": 2032
+    },
+    {
+      "epoch": 1.8767597507500577,
+      "grad_norm": 5.8125,
+      "learning_rate": 8.650433622414944e-05,
+      "loss": 1.7146,
+      "step": 2033
+    },
+    {
+      "epoch": 1.8776828986845142,
+      "grad_norm": 4.44921875,
+      "learning_rate": 8.649766511007339e-05,
+      "loss": 1.7,
+      "step": 2034
+    },
+    {
+      "epoch": 1.8786060466189707,
+      "grad_norm": 3.5703125,
+      "learning_rate": 8.649099399599734e-05,
+      "loss": 1.1168,
+      "step": 2035
+    },
+    {
+      "epoch": 1.8795291945534272,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.648432288192128e-05,
+      "loss": 1.4514,
+      "step": 2036
+    },
+    {
+      "epoch": 1.8804523424878838,
+      "grad_norm": 5.03515625,
+      "learning_rate": 8.647765176784524e-05,
+      "loss": 1.5388,
+      "step": 2037
+    },
+    {
+      "epoch": 1.8813754904223403,
+      "grad_norm": 4.02734375,
+      "learning_rate": 8.647098065376918e-05,
+      "loss": 1.2385,
+      "step": 2038
+    },
+    {
+      "epoch": 1.8822986383567968,
+      "grad_norm": 3.666015625,
+      "learning_rate": 8.646430953969313e-05,
+      "loss": 1.2545,
+      "step": 2039
+    },
+    {
+      "epoch": 1.8832217862912533,
+      "grad_norm": 5.0703125,
+      "learning_rate": 8.645763842561708e-05,
+      "loss": 1.2419,
+      "step": 2040
+    },
+    {
+      "epoch": 1.8841449342257097,
+      "grad_norm": 5.48046875,
+      "learning_rate": 8.645096731154103e-05,
+      "loss": 1.5591,
+      "step": 2041
+    },
+    {
+      "epoch": 1.8850680821601662,
+      "grad_norm": 5.37109375,
+      "learning_rate": 8.644429619746498e-05,
+      "loss": 1.6174,
+      "step": 2042
+    },
+    {
+      "epoch": 1.8859912300946227,
+      "grad_norm": 6.3671875,
+      "learning_rate": 8.643762508338893e-05,
+      "loss": 1.7532,
+      "step": 2043
+    },
+    {
+      "epoch": 1.8869143780290791,
+      "grad_norm": 4.234375,
+      "learning_rate": 8.643095396931288e-05,
+      "loss": 1.4471,
+      "step": 2044
+    },
+    {
+      "epoch": 1.8878375259635356,
+      "grad_norm": 4.25,
+      "learning_rate": 8.642428285523683e-05,
+      "loss": 1.6848,
+      "step": 2045
+    },
+    {
+      "epoch": 1.888760673897992,
+      "grad_norm": 4.4765625,
+      "learning_rate": 8.641761174116078e-05,
+      "loss": 1.7224,
+      "step": 2046
+    },
+    {
+      "epoch": 1.8896838218324485,
+      "grad_norm": 4.421875,
+      "learning_rate": 8.641094062708472e-05,
+      "loss": 1.6724,
+      "step": 2047
+    },
+    {
+      "epoch": 1.890606969766905,
+      "grad_norm": 4.83984375,
+      "learning_rate": 8.640426951300868e-05,
+      "loss": 1.5093,
+      "step": 2048
+    },
+    {
+      "epoch": 1.8915301177013615,
+      "grad_norm": 3.841796875,
+      "learning_rate": 8.639759839893263e-05,
+      "loss": 1.46,
+      "step": 2049
+    },
+    {
+      "epoch": 1.8924532656358182,
+      "grad_norm": 4.45703125,
+      "learning_rate": 8.639092728485657e-05,
+      "loss": 1.54,
+      "step": 2050
+    },
+    {
+      "epoch": 1.8933764135702746,
+      "grad_norm": 5.48046875,
+      "learning_rate": 8.638425617078052e-05,
+      "loss": 1.2917,
+      "step": 2051
+    },
+    {
+      "epoch": 1.894299561504731,
+      "grad_norm": 4.8984375,
+      "learning_rate": 8.637758505670448e-05,
+      "loss": 1.3701,
+      "step": 2052
+    },
+    {
+      "epoch": 1.8952227094391876,
+      "grad_norm": 5.34375,
+      "learning_rate": 8.637091394262843e-05,
+      "loss": 1.4822,
+      "step": 2053
+    },
+    {
+      "epoch": 1.8961458573736443,
+      "grad_norm": 4.2734375,
+      "learning_rate": 8.636424282855238e-05,
+      "loss": 1.5742,
+      "step": 2054
+    },
+    {
+      "epoch": 1.8970690053081007,
+      "grad_norm": 4.8828125,
+      "learning_rate": 8.635757171447632e-05,
+      "loss": 1.616,
+      "step": 2055
+    },
+    {
+      "epoch": 1.8979921532425572,
+      "grad_norm": 3.966796875,
+      "learning_rate": 8.635090060040027e-05,
+      "loss": 1.21,
+      "step": 2056
+    },
+    {
+      "epoch": 1.8989153011770137,
+      "grad_norm": 4.125,
+      "learning_rate": 8.634422948632423e-05,
+      "loss": 1.3792,
+      "step": 2057
+    },
+    {
+      "epoch": 1.8998384491114702,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.633755837224817e-05,
+      "loss": 1.4573,
+      "step": 2058
+    },
+    {
+      "epoch": 1.9007615970459266,
+      "grad_norm": 3.37109375,
+      "learning_rate": 8.633088725817212e-05,
+      "loss": 1.4846,
+      "step": 2059
+    },
+    {
+      "epoch": 1.901684744980383,
+      "grad_norm": 3.76953125,
+      "learning_rate": 8.632421614409607e-05,
+      "loss": 1.448,
+      "step": 2060
+    },
+    {
+      "epoch": 1.9026078929148396,
+      "grad_norm": 4.37890625,
+      "learning_rate": 8.631754503002002e-05,
+      "loss": 1.6941,
+      "step": 2061
+    },
+    {
+      "epoch": 1.903531040849296,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.631087391594397e-05,
+      "loss": 1.3248,
+      "step": 2062
+    },
+    {
+      "epoch": 1.9044541887837525,
+      "grad_norm": 4.43359375,
+      "learning_rate": 8.630420280186792e-05,
+      "loss": 1.4734,
+      "step": 2063
+    },
+    {
+      "epoch": 1.905377336718209,
+      "grad_norm": 4.9609375,
+      "learning_rate": 8.629753168779186e-05,
+      "loss": 1.4197,
+      "step": 2064
+    },
+    {
+      "epoch": 1.9063004846526654,
+      "grad_norm": 3.9140625,
+      "learning_rate": 8.629086057371582e-05,
+      "loss": 1.4333,
+      "step": 2065
+    },
+    {
+      "epoch": 1.9072236325871221,
+      "grad_norm": 4.3984375,
+      "learning_rate": 8.628418945963977e-05,
+      "loss": 1.8264,
+      "step": 2066
+    },
+    {
+      "epoch": 1.9081467805215786,
+      "grad_norm": 5.078125,
+      "learning_rate": 8.627751834556371e-05,
+      "loss": 1.5081,
+      "step": 2067
+    },
+    {
+      "epoch": 1.909069928456035,
+      "grad_norm": 4.51171875,
+      "learning_rate": 8.627084723148766e-05,
+      "loss": 1.4349,
+      "step": 2068
+    },
+    {
+      "epoch": 1.9099930763904915,
+      "grad_norm": 4.51953125,
+      "learning_rate": 8.626417611741162e-05,
+      "loss": 1.4779,
+      "step": 2069
+    },
+    {
+      "epoch": 1.9109162243249482,
+      "grad_norm": 4.08984375,
+      "learning_rate": 8.625750500333556e-05,
+      "loss": 1.4878,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9118393722594047,
+      "grad_norm": 4.4375,
+      "learning_rate": 8.62508338892595e-05,
+      "loss": 1.3905,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9127625201938612,
+      "grad_norm": 6.08203125,
+      "learning_rate": 8.624416277518345e-05,
+      "loss": 1.6211,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9136856681283176,
+      "grad_norm": 7.125,
+      "learning_rate": 8.623749166110741e-05,
+      "loss": 1.5343,
+      "step": 2073
+    },
+    {
+      "epoch": 1.914608816062774,
+      "grad_norm": 3.572265625,
+      "learning_rate": 8.623082054703136e-05,
+      "loss": 1.1339,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9155319639972306,
+      "grad_norm": 4.22265625,
+      "learning_rate": 8.62241494329553e-05,
+      "loss": 1.2952,
+      "step": 2075
+    },
+    {
+      "epoch": 1.916455111931687,
+      "grad_norm": 5.18359375,
+      "learning_rate": 8.621747831887925e-05,
+      "loss": 1.5156,
+      "step": 2076
+    },
+    {
+      "epoch": 1.9173782598661435,
+      "grad_norm": 4.828125,
+      "learning_rate": 8.62108072048032e-05,
+      "loss": 1.5435,
+      "step": 2077
+    },
+    {
+      "epoch": 1.9183014078006,
+      "grad_norm": 4.6640625,
+      "learning_rate": 8.620413609072715e-05,
+      "loss": 1.2637,
+      "step": 2078
+    },
+    {
+      "epoch": 1.9192245557350565,
+      "grad_norm": 4.56640625,
+      "learning_rate": 8.619746497665111e-05,
+      "loss": 1.3616,
+      "step": 2079
+    },
+    {
+      "epoch": 1.920147703669513,
+      "grad_norm": 4.4140625,
+      "learning_rate": 8.619079386257506e-05,
+      "loss": 1.6125,
+      "step": 2080
+    },
+    {
+      "epoch": 1.9210708516039694,
+      "grad_norm": 4.6484375,
+      "learning_rate": 8.6184122748499e-05,
+      "loss": 1.4722,
+      "step": 2081
+    },
+    {
+      "epoch": 1.9219939995384259,
+      "grad_norm": 5.32421875,
+      "learning_rate": 8.617745163442296e-05,
+      "loss": 1.5388,
+      "step": 2082
+    },
+    {
+      "epoch": 1.9229171474728826,
+      "grad_norm": 5.08203125,
+      "learning_rate": 8.61707805203469e-05,
+      "loss": 1.1285,
+      "step": 2083
+    },
+    {
+      "epoch": 1.923840295407339,
+      "grad_norm": 5.3046875,
+      "learning_rate": 8.616410940627085e-05,
+      "loss": 1.6062,
+      "step": 2084
+    },
+    {
+      "epoch": 1.9247634433417955,
+      "grad_norm": 3.646484375,
+      "learning_rate": 8.615743829219481e-05,
+      "loss": 1.4829,
+      "step": 2085
+    },
+    {
+      "epoch": 1.925686591276252,
+      "grad_norm": 4.0703125,
+      "learning_rate": 8.615076717811876e-05,
+      "loss": 1.8333,
+      "step": 2086
+    },
+    {
+      "epoch": 1.9266097392107087,
+      "grad_norm": 4.78515625,
+      "learning_rate": 8.61440960640427e-05,
+      "loss": 1.4834,
+      "step": 2087
+    },
+    {
+      "epoch": 1.9275328871451651,
+      "grad_norm": 3.685546875,
+      "learning_rate": 8.613742494996665e-05,
+      "loss": 1.5068,
+      "step": 2088
+    },
+    {
+      "epoch": 1.9284560350796216,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.61307538358906e-05,
+      "loss": 1.5757,
+      "step": 2089
+    },
+    {
+      "epoch": 1.929379183014078,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.612408272181455e-05,
+      "loss": 1.6262,
+      "step": 2090
+    },
+    {
+      "epoch": 1.9303023309485345,
+      "grad_norm": 5.42578125,
+      "learning_rate": 8.61174116077385e-05,
+      "loss": 1.6096,
+      "step": 2091
+    },
+    {
+      "epoch": 1.931225478882991,
+      "grad_norm": 4.68359375,
+      "learning_rate": 8.611074049366244e-05,
+      "loss": 1.5579,
+      "step": 2092
+    },
+    {
+      "epoch": 1.9321486268174475,
+      "grad_norm": 6.39453125,
+      "learning_rate": 8.61040693795864e-05,
+      "loss": 1.5894,
+      "step": 2093
+    },
+    {
+      "epoch": 1.933071774751904,
+      "grad_norm": 5.7421875,
+      "learning_rate": 8.609739826551035e-05,
+      "loss": 1.5513,
+      "step": 2094
+    },
+    {
+      "epoch": 1.9339949226863604,
+      "grad_norm": 4.296875,
+      "learning_rate": 8.609072715143429e-05,
+      "loss": 1.205,
+      "step": 2095
+    },
+    {
+      "epoch": 1.9349180706208169,
+      "grad_norm": 5.33203125,
+      "learning_rate": 8.608405603735824e-05,
+      "loss": 1.5676,
+      "step": 2096
+    },
+    {
+      "epoch": 1.9358412185552734,
+      "grad_norm": 3.9921875,
+      "learning_rate": 8.60773849232822e-05,
+      "loss": 1.7812,
+      "step": 2097
+    },
+    {
+      "epoch": 1.9367643664897298,
+      "grad_norm": 4.75390625,
+      "learning_rate": 8.607071380920614e-05,
+      "loss": 1.8044,
+      "step": 2098
+    },
+    {
+      "epoch": 1.9376875144241865,
+      "grad_norm": 4.6015625,
+      "learning_rate": 8.606404269513009e-05,
+      "loss": 1.431,
+      "step": 2099
+    },
+    {
+      "epoch": 1.938610662358643,
+      "grad_norm": 3.99609375,
+      "learning_rate": 8.605737158105403e-05,
+      "loss": 1.5347,
+      "step": 2100
+    },
+    {
+      "epoch": 1.9395338102930995,
+      "grad_norm": 4.62890625,
+      "learning_rate": 8.605070046697799e-05,
+      "loss": 1.6221,
+      "step": 2101
+    },
+    {
+      "epoch": 1.940456958227556,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.604402935290194e-05,
+      "loss": 1.6443,
+      "step": 2102
+    },
+    {
+      "epoch": 1.9413801061620126,
+      "grad_norm": 4.390625,
+      "learning_rate": 8.603735823882588e-05,
+      "loss": 1.3918,
+      "step": 2103
+    },
+    {
+      "epoch": 1.942303254096469,
+      "grad_norm": 5.30859375,
+      "learning_rate": 8.603068712474984e-05,
+      "loss": 1.6514,
+      "step": 2104
+    },
+    {
+      "epoch": 1.9432264020309256,
+      "grad_norm": 4.06640625,
+      "learning_rate": 8.602401601067379e-05,
+      "loss": 1.2115,
+      "step": 2105
+    },
+    {
+      "epoch": 1.944149549965382,
+      "grad_norm": 4.4453125,
+      "learning_rate": 8.601734489659775e-05,
+      "loss": 1.3433,
+      "step": 2106
+    },
+    {
+      "epoch": 1.9450726978998385,
+      "grad_norm": 4.23046875,
+      "learning_rate": 8.601067378252169e-05,
+      "loss": 1.3552,
+      "step": 2107
+    },
+    {
+      "epoch": 1.945995845834295,
+      "grad_norm": 5.12109375,
+      "learning_rate": 8.600400266844564e-05,
+      "loss": 1.5596,
+      "step": 2108
+    },
+    {
+      "epoch": 1.9469189937687514,
+      "grad_norm": 4.9609375,
+      "learning_rate": 8.599733155436958e-05,
+      "loss": 1.3351,
+      "step": 2109
+    },
+    {
+      "epoch": 1.947842141703208,
+      "grad_norm": 5.85546875,
+      "learning_rate": 8.599066044029354e-05,
+      "loss": 1.4133,
+      "step": 2110
+    },
+    {
+      "epoch": 1.9487652896376644,
+      "grad_norm": 4.82421875,
+      "learning_rate": 8.598398932621749e-05,
+      "loss": 1.6865,
+      "step": 2111
+    },
+    {
+      "epoch": 1.9496884375721208,
+      "grad_norm": 5.125,
+      "learning_rate": 8.597731821214143e-05,
+      "loss": 1.5315,
+      "step": 2112
+    },
+    {
+      "epoch": 1.9506115855065773,
+      "grad_norm": 4.42578125,
+      "learning_rate": 8.597064709806538e-05,
+      "loss": 1.5703,
+      "step": 2113
+    },
+    {
+      "epoch": 1.9515347334410338,
+      "grad_norm": 5.08203125,
+      "learning_rate": 8.596397598398934e-05,
+      "loss": 1.4072,
+      "step": 2114
+    },
+    {
+      "epoch": 1.9524578813754903,
+      "grad_norm": 4.2734375,
+      "learning_rate": 8.595730486991328e-05,
+      "loss": 1.3877,
+      "step": 2115
+    },
+    {
+      "epoch": 1.953381029309947,
+      "grad_norm": 3.974609375,
+      "learning_rate": 8.595063375583723e-05,
+      "loss": 1.4106,
+      "step": 2116
+    },
+    {
+      "epoch": 1.9543041772444034,
+      "grad_norm": 4.66796875,
+      "learning_rate": 8.594396264176117e-05,
+      "loss": 1.4905,
+      "step": 2117
+    },
+    {
+      "epoch": 1.9552273251788599,
+      "grad_norm": 5.98046875,
+      "learning_rate": 8.593729152768513e-05,
+      "loss": 1.824,
+      "step": 2118
+    },
+    {
+      "epoch": 1.9561504731133164,
+      "grad_norm": 5.50390625,
+      "learning_rate": 8.593062041360908e-05,
+      "loss": 1.7466,
+      "step": 2119
+    },
+    {
+      "epoch": 1.957073621047773,
+      "grad_norm": 5.10546875,
+      "learning_rate": 8.592394929953302e-05,
+      "loss": 1.4734,
+      "step": 2120
+    },
+    {
+      "epoch": 1.9579967689822295,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.591727818545697e-05,
+      "loss": 1.5433,
+      "step": 2121
+    },
+    {
+      "epoch": 1.958919916916686,
+      "grad_norm": 4.578125,
+      "learning_rate": 8.591060707138093e-05,
+      "loss": 1.6206,
+      "step": 2122
+    },
+    {
+      "epoch": 1.9598430648511425,
+      "grad_norm": 5.2734375,
+      "learning_rate": 8.590393595730487e-05,
+      "loss": 1.5381,
+      "step": 2123
+    },
+    {
+      "epoch": 1.960766212785599,
+      "grad_norm": 3.95703125,
+      "learning_rate": 8.589726484322882e-05,
+      "loss": 1.5286,
+      "step": 2124
+    },
+    {
+      "epoch": 1.9616893607200554,
+      "grad_norm": 5.8515625,
+      "learning_rate": 8.589059372915276e-05,
+      "loss": 1.6477,
+      "step": 2125
+    },
+    {
+      "epoch": 1.9626125086545119,
+      "grad_norm": 5.45703125,
+      "learning_rate": 8.588392261507672e-05,
+      "loss": 1.6887,
+      "step": 2126
+    },
+    {
+      "epoch": 1.9635356565889683,
+      "grad_norm": 5.796875,
+      "learning_rate": 8.587725150100067e-05,
+      "loss": 1.6079,
+      "step": 2127
+    },
+    {
+      "epoch": 1.9644588045234248,
+      "grad_norm": 3.94921875,
+      "learning_rate": 8.587058038692461e-05,
+      "loss": 1.5947,
+      "step": 2128
+    },
+    {
+      "epoch": 1.9653819524578813,
+      "grad_norm": 5.15234375,
+      "learning_rate": 8.586390927284857e-05,
+      "loss": 1.9146,
+      "step": 2129
+    },
+    {
+      "epoch": 1.9663051003923377,
+      "grad_norm": 4.30078125,
+      "learning_rate": 8.585723815877253e-05,
+      "loss": 1.6553,
+      "step": 2130
+    },
+    {
+      "epoch": 1.9672282483267942,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.585056704469648e-05,
+      "loss": 1.7451,
+      "step": 2131
+    },
+    {
+      "epoch": 1.968151396261251,
+      "grad_norm": 4.46484375,
+      "learning_rate": 8.584389593062042e-05,
+      "loss": 1.8879,
+      "step": 2132
+    },
+    {
+      "epoch": 1.9690745441957074,
+      "grad_norm": 5.62890625,
+      "learning_rate": 8.583722481654437e-05,
+      "loss": 1.5991,
+      "step": 2133
+    },
+    {
+      "epoch": 1.9699976921301638,
+      "grad_norm": 4.13671875,
+      "learning_rate": 8.583055370246833e-05,
+      "loss": 1.5647,
+      "step": 2134
+    },
+    {
+      "epoch": 1.9709208400646203,
+      "grad_norm": 4.88671875,
+      "learning_rate": 8.582388258839227e-05,
+      "loss": 1.4136,
+      "step": 2135
+    },
+    {
+      "epoch": 1.971843987999077,
+      "grad_norm": 3.853515625,
+      "learning_rate": 8.581721147431622e-05,
+      "loss": 1.425,
+      "step": 2136
+    },
+    {
+      "epoch": 1.9727671359335335,
+      "grad_norm": 4.82421875,
+      "learning_rate": 8.581054036024016e-05,
+      "loss": 1.5496,
+      "step": 2137
+    },
+    {
+      "epoch": 1.97369028386799,
+      "grad_norm": 4.421875,
+      "learning_rate": 8.580386924616412e-05,
+      "loss": 1.741,
+      "step": 2138
+    },
+    {
+      "epoch": 1.9746134318024464,
+      "grad_norm": 4.54296875,
+      "learning_rate": 8.579719813208807e-05,
+      "loss": 1.616,
+      "step": 2139
+    },
+    {
+      "epoch": 1.9755365797369029,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.579052701801201e-05,
+      "loss": 1.4663,
+      "step": 2140
+    },
+    {
+      "epoch": 1.9764597276713594,
+      "grad_norm": 4.1328125,
+      "learning_rate": 8.578385590393596e-05,
+      "loss": 1.3407,
+      "step": 2141
+    },
+    {
+      "epoch": 1.9773828756058158,
+      "grad_norm": 4.45703125,
+      "learning_rate": 8.577718478985992e-05,
+      "loss": 1.6777,
+      "step": 2142
+    },
+    {
+      "epoch": 1.9783060235402723,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.577051367578386e-05,
+      "loss": 1.5227,
+      "step": 2143
+    },
+    {
+      "epoch": 1.9792291714747288,
+      "grad_norm": 5.84375,
+      "learning_rate": 8.576384256170781e-05,
+      "loss": 1.6548,
+      "step": 2144
+    },
+    {
+      "epoch": 1.9801523194091852,
+      "grad_norm": 4.77734375,
+      "learning_rate": 8.575717144763175e-05,
+      "loss": 1.627,
+      "step": 2145
+    },
+    {
+      "epoch": 1.9810754673436417,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.575050033355571e-05,
+      "loss": 1.5171,
+      "step": 2146
+    },
+    {
+      "epoch": 1.9819986152780982,
+      "grad_norm": 4.9921875,
+      "learning_rate": 8.574382921947966e-05,
+      "loss": 1.4094,
+      "step": 2147
+    },
+    {
+      "epoch": 1.9829217632125549,
+      "grad_norm": 4.33984375,
+      "learning_rate": 8.57371581054036e-05,
+      "loss": 1.4396,
+      "step": 2148
+    },
+    {
+      "epoch": 1.9838449111470113,
+      "grad_norm": 4.234375,
+      "learning_rate": 8.573048699132755e-05,
+      "loss": 1.249,
+      "step": 2149
+    },
+    {
+      "epoch": 1.9847680590814678,
+      "grad_norm": 5.25,
+      "learning_rate": 8.57238158772515e-05,
+      "loss": 1.6138,
+      "step": 2150
+    },
+    {
+      "epoch": 1.9856912070159243,
+      "grad_norm": 6.1015625,
+      "learning_rate": 8.571714476317545e-05,
+      "loss": 1.4619,
+      "step": 2151
+    },
+    {
+      "epoch": 1.9866143549503807,
+      "grad_norm": 6.83203125,
+      "learning_rate": 8.57104736490994e-05,
+      "loss": 1.7981,
+      "step": 2152
+    },
+    {
+      "epoch": 1.9875375028848374,
+      "grad_norm": 3.861328125,
+      "learning_rate": 8.570380253502334e-05,
+      "loss": 1.9382,
+      "step": 2153
+    },
+    {
+      "epoch": 1.988460650819294,
+      "grad_norm": 4.0703125,
+      "learning_rate": 8.56971314209473e-05,
+      "loss": 1.4983,
+      "step": 2154
+    },
+    {
+      "epoch": 1.9893837987537504,
+      "grad_norm": 4.5,
+      "learning_rate": 8.569046030687126e-05,
+      "loss": 1.6365,
+      "step": 2155
+    },
+    {
+      "epoch": 1.9903069466882068,
+      "grad_norm": 5.1796875,
+      "learning_rate": 8.56837891927952e-05,
+      "loss": 1.6328,
+      "step": 2156
+    },
+    {
+      "epoch": 1.9912300946226633,
+      "grad_norm": 4.56640625,
+      "learning_rate": 8.567711807871915e-05,
+      "loss": 1.8059,
+      "step": 2157
+    },
+    {
+      "epoch": 1.9921532425571198,
+      "grad_norm": 5.4453125,
+      "learning_rate": 8.56704469646431e-05,
+      "loss": 1.5139,
+      "step": 2158
+    },
+    {
+      "epoch": 1.9930763904915763,
+      "grad_norm": 6.18359375,
+      "learning_rate": 8.566377585056706e-05,
+      "loss": 1.5652,
+      "step": 2159
+    },
+    {
+      "epoch": 1.9939995384260327,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.5657104736491e-05,
+      "loss": 1.572,
+      "step": 2160
+    },
+    {
+      "epoch": 1.9949226863604892,
+      "grad_norm": 3.7578125,
+      "learning_rate": 8.565043362241495e-05,
+      "loss": 1.6018,
+      "step": 2161
+    },
+    {
+      "epoch": 1.9958458342949457,
+      "grad_norm": 5.08203125,
+      "learning_rate": 8.564376250833889e-05,
+      "loss": 1.7935,
+      "step": 2162
+    },
+    {
+      "epoch": 1.9967689822294021,
+      "grad_norm": 6.125,
+      "learning_rate": 8.563709139426285e-05,
+      "loss": 1.6143,
+      "step": 2163
+    },
+    {
+      "epoch": 1.9976921301638586,
+      "grad_norm": 5.171875,
+      "learning_rate": 8.56304202801868e-05,
+      "loss": 1.4597,
+      "step": 2164
+    },
+    {
+      "epoch": 1.9986152780983153,
+      "grad_norm": 4.01953125,
+      "learning_rate": 8.562374916611074e-05,
+      "loss": 1.1812,
+      "step": 2165
+    },
+    {
+      "epoch": 1.9995384260327718,
+      "grad_norm": 6.08984375,
+      "learning_rate": 8.561707805203469e-05,
+      "loss": 1.6858,
+      "step": 2166
+    },
+    {
+      "epoch": 2.0004615739672285,
+      "grad_norm": 6.67578125,
+      "learning_rate": 8.561040693795865e-05,
+      "loss": 1.4653,
+      "step": 2167
+    },
+    {
+      "epoch": 2.001384721901685,
+      "grad_norm": 5.37109375,
+      "learning_rate": 8.560373582388259e-05,
+      "loss": 1.531,
+      "step": 2168
+    },
+    {
+      "epoch": 2.0023078698361414,
+      "grad_norm": 4.7578125,
+      "learning_rate": 8.559706470980654e-05,
+      "loss": 1.2557,
+      "step": 2169
+    },
+    {
+      "epoch": 2.003231017770598,
+      "grad_norm": 5.5390625,
+      "learning_rate": 8.559039359573048e-05,
+      "loss": 1.4819,
+      "step": 2170
+    },
+    {
+      "epoch": 2.0041541657050543,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.558372248165444e-05,
+      "loss": 1.3473,
+      "step": 2171
+    },
+    {
+      "epoch": 2.005077313639511,
+      "grad_norm": 4.78515625,
+      "learning_rate": 8.557705136757839e-05,
+      "loss": 1.5178,
+      "step": 2172
+    },
+    {
+      "epoch": 2.0060004615739673,
+      "grad_norm": 5.0703125,
+      "learning_rate": 8.557038025350233e-05,
+      "loss": 1.6157,
+      "step": 2173
+    },
+    {
+      "epoch": 2.0069236095084237,
+      "grad_norm": 5.16796875,
+      "learning_rate": 8.556370913942628e-05,
+      "loss": 1.7178,
+      "step": 2174
+    },
+    {
+      "epoch": 2.00784675744288,
+      "grad_norm": 6.015625,
+      "learning_rate": 8.555703802535024e-05,
+      "loss": 1.6396,
+      "step": 2175
+    },
+    {
+      "epoch": 2.0087699053773367,
+      "grad_norm": 6.12109375,
+      "learning_rate": 8.555036691127418e-05,
+      "loss": 1.5161,
+      "step": 2176
+    },
+    {
+      "epoch": 2.009693053311793,
+      "grad_norm": 5.2578125,
+      "learning_rate": 8.554369579719813e-05,
+      "loss": 1.5947,
+      "step": 2177
+    },
+    {
+      "epoch": 2.0106162012462496,
+      "grad_norm": 6.91796875,
+      "learning_rate": 8.553702468312207e-05,
+      "loss": 1.4634,
+      "step": 2178
+    },
+    {
+      "epoch": 2.011539349180706,
+      "grad_norm": 5.16015625,
+      "learning_rate": 8.553035356904605e-05,
+      "loss": 1.1517,
+      "step": 2179
+    },
+    {
+      "epoch": 2.0124624971151626,
+      "grad_norm": 5.69921875,
+      "learning_rate": 8.552368245496999e-05,
+      "loss": 1.5493,
+      "step": 2180
+    },
+    {
+      "epoch": 2.013385645049619,
+      "grad_norm": 5.82421875,
+      "learning_rate": 8.551701134089394e-05,
+      "loss": 1.5195,
+      "step": 2181
+    },
+    {
+      "epoch": 2.0143087929840755,
+      "grad_norm": 4.4609375,
+      "learning_rate": 8.551034022681788e-05,
+      "loss": 1.3162,
+      "step": 2182
+    },
+    {
+      "epoch": 2.0152319409185324,
+      "grad_norm": 4.484375,
+      "learning_rate": 8.550366911274184e-05,
+      "loss": 1.2917,
+      "step": 2183
+    },
+    {
+      "epoch": 2.016155088852989,
+      "grad_norm": 3.47265625,
+      "learning_rate": 8.549699799866579e-05,
+      "loss": 1.6162,
+      "step": 2184
+    },
+    {
+      "epoch": 2.0170782367874454,
+      "grad_norm": 3.291015625,
+      "learning_rate": 8.549032688458973e-05,
+      "loss": 1.5686,
+      "step": 2185
+    },
+    {
+      "epoch": 2.018001384721902,
+      "grad_norm": 4.4296875,
+      "learning_rate": 8.548365577051368e-05,
+      "loss": 1.6055,
+      "step": 2186
+    },
+    {
+      "epoch": 2.0189245326563583,
+      "grad_norm": 5.28125,
+      "learning_rate": 8.547698465643764e-05,
+      "loss": 1.6021,
+      "step": 2187
+    },
+    {
+      "epoch": 2.0198476805908148,
+      "grad_norm": 4.70703125,
+      "learning_rate": 8.547031354236158e-05,
+      "loss": 1.6169,
+      "step": 2188
+    },
+    {
+      "epoch": 2.0207708285252712,
+      "grad_norm": 4.08984375,
+      "learning_rate": 8.546364242828553e-05,
+      "loss": 1.5435,
+      "step": 2189
+    },
+    {
+      "epoch": 2.0216939764597277,
+      "grad_norm": 3.966796875,
+      "learning_rate": 8.545697131420947e-05,
+      "loss": 1.4655,
+      "step": 2190
+    },
+    {
+      "epoch": 2.022617124394184,
+      "grad_norm": 4.140625,
+      "learning_rate": 8.545030020013343e-05,
+      "loss": 1.0692,
+      "step": 2191
+    },
+    {
+      "epoch": 2.0235402723286406,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.544362908605738e-05,
+      "loss": 1.502,
+      "step": 2192
+    },
+    {
+      "epoch": 2.024463420263097,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.543695797198132e-05,
+      "loss": 1.4473,
+      "step": 2193
+    },
+    {
+      "epoch": 2.0253865681975536,
+      "grad_norm": 5.2265625,
+      "learning_rate": 8.543028685790527e-05,
+      "loss": 1.1649,
+      "step": 2194
+    },
+    {
+      "epoch": 2.02630971613201,
+      "grad_norm": 5.48828125,
+      "learning_rate": 8.542361574382923e-05,
+      "loss": 1.5864,
+      "step": 2195
+    },
+    {
+      "epoch": 2.0272328640664665,
+      "grad_norm": 4.61328125,
+      "learning_rate": 8.541694462975317e-05,
+      "loss": 1.634,
+      "step": 2196
+    },
+    {
+      "epoch": 2.028156012000923,
+      "grad_norm": 5.01171875,
+      "learning_rate": 8.541027351567712e-05,
+      "loss": 1.4753,
+      "step": 2197
+    },
+    {
+      "epoch": 2.0290791599353795,
+      "grad_norm": 4.91015625,
+      "learning_rate": 8.540360240160106e-05,
+      "loss": 1.4353,
+      "step": 2198
+    },
+    {
+      "epoch": 2.030002307869836,
+      "grad_norm": 4.6484375,
+      "learning_rate": 8.539693128752502e-05,
+      "loss": 1.5662,
+      "step": 2199
+    },
+    {
+      "epoch": 2.030925455804293,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.539026017344897e-05,
+      "loss": 1.3431,
+      "step": 2200
+    },
+    {
+      "epoch": 2.0318486037387493,
+      "grad_norm": 4.1484375,
+      "learning_rate": 8.538358905937291e-05,
+      "loss": 1.5168,
+      "step": 2201
+    },
+    {
+      "epoch": 2.032771751673206,
+      "grad_norm": 4.3515625,
+      "learning_rate": 8.537691794529686e-05,
+      "loss": 1.5144,
+      "step": 2202
+    },
+    {
+      "epoch": 2.0336948996076623,
+      "grad_norm": 5.27734375,
+      "learning_rate": 8.537024683122082e-05,
+      "loss": 1.4519,
+      "step": 2203
+    },
+    {
+      "epoch": 2.0346180475421187,
+      "grad_norm": 5.6796875,
+      "learning_rate": 8.536357571714478e-05,
+      "loss": 1.4365,
+      "step": 2204
+    },
+    {
+      "epoch": 2.035541195476575,
+      "grad_norm": 3.83984375,
+      "learning_rate": 8.535690460306872e-05,
+      "loss": 1.3273,
+      "step": 2205
+    },
+    {
+      "epoch": 2.0364643434110317,
+      "grad_norm": 5.14453125,
+      "learning_rate": 8.535023348899267e-05,
+      "loss": 1.3889,
+      "step": 2206
+    },
+    {
+      "epoch": 2.037387491345488,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.534356237491661e-05,
+      "loss": 1.541,
+      "step": 2207
+    },
+    {
+      "epoch": 2.0383106392799446,
+      "grad_norm": 4.51953125,
+      "learning_rate": 8.533689126084057e-05,
+      "loss": 1.2506,
+      "step": 2208
+    },
+    {
+      "epoch": 2.039233787214401,
+      "grad_norm": 4.19921875,
+      "learning_rate": 8.533022014676452e-05,
+      "loss": 1.3032,
+      "step": 2209
+    },
+    {
+      "epoch": 2.0401569351488575,
+      "grad_norm": 4.359375,
+      "learning_rate": 8.532354903268846e-05,
+      "loss": 1.3906,
+      "step": 2210
+    },
+    {
+      "epoch": 2.041080083083314,
+      "grad_norm": 3.91015625,
+      "learning_rate": 8.531687791861241e-05,
+      "loss": 1.6089,
+      "step": 2211
+    },
+    {
+      "epoch": 2.0420032310177705,
+      "grad_norm": 3.75,
+      "learning_rate": 8.531020680453637e-05,
+      "loss": 1.5916,
+      "step": 2212
+    },
+    {
+      "epoch": 2.042926378952227,
+      "grad_norm": 6.578125,
+      "learning_rate": 8.530353569046031e-05,
+      "loss": 1.5728,
+      "step": 2213
+    },
+    {
+      "epoch": 2.0438495268866834,
+      "grad_norm": 4.21484375,
+      "learning_rate": 8.529686457638426e-05,
+      "loss": 1.5081,
+      "step": 2214
+    },
+    {
+      "epoch": 2.04477267482114,
+      "grad_norm": 4.34765625,
+      "learning_rate": 8.52901934623082e-05,
+      "loss": 1.4957,
+      "step": 2215
+    },
+    {
+      "epoch": 2.0456958227555964,
+      "grad_norm": 4.53515625,
+      "learning_rate": 8.528352234823216e-05,
+      "loss": 1.5785,
+      "step": 2216
+    },
+    {
+      "epoch": 2.0466189706900533,
+      "grad_norm": 4.24609375,
+      "learning_rate": 8.527685123415611e-05,
+      "loss": 1.7727,
+      "step": 2217
+    },
+    {
+      "epoch": 2.0475421186245097,
+      "grad_norm": 4.97265625,
+      "learning_rate": 8.527018012008005e-05,
+      "loss": 1.8245,
+      "step": 2218
+    },
+    {
+      "epoch": 2.048465266558966,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.5263509006004e-05,
+      "loss": 1.3994,
+      "step": 2219
+    },
+    {
+      "epoch": 2.0493884144934227,
+      "grad_norm": 6.46875,
+      "learning_rate": 8.525683789192796e-05,
+      "loss": 1.3662,
+      "step": 2220
+    },
+    {
+      "epoch": 2.050311562427879,
+      "grad_norm": 6.1953125,
+      "learning_rate": 8.52501667778519e-05,
+      "loss": 1.5132,
+      "step": 2221
+    },
+    {
+      "epoch": 2.0512347103623356,
+      "grad_norm": 6.59375,
+      "learning_rate": 8.524349566377585e-05,
+      "loss": 1.3613,
+      "step": 2222
+    },
+    {
+      "epoch": 2.052157858296792,
+      "grad_norm": 5.328125,
+      "learning_rate": 8.52368245496998e-05,
+      "loss": 1.8169,
+      "step": 2223
+    },
+    {
+      "epoch": 2.0530810062312486,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.523015343562375e-05,
+      "loss": 1.6216,
+      "step": 2224
+    },
+    {
+      "epoch": 2.054004154165705,
+      "grad_norm": 4.8671875,
+      "learning_rate": 8.52234823215477e-05,
+      "loss": 1.2705,
+      "step": 2225
+    },
+    {
+      "epoch": 2.0549273021001615,
+      "grad_norm": 4.45703125,
+      "learning_rate": 8.521681120747164e-05,
+      "loss": 1.5537,
+      "step": 2226
+    },
+    {
+      "epoch": 2.055850450034618,
+      "grad_norm": 3.919921875,
+      "learning_rate": 8.521014009339559e-05,
+      "loss": 1.4277,
+      "step": 2227
+    },
+    {
+      "epoch": 2.0567735979690744,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.520346897931955e-05,
+      "loss": 1.2869,
+      "step": 2228
+    },
+    {
+      "epoch": 2.057696745903531,
+      "grad_norm": 4.40625,
+      "learning_rate": 8.519679786524351e-05,
+      "loss": 1.584,
+      "step": 2229
+    },
+    {
+      "epoch": 2.0586198938379874,
+      "grad_norm": 4.765625,
+      "learning_rate": 8.519012675116745e-05,
+      "loss": 1.5471,
+      "step": 2230
+    },
+    {
+      "epoch": 2.059543041772444,
+      "grad_norm": 5.79296875,
+      "learning_rate": 8.51834556370914e-05,
+      "loss": 1.6709,
+      "step": 2231
+    },
+    {
+      "epoch": 2.0604661897069003,
+      "grad_norm": 5.296875,
+      "learning_rate": 8.517678452301536e-05,
+      "loss": 1.3291,
+      "step": 2232
+    },
+    {
+      "epoch": 2.0613893376413572,
+      "grad_norm": 4.75,
+      "learning_rate": 8.51701134089393e-05,
+      "loss": 1.4214,
+      "step": 2233
+    },
+    {
+      "epoch": 2.0623124855758137,
+      "grad_norm": 5.62109375,
+      "learning_rate": 8.516344229486325e-05,
+      "loss": 1.5298,
+      "step": 2234
+    },
+    {
+      "epoch": 2.06323563351027,
+      "grad_norm": 5.1875,
+      "learning_rate": 8.51567711807872e-05,
+      "loss": 1.322,
+      "step": 2235
+    },
+    {
+      "epoch": 2.0641587814447266,
+      "grad_norm": 6.00390625,
+      "learning_rate": 8.515010006671115e-05,
+      "loss": 1.5796,
+      "step": 2236
+    },
+    {
+      "epoch": 2.065081929379183,
+      "grad_norm": 4.97265625,
+      "learning_rate": 8.51434289526351e-05,
+      "loss": 1.2899,
+      "step": 2237
+    },
+    {
+      "epoch": 2.0660050773136396,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.513675783855904e-05,
+      "loss": 1.5515,
+      "step": 2238
+    },
+    {
+      "epoch": 2.066928225248096,
+      "grad_norm": 3.677734375,
+      "learning_rate": 8.513008672448299e-05,
+      "loss": 1.4844,
+      "step": 2239
+    },
+    {
+      "epoch": 2.0678513731825525,
+      "grad_norm": 4.73828125,
+      "learning_rate": 8.512341561040695e-05,
+      "loss": 1.6062,
+      "step": 2240
+    },
+    {
+      "epoch": 2.068774521117009,
+      "grad_norm": 4.96484375,
+      "learning_rate": 8.51167444963309e-05,
+      "loss": 1.4832,
+      "step": 2241
+    },
+    {
+      "epoch": 2.0696976690514655,
+      "grad_norm": 5.6796875,
+      "learning_rate": 8.511007338225484e-05,
+      "loss": 1.5178,
+      "step": 2242
+    },
+    {
+      "epoch": 2.070620816985922,
+      "grad_norm": 5.33203125,
+      "learning_rate": 8.510340226817878e-05,
+      "loss": 1.2419,
+      "step": 2243
+    },
+    {
+      "epoch": 2.0715439649203784,
+      "grad_norm": 5.2109375,
+      "learning_rate": 8.509673115410274e-05,
+      "loss": 1.5791,
+      "step": 2244
+    },
+    {
+      "epoch": 2.072467112854835,
+      "grad_norm": 3.98046875,
+      "learning_rate": 8.509006004002669e-05,
+      "loss": 1.5249,
+      "step": 2245
+    },
+    {
+      "epoch": 2.0733902607892913,
+      "grad_norm": 3.900390625,
+      "learning_rate": 8.508338892595063e-05,
+      "loss": 1.2,
+      "step": 2246
+    },
+    {
+      "epoch": 2.074313408723748,
+      "grad_norm": 4.7578125,
+      "learning_rate": 8.507671781187458e-05,
+      "loss": 1.4009,
+      "step": 2247
+    },
+    {
+      "epoch": 2.0752365566582043,
+      "grad_norm": 4.01171875,
+      "learning_rate": 8.507004669779854e-05,
+      "loss": 0.9307,
+      "step": 2248
+    },
+    {
+      "epoch": 2.076159704592661,
+      "grad_norm": 5.8203125,
+      "learning_rate": 8.506337558372248e-05,
+      "loss": 1.4883,
+      "step": 2249
+    },
+    {
+      "epoch": 2.0770828525271177,
+      "grad_norm": 4.83984375,
+      "learning_rate": 8.505670446964643e-05,
+      "loss": 1.2366,
+      "step": 2250
+    },
+    {
+      "epoch": 2.078006000461574,
+      "grad_norm": 6.33984375,
+      "learning_rate": 8.505003335557037e-05,
+      "loss": 1.6282,
+      "step": 2251
+    },
+    {
+      "epoch": 2.0789291483960306,
+      "grad_norm": 5.21484375,
+      "learning_rate": 8.504336224149433e-05,
+      "loss": 1.6118,
+      "step": 2252
+    },
+    {
+      "epoch": 2.079852296330487,
+      "grad_norm": 4.1484375,
+      "learning_rate": 8.503669112741828e-05,
+      "loss": 1.2433,
+      "step": 2253
+    },
+    {
+      "epoch": 2.0807754442649435,
+      "grad_norm": 4.91015625,
+      "learning_rate": 8.503002001334224e-05,
+      "loss": 1.502,
+      "step": 2254
+    },
+    {
+      "epoch": 2.0816985921994,
+      "grad_norm": 4.3203125,
+      "learning_rate": 8.502334889926618e-05,
+      "loss": 1.5029,
+      "step": 2255
+    },
+    {
+      "epoch": 2.0826217401338565,
+      "grad_norm": 3.896484375,
+      "learning_rate": 8.501667778519013e-05,
+      "loss": 1.248,
+      "step": 2256
+    },
+    {
+      "epoch": 2.083544888068313,
+      "grad_norm": 4.1171875,
+      "learning_rate": 8.501000667111409e-05,
+      "loss": 1.4751,
+      "step": 2257
+    },
+    {
+      "epoch": 2.0844680360027694,
+      "grad_norm": 4.55859375,
+      "learning_rate": 8.500333555703803e-05,
+      "loss": 1.7261,
+      "step": 2258
+    },
+    {
+      "epoch": 2.085391183937226,
+      "grad_norm": 5.30078125,
+      "learning_rate": 8.499666444296198e-05,
+      "loss": 1.5605,
+      "step": 2259
+    },
+    {
+      "epoch": 2.0863143318716824,
+      "grad_norm": 4.453125,
+      "learning_rate": 8.498999332888592e-05,
+      "loss": 1.4194,
+      "step": 2260
+    },
+    {
+      "epoch": 2.087237479806139,
+      "grad_norm": 3.6953125,
+      "learning_rate": 8.498332221480988e-05,
+      "loss": 1.4973,
+      "step": 2261
+    },
+    {
+      "epoch": 2.0881606277405953,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.497665110073383e-05,
+      "loss": 1.7812,
+      "step": 2262
+    },
+    {
+      "epoch": 2.0890837756750518,
+      "grad_norm": 4.02734375,
+      "learning_rate": 8.496997998665777e-05,
+      "loss": 1.6572,
+      "step": 2263
+    },
+    {
+      "epoch": 2.0900069236095082,
+      "grad_norm": 4.29296875,
+      "learning_rate": 8.496330887258172e-05,
+      "loss": 1.6423,
+      "step": 2264
+    },
+    {
+      "epoch": 2.0909300715439647,
+      "grad_norm": 4.80859375,
+      "learning_rate": 8.495663775850568e-05,
+      "loss": 1.4221,
+      "step": 2265
+    },
+    {
+      "epoch": 2.0918532194784216,
+      "grad_norm": 5.22265625,
+      "learning_rate": 8.494996664442962e-05,
+      "loss": 1.5625,
+      "step": 2266
+    },
+    {
+      "epoch": 2.092776367412878,
+      "grad_norm": 6.140625,
+      "learning_rate": 8.494329553035357e-05,
+      "loss": 1.355,
+      "step": 2267
+    },
+    {
+      "epoch": 2.0936995153473346,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.493662441627751e-05,
+      "loss": 1.7498,
+      "step": 2268
+    },
+    {
+      "epoch": 2.094622663281791,
+      "grad_norm": 5.703125,
+      "learning_rate": 8.492995330220147e-05,
+      "loss": 1.5432,
+      "step": 2269
+    },
+    {
+      "epoch": 2.0955458112162475,
+      "grad_norm": 4.98046875,
+      "learning_rate": 8.492328218812542e-05,
+      "loss": 1.3239,
+      "step": 2270
+    },
+    {
+      "epoch": 2.096468959150704,
+      "grad_norm": 6.69140625,
+      "learning_rate": 8.491661107404936e-05,
+      "loss": 1.6768,
+      "step": 2271
+    },
+    {
+      "epoch": 2.0973921070851604,
+      "grad_norm": 3.73046875,
+      "learning_rate": 8.490993995997332e-05,
+      "loss": 1.4241,
+      "step": 2272
+    },
+    {
+      "epoch": 2.098315255019617,
+      "grad_norm": 4.875,
+      "learning_rate": 8.490326884589727e-05,
+      "loss": 1.5879,
+      "step": 2273
+    },
+    {
+      "epoch": 2.0992384029540734,
+      "grad_norm": 6.203125,
+      "learning_rate": 8.489659773182121e-05,
+      "loss": 1.5686,
+      "step": 2274
+    },
+    {
+      "epoch": 2.10016155088853,
+      "grad_norm": 5.69140625,
+      "learning_rate": 8.488992661774516e-05,
+      "loss": 1.338,
+      "step": 2275
+    },
+    {
+      "epoch": 2.1010846988229863,
+      "grad_norm": 4.640625,
+      "learning_rate": 8.488325550366912e-05,
+      "loss": 1.6492,
+      "step": 2276
+    },
+    {
+      "epoch": 2.102007846757443,
+      "grad_norm": 5.73046875,
+      "learning_rate": 8.487658438959306e-05,
+      "loss": 1.374,
+      "step": 2277
+    },
+    {
+      "epoch": 2.1029309946918993,
+      "grad_norm": 4.62109375,
+      "learning_rate": 8.486991327551701e-05,
+      "loss": 1.5093,
+      "step": 2278
+    },
+    {
+      "epoch": 2.1038541426263557,
+      "grad_norm": 5.37109375,
+      "learning_rate": 8.486324216144097e-05,
+      "loss": 1.5852,
+      "step": 2279
+    },
+    {
+      "epoch": 2.104777290560812,
+      "grad_norm": 6.83984375,
+      "learning_rate": 8.485657104736491e-05,
+      "loss": 1.2148,
+      "step": 2280
+    },
+    {
+      "epoch": 2.1057004384952687,
+      "grad_norm": 5.71875,
+      "learning_rate": 8.484989993328887e-05,
+      "loss": 1.4407,
+      "step": 2281
+    },
+    {
+      "epoch": 2.106623586429725,
+      "grad_norm": 5.125,
+      "learning_rate": 8.484322881921282e-05,
+      "loss": 1.418,
+      "step": 2282
+    },
+    {
+      "epoch": 2.107546734364182,
+      "grad_norm": 5.4375,
+      "learning_rate": 8.483655770513676e-05,
+      "loss": 1.3577,
+      "step": 2283
+    },
+    {
+      "epoch": 2.1084698822986385,
+      "grad_norm": 4.3203125,
+      "learning_rate": 8.482988659106071e-05,
+      "loss": 1.8357,
+      "step": 2284
+    },
+    {
+      "epoch": 2.109393030233095,
+      "grad_norm": 4.640625,
+      "learning_rate": 8.482321547698467e-05,
+      "loss": 1.3389,
+      "step": 2285
+    },
+    {
+      "epoch": 2.1103161781675515,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.481654436290861e-05,
+      "loss": 1.5051,
+      "step": 2286
+    },
+    {
+      "epoch": 2.111239326102008,
+      "grad_norm": 4.79296875,
+      "learning_rate": 8.480987324883256e-05,
+      "loss": 1.3828,
+      "step": 2287
+    },
+    {
+      "epoch": 2.1121624740364644,
+      "grad_norm": 4.55859375,
+      "learning_rate": 8.48032021347565e-05,
+      "loss": 1.5161,
+      "step": 2288
+    },
+    {
+      "epoch": 2.113085621970921,
+      "grad_norm": 4.23828125,
+      "learning_rate": 8.479653102068046e-05,
+      "loss": 1.7229,
+      "step": 2289
+    },
+    {
+      "epoch": 2.1140087699053773,
+      "grad_norm": 5.125,
+      "learning_rate": 8.478985990660441e-05,
+      "loss": 1.5779,
+      "step": 2290
+    },
+    {
+      "epoch": 2.114931917839834,
+      "grad_norm": 6.55078125,
+      "learning_rate": 8.478318879252835e-05,
+      "loss": 1.552,
+      "step": 2291
+    },
+    {
+      "epoch": 2.1158550657742903,
+      "grad_norm": 4.6796875,
+      "learning_rate": 8.47765176784523e-05,
+      "loss": 1.4839,
+      "step": 2292
+    },
+    {
+      "epoch": 2.1167782137087467,
+      "grad_norm": 5.8125,
+      "learning_rate": 8.476984656437626e-05,
+      "loss": 1.7566,
+      "step": 2293
+    },
+    {
+      "epoch": 2.117701361643203,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.47631754503002e-05,
+      "loss": 1.3772,
+      "step": 2294
+    },
+    {
+      "epoch": 2.1186245095776597,
+      "grad_norm": 6.09375,
+      "learning_rate": 8.475650433622415e-05,
+      "loss": 1.3228,
+      "step": 2295
+    },
+    {
+      "epoch": 2.119547657512116,
+      "grad_norm": 5.0,
+      "learning_rate": 8.47498332221481e-05,
+      "loss": 1.6104,
+      "step": 2296
+    },
+    {
+      "epoch": 2.1204708054465726,
+      "grad_norm": 4.84765625,
+      "learning_rate": 8.474316210807205e-05,
+      "loss": 1.4751,
+      "step": 2297
+    },
+    {
+      "epoch": 2.121393953381029,
+      "grad_norm": 5.125,
+      "learning_rate": 8.4736490993996e-05,
+      "loss": 1.4739,
+      "step": 2298
+    },
+    {
+      "epoch": 2.122317101315486,
+      "grad_norm": 3.58203125,
+      "learning_rate": 8.472981987991995e-05,
+      "loss": 1.4316,
+      "step": 2299
+    },
+    {
+      "epoch": 2.1232402492499425,
+      "grad_norm": 4.4765625,
+      "learning_rate": 8.472314876584389e-05,
+      "loss": 1.4153,
+      "step": 2300
+    },
+    {
+      "epoch": 2.124163397184399,
+      "grad_norm": 4.5234375,
+      "learning_rate": 8.471647765176785e-05,
+      "loss": 1.6331,
+      "step": 2301
+    },
+    {
+      "epoch": 2.1250865451188554,
+      "grad_norm": 3.623046875,
+      "learning_rate": 8.47098065376918e-05,
+      "loss": 1.5399,
+      "step": 2302
+    },
+    {
+      "epoch": 2.126009693053312,
+      "grad_norm": 4.55859375,
+      "learning_rate": 8.470313542361574e-05,
+      "loss": 1.3274,
+      "step": 2303
+    },
+    {
+      "epoch": 2.1269328409877684,
+      "grad_norm": 4.33203125,
+      "learning_rate": 8.46964643095397e-05,
+      "loss": 1.3999,
+      "step": 2304
+    },
+    {
+      "epoch": 2.127855988922225,
+      "grad_norm": 4.83203125,
+      "learning_rate": 8.468979319546364e-05,
+      "loss": 1.0906,
+      "step": 2305
+    },
+    {
+      "epoch": 2.1287791368566813,
+      "grad_norm": 4.01171875,
+      "learning_rate": 8.46831220813876e-05,
+      "loss": 1.1834,
+      "step": 2306
+    },
+    {
+      "epoch": 2.1297022847911378,
+      "grad_norm": 5.1875,
+      "learning_rate": 8.467645096731155e-05,
+      "loss": 1.4104,
+      "step": 2307
+    },
+    {
+      "epoch": 2.1306254327255942,
+      "grad_norm": 5.08203125,
+      "learning_rate": 8.46697798532355e-05,
+      "loss": 1.5337,
+      "step": 2308
+    },
+    {
+      "epoch": 2.1315485806600507,
+      "grad_norm": 5.57421875,
+      "learning_rate": 8.466310873915944e-05,
+      "loss": 1.3882,
+      "step": 2309
+    },
+    {
+      "epoch": 2.132471728594507,
+      "grad_norm": 4.56640625,
+      "learning_rate": 8.46564376250834e-05,
+      "loss": 1.4956,
+      "step": 2310
+    },
+    {
+      "epoch": 2.1333948765289636,
+      "grad_norm": 3.587890625,
+      "learning_rate": 8.464976651100734e-05,
+      "loss": 1.2059,
+      "step": 2311
+    },
+    {
+      "epoch": 2.13431802446342,
+      "grad_norm": 5.015625,
+      "learning_rate": 8.464309539693129e-05,
+      "loss": 1.5679,
+      "step": 2312
+    },
+    {
+      "epoch": 2.1352411723978766,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.463642428285525e-05,
+      "loss": 1.7383,
+      "step": 2313
+    },
+    {
+      "epoch": 2.136164320332333,
+      "grad_norm": 4.52734375,
+      "learning_rate": 8.46297531687792e-05,
+      "loss": 1.5017,
+      "step": 2314
+    },
+    {
+      "epoch": 2.13708746826679,
+      "grad_norm": 4.9140625,
+      "learning_rate": 8.462308205470314e-05,
+      "loss": 1.4723,
+      "step": 2315
+    },
+    {
+      "epoch": 2.1380106162012464,
+      "grad_norm": 5.234375,
+      "learning_rate": 8.461641094062709e-05,
+      "loss": 1.6321,
+      "step": 2316
+    },
+    {
+      "epoch": 2.138933764135703,
+      "grad_norm": 4.73046875,
+      "learning_rate": 8.460973982655104e-05,
+      "loss": 1.4231,
+      "step": 2317
+    },
+    {
+      "epoch": 2.1398569120701594,
+      "grad_norm": 3.9765625,
+      "learning_rate": 8.460306871247499e-05,
+      "loss": 1.2401,
+      "step": 2318
+    },
+    {
+      "epoch": 2.140780060004616,
+      "grad_norm": 4.56640625,
+      "learning_rate": 8.459639759839894e-05,
+      "loss": 1.3024,
+      "step": 2319
+    },
+    {
+      "epoch": 2.1417032079390723,
+      "grad_norm": 5.078125,
+      "learning_rate": 8.458972648432288e-05,
+      "loss": 1.521,
+      "step": 2320
+    },
+    {
+      "epoch": 2.142626355873529,
+      "grad_norm": 4.02734375,
+      "learning_rate": 8.458305537024684e-05,
+      "loss": 1.6719,
+      "step": 2321
+    },
+    {
+      "epoch": 2.1435495038079853,
+      "grad_norm": 4.72265625,
+      "learning_rate": 8.457638425617078e-05,
+      "loss": 1.6284,
+      "step": 2322
+    },
+    {
+      "epoch": 2.1444726517424417,
+      "grad_norm": 4.4765625,
+      "learning_rate": 8.456971314209473e-05,
+      "loss": 1.4614,
+      "step": 2323
+    },
+    {
+      "epoch": 2.145395799676898,
+      "grad_norm": 4.5078125,
+      "learning_rate": 8.456304202801868e-05,
+      "loss": 1.6829,
+      "step": 2324
+    },
+    {
+      "epoch": 2.1463189476113547,
+      "grad_norm": 3.5703125,
+      "learning_rate": 8.455637091394263e-05,
+      "loss": 1.5442,
+      "step": 2325
+    },
+    {
+      "epoch": 2.147242095545811,
+      "grad_norm": 4.0859375,
+      "learning_rate": 8.454969979986658e-05,
+      "loss": 1.8213,
+      "step": 2326
+    },
+    {
+      "epoch": 2.1481652434802676,
+      "grad_norm": 4.90234375,
+      "learning_rate": 8.454302868579053e-05,
+      "loss": 1.5061,
+      "step": 2327
+    },
+    {
+      "epoch": 2.149088391414724,
+      "grad_norm": 4.76953125,
+      "learning_rate": 8.453635757171447e-05,
+      "loss": 1.4937,
+      "step": 2328
+    },
+    {
+      "epoch": 2.1500115393491805,
+      "grad_norm": 4.5078125,
+      "learning_rate": 8.452968645763843e-05,
+      "loss": 1.3887,
+      "step": 2329
+    },
+    {
+      "epoch": 2.150934687283637,
+      "grad_norm": 3.787109375,
+      "learning_rate": 8.452301534356239e-05,
+      "loss": 1.1946,
+      "step": 2330
+    },
+    {
+      "epoch": 2.1518578352180935,
+      "grad_norm": 3.99609375,
+      "learning_rate": 8.451634422948633e-05,
+      "loss": 1.4929,
+      "step": 2331
+    },
+    {
+      "epoch": 2.1527809831525504,
+      "grad_norm": 4.39453125,
+      "learning_rate": 8.450967311541028e-05,
+      "loss": 1.4849,
+      "step": 2332
+    },
+    {
+      "epoch": 2.153704131087007,
+      "grad_norm": 5.30859375,
+      "learning_rate": 8.450300200133423e-05,
+      "loss": 1.5713,
+      "step": 2333
+    },
+    {
+      "epoch": 2.1546272790214633,
+      "grad_norm": 6.4296875,
+      "learning_rate": 8.449633088725818e-05,
+      "loss": 1.5342,
+      "step": 2334
+    },
+    {
+      "epoch": 2.15555042695592,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.448965977318213e-05,
+      "loss": 1.4374,
+      "step": 2335
+    },
+    {
+      "epoch": 2.1564735748903763,
+      "grad_norm": 4.9765625,
+      "learning_rate": 8.448298865910608e-05,
+      "loss": 1.3602,
+      "step": 2336
+    },
+    {
+      "epoch": 2.1573967228248327,
+      "grad_norm": 3.732421875,
+      "learning_rate": 8.447631754503002e-05,
+      "loss": 1.5166,
+      "step": 2337
+    },
+    {
+      "epoch": 2.158319870759289,
+      "grad_norm": 5.08984375,
+      "learning_rate": 8.446964643095398e-05,
+      "loss": 0.9401,
+      "step": 2338
+    },
+    {
+      "epoch": 2.1592430186937457,
+      "grad_norm": 4.55078125,
+      "learning_rate": 8.446297531687792e-05,
+      "loss": 1.6052,
+      "step": 2339
+    },
+    {
+      "epoch": 2.160166166628202,
+      "grad_norm": 5.73828125,
+      "learning_rate": 8.445630420280187e-05,
+      "loss": 1.6306,
+      "step": 2340
+    },
+    {
+      "epoch": 2.1610893145626586,
+      "grad_norm": 5.265625,
+      "learning_rate": 8.444963308872582e-05,
+      "loss": 1.4319,
+      "step": 2341
+    },
+    {
+      "epoch": 2.162012462497115,
+      "grad_norm": 5.609375,
+      "learning_rate": 8.444296197464977e-05,
+      "loss": 1.447,
+      "step": 2342
+    },
+    {
+      "epoch": 2.1629356104315716,
+      "grad_norm": 4.51953125,
+      "learning_rate": 8.443629086057372e-05,
+      "loss": 1.2761,
+      "step": 2343
+    },
+    {
+      "epoch": 2.163858758366028,
+      "grad_norm": 4.91796875,
+      "learning_rate": 8.442961974649767e-05,
+      "loss": 1.5227,
+      "step": 2344
+    },
+    {
+      "epoch": 2.1647819063004845,
+      "grad_norm": 4.07421875,
+      "learning_rate": 8.442294863242161e-05,
+      "loss": 1.519,
+      "step": 2345
+    },
+    {
+      "epoch": 2.165705054234941,
+      "grad_norm": 4.0703125,
+      "learning_rate": 8.441627751834557e-05,
+      "loss": 1.6726,
+      "step": 2346
+    },
+    {
+      "epoch": 2.1666282021693974,
+      "grad_norm": 4.13671875,
+      "learning_rate": 8.440960640426952e-05,
+      "loss": 1.5347,
+      "step": 2347
+    },
+    {
+      "epoch": 2.167551350103854,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.440293529019346e-05,
+      "loss": 1.3662,
+      "step": 2348
+    },
+    {
+      "epoch": 2.168474498038311,
+      "grad_norm": 4.4765625,
+      "learning_rate": 8.43962641761174e-05,
+      "loss": 1.46,
+      "step": 2349
+    },
+    {
+      "epoch": 2.1693976459727673,
+      "grad_norm": 4.078125,
+      "learning_rate": 8.438959306204137e-05,
+      "loss": 1.3666,
+      "step": 2350
+    },
+    {
+      "epoch": 2.1703207939072238,
+      "grad_norm": 4.46484375,
+      "learning_rate": 8.438292194796531e-05,
+      "loss": 1.6804,
+      "step": 2351
+    },
+    {
+      "epoch": 2.1712439418416802,
+      "grad_norm": 4.58984375,
+      "learning_rate": 8.437625083388926e-05,
+      "loss": 1.8743,
+      "step": 2352
+    },
+    {
+      "epoch": 2.1721670897761367,
+      "grad_norm": 4.609375,
+      "learning_rate": 8.43695797198132e-05,
+      "loss": 1.4819,
+      "step": 2353
+    },
+    {
+      "epoch": 2.173090237710593,
+      "grad_norm": 4.625,
+      "learning_rate": 8.436290860573717e-05,
+      "loss": 1.6121,
+      "step": 2354
+    },
+    {
+      "epoch": 2.1740133856450496,
+      "grad_norm": 4.07421875,
+      "learning_rate": 8.435623749166112e-05,
+      "loss": 1.5459,
+      "step": 2355
+    },
+    {
+      "epoch": 2.174936533579506,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.434956637758507e-05,
+      "loss": 1.254,
+      "step": 2356
+    },
+    {
+      "epoch": 2.1758596815139626,
+      "grad_norm": 5.140625,
+      "learning_rate": 8.434289526350901e-05,
+      "loss": 1.4644,
+      "step": 2357
+    },
+    {
+      "epoch": 2.176782829448419,
+      "grad_norm": 4.7890625,
+      "learning_rate": 8.433622414943297e-05,
+      "loss": 1.6201,
+      "step": 2358
+    },
+    {
+      "epoch": 2.1777059773828755,
+      "grad_norm": 3.611328125,
+      "learning_rate": 8.432955303535691e-05,
+      "loss": 1.4456,
+      "step": 2359
+    },
+    {
+      "epoch": 2.178629125317332,
+      "grad_norm": 4.70703125,
+      "learning_rate": 8.432288192128086e-05,
+      "loss": 1.9114,
+      "step": 2360
+    },
+    {
+      "epoch": 2.1795522732517885,
+      "grad_norm": 4.55859375,
+      "learning_rate": 8.43162108072048e-05,
+      "loss": 1.4333,
+      "step": 2361
+    },
+    {
+      "epoch": 2.180475421186245,
+      "grad_norm": 5.421875,
+      "learning_rate": 8.430953969312876e-05,
+      "loss": 1.6689,
+      "step": 2362
+    },
+    {
+      "epoch": 2.1813985691207014,
+      "grad_norm": 5.44140625,
+      "learning_rate": 8.430286857905271e-05,
+      "loss": 1.646,
+      "step": 2363
+    },
+    {
+      "epoch": 2.1823217170551583,
+      "grad_norm": 5.1640625,
+      "learning_rate": 8.429619746497666e-05,
+      "loss": 1.4409,
+      "step": 2364
+    },
+    {
+      "epoch": 2.183244864989615,
+      "grad_norm": 3.880859375,
+      "learning_rate": 8.42895263509006e-05,
+      "loss": 1.5034,
+      "step": 2365
+    },
+    {
+      "epoch": 2.1841680129240713,
+      "grad_norm": 3.724609375,
+      "learning_rate": 8.428285523682456e-05,
+      "loss": 1.6375,
+      "step": 2366
+    },
+    {
+      "epoch": 2.1850911608585277,
+      "grad_norm": 4.16796875,
+      "learning_rate": 8.42761841227485e-05,
+      "loss": 1.5552,
+      "step": 2367
+    },
+    {
+      "epoch": 2.186014308792984,
+      "grad_norm": 4.765625,
+      "learning_rate": 8.426951300867245e-05,
+      "loss": 1.2758,
+      "step": 2368
+    },
+    {
+      "epoch": 2.1869374567274407,
+      "grad_norm": 5.55078125,
+      "learning_rate": 8.42628418945964e-05,
+      "loss": 1.6489,
+      "step": 2369
+    },
+    {
+      "epoch": 2.187860604661897,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.425617078052036e-05,
+      "loss": 1.3535,
+      "step": 2370
+    },
+    {
+      "epoch": 2.1887837525963536,
+      "grad_norm": 4.48828125,
+      "learning_rate": 8.42494996664443e-05,
+      "loss": 1.2598,
+      "step": 2371
+    },
+    {
+      "epoch": 2.18970690053081,
+      "grad_norm": 5.0234375,
+      "learning_rate": 8.424282855236825e-05,
+      "loss": 1.429,
+      "step": 2372
+    },
+    {
+      "epoch": 2.1906300484652665,
+      "grad_norm": 4.98046875,
+      "learning_rate": 8.423615743829219e-05,
+      "loss": 1.4668,
+      "step": 2373
+    },
+    {
+      "epoch": 2.191553196399723,
+      "grad_norm": 5.703125,
+      "learning_rate": 8.422948632421615e-05,
+      "loss": 1.2263,
+      "step": 2374
+    },
+    {
+      "epoch": 2.1924763443341795,
+      "grad_norm": 4.37890625,
+      "learning_rate": 8.42228152101401e-05,
+      "loss": 1.5513,
+      "step": 2375
+    },
+    {
+      "epoch": 2.193399492268636,
+      "grad_norm": 4.46484375,
+      "learning_rate": 8.421614409606404e-05,
+      "loss": 1.6943,
+      "step": 2376
+    },
+    {
+      "epoch": 2.1943226402030924,
+      "grad_norm": 4.17578125,
+      "learning_rate": 8.420947298198799e-05,
+      "loss": 1.4299,
+      "step": 2377
+    },
+    {
+      "epoch": 2.195245788137549,
+      "grad_norm": 4.5703125,
+      "learning_rate": 8.420280186791195e-05,
+      "loss": 1.5957,
+      "step": 2378
+    },
+    {
+      "epoch": 2.1961689360720054,
+      "grad_norm": 4.765625,
+      "learning_rate": 8.41961307538359e-05,
+      "loss": 1.5867,
+      "step": 2379
+    },
+    {
+      "epoch": 2.197092084006462,
+      "grad_norm": 4.9765625,
+      "learning_rate": 8.418945963975985e-05,
+      "loss": 1.4575,
+      "step": 2380
+    },
+    {
+      "epoch": 2.1980152319409187,
+      "grad_norm": 5.1484375,
+      "learning_rate": 8.41827885256838e-05,
+      "loss": 1.575,
+      "step": 2381
+    },
+    {
+      "epoch": 2.198938379875375,
+      "grad_norm": 5.58984375,
+      "learning_rate": 8.417611741160774e-05,
+      "loss": 1.6509,
+      "step": 2382
+    },
+    {
+      "epoch": 2.1998615278098317,
+      "grad_norm": 3.95703125,
+      "learning_rate": 8.41694462975317e-05,
+      "loss": 1.8569,
+      "step": 2383
+    },
+    {
+      "epoch": 2.200784675744288,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.416277518345565e-05,
+      "loss": 1.4727,
+      "step": 2384
+    },
+    {
+      "epoch": 2.2017078236787446,
+      "grad_norm": 4.3515625,
+      "learning_rate": 8.415610406937959e-05,
+      "loss": 1.3162,
+      "step": 2385
+    },
+    {
+      "epoch": 2.202630971613201,
+      "grad_norm": 3.74609375,
+      "learning_rate": 8.414943295530354e-05,
+      "loss": 1.5483,
+      "step": 2386
+    },
+    {
+      "epoch": 2.2035541195476576,
+      "grad_norm": 4.09765625,
+      "learning_rate": 8.41427618412275e-05,
+      "loss": 1.5181,
+      "step": 2387
+    },
+    {
+      "epoch": 2.204477267482114,
+      "grad_norm": 6.51171875,
+      "learning_rate": 8.413609072715144e-05,
+      "loss": 1.8325,
+      "step": 2388
+    },
+    {
+      "epoch": 2.2054004154165705,
+      "grad_norm": 5.3359375,
+      "learning_rate": 8.412941961307539e-05,
+      "loss": 1.6794,
+      "step": 2389
+    },
+    {
+      "epoch": 2.206323563351027,
+      "grad_norm": 6.26953125,
+      "learning_rate": 8.412274849899933e-05,
+      "loss": 1.613,
+      "step": 2390
+    },
+    {
+      "epoch": 2.2072467112854834,
+      "grad_norm": 3.98828125,
+      "learning_rate": 8.411607738492329e-05,
+      "loss": 1.5112,
+      "step": 2391
+    },
+    {
+      "epoch": 2.20816985921994,
+      "grad_norm": 4.80859375,
+      "learning_rate": 8.410940627084724e-05,
+      "loss": 1.5128,
+      "step": 2392
+    },
+    {
+      "epoch": 2.2090930071543964,
+      "grad_norm": 4.9375,
+      "learning_rate": 8.410273515677118e-05,
+      "loss": 1.2646,
+      "step": 2393
+    },
+    {
+      "epoch": 2.210016155088853,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.409606404269513e-05,
+      "loss": 1.5688,
+      "step": 2394
+    },
+    {
+      "epoch": 2.2109393030233093,
+      "grad_norm": 4.04296875,
+      "learning_rate": 8.408939292861909e-05,
+      "loss": 1.5505,
+      "step": 2395
+    },
+    {
+      "epoch": 2.211862450957766,
+      "grad_norm": 4.19140625,
+      "learning_rate": 8.408272181454303e-05,
+      "loss": 1.5156,
+      "step": 2396
+    },
+    {
+      "epoch": 2.2127855988922223,
+      "grad_norm": 4.16796875,
+      "learning_rate": 8.407605070046698e-05,
+      "loss": 1.4268,
+      "step": 2397
+    },
+    {
+      "epoch": 2.213708746826679,
+      "grad_norm": 4.5234375,
+      "learning_rate": 8.406937958639092e-05,
+      "loss": 1.5071,
+      "step": 2398
+    },
+    {
+      "epoch": 2.2146318947611356,
+      "grad_norm": 5.2890625,
+      "learning_rate": 8.406270847231488e-05,
+      "loss": 1.2993,
+      "step": 2399
+    },
+    {
+      "epoch": 2.215555042695592,
+      "grad_norm": 4.14453125,
+      "learning_rate": 8.405603735823883e-05,
+      "loss": 1.3821,
+      "step": 2400
+    },
+    {
+      "epoch": 2.2164781906300486,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.404936624416277e-05,
+      "loss": 1.1528,
+      "step": 2401
+    },
+    {
+      "epoch": 2.217401338564505,
+      "grad_norm": 3.958984375,
+      "learning_rate": 8.404269513008672e-05,
+      "loss": 1.2159,
+      "step": 2402
+    },
+    {
+      "epoch": 2.2183244864989615,
+      "grad_norm": 4.53515625,
+      "learning_rate": 8.403602401601068e-05,
+      "loss": 1.1405,
+      "step": 2403
+    },
+    {
+      "epoch": 2.219247634433418,
+      "grad_norm": 5.98828125,
+      "learning_rate": 8.402935290193464e-05,
+      "loss": 1.7131,
+      "step": 2404
+    },
+    {
+      "epoch": 2.2201707823678745,
+      "grad_norm": 4.33203125,
+      "learning_rate": 8.402268178785858e-05,
+      "loss": 1.4014,
+      "step": 2405
+    },
+    {
+      "epoch": 2.221093930302331,
+      "grad_norm": 3.85546875,
+      "learning_rate": 8.401601067378253e-05,
+      "loss": 1.4309,
+      "step": 2406
+    },
+    {
+      "epoch": 2.2220170782367874,
+      "grad_norm": 4.08984375,
+      "learning_rate": 8.400933955970649e-05,
+      "loss": 1.3591,
+      "step": 2407
+    },
+    {
+      "epoch": 2.222940226171244,
+      "grad_norm": 4.15234375,
+      "learning_rate": 8.400266844563043e-05,
+      "loss": 1.4993,
+      "step": 2408
+    },
+    {
+      "epoch": 2.2238633741057003,
+      "grad_norm": 4.4375,
+      "learning_rate": 8.399599733155438e-05,
+      "loss": 1.531,
+      "step": 2409
+    },
+    {
+      "epoch": 2.224786522040157,
+      "grad_norm": 4.546875,
+      "learning_rate": 8.398932621747832e-05,
+      "loss": 1.5879,
+      "step": 2410
+    },
+    {
+      "epoch": 2.2257096699746133,
+      "grad_norm": 4.125,
+      "learning_rate": 8.398265510340228e-05,
+      "loss": 1.4006,
+      "step": 2411
+    },
+    {
+      "epoch": 2.2266328179090698,
+      "grad_norm": 5.12109375,
+      "learning_rate": 8.397598398932623e-05,
+      "loss": 1.3184,
+      "step": 2412
+    },
+    {
+      "epoch": 2.227555965843526,
+      "grad_norm": 5.11328125,
+      "learning_rate": 8.396931287525017e-05,
+      "loss": 1.3762,
+      "step": 2413
+    },
+    {
+      "epoch": 2.2284791137779827,
+      "grad_norm": 5.4921875,
+      "learning_rate": 8.396264176117412e-05,
+      "loss": 1.457,
+      "step": 2414
+    },
+    {
+      "epoch": 2.2294022617124396,
+      "grad_norm": 4.10546875,
+      "learning_rate": 8.395597064709808e-05,
+      "loss": 1.4197,
+      "step": 2415
+    },
+    {
+      "epoch": 2.230325409646896,
+      "grad_norm": 3.7890625,
+      "learning_rate": 8.394929953302202e-05,
+      "loss": 1.3253,
+      "step": 2416
+    },
+    {
+      "epoch": 2.2312485575813525,
+      "grad_norm": 4.0234375,
+      "learning_rate": 8.394262841894597e-05,
+      "loss": 1.4452,
+      "step": 2417
+    },
+    {
+      "epoch": 2.232171705515809,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.393595730486991e-05,
+      "loss": 1.5981,
+      "step": 2418
+    },
+    {
+      "epoch": 2.2330948534502655,
+      "grad_norm": 5.453125,
+      "learning_rate": 8.392928619079387e-05,
+      "loss": 1.2185,
+      "step": 2419
+    },
+    {
+      "epoch": 2.234018001384722,
+      "grad_norm": 5.3203125,
+      "learning_rate": 8.392261507671782e-05,
+      "loss": 1.4294,
+      "step": 2420
+    },
+    {
+      "epoch": 2.2349411493191784,
+      "grad_norm": 4.17578125,
+      "learning_rate": 8.391594396264176e-05,
+      "loss": 1.511,
+      "step": 2421
+    },
+    {
+      "epoch": 2.235864297253635,
+      "grad_norm": 4.1328125,
+      "learning_rate": 8.390927284856571e-05,
+      "loss": 1.3521,
+      "step": 2422
+    },
+    {
+      "epoch": 2.2367874451880914,
+      "grad_norm": 4.1640625,
+      "learning_rate": 8.390260173448967e-05,
+      "loss": 1.3657,
+      "step": 2423
+    },
+    {
+      "epoch": 2.237710593122548,
+      "grad_norm": 4.6796875,
+      "learning_rate": 8.389593062041361e-05,
+      "loss": 1.3954,
+      "step": 2424
+    },
+    {
+      "epoch": 2.2386337410570043,
+      "grad_norm": 5.1953125,
+      "learning_rate": 8.388925950633756e-05,
+      "loss": 1.4888,
+      "step": 2425
+    },
+    {
+      "epoch": 2.2395568889914608,
+      "grad_norm": 4.5078125,
+      "learning_rate": 8.38825883922615e-05,
+      "loss": 1.3418,
+      "step": 2426
+    },
+    {
+      "epoch": 2.2404800369259172,
+      "grad_norm": 3.783203125,
+      "learning_rate": 8.387591727818546e-05,
+      "loss": 1.6799,
+      "step": 2427
+    },
+    {
+      "epoch": 2.2414031848603737,
+      "grad_norm": 4.40625,
+      "learning_rate": 8.386924616410941e-05,
+      "loss": 1.4065,
+      "step": 2428
+    },
+    {
+      "epoch": 2.24232633279483,
+      "grad_norm": 3.974609375,
+      "learning_rate": 8.386257505003337e-05,
+      "loss": 1.5503,
+      "step": 2429
+    },
+    {
+      "epoch": 2.243249480729287,
+      "grad_norm": 5.3203125,
+      "learning_rate": 8.385590393595731e-05,
+      "loss": 1.6384,
+      "step": 2430
+    },
+    {
+      "epoch": 2.2441726286637436,
+      "grad_norm": 4.25390625,
+      "learning_rate": 8.384923282188126e-05,
+      "loss": 1.6843,
+      "step": 2431
+    },
+    {
+      "epoch": 2.2450957765982,
+      "grad_norm": 6.2265625,
+      "learning_rate": 8.384256170780522e-05,
+      "loss": 1.5165,
+      "step": 2432
+    },
+    {
+      "epoch": 2.2460189245326565,
+      "grad_norm": 4.484375,
+      "learning_rate": 8.383589059372916e-05,
+      "loss": 1.5437,
+      "step": 2433
+    },
+    {
+      "epoch": 2.246942072467113,
+      "grad_norm": 4.84765625,
+      "learning_rate": 8.382921947965311e-05,
+      "loss": 1.3677,
+      "step": 2434
+    },
+    {
+      "epoch": 2.2478652204015694,
+      "grad_norm": 4.5546875,
+      "learning_rate": 8.382254836557705e-05,
+      "loss": 1.686,
+      "step": 2435
+    },
+    {
+      "epoch": 2.248788368336026,
+      "grad_norm": 5.32421875,
+      "learning_rate": 8.381587725150101e-05,
+      "loss": 1.2866,
+      "step": 2436
+    },
+    {
+      "epoch": 2.2497115162704824,
+      "grad_norm": 5.22265625,
+      "learning_rate": 8.380920613742496e-05,
+      "loss": 1.4436,
+      "step": 2437
+    },
+    {
+      "epoch": 2.250634664204939,
+      "grad_norm": 3.6796875,
+      "learning_rate": 8.38025350233489e-05,
+      "loss": 1.6453,
+      "step": 2438
+    },
+    {
+      "epoch": 2.2515578121393953,
+      "grad_norm": 4.734375,
+      "learning_rate": 8.379586390927285e-05,
+      "loss": 1.3206,
+      "step": 2439
+    },
+    {
+      "epoch": 2.252480960073852,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.37891927951968e-05,
+      "loss": 1.2177,
+      "step": 2440
+    },
+    {
+      "epoch": 2.2534041080083083,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.378252168112075e-05,
+      "loss": 1.396,
+      "step": 2441
+    },
+    {
+      "epoch": 2.2543272559427647,
+      "grad_norm": 6.59765625,
+      "learning_rate": 8.37758505670447e-05,
+      "loss": 1.5542,
+      "step": 2442
+    },
+    {
+      "epoch": 2.255250403877221,
+      "grad_norm": 3.755859375,
+      "learning_rate": 8.376917945296864e-05,
+      "loss": 1.3971,
+      "step": 2443
+    },
+    {
+      "epoch": 2.2561735518116777,
+      "grad_norm": 3.951171875,
+      "learning_rate": 8.37625083388926e-05,
+      "loss": 1.405,
+      "step": 2444
+    },
+    {
+      "epoch": 2.257096699746134,
+      "grad_norm": 4.48828125,
+      "learning_rate": 8.375583722481655e-05,
+      "loss": 1.3469,
+      "step": 2445
+    },
+    {
+      "epoch": 2.2580198476805906,
+      "grad_norm": 4.41796875,
+      "learning_rate": 8.374916611074049e-05,
+      "loss": 1.385,
+      "step": 2446
+    },
+    {
+      "epoch": 2.2589429956150475,
+      "grad_norm": 4.98046875,
+      "learning_rate": 8.374249499666444e-05,
+      "loss": 1.6958,
+      "step": 2447
+    },
+    {
+      "epoch": 2.259866143549504,
+      "grad_norm": 4.93359375,
+      "learning_rate": 8.37358238825884e-05,
+      "loss": 1.2148,
+      "step": 2448
+    },
+    {
+      "epoch": 2.2607892914839605,
+      "grad_norm": 5.41796875,
+      "learning_rate": 8.372915276851234e-05,
+      "loss": 1.5354,
+      "step": 2449
+    },
+    {
+      "epoch": 2.261712439418417,
+      "grad_norm": 5.6640625,
+      "learning_rate": 8.372248165443629e-05,
+      "loss": 1.4514,
+      "step": 2450
+    },
+    {
+      "epoch": 2.2626355873528734,
+      "grad_norm": 5.859375,
+      "learning_rate": 8.371581054036023e-05,
+      "loss": 1.656,
+      "step": 2451
+    },
+    {
+      "epoch": 2.26355873528733,
+      "grad_norm": 6.1171875,
+      "learning_rate": 8.370913942628419e-05,
+      "loss": 1.4805,
+      "step": 2452
+    },
+    {
+      "epoch": 2.2644818832217863,
+      "grad_norm": 4.7421875,
+      "learning_rate": 8.370246831220814e-05,
+      "loss": 1.6006,
+      "step": 2453
+    },
+    {
+      "epoch": 2.265405031156243,
+      "grad_norm": 3.80859375,
+      "learning_rate": 8.369579719813208e-05,
+      "loss": 1.1958,
+      "step": 2454
+    },
+    {
+      "epoch": 2.2663281790906993,
+      "grad_norm": 5.26953125,
+      "learning_rate": 8.368912608405604e-05,
+      "loss": 1.3326,
+      "step": 2455
+    },
+    {
+      "epoch": 2.2672513270251557,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.368245496998e-05,
+      "loss": 1.4231,
+      "step": 2456
+    },
+    {
+      "epoch": 2.268174474959612,
+      "grad_norm": 4.89453125,
+      "learning_rate": 8.367578385590395e-05,
+      "loss": 1.4675,
+      "step": 2457
+    },
+    {
+      "epoch": 2.2690976228940687,
+      "grad_norm": 5.83203125,
+      "learning_rate": 8.366911274182789e-05,
+      "loss": 1.4978,
+      "step": 2458
+    },
+    {
+      "epoch": 2.270020770828525,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.366244162775184e-05,
+      "loss": 1.3333,
+      "step": 2459
+    },
+    {
+      "epoch": 2.2709439187629816,
+      "grad_norm": 6.1171875,
+      "learning_rate": 8.36557705136758e-05,
+      "loss": 1.6372,
+      "step": 2460
+    },
+    {
+      "epoch": 2.271867066697438,
+      "grad_norm": 4.57421875,
+      "learning_rate": 8.364909939959974e-05,
+      "loss": 1.4822,
+      "step": 2461
+    },
+    {
+      "epoch": 2.2727902146318946,
+      "grad_norm": 5.2578125,
+      "learning_rate": 8.364242828552369e-05,
+      "loss": 1.6289,
+      "step": 2462
+    },
+    {
+      "epoch": 2.273713362566351,
+      "grad_norm": 5.14453125,
+      "learning_rate": 8.363575717144763e-05,
+      "loss": 1.5869,
+      "step": 2463
+    },
+    {
+      "epoch": 2.274636510500808,
+      "grad_norm": 5.140625,
+      "learning_rate": 8.362908605737159e-05,
+      "loss": 1.6111,
+      "step": 2464
+    },
+    {
+      "epoch": 2.2755596584352644,
+      "grad_norm": 5.390625,
+      "learning_rate": 8.362241494329554e-05,
+      "loss": 1.5205,
+      "step": 2465
+    },
+    {
+      "epoch": 2.276482806369721,
+      "grad_norm": 4.97265625,
+      "learning_rate": 8.361574382921948e-05,
+      "loss": 1.3024,
+      "step": 2466
+    },
+    {
+      "epoch": 2.2774059543041774,
+      "grad_norm": 4.64453125,
+      "learning_rate": 8.360907271514343e-05,
+      "loss": 1.4202,
+      "step": 2467
+    },
+    {
+      "epoch": 2.278329102238634,
+      "grad_norm": 5.63671875,
+      "learning_rate": 8.360240160106739e-05,
+      "loss": 1.615,
+      "step": 2468
+    },
+    {
+      "epoch": 2.2792522501730903,
+      "grad_norm": 5.11328125,
+      "learning_rate": 8.359573048699133e-05,
+      "loss": 1.5874,
+      "step": 2469
+    },
+    {
+      "epoch": 2.2801753981075468,
+      "grad_norm": 4.30859375,
+      "learning_rate": 8.358905937291528e-05,
+      "loss": 1.3887,
+      "step": 2470
+    },
+    {
+      "epoch": 2.2810985460420032,
+      "grad_norm": 4.61328125,
+      "learning_rate": 8.358238825883922e-05,
+      "loss": 1.365,
+      "step": 2471
+    },
+    {
+      "epoch": 2.2820216939764597,
+      "grad_norm": 4.74609375,
+      "learning_rate": 8.357571714476318e-05,
+      "loss": 1.6296,
+      "step": 2472
+    },
+    {
+      "epoch": 2.282944841910916,
+      "grad_norm": 4.10546875,
+      "learning_rate": 8.356904603068713e-05,
+      "loss": 1.4961,
+      "step": 2473
+    },
+    {
+      "epoch": 2.2838679898453726,
+      "grad_norm": 3.6171875,
+      "learning_rate": 8.356237491661107e-05,
+      "loss": 1.2499,
+      "step": 2474
+    },
+    {
+      "epoch": 2.284791137779829,
+      "grad_norm": 3.837890625,
+      "learning_rate": 8.355570380253502e-05,
+      "loss": 1.4819,
+      "step": 2475
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 3.791015625,
+      "learning_rate": 8.354903268845898e-05,
+      "loss": 1.5034,
+      "step": 2476
+    },
+    {
+      "epoch": 2.286637433648742,
+      "grad_norm": 3.830078125,
+      "learning_rate": 8.354236157438292e-05,
+      "loss": 1.6208,
+      "step": 2477
+    },
+    {
+      "epoch": 2.2875605815831985,
+      "grad_norm": 3.744140625,
+      "learning_rate": 8.353569046030687e-05,
+      "loss": 1.3408,
+      "step": 2478
+    },
+    {
+      "epoch": 2.2884837295176554,
+      "grad_norm": 4.6015625,
+      "learning_rate": 8.352901934623081e-05,
+      "loss": 1.6577,
+      "step": 2479
+    },
+    {
+      "epoch": 2.2894068774521115,
+      "grad_norm": 4.4296875,
+      "learning_rate": 8.352234823215477e-05,
+      "loss": 1.3862,
+      "step": 2480
+    },
+    {
+      "epoch": 2.2903300253865684,
+      "grad_norm": 4.5,
+      "learning_rate": 8.351567711807873e-05,
+      "loss": 1.427,
+      "step": 2481
+    },
+    {
+      "epoch": 2.291253173321025,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.350900600400268e-05,
+      "loss": 1.3877,
+      "step": 2482
+    },
+    {
+      "epoch": 2.2921763212554813,
+      "grad_norm": 4.0859375,
+      "learning_rate": 8.350233488992662e-05,
+      "loss": 1.1031,
+      "step": 2483
+    },
+    {
+      "epoch": 2.293099469189938,
+      "grad_norm": 4.08203125,
+      "learning_rate": 8.349566377585057e-05,
+      "loss": 1.4341,
+      "step": 2484
+    },
+    {
+      "epoch": 2.2940226171243943,
+      "grad_norm": 4.2109375,
+      "learning_rate": 8.348899266177453e-05,
+      "loss": 1.5333,
+      "step": 2485
+    },
+    {
+      "epoch": 2.2949457650588507,
+      "grad_norm": 3.916015625,
+      "learning_rate": 8.348232154769847e-05,
+      "loss": 1.4875,
+      "step": 2486
+    },
+    {
+      "epoch": 2.295868912993307,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.347565043362242e-05,
+      "loss": 1.2152,
+      "step": 2487
+    },
+    {
+      "epoch": 2.2967920609277637,
+      "grad_norm": 4.64453125,
+      "learning_rate": 8.346897931954636e-05,
+      "loss": 1.4258,
+      "step": 2488
+    },
+    {
+      "epoch": 2.29771520886222,
+      "grad_norm": 4.640625,
+      "learning_rate": 8.346230820547032e-05,
+      "loss": 1.304,
+      "step": 2489
+    },
+    {
+      "epoch": 2.2986383567966766,
+      "grad_norm": 4.3203125,
+      "learning_rate": 8.345563709139427e-05,
+      "loss": 1.4202,
+      "step": 2490
+    },
+    {
+      "epoch": 2.299561504731133,
+      "grad_norm": 4.30859375,
+      "learning_rate": 8.344896597731821e-05,
+      "loss": 1.3083,
+      "step": 2491
+    },
+    {
+      "epoch": 2.3004846526655895,
+      "grad_norm": 5.75390625,
+      "learning_rate": 8.344229486324216e-05,
+      "loss": 1.4697,
+      "step": 2492
+    },
+    {
+      "epoch": 2.301407800600046,
+      "grad_norm": 5.03515625,
+      "learning_rate": 8.343562374916612e-05,
+      "loss": 1.5564,
+      "step": 2493
+    },
+    {
+      "epoch": 2.3023309485345025,
+      "grad_norm": 6.22265625,
+      "learning_rate": 8.342895263509006e-05,
+      "loss": 1.2419,
+      "step": 2494
+    },
+    {
+      "epoch": 2.303254096468959,
+      "grad_norm": 4.6328125,
+      "learning_rate": 8.342228152101401e-05,
+      "loss": 1.6042,
+      "step": 2495
+    },
+    {
+      "epoch": 2.304177244403416,
+      "grad_norm": 5.203125,
+      "learning_rate": 8.341561040693797e-05,
+      "loss": 1.7007,
+      "step": 2496
+    },
+    {
+      "epoch": 2.305100392337872,
+      "grad_norm": 5.3984375,
+      "learning_rate": 8.340893929286191e-05,
+      "loss": 1.1343,
+      "step": 2497
+    },
+    {
+      "epoch": 2.306023540272329,
+      "grad_norm": 4.984375,
+      "learning_rate": 8.340226817878586e-05,
+      "loss": 1.2158,
+      "step": 2498
+    },
+    {
+      "epoch": 2.3069466882067853,
+      "grad_norm": 5.50390625,
+      "learning_rate": 8.33955970647098e-05,
+      "loss": 1.5098,
+      "step": 2499
+    },
+    {
+      "epoch": 2.3078698361412417,
+      "grad_norm": 4.20703125,
+      "learning_rate": 8.338892595063376e-05,
+      "loss": 1.8418,
+      "step": 2500
+    },
+    {
+      "epoch": 2.308792984075698,
+      "grad_norm": 4.5859375,
+      "learning_rate": 8.338225483655771e-05,
+      "loss": 1.6401,
+      "step": 2501
+    },
+    {
+      "epoch": 2.3097161320101547,
+      "grad_norm": 4.06640625,
+      "learning_rate": 8.337558372248165e-05,
+      "loss": 1.5615,
+      "step": 2502
+    },
+    {
+      "epoch": 2.310639279944611,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.33689126084056e-05,
+      "loss": 1.1228,
+      "step": 2503
+    },
+    {
+      "epoch": 2.3115624278790676,
+      "grad_norm": 4.921875,
+      "learning_rate": 8.336224149432956e-05,
+      "loss": 1.5657,
+      "step": 2504
+    },
+    {
+      "epoch": 2.312485575813524,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.335557038025352e-05,
+      "loss": 1.4546,
+      "step": 2505
+    },
+    {
+      "epoch": 2.3134087237479806,
+      "grad_norm": 4.50390625,
+      "learning_rate": 8.334889926617746e-05,
+      "loss": 1.1039,
+      "step": 2506
+    },
+    {
+      "epoch": 2.314331871682437,
+      "grad_norm": 4.50390625,
+      "learning_rate": 8.334222815210141e-05,
+      "loss": 1.7339,
+      "step": 2507
+    },
+    {
+      "epoch": 2.3152550196168935,
+      "grad_norm": 5.64453125,
+      "learning_rate": 8.333555703802535e-05,
+      "loss": 1.6208,
+      "step": 2508
+    },
+    {
+      "epoch": 2.31617816755135,
+      "grad_norm": 4.20703125,
+      "learning_rate": 8.332888592394931e-05,
+      "loss": 1.5115,
+      "step": 2509
+    },
+    {
+      "epoch": 2.3171013154858064,
+      "grad_norm": 5.20703125,
+      "learning_rate": 8.332221480987326e-05,
+      "loss": 1.7336,
+      "step": 2510
+    },
+    {
+      "epoch": 2.318024463420263,
+      "grad_norm": 5.484375,
+      "learning_rate": 8.33155436957972e-05,
+      "loss": 1.4597,
+      "step": 2511
+    },
+    {
+      "epoch": 2.3189476113547194,
+      "grad_norm": 4.0234375,
+      "learning_rate": 8.330887258172115e-05,
+      "loss": 1.2263,
+      "step": 2512
+    },
+    {
+      "epoch": 2.3198707592891763,
+      "grad_norm": 4.69921875,
+      "learning_rate": 8.330220146764511e-05,
+      "loss": 1.6284,
+      "step": 2513
+    },
+    {
+      "epoch": 2.3207939072236328,
+      "grad_norm": 3.767578125,
+      "learning_rate": 8.329553035356905e-05,
+      "loss": 1.5332,
+      "step": 2514
+    },
+    {
+      "epoch": 2.3217170551580892,
+      "grad_norm": 4.11328125,
+      "learning_rate": 8.3288859239493e-05,
+      "loss": 1.4338,
+      "step": 2515
+    },
+    {
+      "epoch": 2.3226402030925457,
+      "grad_norm": 3.6796875,
+      "learning_rate": 8.328218812541694e-05,
+      "loss": 1.3147,
+      "step": 2516
+    },
+    {
+      "epoch": 2.323563351027002,
+      "grad_norm": 4.0390625,
+      "learning_rate": 8.32755170113409e-05,
+      "loss": 1.4702,
+      "step": 2517
+    },
+    {
+      "epoch": 2.3244864989614586,
+      "grad_norm": 3.548828125,
+      "learning_rate": 8.326884589726485e-05,
+      "loss": 1.3555,
+      "step": 2518
+    },
+    {
+      "epoch": 2.325409646895915,
+      "grad_norm": 4.21484375,
+      "learning_rate": 8.32621747831888e-05,
+      "loss": 1.4385,
+      "step": 2519
+    },
+    {
+      "epoch": 2.3263327948303716,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.325550366911274e-05,
+      "loss": 1.6577,
+      "step": 2520
+    },
+    {
+      "epoch": 2.327255942764828,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.32488325550367e-05,
+      "loss": 1.5259,
+      "step": 2521
+    },
+    {
+      "epoch": 2.3281790906992845,
+      "grad_norm": 4.05078125,
+      "learning_rate": 8.324216144096064e-05,
+      "loss": 1.4663,
+      "step": 2522
+    },
+    {
+      "epoch": 2.329102238633741,
+      "grad_norm": 5.33984375,
+      "learning_rate": 8.323549032688459e-05,
+      "loss": 1.5249,
+      "step": 2523
+    },
+    {
+      "epoch": 2.3300253865681975,
+      "grad_norm": 5.26953125,
+      "learning_rate": 8.322881921280853e-05,
+      "loss": 1.5959,
+      "step": 2524
+    },
+    {
+      "epoch": 2.330948534502654,
+      "grad_norm": 4.1640625,
+      "learning_rate": 8.32221480987325e-05,
+      "loss": 1.2175,
+      "step": 2525
+    },
+    {
+      "epoch": 2.3318716824371104,
+      "grad_norm": 4.33984375,
+      "learning_rate": 8.321547698465644e-05,
+      "loss": 1.1589,
+      "step": 2526
+    },
+    {
+      "epoch": 2.332794830371567,
+      "grad_norm": 5.4921875,
+      "learning_rate": 8.320880587058038e-05,
+      "loss": 1.6216,
+      "step": 2527
+    },
+    {
+      "epoch": 2.333717978306024,
+      "grad_norm": 4.75390625,
+      "learning_rate": 8.320213475650433e-05,
+      "loss": 1.4443,
+      "step": 2528
+    },
+    {
+      "epoch": 2.33464112624048,
+      "grad_norm": 4.48046875,
+      "learning_rate": 8.319546364242829e-05,
+      "loss": 1.7656,
+      "step": 2529
+    },
+    {
+      "epoch": 2.3355642741749367,
+      "grad_norm": 4.9453125,
+      "learning_rate": 8.318879252835225e-05,
+      "loss": 1.53,
+      "step": 2530
+    },
+    {
+      "epoch": 2.336487422109393,
+      "grad_norm": 4.82421875,
+      "learning_rate": 8.318212141427619e-05,
+      "loss": 1.6785,
+      "step": 2531
+    },
+    {
+      "epoch": 2.3374105700438497,
+      "grad_norm": 4.55859375,
+      "learning_rate": 8.317545030020014e-05,
+      "loss": 1.749,
+      "step": 2532
+    },
+    {
+      "epoch": 2.338333717978306,
+      "grad_norm": 3.591796875,
+      "learning_rate": 8.316877918612408e-05,
+      "loss": 1.5972,
+      "step": 2533
+    },
+    {
+      "epoch": 2.3392568659127626,
+      "grad_norm": 4.2890625,
+      "learning_rate": 8.316210807204804e-05,
+      "loss": 1.614,
+      "step": 2534
+    },
+    {
+      "epoch": 2.340180013847219,
+      "grad_norm": 3.716796875,
+      "learning_rate": 8.315543695797199e-05,
+      "loss": 1.6384,
+      "step": 2535
+    },
+    {
+      "epoch": 2.3411031617816755,
+      "grad_norm": 5.1796875,
+      "learning_rate": 8.314876584389593e-05,
+      "loss": 1.3372,
+      "step": 2536
+    },
+    {
+      "epoch": 2.342026309716132,
+      "grad_norm": 5.08203125,
+      "learning_rate": 8.314209472981989e-05,
+      "loss": 1.4797,
+      "step": 2537
+    },
+    {
+      "epoch": 2.3429494576505885,
+      "grad_norm": 3.326171875,
+      "learning_rate": 8.313542361574384e-05,
+      "loss": 1.1672,
+      "step": 2538
+    },
+    {
+      "epoch": 2.343872605585045,
+      "grad_norm": 4.0546875,
+      "learning_rate": 8.312875250166778e-05,
+      "loss": 1.3865,
+      "step": 2539
+    },
+    {
+      "epoch": 2.3447957535195014,
+      "grad_norm": 4.390625,
+      "learning_rate": 8.312208138759173e-05,
+      "loss": 1.5898,
+      "step": 2540
+    },
+    {
+      "epoch": 2.345718901453958,
+      "grad_norm": 3.58203125,
+      "learning_rate": 8.311541027351569e-05,
+      "loss": 1.7739,
+      "step": 2541
+    },
+    {
+      "epoch": 2.3466420493884144,
+      "grad_norm": 4.13671875,
+      "learning_rate": 8.310873915943963e-05,
+      "loss": 1.325,
+      "step": 2542
+    },
+    {
+      "epoch": 2.347565197322871,
+      "grad_norm": 3.134765625,
+      "learning_rate": 8.310206804536358e-05,
+      "loss": 1.3129,
+      "step": 2543
+    },
+    {
+      "epoch": 2.3484883452573273,
+      "grad_norm": 4.39453125,
+      "learning_rate": 8.309539693128752e-05,
+      "loss": 1.1141,
+      "step": 2544
+    },
+    {
+      "epoch": 2.349411493191784,
+      "grad_norm": 4.5703125,
+      "learning_rate": 8.308872581721148e-05,
+      "loss": 1.6833,
+      "step": 2545
+    },
+    {
+      "epoch": 2.3503346411262402,
+      "grad_norm": 4.0703125,
+      "learning_rate": 8.308205470313543e-05,
+      "loss": 1.8108,
+      "step": 2546
+    },
+    {
+      "epoch": 2.351257789060697,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.307538358905937e-05,
+      "loss": 1.5891,
+      "step": 2547
+    },
+    {
+      "epoch": 2.3521809369951536,
+      "grad_norm": 4.12890625,
+      "learning_rate": 8.306871247498332e-05,
+      "loss": 1.1308,
+      "step": 2548
+    },
+    {
+      "epoch": 2.35310408492961,
+      "grad_norm": 4.671875,
+      "learning_rate": 8.306204136090728e-05,
+      "loss": 1.6406,
+      "step": 2549
+    },
+    {
+      "epoch": 2.3540272328640666,
+      "grad_norm": 4.21484375,
+      "learning_rate": 8.305537024683122e-05,
+      "loss": 1.2584,
+      "step": 2550
+    },
+    {
+      "epoch": 2.354950380798523,
+      "grad_norm": 5.3671875,
+      "learning_rate": 8.304869913275517e-05,
+      "loss": 1.4524,
+      "step": 2551
+    },
+    {
+      "epoch": 2.3558735287329795,
+      "grad_norm": 4.79296875,
+      "learning_rate": 8.304202801867911e-05,
+      "loss": 1.4866,
+      "step": 2552
+    },
+    {
+      "epoch": 2.356796676667436,
+      "grad_norm": 4.03515625,
+      "learning_rate": 8.303535690460307e-05,
+      "loss": 1.6318,
+      "step": 2553
+    },
+    {
+      "epoch": 2.3577198246018924,
+      "grad_norm": 5.10546875,
+      "learning_rate": 8.302868579052702e-05,
+      "loss": 1.6743,
+      "step": 2554
+    },
+    {
+      "epoch": 2.358642972536349,
+      "grad_norm": 4.921875,
+      "learning_rate": 8.302201467645098e-05,
+      "loss": 1.5996,
+      "step": 2555
+    },
+    {
+      "epoch": 2.3595661204708054,
+      "grad_norm": 5.25,
+      "learning_rate": 8.301534356237492e-05,
+      "loss": 1.4641,
+      "step": 2556
+    },
+    {
+      "epoch": 2.360489268405262,
+      "grad_norm": 3.1171875,
+      "learning_rate": 8.300867244829887e-05,
+      "loss": 1.5461,
+      "step": 2557
+    },
+    {
+      "epoch": 2.3614124163397183,
+      "grad_norm": 5.85546875,
+      "learning_rate": 8.300200133422283e-05,
+      "loss": 1.3306,
+      "step": 2558
+    },
+    {
+      "epoch": 2.362335564274175,
+      "grad_norm": 5.1015625,
+      "learning_rate": 8.299533022014677e-05,
+      "loss": 1.4495,
+      "step": 2559
+    },
+    {
+      "epoch": 2.3632587122086313,
+      "grad_norm": 3.38671875,
+      "learning_rate": 8.298865910607072e-05,
+      "loss": 1.553,
+      "step": 2560
+    },
+    {
+      "epoch": 2.3641818601430877,
+      "grad_norm": 3.45703125,
+      "learning_rate": 8.298198799199466e-05,
+      "loss": 1.5244,
+      "step": 2561
+    },
+    {
+      "epoch": 2.3651050080775446,
+      "grad_norm": 4.56640625,
+      "learning_rate": 8.297531687791862e-05,
+      "loss": 1.4961,
+      "step": 2562
+    },
+    {
+      "epoch": 2.366028156012001,
+      "grad_norm": 5.984375,
+      "learning_rate": 8.296864576384257e-05,
+      "loss": 1.5383,
+      "step": 2563
+    },
+    {
+      "epoch": 2.3669513039464576,
+      "grad_norm": 4.13671875,
+      "learning_rate": 8.296197464976651e-05,
+      "loss": 1.2318,
+      "step": 2564
+    },
+    {
+      "epoch": 2.367874451880914,
+      "grad_norm": 5.3515625,
+      "learning_rate": 8.295530353569046e-05,
+      "loss": 1.7156,
+      "step": 2565
+    },
+    {
+      "epoch": 2.3687975998153705,
+      "grad_norm": 6.07421875,
+      "learning_rate": 8.294863242161442e-05,
+      "loss": 1.4268,
+      "step": 2566
+    },
+    {
+      "epoch": 2.369720747749827,
+      "grad_norm": 4.1171875,
+      "learning_rate": 8.294196130753836e-05,
+      "loss": 1.4395,
+      "step": 2567
+    },
+    {
+      "epoch": 2.3706438956842835,
+      "grad_norm": 4.4375,
+      "learning_rate": 8.293529019346231e-05,
+      "loss": 1.4507,
+      "step": 2568
+    },
+    {
+      "epoch": 2.37156704361874,
+      "grad_norm": 4.921875,
+      "learning_rate": 8.292861907938625e-05,
+      "loss": 1.5845,
+      "step": 2569
+    },
+    {
+      "epoch": 2.3724901915531964,
+      "grad_norm": 5.51953125,
+      "learning_rate": 8.292194796531021e-05,
+      "loss": 1.2727,
+      "step": 2570
+    },
+    {
+      "epoch": 2.373413339487653,
+      "grad_norm": 5.296875,
+      "learning_rate": 8.291527685123416e-05,
+      "loss": 1.5627,
+      "step": 2571
+    },
+    {
+      "epoch": 2.3743364874221093,
+      "grad_norm": 5.12890625,
+      "learning_rate": 8.29086057371581e-05,
+      "loss": 1.4028,
+      "step": 2572
+    },
+    {
+      "epoch": 2.375259635356566,
+      "grad_norm": 5.546875,
+      "learning_rate": 8.290193462308205e-05,
+      "loss": 1.6587,
+      "step": 2573
+    },
+    {
+      "epoch": 2.3761827832910223,
+      "grad_norm": 4.8359375,
+      "learning_rate": 8.289526350900601e-05,
+      "loss": 1.3877,
+      "step": 2574
+    },
+    {
+      "epoch": 2.3771059312254788,
+      "grad_norm": 4.62109375,
+      "learning_rate": 8.288859239492995e-05,
+      "loss": 1.1573,
+      "step": 2575
+    },
+    {
+      "epoch": 2.378029079159935,
+      "grad_norm": 4.8203125,
+      "learning_rate": 8.28819212808539e-05,
+      "loss": 1.6995,
+      "step": 2576
+    },
+    {
+      "epoch": 2.3789522270943917,
+      "grad_norm": 6.0390625,
+      "learning_rate": 8.287525016677785e-05,
+      "loss": 1.3068,
+      "step": 2577
+    },
+    {
+      "epoch": 2.379875375028848,
+      "grad_norm": 3.580078125,
+      "learning_rate": 8.28685790527018e-05,
+      "loss": 1.1746,
+      "step": 2578
+    },
+    {
+      "epoch": 2.380798522963305,
+      "grad_norm": 5.08984375,
+      "learning_rate": 8.286190793862575e-05,
+      "loss": 1.5117,
+      "step": 2579
+    },
+    {
+      "epoch": 2.3817216708977615,
+      "grad_norm": 4.1015625,
+      "learning_rate": 8.285523682454971e-05,
+      "loss": 1.209,
+      "step": 2580
+    },
+    {
+      "epoch": 2.382644818832218,
+      "grad_norm": 3.67578125,
+      "learning_rate": 8.284856571047365e-05,
+      "loss": 1.498,
+      "step": 2581
+    },
+    {
+      "epoch": 2.3835679667666745,
+      "grad_norm": 4.01953125,
+      "learning_rate": 8.284189459639761e-05,
+      "loss": 1.7107,
+      "step": 2582
+    },
+    {
+      "epoch": 2.384491114701131,
+      "grad_norm": 4.578125,
+      "learning_rate": 8.283522348232156e-05,
+      "loss": 1.5769,
+      "step": 2583
+    },
+    {
+      "epoch": 2.3854142626355874,
+      "grad_norm": 5.11328125,
+      "learning_rate": 8.28285523682455e-05,
+      "loss": 1.4373,
+      "step": 2584
+    },
+    {
+      "epoch": 2.386337410570044,
+      "grad_norm": 4.6796875,
+      "learning_rate": 8.282188125416945e-05,
+      "loss": 1.4575,
+      "step": 2585
+    },
+    {
+      "epoch": 2.3872605585045004,
+      "grad_norm": 5.36328125,
+      "learning_rate": 8.281521014009341e-05,
+      "loss": 1.3774,
+      "step": 2586
+    },
+    {
+      "epoch": 2.388183706438957,
+      "grad_norm": 5.2109375,
+      "learning_rate": 8.280853902601735e-05,
+      "loss": 1.3679,
+      "step": 2587
+    },
+    {
+      "epoch": 2.3891068543734133,
+      "grad_norm": 4.52734375,
+      "learning_rate": 8.28018679119413e-05,
+      "loss": 1.7366,
+      "step": 2588
+    },
+    {
+      "epoch": 2.3900300023078698,
+      "grad_norm": 4.31640625,
+      "learning_rate": 8.279519679786524e-05,
+      "loss": 1.5759,
+      "step": 2589
+    },
+    {
+      "epoch": 2.3909531502423262,
+      "grad_norm": 4.9296875,
+      "learning_rate": 8.27885256837892e-05,
+      "loss": 1.6062,
+      "step": 2590
+    },
+    {
+      "epoch": 2.3918762981767827,
+      "grad_norm": 3.978515625,
+      "learning_rate": 8.278185456971315e-05,
+      "loss": 1.4033,
+      "step": 2591
+    },
+    {
+      "epoch": 2.392799446111239,
+      "grad_norm": 5.1171875,
+      "learning_rate": 8.27751834556371e-05,
+      "loss": 1.4106,
+      "step": 2592
+    },
+    {
+      "epoch": 2.3937225940456957,
+      "grad_norm": 4.390625,
+      "learning_rate": 8.276851234156104e-05,
+      "loss": 1.6484,
+      "step": 2593
+    },
+    {
+      "epoch": 2.3946457419801526,
+      "grad_norm": 4.20703125,
+      "learning_rate": 8.2761841227485e-05,
+      "loss": 1.2908,
+      "step": 2594
+    },
+    {
+      "epoch": 2.3955688899146086,
+      "grad_norm": 5.35546875,
+      "learning_rate": 8.275517011340894e-05,
+      "loss": 1.3842,
+      "step": 2595
+    },
+    {
+      "epoch": 2.3964920378490655,
+      "grad_norm": 4.91015625,
+      "learning_rate": 8.274849899933289e-05,
+      "loss": 1.6958,
+      "step": 2596
+    },
+    {
+      "epoch": 2.397415185783522,
+      "grad_norm": 4.703125,
+      "learning_rate": 8.274182788525684e-05,
+      "loss": 1.4146,
+      "step": 2597
+    },
+    {
+      "epoch": 2.3983383337179784,
+      "grad_norm": 5.2890625,
+      "learning_rate": 8.27351567711808e-05,
+      "loss": 1.3783,
+      "step": 2598
+    },
+    {
+      "epoch": 2.399261481652435,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.272848565710474e-05,
+      "loss": 1.6328,
+      "step": 2599
+    },
+    {
+      "epoch": 2.4001846295868914,
+      "grad_norm": 4.25390625,
+      "learning_rate": 8.272181454302869e-05,
+      "loss": 1.6062,
+      "step": 2600
+    },
+    {
+      "epoch": 2.401107777521348,
+      "grad_norm": 4.81640625,
+      "learning_rate": 8.271514342895263e-05,
+      "loss": 1.314,
+      "step": 2601
+    },
+    {
+      "epoch": 2.4020309254558043,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.270847231487659e-05,
+      "loss": 1.6316,
+      "step": 2602
+    },
+    {
+      "epoch": 2.402954073390261,
+      "grad_norm": 4.328125,
+      "learning_rate": 8.270180120080054e-05,
+      "loss": 1.4729,
+      "step": 2603
+    },
+    {
+      "epoch": 2.4038772213247173,
+      "grad_norm": 4.296875,
+      "learning_rate": 8.269513008672448e-05,
+      "loss": 1.553,
+      "step": 2604
+    },
+    {
+      "epoch": 2.4048003692591737,
+      "grad_norm": 4.98828125,
+      "learning_rate": 8.268845897264844e-05,
+      "loss": 1.5952,
+      "step": 2605
+    },
+    {
+      "epoch": 2.40572351719363,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.268178785857238e-05,
+      "loss": 1.4155,
+      "step": 2606
+    },
+    {
+      "epoch": 2.4066466651280867,
+      "grad_norm": 4.8671875,
+      "learning_rate": 8.267511674449634e-05,
+      "loss": 1.5798,
+      "step": 2607
+    },
+    {
+      "epoch": 2.407569813062543,
+      "grad_norm": 4.65234375,
+      "learning_rate": 8.266844563042029e-05,
+      "loss": 1.1147,
+      "step": 2608
+    },
+    {
+      "epoch": 2.4084929609969996,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.266177451634423e-05,
+      "loss": 0.9905,
+      "step": 2609
+    },
+    {
+      "epoch": 2.409416108931456,
+      "grad_norm": 3.392578125,
+      "learning_rate": 8.265510340226818e-05,
+      "loss": 0.9475,
+      "step": 2610
+    },
+    {
+      "epoch": 2.410339256865913,
+      "grad_norm": 3.8828125,
+      "learning_rate": 8.264843228819214e-05,
+      "loss": 1.3038,
+      "step": 2611
+    },
+    {
+      "epoch": 2.411262404800369,
+      "grad_norm": 4.12109375,
+      "learning_rate": 8.264176117411608e-05,
+      "loss": 1.5566,
+      "step": 2612
+    },
+    {
+      "epoch": 2.412185552734826,
+      "grad_norm": 3.943359375,
+      "learning_rate": 8.263509006004003e-05,
+      "loss": 1.4854,
+      "step": 2613
+    },
+    {
+      "epoch": 2.4131087006692824,
+      "grad_norm": 3.7734375,
+      "learning_rate": 8.262841894596398e-05,
+      "loss": 1.1405,
+      "step": 2614
+    },
+    {
+      "epoch": 2.414031848603739,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.262174783188793e-05,
+      "loss": 1.6189,
+      "step": 2615
+    },
+    {
+      "epoch": 2.4149549965381953,
+      "grad_norm": 3.849609375,
+      "learning_rate": 8.261507671781188e-05,
+      "loss": 1.0947,
+      "step": 2616
+    },
+    {
+      "epoch": 2.415878144472652,
+      "grad_norm": 4.421875,
+      "learning_rate": 8.260840560373583e-05,
+      "loss": 1.4683,
+      "step": 2617
+    },
+    {
+      "epoch": 2.4168012924071083,
+      "grad_norm": 3.791015625,
+      "learning_rate": 8.260173448965977e-05,
+      "loss": 1.3938,
+      "step": 2618
+    },
+    {
+      "epoch": 2.4177244403415648,
+      "grad_norm": 4.20703125,
+      "learning_rate": 8.259506337558373e-05,
+      "loss": 1.3314,
+      "step": 2619
+    },
+    {
+      "epoch": 2.418647588276021,
+      "grad_norm": 4.42578125,
+      "learning_rate": 8.258839226150768e-05,
+      "loss": 1.5654,
+      "step": 2620
+    },
+    {
+      "epoch": 2.4195707362104777,
+      "grad_norm": 4.2265625,
+      "learning_rate": 8.258172114743162e-05,
+      "loss": 1.6538,
+      "step": 2621
+    },
+    {
+      "epoch": 2.420493884144934,
+      "grad_norm": 4.6796875,
+      "learning_rate": 8.257505003335557e-05,
+      "loss": 1.3022,
+      "step": 2622
+    },
+    {
+      "epoch": 2.4214170320793906,
+      "grad_norm": 4.7421875,
+      "learning_rate": 8.256837891927953e-05,
+      "loss": 1.3138,
+      "step": 2623
+    },
+    {
+      "epoch": 2.422340180013847,
+      "grad_norm": 5.58203125,
+      "learning_rate": 8.256170780520347e-05,
+      "loss": 1.5969,
+      "step": 2624
+    },
+    {
+      "epoch": 2.4232633279483036,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.255503669112742e-05,
+      "loss": 1.3865,
+      "step": 2625
+    },
+    {
+      "epoch": 2.42418647588276,
+      "grad_norm": 5.23046875,
+      "learning_rate": 8.254836557705136e-05,
+      "loss": 1.1835,
+      "step": 2626
+    },
+    {
+      "epoch": 2.4251096238172165,
+      "grad_norm": 4.98046875,
+      "learning_rate": 8.254169446297532e-05,
+      "loss": 1.4111,
+      "step": 2627
+    },
+    {
+      "epoch": 2.4260327717516734,
+      "grad_norm": 4.79296875,
+      "learning_rate": 8.253502334889927e-05,
+      "loss": 1.5757,
+      "step": 2628
+    },
+    {
+      "epoch": 2.42695591968613,
+      "grad_norm": 4.171875,
+      "learning_rate": 8.252835223482321e-05,
+      "loss": 1.5874,
+      "step": 2629
+    },
+    {
+      "epoch": 2.4278790676205864,
+      "grad_norm": 5.40234375,
+      "learning_rate": 8.252168112074717e-05,
+      "loss": 1.3194,
+      "step": 2630
+    },
+    {
+      "epoch": 2.428802215555043,
+      "grad_norm": 4.29296875,
+      "learning_rate": 8.251501000667113e-05,
+      "loss": 1.2449,
+      "step": 2631
+    },
+    {
+      "epoch": 2.4297253634894993,
+      "grad_norm": 4.87109375,
+      "learning_rate": 8.250833889259507e-05,
+      "loss": 1.4387,
+      "step": 2632
+    },
+    {
+      "epoch": 2.4306485114239558,
+      "grad_norm": 4.82421875,
+      "learning_rate": 8.250166777851902e-05,
+      "loss": 1.3999,
+      "step": 2633
+    },
+    {
+      "epoch": 2.4315716593584122,
+      "grad_norm": 5.125,
+      "learning_rate": 8.249499666444297e-05,
+      "loss": 1.3928,
+      "step": 2634
+    },
+    {
+      "epoch": 2.4324948072928687,
+      "grad_norm": 4.7578125,
+      "learning_rate": 8.248832555036692e-05,
+      "loss": 1.2054,
+      "step": 2635
+    },
+    {
+      "epoch": 2.433417955227325,
+      "grad_norm": 6.35546875,
+      "learning_rate": 8.248165443629087e-05,
+      "loss": 1.4094,
+      "step": 2636
+    },
+    {
+      "epoch": 2.4343411031617816,
+      "grad_norm": 4.2578125,
+      "learning_rate": 8.247498332221482e-05,
+      "loss": 1.3899,
+      "step": 2637
+    },
+    {
+      "epoch": 2.435264251096238,
+      "grad_norm": 4.48828125,
+      "learning_rate": 8.246831220813876e-05,
+      "loss": 1.582,
+      "step": 2638
+    },
+    {
+      "epoch": 2.4361873990306946,
+      "grad_norm": 4.8515625,
+      "learning_rate": 8.246164109406272e-05,
+      "loss": 1.4131,
+      "step": 2639
+    },
+    {
+      "epoch": 2.437110546965151,
+      "grad_norm": 5.18359375,
+      "learning_rate": 8.245496997998667e-05,
+      "loss": 1.7012,
+      "step": 2640
+    },
+    {
+      "epoch": 2.4380336948996075,
+      "grad_norm": 5.47265625,
+      "learning_rate": 8.244829886591061e-05,
+      "loss": 1.5225,
+      "step": 2641
+    },
+    {
+      "epoch": 2.438956842834064,
+      "grad_norm": 5.234375,
+      "learning_rate": 8.244162775183456e-05,
+      "loss": 1.449,
+      "step": 2642
+    },
+    {
+      "epoch": 2.4398799907685205,
+      "grad_norm": 6.01171875,
+      "learning_rate": 8.243495663775851e-05,
+      "loss": 1.468,
+      "step": 2643
+    },
+    {
+      "epoch": 2.440803138702977,
+      "grad_norm": 4.37890625,
+      "learning_rate": 8.242828552368246e-05,
+      "loss": 1.3436,
+      "step": 2644
+    },
+    {
+      "epoch": 2.441726286637434,
+      "grad_norm": 4.1796875,
+      "learning_rate": 8.24216144096064e-05,
+      "loss": 1.3053,
+      "step": 2645
+    },
+    {
+      "epoch": 2.4426494345718903,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.241494329553035e-05,
+      "loss": 1.5688,
+      "step": 2646
+    },
+    {
+      "epoch": 2.443572582506347,
+      "grad_norm": 4.55859375,
+      "learning_rate": 8.240827218145431e-05,
+      "loss": 1.4315,
+      "step": 2647
+    },
+    {
+      "epoch": 2.4444957304408033,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.240160106737826e-05,
+      "loss": 1.5847,
+      "step": 2648
+    },
+    {
+      "epoch": 2.4454188783752597,
+      "grad_norm": 4.609375,
+      "learning_rate": 8.23949299533022e-05,
+      "loss": 1.4973,
+      "step": 2649
+    },
+    {
+      "epoch": 2.446342026309716,
+      "grad_norm": 4.921875,
+      "learning_rate": 8.238825883922615e-05,
+      "loss": 1.6292,
+      "step": 2650
+    },
+    {
+      "epoch": 2.4472651742441727,
+      "grad_norm": 4.921875,
+      "learning_rate": 8.23815877251501e-05,
+      "loss": 1.2784,
+      "step": 2651
+    },
+    {
+      "epoch": 2.448188322178629,
+      "grad_norm": 4.32421875,
+      "learning_rate": 8.237491661107405e-05,
+      "loss": 1.4722,
+      "step": 2652
+    },
+    {
+      "epoch": 2.4491114701130856,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.2368245496998e-05,
+      "loss": 1.3978,
+      "step": 2653
+    },
+    {
+      "epoch": 2.450034618047542,
+      "grad_norm": 6.046875,
+      "learning_rate": 8.236157438292194e-05,
+      "loss": 1.4624,
+      "step": 2654
+    },
+    {
+      "epoch": 2.4509577659819985,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.23549032688459e-05,
+      "loss": 1.5151,
+      "step": 2655
+    },
+    {
+      "epoch": 2.451880913916455,
+      "grad_norm": 5.02734375,
+      "learning_rate": 8.234823215476986e-05,
+      "loss": 1.028,
+      "step": 2656
+    },
+    {
+      "epoch": 2.4528040618509115,
+      "grad_norm": 5.49609375,
+      "learning_rate": 8.23415610406938e-05,
+      "loss": 1.5171,
+      "step": 2657
+    },
+    {
+      "epoch": 2.453727209785368,
+      "grad_norm": 4.33203125,
+      "learning_rate": 8.233488992661775e-05,
+      "loss": 1.6499,
+      "step": 2658
+    },
+    {
+      "epoch": 2.4546503577198244,
+      "grad_norm": 4.640625,
+      "learning_rate": 8.23282188125417e-05,
+      "loss": 1.3381,
+      "step": 2659
+    },
+    {
+      "epoch": 2.4555735056542813,
+      "grad_norm": 4.4921875,
+      "learning_rate": 8.232154769846566e-05,
+      "loss": 1.2563,
+      "step": 2660
+    },
+    {
+      "epoch": 2.4564966535887374,
+      "grad_norm": 4.3828125,
+      "learning_rate": 8.23148765843896e-05,
+      "loss": 1.1259,
+      "step": 2661
+    },
+    {
+      "epoch": 2.4574198015231943,
+      "grad_norm": 5.4765625,
+      "learning_rate": 8.230820547031355e-05,
+      "loss": 1.5242,
+      "step": 2662
+    },
+    {
+      "epoch": 2.4583429494576508,
+      "grad_norm": 5.3984375,
+      "learning_rate": 8.230153435623749e-05,
+      "loss": 1.6848,
+      "step": 2663
+    },
+    {
+      "epoch": 2.459266097392107,
+      "grad_norm": 3.87890625,
+      "learning_rate": 8.229486324216145e-05,
+      "loss": 1.4933,
+      "step": 2664
+    },
+    {
+      "epoch": 2.4601892453265637,
+      "grad_norm": 4.10546875,
+      "learning_rate": 8.22881921280854e-05,
+      "loss": 1.2277,
+      "step": 2665
+    },
+    {
+      "epoch": 2.46111239326102,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.228152101400934e-05,
+      "loss": 1.7683,
+      "step": 2666
+    },
+    {
+      "epoch": 2.4620355411954766,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.227484989993329e-05,
+      "loss": 1.738,
+      "step": 2667
+    },
+    {
+      "epoch": 2.462958689129933,
+      "grad_norm": 4.734375,
+      "learning_rate": 8.226817878585725e-05,
+      "loss": 1.4978,
+      "step": 2668
+    },
+    {
+      "epoch": 2.4638818370643896,
+      "grad_norm": 4.515625,
+      "learning_rate": 8.226150767178119e-05,
+      "loss": 1.5389,
+      "step": 2669
+    },
+    {
+      "epoch": 2.464804984998846,
+      "grad_norm": 4.08203125,
+      "learning_rate": 8.225483655770514e-05,
+      "loss": 1.4365,
+      "step": 2670
+    },
+    {
+      "epoch": 2.4657281329333025,
+      "grad_norm": 5.29296875,
+      "learning_rate": 8.224816544362908e-05,
+      "loss": 1.6262,
+      "step": 2671
+    },
+    {
+      "epoch": 2.466651280867759,
+      "grad_norm": 5.39453125,
+      "learning_rate": 8.224149432955304e-05,
+      "loss": 1.1523,
+      "step": 2672
+    },
+    {
+      "epoch": 2.4675744288022154,
+      "grad_norm": 6.21484375,
+      "learning_rate": 8.223482321547699e-05,
+      "loss": 1.6248,
+      "step": 2673
+    },
+    {
+      "epoch": 2.468497576736672,
+      "grad_norm": 3.8359375,
+      "learning_rate": 8.222815210140093e-05,
+      "loss": 1.5149,
+      "step": 2674
+    },
+    {
+      "epoch": 2.4694207246711284,
+      "grad_norm": 4.1640625,
+      "learning_rate": 8.222148098732488e-05,
+      "loss": 1.585,
+      "step": 2675
+    },
+    {
+      "epoch": 2.470343872605585,
+      "grad_norm": 3.615234375,
+      "learning_rate": 8.221480987324884e-05,
+      "loss": 1.469,
+      "step": 2676
+    },
+    {
+      "epoch": 2.4712670205400418,
+      "grad_norm": 3.95703125,
+      "learning_rate": 8.220813875917278e-05,
+      "loss": 1.0502,
+      "step": 2677
+    },
+    {
+      "epoch": 2.472190168474498,
+      "grad_norm": 3.1484375,
+      "learning_rate": 8.220146764509673e-05,
+      "loss": 0.8159,
+      "step": 2678
+    },
+    {
+      "epoch": 2.4731133164089547,
+      "grad_norm": 4.07421875,
+      "learning_rate": 8.219479653102067e-05,
+      "loss": 1.5977,
+      "step": 2679
+    },
+    {
+      "epoch": 2.474036464343411,
+      "grad_norm": 4.04296875,
+      "learning_rate": 8.218812541694464e-05,
+      "loss": 1.6912,
+      "step": 2680
+    },
+    {
+      "epoch": 2.4749596122778676,
+      "grad_norm": 4.06640625,
+      "learning_rate": 8.218145430286859e-05,
+      "loss": 1.582,
+      "step": 2681
+    },
+    {
+      "epoch": 2.475882760212324,
+      "grad_norm": 4.10546875,
+      "learning_rate": 8.217478318879254e-05,
+      "loss": 1.2616,
+      "step": 2682
+    },
+    {
+      "epoch": 2.4768059081467806,
+      "grad_norm": 4.515625,
+      "learning_rate": 8.216811207471648e-05,
+      "loss": 1.7366,
+      "step": 2683
+    },
+    {
+      "epoch": 2.477729056081237,
+      "grad_norm": 4.0625,
+      "learning_rate": 8.216144096064044e-05,
+      "loss": 1.5774,
+      "step": 2684
+    },
+    {
+      "epoch": 2.4786522040156935,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.215476984656439e-05,
+      "loss": 1.3257,
+      "step": 2685
+    },
+    {
+      "epoch": 2.47957535195015,
+      "grad_norm": 3.625,
+      "learning_rate": 8.214809873248833e-05,
+      "loss": 1.5879,
+      "step": 2686
+    },
+    {
+      "epoch": 2.4804984998846065,
+      "grad_norm": 4.125,
+      "learning_rate": 8.214142761841228e-05,
+      "loss": 1.8196,
+      "step": 2687
+    },
+    {
+      "epoch": 2.481421647819063,
+      "grad_norm": 4.60546875,
+      "learning_rate": 8.213475650433624e-05,
+      "loss": 1.4612,
+      "step": 2688
+    },
+    {
+      "epoch": 2.4823447957535194,
+      "grad_norm": 4.45703125,
+      "learning_rate": 8.212808539026018e-05,
+      "loss": 1.4998,
+      "step": 2689
+    },
+    {
+      "epoch": 2.483267943687976,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.212141427618413e-05,
+      "loss": 1.4036,
+      "step": 2690
+    },
+    {
+      "epoch": 2.4841910916224323,
+      "grad_norm": 5.2734375,
+      "learning_rate": 8.211474316210807e-05,
+      "loss": 1.6787,
+      "step": 2691
+    },
+    {
+      "epoch": 2.485114239556889,
+      "grad_norm": 4.5546875,
+      "learning_rate": 8.210807204803203e-05,
+      "loss": 1.5308,
+      "step": 2692
+    },
+    {
+      "epoch": 2.4860373874913453,
+      "grad_norm": 4.58984375,
+      "learning_rate": 8.210140093395598e-05,
+      "loss": 1.2699,
+      "step": 2693
+    },
+    {
+      "epoch": 2.486960535425802,
+      "grad_norm": 5.25390625,
+      "learning_rate": 8.209472981987992e-05,
+      "loss": 1.5762,
+      "step": 2694
+    },
+    {
+      "epoch": 2.4878836833602587,
+      "grad_norm": 4.90234375,
+      "learning_rate": 8.208805870580387e-05,
+      "loss": 1.7712,
+      "step": 2695
+    },
+    {
+      "epoch": 2.488806831294715,
+      "grad_norm": 5.42578125,
+      "learning_rate": 8.208138759172783e-05,
+      "loss": 1.5413,
+      "step": 2696
+    },
+    {
+      "epoch": 2.4897299792291716,
+      "grad_norm": 5.70703125,
+      "learning_rate": 8.207471647765177e-05,
+      "loss": 1.4031,
+      "step": 2697
+    },
+    {
+      "epoch": 2.490653127163628,
+      "grad_norm": 5.2265625,
+      "learning_rate": 8.206804536357572e-05,
+      "loss": 1.6064,
+      "step": 2698
+    },
+    {
+      "epoch": 2.4915762750980845,
+      "grad_norm": 5.015625,
+      "learning_rate": 8.206137424949966e-05,
+      "loss": 1.4487,
+      "step": 2699
+    },
+    {
+      "epoch": 2.492499423032541,
+      "grad_norm": 4.9296875,
+      "learning_rate": 8.205470313542362e-05,
+      "loss": 1.3704,
+      "step": 2700
+    },
+    {
+      "epoch": 2.4934225709669975,
+      "grad_norm": 5.04296875,
+      "learning_rate": 8.204803202134757e-05,
+      "loss": 1.2991,
+      "step": 2701
+    },
+    {
+      "epoch": 2.494345718901454,
+      "grad_norm": 3.66015625,
+      "learning_rate": 8.204136090727151e-05,
+      "loss": 1.6589,
+      "step": 2702
+    },
+    {
+      "epoch": 2.4952688668359104,
+      "grad_norm": 3.724609375,
+      "learning_rate": 8.203468979319546e-05,
+      "loss": 1.5569,
+      "step": 2703
+    },
+    {
+      "epoch": 2.496192014770367,
+      "grad_norm": 4.203125,
+      "learning_rate": 8.202801867911942e-05,
+      "loss": 1.4961,
+      "step": 2704
+    },
+    {
+      "epoch": 2.4971151627048234,
+      "grad_norm": 5.390625,
+      "learning_rate": 8.202134756504338e-05,
+      "loss": 1.675,
+      "step": 2705
+    },
+    {
+      "epoch": 2.49803831063928,
+      "grad_norm": 4.1015625,
+      "learning_rate": 8.201467645096732e-05,
+      "loss": 1.1785,
+      "step": 2706
+    },
+    {
+      "epoch": 2.4989614585737363,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.200800533689127e-05,
+      "loss": 1.3632,
+      "step": 2707
+    },
+    {
+      "epoch": 2.4998846065081928,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.200133422281521e-05,
+      "loss": 1.1406,
+      "step": 2708
+    },
+    {
+      "epoch": 2.5008077544426497,
+      "grad_norm": 4.66796875,
+      "learning_rate": 8.199466310873917e-05,
+      "loss": 1.8347,
+      "step": 2709
+    },
+    {
+      "epoch": 2.5017309023771057,
+      "grad_norm": 4.51953125,
+      "learning_rate": 8.198799199466312e-05,
+      "loss": 1.7366,
+      "step": 2710
+    },
+    {
+      "epoch": 2.5026540503115626,
+      "grad_norm": 3.712890625,
+      "learning_rate": 8.198132088058706e-05,
+      "loss": 1.2421,
+      "step": 2711
+    },
+    {
+      "epoch": 2.5035771982460187,
+      "grad_norm": 4.98046875,
+      "learning_rate": 8.197464976651101e-05,
+      "loss": 1.3331,
+      "step": 2712
+    },
+    {
+      "epoch": 2.5045003461804756,
+      "grad_norm": 4.40234375,
+      "learning_rate": 8.196797865243497e-05,
+      "loss": 1.1549,
+      "step": 2713
+    },
+    {
+      "epoch": 2.505423494114932,
+      "grad_norm": 4.36328125,
+      "learning_rate": 8.196130753835891e-05,
+      "loss": 1.6301,
+      "step": 2714
+    },
+    {
+      "epoch": 2.5063466420493885,
+      "grad_norm": 5.22265625,
+      "learning_rate": 8.195463642428286e-05,
+      "loss": 1.782,
+      "step": 2715
+    },
+    {
+      "epoch": 2.507269789983845,
+      "grad_norm": 5.34375,
+      "learning_rate": 8.19479653102068e-05,
+      "loss": 1.6665,
+      "step": 2716
+    },
+    {
+      "epoch": 2.5081929379183014,
+      "grad_norm": 4.734375,
+      "learning_rate": 8.194129419613076e-05,
+      "loss": 1.4429,
+      "step": 2717
+    },
+    {
+      "epoch": 2.509116085852758,
+      "grad_norm": 44.65625,
+      "learning_rate": 8.193462308205471e-05,
+      "loss": 2.1982,
+      "step": 2718
+    },
+    {
+      "epoch": 2.5100392337872144,
+      "grad_norm": 4.38671875,
+      "learning_rate": 8.192795196797865e-05,
+      "loss": 1.5288,
+      "step": 2719
+    },
+    {
+      "epoch": 2.510962381721671,
+      "grad_norm": 4.203125,
+      "learning_rate": 8.19212808539026e-05,
+      "loss": 1.6067,
+      "step": 2720
+    },
+    {
+      "epoch": 2.5118855296561273,
+      "grad_norm": 4.30078125,
+      "learning_rate": 8.191460973982656e-05,
+      "loss": 1.3636,
+      "step": 2721
+    },
+    {
+      "epoch": 2.512808677590584,
+      "grad_norm": 4.359375,
+      "learning_rate": 8.19079386257505e-05,
+      "loss": 1.4756,
+      "step": 2722
+    },
+    {
+      "epoch": 2.5137318255250403,
+      "grad_norm": 3.607421875,
+      "learning_rate": 8.190126751167445e-05,
+      "loss": 1.3848,
+      "step": 2723
+    },
+    {
+      "epoch": 2.5146549734594967,
+      "grad_norm": 5.109375,
+      "learning_rate": 8.18945963975984e-05,
+      "loss": 1.5968,
+      "step": 2724
+    },
+    {
+      "epoch": 2.515578121393953,
+      "grad_norm": 4.25390625,
+      "learning_rate": 8.188792528352235e-05,
+      "loss": 1.4465,
+      "step": 2725
+    },
+    {
+      "epoch": 2.51650126932841,
+      "grad_norm": 5.859375,
+      "learning_rate": 8.18812541694463e-05,
+      "loss": 1.4331,
+      "step": 2726
+    },
+    {
+      "epoch": 2.517424417262866,
+      "grad_norm": 4.56640625,
+      "learning_rate": 8.187458305537024e-05,
+      "loss": 1.4956,
+      "step": 2727
+    },
+    {
+      "epoch": 2.518347565197323,
+      "grad_norm": 3.775390625,
+      "learning_rate": 8.18679119412942e-05,
+      "loss": 1.3418,
+      "step": 2728
+    },
+    {
+      "epoch": 2.5192707131317795,
+      "grad_norm": 3.81640625,
+      "learning_rate": 8.186124082721815e-05,
+      "loss": 1.5256,
+      "step": 2729
+    },
+    {
+      "epoch": 2.520193861066236,
+      "grad_norm": 5.18359375,
+      "learning_rate": 8.18545697131421e-05,
+      "loss": 1.4932,
+      "step": 2730
+    },
+    {
+      "epoch": 2.5211170090006925,
+      "grad_norm": 4.265625,
+      "learning_rate": 8.184789859906605e-05,
+      "loss": 1.1827,
+      "step": 2731
+    },
+    {
+      "epoch": 2.522040156935149,
+      "grad_norm": 3.685546875,
+      "learning_rate": 8.184122748499e-05,
+      "loss": 1.5898,
+      "step": 2732
+    },
+    {
+      "epoch": 2.5229633048696054,
+      "grad_norm": 3.91796875,
+      "learning_rate": 8.183455637091396e-05,
+      "loss": 1.3298,
+      "step": 2733
+    },
+    {
+      "epoch": 2.523886452804062,
+      "grad_norm": 4.328125,
+      "learning_rate": 8.18278852568379e-05,
+      "loss": 1.6541,
+      "step": 2734
+    },
+    {
+      "epoch": 2.5248096007385183,
+      "grad_norm": 4.296875,
+      "learning_rate": 8.182121414276185e-05,
+      "loss": 1.564,
+      "step": 2735
+    },
+    {
+      "epoch": 2.525732748672975,
+      "grad_norm": 4.76171875,
+      "learning_rate": 8.181454302868579e-05,
+      "loss": 1.6448,
+      "step": 2736
+    },
+    {
+      "epoch": 2.5266558966074313,
+      "grad_norm": 4.3359375,
+      "learning_rate": 8.180787191460975e-05,
+      "loss": 1.4352,
+      "step": 2737
+    },
+    {
+      "epoch": 2.5275790445418878,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.18012008005337e-05,
+      "loss": 1.3561,
+      "step": 2738
+    },
+    {
+      "epoch": 2.5285021924763442,
+      "grad_norm": 5.296875,
+      "learning_rate": 8.179452968645764e-05,
+      "loss": 1.4741,
+      "step": 2739
+    },
+    {
+      "epoch": 2.5294253404108007,
+      "grad_norm": 4.76953125,
+      "learning_rate": 8.178785857238159e-05,
+      "loss": 1.3278,
+      "step": 2740
+    },
+    {
+      "epoch": 2.530348488345257,
+      "grad_norm": 3.96484375,
+      "learning_rate": 8.178118745830555e-05,
+      "loss": 1.5295,
+      "step": 2741
+    },
+    {
+      "epoch": 2.5312716362797136,
+      "grad_norm": 4.04296875,
+      "learning_rate": 8.177451634422949e-05,
+      "loss": 1.4993,
+      "step": 2742
+    },
+    {
+      "epoch": 2.5321947842141705,
+      "grad_norm": 4.5546875,
+      "learning_rate": 8.176784523015344e-05,
+      "loss": 1.4163,
+      "step": 2743
+    },
+    {
+      "epoch": 2.5331179321486266,
+      "grad_norm": 3.74609375,
+      "learning_rate": 8.176117411607738e-05,
+      "loss": 1.5127,
+      "step": 2744
+    },
+    {
+      "epoch": 2.5340410800830835,
+      "grad_norm": 4.24609375,
+      "learning_rate": 8.175450300200134e-05,
+      "loss": 1.0872,
+      "step": 2745
+    },
+    {
+      "epoch": 2.53496422801754,
+      "grad_norm": 4.6484375,
+      "learning_rate": 8.174783188792529e-05,
+      "loss": 1.6545,
+      "step": 2746
+    },
+    {
+      "epoch": 2.5358873759519964,
+      "grad_norm": 3.91015625,
+      "learning_rate": 8.174116077384923e-05,
+      "loss": 1.5149,
+      "step": 2747
+    },
+    {
+      "epoch": 2.536810523886453,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.173448965977318e-05,
+      "loss": 1.6216,
+      "step": 2748
+    },
+    {
+      "epoch": 2.5377336718209094,
+      "grad_norm": 5.35546875,
+      "learning_rate": 8.172781854569714e-05,
+      "loss": 1.3455,
+      "step": 2749
+    },
+    {
+      "epoch": 2.538656819755366,
+      "grad_norm": 3.73828125,
+      "learning_rate": 8.172114743162108e-05,
+      "loss": 1.7285,
+      "step": 2750
+    },
+    {
+      "epoch": 2.5395799676898223,
+      "grad_norm": 3.626953125,
+      "learning_rate": 8.171447631754503e-05,
+      "loss": 1.8555,
+      "step": 2751
+    },
+    {
+      "epoch": 2.5405031156242788,
+      "grad_norm": 3.783203125,
+      "learning_rate": 8.170780520346897e-05,
+      "loss": 1.7021,
+      "step": 2752
+    },
+    {
+      "epoch": 2.5414262635587352,
+      "grad_norm": 4.66796875,
+      "learning_rate": 8.170113408939293e-05,
+      "loss": 1.2872,
+      "step": 2753
+    },
+    {
+      "epoch": 2.5423494114931917,
+      "grad_norm": 5.1953125,
+      "learning_rate": 8.169446297531688e-05,
+      "loss": 1.3962,
+      "step": 2754
+    },
+    {
+      "epoch": 2.543272559427648,
+      "grad_norm": 4.796875,
+      "learning_rate": 8.168779186124084e-05,
+      "loss": 1.6177,
+      "step": 2755
+    },
+    {
+      "epoch": 2.5441957073621047,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.168112074716478e-05,
+      "loss": 1.3331,
+      "step": 2756
+    },
+    {
+      "epoch": 2.545118855296561,
+      "grad_norm": 2.99609375,
+      "learning_rate": 8.167444963308873e-05,
+      "loss": 0.9553,
+      "step": 2757
+    },
+    {
+      "epoch": 2.546042003231018,
+      "grad_norm": 4.69140625,
+      "learning_rate": 8.166777851901269e-05,
+      "loss": 1.488,
+      "step": 2758
+    },
+    {
+      "epoch": 2.546965151165474,
+      "grad_norm": 3.75390625,
+      "learning_rate": 8.166110740493663e-05,
+      "loss": 1.4982,
+      "step": 2759
+    },
+    {
+      "epoch": 2.547888299099931,
+      "grad_norm": 4.27734375,
+      "learning_rate": 8.165443629086058e-05,
+      "loss": 1.417,
+      "step": 2760
+    },
+    {
+      "epoch": 2.548811447034387,
+      "grad_norm": 5.484375,
+      "learning_rate": 8.164776517678452e-05,
+      "loss": 1.4124,
+      "step": 2761
+    },
+    {
+      "epoch": 2.549734594968844,
+      "grad_norm": 5.6328125,
+      "learning_rate": 8.164109406270848e-05,
+      "loss": 1.707,
+      "step": 2762
+    },
+    {
+      "epoch": 2.5506577429033004,
+      "grad_norm": 4.03515625,
+      "learning_rate": 8.163442294863243e-05,
+      "loss": 1.4185,
+      "step": 2763
+    },
+    {
+      "epoch": 2.551580890837757,
+      "grad_norm": 4.05078125,
+      "learning_rate": 8.162775183455637e-05,
+      "loss": 1.4727,
+      "step": 2764
+    },
+    {
+      "epoch": 2.5525040387722133,
+      "grad_norm": 4.91796875,
+      "learning_rate": 8.162108072048033e-05,
+      "loss": 1.5874,
+      "step": 2765
+    },
+    {
+      "epoch": 2.55342718670667,
+      "grad_norm": 4.19140625,
+      "learning_rate": 8.161440960640428e-05,
+      "loss": 1.8057,
+      "step": 2766
+    },
+    {
+      "epoch": 2.5543503346411263,
+      "grad_norm": 5.37109375,
+      "learning_rate": 8.160773849232822e-05,
+      "loss": 0.9521,
+      "step": 2767
+    },
+    {
+      "epoch": 2.5552734825755827,
+      "grad_norm": 4.6328125,
+      "learning_rate": 8.160106737825217e-05,
+      "loss": 1.5808,
+      "step": 2768
+    },
+    {
+      "epoch": 2.556196630510039,
+      "grad_norm": 4.72265625,
+      "learning_rate": 8.159439626417613e-05,
+      "loss": 1.4519,
+      "step": 2769
+    },
+    {
+      "epoch": 2.5571197784444957,
+      "grad_norm": 4.171875,
+      "learning_rate": 8.158772515010007e-05,
+      "loss": 1.2435,
+      "step": 2770
+    },
+    {
+      "epoch": 2.558042926378952,
+      "grad_norm": 4.2890625,
+      "learning_rate": 8.158105403602402e-05,
+      "loss": 1.5552,
+      "step": 2771
+    },
+    {
+      "epoch": 2.5589660743134086,
+      "grad_norm": 3.869140625,
+      "learning_rate": 8.157438292194796e-05,
+      "loss": 1.54,
+      "step": 2772
+    },
+    {
+      "epoch": 2.559889222247865,
+      "grad_norm": 4.64453125,
+      "learning_rate": 8.156771180787192e-05,
+      "loss": 1.5759,
+      "step": 2773
+    },
+    {
+      "epoch": 2.5608123701823216,
+      "grad_norm": 4.21484375,
+      "learning_rate": 8.156104069379587e-05,
+      "loss": 1.6174,
+      "step": 2774
+    },
+    {
+      "epoch": 2.5617355181167785,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.155436957971981e-05,
+      "loss": 1.5386,
+      "step": 2775
+    },
+    {
+      "epoch": 2.5626586660512345,
+      "grad_norm": 3.623046875,
+      "learning_rate": 8.154769846564376e-05,
+      "loss": 1.7852,
+      "step": 2776
+    },
+    {
+      "epoch": 2.5635818139856914,
+      "grad_norm": 4.515625,
+      "learning_rate": 8.154102735156772e-05,
+      "loss": 1.6204,
+      "step": 2777
+    },
+    {
+      "epoch": 2.5645049619201474,
+      "grad_norm": 3.865234375,
+      "learning_rate": 8.153435623749166e-05,
+      "loss": 1.8606,
+      "step": 2778
+    },
+    {
+      "epoch": 2.5654281098546043,
+      "grad_norm": 4.2265625,
+      "learning_rate": 8.152768512341561e-05,
+      "loss": 1.6299,
+      "step": 2779
+    },
+    {
+      "epoch": 2.566351257789061,
+      "grad_norm": 4.74609375,
+      "learning_rate": 8.152101400933957e-05,
+      "loss": 1.6346,
+      "step": 2780
+    },
+    {
+      "epoch": 2.5672744057235173,
+      "grad_norm": 4.4921875,
+      "learning_rate": 8.151434289526351e-05,
+      "loss": 1.4045,
+      "step": 2781
+    },
+    {
+      "epoch": 2.5681975536579738,
+      "grad_norm": 4.94140625,
+      "learning_rate": 8.150767178118747e-05,
+      "loss": 1.4221,
+      "step": 2782
+    },
+    {
+      "epoch": 2.5691207015924302,
+      "grad_norm": 5.74609375,
+      "learning_rate": 8.150100066711142e-05,
+      "loss": 1.4685,
+      "step": 2783
+    },
+    {
+      "epoch": 2.5700438495268867,
+      "grad_norm": 6.24609375,
+      "learning_rate": 8.149432955303536e-05,
+      "loss": 1.182,
+      "step": 2784
+    },
+    {
+      "epoch": 2.570966997461343,
+      "grad_norm": 6.06640625,
+      "learning_rate": 8.148765843895931e-05,
+      "loss": 1.4441,
+      "step": 2785
+    },
+    {
+      "epoch": 2.5718901453957996,
+      "grad_norm": 4.12109375,
+      "learning_rate": 8.148098732488327e-05,
+      "loss": 1.512,
+      "step": 2786
+    },
+    {
+      "epoch": 2.572813293330256,
+      "grad_norm": 3.755859375,
+      "learning_rate": 8.147431621080721e-05,
+      "loss": 0.9511,
+      "step": 2787
+    },
+    {
+      "epoch": 2.5737364412647126,
+      "grad_norm": 4.9453125,
+      "learning_rate": 8.146764509673116e-05,
+      "loss": 1.5598,
+      "step": 2788
+    },
+    {
+      "epoch": 2.574659589199169,
+      "grad_norm": 3.9921875,
+      "learning_rate": 8.14609739826551e-05,
+      "loss": 1.6726,
+      "step": 2789
+    },
+    {
+      "epoch": 2.5755827371336255,
+      "grad_norm": 4.01171875,
+      "learning_rate": 8.145430286857906e-05,
+      "loss": 1.2477,
+      "step": 2790
+    },
+    {
+      "epoch": 2.576505885068082,
+      "grad_norm": 4.234375,
+      "learning_rate": 8.144763175450301e-05,
+      "loss": 1.199,
+      "step": 2791
+    },
+    {
+      "epoch": 2.577429033002539,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.144096064042695e-05,
+      "loss": 1.4939,
+      "step": 2792
+    },
+    {
+      "epoch": 2.578352180936995,
+      "grad_norm": 3.857421875,
+      "learning_rate": 8.14342895263509e-05,
+      "loss": 1.446,
+      "step": 2793
+    },
+    {
+      "epoch": 2.579275328871452,
+      "grad_norm": 5.421875,
+      "learning_rate": 8.142761841227486e-05,
+      "loss": 1.6868,
+      "step": 2794
+    },
+    {
+      "epoch": 2.5801984768059083,
+      "grad_norm": 4.74609375,
+      "learning_rate": 8.14209472981988e-05,
+      "loss": 1.2843,
+      "step": 2795
+    },
+    {
+      "epoch": 2.5811216247403648,
+      "grad_norm": 3.794921875,
+      "learning_rate": 8.141427618412275e-05,
+      "loss": 1.439,
+      "step": 2796
+    },
+    {
+      "epoch": 2.5820447726748212,
+      "grad_norm": 3.306640625,
+      "learning_rate": 8.14076050700467e-05,
+      "loss": 1.3328,
+      "step": 2797
+    },
+    {
+      "epoch": 2.5829679206092777,
+      "grad_norm": 3.8828125,
+      "learning_rate": 8.140093395597065e-05,
+      "loss": 1.3726,
+      "step": 2798
+    },
+    {
+      "epoch": 2.583891068543734,
+      "grad_norm": 4.375,
+      "learning_rate": 8.13942628418946e-05,
+      "loss": 1.095,
+      "step": 2799
+    },
+    {
+      "epoch": 2.5848142164781907,
+      "grad_norm": 4.87890625,
+      "learning_rate": 8.138759172781854e-05,
+      "loss": 1.4863,
+      "step": 2800
+    },
+    {
+      "epoch": 2.585737364412647,
+      "grad_norm": 4.86328125,
+      "learning_rate": 8.138092061374249e-05,
+      "loss": 1.2666,
+      "step": 2801
+    },
+    {
+      "epoch": 2.5866605123471036,
+      "grad_norm": 4.3515625,
+      "learning_rate": 8.137424949966645e-05,
+      "loss": 1.2578,
+      "step": 2802
+    },
+    {
+      "epoch": 2.58758366028156,
+      "grad_norm": 4.23046875,
+      "learning_rate": 8.13675783855904e-05,
+      "loss": 1.4045,
+      "step": 2803
+    },
+    {
+      "epoch": 2.5885068082160165,
+      "grad_norm": 4.2734375,
+      "learning_rate": 8.136090727151434e-05,
+      "loss": 1.3364,
+      "step": 2804
+    },
+    {
+      "epoch": 2.589429956150473,
+      "grad_norm": 4.64453125,
+      "learning_rate": 8.13542361574383e-05,
+      "loss": 1.4773,
+      "step": 2805
+    },
+    {
+      "epoch": 2.5903531040849295,
+      "grad_norm": 3.92578125,
+      "learning_rate": 8.134756504336226e-05,
+      "loss": 1.3313,
+      "step": 2806
+    },
+    {
+      "epoch": 2.5912762520193864,
+      "grad_norm": 5.30078125,
+      "learning_rate": 8.13408939292862e-05,
+      "loss": 1.4749,
+      "step": 2807
+    },
+    {
+      "epoch": 2.5921993999538424,
+      "grad_norm": 5.11328125,
+      "learning_rate": 8.133422281521015e-05,
+      "loss": 1.2012,
+      "step": 2808
+    },
+    {
+      "epoch": 2.5931225478882993,
+      "grad_norm": 4.08984375,
+      "learning_rate": 8.13275517011341e-05,
+      "loss": 1.2406,
+      "step": 2809
+    },
+    {
+      "epoch": 2.5940456958227553,
+      "grad_norm": 5.31640625,
+      "learning_rate": 8.132088058705805e-05,
+      "loss": 1.1251,
+      "step": 2810
+    },
+    {
+      "epoch": 2.5949688437572123,
+      "grad_norm": 4.1328125,
+      "learning_rate": 8.1314209472982e-05,
+      "loss": 1.3622,
+      "step": 2811
+    },
+    {
+      "epoch": 2.5958919916916687,
+      "grad_norm": 3.978515625,
+      "learning_rate": 8.130753835890594e-05,
+      "loss": 1.5378,
+      "step": 2812
+    },
+    {
+      "epoch": 2.596815139626125,
+      "grad_norm": 4.18359375,
+      "learning_rate": 8.130086724482989e-05,
+      "loss": 1.7065,
+      "step": 2813
+    },
+    {
+      "epoch": 2.5977382875605817,
+      "grad_norm": 4.0859375,
+      "learning_rate": 8.129419613075385e-05,
+      "loss": 1.2371,
+      "step": 2814
+    },
+    {
+      "epoch": 2.598661435495038,
+      "grad_norm": 4.0703125,
+      "learning_rate": 8.128752501667779e-05,
+      "loss": 1.1396,
+      "step": 2815
+    },
+    {
+      "epoch": 2.5995845834294946,
+      "grad_norm": 4.0546875,
+      "learning_rate": 8.128085390260174e-05,
+      "loss": 1.3772,
+      "step": 2816
+    },
+    {
+      "epoch": 2.600507731363951,
+      "grad_norm": 4.71484375,
+      "learning_rate": 8.127418278852568e-05,
+      "loss": 1.709,
+      "step": 2817
+    },
+    {
+      "epoch": 2.6014308792984076,
+      "grad_norm": 4.08203125,
+      "learning_rate": 8.126751167444964e-05,
+      "loss": 1.2528,
+      "step": 2818
+    },
+    {
+      "epoch": 2.602354027232864,
+      "grad_norm": 3.986328125,
+      "learning_rate": 8.126084056037359e-05,
+      "loss": 1.3867,
+      "step": 2819
+    },
+    {
+      "epoch": 2.6032771751673205,
+      "grad_norm": 4.58203125,
+      "learning_rate": 8.125416944629753e-05,
+      "loss": 1.4775,
+      "step": 2820
+    },
+    {
+      "epoch": 2.604200323101777,
+      "grad_norm": 3.96875,
+      "learning_rate": 8.124749833222148e-05,
+      "loss": 1.3553,
+      "step": 2821
+    },
+    {
+      "epoch": 2.6051234710362334,
+      "grad_norm": 4.05859375,
+      "learning_rate": 8.124082721814544e-05,
+      "loss": 1.5845,
+      "step": 2822
+    },
+    {
+      "epoch": 2.60604661897069,
+      "grad_norm": 3.869140625,
+      "learning_rate": 8.123415610406938e-05,
+      "loss": 1.4312,
+      "step": 2823
+    },
+    {
+      "epoch": 2.606969766905147,
+      "grad_norm": 4.2421875,
+      "learning_rate": 8.122748498999333e-05,
+      "loss": 1.5161,
+      "step": 2824
+    },
+    {
+      "epoch": 2.607892914839603,
+      "grad_norm": 4.859375,
+      "learning_rate": 8.122081387591727e-05,
+      "loss": 1.5579,
+      "step": 2825
+    },
+    {
+      "epoch": 2.6088160627740598,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.121414276184123e-05,
+      "loss": 1.6077,
+      "step": 2826
+    },
+    {
+      "epoch": 2.6097392107085158,
+      "grad_norm": 4.51171875,
+      "learning_rate": 8.120747164776518e-05,
+      "loss": 1.458,
+      "step": 2827
+    },
+    {
+      "epoch": 2.6106623586429727,
+      "grad_norm": 4.68359375,
+      "learning_rate": 8.120080053368912e-05,
+      "loss": 1.5894,
+      "step": 2828
+    },
+    {
+      "epoch": 2.611585506577429,
+      "grad_norm": 7.1484375,
+      "learning_rate": 8.119412941961307e-05,
+      "loss": 1.5627,
+      "step": 2829
+    },
+    {
+      "epoch": 2.6125086545118856,
+      "grad_norm": 3.220703125,
+      "learning_rate": 8.118745830553703e-05,
+      "loss": 1.3048,
+      "step": 2830
+    },
+    {
+      "epoch": 2.613431802446342,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.118078719146099e-05,
+      "loss": 1.1713,
+      "step": 2831
+    },
+    {
+      "epoch": 2.6143549503807986,
+      "grad_norm": 5.55078125,
+      "learning_rate": 8.117411607738493e-05,
+      "loss": 1.8269,
+      "step": 2832
+    },
+    {
+      "epoch": 2.615278098315255,
+      "grad_norm": 6.3828125,
+      "learning_rate": 8.116744496330888e-05,
+      "loss": 1.3862,
+      "step": 2833
+    },
+    {
+      "epoch": 2.6162012462497115,
+      "grad_norm": 5.73046875,
+      "learning_rate": 8.116077384923282e-05,
+      "loss": 1.4813,
+      "step": 2834
+    },
+    {
+      "epoch": 2.617124394184168,
+      "grad_norm": 5.296875,
+      "learning_rate": 8.115410273515678e-05,
+      "loss": 1.5037,
+      "step": 2835
+    },
+    {
+      "epoch": 2.6180475421186244,
+      "grad_norm": 5.58203125,
+      "learning_rate": 8.114743162108073e-05,
+      "loss": 1.4653,
+      "step": 2836
+    },
+    {
+      "epoch": 2.618970690053081,
+      "grad_norm": 3.80078125,
+      "learning_rate": 8.114076050700467e-05,
+      "loss": 1.5056,
+      "step": 2837
+    },
+    {
+      "epoch": 2.6198938379875374,
+      "grad_norm": 5.07421875,
+      "learning_rate": 8.113408939292862e-05,
+      "loss": 1.1646,
+      "step": 2838
+    },
+    {
+      "epoch": 2.620816985921994,
+      "grad_norm": 5.6328125,
+      "learning_rate": 8.112741827885258e-05,
+      "loss": 1.4332,
+      "step": 2839
+    },
+    {
+      "epoch": 2.6217401338564503,
+      "grad_norm": 5.04296875,
+      "learning_rate": 8.112074716477652e-05,
+      "loss": 1.4924,
+      "step": 2840
+    },
+    {
+      "epoch": 2.6226632817909072,
+      "grad_norm": 5.1328125,
+      "learning_rate": 8.111407605070047e-05,
+      "loss": 1.533,
+      "step": 2841
+    },
+    {
+      "epoch": 2.6235864297253633,
+      "grad_norm": 5.02734375,
+      "learning_rate": 8.110740493662441e-05,
+      "loss": 1.5002,
+      "step": 2842
+    },
+    {
+      "epoch": 2.62450957765982,
+      "grad_norm": 4.66796875,
+      "learning_rate": 8.110073382254837e-05,
+      "loss": 1.2148,
+      "step": 2843
+    },
+    {
+      "epoch": 2.625432725594276,
+      "grad_norm": 3.798828125,
+      "learning_rate": 8.109406270847232e-05,
+      "loss": 1.2716,
+      "step": 2844
+    },
+    {
+      "epoch": 2.626355873528733,
+      "grad_norm": 4.8046875,
+      "learning_rate": 8.108739159439626e-05,
+      "loss": 1.6118,
+      "step": 2845
+    },
+    {
+      "epoch": 2.6272790214631896,
+      "grad_norm": 4.37890625,
+      "learning_rate": 8.108072048032021e-05,
+      "loss": 1.45,
+      "step": 2846
+    },
+    {
+      "epoch": 2.628202169397646,
+      "grad_norm": 5.1796875,
+      "learning_rate": 8.107404936624417e-05,
+      "loss": 1.5259,
+      "step": 2847
+    },
+    {
+      "epoch": 2.6291253173321025,
+      "grad_norm": 4.9609375,
+      "learning_rate": 8.106737825216811e-05,
+      "loss": 1.5132,
+      "step": 2848
+    },
+    {
+      "epoch": 2.630048465266559,
+      "grad_norm": 7.0859375,
+      "learning_rate": 8.106070713809206e-05,
+      "loss": 1.4602,
+      "step": 2849
+    },
+    {
+      "epoch": 2.6309716132010155,
+      "grad_norm": 3.787109375,
+      "learning_rate": 8.1054036024016e-05,
+      "loss": 1.4783,
+      "step": 2850
+    },
+    {
+      "epoch": 2.631894761135472,
+      "grad_norm": 4.7890625,
+      "learning_rate": 8.104736490993996e-05,
+      "loss": 1.5562,
+      "step": 2851
+    },
+    {
+      "epoch": 2.6328179090699284,
+      "grad_norm": 4.71484375,
+      "learning_rate": 8.104069379586391e-05,
+      "loss": 1.3136,
+      "step": 2852
+    },
+    {
+      "epoch": 2.633741057004385,
+      "grad_norm": 4.328125,
+      "learning_rate": 8.103402268178785e-05,
+      "loss": 1.3621,
+      "step": 2853
+    },
+    {
+      "epoch": 2.6346642049388413,
+      "grad_norm": 4.125,
+      "learning_rate": 8.10273515677118e-05,
+      "loss": 1.1642,
+      "step": 2854
+    },
+    {
+      "epoch": 2.635587352873298,
+      "grad_norm": 4.6171875,
+      "learning_rate": 8.102068045363576e-05,
+      "loss": 1.4047,
+      "step": 2855
+    },
+    {
+      "epoch": 2.6365105008077543,
+      "grad_norm": 4.4609375,
+      "learning_rate": 8.101400933955972e-05,
+      "loss": 1.437,
+      "step": 2856
+    },
+    {
+      "epoch": 2.6374336487422108,
+      "grad_norm": 3.759765625,
+      "learning_rate": 8.100733822548366e-05,
+      "loss": 1.3838,
+      "step": 2857
+    },
+    {
+      "epoch": 2.6383567966766677,
+      "grad_norm": 4.7890625,
+      "learning_rate": 8.100066711140761e-05,
+      "loss": 1.635,
+      "step": 2858
+    },
+    {
+      "epoch": 2.6392799446111237,
+      "grad_norm": 4.69921875,
+      "learning_rate": 8.099399599733157e-05,
+      "loss": 1.5483,
+      "step": 2859
+    },
+    {
+      "epoch": 2.6402030925455806,
+      "grad_norm": 4.609375,
+      "learning_rate": 8.098732488325551e-05,
+      "loss": 1.423,
+      "step": 2860
+    },
+    {
+      "epoch": 2.641126240480037,
+      "grad_norm": 3.818359375,
+      "learning_rate": 8.098065376917946e-05,
+      "loss": 1.6047,
+      "step": 2861
+    },
+    {
+      "epoch": 2.6420493884144935,
+      "grad_norm": 3.861328125,
+      "learning_rate": 8.09739826551034e-05,
+      "loss": 1.2986,
+      "step": 2862
+    },
+    {
+      "epoch": 2.64297253634895,
+      "grad_norm": 4.84765625,
+      "learning_rate": 8.096731154102736e-05,
+      "loss": 1.3867,
+      "step": 2863
+    },
+    {
+      "epoch": 2.6438956842834065,
+      "grad_norm": 4.296875,
+      "learning_rate": 8.096064042695131e-05,
+      "loss": 1.4014,
+      "step": 2864
+    },
+    {
+      "epoch": 2.644818832217863,
+      "grad_norm": 4.828125,
+      "learning_rate": 8.095396931287525e-05,
+      "loss": 1.3636,
+      "step": 2865
+    },
+    {
+      "epoch": 2.6457419801523194,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.09472981987992e-05,
+      "loss": 1.392,
+      "step": 2866
+    },
+    {
+      "epoch": 2.646665128086776,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.094062708472316e-05,
+      "loss": 1.4543,
+      "step": 2867
+    },
+    {
+      "epoch": 2.6475882760212324,
+      "grad_norm": 5.1015625,
+      "learning_rate": 8.09339559706471e-05,
+      "loss": 1.5229,
+      "step": 2868
+    },
+    {
+      "epoch": 2.648511423955689,
+      "grad_norm": 4.4375,
+      "learning_rate": 8.092728485657105e-05,
+      "loss": 1.7454,
+      "step": 2869
+    },
+    {
+      "epoch": 2.6494345718901453,
+      "grad_norm": 3.353515625,
+      "learning_rate": 8.0920613742495e-05,
+      "loss": 1.2609,
+      "step": 2870
+    },
+    {
+      "epoch": 2.6503577198246018,
+      "grad_norm": 3.962890625,
+      "learning_rate": 8.091394262841895e-05,
+      "loss": 1.6587,
+      "step": 2871
+    },
+    {
+      "epoch": 2.6512808677590582,
+      "grad_norm": 4.0546875,
+      "learning_rate": 8.09072715143429e-05,
+      "loss": 1.6035,
+      "step": 2872
+    },
+    {
+      "epoch": 2.652204015693515,
+      "grad_norm": 5.28125,
+      "learning_rate": 8.090060040026684e-05,
+      "loss": 1.5891,
+      "step": 2873
+    },
+    {
+      "epoch": 2.653127163627971,
+      "grad_norm": 6.20703125,
+      "learning_rate": 8.089392928619079e-05,
+      "loss": 1.3992,
+      "step": 2874
+    },
+    {
+      "epoch": 2.654050311562428,
+      "grad_norm": 5.20703125,
+      "learning_rate": 8.088725817211475e-05,
+      "loss": 1.5176,
+      "step": 2875
+    },
+    {
+      "epoch": 2.654973459496884,
+      "grad_norm": 4.68359375,
+      "learning_rate": 8.08805870580387e-05,
+      "loss": 1.3296,
+      "step": 2876
+    },
+    {
+      "epoch": 2.655896607431341,
+      "grad_norm": 3.998046875,
+      "learning_rate": 8.087391594396264e-05,
+      "loss": 1.4095,
+      "step": 2877
+    },
+    {
+      "epoch": 2.6568197553657975,
+      "grad_norm": 3.763671875,
+      "learning_rate": 8.086724482988659e-05,
+      "loss": 1.1417,
+      "step": 2878
+    },
+    {
+      "epoch": 2.657742903300254,
+      "grad_norm": 4.48046875,
+      "learning_rate": 8.086057371581054e-05,
+      "loss": 1.554,
+      "step": 2879
+    },
+    {
+      "epoch": 2.6586660512347104,
+      "grad_norm": 4.3359375,
+      "learning_rate": 8.085390260173449e-05,
+      "loss": 1.8213,
+      "step": 2880
+    },
+    {
+      "epoch": 2.659589199169167,
+      "grad_norm": 4.40234375,
+      "learning_rate": 8.084723148765845e-05,
+      "loss": 1.2975,
+      "step": 2881
+    },
+    {
+      "epoch": 2.6605123471036234,
+      "grad_norm": 5.6171875,
+      "learning_rate": 8.08405603735824e-05,
+      "loss": 1.7227,
+      "step": 2882
+    },
+    {
+      "epoch": 2.66143549503808,
+      "grad_norm": 5.8046875,
+      "learning_rate": 8.083388925950634e-05,
+      "loss": 1.468,
+      "step": 2883
+    },
+    {
+      "epoch": 2.6623586429725363,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.08272181454303e-05,
+      "loss": 1.3301,
+      "step": 2884
+    },
+    {
+      "epoch": 2.663281790906993,
+      "grad_norm": 4.7734375,
+      "learning_rate": 8.082054703135424e-05,
+      "loss": 1.5779,
+      "step": 2885
+    },
+    {
+      "epoch": 2.6642049388414493,
+      "grad_norm": 4.22265625,
+      "learning_rate": 8.081387591727819e-05,
+      "loss": 1.2297,
+      "step": 2886
+    },
+    {
+      "epoch": 2.6651280867759057,
+      "grad_norm": 3.833984375,
+      "learning_rate": 8.080720480320214e-05,
+      "loss": 1.3931,
+      "step": 2887
+    },
+    {
+      "epoch": 2.666051234710362,
+      "grad_norm": 4.70703125,
+      "learning_rate": 8.08005336891261e-05,
+      "loss": 1.4626,
+      "step": 2888
+    },
+    {
+      "epoch": 2.6669743826448187,
+      "grad_norm": 4.31640625,
+      "learning_rate": 8.079386257505004e-05,
+      "loss": 1.3386,
+      "step": 2889
+    },
+    {
+      "epoch": 2.6678975305792756,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.078719146097398e-05,
+      "loss": 1.5339,
+      "step": 2890
+    },
+    {
+      "epoch": 2.6688206785137316,
+      "grad_norm": 4.66796875,
+      "learning_rate": 8.078052034689793e-05,
+      "loss": 1.5344,
+      "step": 2891
+    },
+    {
+      "epoch": 2.6697438264481885,
+      "grad_norm": 5.1640625,
+      "learning_rate": 8.077384923282189e-05,
+      "loss": 1.4219,
+      "step": 2892
+    },
+    {
+      "epoch": 2.6706669743826446,
+      "grad_norm": 4.16796875,
+      "learning_rate": 8.076717811874583e-05,
+      "loss": 1.5151,
+      "step": 2893
+    },
+    {
+      "epoch": 2.6715901223171015,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.076050700466978e-05,
+      "loss": 1.4705,
+      "step": 2894
+    },
+    {
+      "epoch": 2.672513270251558,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.075383589059373e-05,
+      "loss": 1.7214,
+      "step": 2895
+    },
+    {
+      "epoch": 2.6734364181860144,
+      "grad_norm": 3.66015625,
+      "learning_rate": 8.074716477651768e-05,
+      "loss": 1.3899,
+      "step": 2896
+    },
+    {
+      "epoch": 2.674359566120471,
+      "grad_norm": 4.1015625,
+      "learning_rate": 8.074049366244163e-05,
+      "loss": 1.7043,
+      "step": 2897
+    },
+    {
+      "epoch": 2.6752827140549273,
+      "grad_norm": 4.83203125,
+      "learning_rate": 8.073382254836558e-05,
+      "loss": 1.5007,
+      "step": 2898
+    },
+    {
+      "epoch": 2.676205861989384,
+      "grad_norm": 4.73828125,
+      "learning_rate": 8.072715143428952e-05,
+      "loss": 1.4382,
+      "step": 2899
+    },
+    {
+      "epoch": 2.6771290099238403,
+      "grad_norm": 5.57421875,
+      "learning_rate": 8.072048032021348e-05,
+      "loss": 1.4292,
+      "step": 2900
+    },
+    {
+      "epoch": 2.6780521578582968,
+      "grad_norm": 4.1328125,
+      "learning_rate": 8.071380920613743e-05,
+      "loss": 1.2105,
+      "step": 2901
+    },
+    {
+      "epoch": 2.6789753057927532,
+      "grad_norm": 3.583984375,
+      "learning_rate": 8.070713809206137e-05,
+      "loss": 1.0671,
+      "step": 2902
+    },
+    {
+      "epoch": 2.6798984537272097,
+      "grad_norm": 3.64453125,
+      "learning_rate": 8.070046697798532e-05,
+      "loss": 1.4302,
+      "step": 2903
+    },
+    {
+      "epoch": 2.680821601661666,
+      "grad_norm": 3.712890625,
+      "learning_rate": 8.069379586390928e-05,
+      "loss": 1.5376,
+      "step": 2904
+    },
+    {
+      "epoch": 2.6817447495961226,
+      "grad_norm": 4.73046875,
+      "learning_rate": 8.068712474983322e-05,
+      "loss": 1.4299,
+      "step": 2905
+    },
+    {
+      "epoch": 2.682667897530579,
+      "grad_norm": 5.8671875,
+      "learning_rate": 8.068045363575718e-05,
+      "loss": 1.2048,
+      "step": 2906
+    },
+    {
+      "epoch": 2.683591045465036,
+      "grad_norm": 6.4375,
+      "learning_rate": 8.067378252168113e-05,
+      "loss": 1.4814,
+      "step": 2907
+    },
+    {
+      "epoch": 2.684514193399492,
+      "grad_norm": 4.36328125,
+      "learning_rate": 8.066711140760508e-05,
+      "loss": 1.3799,
+      "step": 2908
+    },
+    {
+      "epoch": 2.685437341333949,
+      "grad_norm": 5.19140625,
+      "learning_rate": 8.066044029352903e-05,
+      "loss": 1.6846,
+      "step": 2909
+    },
+    {
+      "epoch": 2.6863604892684054,
+      "grad_norm": 5.359375,
+      "learning_rate": 8.065376917945297e-05,
+      "loss": 1.0939,
+      "step": 2910
+    },
+    {
+      "epoch": 2.687283637202862,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.064709806537692e-05,
+      "loss": 1.5125,
+      "step": 2911
+    },
+    {
+      "epoch": 2.6882067851373184,
+      "grad_norm": 3.7109375,
+      "learning_rate": 8.064042695130088e-05,
+      "loss": 1.4348,
+      "step": 2912
+    },
+    {
+      "epoch": 2.689129933071775,
+      "grad_norm": 3.7109375,
+      "learning_rate": 8.063375583722482e-05,
+      "loss": 1.0852,
+      "step": 2913
+    },
+    {
+      "epoch": 2.6900530810062313,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.062708472314877e-05,
+      "loss": 1.5083,
+      "step": 2914
+    },
+    {
+      "epoch": 2.6909762289406878,
+      "grad_norm": 4.30859375,
+      "learning_rate": 8.062041360907272e-05,
+      "loss": 1.4761,
+      "step": 2915
+    },
+    {
+      "epoch": 2.6918993768751442,
+      "grad_norm": 4.20703125,
+      "learning_rate": 8.061374249499667e-05,
+      "loss": 1.4841,
+      "step": 2916
+    },
+    {
+      "epoch": 2.6928225248096007,
+      "grad_norm": 4.45703125,
+      "learning_rate": 8.060707138092062e-05,
+      "loss": 1.4371,
+      "step": 2917
+    },
+    {
+      "epoch": 2.693745672744057,
+      "grad_norm": 4.63671875,
+      "learning_rate": 8.060040026684457e-05,
+      "loss": 1.4978,
+      "step": 2918
+    },
+    {
+      "epoch": 2.6946688206785137,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.059372915276851e-05,
+      "loss": 1.5,
+      "step": 2919
+    },
+    {
+      "epoch": 2.69559196861297,
+      "grad_norm": 5.0390625,
+      "learning_rate": 8.058705803869247e-05,
+      "loss": 1.6108,
+      "step": 2920
+    },
+    {
+      "epoch": 2.6965151165474266,
+      "grad_norm": 4.54296875,
+      "learning_rate": 8.058038692461642e-05,
+      "loss": 1.8088,
+      "step": 2921
+    },
+    {
+      "epoch": 2.697438264481883,
+      "grad_norm": 4.83984375,
+      "learning_rate": 8.057371581054036e-05,
+      "loss": 1.7092,
+      "step": 2922
+    },
+    {
+      "epoch": 2.6983614124163395,
+      "grad_norm": 4.72265625,
+      "learning_rate": 8.05670446964643e-05,
+      "loss": 1.3564,
+      "step": 2923
+    },
+    {
+      "epoch": 2.6992845603507964,
+      "grad_norm": 3.76171875,
+      "learning_rate": 8.056037358238827e-05,
+      "loss": 1.5005,
+      "step": 2924
+    },
+    {
+      "epoch": 2.7002077082852525,
+      "grad_norm": 4.21484375,
+      "learning_rate": 8.055370246831221e-05,
+      "loss": 1.7368,
+      "step": 2925
+    },
+    {
+      "epoch": 2.7011308562197094,
+      "grad_norm": 4.859375,
+      "learning_rate": 8.054703135423616e-05,
+      "loss": 1.4758,
+      "step": 2926
+    },
+    {
+      "epoch": 2.702054004154166,
+      "grad_norm": 5.1328125,
+      "learning_rate": 8.05403602401601e-05,
+      "loss": 1.6182,
+      "step": 2927
+    },
+    {
+      "epoch": 2.7029771520886223,
+      "grad_norm": 4.36328125,
+      "learning_rate": 8.053368912608406e-05,
+      "loss": 1.1572,
+      "step": 2928
+    },
+    {
+      "epoch": 2.703900300023079,
+      "grad_norm": 4.296875,
+      "learning_rate": 8.0527018012008e-05,
+      "loss": 1.436,
+      "step": 2929
+    },
+    {
+      "epoch": 2.7048234479575353,
+      "grad_norm": 4.4765625,
+      "learning_rate": 8.052034689793195e-05,
+      "loss": 1.5447,
+      "step": 2930
+    },
+    {
+      "epoch": 2.7057465958919917,
+      "grad_norm": 5.203125,
+      "learning_rate": 8.051367578385591e-05,
+      "loss": 1.213,
+      "step": 2931
+    },
+    {
+      "epoch": 2.706669743826448,
+      "grad_norm": 3.0078125,
+      "learning_rate": 8.050700466977986e-05,
+      "loss": 1.1336,
+      "step": 2932
+    },
+    {
+      "epoch": 2.7075928917609047,
+      "grad_norm": 4.16015625,
+      "learning_rate": 8.050033355570381e-05,
+      "loss": 1.344,
+      "step": 2933
+    },
+    {
+      "epoch": 2.708516039695361,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.049366244162776e-05,
+      "loss": 1.4651,
+      "step": 2934
+    },
+    {
+      "epoch": 2.7094391876298176,
+      "grad_norm": 4.66015625,
+      "learning_rate": 8.04869913275517e-05,
+      "loss": 1.6338,
+      "step": 2935
+    },
+    {
+      "epoch": 2.710362335564274,
+      "grad_norm": 5.1171875,
+      "learning_rate": 8.048032021347565e-05,
+      "loss": 1.6042,
+      "step": 2936
+    },
+    {
+      "epoch": 2.7112854834987306,
+      "grad_norm": 5.953125,
+      "learning_rate": 8.047364909939961e-05,
+      "loss": 1.5095,
+      "step": 2937
+    },
+    {
+      "epoch": 2.712208631433187,
+      "grad_norm": 5.2109375,
+      "learning_rate": 8.046697798532356e-05,
+      "loss": 1.1489,
+      "step": 2938
+    },
+    {
+      "epoch": 2.713131779367644,
+      "grad_norm": 5.58984375,
+      "learning_rate": 8.04603068712475e-05,
+      "loss": 1.6794,
+      "step": 2939
+    },
+    {
+      "epoch": 2.7140549273021,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.045363575717145e-05,
+      "loss": 1.3555,
+      "step": 2940
+    },
+    {
+      "epoch": 2.714978075236557,
+      "grad_norm": 5.6796875,
+      "learning_rate": 8.04469646430954e-05,
+      "loss": 1.7373,
+      "step": 2941
+    },
+    {
+      "epoch": 2.715901223171013,
+      "grad_norm": 4.61328125,
+      "learning_rate": 8.044029352901935e-05,
+      "loss": 1.6501,
+      "step": 2942
+    },
+    {
+      "epoch": 2.71682437110547,
+      "grad_norm": 4.52734375,
+      "learning_rate": 8.04336224149433e-05,
+      "loss": 1.6292,
+      "step": 2943
+    },
+    {
+      "epoch": 2.7177475190399263,
+      "grad_norm": 4.48046875,
+      "learning_rate": 8.042695130086724e-05,
+      "loss": 1.646,
+      "step": 2944
+    },
+    {
+      "epoch": 2.7186706669743828,
+      "grad_norm": 3.8359375,
+      "learning_rate": 8.04202801867912e-05,
+      "loss": 1.7534,
+      "step": 2945
+    },
+    {
+      "epoch": 2.7195938149088392,
+      "grad_norm": 4.92578125,
+      "learning_rate": 8.041360907271515e-05,
+      "loss": 1.4316,
+      "step": 2946
+    },
+    {
+      "epoch": 2.7205169628432957,
+      "grad_norm": 5.2890625,
+      "learning_rate": 8.040693795863909e-05,
+      "loss": 1.2993,
+      "step": 2947
+    },
+    {
+      "epoch": 2.721440110777752,
+      "grad_norm": 5.42578125,
+      "learning_rate": 8.040026684456305e-05,
+      "loss": 1.3884,
+      "step": 2948
+    },
+    {
+      "epoch": 2.7223632587122086,
+      "grad_norm": 4.23046875,
+      "learning_rate": 8.0393595730487e-05,
+      "loss": 1.4124,
+      "step": 2949
+    },
+    {
+      "epoch": 2.723286406646665,
+      "grad_norm": 4.33984375,
+      "learning_rate": 8.038692461641094e-05,
+      "loss": 1.4229,
+      "step": 2950
+    },
+    {
+      "epoch": 2.7242095545811216,
+      "grad_norm": 4.8828125,
+      "learning_rate": 8.038025350233489e-05,
+      "loss": 1.4211,
+      "step": 2951
+    },
+    {
+      "epoch": 2.725132702515578,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.037358238825885e-05,
+      "loss": 1.6243,
+      "step": 2952
+    },
+    {
+      "epoch": 2.7260558504500345,
+      "grad_norm": 4.68359375,
+      "learning_rate": 8.036691127418279e-05,
+      "loss": 1.6033,
+      "step": 2953
+    },
+    {
+      "epoch": 2.726978998384491,
+      "grad_norm": 4.703125,
+      "learning_rate": 8.036024016010674e-05,
+      "loss": 1.4843,
+      "step": 2954
+    },
+    {
+      "epoch": 2.7279021463189475,
+      "grad_norm": 4.53515625,
+      "learning_rate": 8.035356904603068e-05,
+      "loss": 1.4834,
+      "step": 2955
+    },
+    {
+      "epoch": 2.7288252942534044,
+      "grad_norm": 4.51171875,
+      "learning_rate": 8.034689793195464e-05,
+      "loss": 1.5686,
+      "step": 2956
+    },
+    {
+      "epoch": 2.7297484421878604,
+      "grad_norm": 4.0546875,
+      "learning_rate": 8.03402268178786e-05,
+      "loss": 1.6052,
+      "step": 2957
+    },
+    {
+      "epoch": 2.7306715901223173,
+      "grad_norm": 4.01171875,
+      "learning_rate": 8.033355570380255e-05,
+      "loss": 1.5996,
+      "step": 2958
+    },
+    {
+      "epoch": 2.7315947380567733,
+      "grad_norm": 3.869140625,
+      "learning_rate": 8.032688458972649e-05,
+      "loss": 1.531,
+      "step": 2959
+    },
+    {
+      "epoch": 2.7325178859912302,
+      "grad_norm": 3.7421875,
+      "learning_rate": 8.032021347565044e-05,
+      "loss": 1.4917,
+      "step": 2960
+    },
+    {
+      "epoch": 2.7334410339256867,
+      "grad_norm": 4.53515625,
+      "learning_rate": 8.03135423615744e-05,
+      "loss": 1.74,
+      "step": 2961
+    },
+    {
+      "epoch": 2.734364181860143,
+      "grad_norm": 3.830078125,
+      "learning_rate": 8.030687124749834e-05,
+      "loss": 1.4828,
+      "step": 2962
+    },
+    {
+      "epoch": 2.7352873297945997,
+      "grad_norm": 3.822265625,
+      "learning_rate": 8.030020013342229e-05,
+      "loss": 1.7097,
+      "step": 2963
+    },
+    {
+      "epoch": 2.736210477729056,
+      "grad_norm": 4.3671875,
+      "learning_rate": 8.029352901934623e-05,
+      "loss": 1.2605,
+      "step": 2964
+    },
+    {
+      "epoch": 2.7371336256635126,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.028685790527019e-05,
+      "loss": 1.5151,
+      "step": 2965
+    },
+    {
+      "epoch": 2.738056773597969,
+      "grad_norm": 4.703125,
+      "learning_rate": 8.028018679119414e-05,
+      "loss": 1.3539,
+      "step": 2966
+    },
+    {
+      "epoch": 2.7389799215324255,
+      "grad_norm": 4.2421875,
+      "learning_rate": 8.027351567711808e-05,
+      "loss": 1.2014,
+      "step": 2967
+    },
+    {
+      "epoch": 2.739903069466882,
+      "grad_norm": 5.6796875,
+      "learning_rate": 8.026684456304203e-05,
+      "loss": 1.4695,
+      "step": 2968
+    },
+    {
+      "epoch": 2.7408262174013385,
+      "grad_norm": 4.64453125,
+      "learning_rate": 8.026017344896599e-05,
+      "loss": 1.1456,
+      "step": 2969
+    },
+    {
+      "epoch": 2.741749365335795,
+      "grad_norm": 4.2890625,
+      "learning_rate": 8.025350233488993e-05,
+      "loss": 1.6548,
+      "step": 2970
+    },
+    {
+      "epoch": 2.7426725132702514,
+      "grad_norm": 4.1640625,
+      "learning_rate": 8.024683122081388e-05,
+      "loss": 1.4988,
+      "step": 2971
+    },
+    {
+      "epoch": 2.743595661204708,
+      "grad_norm": 3.6953125,
+      "learning_rate": 8.024016010673782e-05,
+      "loss": 1.3997,
+      "step": 2972
+    },
+    {
+      "epoch": 2.744518809139165,
+      "grad_norm": 4.26171875,
+      "learning_rate": 8.023348899266178e-05,
+      "loss": 1.3767,
+      "step": 2973
+    },
+    {
+      "epoch": 2.745441957073621,
+      "grad_norm": 4.65234375,
+      "learning_rate": 8.022681787858573e-05,
+      "loss": 1.6104,
+      "step": 2974
+    },
+    {
+      "epoch": 2.7463651050080777,
+      "grad_norm": 5.97265625,
+      "learning_rate": 8.022014676450967e-05,
+      "loss": 1.4944,
+      "step": 2975
+    },
+    {
+      "epoch": 2.747288252942534,
+      "grad_norm": 4.44140625,
+      "learning_rate": 8.021347565043362e-05,
+      "loss": 1.6833,
+      "step": 2976
+    },
+    {
+      "epoch": 2.7482114008769907,
+      "grad_norm": 4.71484375,
+      "learning_rate": 8.020680453635758e-05,
+      "loss": 1.3259,
+      "step": 2977
+    },
+    {
+      "epoch": 2.749134548811447,
+      "grad_norm": 5.7265625,
+      "learning_rate": 8.020013342228152e-05,
+      "loss": 1.592,
+      "step": 2978
+    },
+    {
+      "epoch": 2.7500576967459036,
+      "grad_norm": 4.390625,
+      "learning_rate": 8.019346230820547e-05,
+      "loss": 1.7302,
+      "step": 2979
+    },
+    {
+      "epoch": 2.75098084468036,
+      "grad_norm": 3.69921875,
+      "learning_rate": 8.018679119412941e-05,
+      "loss": 1.7383,
+      "step": 2980
+    },
+    {
+      "epoch": 2.7519039926148166,
+      "grad_norm": 3.984375,
+      "learning_rate": 8.018012008005337e-05,
+      "loss": 1.5381,
+      "step": 2981
+    },
+    {
+      "epoch": 2.752827140549273,
+      "grad_norm": 3.962890625,
+      "learning_rate": 8.017344896597733e-05,
+      "loss": 1.3213,
+      "step": 2982
+    },
+    {
+      "epoch": 2.7537502884837295,
+      "grad_norm": 4.19921875,
+      "learning_rate": 8.016677785190128e-05,
+      "loss": 1.4248,
+      "step": 2983
+    },
+    {
+      "epoch": 2.754673436418186,
+      "grad_norm": 5.58203125,
+      "learning_rate": 8.016010673782522e-05,
+      "loss": 1.5579,
+      "step": 2984
+    },
+    {
+      "epoch": 2.7555965843526424,
+      "grad_norm": 4.140625,
+      "learning_rate": 8.015343562374917e-05,
+      "loss": 1.8506,
+      "step": 2985
+    },
+    {
+      "epoch": 2.756519732287099,
+      "grad_norm": 3.525390625,
+      "learning_rate": 8.014676450967313e-05,
+      "loss": 1.6011,
+      "step": 2986
+    },
+    {
+      "epoch": 2.7574428802215554,
+      "grad_norm": 3.6484375,
+      "learning_rate": 8.014009339559707e-05,
+      "loss": 1.4237,
+      "step": 2987
+    },
+    {
+      "epoch": 2.758366028156012,
+      "grad_norm": 4.046875,
+      "learning_rate": 8.013342228152102e-05,
+      "loss": 1.3433,
+      "step": 2988
+    },
+    {
+      "epoch": 2.7592891760904683,
+      "grad_norm": 4.2109375,
+      "learning_rate": 8.012675116744498e-05,
+      "loss": 1.5989,
+      "step": 2989
+    },
+    {
+      "epoch": 2.7602123240249252,
+      "grad_norm": 4.9765625,
+      "learning_rate": 8.012008005336892e-05,
+      "loss": 1.3562,
+      "step": 2990
+    },
+    {
+      "epoch": 2.7611354719593812,
+      "grad_norm": 5.31640625,
+      "learning_rate": 8.011340893929287e-05,
+      "loss": 1.3348,
+      "step": 2991
+    },
+    {
+      "epoch": 2.762058619893838,
+      "grad_norm": 4.69921875,
+      "learning_rate": 8.010673782521681e-05,
+      "loss": 1.3347,
+      "step": 2992
+    },
+    {
+      "epoch": 2.7629817678282946,
+      "grad_norm": 4.3515625,
+      "learning_rate": 8.010006671114077e-05,
+      "loss": 1.2853,
+      "step": 2993
+    },
+    {
+      "epoch": 2.763904915762751,
+      "grad_norm": 4.1796875,
+      "learning_rate": 8.009339559706472e-05,
+      "loss": 1.708,
+      "step": 2994
+    },
+    {
+      "epoch": 2.7648280636972076,
+      "grad_norm": 5.35546875,
+      "learning_rate": 8.008672448298866e-05,
+      "loss": 1.4734,
+      "step": 2995
+    },
+    {
+      "epoch": 2.765751211631664,
+      "grad_norm": 4.81640625,
+      "learning_rate": 8.008005336891261e-05,
+      "loss": 1.5682,
+      "step": 2996
+    },
+    {
+      "epoch": 2.7666743595661205,
+      "grad_norm": 4.43359375,
+      "learning_rate": 8.007338225483657e-05,
+      "loss": 1.2998,
+      "step": 2997
+    },
+    {
+      "epoch": 2.767597507500577,
+      "grad_norm": 5.17578125,
+      "learning_rate": 8.006671114076051e-05,
+      "loss": 1.2651,
+      "step": 2998
+    },
+    {
+      "epoch": 2.7685206554350335,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.006004002668446e-05,
+      "loss": 1.3644,
+      "step": 2999
+    },
+    {
+      "epoch": 2.76944380336949,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.00533689126084e-05,
+      "loss": 1.5076,
+      "step": 3000
+    },
+    {
+      "epoch": 2.76944380336949,
+      "eval_loss": 0.88525390625,
+      "eval_runtime": 23.3445,
+      "eval_samples_per_second": 976.717,
+      "eval_steps_per_second": 61.085,
+      "step": 3000
+    },
+    {
+      "epoch": 2.7703669513039464,
+      "grad_norm": 4.47265625,
+      "learning_rate": 8.004669779853236e-05,
+      "loss": 1.2139,
+      "step": 3001
+    },
+    {
+      "epoch": 2.771290099238403,
+      "grad_norm": 5.01171875,
+      "learning_rate": 8.004002668445631e-05,
+      "loss": 1.6943,
+      "step": 3002
+    },
+    {
+      "epoch": 2.7722132471728593,
+      "grad_norm": 3.98046875,
+      "learning_rate": 8.003335557038025e-05,
+      "loss": 1.4731,
+      "step": 3003
+    },
+    {
+      "epoch": 2.773136395107316,
+      "grad_norm": 4.62109375,
+      "learning_rate": 8.00266844563042e-05,
+      "loss": 1.6038,
+      "step": 3004
+    },
+    {
+      "epoch": 2.7740595430417727,
+      "grad_norm": 4.10546875,
+      "learning_rate": 8.002001334222816e-05,
+      "loss": 1.5725,
+      "step": 3005
+    },
+    {
+      "epoch": 2.7749826909762287,
+      "grad_norm": 5.21484375,
+      "learning_rate": 8.001334222815212e-05,
+      "loss": 1.6038,
+      "step": 3006
+    },
+    {
+      "epoch": 2.7759058389106857,
+      "grad_norm": 4.75,
+      "learning_rate": 8.000667111407606e-05,
+      "loss": 1.2842,
+      "step": 3007
+    },
+    {
+      "epoch": 2.7768289868451417,
+      "grad_norm": 4.6328125,
+      "learning_rate": 8e-05,
+      "loss": 1.6304,
+      "step": 3008
+    },
+    {
+      "epoch": 2.7777521347795986,
+      "grad_norm": 3.814453125,
+      "learning_rate": 7.999332888592395e-05,
+      "loss": 1.4319,
+      "step": 3009
+    },
+    {
+      "epoch": 2.778675282714055,
+      "grad_norm": 4.3828125,
+      "learning_rate": 7.998665777184791e-05,
+      "loss": 1.4135,
+      "step": 3010
+    },
+    {
+      "epoch": 2.7795984306485115,
+      "grad_norm": 4.87890625,
+      "learning_rate": 7.997998665777186e-05,
+      "loss": 1.5347,
+      "step": 3011
+    },
+    {
+      "epoch": 2.780521578582968,
+      "grad_norm": 3.580078125,
+      "learning_rate": 7.99733155436958e-05,
+      "loss": 1.4031,
+      "step": 3012
+    },
+    {
+      "epoch": 2.7814447265174245,
+      "grad_norm": 3.82421875,
+      "learning_rate": 7.996664442961975e-05,
+      "loss": 1.0766,
+      "step": 3013
+    },
+    {
+      "epoch": 2.782367874451881,
+      "grad_norm": 4.48046875,
+      "learning_rate": 7.99599733155437e-05,
+      "loss": 1.5474,
+      "step": 3014
+    },
+    {
+      "epoch": 2.7832910223863374,
+      "grad_norm": 3.998046875,
+      "learning_rate": 7.995330220146765e-05,
+      "loss": 1.4265,
+      "step": 3015
+    },
+    {
+      "epoch": 2.784214170320794,
+      "grad_norm": 4.55078125,
+      "learning_rate": 7.99466310873916e-05,
+      "loss": 0.9846,
+      "step": 3016
+    },
+    {
+      "epoch": 2.7851373182552503,
+      "grad_norm": 5.74609375,
+      "learning_rate": 7.993995997331554e-05,
+      "loss": 1.2429,
+      "step": 3017
+    },
+    {
+      "epoch": 2.786060466189707,
+      "grad_norm": 4.25390625,
+      "learning_rate": 7.99332888592395e-05,
+      "loss": 1.3086,
+      "step": 3018
+    },
+    {
+      "epoch": 2.7869836141241633,
+      "grad_norm": 4.234375,
+      "learning_rate": 7.992661774516345e-05,
+      "loss": 1.561,
+      "step": 3019
+    },
+    {
+      "epoch": 2.7879067620586198,
+      "grad_norm": 5.046875,
+      "learning_rate": 7.991994663108739e-05,
+      "loss": 1.2543,
+      "step": 3020
+    },
+    {
+      "epoch": 2.7888299099930762,
+      "grad_norm": 4.1328125,
+      "learning_rate": 7.991327551701134e-05,
+      "loss": 1.4614,
+      "step": 3021
+    },
+    {
+      "epoch": 2.789753057927533,
+      "grad_norm": 4.7734375,
+      "learning_rate": 7.99066044029353e-05,
+      "loss": 1.3506,
+      "step": 3022
+    },
+    {
+      "epoch": 2.790676205861989,
+      "grad_norm": 4.20703125,
+      "learning_rate": 7.989993328885924e-05,
+      "loss": 1.2957,
+      "step": 3023
+    },
+    {
+      "epoch": 2.791599353796446,
+      "grad_norm": 5.0859375,
+      "learning_rate": 7.989326217478319e-05,
+      "loss": 1.5923,
+      "step": 3024
+    },
+    {
+      "epoch": 2.792522501730902,
+      "grad_norm": 4.13671875,
+      "learning_rate": 7.988659106070713e-05,
+      "loss": 1.6541,
+      "step": 3025
+    },
+    {
+      "epoch": 2.793445649665359,
+      "grad_norm": 6.4921875,
+      "learning_rate": 7.987991994663109e-05,
+      "loss": 1.1464,
+      "step": 3026
+    },
+    {
+      "epoch": 2.7943687975998155,
+      "grad_norm": 5.17578125,
+      "learning_rate": 7.987324883255504e-05,
+      "loss": 1.7439,
+      "step": 3027
+    },
+    {
+      "epoch": 2.795291945534272,
+      "grad_norm": 4.87109375,
+      "learning_rate": 7.986657771847898e-05,
+      "loss": 1.8062,
+      "step": 3028
+    },
+    {
+      "epoch": 2.7962150934687284,
+      "grad_norm": 4.35546875,
+      "learning_rate": 7.985990660440293e-05,
+      "loss": 1.6379,
+      "step": 3029
+    },
+    {
+      "epoch": 2.797138241403185,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.985323549032689e-05,
+      "loss": 1.6523,
+      "step": 3030
+    },
+    {
+      "epoch": 2.7980613893376414,
+      "grad_norm": 4.4375,
+      "learning_rate": 7.984656437625085e-05,
+      "loss": 1.554,
+      "step": 3031
+    },
+    {
+      "epoch": 2.798984537272098,
+      "grad_norm": 4.46875,
+      "learning_rate": 7.983989326217479e-05,
+      "loss": 1.2727,
+      "step": 3032
+    },
+    {
+      "epoch": 2.7999076852065543,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.983322214809874e-05,
+      "loss": 1.4897,
+      "step": 3033
+    },
+    {
+      "epoch": 2.800830833141011,
+      "grad_norm": 5.390625,
+      "learning_rate": 7.98265510340227e-05,
+      "loss": 1.4395,
+      "step": 3034
+    },
+    {
+      "epoch": 2.8017539810754672,
+      "grad_norm": 4.69921875,
+      "learning_rate": 7.981987991994664e-05,
+      "loss": 1.3757,
+      "step": 3035
+    },
+    {
+      "epoch": 2.8026771290099237,
+      "grad_norm": 4.02734375,
+      "learning_rate": 7.981320880587059e-05,
+      "loss": 0.9637,
+      "step": 3036
+    },
+    {
+      "epoch": 2.80360027694438,
+      "grad_norm": 5.16015625,
+      "learning_rate": 7.980653769179453e-05,
+      "loss": 1.3699,
+      "step": 3037
+    },
+    {
+      "epoch": 2.8045234248788367,
+      "grad_norm": 4.74609375,
+      "learning_rate": 7.979986657771849e-05,
+      "loss": 1.3933,
+      "step": 3038
+    },
+    {
+      "epoch": 2.8054465728132936,
+      "grad_norm": 4.25,
+      "learning_rate": 7.979319546364244e-05,
+      "loss": 1.5476,
+      "step": 3039
+    },
+    {
+      "epoch": 2.8063697207477496,
+      "grad_norm": 4.73046875,
+      "learning_rate": 7.978652434956638e-05,
+      "loss": 1.3389,
+      "step": 3040
+    },
+    {
+      "epoch": 2.8072928686822065,
+      "grad_norm": 3.818359375,
+      "learning_rate": 7.977985323549033e-05,
+      "loss": 1.4015,
+      "step": 3041
+    },
+    {
+      "epoch": 2.808216016616663,
+      "grad_norm": 4.1640625,
+      "learning_rate": 7.977318212141429e-05,
+      "loss": 1.6948,
+      "step": 3042
+    },
+    {
+      "epoch": 2.8091391645511194,
+      "grad_norm": 4.05078125,
+      "learning_rate": 7.976651100733823e-05,
+      "loss": 1.3669,
+      "step": 3043
+    },
+    {
+      "epoch": 2.810062312485576,
+      "grad_norm": 4.84375,
+      "learning_rate": 7.975983989326218e-05,
+      "loss": 1.5364,
+      "step": 3044
+    },
+    {
+      "epoch": 2.8109854604200324,
+      "grad_norm": 5.16015625,
+      "learning_rate": 7.975316877918612e-05,
+      "loss": 1.8181,
+      "step": 3045
+    },
+    {
+      "epoch": 2.811908608354489,
+      "grad_norm": 5.421875,
+      "learning_rate": 7.974649766511008e-05,
+      "loss": 1.1215,
+      "step": 3046
+    },
+    {
+      "epoch": 2.8128317562889453,
+      "grad_norm": 5.44140625,
+      "learning_rate": 7.973982655103403e-05,
+      "loss": 1.592,
+      "step": 3047
+    },
+    {
+      "epoch": 2.813754904223402,
+      "grad_norm": 3.43359375,
+      "learning_rate": 7.973315543695797e-05,
+      "loss": 1.6582,
+      "step": 3048
+    },
+    {
+      "epoch": 2.8146780521578583,
+      "grad_norm": 3.857421875,
+      "learning_rate": 7.972648432288192e-05,
+      "loss": 1.2366,
+      "step": 3049
+    },
+    {
+      "epoch": 2.8156012000923147,
+      "grad_norm": 3.736328125,
+      "learning_rate": 7.971981320880588e-05,
+      "loss": 1.088,
+      "step": 3050
+    },
+    {
+      "epoch": 2.816524348026771,
+      "grad_norm": 3.775390625,
+      "learning_rate": 7.971314209472982e-05,
+      "loss": 1.5354,
+      "step": 3051
+    },
+    {
+      "epoch": 2.8174474959612277,
+      "grad_norm": 4.453125,
+      "learning_rate": 7.970647098065377e-05,
+      "loss": 1.5538,
+      "step": 3052
+    },
+    {
+      "epoch": 2.818370643895684,
+      "grad_norm": 4.2734375,
+      "learning_rate": 7.969979986657771e-05,
+      "loss": 1.5542,
+      "step": 3053
+    },
+    {
+      "epoch": 2.8192937918301406,
+      "grad_norm": 4.72265625,
+      "learning_rate": 7.969312875250167e-05,
+      "loss": 1.554,
+      "step": 3054
+    },
+    {
+      "epoch": 2.820216939764597,
+      "grad_norm": 4.37890625,
+      "learning_rate": 7.968645763842562e-05,
+      "loss": 1.4104,
+      "step": 3055
+    },
+    {
+      "epoch": 2.821140087699054,
+      "grad_norm": 4.359375,
+      "learning_rate": 7.967978652434958e-05,
+      "loss": 1.3843,
+      "step": 3056
+    },
+    {
+      "epoch": 2.82206323563351,
+      "grad_norm": 4.87890625,
+      "learning_rate": 7.967311541027352e-05,
+      "loss": 1.3992,
+      "step": 3057
+    },
+    {
+      "epoch": 2.822986383567967,
+      "grad_norm": 5.265625,
+      "learning_rate": 7.966644429619747e-05,
+      "loss": 1.3093,
+      "step": 3058
+    },
+    {
+      "epoch": 2.8239095315024234,
+      "grad_norm": 4.47265625,
+      "learning_rate": 7.965977318212143e-05,
+      "loss": 1.4583,
+      "step": 3059
+    },
+    {
+      "epoch": 2.82483267943688,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.965310206804537e-05,
+      "loss": 1.4933,
+      "step": 3060
+    },
+    {
+      "epoch": 2.8257558273713363,
+      "grad_norm": 4.22265625,
+      "learning_rate": 7.964643095396932e-05,
+      "loss": 1.5781,
+      "step": 3061
+    },
+    {
+      "epoch": 2.826678975305793,
+      "grad_norm": 4.32421875,
+      "learning_rate": 7.963975983989326e-05,
+      "loss": 1.5938,
+      "step": 3062
+    },
+    {
+      "epoch": 2.8276021232402493,
+      "grad_norm": 4.46484375,
+      "learning_rate": 7.963308872581722e-05,
+      "loss": 1.3395,
+      "step": 3063
+    },
+    {
+      "epoch": 2.8285252711747058,
+      "grad_norm": 3.58984375,
+      "learning_rate": 7.962641761174117e-05,
+      "loss": 1.3961,
+      "step": 3064
+    },
+    {
+      "epoch": 2.8294484191091622,
+      "grad_norm": 4.73046875,
+      "learning_rate": 7.961974649766511e-05,
+      "loss": 1.6785,
+      "step": 3065
+    },
+    {
+      "epoch": 2.8303715670436187,
+      "grad_norm": 3.82421875,
+      "learning_rate": 7.961307538358906e-05,
+      "loss": 1.4149,
+      "step": 3066
+    },
+    {
+      "epoch": 2.831294714978075,
+      "grad_norm": 4.61328125,
+      "learning_rate": 7.960640426951302e-05,
+      "loss": 1.429,
+      "step": 3067
+    },
+    {
+      "epoch": 2.8322178629125316,
+      "grad_norm": 3.767578125,
+      "learning_rate": 7.959973315543696e-05,
+      "loss": 1.3718,
+      "step": 3068
+    },
+    {
+      "epoch": 2.833141010846988,
+      "grad_norm": 3.95703125,
+      "learning_rate": 7.959306204136091e-05,
+      "loss": 1.2856,
+      "step": 3069
+    },
+    {
+      "epoch": 2.8340641587814446,
+      "grad_norm": 3.587890625,
+      "learning_rate": 7.958639092728485e-05,
+      "loss": 1.0807,
+      "step": 3070
+    },
+    {
+      "epoch": 2.8349873067159015,
+      "grad_norm": 3.115234375,
+      "learning_rate": 7.957971981320881e-05,
+      "loss": 1.0566,
+      "step": 3071
+    },
+    {
+      "epoch": 2.8359104546503575,
+      "grad_norm": 4.2421875,
+      "learning_rate": 7.957304869913276e-05,
+      "loss": 1.4578,
+      "step": 3072
+    },
+    {
+      "epoch": 2.8368336025848144,
+      "grad_norm": 3.818359375,
+      "learning_rate": 7.95663775850567e-05,
+      "loss": 1.7263,
+      "step": 3073
+    },
+    {
+      "epoch": 2.8377567505192705,
+      "grad_norm": 3.595703125,
+      "learning_rate": 7.955970647098065e-05,
+      "loss": 1.8,
+      "step": 3074
+    },
+    {
+      "epoch": 2.8386798984537274,
+      "grad_norm": 4.0859375,
+      "learning_rate": 7.955303535690461e-05,
+      "loss": 1.6299,
+      "step": 3075
+    },
+    {
+      "epoch": 2.839603046388184,
+      "grad_norm": 4.97265625,
+      "learning_rate": 7.954636424282855e-05,
+      "loss": 1.5046,
+      "step": 3076
+    },
+    {
+      "epoch": 2.8405261943226403,
+      "grad_norm": 4.3046875,
+      "learning_rate": 7.95396931287525e-05,
+      "loss": 1.172,
+      "step": 3077
+    },
+    {
+      "epoch": 2.8414493422570968,
+      "grad_norm": 4.88671875,
+      "learning_rate": 7.953302201467644e-05,
+      "loss": 1.3704,
+      "step": 3078
+    },
+    {
+      "epoch": 2.8423724901915532,
+      "grad_norm": 4.609375,
+      "learning_rate": 7.95263509006004e-05,
+      "loss": 1.5088,
+      "step": 3079
+    },
+    {
+      "epoch": 2.8432956381260097,
+      "grad_norm": 4.73828125,
+      "learning_rate": 7.951967978652435e-05,
+      "loss": 1.5774,
+      "step": 3080
+    },
+    {
+      "epoch": 2.844218786060466,
+      "grad_norm": 3.51171875,
+      "learning_rate": 7.951300867244831e-05,
+      "loss": 1.4055,
+      "step": 3081
+    },
+    {
+      "epoch": 2.8451419339949227,
+      "grad_norm": 3.75390625,
+      "learning_rate": 7.950633755837225e-05,
+      "loss": 1.625,
+      "step": 3082
+    },
+    {
+      "epoch": 2.846065081929379,
+      "grad_norm": 3.958984375,
+      "learning_rate": 7.949966644429621e-05,
+      "loss": 1.3958,
+      "step": 3083
+    },
+    {
+      "epoch": 2.8469882298638356,
+      "grad_norm": 5.03515625,
+      "learning_rate": 7.949299533022016e-05,
+      "loss": 1.6287,
+      "step": 3084
+    },
+    {
+      "epoch": 2.847911377798292,
+      "grad_norm": 4.33984375,
+      "learning_rate": 7.94863242161441e-05,
+      "loss": 1.5928,
+      "step": 3085
+    },
+    {
+      "epoch": 2.8488345257327485,
+      "grad_norm": 6.01171875,
+      "learning_rate": 7.947965310206805e-05,
+      "loss": 1.4016,
+      "step": 3086
+    },
+    {
+      "epoch": 2.849757673667205,
+      "grad_norm": 4.10546875,
+      "learning_rate": 7.947298198799201e-05,
+      "loss": 1.6836,
+      "step": 3087
+    },
+    {
+      "epoch": 2.850680821601662,
+      "grad_norm": 4.89453125,
+      "learning_rate": 7.946631087391595e-05,
+      "loss": 1.3638,
+      "step": 3088
+    },
+    {
+      "epoch": 2.851603969536118,
+      "grad_norm": 4.359375,
+      "learning_rate": 7.94596397598399e-05,
+      "loss": 1.4561,
+      "step": 3089
+    },
+    {
+      "epoch": 2.852527117470575,
+      "grad_norm": 4.0546875,
+      "learning_rate": 7.945296864576384e-05,
+      "loss": 1.8445,
+      "step": 3090
+    },
+    {
+      "epoch": 2.853450265405031,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.94462975316878e-05,
+      "loss": 1.2035,
+      "step": 3091
+    },
+    {
+      "epoch": 2.854373413339488,
+      "grad_norm": 4.06640625,
+      "learning_rate": 7.943962641761175e-05,
+      "loss": 1.3899,
+      "step": 3092
+    },
+    {
+      "epoch": 2.8552965612739443,
+      "grad_norm": 3.625,
+      "learning_rate": 7.94329553035357e-05,
+      "loss": 1.3829,
+      "step": 3093
+    },
+    {
+      "epoch": 2.8562197092084007,
+      "grad_norm": 3.96875,
+      "learning_rate": 7.942628418945964e-05,
+      "loss": 1.3972,
+      "step": 3094
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 4.50390625,
+      "learning_rate": 7.94196130753836e-05,
+      "loss": 1.5288,
+      "step": 3095
+    },
+    {
+      "epoch": 2.8580660050773137,
+      "grad_norm": 4.046875,
+      "learning_rate": 7.941294196130754e-05,
+      "loss": 1.5315,
+      "step": 3096
+    },
+    {
+      "epoch": 2.85898915301177,
+      "grad_norm": 3.447265625,
+      "learning_rate": 7.940627084723149e-05,
+      "loss": 1.2685,
+      "step": 3097
+    },
+    {
+      "epoch": 2.8599123009462266,
+      "grad_norm": 3.884765625,
+      "learning_rate": 7.939959973315543e-05,
+      "loss": 1.749,
+      "step": 3098
+    },
+    {
+      "epoch": 2.860835448880683,
+      "grad_norm": 4.2421875,
+      "learning_rate": 7.939292861907939e-05,
+      "loss": 1.6567,
+      "step": 3099
+    },
+    {
+      "epoch": 2.8617585968151396,
+      "grad_norm": 3.806640625,
+      "learning_rate": 7.938625750500334e-05,
+      "loss": 1.6604,
+      "step": 3100
+    },
+    {
+      "epoch": 2.862681744749596,
+      "grad_norm": 4.24609375,
+      "learning_rate": 7.937958639092728e-05,
+      "loss": 1.4004,
+      "step": 3101
+    },
+    {
+      "epoch": 2.8636048926840525,
+      "grad_norm": 4.20703125,
+      "learning_rate": 7.937291527685123e-05,
+      "loss": 1.5156,
+      "step": 3102
+    },
+    {
+      "epoch": 2.864528040618509,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.936624416277519e-05,
+      "loss": 1.3304,
+      "step": 3103
+    },
+    {
+      "epoch": 2.8654511885529654,
+      "grad_norm": 4.59765625,
+      "learning_rate": 7.935957304869913e-05,
+      "loss": 1.6387,
+      "step": 3104
+    },
+    {
+      "epoch": 2.8663743364874223,
+      "grad_norm": 4.2734375,
+      "learning_rate": 7.935290193462308e-05,
+      "loss": 0.9982,
+      "step": 3105
+    },
+    {
+      "epoch": 2.8672974844218784,
+      "grad_norm": 4.83984375,
+      "learning_rate": 7.934623082054704e-05,
+      "loss": 1.4626,
+      "step": 3106
+    },
+    {
+      "epoch": 2.8682206323563353,
+      "grad_norm": 4.44140625,
+      "learning_rate": 7.933955970647098e-05,
+      "loss": 1.5605,
+      "step": 3107
+    },
+    {
+      "epoch": 2.8691437802907918,
+      "grad_norm": 3.548828125,
+      "learning_rate": 7.933288859239494e-05,
+      "loss": 1.6394,
+      "step": 3108
+    },
+    {
+      "epoch": 2.8700669282252482,
+      "grad_norm": 3.716796875,
+      "learning_rate": 7.932621747831889e-05,
+      "loss": 1.3804,
+      "step": 3109
+    },
+    {
+      "epoch": 2.8709900761597047,
+      "grad_norm": 5.546875,
+      "learning_rate": 7.931954636424283e-05,
+      "loss": 1.3906,
+      "step": 3110
+    },
+    {
+      "epoch": 2.871913224094161,
+      "grad_norm": 3.833984375,
+      "learning_rate": 7.931287525016678e-05,
+      "loss": 1.3103,
+      "step": 3111
+    },
+    {
+      "epoch": 2.8728363720286176,
+      "grad_norm": 4.5390625,
+      "learning_rate": 7.930620413609074e-05,
+      "loss": 1.6118,
+      "step": 3112
+    },
+    {
+      "epoch": 2.873759519963074,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.929953302201468e-05,
+      "loss": 1.6033,
+      "step": 3113
+    },
+    {
+      "epoch": 2.8746826678975306,
+      "grad_norm": 3.904296875,
+      "learning_rate": 7.929286190793863e-05,
+      "loss": 1.6411,
+      "step": 3114
+    },
+    {
+      "epoch": 2.875605815831987,
+      "grad_norm": 3.654296875,
+      "learning_rate": 7.928619079386257e-05,
+      "loss": 1.5898,
+      "step": 3115
+    },
+    {
+      "epoch": 2.8765289637664435,
+      "grad_norm": 5.03515625,
+      "learning_rate": 7.927951967978653e-05,
+      "loss": 1.1406,
+      "step": 3116
+    },
+    {
+      "epoch": 2.8774521117009,
+      "grad_norm": 4.6171875,
+      "learning_rate": 7.927284856571048e-05,
+      "loss": 1.5261,
+      "step": 3117
+    },
+    {
+      "epoch": 2.8783752596353565,
+      "grad_norm": 3.958984375,
+      "learning_rate": 7.926617745163442e-05,
+      "loss": 1.6729,
+      "step": 3118
+    },
+    {
+      "epoch": 2.879298407569813,
+      "grad_norm": 3.185546875,
+      "learning_rate": 7.925950633755837e-05,
+      "loss": 1.3055,
+      "step": 3119
+    },
+    {
+      "epoch": 2.8802215555042694,
+      "grad_norm": 4.01171875,
+      "learning_rate": 7.925283522348233e-05,
+      "loss": 1.4153,
+      "step": 3120
+    },
+    {
+      "epoch": 2.881144703438726,
+      "grad_norm": 4.5703125,
+      "learning_rate": 7.924616410940627e-05,
+      "loss": 1.5024,
+      "step": 3121
+    },
+    {
+      "epoch": 2.8820678513731828,
+      "grad_norm": 3.7578125,
+      "learning_rate": 7.923949299533022e-05,
+      "loss": 1.6348,
+      "step": 3122
+    },
+    {
+      "epoch": 2.882990999307639,
+      "grad_norm": 4.06640625,
+      "learning_rate": 7.923282188125416e-05,
+      "loss": 1.6033,
+      "step": 3123
+    },
+    {
+      "epoch": 2.8839141472420957,
+      "grad_norm": 4.30859375,
+      "learning_rate": 7.922615076717812e-05,
+      "loss": 1.2721,
+      "step": 3124
+    },
+    {
+      "epoch": 2.884837295176552,
+      "grad_norm": 5.7265625,
+      "learning_rate": 7.921947965310207e-05,
+      "loss": 1.4175,
+      "step": 3125
+    },
+    {
+      "epoch": 2.8857604431110087,
+      "grad_norm": 4.890625,
+      "learning_rate": 7.921280853902601e-05,
+      "loss": 1.5071,
+      "step": 3126
+    },
+    {
+      "epoch": 2.886683591045465,
+      "grad_norm": 4.78515625,
+      "learning_rate": 7.920613742494996e-05,
+      "loss": 1.4751,
+      "step": 3127
+    },
+    {
+      "epoch": 2.8876067389799216,
+      "grad_norm": 4.05859375,
+      "learning_rate": 7.919946631087392e-05,
+      "loss": 1.4609,
+      "step": 3128
+    },
+    {
+      "epoch": 2.888529886914378,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.919279519679786e-05,
+      "loss": 1.364,
+      "step": 3129
+    },
+    {
+      "epoch": 2.8894530348488345,
+      "grad_norm": 3.8671875,
+      "learning_rate": 7.918612408272181e-05,
+      "loss": 1.4114,
+      "step": 3130
+    },
+    {
+      "epoch": 2.890376182783291,
+      "grad_norm": 3.98828125,
+      "learning_rate": 7.917945296864577e-05,
+      "loss": 1.6018,
+      "step": 3131
+    },
+    {
+      "epoch": 2.8912993307177475,
+      "grad_norm": 3.833984375,
+      "learning_rate": 7.917278185456973e-05,
+      "loss": 1.4075,
+      "step": 3132
+    },
+    {
+      "epoch": 2.892222478652204,
+      "grad_norm": 4.9453125,
+      "learning_rate": 7.916611074049367e-05,
+      "loss": 1.7815,
+      "step": 3133
+    },
+    {
+      "epoch": 2.8931456265866604,
+      "grad_norm": 5.36328125,
+      "learning_rate": 7.915943962641762e-05,
+      "loss": 1.5366,
+      "step": 3134
+    },
+    {
+      "epoch": 2.894068774521117,
+      "grad_norm": 5.1328125,
+      "learning_rate": 7.915276851234156e-05,
+      "loss": 1.5952,
+      "step": 3135
+    },
+    {
+      "epoch": 2.8949919224555734,
+      "grad_norm": 3.9765625,
+      "learning_rate": 7.914609739826552e-05,
+      "loss": 1.3668,
+      "step": 3136
+    },
+    {
+      "epoch": 2.8959150703900303,
+      "grad_norm": 3.447265625,
+      "learning_rate": 7.913942628418947e-05,
+      "loss": 1.2738,
+      "step": 3137
+    },
+    {
+      "epoch": 2.8968382183244863,
+      "grad_norm": 4.28125,
+      "learning_rate": 7.913275517011341e-05,
+      "loss": 1.5669,
+      "step": 3138
+    },
+    {
+      "epoch": 2.897761366258943,
+      "grad_norm": 4.6953125,
+      "learning_rate": 7.912608405603736e-05,
+      "loss": 1.7588,
+      "step": 3139
+    },
+    {
+      "epoch": 2.8986845141933992,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.911941294196132e-05,
+      "loss": 1.6252,
+      "step": 3140
+    },
+    {
+      "epoch": 2.899607662127856,
+      "grad_norm": 3.73046875,
+      "learning_rate": 7.911274182788526e-05,
+      "loss": 1.6819,
+      "step": 3141
+    },
+    {
+      "epoch": 2.9005308100623126,
+      "grad_norm": 4.328125,
+      "learning_rate": 7.910607071380921e-05,
+      "loss": 1.6833,
+      "step": 3142
+    },
+    {
+      "epoch": 2.901453957996769,
+      "grad_norm": 3.216796875,
+      "learning_rate": 7.909939959973315e-05,
+      "loss": 1.4573,
+      "step": 3143
+    },
+    {
+      "epoch": 2.9023771059312256,
+      "grad_norm": 3.982421875,
+      "learning_rate": 7.909272848565711e-05,
+      "loss": 1.3989,
+      "step": 3144
+    },
+    {
+      "epoch": 2.903300253865682,
+      "grad_norm": 5.31640625,
+      "learning_rate": 7.908605737158106e-05,
+      "loss": 1.4824,
+      "step": 3145
+    },
+    {
+      "epoch": 2.9042234018001385,
+      "grad_norm": 5.55859375,
+      "learning_rate": 7.9079386257505e-05,
+      "loss": 1.3501,
+      "step": 3146
+    },
+    {
+      "epoch": 2.905146549734595,
+      "grad_norm": 4.67578125,
+      "learning_rate": 7.907271514342895e-05,
+      "loss": 1.5884,
+      "step": 3147
+    },
+    {
+      "epoch": 2.9060696976690514,
+      "grad_norm": 4.1171875,
+      "learning_rate": 7.906604402935291e-05,
+      "loss": 1.4775,
+      "step": 3148
+    },
+    {
+      "epoch": 2.906992845603508,
+      "grad_norm": 3.666015625,
+      "learning_rate": 7.905937291527685e-05,
+      "loss": 1.4749,
+      "step": 3149
+    },
+    {
+      "epoch": 2.9079159935379644,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.90527018012008e-05,
+      "loss": 1.2062,
+      "step": 3150
+    },
+    {
+      "epoch": 2.908839141472421,
+      "grad_norm": 3.677734375,
+      "learning_rate": 7.904603068712475e-05,
+      "loss": 1.4761,
+      "step": 3151
+    },
+    {
+      "epoch": 2.9097622894068773,
+      "grad_norm": 3.62890625,
+      "learning_rate": 7.90393595730487e-05,
+      "loss": 1.478,
+      "step": 3152
+    },
+    {
+      "epoch": 2.910685437341334,
+      "grad_norm": 4.64453125,
+      "learning_rate": 7.903268845897265e-05,
+      "loss": 1.7812,
+      "step": 3153
+    },
+    {
+      "epoch": 2.9116085852757907,
+      "grad_norm": 4.11328125,
+      "learning_rate": 7.90260173448966e-05,
+      "loss": 1.7771,
+      "step": 3154
+    },
+    {
+      "epoch": 2.9125317332102467,
+      "grad_norm": 5.26171875,
+      "learning_rate": 7.901934623082054e-05,
+      "loss": 1.3893,
+      "step": 3155
+    },
+    {
+      "epoch": 2.9134548811447036,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.90126751167445e-05,
+      "loss": 1.5933,
+      "step": 3156
+    },
+    {
+      "epoch": 2.9143780290791597,
+      "grad_norm": 3.88671875,
+      "learning_rate": 7.900600400266846e-05,
+      "loss": 1.3904,
+      "step": 3157
+    },
+    {
+      "epoch": 2.9153011770136166,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.89993328885924e-05,
+      "loss": 1.5576,
+      "step": 3158
+    },
+    {
+      "epoch": 2.916224324948073,
+      "grad_norm": 3.8671875,
+      "learning_rate": 7.899266177451635e-05,
+      "loss": 1.4923,
+      "step": 3159
+    },
+    {
+      "epoch": 2.9171474728825295,
+      "grad_norm": 3.552734375,
+      "learning_rate": 7.89859906604403e-05,
+      "loss": 1.2365,
+      "step": 3160
+    },
+    {
+      "epoch": 2.918070620816986,
+      "grad_norm": 4.58203125,
+      "learning_rate": 7.897931954636425e-05,
+      "loss": 1.6409,
+      "step": 3161
+    },
+    {
+      "epoch": 2.9189937687514425,
+      "grad_norm": 4.65234375,
+      "learning_rate": 7.89726484322882e-05,
+      "loss": 1.5164,
+      "step": 3162
+    },
+    {
+      "epoch": 2.919916916685899,
+      "grad_norm": 4.76171875,
+      "learning_rate": 7.896597731821214e-05,
+      "loss": 1.5242,
+      "step": 3163
+    },
+    {
+      "epoch": 2.9208400646203554,
+      "grad_norm": 4.92578125,
+      "learning_rate": 7.895930620413609e-05,
+      "loss": 1.675,
+      "step": 3164
+    },
+    {
+      "epoch": 2.921763212554812,
+      "grad_norm": 4.453125,
+      "learning_rate": 7.895263509006005e-05,
+      "loss": 1.4365,
+      "step": 3165
+    },
+    {
+      "epoch": 2.9226863604892683,
+      "grad_norm": 4.82421875,
+      "learning_rate": 7.8945963975984e-05,
+      "loss": 1.6724,
+      "step": 3166
+    },
+    {
+      "epoch": 2.923609508423725,
+      "grad_norm": 5.27734375,
+      "learning_rate": 7.893929286190794e-05,
+      "loss": 1.2094,
+      "step": 3167
+    },
+    {
+      "epoch": 2.9245326563581813,
+      "grad_norm": 4.8203125,
+      "learning_rate": 7.893262174783189e-05,
+      "loss": 1.636,
+      "step": 3168
+    },
+    {
+      "epoch": 2.9254558042926377,
+      "grad_norm": 4.6484375,
+      "learning_rate": 7.892595063375584e-05,
+      "loss": 1.5591,
+      "step": 3169
+    },
+    {
+      "epoch": 2.926378952227094,
+      "grad_norm": 4.0078125,
+      "learning_rate": 7.891927951967979e-05,
+      "loss": 1.5181,
+      "step": 3170
+    },
+    {
+      "epoch": 2.927302100161551,
+      "grad_norm": 4.91796875,
+      "learning_rate": 7.891260840560374e-05,
+      "loss": 1.1419,
+      "step": 3171
+    },
+    {
+      "epoch": 2.928225248096007,
+      "grad_norm": 5.44140625,
+      "learning_rate": 7.890593729152768e-05,
+      "loss": 1.9265,
+      "step": 3172
+    },
+    {
+      "epoch": 2.929148396030464,
+      "grad_norm": 3.99609375,
+      "learning_rate": 7.889926617745164e-05,
+      "loss": 1.4272,
+      "step": 3173
+    },
+    {
+      "epoch": 2.9300715439649205,
+      "grad_norm": 2.912109375,
+      "learning_rate": 7.889259506337558e-05,
+      "loss": 1.7136,
+      "step": 3174
+    },
+    {
+      "epoch": 2.930994691899377,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.888592394929953e-05,
+      "loss": 1.2913,
+      "step": 3175
+    },
+    {
+      "epoch": 2.9319178398338335,
+      "grad_norm": 4.00390625,
+      "learning_rate": 7.887925283522349e-05,
+      "loss": 1.3015,
+      "step": 3176
+    },
+    {
+      "epoch": 2.93284098776829,
+      "grad_norm": 4.46875,
+      "learning_rate": 7.887258172114743e-05,
+      "loss": 1.552,
+      "step": 3177
+    },
+    {
+      "epoch": 2.9337641357027464,
+      "grad_norm": 3.466796875,
+      "learning_rate": 7.886591060707138e-05,
+      "loss": 1.5378,
+      "step": 3178
+    },
+    {
+      "epoch": 2.934687283637203,
+      "grad_norm": 3.52734375,
+      "learning_rate": 7.885923949299533e-05,
+      "loss": 1.4426,
+      "step": 3179
+    },
+    {
+      "epoch": 2.9356104315716594,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.885256837891928e-05,
+      "loss": 1.3804,
+      "step": 3180
+    },
+    {
+      "epoch": 2.936533579506116,
+      "grad_norm": 4.04296875,
+      "learning_rate": 7.884589726484324e-05,
+      "loss": 1.4734,
+      "step": 3181
+    },
+    {
+      "epoch": 2.9374567274405723,
+      "grad_norm": 3.37109375,
+      "learning_rate": 7.883922615076719e-05,
+      "loss": 1.1906,
+      "step": 3182
+    },
+    {
+      "epoch": 2.9383798753750288,
+      "grad_norm": 3.88671875,
+      "learning_rate": 7.883255503669113e-05,
+      "loss": 1.5454,
+      "step": 3183
+    },
+    {
+      "epoch": 2.9393030233094852,
+      "grad_norm": 3.697265625,
+      "learning_rate": 7.882588392261508e-05,
+      "loss": 1.3538,
+      "step": 3184
+    },
+    {
+      "epoch": 2.9402261712439417,
+      "grad_norm": 3.56640625,
+      "learning_rate": 7.881921280853904e-05,
+      "loss": 1.6204,
+      "step": 3185
+    },
+    {
+      "epoch": 2.9411493191783986,
+      "grad_norm": 3.958984375,
+      "learning_rate": 7.881254169446298e-05,
+      "loss": 1.4197,
+      "step": 3186
+    },
+    {
+      "epoch": 2.9420724671128546,
+      "grad_norm": 3.7734375,
+      "learning_rate": 7.880587058038693e-05,
+      "loss": 1.4875,
+      "step": 3187
+    },
+    {
+      "epoch": 2.9429956150473116,
+      "grad_norm": 4.5234375,
+      "learning_rate": 7.879919946631088e-05,
+      "loss": 1.6089,
+      "step": 3188
+    },
+    {
+      "epoch": 2.9439187629817676,
+      "grad_norm": 3.982421875,
+      "learning_rate": 7.879252835223483e-05,
+      "loss": 1.4949,
+      "step": 3189
+    },
+    {
+      "epoch": 2.9448419109162245,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.878585723815878e-05,
+      "loss": 1.4954,
+      "step": 3190
+    },
+    {
+      "epoch": 2.945765058850681,
+      "grad_norm": 4.4453125,
+      "learning_rate": 7.877918612408273e-05,
+      "loss": 1.7783,
+      "step": 3191
+    },
+    {
+      "epoch": 2.9466882067851374,
+      "grad_norm": 4.421875,
+      "learning_rate": 7.877251501000667e-05,
+      "loss": 1.2246,
+      "step": 3192
+    },
+    {
+      "epoch": 2.947611354719594,
+      "grad_norm": 4.29296875,
+      "learning_rate": 7.876584389593063e-05,
+      "loss": 1.5298,
+      "step": 3193
+    },
+    {
+      "epoch": 2.9485345026540504,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.875917278185457e-05,
+      "loss": 1.4321,
+      "step": 3194
+    },
+    {
+      "epoch": 2.949457650588507,
+      "grad_norm": 3.8359375,
+      "learning_rate": 7.875250166777852e-05,
+      "loss": 1.6208,
+      "step": 3195
+    },
+    {
+      "epoch": 2.9503807985229633,
+      "grad_norm": 4.48046875,
+      "learning_rate": 7.874583055370247e-05,
+      "loss": 1.6431,
+      "step": 3196
+    },
+    {
+      "epoch": 2.95130394645742,
+      "grad_norm": 5.33984375,
+      "learning_rate": 7.873915943962642e-05,
+      "loss": 1.3774,
+      "step": 3197
+    },
+    {
+      "epoch": 2.9522270943918762,
+      "grad_norm": 3.916015625,
+      "learning_rate": 7.873248832555037e-05,
+      "loss": 1.3392,
+      "step": 3198
+    },
+    {
+      "epoch": 2.9531502423263327,
+      "grad_norm": 3.79296875,
+      "learning_rate": 7.872581721147432e-05,
+      "loss": 1.2075,
+      "step": 3199
+    },
+    {
+      "epoch": 2.954073390260789,
+      "grad_norm": 4.59765625,
+      "learning_rate": 7.871914609739826e-05,
+      "loss": 1.6165,
+      "step": 3200
+    },
+    {
+      "epoch": 2.9549965381952457,
+      "grad_norm": 3.76171875,
+      "learning_rate": 7.871247498332222e-05,
+      "loss": 1.6392,
+      "step": 3201
+    },
+    {
+      "epoch": 2.955919686129702,
+      "grad_norm": 3.73828125,
+      "learning_rate": 7.870580386924617e-05,
+      "loss": 1.4727,
+      "step": 3202
+    },
+    {
+      "epoch": 2.956842834064159,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.869913275517011e-05,
+      "loss": 1.5881,
+      "step": 3203
+    },
+    {
+      "epoch": 2.957765981998615,
+      "grad_norm": 5.09375,
+      "learning_rate": 7.869246164109406e-05,
+      "loss": 1.4442,
+      "step": 3204
+    },
+    {
+      "epoch": 2.958689129933072,
+      "grad_norm": 4.17578125,
+      "learning_rate": 7.868579052701802e-05,
+      "loss": 1.0582,
+      "step": 3205
+    },
+    {
+      "epoch": 2.959612277867528,
+      "grad_norm": 5.40234375,
+      "learning_rate": 7.867911941294196e-05,
+      "loss": 1.4165,
+      "step": 3206
+    },
+    {
+      "epoch": 2.960535425801985,
+      "grad_norm": 4.890625,
+      "learning_rate": 7.867244829886592e-05,
+      "loss": 1.5154,
+      "step": 3207
+    },
+    {
+      "epoch": 2.9614585737364414,
+      "grad_norm": 3.75,
+      "learning_rate": 7.866577718478987e-05,
+      "loss": 1.6907,
+      "step": 3208
+    },
+    {
+      "epoch": 2.962381721670898,
+      "grad_norm": 3.876953125,
+      "learning_rate": 7.865910607071381e-05,
+      "loss": 1.45,
+      "step": 3209
+    },
+    {
+      "epoch": 2.9633048696053543,
+      "grad_norm": 4.90234375,
+      "learning_rate": 7.865243495663777e-05,
+      "loss": 1.5486,
+      "step": 3210
+    },
+    {
+      "epoch": 2.964228017539811,
+      "grad_norm": 4.69921875,
+      "learning_rate": 7.864576384256171e-05,
+      "loss": 1.6277,
+      "step": 3211
+    },
+    {
+      "epoch": 2.9651511654742673,
+      "grad_norm": 5.37109375,
+      "learning_rate": 7.863909272848566e-05,
+      "loss": 1.7683,
+      "step": 3212
+    },
+    {
+      "epoch": 2.9660743134087237,
+      "grad_norm": 5.01171875,
+      "learning_rate": 7.863242161440962e-05,
+      "loss": 1.2115,
+      "step": 3213
+    },
+    {
+      "epoch": 2.96699746134318,
+      "grad_norm": 4.50390625,
+      "learning_rate": 7.862575050033356e-05,
+      "loss": 1.3572,
+      "step": 3214
+    },
+    {
+      "epoch": 2.9679206092776367,
+      "grad_norm": 3.83984375,
+      "learning_rate": 7.861907938625751e-05,
+      "loss": 1.7534,
+      "step": 3215
+    },
+    {
+      "epoch": 2.968843757212093,
+      "grad_norm": 4.1015625,
+      "learning_rate": 7.861240827218146e-05,
+      "loss": 1.5359,
+      "step": 3216
+    },
+    {
+      "epoch": 2.9697669051465496,
+      "grad_norm": 4.1640625,
+      "learning_rate": 7.860573715810541e-05,
+      "loss": 1.6565,
+      "step": 3217
+    },
+    {
+      "epoch": 2.970690053081006,
+      "grad_norm": 3.8828125,
+      "learning_rate": 7.859906604402936e-05,
+      "loss": 1.5588,
+      "step": 3218
+    },
+    {
+      "epoch": 2.9716132010154626,
+      "grad_norm": 3.78515625,
+      "learning_rate": 7.85923949299533e-05,
+      "loss": 1.3296,
+      "step": 3219
+    },
+    {
+      "epoch": 2.9725363489499195,
+      "grad_norm": 4.25,
+      "learning_rate": 7.858572381587725e-05,
+      "loss": 1.5388,
+      "step": 3220
+    },
+    {
+      "epoch": 2.9734594968843755,
+      "grad_norm": 4.67578125,
+      "learning_rate": 7.857905270180121e-05,
+      "loss": 1.3899,
+      "step": 3221
+    },
+    {
+      "epoch": 2.9743826448188324,
+      "grad_norm": 3.87890625,
+      "learning_rate": 7.857238158772516e-05,
+      "loss": 1.417,
+      "step": 3222
+    },
+    {
+      "epoch": 2.975305792753289,
+      "grad_norm": 3.86328125,
+      "learning_rate": 7.85657104736491e-05,
+      "loss": 1.5063,
+      "step": 3223
+    },
+    {
+      "epoch": 2.9762289406877454,
+      "grad_norm": 4.375,
+      "learning_rate": 7.855903935957305e-05,
+      "loss": 1.4639,
+      "step": 3224
+    },
+    {
+      "epoch": 2.977152088622202,
+      "grad_norm": 4.2109375,
+      "learning_rate": 7.8552368245497e-05,
+      "loss": 1.4539,
+      "step": 3225
+    },
+    {
+      "epoch": 2.9780752365566583,
+      "grad_norm": 3.87890625,
+      "learning_rate": 7.854569713142095e-05,
+      "loss": 1.4871,
+      "step": 3226
+    },
+    {
+      "epoch": 2.9789983844911148,
+      "grad_norm": 3.52734375,
+      "learning_rate": 7.85390260173449e-05,
+      "loss": 1.6479,
+      "step": 3227
+    },
+    {
+      "epoch": 2.9799215324255712,
+      "grad_norm": 3.708984375,
+      "learning_rate": 7.853235490326884e-05,
+      "loss": 1.5166,
+      "step": 3228
+    },
+    {
+      "epoch": 2.9808446803600277,
+      "grad_norm": 3.48046875,
+      "learning_rate": 7.85256837891928e-05,
+      "loss": 1.5291,
+      "step": 3229
+    },
+    {
+      "epoch": 2.981767828294484,
+      "grad_norm": 4.125,
+      "learning_rate": 7.851901267511675e-05,
+      "loss": 1.4678,
+      "step": 3230
+    },
+    {
+      "epoch": 2.9826909762289406,
+      "grad_norm": 4.6953125,
+      "learning_rate": 7.851234156104069e-05,
+      "loss": 1.7097,
+      "step": 3231
+    },
+    {
+      "epoch": 2.983614124163397,
+      "grad_norm": 4.30078125,
+      "learning_rate": 7.850567044696465e-05,
+      "loss": 1.5928,
+      "step": 3232
+    },
+    {
+      "epoch": 2.9845372720978536,
+      "grad_norm": 5.3203125,
+      "learning_rate": 7.84989993328886e-05,
+      "loss": 1.275,
+      "step": 3233
+    },
+    {
+      "epoch": 2.98546042003231,
+      "grad_norm": 4.6796875,
+      "learning_rate": 7.849232821881255e-05,
+      "loss": 1.479,
+      "step": 3234
+    },
+    {
+      "epoch": 2.9863835679667665,
+      "grad_norm": 3.87890625,
+      "learning_rate": 7.84856571047365e-05,
+      "loss": 1.2788,
+      "step": 3235
+    },
+    {
+      "epoch": 2.987306715901223,
+      "grad_norm": 4.80859375,
+      "learning_rate": 7.847898599066045e-05,
+      "loss": 1.4663,
+      "step": 3236
+    },
+    {
+      "epoch": 2.98822986383568,
+      "grad_norm": 5.01171875,
+      "learning_rate": 7.847231487658439e-05,
+      "loss": 1.8135,
+      "step": 3237
+    },
+    {
+      "epoch": 2.989153011770136,
+      "grad_norm": 4.9921875,
+      "learning_rate": 7.846564376250835e-05,
+      "loss": 1.2982,
+      "step": 3238
+    },
+    {
+      "epoch": 2.990076159704593,
+      "grad_norm": 6.12109375,
+      "learning_rate": 7.84589726484323e-05,
+      "loss": 1.446,
+      "step": 3239
+    },
+    {
+      "epoch": 2.9909993076390493,
+      "grad_norm": 4.69140625,
+      "learning_rate": 7.845230153435624e-05,
+      "loss": 1.5078,
+      "step": 3240
+    },
+    {
+      "epoch": 2.991922455573506,
+      "grad_norm": 3.8828125,
+      "learning_rate": 7.844563042028019e-05,
+      "loss": 1.4113,
+      "step": 3241
+    },
+    {
+      "epoch": 2.9928456035079622,
+      "grad_norm": 4.19921875,
+      "learning_rate": 7.843895930620415e-05,
+      "loss": 1.3582,
+      "step": 3242
+    },
+    {
+      "epoch": 2.9937687514424187,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.843228819212809e-05,
+      "loss": 1.5898,
+      "step": 3243
+    },
+    {
+      "epoch": 2.994691899376875,
+      "grad_norm": 3.353515625,
+      "learning_rate": 7.842561707805204e-05,
+      "loss": 1.3921,
+      "step": 3244
+    },
+    {
+      "epoch": 2.9956150473113317,
+      "grad_norm": 4.484375,
+      "learning_rate": 7.841894596397598e-05,
+      "loss": 1.3379,
+      "step": 3245
+    },
+    {
+      "epoch": 2.996538195245788,
+      "grad_norm": 4.4453125,
+      "learning_rate": 7.841227484989994e-05,
+      "loss": 1.5984,
+      "step": 3246
+    },
+    {
+      "epoch": 2.9974613431802446,
+      "grad_norm": 3.40234375,
+      "learning_rate": 7.840560373582389e-05,
+      "loss": 1.1802,
+      "step": 3247
+    },
+    {
+      "epoch": 2.998384491114701,
+      "grad_norm": 4.74609375,
+      "learning_rate": 7.839893262174783e-05,
+      "loss": 1.5679,
+      "step": 3248
+    },
+    {
+      "epoch": 2.9993076390491575,
+      "grad_norm": 3.9609375,
+      "learning_rate": 7.839226150767178e-05,
+      "loss": 1.2305,
+      "step": 3249
+    },
+    {
+      "epoch": 3.000230786983614,
+      "grad_norm": 3.134765625,
+      "learning_rate": 7.838559039359574e-05,
+      "loss": 1.1976,
+      "step": 3250
+    },
+    {
+      "epoch": 3.0011539349180705,
+      "grad_norm": 4.57421875,
+      "learning_rate": 7.837891927951968e-05,
+      "loss": 1.3855,
+      "step": 3251
+    },
+    {
+      "epoch": 3.002077082852527,
+      "grad_norm": 4.26953125,
+      "learning_rate": 7.837224816544363e-05,
+      "loss": 1.0769,
+      "step": 3252
+    },
+    {
+      "epoch": 3.0030002307869834,
+      "grad_norm": 4.390625,
+      "learning_rate": 7.836557705136757e-05,
+      "loss": 1.3547,
+      "step": 3253
+    },
+    {
+      "epoch": 3.0039233787214403,
+      "grad_norm": 4.04296875,
+      "learning_rate": 7.835890593729153e-05,
+      "loss": 1.4004,
+      "step": 3254
+    },
+    {
+      "epoch": 3.004846526655897,
+      "grad_norm": 4.32421875,
+      "learning_rate": 7.835223482321548e-05,
+      "loss": 1.5625,
+      "step": 3255
+    },
+    {
+      "epoch": 3.0057696745903533,
+      "grad_norm": 4.5390625,
+      "learning_rate": 7.834556370913942e-05,
+      "loss": 1.4487,
+      "step": 3256
+    },
+    {
+      "epoch": 3.0066928225248097,
+      "grad_norm": 4.15234375,
+      "learning_rate": 7.833889259506338e-05,
+      "loss": 1.3845,
+      "step": 3257
+    },
+    {
+      "epoch": 3.007615970459266,
+      "grad_norm": 4.52734375,
+      "learning_rate": 7.833222148098734e-05,
+      "loss": 1.606,
+      "step": 3258
+    },
+    {
+      "epoch": 3.0085391183937227,
+      "grad_norm": 5.80078125,
+      "learning_rate": 7.832555036691129e-05,
+      "loss": 1.5532,
+      "step": 3259
+    },
+    {
+      "epoch": 3.009462266328179,
+      "grad_norm": 4.78125,
+      "learning_rate": 7.831887925283523e-05,
+      "loss": 1.4971,
+      "step": 3260
+    },
+    {
+      "epoch": 3.0103854142626356,
+      "grad_norm": 5.546875,
+      "learning_rate": 7.831220813875918e-05,
+      "loss": 1.5359,
+      "step": 3261
+    },
+    {
+      "epoch": 3.011308562197092,
+      "grad_norm": 5.3125,
+      "learning_rate": 7.830553702468314e-05,
+      "loss": 1.4449,
+      "step": 3262
+    },
+    {
+      "epoch": 3.0122317101315486,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.829886591060708e-05,
+      "loss": 1.4551,
+      "step": 3263
+    },
+    {
+      "epoch": 3.013154858066005,
+      "grad_norm": 4.69140625,
+      "learning_rate": 7.829219479653103e-05,
+      "loss": 1.4292,
+      "step": 3264
+    },
+    {
+      "epoch": 3.0140780060004615,
+      "grad_norm": 4.2734375,
+      "learning_rate": 7.828552368245497e-05,
+      "loss": 1.5762,
+      "step": 3265
+    },
+    {
+      "epoch": 3.015001153934918,
+      "grad_norm": 5.5859375,
+      "learning_rate": 7.827885256837893e-05,
+      "loss": 1.5317,
+      "step": 3266
+    },
+    {
+      "epoch": 3.0159243018693744,
+      "grad_norm": 5.10546875,
+      "learning_rate": 7.827218145430288e-05,
+      "loss": 1.6616,
+      "step": 3267
+    },
+    {
+      "epoch": 3.016847449803831,
+      "grad_norm": 6.2109375,
+      "learning_rate": 7.826551034022682e-05,
+      "loss": 1.6277,
+      "step": 3268
+    },
+    {
+      "epoch": 3.0177705977382874,
+      "grad_norm": 4.2578125,
+      "learning_rate": 7.825883922615077e-05,
+      "loss": 1.4518,
+      "step": 3269
+    },
+    {
+      "epoch": 3.018693745672744,
+      "grad_norm": 4.8828125,
+      "learning_rate": 7.825216811207473e-05,
+      "loss": 1.4839,
+      "step": 3270
+    },
+    {
+      "epoch": 3.0196168936072008,
+      "grad_norm": 4.69921875,
+      "learning_rate": 7.824549699799867e-05,
+      "loss": 1.5698,
+      "step": 3271
+    },
+    {
+      "epoch": 3.0205400415416572,
+      "grad_norm": 4.58203125,
+      "learning_rate": 7.823882588392262e-05,
+      "loss": 1.748,
+      "step": 3272
+    },
+    {
+      "epoch": 3.0214631894761137,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.823215476984656e-05,
+      "loss": 1.6248,
+      "step": 3273
+    },
+    {
+      "epoch": 3.02238633741057,
+      "grad_norm": 4.5859375,
+      "learning_rate": 7.822548365577052e-05,
+      "loss": 1.3621,
+      "step": 3274
+    },
+    {
+      "epoch": 3.0233094853450266,
+      "grad_norm": 5.36328125,
+      "learning_rate": 7.821881254169447e-05,
+      "loss": 1.7097,
+      "step": 3275
+    },
+    {
+      "epoch": 3.024232633279483,
+      "grad_norm": 5.24609375,
+      "learning_rate": 7.821214142761841e-05,
+      "loss": 1.2319,
+      "step": 3276
+    },
+    {
+      "epoch": 3.0251557812139396,
+      "grad_norm": 4.14453125,
+      "learning_rate": 7.820547031354236e-05,
+      "loss": 1.3525,
+      "step": 3277
+    },
+    {
+      "epoch": 3.026078929148396,
+      "grad_norm": 4.41015625,
+      "learning_rate": 7.819879919946632e-05,
+      "loss": 1.3308,
+      "step": 3278
+    },
+    {
+      "epoch": 3.0270020770828525,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.819212808539026e-05,
+      "loss": 1.5557,
+      "step": 3279
+    },
+    {
+      "epoch": 3.027925225017309,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.818545697131421e-05,
+      "loss": 1.2147,
+      "step": 3280
+    },
+    {
+      "epoch": 3.0288483729517655,
+      "grad_norm": 4.6171875,
+      "learning_rate": 7.817878585723815e-05,
+      "loss": 1.4978,
+      "step": 3281
+    },
+    {
+      "epoch": 3.029771520886222,
+      "grad_norm": 3.76171875,
+      "learning_rate": 7.817211474316211e-05,
+      "loss": 1.4248,
+      "step": 3282
+    },
+    {
+      "epoch": 3.0306946688206784,
+      "grad_norm": 3.865234375,
+      "learning_rate": 7.816544362908607e-05,
+      "loss": 1.3845,
+      "step": 3283
+    },
+    {
+      "epoch": 3.031617816755135,
+      "grad_norm": 4.515625,
+      "learning_rate": 7.815877251501002e-05,
+      "loss": 1.3286,
+      "step": 3284
+    },
+    {
+      "epoch": 3.0325409646895913,
+      "grad_norm": 4.4453125,
+      "learning_rate": 7.815210140093396e-05,
+      "loss": 0.7297,
+      "step": 3285
+    },
+    {
+      "epoch": 3.033464112624048,
+      "grad_norm": 5.74609375,
+      "learning_rate": 7.814543028685791e-05,
+      "loss": 1.6931,
+      "step": 3286
+    },
+    {
+      "epoch": 3.0343872605585047,
+      "grad_norm": 4.62109375,
+      "learning_rate": 7.813875917278187e-05,
+      "loss": 1.3962,
+      "step": 3287
+    },
+    {
+      "epoch": 3.035310408492961,
+      "grad_norm": 4.30859375,
+      "learning_rate": 7.813208805870581e-05,
+      "loss": 1.3394,
+      "step": 3288
+    },
+    {
+      "epoch": 3.0362335564274177,
+      "grad_norm": 5.0703125,
+      "learning_rate": 7.812541694462976e-05,
+      "loss": 1.5347,
+      "step": 3289
+    },
+    {
+      "epoch": 3.037156704361874,
+      "grad_norm": 5.26953125,
+      "learning_rate": 7.81187458305537e-05,
+      "loss": 1.4995,
+      "step": 3290
+    },
+    {
+      "epoch": 3.0380798522963306,
+      "grad_norm": 4.640625,
+      "learning_rate": 7.811207471647766e-05,
+      "loss": 1.418,
+      "step": 3291
+    },
+    {
+      "epoch": 3.039003000230787,
+      "grad_norm": 3.935546875,
+      "learning_rate": 7.81054036024016e-05,
+      "loss": 1.7817,
+      "step": 3292
+    },
+    {
+      "epoch": 3.0399261481652435,
+      "grad_norm": 4.80859375,
+      "learning_rate": 7.809873248832555e-05,
+      "loss": 1.5571,
+      "step": 3293
+    },
+    {
+      "epoch": 3.0408492960997,
+      "grad_norm": 5.09375,
+      "learning_rate": 7.80920613742495e-05,
+      "loss": 1.3513,
+      "step": 3294
+    },
+    {
+      "epoch": 3.0417724440341565,
+      "grad_norm": 5.50390625,
+      "learning_rate": 7.808539026017346e-05,
+      "loss": 1.1833,
+      "step": 3295
+    },
+    {
+      "epoch": 3.042695591968613,
+      "grad_norm": 4.52734375,
+      "learning_rate": 7.80787191460974e-05,
+      "loss": 1.5427,
+      "step": 3296
+    },
+    {
+      "epoch": 3.0436187399030694,
+      "grad_norm": 3.845703125,
+      "learning_rate": 7.807204803202135e-05,
+      "loss": 1.5073,
+      "step": 3297
+    },
+    {
+      "epoch": 3.044541887837526,
+      "grad_norm": 4.28515625,
+      "learning_rate": 7.806537691794529e-05,
+      "loss": 1.3259,
+      "step": 3298
+    },
+    {
+      "epoch": 3.0454650357719824,
+      "grad_norm": 5.72265625,
+      "learning_rate": 7.805870580386925e-05,
+      "loss": 1.5762,
+      "step": 3299
+    },
+    {
+      "epoch": 3.046388183706439,
+      "grad_norm": 3.943359375,
+      "learning_rate": 7.80520346897932e-05,
+      "loss": 1.4028,
+      "step": 3300
+    },
+    {
+      "epoch": 3.0473113316408953,
+      "grad_norm": 5.19921875,
+      "learning_rate": 7.804536357571714e-05,
+      "loss": 1.3118,
+      "step": 3301
+    },
+    {
+      "epoch": 3.0482344795753518,
+      "grad_norm": 3.783203125,
+      "learning_rate": 7.803869246164109e-05,
+      "loss": 1.1378,
+      "step": 3302
+    },
+    {
+      "epoch": 3.0491576275098082,
+      "grad_norm": 3.7109375,
+      "learning_rate": 7.803202134756505e-05,
+      "loss": 1.5282,
+      "step": 3303
+    },
+    {
+      "epoch": 3.050080775444265,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.802535023348899e-05,
+      "loss": 1.4065,
+      "step": 3304
+    },
+    {
+      "epoch": 3.0510039233787216,
+      "grad_norm": 4.71484375,
+      "learning_rate": 7.801867911941294e-05,
+      "loss": 1.3363,
+      "step": 3305
+    },
+    {
+      "epoch": 3.051927071313178,
+      "grad_norm": 4.43359375,
+      "learning_rate": 7.801200800533688e-05,
+      "loss": 1.6064,
+      "step": 3306
+    },
+    {
+      "epoch": 3.0528502192476346,
+      "grad_norm": 3.841796875,
+      "learning_rate": 7.800533689126086e-05,
+      "loss": 1.3359,
+      "step": 3307
+    },
+    {
+      "epoch": 3.053773367182091,
+      "grad_norm": 3.4296875,
+      "learning_rate": 7.79986657771848e-05,
+      "loss": 1.5024,
+      "step": 3308
+    },
+    {
+      "epoch": 3.0546965151165475,
+      "grad_norm": 3.498046875,
+      "learning_rate": 7.799199466310875e-05,
+      "loss": 1.2792,
+      "step": 3309
+    },
+    {
+      "epoch": 3.055619663051004,
+      "grad_norm": 4.8359375,
+      "learning_rate": 7.798532354903269e-05,
+      "loss": 1.4526,
+      "step": 3310
+    },
+    {
+      "epoch": 3.0565428109854604,
+      "grad_norm": 4.19921875,
+      "learning_rate": 7.797865243495665e-05,
+      "loss": 1.5328,
+      "step": 3311
+    },
+    {
+      "epoch": 3.057465958919917,
+      "grad_norm": 4.58984375,
+      "learning_rate": 7.79719813208806e-05,
+      "loss": 1.5757,
+      "step": 3312
+    },
+    {
+      "epoch": 3.0583891068543734,
+      "grad_norm": 4.671875,
+      "learning_rate": 7.796531020680454e-05,
+      "loss": 1.2783,
+      "step": 3313
+    },
+    {
+      "epoch": 3.05931225478883,
+      "grad_norm": 4.82421875,
+      "learning_rate": 7.795863909272849e-05,
+      "loss": 1.3467,
+      "step": 3314
+    },
+    {
+      "epoch": 3.0602354027232863,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.795196797865245e-05,
+      "loss": 1.563,
+      "step": 3315
+    },
+    {
+      "epoch": 3.061158550657743,
+      "grad_norm": 4.33984375,
+      "learning_rate": 7.794529686457639e-05,
+      "loss": 1.175,
+      "step": 3316
+    },
+    {
+      "epoch": 3.0620816985921993,
+      "grad_norm": 4.67578125,
+      "learning_rate": 7.793862575050034e-05,
+      "loss": 1.4663,
+      "step": 3317
+    },
+    {
+      "epoch": 3.0630048465266557,
+      "grad_norm": 4.0078125,
+      "learning_rate": 7.793195463642428e-05,
+      "loss": 1.2037,
+      "step": 3318
+    },
+    {
+      "epoch": 3.063927994461112,
+      "grad_norm": 5.296875,
+      "learning_rate": 7.792528352234824e-05,
+      "loss": 1.3066,
+      "step": 3319
+    },
+    {
+      "epoch": 3.064851142395569,
+      "grad_norm": 5.3046875,
+      "learning_rate": 7.791861240827219e-05,
+      "loss": 1.3552,
+      "step": 3320
+    },
+    {
+      "epoch": 3.0657742903300256,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.791194129419613e-05,
+      "loss": 1.5608,
+      "step": 3321
+    },
+    {
+      "epoch": 3.066697438264482,
+      "grad_norm": 3.279296875,
+      "learning_rate": 7.790527018012008e-05,
+      "loss": 1.2496,
+      "step": 3322
+    },
+    {
+      "epoch": 3.0676205861989385,
+      "grad_norm": 4.11328125,
+      "learning_rate": 7.789859906604404e-05,
+      "loss": 1.1975,
+      "step": 3323
+    },
+    {
+      "epoch": 3.068543734133395,
+      "grad_norm": 4.33203125,
+      "learning_rate": 7.789192795196798e-05,
+      "loss": 1.3057,
+      "step": 3324
+    },
+    {
+      "epoch": 3.0694668820678515,
+      "grad_norm": 3.76171875,
+      "learning_rate": 7.788525683789193e-05,
+      "loss": 1.4132,
+      "step": 3325
+    },
+    {
+      "epoch": 3.070390030002308,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.787858572381587e-05,
+      "loss": 1.4595,
+      "step": 3326
+    },
+    {
+      "epoch": 3.0713131779367644,
+      "grad_norm": 4.55859375,
+      "learning_rate": 7.787191460973983e-05,
+      "loss": 1.484,
+      "step": 3327
+    },
+    {
+      "epoch": 3.072236325871221,
+      "grad_norm": 3.82421875,
+      "learning_rate": 7.786524349566378e-05,
+      "loss": 1.4657,
+      "step": 3328
+    },
+    {
+      "epoch": 3.0731594738056773,
+      "grad_norm": 4.046875,
+      "learning_rate": 7.785857238158772e-05,
+      "loss": 1.3792,
+      "step": 3329
+    },
+    {
+      "epoch": 3.074082621740134,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.785190126751167e-05,
+      "loss": 1.0896,
+      "step": 3330
+    },
+    {
+      "epoch": 3.0750057696745903,
+      "grad_norm": 3.693359375,
+      "learning_rate": 7.784523015343563e-05,
+      "loss": 1.6128,
+      "step": 3331
+    },
+    {
+      "epoch": 3.0759289176090467,
+      "grad_norm": 3.5390625,
+      "learning_rate": 7.783855903935959e-05,
+      "loss": 1.2339,
+      "step": 3332
+    },
+    {
+      "epoch": 3.076852065543503,
+      "grad_norm": 4.05859375,
+      "learning_rate": 7.783188792528353e-05,
+      "loss": 1.6379,
+      "step": 3333
+    },
+    {
+      "epoch": 3.0777752134779597,
+      "grad_norm": 3.44140625,
+      "learning_rate": 7.782521681120748e-05,
+      "loss": 1.6003,
+      "step": 3334
+    },
+    {
+      "epoch": 3.078698361412416,
+      "grad_norm": 4.24609375,
+      "learning_rate": 7.781854569713142e-05,
+      "loss": 1.3132,
+      "step": 3335
+    },
+    {
+      "epoch": 3.0796215093468726,
+      "grad_norm": 4.8046875,
+      "learning_rate": 7.781187458305538e-05,
+      "loss": 1.4583,
+      "step": 3336
+    },
+    {
+      "epoch": 3.0805446572813295,
+      "grad_norm": 3.705078125,
+      "learning_rate": 7.780520346897933e-05,
+      "loss": 1.3788,
+      "step": 3337
+    },
+    {
+      "epoch": 3.081467805215786,
+      "grad_norm": 4.05859375,
+      "learning_rate": 7.779853235490327e-05,
+      "loss": 1.6194,
+      "step": 3338
+    },
+    {
+      "epoch": 3.0823909531502425,
+      "grad_norm": 4.62890625,
+      "learning_rate": 7.779186124082722e-05,
+      "loss": 1.4854,
+      "step": 3339
+    },
+    {
+      "epoch": 3.083314101084699,
+      "grad_norm": 5.01953125,
+      "learning_rate": 7.778519012675118e-05,
+      "loss": 1.6667,
+      "step": 3340
+    },
+    {
+      "epoch": 3.0842372490191554,
+      "grad_norm": 4.4453125,
+      "learning_rate": 7.777851901267512e-05,
+      "loss": 1.2654,
+      "step": 3341
+    },
+    {
+      "epoch": 3.085160396953612,
+      "grad_norm": 5.1640625,
+      "learning_rate": 7.777184789859907e-05,
+      "loss": 1.5369,
+      "step": 3342
+    },
+    {
+      "epoch": 3.0860835448880684,
+      "grad_norm": 4.734375,
+      "learning_rate": 7.776517678452301e-05,
+      "loss": 1.4685,
+      "step": 3343
+    },
+    {
+      "epoch": 3.087006692822525,
+      "grad_norm": 4.5390625,
+      "learning_rate": 7.775850567044697e-05,
+      "loss": 1.6453,
+      "step": 3344
+    },
+    {
+      "epoch": 3.0879298407569813,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.775183455637092e-05,
+      "loss": 1.4385,
+      "step": 3345
+    },
+    {
+      "epoch": 3.0888529886914378,
+      "grad_norm": 4.62109375,
+      "learning_rate": 7.774516344229486e-05,
+      "loss": 1.436,
+      "step": 3346
+    },
+    {
+      "epoch": 3.0897761366258942,
+      "grad_norm": 5.15625,
+      "learning_rate": 7.773849232821881e-05,
+      "loss": 1.6238,
+      "step": 3347
+    },
+    {
+      "epoch": 3.0906992845603507,
+      "grad_norm": 4.671875,
+      "learning_rate": 7.773182121414277e-05,
+      "loss": 1.3718,
+      "step": 3348
+    },
+    {
+      "epoch": 3.091622432494807,
+      "grad_norm": 4.38671875,
+      "learning_rate": 7.772515010006671e-05,
+      "loss": 1.4055,
+      "step": 3349
+    },
+    {
+      "epoch": 3.0925455804292636,
+      "grad_norm": 3.888671875,
+      "learning_rate": 7.771847898599066e-05,
+      "loss": 1.5078,
+      "step": 3350
+    },
+    {
+      "epoch": 3.09346872836372,
+      "grad_norm": 4.62109375,
+      "learning_rate": 7.77118078719146e-05,
+      "loss": 1.3257,
+      "step": 3351
+    },
+    {
+      "epoch": 3.0943918762981766,
+      "grad_norm": 5.078125,
+      "learning_rate": 7.770513675783856e-05,
+      "loss": 1.5549,
+      "step": 3352
+    },
+    {
+      "epoch": 3.0953150242326335,
+      "grad_norm": 4.94140625,
+      "learning_rate": 7.769846564376251e-05,
+      "loss": 1.149,
+      "step": 3353
+    },
+    {
+      "epoch": 3.09623817216709,
+      "grad_norm": 4.75390625,
+      "learning_rate": 7.769179452968645e-05,
+      "loss": 1.5251,
+      "step": 3354
+    },
+    {
+      "epoch": 3.0971613201015464,
+      "grad_norm": 3.912109375,
+      "learning_rate": 7.76851234156104e-05,
+      "loss": 1.5835,
+      "step": 3355
+    },
+    {
+      "epoch": 3.098084468036003,
+      "grad_norm": 4.3828125,
+      "learning_rate": 7.767845230153436e-05,
+      "loss": 1.3313,
+      "step": 3356
+    },
+    {
+      "epoch": 3.0990076159704594,
+      "grad_norm": 4.07421875,
+      "learning_rate": 7.767178118745832e-05,
+      "loss": 1.1337,
+      "step": 3357
+    },
+    {
+      "epoch": 3.099930763904916,
+      "grad_norm": 4.51953125,
+      "learning_rate": 7.766511007338226e-05,
+      "loss": 1.4321,
+      "step": 3358
+    },
+    {
+      "epoch": 3.1008539118393723,
+      "grad_norm": 3.994140625,
+      "learning_rate": 7.765843895930621e-05,
+      "loss": 1.6741,
+      "step": 3359
+    },
+    {
+      "epoch": 3.101777059773829,
+      "grad_norm": 4.27734375,
+      "learning_rate": 7.765176784523017e-05,
+      "loss": 1.5623,
+      "step": 3360
+    },
+    {
+      "epoch": 3.1027002077082853,
+      "grad_norm": 4.48828125,
+      "learning_rate": 7.764509673115411e-05,
+      "loss": 1.7073,
+      "step": 3361
+    },
+    {
+      "epoch": 3.1036233556427417,
+      "grad_norm": 3.71484375,
+      "learning_rate": 7.763842561707806e-05,
+      "loss": 1.329,
+      "step": 3362
+    },
+    {
+      "epoch": 3.104546503577198,
+      "grad_norm": 3.83203125,
+      "learning_rate": 7.7631754503002e-05,
+      "loss": 1.4263,
+      "step": 3363
+    },
+    {
+      "epoch": 3.1054696515116547,
+      "grad_norm": 4.2578125,
+      "learning_rate": 7.762508338892596e-05,
+      "loss": 1.5039,
+      "step": 3364
+    },
+    {
+      "epoch": 3.106392799446111,
+      "grad_norm": 3.67578125,
+      "learning_rate": 7.761841227484991e-05,
+      "loss": 1.3977,
+      "step": 3365
+    },
+    {
+      "epoch": 3.1073159473805676,
+      "grad_norm": 4.11328125,
+      "learning_rate": 7.761174116077385e-05,
+      "loss": 1.4182,
+      "step": 3366
+    },
+    {
+      "epoch": 3.108239095315024,
+      "grad_norm": 4.48828125,
+      "learning_rate": 7.76050700466978e-05,
+      "loss": 1.3202,
+      "step": 3367
+    },
+    {
+      "epoch": 3.1091622432494805,
+      "grad_norm": 5.3515625,
+      "learning_rate": 7.759839893262176e-05,
+      "loss": 1.2766,
+      "step": 3368
+    },
+    {
+      "epoch": 3.1100853911839375,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.75917278185457e-05,
+      "loss": 1.1956,
+      "step": 3369
+    },
+    {
+      "epoch": 3.111008539118394,
+      "grad_norm": 3.72265625,
+      "learning_rate": 7.758505670446965e-05,
+      "loss": 1.248,
+      "step": 3370
+    },
+    {
+      "epoch": 3.1119316870528504,
+      "grad_norm": 4.4140625,
+      "learning_rate": 7.75783855903936e-05,
+      "loss": 1.4763,
+      "step": 3371
+    },
+    {
+      "epoch": 3.112854834987307,
+      "grad_norm": 5.41015625,
+      "learning_rate": 7.757171447631755e-05,
+      "loss": 1.2874,
+      "step": 3372
+    },
+    {
+      "epoch": 3.1137779829217633,
+      "grad_norm": 4.24609375,
+      "learning_rate": 7.75650433622415e-05,
+      "loss": 1.3281,
+      "step": 3373
+    },
+    {
+      "epoch": 3.11470113085622,
+      "grad_norm": 4.28125,
+      "learning_rate": 7.755837224816544e-05,
+      "loss": 1.4956,
+      "step": 3374
+    },
+    {
+      "epoch": 3.1156242787906763,
+      "grad_norm": 4.96484375,
+      "learning_rate": 7.755170113408939e-05,
+      "loss": 1.5295,
+      "step": 3375
+    },
+    {
+      "epoch": 3.1165474267251327,
+      "grad_norm": 5.19140625,
+      "learning_rate": 7.754503002001335e-05,
+      "loss": 1.4277,
+      "step": 3376
+    },
+    {
+      "epoch": 3.117470574659589,
+      "grad_norm": 4.38671875,
+      "learning_rate": 7.75383589059373e-05,
+      "loss": 1.5854,
+      "step": 3377
+    },
+    {
+      "epoch": 3.1183937225940457,
+      "grad_norm": 3.779296875,
+      "learning_rate": 7.753168779186124e-05,
+      "loss": 1.3582,
+      "step": 3378
+    },
+    {
+      "epoch": 3.119316870528502,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.752501667778518e-05,
+      "loss": 1.5913,
+      "step": 3379
+    },
+    {
+      "epoch": 3.1202400184629586,
+      "grad_norm": 5.19921875,
+      "learning_rate": 7.751834556370914e-05,
+      "loss": 1.5178,
+      "step": 3380
+    },
+    {
+      "epoch": 3.121163166397415,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.751167444963309e-05,
+      "loss": 1.4265,
+      "step": 3381
+    },
+    {
+      "epoch": 3.1220863143318716,
+      "grad_norm": 4.62109375,
+      "learning_rate": 7.750500333555705e-05,
+      "loss": 1.6289,
+      "step": 3382
+    },
+    {
+      "epoch": 3.123009462266328,
+      "grad_norm": 4.07421875,
+      "learning_rate": 7.749833222148099e-05,
+      "loss": 1.4905,
+      "step": 3383
+    },
+    {
+      "epoch": 3.1239326102007845,
+      "grad_norm": 3.814453125,
+      "learning_rate": 7.749166110740494e-05,
+      "loss": 1.6379,
+      "step": 3384
+    },
+    {
+      "epoch": 3.124855758135241,
+      "grad_norm": 3.662109375,
+      "learning_rate": 7.74849899933289e-05,
+      "loss": 1.2957,
+      "step": 3385
+    },
+    {
+      "epoch": 3.125778906069698,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.747831887925284e-05,
+      "loss": 1.4485,
+      "step": 3386
+    },
+    {
+      "epoch": 3.1267020540041544,
+      "grad_norm": 4.203125,
+      "learning_rate": 7.747164776517679e-05,
+      "loss": 1.4146,
+      "step": 3387
+    },
+    {
+      "epoch": 3.127625201938611,
+      "grad_norm": 4.484375,
+      "learning_rate": 7.746497665110073e-05,
+      "loss": 1.4585,
+      "step": 3388
+    },
+    {
+      "epoch": 3.1285483498730673,
+      "grad_norm": 4.00390625,
+      "learning_rate": 7.745830553702469e-05,
+      "loss": 1.5745,
+      "step": 3389
+    },
+    {
+      "epoch": 3.1294714978075238,
+      "grad_norm": 4.9140625,
+      "learning_rate": 7.745163442294864e-05,
+      "loss": 1.6094,
+      "step": 3390
+    },
+    {
+      "epoch": 3.1303946457419802,
+      "grad_norm": 4.68359375,
+      "learning_rate": 7.744496330887258e-05,
+      "loss": 1.5684,
+      "step": 3391
+    },
+    {
+      "epoch": 3.1313177936764367,
+      "grad_norm": 3.912109375,
+      "learning_rate": 7.743829219479653e-05,
+      "loss": 1.1422,
+      "step": 3392
+    },
+    {
+      "epoch": 3.132240941610893,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.743162108072049e-05,
+      "loss": 1.3989,
+      "step": 3393
+    },
+    {
+      "epoch": 3.1331640895453496,
+      "grad_norm": 5.390625,
+      "learning_rate": 7.742494996664443e-05,
+      "loss": 1.4929,
+      "step": 3394
+    },
+    {
+      "epoch": 3.134087237479806,
+      "grad_norm": 5.328125,
+      "learning_rate": 7.741827885256838e-05,
+      "loss": 1.54,
+      "step": 3395
+    },
+    {
+      "epoch": 3.1350103854142626,
+      "grad_norm": 5.28515625,
+      "learning_rate": 7.741160773849232e-05,
+      "loss": 1.4421,
+      "step": 3396
+    },
+    {
+      "epoch": 3.135933533348719,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.740493662441628e-05,
+      "loss": 1.1073,
+      "step": 3397
+    },
+    {
+      "epoch": 3.1368566812831755,
+      "grad_norm": 3.994140625,
+      "learning_rate": 7.739826551034023e-05,
+      "loss": 1.5354,
+      "step": 3398
+    },
+    {
+      "epoch": 3.137779829217632,
+      "grad_norm": 5.0078125,
+      "learning_rate": 7.739159439626417e-05,
+      "loss": 1.3658,
+      "step": 3399
+    },
+    {
+      "epoch": 3.1387029771520885,
+      "grad_norm": 4.92578125,
+      "learning_rate": 7.738492328218813e-05,
+      "loss": 1.6506,
+      "step": 3400
+    },
+    {
+      "epoch": 3.139626125086545,
+      "grad_norm": 4.72265625,
+      "learning_rate": 7.737825216811208e-05,
+      "loss": 1.1625,
+      "step": 3401
+    },
+    {
+      "epoch": 3.1405492730210014,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.737158105403602e-05,
+      "loss": 1.4347,
+      "step": 3402
+    },
+    {
+      "epoch": 3.1414724209554583,
+      "grad_norm": 3.5859375,
+      "learning_rate": 7.736490993995997e-05,
+      "loss": 1.278,
+      "step": 3403
+    },
+    {
+      "epoch": 3.142395568889915,
+      "grad_norm": 4.6171875,
+      "learning_rate": 7.735823882588393e-05,
+      "loss": 1.6462,
+      "step": 3404
+    },
+    {
+      "epoch": 3.1433187168243713,
+      "grad_norm": 4.0078125,
+      "learning_rate": 7.735156771180787e-05,
+      "loss": 1.425,
+      "step": 3405
+    },
+    {
+      "epoch": 3.1442418647588277,
+      "grad_norm": 4.38671875,
+      "learning_rate": 7.734489659773182e-05,
+      "loss": 1.3948,
+      "step": 3406
+    },
+    {
+      "epoch": 3.145165012693284,
+      "grad_norm": 5.01953125,
+      "learning_rate": 7.733822548365578e-05,
+      "loss": 1.3845,
+      "step": 3407
+    },
+    {
+      "epoch": 3.1460881606277407,
+      "grad_norm": 4.37109375,
+      "learning_rate": 7.733155436957972e-05,
+      "loss": 1.5654,
+      "step": 3408
+    },
+    {
+      "epoch": 3.147011308562197,
+      "grad_norm": 3.669921875,
+      "learning_rate": 7.732488325550368e-05,
+      "loss": 1.1489,
+      "step": 3409
+    },
+    {
+      "epoch": 3.1479344564966536,
+      "grad_norm": 3.775390625,
+      "learning_rate": 7.731821214142763e-05,
+      "loss": 1.3346,
+      "step": 3410
+    },
+    {
+      "epoch": 3.14885760443111,
+      "grad_norm": 3.5546875,
+      "learning_rate": 7.731154102735157e-05,
+      "loss": 1.5227,
+      "step": 3411
+    },
+    {
+      "epoch": 3.1497807523655665,
+      "grad_norm": 4.2734375,
+      "learning_rate": 7.730486991327552e-05,
+      "loss": 1.6074,
+      "step": 3412
+    },
+    {
+      "epoch": 3.150703900300023,
+      "grad_norm": 4.84375,
+      "learning_rate": 7.729819879919948e-05,
+      "loss": 1.356,
+      "step": 3413
+    },
+    {
+      "epoch": 3.1516270482344795,
+      "grad_norm": 4.6328125,
+      "learning_rate": 7.729152768512342e-05,
+      "loss": 1.2917,
+      "step": 3414
+    },
+    {
+      "epoch": 3.152550196168936,
+      "grad_norm": 4.58984375,
+      "learning_rate": 7.728485657104737e-05,
+      "loss": 1.5437,
+      "step": 3415
+    },
+    {
+      "epoch": 3.1534733441033924,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.727818545697131e-05,
+      "loss": 1.4807,
+      "step": 3416
+    },
+    {
+      "epoch": 3.154396492037849,
+      "grad_norm": 4.8671875,
+      "learning_rate": 7.727151434289527e-05,
+      "loss": 1.4216,
+      "step": 3417
+    },
+    {
+      "epoch": 3.1553196399723054,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.726484322881922e-05,
+      "loss": 1.1973,
+      "step": 3418
+    },
+    {
+      "epoch": 3.156242787906762,
+      "grad_norm": 4.39453125,
+      "learning_rate": 7.725817211474316e-05,
+      "loss": 1.3091,
+      "step": 3419
+    },
+    {
+      "epoch": 3.1571659358412187,
+      "grad_norm": 3.8359375,
+      "learning_rate": 7.725150100066711e-05,
+      "loss": 1.245,
+      "step": 3420
+    },
+    {
+      "epoch": 3.158089083775675,
+      "grad_norm": 3.85546875,
+      "learning_rate": 7.724482988659107e-05,
+      "loss": 1.3859,
+      "step": 3421
+    },
+    {
+      "epoch": 3.1590122317101317,
+      "grad_norm": 4.25390625,
+      "learning_rate": 7.723815877251501e-05,
+      "loss": 1.2034,
+      "step": 3422
+    },
+    {
+      "epoch": 3.159935379644588,
+      "grad_norm": 4.6328125,
+      "learning_rate": 7.723148765843896e-05,
+      "loss": 1.7993,
+      "step": 3423
+    },
+    {
+      "epoch": 3.1608585275790446,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.72248165443629e-05,
+      "loss": 1.3171,
+      "step": 3424
+    },
+    {
+      "epoch": 3.161781675513501,
+      "grad_norm": 3.716796875,
+      "learning_rate": 7.721814543028686e-05,
+      "loss": 1.3774,
+      "step": 3425
+    },
+    {
+      "epoch": 3.1627048234479576,
+      "grad_norm": 3.75,
+      "learning_rate": 7.721147431621081e-05,
+      "loss": 1.1776,
+      "step": 3426
+    },
+    {
+      "epoch": 3.163627971382414,
+      "grad_norm": 4.4375,
+      "learning_rate": 7.720480320213475e-05,
+      "loss": 1.6357,
+      "step": 3427
+    },
+    {
+      "epoch": 3.1645511193168705,
+      "grad_norm": 3.865234375,
+      "learning_rate": 7.71981320880587e-05,
+      "loss": 1.4739,
+      "step": 3428
+    },
+    {
+      "epoch": 3.165474267251327,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.719146097398266e-05,
+      "loss": 1.5203,
+      "step": 3429
+    },
+    {
+      "epoch": 3.1663974151857834,
+      "grad_norm": 4.78515625,
+      "learning_rate": 7.71847898599066e-05,
+      "loss": 1.4846,
+      "step": 3430
+    },
+    {
+      "epoch": 3.16732056312024,
+      "grad_norm": 5.5078125,
+      "learning_rate": 7.717811874583055e-05,
+      "loss": 1.3081,
+      "step": 3431
+    },
+    {
+      "epoch": 3.1682437110546964,
+      "grad_norm": 4.24609375,
+      "learning_rate": 7.717144763175451e-05,
+      "loss": 1.3428,
+      "step": 3432
+    },
+    {
+      "epoch": 3.169166858989153,
+      "grad_norm": 3.4765625,
+      "learning_rate": 7.716477651767845e-05,
+      "loss": 1.5085,
+      "step": 3433
+    },
+    {
+      "epoch": 3.1700900069236093,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.715810540360241e-05,
+      "loss": 1.6528,
+      "step": 3434
+    },
+    {
+      "epoch": 3.1710131548580662,
+      "grad_norm": 4.1484375,
+      "learning_rate": 7.715143428952636e-05,
+      "loss": 1.575,
+      "step": 3435
+    },
+    {
+      "epoch": 3.1719363027925227,
+      "grad_norm": 3.875,
+      "learning_rate": 7.71447631754503e-05,
+      "loss": 1.4263,
+      "step": 3436
+    },
+    {
+      "epoch": 3.172859450726979,
+      "grad_norm": 4.640625,
+      "learning_rate": 7.713809206137425e-05,
+      "loss": 1.1973,
+      "step": 3437
+    },
+    {
+      "epoch": 3.1737825986614356,
+      "grad_norm": 5.58203125,
+      "learning_rate": 7.713142094729821e-05,
+      "loss": 1.8657,
+      "step": 3438
+    },
+    {
+      "epoch": 3.174705746595892,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.712474983322215e-05,
+      "loss": 1.363,
+      "step": 3439
+    },
+    {
+      "epoch": 3.1756288945303486,
+      "grad_norm": 3.794921875,
+      "learning_rate": 7.71180787191461e-05,
+      "loss": 1.2808,
+      "step": 3440
+    },
+    {
+      "epoch": 3.176552042464805,
+      "grad_norm": 3.435546875,
+      "learning_rate": 7.711140760507006e-05,
+      "loss": 1.504,
+      "step": 3441
+    },
+    {
+      "epoch": 3.1774751903992615,
+      "grad_norm": 3.80859375,
+      "learning_rate": 7.7104736490994e-05,
+      "loss": 1.46,
+      "step": 3442
+    },
+    {
+      "epoch": 3.178398338333718,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.709806537691795e-05,
+      "loss": 1.6702,
+      "step": 3443
+    },
+    {
+      "epoch": 3.1793214862681745,
+      "grad_norm": 4.18359375,
+      "learning_rate": 7.70913942628419e-05,
+      "loss": 1.4915,
+      "step": 3444
+    },
+    {
+      "epoch": 3.180244634202631,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.708472314876585e-05,
+      "loss": 1.2839,
+      "step": 3445
+    },
+    {
+      "epoch": 3.1811677821370874,
+      "grad_norm": 4.05859375,
+      "learning_rate": 7.70780520346898e-05,
+      "loss": 1.3828,
+      "step": 3446
+    },
+    {
+      "epoch": 3.182090930071544,
+      "grad_norm": 3.6796875,
+      "learning_rate": 7.707138092061374e-05,
+      "loss": 1.3333,
+      "step": 3447
+    },
+    {
+      "epoch": 3.1830140780060003,
+      "grad_norm": 5.2734375,
+      "learning_rate": 7.706470980653769e-05,
+      "loss": 1.4558,
+      "step": 3448
+    },
+    {
+      "epoch": 3.183937225940457,
+      "grad_norm": 5.078125,
+      "learning_rate": 7.705803869246165e-05,
+      "loss": 1.3225,
+      "step": 3449
+    },
+    {
+      "epoch": 3.1848603738749133,
+      "grad_norm": 4.6484375,
+      "learning_rate": 7.70513675783856e-05,
+      "loss": 1.7078,
+      "step": 3450
+    },
+    {
+      "epoch": 3.1857835218093697,
+      "grad_norm": 3.91796875,
+      "learning_rate": 7.704469646430954e-05,
+      "loss": 1.5125,
+      "step": 3451
+    },
+    {
+      "epoch": 3.1867066697438267,
+      "grad_norm": 3.73828125,
+      "learning_rate": 7.703802535023349e-05,
+      "loss": 1.1694,
+      "step": 3452
+    },
+    {
+      "epoch": 3.187629817678283,
+      "grad_norm": 3.78515625,
+      "learning_rate": 7.703135423615744e-05,
+      "loss": 1.6013,
+      "step": 3453
+    },
+    {
+      "epoch": 3.1885529656127396,
+      "grad_norm": 3.873046875,
+      "learning_rate": 7.702468312208139e-05,
+      "loss": 1.2723,
+      "step": 3454
+    },
+    {
+      "epoch": 3.189476113547196,
+      "grad_norm": 3.888671875,
+      "learning_rate": 7.701801200800534e-05,
+      "loss": 1.4036,
+      "step": 3455
+    },
+    {
+      "epoch": 3.1903992614816525,
+      "grad_norm": 5.17578125,
+      "learning_rate": 7.701134089392928e-05,
+      "loss": 1.532,
+      "step": 3456
+    },
+    {
+      "epoch": 3.191322409416109,
+      "grad_norm": 5.07421875,
+      "learning_rate": 7.700466977985324e-05,
+      "loss": 1.4961,
+      "step": 3457
+    },
+    {
+      "epoch": 3.1922455573505655,
+      "grad_norm": 3.6328125,
+      "learning_rate": 7.69979986657772e-05,
+      "loss": 1.5845,
+      "step": 3458
+    },
+    {
+      "epoch": 3.193168705285022,
+      "grad_norm": 4.3359375,
+      "learning_rate": 7.699132755170114e-05,
+      "loss": 1.1884,
+      "step": 3459
+    },
+    {
+      "epoch": 3.1940918532194784,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.698465643762509e-05,
+      "loss": 1.5735,
+      "step": 3460
+    },
+    {
+      "epoch": 3.195015001153935,
+      "grad_norm": 5.3671875,
+      "learning_rate": 7.697798532354903e-05,
+      "loss": 1.6206,
+      "step": 3461
+    },
+    {
+      "epoch": 3.1959381490883914,
+      "grad_norm": 3.3671875,
+      "learning_rate": 7.6971314209473e-05,
+      "loss": 1.3611,
+      "step": 3462
+    },
+    {
+      "epoch": 3.196861297022848,
+      "grad_norm": 4.6015625,
+      "learning_rate": 7.696464309539694e-05,
+      "loss": 1.468,
+      "step": 3463
+    },
+    {
+      "epoch": 3.1977844449573043,
+      "grad_norm": 3.75390625,
+      "learning_rate": 7.695797198132088e-05,
+      "loss": 1.4414,
+      "step": 3464
+    },
+    {
+      "epoch": 3.1987075928917608,
+      "grad_norm": 4.6328125,
+      "learning_rate": 7.695130086724483e-05,
+      "loss": 1.3491,
+      "step": 3465
+    },
+    {
+      "epoch": 3.1996307408262172,
+      "grad_norm": 3.865234375,
+      "learning_rate": 7.694462975316879e-05,
+      "loss": 1.5383,
+      "step": 3466
+    },
+    {
+      "epoch": 3.2005538887606737,
+      "grad_norm": 4.41015625,
+      "learning_rate": 7.693795863909273e-05,
+      "loss": 1.6592,
+      "step": 3467
+    },
+    {
+      "epoch": 3.20147703669513,
+      "grad_norm": 4.12890625,
+      "learning_rate": 7.693128752501668e-05,
+      "loss": 1.4866,
+      "step": 3468
+    },
+    {
+      "epoch": 3.202400184629587,
+      "grad_norm": 4.0078125,
+      "learning_rate": 7.692461641094063e-05,
+      "loss": 1.5334,
+      "step": 3469
+    },
+    {
+      "epoch": 3.2033233325640436,
+      "grad_norm": 4.2734375,
+      "learning_rate": 7.691794529686458e-05,
+      "loss": 1.3699,
+      "step": 3470
+    },
+    {
+      "epoch": 3.2042464804985,
+      "grad_norm": 5.0,
+      "learning_rate": 7.691127418278853e-05,
+      "loss": 1.5291,
+      "step": 3471
+    },
+    {
+      "epoch": 3.2051696284329565,
+      "grad_norm": 4.07421875,
+      "learning_rate": 7.690460306871248e-05,
+      "loss": 1.426,
+      "step": 3472
+    },
+    {
+      "epoch": 3.206092776367413,
+      "grad_norm": 4.21484375,
+      "learning_rate": 7.689793195463642e-05,
+      "loss": 1.5256,
+      "step": 3473
+    },
+    {
+      "epoch": 3.2070159243018694,
+      "grad_norm": 4.0390625,
+      "learning_rate": 7.689126084056038e-05,
+      "loss": 1.3167,
+      "step": 3474
+    },
+    {
+      "epoch": 3.207939072236326,
+      "grad_norm": 4.2265625,
+      "learning_rate": 7.688458972648433e-05,
+      "loss": 1.373,
+      "step": 3475
+    },
+    {
+      "epoch": 3.2088622201707824,
+      "grad_norm": 3.88671875,
+      "learning_rate": 7.687791861240827e-05,
+      "loss": 1.4836,
+      "step": 3476
+    },
+    {
+      "epoch": 3.209785368105239,
+      "grad_norm": 4.77734375,
+      "learning_rate": 7.687124749833222e-05,
+      "loss": 1.2058,
+      "step": 3477
+    },
+    {
+      "epoch": 3.2107085160396953,
+      "grad_norm": 3.86328125,
+      "learning_rate": 7.686457638425617e-05,
+      "loss": 1.3042,
+      "step": 3478
+    },
+    {
+      "epoch": 3.211631663974152,
+      "grad_norm": 4.1015625,
+      "learning_rate": 7.685790527018012e-05,
+      "loss": 1.4314,
+      "step": 3479
+    },
+    {
+      "epoch": 3.2125548119086083,
+      "grad_norm": 3.513671875,
+      "learning_rate": 7.685123415610407e-05,
+      "loss": 0.9984,
+      "step": 3480
+    },
+    {
+      "epoch": 3.2134779598430647,
+      "grad_norm": 5.51953125,
+      "learning_rate": 7.684456304202801e-05,
+      "loss": 1.4388,
+      "step": 3481
+    },
+    {
+      "epoch": 3.214401107777521,
+      "grad_norm": 5.6953125,
+      "learning_rate": 7.683789192795198e-05,
+      "loss": 1.6523,
+      "step": 3482
+    },
+    {
+      "epoch": 3.2153242557119777,
+      "grad_norm": 3.79296875,
+      "learning_rate": 7.683122081387593e-05,
+      "loss": 1.5103,
+      "step": 3483
+    },
+    {
+      "epoch": 3.216247403646434,
+      "grad_norm": 3.908203125,
+      "learning_rate": 7.682454969979987e-05,
+      "loss": 1.3538,
+      "step": 3484
+    },
+    {
+      "epoch": 3.2171705515808906,
+      "grad_norm": 4.63671875,
+      "learning_rate": 7.681787858572382e-05,
+      "loss": 1.4846,
+      "step": 3485
+    },
+    {
+      "epoch": 3.2180936995153475,
+      "grad_norm": 4.91796875,
+      "learning_rate": 7.681120747164778e-05,
+      "loss": 1.3402,
+      "step": 3486
+    },
+    {
+      "epoch": 3.219016847449804,
+      "grad_norm": 4.93359375,
+      "learning_rate": 7.680453635757172e-05,
+      "loss": 1.3379,
+      "step": 3487
+    },
+    {
+      "epoch": 3.2199399953842605,
+      "grad_norm": 3.837890625,
+      "learning_rate": 7.679786524349567e-05,
+      "loss": 1.278,
+      "step": 3488
+    },
+    {
+      "epoch": 3.220863143318717,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.679119412941962e-05,
+      "loss": 1.6331,
+      "step": 3489
+    },
+    {
+      "epoch": 3.2217862912531734,
+      "grad_norm": 5.02734375,
+      "learning_rate": 7.678452301534357e-05,
+      "loss": 1.201,
+      "step": 3490
+    },
+    {
+      "epoch": 3.22270943918763,
+      "grad_norm": 4.33203125,
+      "learning_rate": 7.677785190126752e-05,
+      "loss": 1.3464,
+      "step": 3491
+    },
+    {
+      "epoch": 3.2236325871220863,
+      "grad_norm": 4.2890625,
+      "learning_rate": 7.677118078719147e-05,
+      "loss": 1.575,
+      "step": 3492
+    },
+    {
+      "epoch": 3.224555735056543,
+      "grad_norm": 4.1484375,
+      "learning_rate": 7.676450967311541e-05,
+      "loss": 1.5312,
+      "step": 3493
+    },
+    {
+      "epoch": 3.2254788829909993,
+      "grad_norm": 5.11328125,
+      "learning_rate": 7.675783855903937e-05,
+      "loss": 1.2353,
+      "step": 3494
+    },
+    {
+      "epoch": 3.2264020309254557,
+      "grad_norm": 3.779296875,
+      "learning_rate": 7.675116744496332e-05,
+      "loss": 1.5603,
+      "step": 3495
+    },
+    {
+      "epoch": 3.227325178859912,
+      "grad_norm": 3.076171875,
+      "learning_rate": 7.674449633088726e-05,
+      "loss": 1.1334,
+      "step": 3496
+    },
+    {
+      "epoch": 3.2282483267943687,
+      "grad_norm": 4.83984375,
+      "learning_rate": 7.67378252168112e-05,
+      "loss": 1.4817,
+      "step": 3497
+    },
+    {
+      "epoch": 3.229171474728825,
+      "grad_norm": 4.56640625,
+      "learning_rate": 7.673115410273516e-05,
+      "loss": 1.5627,
+      "step": 3498
+    },
+    {
+      "epoch": 3.2300946226632816,
+      "grad_norm": 5.59765625,
+      "learning_rate": 7.672448298865911e-05,
+      "loss": 1.7263,
+      "step": 3499
+    },
+    {
+      "epoch": 3.231017770597738,
+      "grad_norm": 4.47265625,
+      "learning_rate": 7.671781187458306e-05,
+      "loss": 1.1758,
+      "step": 3500
+    },
+    {
+      "epoch": 3.231940918532195,
+      "grad_norm": 5.29296875,
+      "learning_rate": 7.6711140760507e-05,
+      "loss": 1.425,
+      "step": 3501
+    },
+    {
+      "epoch": 3.2328640664666515,
+      "grad_norm": 5.8984375,
+      "learning_rate": 7.670446964643096e-05,
+      "loss": 1.4834,
+      "step": 3502
+    },
+    {
+      "epoch": 3.233787214401108,
+      "grad_norm": 4.25390625,
+      "learning_rate": 7.66977985323549e-05,
+      "loss": 1.4695,
+      "step": 3503
+    },
+    {
+      "epoch": 3.2347103623355644,
+      "grad_norm": 5.21875,
+      "learning_rate": 7.669112741827885e-05,
+      "loss": 1.2659,
+      "step": 3504
+    },
+    {
+      "epoch": 3.235633510270021,
+      "grad_norm": 4.2265625,
+      "learning_rate": 7.66844563042028e-05,
+      "loss": 1.511,
+      "step": 3505
+    },
+    {
+      "epoch": 3.2365566582044774,
+      "grad_norm": 4.07421875,
+      "learning_rate": 7.667778519012676e-05,
+      "loss": 1.5295,
+      "step": 3506
+    },
+    {
+      "epoch": 3.237479806138934,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.667111407605071e-05,
+      "loss": 1.2488,
+      "step": 3507
+    },
+    {
+      "epoch": 3.2384029540733903,
+      "grad_norm": 4.015625,
+      "learning_rate": 7.666444296197466e-05,
+      "loss": 1.4858,
+      "step": 3508
+    },
+    {
+      "epoch": 3.2393261020078468,
+      "grad_norm": 3.78515625,
+      "learning_rate": 7.66577718478986e-05,
+      "loss": 1.3506,
+      "step": 3509
+    },
+    {
+      "epoch": 3.2402492499423032,
+      "grad_norm": 4.328125,
+      "learning_rate": 7.665110073382255e-05,
+      "loss": 1.4492,
+      "step": 3510
+    },
+    {
+      "epoch": 3.2411723978767597,
+      "grad_norm": 3.892578125,
+      "learning_rate": 7.664442961974651e-05,
+      "loss": 1.5815,
+      "step": 3511
+    },
+    {
+      "epoch": 3.242095545811216,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.663775850567046e-05,
+      "loss": 1.5066,
+      "step": 3512
+    },
+    {
+      "epoch": 3.2430186937456726,
+      "grad_norm": 3.857421875,
+      "learning_rate": 7.66310873915944e-05,
+      "loss": 1.626,
+      "step": 3513
+    },
+    {
+      "epoch": 3.243941841680129,
+      "grad_norm": 4.046875,
+      "learning_rate": 7.662441627751835e-05,
+      "loss": 1.6987,
+      "step": 3514
+    },
+    {
+      "epoch": 3.2448649896145856,
+      "grad_norm": 4.48046875,
+      "learning_rate": 7.66177451634423e-05,
+      "loss": 1.4424,
+      "step": 3515
+    },
+    {
+      "epoch": 3.245788137549042,
+      "grad_norm": 3.337890625,
+      "learning_rate": 7.661107404936625e-05,
+      "loss": 1.4583,
+      "step": 3516
+    },
+    {
+      "epoch": 3.2467112854834985,
+      "grad_norm": 4.125,
+      "learning_rate": 7.66044029352902e-05,
+      "loss": 1.5046,
+      "step": 3517
+    },
+    {
+      "epoch": 3.2476344334179554,
+      "grad_norm": 5.609375,
+      "learning_rate": 7.659773182121414e-05,
+      "loss": 1.6455,
+      "step": 3518
+    },
+    {
+      "epoch": 3.248557581352412,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.65910607071381e-05,
+      "loss": 1.3268,
+      "step": 3519
+    },
+    {
+      "epoch": 3.2494807292868684,
+      "grad_norm": 4.30859375,
+      "learning_rate": 7.658438959306205e-05,
+      "loss": 1.4626,
+      "step": 3520
+    },
+    {
+      "epoch": 3.250403877221325,
+      "grad_norm": 3.607421875,
+      "learning_rate": 7.657771847898599e-05,
+      "loss": 0.9515,
+      "step": 3521
+    },
+    {
+      "epoch": 3.2513270251557813,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.657104736490994e-05,
+      "loss": 1.2744,
+      "step": 3522
+    },
+    {
+      "epoch": 3.252250173090238,
+      "grad_norm": 3.970703125,
+      "learning_rate": 7.65643762508339e-05,
+      "loss": 1.4331,
+      "step": 3523
+    },
+    {
+      "epoch": 3.2531733210246943,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.655770513675784e-05,
+      "loss": 1.45,
+      "step": 3524
+    },
+    {
+      "epoch": 3.2540964689591507,
+      "grad_norm": 4.7734375,
+      "learning_rate": 7.655103402268179e-05,
+      "loss": 1.3605,
+      "step": 3525
+    },
+    {
+      "epoch": 3.255019616893607,
+      "grad_norm": 3.873046875,
+      "learning_rate": 7.654436290860573e-05,
+      "loss": 1.1212,
+      "step": 3526
+    },
+    {
+      "epoch": 3.2559427648280637,
+      "grad_norm": 4.35546875,
+      "learning_rate": 7.653769179452969e-05,
+      "loss": 1.5,
+      "step": 3527
+    },
+    {
+      "epoch": 3.25686591276252,
+      "grad_norm": 4.9921875,
+      "learning_rate": 7.653102068045364e-05,
+      "loss": 1.6133,
+      "step": 3528
+    },
+    {
+      "epoch": 3.2577890606969766,
+      "grad_norm": 3.8203125,
+      "learning_rate": 7.652434956637758e-05,
+      "loss": 1.3198,
+      "step": 3529
+    },
+    {
+      "epoch": 3.258712208631433,
+      "grad_norm": 3.927734375,
+      "learning_rate": 7.651767845230153e-05,
+      "loss": 1.3777,
+      "step": 3530
+    },
+    {
+      "epoch": 3.2596353565658895,
+      "grad_norm": 4.12890625,
+      "learning_rate": 7.651100733822549e-05,
+      "loss": 1.3199,
+      "step": 3531
+    },
+    {
+      "epoch": 3.260558504500346,
+      "grad_norm": 3.69140625,
+      "learning_rate": 7.650433622414945e-05,
+      "loss": 1.2041,
+      "step": 3532
+    },
+    {
+      "epoch": 3.261481652434803,
+      "grad_norm": 3.478515625,
+      "learning_rate": 7.649766511007339e-05,
+      "loss": 1.4016,
+      "step": 3533
+    },
+    {
+      "epoch": 3.262404800369259,
+      "grad_norm": 4.6796875,
+      "learning_rate": 7.649099399599734e-05,
+      "loss": 1.6816,
+      "step": 3534
+    },
+    {
+      "epoch": 3.263327948303716,
+      "grad_norm": 4.2421875,
+      "learning_rate": 7.64843228819213e-05,
+      "loss": 1.3567,
+      "step": 3535
+    },
+    {
+      "epoch": 3.2642510962381723,
+      "grad_norm": 5.33203125,
+      "learning_rate": 7.647765176784524e-05,
+      "loss": 1.5261,
+      "step": 3536
+    },
+    {
+      "epoch": 3.265174244172629,
+      "grad_norm": 4.24609375,
+      "learning_rate": 7.647098065376919e-05,
+      "loss": 1.5007,
+      "step": 3537
+    },
+    {
+      "epoch": 3.2660973921070853,
+      "grad_norm": 4.7265625,
+      "learning_rate": 7.646430953969313e-05,
+      "loss": 1.1987,
+      "step": 3538
+    },
+    {
+      "epoch": 3.2670205400415417,
+      "grad_norm": 3.955078125,
+      "learning_rate": 7.645763842561709e-05,
+      "loss": 1.3652,
+      "step": 3539
+    },
+    {
+      "epoch": 3.267943687975998,
+      "grad_norm": 5.0703125,
+      "learning_rate": 7.645096731154104e-05,
+      "loss": 1.5125,
+      "step": 3540
+    },
+    {
+      "epoch": 3.2688668359104547,
+      "grad_norm": 3.923828125,
+      "learning_rate": 7.644429619746498e-05,
+      "loss": 1.2837,
+      "step": 3541
+    },
+    {
+      "epoch": 3.269789983844911,
+      "grad_norm": 3.8984375,
+      "learning_rate": 7.643762508338893e-05,
+      "loss": 1.4031,
+      "step": 3542
+    },
+    {
+      "epoch": 3.2707131317793676,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.643095396931289e-05,
+      "loss": 1.5588,
+      "step": 3543
+    },
+    {
+      "epoch": 3.271636279713824,
+      "grad_norm": 4.484375,
+      "learning_rate": 7.642428285523683e-05,
+      "loss": 1.3062,
+      "step": 3544
+    },
+    {
+      "epoch": 3.2725594276482806,
+      "grad_norm": 5.30859375,
+      "learning_rate": 7.641761174116078e-05,
+      "loss": 1.5144,
+      "step": 3545
+    },
+    {
+      "epoch": 3.273482575582737,
+      "grad_norm": 4.98828125,
+      "learning_rate": 7.641094062708472e-05,
+      "loss": 1.5249,
+      "step": 3546
+    },
+    {
+      "epoch": 3.2744057235171935,
+      "grad_norm": 4.1640625,
+      "learning_rate": 7.640426951300868e-05,
+      "loss": 1.6738,
+      "step": 3547
+    },
+    {
+      "epoch": 3.27532887145165,
+      "grad_norm": 4.359375,
+      "learning_rate": 7.639759839893263e-05,
+      "loss": 1.6211,
+      "step": 3548
+    },
+    {
+      "epoch": 3.2762520193861064,
+      "grad_norm": 4.05859375,
+      "learning_rate": 7.639092728485657e-05,
+      "loss": 1.3367,
+      "step": 3549
+    },
+    {
+      "epoch": 3.2771751673205634,
+      "grad_norm": 3.96484375,
+      "learning_rate": 7.638425617078052e-05,
+      "loss": 1.304,
+      "step": 3550
+    },
+    {
+      "epoch": 3.2780983152550194,
+      "grad_norm": 3.625,
+      "learning_rate": 7.637758505670448e-05,
+      "loss": 1.5503,
+      "step": 3551
+    },
+    {
+      "epoch": 3.2790214631894763,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.637091394262842e-05,
+      "loss": 1.4248,
+      "step": 3552
+    },
+    {
+      "epoch": 3.2799446111239328,
+      "grad_norm": 3.529296875,
+      "learning_rate": 7.636424282855237e-05,
+      "loss": 1.6436,
+      "step": 3553
+    },
+    {
+      "epoch": 3.2808677590583892,
+      "grad_norm": 3.58203125,
+      "learning_rate": 7.635757171447631e-05,
+      "loss": 1.6482,
+      "step": 3554
+    },
+    {
+      "epoch": 3.2817909069928457,
+      "grad_norm": 4.31640625,
+      "learning_rate": 7.635090060040027e-05,
+      "loss": 1.1746,
+      "step": 3555
+    },
+    {
+      "epoch": 3.282714054927302,
+      "grad_norm": 3.53515625,
+      "learning_rate": 7.634422948632422e-05,
+      "loss": 1.2509,
+      "step": 3556
+    },
+    {
+      "epoch": 3.2836372028617586,
+      "grad_norm": 3.4296875,
+      "learning_rate": 7.633755837224818e-05,
+      "loss": 1.2681,
+      "step": 3557
+    },
+    {
+      "epoch": 3.284560350796215,
+      "grad_norm": 4.05078125,
+      "learning_rate": 7.633088725817212e-05,
+      "loss": 1.5295,
+      "step": 3558
+    },
+    {
+      "epoch": 3.2854834987306716,
+      "grad_norm": 3.998046875,
+      "learning_rate": 7.632421614409607e-05,
+      "loss": 1.5261,
+      "step": 3559
+    },
+    {
+      "epoch": 3.286406646665128,
+      "grad_norm": 3.94140625,
+      "learning_rate": 7.631754503002003e-05,
+      "loss": 1.0209,
+      "step": 3560
+    },
+    {
+      "epoch": 3.2873297945995845,
+      "grad_norm": 4.6328125,
+      "learning_rate": 7.631087391594397e-05,
+      "loss": 1.3218,
+      "step": 3561
+    },
+    {
+      "epoch": 3.288252942534041,
+      "grad_norm": 4.20703125,
+      "learning_rate": 7.630420280186792e-05,
+      "loss": 1.4592,
+      "step": 3562
+    },
+    {
+      "epoch": 3.2891760904684975,
+      "grad_norm": 5.375,
+      "learning_rate": 7.629753168779186e-05,
+      "loss": 1.6743,
+      "step": 3563
+    },
+    {
+      "epoch": 3.290099238402954,
+      "grad_norm": 5.5625,
+      "learning_rate": 7.629086057371582e-05,
+      "loss": 1.4666,
+      "step": 3564
+    },
+    {
+      "epoch": 3.2910223863374104,
+      "grad_norm": 4.59765625,
+      "learning_rate": 7.628418945963977e-05,
+      "loss": 1.1548,
+      "step": 3565
+    },
+    {
+      "epoch": 3.291945534271867,
+      "grad_norm": 4.84765625,
+      "learning_rate": 7.627751834556371e-05,
+      "loss": 1.4241,
+      "step": 3566
+    },
+    {
+      "epoch": 3.292868682206324,
+      "grad_norm": 4.91015625,
+      "learning_rate": 7.627084723148766e-05,
+      "loss": 1.4519,
+      "step": 3567
+    },
+    {
+      "epoch": 3.2937918301407803,
+      "grad_norm": 4.1015625,
+      "learning_rate": 7.626417611741162e-05,
+      "loss": 1.4519,
+      "step": 3568
+    },
+    {
+      "epoch": 3.2947149780752367,
+      "grad_norm": 3.83984375,
+      "learning_rate": 7.625750500333556e-05,
+      "loss": 1.4143,
+      "step": 3569
+    },
+    {
+      "epoch": 3.295638126009693,
+      "grad_norm": 4.01171875,
+      "learning_rate": 7.625083388925951e-05,
+      "loss": 1.3806,
+      "step": 3570
+    },
+    {
+      "epoch": 3.2965612739441497,
+      "grad_norm": 5.58203125,
+      "learning_rate": 7.624416277518345e-05,
+      "loss": 1.4255,
+      "step": 3571
+    },
+    {
+      "epoch": 3.297484421878606,
+      "grad_norm": 3.849609375,
+      "learning_rate": 7.623749166110741e-05,
+      "loss": 1.6714,
+      "step": 3572
+    },
+    {
+      "epoch": 3.2984075698130626,
+      "grad_norm": 4.12890625,
+      "learning_rate": 7.623082054703136e-05,
+      "loss": 1.7786,
+      "step": 3573
+    },
+    {
+      "epoch": 3.299330717747519,
+      "grad_norm": 4.66015625,
+      "learning_rate": 7.62241494329553e-05,
+      "loss": 1.4941,
+      "step": 3574
+    },
+    {
+      "epoch": 3.3002538656819755,
+      "grad_norm": 5.18359375,
+      "learning_rate": 7.621747831887925e-05,
+      "loss": 1.3354,
+      "step": 3575
+    },
+    {
+      "epoch": 3.301177013616432,
+      "grad_norm": 3.478515625,
+      "learning_rate": 7.62108072048032e-05,
+      "loss": 1.2516,
+      "step": 3576
+    },
+    {
+      "epoch": 3.3021001615508885,
+      "grad_norm": 3.7265625,
+      "learning_rate": 7.620413609072715e-05,
+      "loss": 1.2635,
+      "step": 3577
+    },
+    {
+      "epoch": 3.303023309485345,
+      "grad_norm": 4.02734375,
+      "learning_rate": 7.61974649766511e-05,
+      "loss": 1.4939,
+      "step": 3578
+    },
+    {
+      "epoch": 3.3039464574198014,
+      "grad_norm": 4.4921875,
+      "learning_rate": 7.619079386257504e-05,
+      "loss": 1.3704,
+      "step": 3579
+    },
+    {
+      "epoch": 3.304869605354258,
+      "grad_norm": 5.66015625,
+      "learning_rate": 7.6184122748499e-05,
+      "loss": 1.686,
+      "step": 3580
+    },
+    {
+      "epoch": 3.3057927532887144,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.617745163442295e-05,
+      "loss": 1.6047,
+      "step": 3581
+    },
+    {
+      "epoch": 3.306715901223171,
+      "grad_norm": 3.302734375,
+      "learning_rate": 7.617078052034689e-05,
+      "loss": 1.4258,
+      "step": 3582
+    },
+    {
+      "epoch": 3.3076390491576273,
+      "grad_norm": 3.849609375,
+      "learning_rate": 7.616410940627085e-05,
+      "loss": 1.2883,
+      "step": 3583
+    },
+    {
+      "epoch": 3.308562197092084,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.615743829219481e-05,
+      "loss": 1.0192,
+      "step": 3584
+    },
+    {
+      "epoch": 3.3094853450265407,
+      "grad_norm": 3.9765625,
+      "learning_rate": 7.615076717811876e-05,
+      "loss": 1.246,
+      "step": 3585
+    },
+    {
+      "epoch": 3.310408492960997,
+      "grad_norm": 4.03515625,
+      "learning_rate": 7.61440960640427e-05,
+      "loss": 1.7793,
+      "step": 3586
+    },
+    {
+      "epoch": 3.3113316408954536,
+      "grad_norm": 3.93359375,
+      "learning_rate": 7.613742494996665e-05,
+      "loss": 1.4292,
+      "step": 3587
+    },
+    {
+      "epoch": 3.31225478882991,
+      "grad_norm": 4.01171875,
+      "learning_rate": 7.61307538358906e-05,
+      "loss": 1.2882,
+      "step": 3588
+    },
+    {
+      "epoch": 3.3131779367643666,
+      "grad_norm": 4.1328125,
+      "learning_rate": 7.612408272181455e-05,
+      "loss": 1.6152,
+      "step": 3589
+    },
+    {
+      "epoch": 3.314101084698823,
+      "grad_norm": 4.546875,
+      "learning_rate": 7.61174116077385e-05,
+      "loss": 1.269,
+      "step": 3590
+    },
+    {
+      "epoch": 3.3150242326332795,
+      "grad_norm": 4.48828125,
+      "learning_rate": 7.611074049366244e-05,
+      "loss": 1.3301,
+      "step": 3591
+    },
+    {
+      "epoch": 3.315947380567736,
+      "grad_norm": 4.44140625,
+      "learning_rate": 7.61040693795864e-05,
+      "loss": 1.2361,
+      "step": 3592
+    },
+    {
+      "epoch": 3.3168705285021924,
+      "grad_norm": 4.9296875,
+      "learning_rate": 7.609739826551035e-05,
+      "loss": 1.7686,
+      "step": 3593
+    },
+    {
+      "epoch": 3.317793676436649,
+      "grad_norm": 4.02734375,
+      "learning_rate": 7.609072715143429e-05,
+      "loss": 1.2909,
+      "step": 3594
+    },
+    {
+      "epoch": 3.3187168243711054,
+      "grad_norm": 4.51171875,
+      "learning_rate": 7.608405603735824e-05,
+      "loss": 1.4706,
+      "step": 3595
+    },
+    {
+      "epoch": 3.319639972305562,
+      "grad_norm": 3.66796875,
+      "learning_rate": 7.60773849232822e-05,
+      "loss": 0.983,
+      "step": 3596
+    },
+    {
+      "epoch": 3.3205631202400183,
+      "grad_norm": 4.00390625,
+      "learning_rate": 7.607071380920614e-05,
+      "loss": 1.4658,
+      "step": 3597
+    },
+    {
+      "epoch": 3.321486268174475,
+      "grad_norm": 4.55078125,
+      "learning_rate": 7.606404269513009e-05,
+      "loss": 1.8,
+      "step": 3598
+    },
+    {
+      "epoch": 3.3224094161089317,
+      "grad_norm": 5.12890625,
+      "learning_rate": 7.605737158105403e-05,
+      "loss": 1.4805,
+      "step": 3599
+    },
+    {
+      "epoch": 3.3233325640433877,
+      "grad_norm": 4.546875,
+      "learning_rate": 7.605070046697799e-05,
+      "loss": 1.3318,
+      "step": 3600
+    },
+    {
+      "epoch": 3.3242557119778446,
+      "grad_norm": 4.77734375,
+      "learning_rate": 7.604402935290194e-05,
+      "loss": 1.5276,
+      "step": 3601
+    },
+    {
+      "epoch": 3.325178859912301,
+      "grad_norm": 4.92578125,
+      "learning_rate": 7.603735823882588e-05,
+      "loss": 1.6125,
+      "step": 3602
+    },
+    {
+      "epoch": 3.3261020078467576,
+      "grad_norm": 2.8203125,
+      "learning_rate": 7.603068712474983e-05,
+      "loss": 1.3647,
+      "step": 3603
+    },
+    {
+      "epoch": 3.327025155781214,
+      "grad_norm": 3.962890625,
+      "learning_rate": 7.602401601067379e-05,
+      "loss": 1.4297,
+      "step": 3604
+    },
+    {
+      "epoch": 3.3279483037156705,
+      "grad_norm": 3.615234375,
+      "learning_rate": 7.601734489659773e-05,
+      "loss": 1.5181,
+      "step": 3605
+    },
+    {
+      "epoch": 3.328871451650127,
+      "grad_norm": 3.763671875,
+      "learning_rate": 7.601067378252168e-05,
+      "loss": 1.3232,
+      "step": 3606
+    },
+    {
+      "epoch": 3.3297945995845835,
+      "grad_norm": 4.19921875,
+      "learning_rate": 7.600400266844562e-05,
+      "loss": 1.1096,
+      "step": 3607
+    },
+    {
+      "epoch": 3.33071774751904,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.599733155436958e-05,
+      "loss": 1.2205,
+      "step": 3608
+    },
+    {
+      "epoch": 3.3316408954534964,
+      "grad_norm": 4.0859375,
+      "learning_rate": 7.599066044029354e-05,
+      "loss": 1.2429,
+      "step": 3609
+    },
+    {
+      "epoch": 3.332564043387953,
+      "grad_norm": 3.287109375,
+      "learning_rate": 7.598398932621749e-05,
+      "loss": 1.3857,
+      "step": 3610
+    },
+    {
+      "epoch": 3.3334871913224093,
+      "grad_norm": 4.73828125,
+      "learning_rate": 7.597731821214143e-05,
+      "loss": 1.2833,
+      "step": 3611
+    },
+    {
+      "epoch": 3.334410339256866,
+      "grad_norm": 3.875,
+      "learning_rate": 7.597064709806538e-05,
+      "loss": 1.4736,
+      "step": 3612
+    },
+    {
+      "epoch": 3.3353334871913223,
+      "grad_norm": 3.64453125,
+      "learning_rate": 7.596397598398934e-05,
+      "loss": 1.5278,
+      "step": 3613
+    },
+    {
+      "epoch": 3.3362566351257787,
+      "grad_norm": 3.4140625,
+      "learning_rate": 7.595730486991328e-05,
+      "loss": 1.5717,
+      "step": 3614
+    },
+    {
+      "epoch": 3.337179783060235,
+      "grad_norm": 4.67578125,
+      "learning_rate": 7.595063375583723e-05,
+      "loss": 1.6169,
+      "step": 3615
+    },
+    {
+      "epoch": 3.338102930994692,
+      "grad_norm": 3.81640625,
+      "learning_rate": 7.594396264176117e-05,
+      "loss": 1.3662,
+      "step": 3616
+    },
+    {
+      "epoch": 3.339026078929148,
+      "grad_norm": 3.376953125,
+      "learning_rate": 7.593729152768513e-05,
+      "loss": 1.7024,
+      "step": 3617
+    },
+    {
+      "epoch": 3.339949226863605,
+      "grad_norm": 4.1015625,
+      "learning_rate": 7.593062041360908e-05,
+      "loss": 1.562,
+      "step": 3618
+    },
+    {
+      "epoch": 3.3408723747980615,
+      "grad_norm": 3.513671875,
+      "learning_rate": 7.592394929953302e-05,
+      "loss": 1.381,
+      "step": 3619
+    },
+    {
+      "epoch": 3.341795522732518,
+      "grad_norm": 3.958984375,
+      "learning_rate": 7.591727818545697e-05,
+      "loss": 1.5017,
+      "step": 3620
+    },
+    {
+      "epoch": 3.3427186706669745,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.591060707138093e-05,
+      "loss": 1.3379,
+      "step": 3621
+    },
+    {
+      "epoch": 3.343641818601431,
+      "grad_norm": 3.869140625,
+      "learning_rate": 7.590393595730487e-05,
+      "loss": 1.3455,
+      "step": 3622
+    },
+    {
+      "epoch": 3.3445649665358874,
+      "grad_norm": 4.15234375,
+      "learning_rate": 7.589726484322882e-05,
+      "loss": 1.4792,
+      "step": 3623
+    },
+    {
+      "epoch": 3.345488114470344,
+      "grad_norm": 4.75390625,
+      "learning_rate": 7.589059372915276e-05,
+      "loss": 1.593,
+      "step": 3624
+    },
+    {
+      "epoch": 3.3464112624048004,
+      "grad_norm": 4.49609375,
+      "learning_rate": 7.588392261507672e-05,
+      "loss": 1.7397,
+      "step": 3625
+    },
+    {
+      "epoch": 3.347334410339257,
+      "grad_norm": 3.755859375,
+      "learning_rate": 7.587725150100067e-05,
+      "loss": 1.2142,
+      "step": 3626
+    },
+    {
+      "epoch": 3.3482575582737133,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.587058038692461e-05,
+      "loss": 1.3503,
+      "step": 3627
+    },
+    {
+      "epoch": 3.3491807062081698,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.586390927284857e-05,
+      "loss": 1.1407,
+      "step": 3628
+    },
+    {
+      "epoch": 3.3501038541426262,
+      "grad_norm": 3.876953125,
+      "learning_rate": 7.585723815877252e-05,
+      "loss": 1.3252,
+      "step": 3629
+    },
+    {
+      "epoch": 3.3510270020770827,
+      "grad_norm": 4.234375,
+      "learning_rate": 7.585056704469646e-05,
+      "loss": 1.5452,
+      "step": 3630
+    },
+    {
+      "epoch": 3.351950150011539,
+      "grad_norm": 4.0703125,
+      "learning_rate": 7.584389593062041e-05,
+      "loss": 1.5486,
+      "step": 3631
+    },
+    {
+      "epoch": 3.3528732979459956,
+      "grad_norm": 4.04296875,
+      "learning_rate": 7.583722481654437e-05,
+      "loss": 1.2886,
+      "step": 3632
+    },
+    {
+      "epoch": 3.3537964458804526,
+      "grad_norm": 4.33984375,
+      "learning_rate": 7.583055370246833e-05,
+      "loss": 1.2781,
+      "step": 3633
+    },
+    {
+      "epoch": 3.354719593814909,
+      "grad_norm": 4.046875,
+      "learning_rate": 7.582388258839227e-05,
+      "loss": 0.9916,
+      "step": 3634
+    },
+    {
+      "epoch": 3.3556427417493655,
+      "grad_norm": 4.3046875,
+      "learning_rate": 7.581721147431622e-05,
+      "loss": 1.5149,
+      "step": 3635
+    },
+    {
+      "epoch": 3.356565889683822,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.581054036024016e-05,
+      "loss": 1.6282,
+      "step": 3636
+    },
+    {
+      "epoch": 3.3574890376182784,
+      "grad_norm": 3.599609375,
+      "learning_rate": 7.580386924616412e-05,
+      "loss": 1.5068,
+      "step": 3637
+    },
+    {
+      "epoch": 3.358412185552735,
+      "grad_norm": 3.927734375,
+      "learning_rate": 7.579719813208807e-05,
+      "loss": 1.4592,
+      "step": 3638
+    },
+    {
+      "epoch": 3.3593353334871914,
+      "grad_norm": 3.666015625,
+      "learning_rate": 7.579052701801201e-05,
+      "loss": 1.098,
+      "step": 3639
+    },
+    {
+      "epoch": 3.360258481421648,
+      "grad_norm": 4.1328125,
+      "learning_rate": 7.578385590393596e-05,
+      "loss": 1.5681,
+      "step": 3640
+    },
+    {
+      "epoch": 3.3611816293561043,
+      "grad_norm": 3.3515625,
+      "learning_rate": 7.577718478985992e-05,
+      "loss": 1.3136,
+      "step": 3641
+    },
+    {
+      "epoch": 3.362104777290561,
+      "grad_norm": 2.841796875,
+      "learning_rate": 7.577051367578386e-05,
+      "loss": 1.2198,
+      "step": 3642
+    },
+    {
+      "epoch": 3.3630279252250173,
+      "grad_norm": 3.673828125,
+      "learning_rate": 7.576384256170781e-05,
+      "loss": 1.5288,
+      "step": 3643
+    },
+    {
+      "epoch": 3.3639510731594737,
+      "grad_norm": 3.982421875,
+      "learning_rate": 7.575717144763175e-05,
+      "loss": 1.5579,
+      "step": 3644
+    },
+    {
+      "epoch": 3.36487422109393,
+      "grad_norm": 4.1171875,
+      "learning_rate": 7.575050033355571e-05,
+      "loss": 1.051,
+      "step": 3645
+    },
+    {
+      "epoch": 3.3657973690283867,
+      "grad_norm": 4.30078125,
+      "learning_rate": 7.574382921947966e-05,
+      "loss": 1.4187,
+      "step": 3646
+    },
+    {
+      "epoch": 3.366720516962843,
+      "grad_norm": 4.86328125,
+      "learning_rate": 7.57371581054036e-05,
+      "loss": 1.4355,
+      "step": 3647
+    },
+    {
+      "epoch": 3.3676436648972996,
+      "grad_norm": 5.40234375,
+      "learning_rate": 7.573048699132755e-05,
+      "loss": 1.3945,
+      "step": 3648
+    },
+    {
+      "epoch": 3.368566812831756,
+      "grad_norm": 5.5078125,
+      "learning_rate": 7.572381587725151e-05,
+      "loss": 1.4307,
+      "step": 3649
+    },
+    {
+      "epoch": 3.369489960766213,
+      "grad_norm": 4.30078125,
+      "learning_rate": 7.571714476317545e-05,
+      "loss": 1.6907,
+      "step": 3650
+    },
+    {
+      "epoch": 3.3704131087006695,
+      "grad_norm": 4.2890625,
+      "learning_rate": 7.57104736490994e-05,
+      "loss": 1.4199,
+      "step": 3651
+    },
+    {
+      "epoch": 3.371336256635126,
+      "grad_norm": 4.2890625,
+      "learning_rate": 7.570380253502334e-05,
+      "loss": 1.4968,
+      "step": 3652
+    },
+    {
+      "epoch": 3.3722594045695824,
+      "grad_norm": 4.23828125,
+      "learning_rate": 7.56971314209473e-05,
+      "loss": 1.4656,
+      "step": 3653
+    },
+    {
+      "epoch": 3.373182552504039,
+      "grad_norm": 3.92578125,
+      "learning_rate": 7.569046030687125e-05,
+      "loss": 1.4521,
+      "step": 3654
+    },
+    {
+      "epoch": 3.3741057004384953,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.56837891927952e-05,
+      "loss": 1.4973,
+      "step": 3655
+    },
+    {
+      "epoch": 3.375028848372952,
+      "grad_norm": 3.673828125,
+      "learning_rate": 7.567711807871914e-05,
+      "loss": 1.6008,
+      "step": 3656
+    },
+    {
+      "epoch": 3.3759519963074083,
+      "grad_norm": 4.328125,
+      "learning_rate": 7.56704469646431e-05,
+      "loss": 1.2377,
+      "step": 3657
+    },
+    {
+      "epoch": 3.3768751442418647,
+      "grad_norm": 4.41015625,
+      "learning_rate": 7.566377585056706e-05,
+      "loss": 1.5225,
+      "step": 3658
+    },
+    {
+      "epoch": 3.377798292176321,
+      "grad_norm": 4.2734375,
+      "learning_rate": 7.5657104736491e-05,
+      "loss": 1.2766,
+      "step": 3659
+    },
+    {
+      "epoch": 3.3787214401107777,
+      "grad_norm": 3.634765625,
+      "learning_rate": 7.565043362241495e-05,
+      "loss": 1.2983,
+      "step": 3660
+    },
+    {
+      "epoch": 3.379644588045234,
+      "grad_norm": 3.63671875,
+      "learning_rate": 7.56437625083389e-05,
+      "loss": 1.3904,
+      "step": 3661
+    },
+    {
+      "epoch": 3.3805677359796906,
+      "grad_norm": 4.125,
+      "learning_rate": 7.563709139426285e-05,
+      "loss": 1.3774,
+      "step": 3662
+    },
+    {
+      "epoch": 3.381490883914147,
+      "grad_norm": 4.30859375,
+      "learning_rate": 7.56304202801868e-05,
+      "loss": 1.5818,
+      "step": 3663
+    },
+    {
+      "epoch": 3.3824140318486036,
+      "grad_norm": 4.046875,
+      "learning_rate": 7.562374916611074e-05,
+      "loss": 1.2126,
+      "step": 3664
+    },
+    {
+      "epoch": 3.3833371797830605,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.56170780520347e-05,
+      "loss": 1.1656,
+      "step": 3665
+    },
+    {
+      "epoch": 3.3842603277175165,
+      "grad_norm": 3.767578125,
+      "learning_rate": 7.561040693795865e-05,
+      "loss": 1.2899,
+      "step": 3666
+    },
+    {
+      "epoch": 3.3851834756519734,
+      "grad_norm": 4.22265625,
+      "learning_rate": 7.560373582388259e-05,
+      "loss": 1.2544,
+      "step": 3667
+    },
+    {
+      "epoch": 3.38610662358643,
+      "grad_norm": 3.4453125,
+      "learning_rate": 7.559706470980654e-05,
+      "loss": 1.3086,
+      "step": 3668
+    },
+    {
+      "epoch": 3.3870297715208864,
+      "grad_norm": 4.078125,
+      "learning_rate": 7.55903935957305e-05,
+      "loss": 1.5566,
+      "step": 3669
+    },
+    {
+      "epoch": 3.387952919455343,
+      "grad_norm": 3.88671875,
+      "learning_rate": 7.558372248165444e-05,
+      "loss": 1.5481,
+      "step": 3670
+    },
+    {
+      "epoch": 3.3888760673897993,
+      "grad_norm": 4.1796875,
+      "learning_rate": 7.557705136757839e-05,
+      "loss": 1.3694,
+      "step": 3671
+    },
+    {
+      "epoch": 3.3897992153242558,
+      "grad_norm": 4.140625,
+      "learning_rate": 7.557038025350233e-05,
+      "loss": 1.3953,
+      "step": 3672
+    },
+    {
+      "epoch": 3.3907223632587122,
+      "grad_norm": 4.5390625,
+      "learning_rate": 7.556370913942629e-05,
+      "loss": 1.5073,
+      "step": 3673
+    },
+    {
+      "epoch": 3.3916455111931687,
+      "grad_norm": 4.62109375,
+      "learning_rate": 7.555703802535024e-05,
+      "loss": 1.5759,
+      "step": 3674
+    },
+    {
+      "epoch": 3.392568659127625,
+      "grad_norm": 5.02734375,
+      "learning_rate": 7.555036691127418e-05,
+      "loss": 1.708,
+      "step": 3675
+    },
+    {
+      "epoch": 3.3934918070620816,
+      "grad_norm": 4.33984375,
+      "learning_rate": 7.554369579719813e-05,
+      "loss": 1.5247,
+      "step": 3676
+    },
+    {
+      "epoch": 3.394414954996538,
+      "grad_norm": 3.724609375,
+      "learning_rate": 7.553702468312209e-05,
+      "loss": 1.4707,
+      "step": 3677
+    },
+    {
+      "epoch": 3.3953381029309946,
+      "grad_norm": 4.015625,
+      "learning_rate": 7.553035356904603e-05,
+      "loss": 1.4155,
+      "step": 3678
+    },
+    {
+      "epoch": 3.396261250865451,
+      "grad_norm": 4.32421875,
+      "learning_rate": 7.552368245496998e-05,
+      "loss": 1.1997,
+      "step": 3679
+    },
+    {
+      "epoch": 3.3971843987999075,
+      "grad_norm": 5.03125,
+      "learning_rate": 7.551701134089392e-05,
+      "loss": 1.4133,
+      "step": 3680
+    },
+    {
+      "epoch": 3.398107546734364,
+      "grad_norm": 4.0390625,
+      "learning_rate": 7.551034022681788e-05,
+      "loss": 1.5708,
+      "step": 3681
+    },
+    {
+      "epoch": 3.399030694668821,
+      "grad_norm": 5.24609375,
+      "learning_rate": 7.550366911274183e-05,
+      "loss": 1.2109,
+      "step": 3682
+    },
+    {
+      "epoch": 3.399953842603277,
+      "grad_norm": 3.806640625,
+      "learning_rate": 7.549699799866579e-05,
+      "loss": 1.4919,
+      "step": 3683
+    },
+    {
+      "epoch": 3.400876990537734,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.549032688458973e-05,
+      "loss": 1.2249,
+      "step": 3684
+    },
+    {
+      "epoch": 3.4018001384721903,
+      "grad_norm": 4.328125,
+      "learning_rate": 7.548365577051368e-05,
+      "loss": 1.415,
+      "step": 3685
+    },
+    {
+      "epoch": 3.402723286406647,
+      "grad_norm": 5.37109375,
+      "learning_rate": 7.547698465643764e-05,
+      "loss": 1.7175,
+      "step": 3686
+    },
+    {
+      "epoch": 3.4036464343411033,
+      "grad_norm": 3.59765625,
+      "learning_rate": 7.547031354236158e-05,
+      "loss": 1.562,
+      "step": 3687
+    },
+    {
+      "epoch": 3.4045695822755597,
+      "grad_norm": 4.2734375,
+      "learning_rate": 7.546364242828553e-05,
+      "loss": 1.5642,
+      "step": 3688
+    },
+    {
+      "epoch": 3.405492730210016,
+      "grad_norm": 4.78515625,
+      "learning_rate": 7.545697131420947e-05,
+      "loss": 1.2876,
+      "step": 3689
+    },
+    {
+      "epoch": 3.4064158781444727,
+      "grad_norm": 4.90234375,
+      "learning_rate": 7.545030020013343e-05,
+      "loss": 1.3491,
+      "step": 3690
+    },
+    {
+      "epoch": 3.407339026078929,
+      "grad_norm": 4.2265625,
+      "learning_rate": 7.544362908605738e-05,
+      "loss": 1.4386,
+      "step": 3691
+    },
+    {
+      "epoch": 3.4082621740133856,
+      "grad_norm": 3.708984375,
+      "learning_rate": 7.543695797198132e-05,
+      "loss": 1.448,
+      "step": 3692
+    },
+    {
+      "epoch": 3.409185321947842,
+      "grad_norm": 3.642578125,
+      "learning_rate": 7.543028685790527e-05,
+      "loss": 1.5642,
+      "step": 3693
+    },
+    {
+      "epoch": 3.4101084698822985,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.542361574382923e-05,
+      "loss": 1.4824,
+      "step": 3694
+    },
+    {
+      "epoch": 3.411031617816755,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.541694462975317e-05,
+      "loss": 1.5029,
+      "step": 3695
+    },
+    {
+      "epoch": 3.4119547657512115,
+      "grad_norm": 4.953125,
+      "learning_rate": 7.541027351567712e-05,
+      "loss": 1.3649,
+      "step": 3696
+    },
+    {
+      "epoch": 3.412877913685668,
+      "grad_norm": 4.55859375,
+      "learning_rate": 7.540360240160106e-05,
+      "loss": 1.5027,
+      "step": 3697
+    },
+    {
+      "epoch": 3.4138010616201244,
+      "grad_norm": 4.5859375,
+      "learning_rate": 7.539693128752502e-05,
+      "loss": 1.177,
+      "step": 3698
+    },
+    {
+      "epoch": 3.4147242095545813,
+      "grad_norm": 4.2109375,
+      "learning_rate": 7.539026017344897e-05,
+      "loss": 1.5464,
+      "step": 3699
+    },
+    {
+      "epoch": 3.415647357489038,
+      "grad_norm": 3.763671875,
+      "learning_rate": 7.538358905937291e-05,
+      "loss": 1.488,
+      "step": 3700
+    },
+    {
+      "epoch": 3.4165705054234943,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.537691794529686e-05,
+      "loss": 0.9857,
+      "step": 3701
+    },
+    {
+      "epoch": 3.4174936533579507,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.537024683122082e-05,
+      "loss": 1.266,
+      "step": 3702
+    },
+    {
+      "epoch": 3.418416801292407,
+      "grad_norm": 5.015625,
+      "learning_rate": 7.536357571714476e-05,
+      "loss": 1.3281,
+      "step": 3703
+    },
+    {
+      "epoch": 3.4193399492268637,
+      "grad_norm": 5.20703125,
+      "learning_rate": 7.535690460306871e-05,
+      "loss": 1.4124,
+      "step": 3704
+    },
+    {
+      "epoch": 3.42026309716132,
+      "grad_norm": 4.11328125,
+      "learning_rate": 7.535023348899266e-05,
+      "loss": 1.5696,
+      "step": 3705
+    },
+    {
+      "epoch": 3.4211862450957766,
+      "grad_norm": 3.435546875,
+      "learning_rate": 7.534356237491661e-05,
+      "loss": 1.286,
+      "step": 3706
+    },
+    {
+      "epoch": 3.422109393030233,
+      "grad_norm": 4.41015625,
+      "learning_rate": 7.533689126084056e-05,
+      "loss": 1.4575,
+      "step": 3707
+    },
+    {
+      "epoch": 3.4230325409646896,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.533022014676452e-05,
+      "loss": 1.4497,
+      "step": 3708
+    },
+    {
+      "epoch": 3.423955688899146,
+      "grad_norm": 4.8125,
+      "learning_rate": 7.532354903268846e-05,
+      "loss": 1.4397,
+      "step": 3709
+    },
+    {
+      "epoch": 3.4248788368336025,
+      "grad_norm": 4.25,
+      "learning_rate": 7.531687791861242e-05,
+      "loss": 1.2095,
+      "step": 3710
+    },
+    {
+      "epoch": 3.425801984768059,
+      "grad_norm": 4.61328125,
+      "learning_rate": 7.531020680453637e-05,
+      "loss": 1.2473,
+      "step": 3711
+    },
+    {
+      "epoch": 3.4267251327025154,
+      "grad_norm": 3.337890625,
+      "learning_rate": 7.530353569046031e-05,
+      "loss": 1.5383,
+      "step": 3712
+    },
+    {
+      "epoch": 3.427648280636972,
+      "grad_norm": 4.01953125,
+      "learning_rate": 7.529686457638426e-05,
+      "loss": 1.3955,
+      "step": 3713
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 3.71484375,
+      "learning_rate": 7.529019346230822e-05,
+      "loss": 1.2777,
+      "step": 3714
+    },
+    {
+      "epoch": 3.429494576505885,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.528352234823216e-05,
+      "loss": 1.4307,
+      "step": 3715
+    },
+    {
+      "epoch": 3.4304177244403418,
+      "grad_norm": 4.12890625,
+      "learning_rate": 7.527685123415611e-05,
+      "loss": 1.3147,
+      "step": 3716
+    },
+    {
+      "epoch": 3.4313408723747982,
+      "grad_norm": 3.73828125,
+      "learning_rate": 7.527018012008005e-05,
+      "loss": 1.3484,
+      "step": 3717
+    },
+    {
+      "epoch": 3.4322640203092547,
+      "grad_norm": 4.125,
+      "learning_rate": 7.526350900600401e-05,
+      "loss": 1.2023,
+      "step": 3718
+    },
+    {
+      "epoch": 3.433187168243711,
+      "grad_norm": 4.51171875,
+      "learning_rate": 7.525683789192796e-05,
+      "loss": 1.5952,
+      "step": 3719
+    },
+    {
+      "epoch": 3.4341103161781676,
+      "grad_norm": 4.5703125,
+      "learning_rate": 7.52501667778519e-05,
+      "loss": 1.5369,
+      "step": 3720
+    },
+    {
+      "epoch": 3.435033464112624,
+      "grad_norm": 4.35546875,
+      "learning_rate": 7.524349566377585e-05,
+      "loss": 1.4055,
+      "step": 3721
+    },
+    {
+      "epoch": 3.4359566120470806,
+      "grad_norm": 5.171875,
+      "learning_rate": 7.523682454969981e-05,
+      "loss": 1.3469,
+      "step": 3722
+    },
+    {
+      "epoch": 3.436879759981537,
+      "grad_norm": 5.0390625,
+      "learning_rate": 7.523015343562375e-05,
+      "loss": 1.7979,
+      "step": 3723
+    },
+    {
+      "epoch": 3.4378029079159935,
+      "grad_norm": 3.857421875,
+      "learning_rate": 7.52234823215477e-05,
+      "loss": 1.1268,
+      "step": 3724
+    },
+    {
+      "epoch": 3.43872605585045,
+      "grad_norm": 4.38671875,
+      "learning_rate": 7.521681120747164e-05,
+      "loss": 1.5759,
+      "step": 3725
+    },
+    {
+      "epoch": 3.4396492037849065,
+      "grad_norm": 3.8359375,
+      "learning_rate": 7.52101400933956e-05,
+      "loss": 1.4309,
+      "step": 3726
+    },
+    {
+      "epoch": 3.440572351719363,
+      "grad_norm": 3.736328125,
+      "learning_rate": 7.520346897931955e-05,
+      "loss": 1.3955,
+      "step": 3727
+    },
+    {
+      "epoch": 3.4414954996538194,
+      "grad_norm": 3.806640625,
+      "learning_rate": 7.51967978652435e-05,
+      "loss": 1.1724,
+      "step": 3728
+    },
+    {
+      "epoch": 3.442418647588276,
+      "grad_norm": 5.36328125,
+      "learning_rate": 7.519012675116744e-05,
+      "loss": 1.4319,
+      "step": 3729
+    },
+    {
+      "epoch": 3.4433417955227323,
+      "grad_norm": 4.5,
+      "learning_rate": 7.51834556370914e-05,
+      "loss": 1.2239,
+      "step": 3730
+    },
+    {
+      "epoch": 3.4442649434571893,
+      "grad_norm": 3.947265625,
+      "learning_rate": 7.517678452301534e-05,
+      "loss": 1.6208,
+      "step": 3731
+    },
+    {
+      "epoch": 3.4451880913916453,
+      "grad_norm": 3.73046875,
+      "learning_rate": 7.517011340893929e-05,
+      "loss": 1.3584,
+      "step": 3732
+    },
+    {
+      "epoch": 3.446111239326102,
+      "grad_norm": 3.423828125,
+      "learning_rate": 7.516344229486325e-05,
+      "loss": 1.3745,
+      "step": 3733
+    },
+    {
+      "epoch": 3.4470343872605587,
+      "grad_norm": 3.708984375,
+      "learning_rate": 7.51567711807872e-05,
+      "loss": 1.5291,
+      "step": 3734
+    },
+    {
+      "epoch": 3.447957535195015,
+      "grad_norm": 4.125,
+      "learning_rate": 7.515010006671115e-05,
+      "loss": 1.13,
+      "step": 3735
+    },
+    {
+      "epoch": 3.4488806831294716,
+      "grad_norm": 4.31640625,
+      "learning_rate": 7.51434289526351e-05,
+      "loss": 1.2985,
+      "step": 3736
+    },
+    {
+      "epoch": 3.449803831063928,
+      "grad_norm": 3.560546875,
+      "learning_rate": 7.513675783855904e-05,
+      "loss": 1.4679,
+      "step": 3737
+    },
+    {
+      "epoch": 3.4507269789983845,
+      "grad_norm": 4.3046875,
+      "learning_rate": 7.513008672448299e-05,
+      "loss": 1.3684,
+      "step": 3738
+    },
+    {
+      "epoch": 3.451650126932841,
+      "grad_norm": 4.42578125,
+      "learning_rate": 7.512341561040695e-05,
+      "loss": 1.4373,
+      "step": 3739
+    },
+    {
+      "epoch": 3.4525732748672975,
+      "grad_norm": 4.13671875,
+      "learning_rate": 7.51167444963309e-05,
+      "loss": 1.4113,
+      "step": 3740
+    },
+    {
+      "epoch": 3.453496422801754,
+      "grad_norm": 5.203125,
+      "learning_rate": 7.511007338225484e-05,
+      "loss": 1.1451,
+      "step": 3741
+    },
+    {
+      "epoch": 3.4544195707362104,
+      "grad_norm": 4.5390625,
+      "learning_rate": 7.510340226817879e-05,
+      "loss": 1.5181,
+      "step": 3742
+    },
+    {
+      "epoch": 3.455342718670667,
+      "grad_norm": 4.9296875,
+      "learning_rate": 7.509673115410274e-05,
+      "loss": 1.4053,
+      "step": 3743
+    },
+    {
+      "epoch": 3.4562658666051234,
+      "grad_norm": 5.10546875,
+      "learning_rate": 7.509006004002669e-05,
+      "loss": 1.7007,
+      "step": 3744
+    },
+    {
+      "epoch": 3.45718901453958,
+      "grad_norm": 5.17578125,
+      "learning_rate": 7.508338892595063e-05,
+      "loss": 1.5132,
+      "step": 3745
+    },
+    {
+      "epoch": 3.4581121624740363,
+      "grad_norm": 5.1328125,
+      "learning_rate": 7.507671781187458e-05,
+      "loss": 1.6421,
+      "step": 3746
+    },
+    {
+      "epoch": 3.4590353104084928,
+      "grad_norm": 4.62109375,
+      "learning_rate": 7.507004669779854e-05,
+      "loss": 1.2458,
+      "step": 3747
+    },
+    {
+      "epoch": 3.4599584583429497,
+      "grad_norm": 3.8984375,
+      "learning_rate": 7.506337558372248e-05,
+      "loss": 1.5842,
+      "step": 3748
+    },
+    {
+      "epoch": 3.4608816062774057,
+      "grad_norm": 3.8046875,
+      "learning_rate": 7.505670446964643e-05,
+      "loss": 1.3916,
+      "step": 3749
+    },
+    {
+      "epoch": 3.4618047542118626,
+      "grad_norm": 4.75390625,
+      "learning_rate": 7.505003335557038e-05,
+      "loss": 1.7798,
+      "step": 3750
+    },
+    {
+      "epoch": 3.462727902146319,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.504336224149433e-05,
+      "loss": 1.4829,
+      "step": 3751
+    },
+    {
+      "epoch": 3.4636510500807756,
+      "grad_norm": 4.7265625,
+      "learning_rate": 7.503669112741828e-05,
+      "loss": 1.3662,
+      "step": 3752
+    },
+    {
+      "epoch": 3.464574198015232,
+      "grad_norm": 4.0,
+      "learning_rate": 7.503002001334223e-05,
+      "loss": 1.6091,
+      "step": 3753
+    },
+    {
+      "epoch": 3.4654973459496885,
+      "grad_norm": 3.7421875,
+      "learning_rate": 7.502334889926617e-05,
+      "loss": 1.3191,
+      "step": 3754
+    },
+    {
+      "epoch": 3.466420493884145,
+      "grad_norm": 4.30859375,
+      "learning_rate": 7.501667778519013e-05,
+      "loss": 1.5822,
+      "step": 3755
+    },
+    {
+      "epoch": 3.4673436418186014,
+      "grad_norm": 4.69140625,
+      "learning_rate": 7.501000667111408e-05,
+      "loss": 1.177,
+      "step": 3756
+    },
+    {
+      "epoch": 3.468266789753058,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.500333555703802e-05,
+      "loss": 1.552,
+      "step": 3757
+    },
+    {
+      "epoch": 3.4691899376875144,
+      "grad_norm": 4.4453125,
+      "learning_rate": 7.499666444296198e-05,
+      "loss": 1.5828,
+      "step": 3758
+    },
+    {
+      "epoch": 3.470113085621971,
+      "grad_norm": 5.22265625,
+      "learning_rate": 7.498999332888594e-05,
+      "loss": 1.3765,
+      "step": 3759
+    },
+    {
+      "epoch": 3.4710362335564273,
+      "grad_norm": 5.65234375,
+      "learning_rate": 7.498332221480988e-05,
+      "loss": 1.3794,
+      "step": 3760
+    },
+    {
+      "epoch": 3.471959381490884,
+      "grad_norm": 3.8984375,
+      "learning_rate": 7.497665110073383e-05,
+      "loss": 1.1438,
+      "step": 3761
+    },
+    {
+      "epoch": 3.4728825294253403,
+      "grad_norm": 4.79296875,
+      "learning_rate": 7.496997998665777e-05,
+      "loss": 1.4636,
+      "step": 3762
+    },
+    {
+      "epoch": 3.4738056773597967,
+      "grad_norm": 4.48046875,
+      "learning_rate": 7.496330887258173e-05,
+      "loss": 1.1045,
+      "step": 3763
+    },
+    {
+      "epoch": 3.474728825294253,
+      "grad_norm": 5.36328125,
+      "learning_rate": 7.495663775850568e-05,
+      "loss": 1.6401,
+      "step": 3764
+    },
+    {
+      "epoch": 3.47565197322871,
+      "grad_norm": 3.798828125,
+      "learning_rate": 7.494996664442962e-05,
+      "loss": 1.4348,
+      "step": 3765
+    },
+    {
+      "epoch": 3.4765751211631666,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.494329553035357e-05,
+      "loss": 1.3872,
+      "step": 3766
+    },
+    {
+      "epoch": 3.477498269097623,
+      "grad_norm": 3.697265625,
+      "learning_rate": 7.493662441627753e-05,
+      "loss": 1.473,
+      "step": 3767
+    },
+    {
+      "epoch": 3.4784214170320795,
+      "grad_norm": 3.9765625,
+      "learning_rate": 7.492995330220147e-05,
+      "loss": 1.5098,
+      "step": 3768
+    },
+    {
+      "epoch": 3.479344564966536,
+      "grad_norm": 4.07421875,
+      "learning_rate": 7.492328218812542e-05,
+      "loss": 1.3278,
+      "step": 3769
+    },
+    {
+      "epoch": 3.4802677129009925,
+      "grad_norm": 3.814453125,
+      "learning_rate": 7.491661107404937e-05,
+      "loss": 1.6306,
+      "step": 3770
+    },
+    {
+      "epoch": 3.481190860835449,
+      "grad_norm": 3.931640625,
+      "learning_rate": 7.490993995997332e-05,
+      "loss": 1.2736,
+      "step": 3771
+    },
+    {
+      "epoch": 3.4821140087699054,
+      "grad_norm": 3.767578125,
+      "learning_rate": 7.490326884589727e-05,
+      "loss": 1.5144,
+      "step": 3772
+    },
+    {
+      "epoch": 3.483037156704362,
+      "grad_norm": 3.939453125,
+      "learning_rate": 7.489659773182122e-05,
+      "loss": 1.3804,
+      "step": 3773
+    },
+    {
+      "epoch": 3.4839603046388183,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.488992661774516e-05,
+      "loss": 1.3246,
+      "step": 3774
+    },
+    {
+      "epoch": 3.484883452573275,
+      "grad_norm": 5.4921875,
+      "learning_rate": 7.488325550366912e-05,
+      "loss": 1.6355,
+      "step": 3775
+    },
+    {
+      "epoch": 3.4858066005077313,
+      "grad_norm": 4.17578125,
+      "learning_rate": 7.487658438959307e-05,
+      "loss": 1.4448,
+      "step": 3776
+    },
+    {
+      "epoch": 3.4867297484421877,
+      "grad_norm": 5.51953125,
+      "learning_rate": 7.486991327551701e-05,
+      "loss": 1.1517,
+      "step": 3777
+    },
+    {
+      "epoch": 3.487652896376644,
+      "grad_norm": 3.99609375,
+      "learning_rate": 7.486324216144096e-05,
+      "loss": 1.5278,
+      "step": 3778
+    },
+    {
+      "epoch": 3.4885760443111007,
+      "grad_norm": 4.47265625,
+      "learning_rate": 7.485657104736492e-05,
+      "loss": 1.3071,
+      "step": 3779
+    },
+    {
+      "epoch": 3.489499192245557,
+      "grad_norm": 3.646484375,
+      "learning_rate": 7.484989993328886e-05,
+      "loss": 1.4287,
+      "step": 3780
+    },
+    {
+      "epoch": 3.4904223401800136,
+      "grad_norm": 5.01953125,
+      "learning_rate": 7.48432288192128e-05,
+      "loss": 1.6604,
+      "step": 3781
+    },
+    {
+      "epoch": 3.4913454881144705,
+      "grad_norm": 3.658203125,
+      "learning_rate": 7.483655770513675e-05,
+      "loss": 1.3008,
+      "step": 3782
+    },
+    {
+      "epoch": 3.492268636048927,
+      "grad_norm": 3.51953125,
+      "learning_rate": 7.482988659106071e-05,
+      "loss": 1.2991,
+      "step": 3783
+    },
+    {
+      "epoch": 3.4931917839833835,
+      "grad_norm": 3.755859375,
+      "learning_rate": 7.482321547698467e-05,
+      "loss": 1.457,
+      "step": 3784
+    },
+    {
+      "epoch": 3.49411493191784,
+      "grad_norm": 3.974609375,
+      "learning_rate": 7.481654436290861e-05,
+      "loss": 1.5154,
+      "step": 3785
+    },
+    {
+      "epoch": 3.4950380798522964,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.480987324883256e-05,
+      "loss": 1.5767,
+      "step": 3786
+    },
+    {
+      "epoch": 3.495961227786753,
+      "grad_norm": 4.3203125,
+      "learning_rate": 7.48032021347565e-05,
+      "loss": 1.0872,
+      "step": 3787
+    },
+    {
+      "epoch": 3.4968843757212094,
+      "grad_norm": 4.05859375,
+      "learning_rate": 7.479653102068046e-05,
+      "loss": 1.6577,
+      "step": 3788
+    },
+    {
+      "epoch": 3.497807523655666,
+      "grad_norm": 4.16796875,
+      "learning_rate": 7.478985990660441e-05,
+      "loss": 1.5654,
+      "step": 3789
+    },
+    {
+      "epoch": 3.4987306715901223,
+      "grad_norm": 3.966796875,
+      "learning_rate": 7.478318879252836e-05,
+      "loss": 1.2068,
+      "step": 3790
+    },
+    {
+      "epoch": 3.4996538195245788,
+      "grad_norm": 4.453125,
+      "learning_rate": 7.47765176784523e-05,
+      "loss": 1.0741,
+      "step": 3791
+    },
+    {
+      "epoch": 3.5005769674590352,
+      "grad_norm": 4.69921875,
+      "learning_rate": 7.476984656437626e-05,
+      "loss": 1.5579,
+      "step": 3792
+    },
+    {
+      "epoch": 3.5015001153934917,
+      "grad_norm": 4.375,
+      "learning_rate": 7.47631754503002e-05,
+      "loss": 1.4248,
+      "step": 3793
+    },
+    {
+      "epoch": 3.502423263327948,
+      "grad_norm": 4.83984375,
+      "learning_rate": 7.475650433622415e-05,
+      "loss": 1.3389,
+      "step": 3794
+    },
+    {
+      "epoch": 3.5033464112624046,
+      "grad_norm": 5.84765625,
+      "learning_rate": 7.47498332221481e-05,
+      "loss": 1.5212,
+      "step": 3795
+    },
+    {
+      "epoch": 3.504269559196861,
+      "grad_norm": 3.634765625,
+      "learning_rate": 7.474316210807206e-05,
+      "loss": 1.5107,
+      "step": 3796
+    },
+    {
+      "epoch": 3.505192707131318,
+      "grad_norm": 4.28125,
+      "learning_rate": 7.4736490993996e-05,
+      "loss": 1.6577,
+      "step": 3797
+    },
+    {
+      "epoch": 3.506115855065774,
+      "grad_norm": 3.341796875,
+      "learning_rate": 7.472981987991995e-05,
+      "loss": 1.4841,
+      "step": 3798
+    },
+    {
+      "epoch": 3.507039003000231,
+      "grad_norm": 4.50390625,
+      "learning_rate": 7.472314876584389e-05,
+      "loss": 1.4497,
+      "step": 3799
+    },
+    {
+      "epoch": 3.5079621509346874,
+      "grad_norm": 3.59765625,
+      "learning_rate": 7.471647765176785e-05,
+      "loss": 1.3807,
+      "step": 3800
+    },
+    {
+      "epoch": 3.508885298869144,
+      "grad_norm": 3.53515625,
+      "learning_rate": 7.47098065376918e-05,
+      "loss": 1.3835,
+      "step": 3801
+    },
+    {
+      "epoch": 3.5098084468036004,
+      "grad_norm": 3.22265625,
+      "learning_rate": 7.470313542361574e-05,
+      "loss": 1.3375,
+      "step": 3802
+    },
+    {
+      "epoch": 3.510731594738057,
+      "grad_norm": 4.359375,
+      "learning_rate": 7.469646430953969e-05,
+      "loss": 1.4534,
+      "step": 3803
+    },
+    {
+      "epoch": 3.5116547426725133,
+      "grad_norm": 3.939453125,
+      "learning_rate": 7.468979319546365e-05,
+      "loss": 1.1533,
+      "step": 3804
+    },
+    {
+      "epoch": 3.51257789060697,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.468312208138759e-05,
+      "loss": 1.3478,
+      "step": 3805
+    },
+    {
+      "epoch": 3.5135010385414263,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.467645096731154e-05,
+      "loss": 1.6021,
+      "step": 3806
+    },
+    {
+      "epoch": 3.5144241864758827,
+      "grad_norm": 3.814453125,
+      "learning_rate": 7.466977985323548e-05,
+      "loss": 1.4236,
+      "step": 3807
+    },
+    {
+      "epoch": 3.515347334410339,
+      "grad_norm": 4.375,
+      "learning_rate": 7.466310873915945e-05,
+      "loss": 1.259,
+      "step": 3808
+    },
+    {
+      "epoch": 3.5162704823447957,
+      "grad_norm": 4.50390625,
+      "learning_rate": 7.46564376250834e-05,
+      "loss": 1.3892,
+      "step": 3809
+    },
+    {
+      "epoch": 3.517193630279252,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.464976651100735e-05,
+      "loss": 1.4661,
+      "step": 3810
+    },
+    {
+      "epoch": 3.5181167782137086,
+      "grad_norm": 5.578125,
+      "learning_rate": 7.464309539693129e-05,
+      "loss": 1.4333,
+      "step": 3811
+    },
+    {
+      "epoch": 3.5190399261481655,
+      "grad_norm": 4.01171875,
+      "learning_rate": 7.463642428285525e-05,
+      "loss": 1.3462,
+      "step": 3812
+    },
+    {
+      "epoch": 3.5199630740826215,
+      "grad_norm": 4.7734375,
+      "learning_rate": 7.46297531687792e-05,
+      "loss": 1.4519,
+      "step": 3813
+    },
+    {
+      "epoch": 3.5208862220170785,
+      "grad_norm": 3.728515625,
+      "learning_rate": 7.462308205470314e-05,
+      "loss": 1.311,
+      "step": 3814
+    },
+    {
+      "epoch": 3.5218093699515345,
+      "grad_norm": 4.1171875,
+      "learning_rate": 7.461641094062709e-05,
+      "loss": 1.2437,
+      "step": 3815
+    },
+    {
+      "epoch": 3.5227325178859914,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.460973982655105e-05,
+      "loss": 1.3102,
+      "step": 3816
+    },
+    {
+      "epoch": 3.523655665820448,
+      "grad_norm": 5.296875,
+      "learning_rate": 7.460306871247499e-05,
+      "loss": 1.603,
+      "step": 3817
+    },
+    {
+      "epoch": 3.5245788137549043,
+      "grad_norm": 3.80859375,
+      "learning_rate": 7.459639759839894e-05,
+      "loss": 1.3018,
+      "step": 3818
+    },
+    {
+      "epoch": 3.525501961689361,
+      "grad_norm": 4.6171875,
+      "learning_rate": 7.458972648432288e-05,
+      "loss": 1.3091,
+      "step": 3819
+    },
+    {
+      "epoch": 3.5264251096238173,
+      "grad_norm": 4.453125,
+      "learning_rate": 7.458305537024684e-05,
+      "loss": 1.384,
+      "step": 3820
+    },
+    {
+      "epoch": 3.5273482575582737,
+      "grad_norm": 4.51171875,
+      "learning_rate": 7.457638425617079e-05,
+      "loss": 1.6055,
+      "step": 3821
+    },
+    {
+      "epoch": 3.52827140549273,
+      "grad_norm": 4.30859375,
+      "learning_rate": 7.456971314209473e-05,
+      "loss": 1.1517,
+      "step": 3822
+    },
+    {
+      "epoch": 3.5291945534271867,
+      "grad_norm": 4.26953125,
+      "learning_rate": 7.456304202801868e-05,
+      "loss": 1.3525,
+      "step": 3823
+    },
+    {
+      "epoch": 3.530117701361643,
+      "grad_norm": 3.73828125,
+      "learning_rate": 7.455637091394264e-05,
+      "loss": 1.4956,
+      "step": 3824
+    },
+    {
+      "epoch": 3.5310408492960996,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.454969979986658e-05,
+      "loss": 1.6133,
+      "step": 3825
+    },
+    {
+      "epoch": 3.531963997230556,
+      "grad_norm": 3.84765625,
+      "learning_rate": 7.454302868579053e-05,
+      "loss": 1.5139,
+      "step": 3826
+    },
+    {
+      "epoch": 3.5328871451650126,
+      "grad_norm": 3.9609375,
+      "learning_rate": 7.453635757171447e-05,
+      "loss": 1.5579,
+      "step": 3827
+    },
+    {
+      "epoch": 3.533810293099469,
+      "grad_norm": 3.7734375,
+      "learning_rate": 7.452968645763843e-05,
+      "loss": 1.1534,
+      "step": 3828
+    },
+    {
+      "epoch": 3.534733441033926,
+      "grad_norm": 5.06640625,
+      "learning_rate": 7.452301534356238e-05,
+      "loss": 1.604,
+      "step": 3829
+    },
+    {
+      "epoch": 3.535656588968382,
+      "grad_norm": 3.822265625,
+      "learning_rate": 7.451634422948632e-05,
+      "loss": 1.4243,
+      "step": 3830
+    },
+    {
+      "epoch": 3.536579736902839,
+      "grad_norm": 4.53515625,
+      "learning_rate": 7.450967311541027e-05,
+      "loss": 1.6345,
+      "step": 3831
+    },
+    {
+      "epoch": 3.537502884837295,
+      "grad_norm": 3.693359375,
+      "learning_rate": 7.450300200133423e-05,
+      "loss": 1.49,
+      "step": 3832
+    },
+    {
+      "epoch": 3.538426032771752,
+      "grad_norm": 3.6640625,
+      "learning_rate": 7.449633088725819e-05,
+      "loss": 1.2749,
+      "step": 3833
+    },
+    {
+      "epoch": 3.5393491807062083,
+      "grad_norm": 4.49609375,
+      "learning_rate": 7.448965977318213e-05,
+      "loss": 1.3591,
+      "step": 3834
+    },
+    {
+      "epoch": 3.5402723286406648,
+      "grad_norm": 4.75,
+      "learning_rate": 7.448298865910608e-05,
+      "loss": 1.5322,
+      "step": 3835
+    },
+    {
+      "epoch": 3.5411954765751212,
+      "grad_norm": 5.30859375,
+      "learning_rate": 7.447631754503002e-05,
+      "loss": 1.4001,
+      "step": 3836
+    },
+    {
+      "epoch": 3.5421186245095777,
+      "grad_norm": 4.35546875,
+      "learning_rate": 7.446964643095398e-05,
+      "loss": 1.366,
+      "step": 3837
+    },
+    {
+      "epoch": 3.543041772444034,
+      "grad_norm": 3.755859375,
+      "learning_rate": 7.446297531687793e-05,
+      "loss": 1.5813,
+      "step": 3838
+    },
+    {
+      "epoch": 3.5439649203784906,
+      "grad_norm": 3.689453125,
+      "learning_rate": 7.445630420280187e-05,
+      "loss": 1.0376,
+      "step": 3839
+    },
+    {
+      "epoch": 3.544888068312947,
+      "grad_norm": 3.458984375,
+      "learning_rate": 7.444963308872582e-05,
+      "loss": 1.3374,
+      "step": 3840
+    },
+    {
+      "epoch": 3.5458112162474036,
+      "grad_norm": 4.6484375,
+      "learning_rate": 7.444296197464978e-05,
+      "loss": 1.427,
+      "step": 3841
+    },
+    {
+      "epoch": 3.54673436418186,
+      "grad_norm": 3.978515625,
+      "learning_rate": 7.443629086057372e-05,
+      "loss": 1.0121,
+      "step": 3842
+    },
+    {
+      "epoch": 3.5476575121163165,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.442961974649767e-05,
+      "loss": 1.1056,
+      "step": 3843
+    },
+    {
+      "epoch": 3.548580660050773,
+      "grad_norm": 4.49609375,
+      "learning_rate": 7.442294863242161e-05,
+      "loss": 1.3936,
+      "step": 3844
+    },
+    {
+      "epoch": 3.5495038079852295,
+      "grad_norm": 4.26953125,
+      "learning_rate": 7.441627751834557e-05,
+      "loss": 1.4993,
+      "step": 3845
+    },
+    {
+      "epoch": 3.5504269559196864,
+      "grad_norm": 3.353515625,
+      "learning_rate": 7.440960640426952e-05,
+      "loss": 1.6467,
+      "step": 3846
+    },
+    {
+      "epoch": 3.5513501038541424,
+      "grad_norm": 4.359375,
+      "learning_rate": 7.440293529019346e-05,
+      "loss": 1.4182,
+      "step": 3847
+    },
+    {
+      "epoch": 3.5522732517885993,
+      "grad_norm": 4.875,
+      "learning_rate": 7.439626417611741e-05,
+      "loss": 1.3862,
+      "step": 3848
+    },
+    {
+      "epoch": 3.5531963997230553,
+      "grad_norm": 4.0859375,
+      "learning_rate": 7.438959306204137e-05,
+      "loss": 1.5732,
+      "step": 3849
+    },
+    {
+      "epoch": 3.5541195476575123,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.438292194796531e-05,
+      "loss": 1.7739,
+      "step": 3850
+    },
+    {
+      "epoch": 3.5550426955919687,
+      "grad_norm": 3.634765625,
+      "learning_rate": 7.437625083388926e-05,
+      "loss": 1.5632,
+      "step": 3851
+    },
+    {
+      "epoch": 3.555965843526425,
+      "grad_norm": 4.03515625,
+      "learning_rate": 7.436957971981322e-05,
+      "loss": 1.4109,
+      "step": 3852
+    },
+    {
+      "epoch": 3.5568889914608817,
+      "grad_norm": 5.19921875,
+      "learning_rate": 7.436290860573716e-05,
+      "loss": 1.6987,
+      "step": 3853
+    },
+    {
+      "epoch": 3.557812139395338,
+      "grad_norm": 5.2421875,
+      "learning_rate": 7.435623749166111e-05,
+      "loss": 1.2162,
+      "step": 3854
+    },
+    {
+      "epoch": 3.5587352873297946,
+      "grad_norm": 4.87109375,
+      "learning_rate": 7.434956637758505e-05,
+      "loss": 1.2191,
+      "step": 3855
+    },
+    {
+      "epoch": 3.559658435264251,
+      "grad_norm": 3.9765625,
+      "learning_rate": 7.434289526350901e-05,
+      "loss": 1.5527,
+      "step": 3856
+    },
+    {
+      "epoch": 3.5605815831987075,
+      "grad_norm": 4.6796875,
+      "learning_rate": 7.433622414943296e-05,
+      "loss": 1.5535,
+      "step": 3857
+    },
+    {
+      "epoch": 3.561504731133164,
+      "grad_norm": 4.26171875,
+      "learning_rate": 7.432955303535692e-05,
+      "loss": 1.3618,
+      "step": 3858
+    },
+    {
+      "epoch": 3.5624278790676205,
+      "grad_norm": 4.2265625,
+      "learning_rate": 7.432288192128086e-05,
+      "loss": 1.5381,
+      "step": 3859
+    },
+    {
+      "epoch": 3.563351027002077,
+      "grad_norm": 3.31640625,
+      "learning_rate": 7.43162108072048e-05,
+      "loss": 1.5186,
+      "step": 3860
+    },
+    {
+      "epoch": 3.5642741749365334,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.430953969312877e-05,
+      "loss": 1.4338,
+      "step": 3861
+    },
+    {
+      "epoch": 3.56519732287099,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.430286857905271e-05,
+      "loss": 1.2607,
+      "step": 3862
+    },
+    {
+      "epoch": 3.566120470805447,
+      "grad_norm": 4.74609375,
+      "learning_rate": 7.429619746497666e-05,
+      "loss": 1.1243,
+      "step": 3863
+    },
+    {
+      "epoch": 3.567043618739903,
+      "grad_norm": 3.845703125,
+      "learning_rate": 7.42895263509006e-05,
+      "loss": 1.2717,
+      "step": 3864
+    },
+    {
+      "epoch": 3.5679667666743597,
+      "grad_norm": 4.48828125,
+      "learning_rate": 7.428285523682456e-05,
+      "loss": 1.3616,
+      "step": 3865
+    },
+    {
+      "epoch": 3.568889914608816,
+      "grad_norm": 5.05078125,
+      "learning_rate": 7.42761841227485e-05,
+      "loss": 1.2556,
+      "step": 3866
+    },
+    {
+      "epoch": 3.5698130625432727,
+      "grad_norm": 4.49609375,
+      "learning_rate": 7.426951300867245e-05,
+      "loss": 1.2909,
+      "step": 3867
+    },
+    {
+      "epoch": 3.570736210477729,
+      "grad_norm": 4.60546875,
+      "learning_rate": 7.42628418945964e-05,
+      "loss": 1.4656,
+      "step": 3868
+    },
+    {
+      "epoch": 3.5716593584121856,
+      "grad_norm": 5.625,
+      "learning_rate": 7.425617078052036e-05,
+      "loss": 1.377,
+      "step": 3869
+    },
+    {
+      "epoch": 3.572582506346642,
+      "grad_norm": 5.07421875,
+      "learning_rate": 7.42494996664443e-05,
+      "loss": 1.3464,
+      "step": 3870
+    },
+    {
+      "epoch": 3.5735056542810986,
+      "grad_norm": 4.8828125,
+      "learning_rate": 7.424282855236825e-05,
+      "loss": 1.4153,
+      "step": 3871
+    },
+    {
+      "epoch": 3.574428802215555,
+      "grad_norm": 5.75390625,
+      "learning_rate": 7.423615743829219e-05,
+      "loss": 1.5972,
+      "step": 3872
+    },
+    {
+      "epoch": 3.5753519501500115,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.422948632421615e-05,
+      "loss": 1.5083,
+      "step": 3873
+    },
+    {
+      "epoch": 3.576275098084468,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.42228152101401e-05,
+      "loss": 1.3408,
+      "step": 3874
+    },
+    {
+      "epoch": 3.5771982460189244,
+      "grad_norm": 6.36328125,
+      "learning_rate": 7.421614409606404e-05,
+      "loss": 1.0304,
+      "step": 3875
+    },
+    {
+      "epoch": 3.578121393953381,
+      "grad_norm": 4.5390625,
+      "learning_rate": 7.420947298198799e-05,
+      "loss": 1.7007,
+      "step": 3876
+    },
+    {
+      "epoch": 3.5790445418878374,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.420280186791195e-05,
+      "loss": 1.6128,
+      "step": 3877
+    },
+    {
+      "epoch": 3.5799676898222943,
+      "grad_norm": 3.48046875,
+      "learning_rate": 7.419613075383589e-05,
+      "loss": 1.4877,
+      "step": 3878
+    },
+    {
+      "epoch": 3.5808908377567503,
+      "grad_norm": 3.375,
+      "learning_rate": 7.418945963975984e-05,
+      "loss": 1.6072,
+      "step": 3879
+    },
+    {
+      "epoch": 3.5818139856912072,
+      "grad_norm": 3.751953125,
+      "learning_rate": 7.418278852568378e-05,
+      "loss": 1.3513,
+      "step": 3880
+    },
+    {
+      "epoch": 3.5827371336256633,
+      "grad_norm": 4.4609375,
+      "learning_rate": 7.417611741160774e-05,
+      "loss": 1.4458,
+      "step": 3881
+    },
+    {
+      "epoch": 3.58366028156012,
+      "grad_norm": 3.826171875,
+      "learning_rate": 7.416944629753169e-05,
+      "loss": 1.4905,
+      "step": 3882
+    },
+    {
+      "epoch": 3.5845834294945766,
+      "grad_norm": 3.90234375,
+      "learning_rate": 7.416277518345565e-05,
+      "loss": 1.4283,
+      "step": 3883
+    },
+    {
+      "epoch": 3.585506577429033,
+      "grad_norm": 5.19921875,
+      "learning_rate": 7.415610406937959e-05,
+      "loss": 1.6045,
+      "step": 3884
+    },
+    {
+      "epoch": 3.5864297253634896,
+      "grad_norm": 4.31640625,
+      "learning_rate": 7.414943295530354e-05,
+      "loss": 1.7583,
+      "step": 3885
+    },
+    {
+      "epoch": 3.587352873297946,
+      "grad_norm": 3.23828125,
+      "learning_rate": 7.41427618412275e-05,
+      "loss": 1.7197,
+      "step": 3886
+    },
+    {
+      "epoch": 3.5882760212324025,
+      "grad_norm": 3.96484375,
+      "learning_rate": 7.413609072715144e-05,
+      "loss": 1.2323,
+      "step": 3887
+    },
+    {
+      "epoch": 3.589199169166859,
+      "grad_norm": 4.0390625,
+      "learning_rate": 7.412941961307539e-05,
+      "loss": 1.4758,
+      "step": 3888
+    },
+    {
+      "epoch": 3.5901223171013155,
+      "grad_norm": 5.01953125,
+      "learning_rate": 7.412274849899933e-05,
+      "loss": 1.6799,
+      "step": 3889
+    },
+    {
+      "epoch": 3.591045465035772,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.411607738492329e-05,
+      "loss": 1.179,
+      "step": 3890
+    },
+    {
+      "epoch": 3.5919686129702284,
+      "grad_norm": 4.875,
+      "learning_rate": 7.410940627084724e-05,
+      "loss": 1.7554,
+      "step": 3891
+    },
+    {
+      "epoch": 3.592891760904685,
+      "grad_norm": 4.3203125,
+      "learning_rate": 7.410273515677118e-05,
+      "loss": 1.4963,
+      "step": 3892
+    },
+    {
+      "epoch": 3.5938149088391413,
+      "grad_norm": 4.56640625,
+      "learning_rate": 7.409606404269514e-05,
+      "loss": 1.4768,
+      "step": 3893
+    },
+    {
+      "epoch": 3.594738056773598,
+      "grad_norm": 3.345703125,
+      "learning_rate": 7.408939292861909e-05,
+      "loss": 1.4801,
+      "step": 3894
+    },
+    {
+      "epoch": 3.5956612047080547,
+      "grad_norm": 4.32421875,
+      "learning_rate": 7.408272181454303e-05,
+      "loss": 1.3413,
+      "step": 3895
+    },
+    {
+      "epoch": 3.5965843526425108,
+      "grad_norm": 3.908203125,
+      "learning_rate": 7.407605070046698e-05,
+      "loss": 1.4644,
+      "step": 3896
+    },
+    {
+      "epoch": 3.5975075005769677,
+      "grad_norm": 3.626953125,
+      "learning_rate": 7.406937958639094e-05,
+      "loss": 1.5574,
+      "step": 3897
+    },
+    {
+      "epoch": 3.5984306485114237,
+      "grad_norm": 3.72265625,
+      "learning_rate": 7.406270847231488e-05,
+      "loss": 1.5942,
+      "step": 3898
+    },
+    {
+      "epoch": 3.5993537964458806,
+      "grad_norm": 3.970703125,
+      "learning_rate": 7.405603735823883e-05,
+      "loss": 1.7195,
+      "step": 3899
+    },
+    {
+      "epoch": 3.600276944380337,
+      "grad_norm": 3.763671875,
+      "learning_rate": 7.404936624416277e-05,
+      "loss": 1.6467,
+      "step": 3900
+    },
+    {
+      "epoch": 3.6012000923147935,
+      "grad_norm": 4.27734375,
+      "learning_rate": 7.404269513008673e-05,
+      "loss": 1.6277,
+      "step": 3901
+    },
+    {
+      "epoch": 3.60212324024925,
+      "grad_norm": 4.76953125,
+      "learning_rate": 7.403602401601068e-05,
+      "loss": 1.2145,
+      "step": 3902
+    },
+    {
+      "epoch": 3.6030463881837065,
+      "grad_norm": 4.421875,
+      "learning_rate": 7.402935290193462e-05,
+      "loss": 1.4587,
+      "step": 3903
+    },
+    {
+      "epoch": 3.603969536118163,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.402268178785857e-05,
+      "loss": 1.401,
+      "step": 3904
+    },
+    {
+      "epoch": 3.6048926840526194,
+      "grad_norm": 3.439453125,
+      "learning_rate": 7.401601067378253e-05,
+      "loss": 1.6318,
+      "step": 3905
+    },
+    {
+      "epoch": 3.605815831987076,
+      "grad_norm": 4.50390625,
+      "learning_rate": 7.400933955970647e-05,
+      "loss": 1.3171,
+      "step": 3906
+    },
+    {
+      "epoch": 3.6067389799215324,
+      "grad_norm": 4.61328125,
+      "learning_rate": 7.400266844563042e-05,
+      "loss": 1.7124,
+      "step": 3907
+    },
+    {
+      "epoch": 3.607662127855989,
+      "grad_norm": 3.958984375,
+      "learning_rate": 7.399599733155438e-05,
+      "loss": 1.4739,
+      "step": 3908
+    },
+    {
+      "epoch": 3.6085852757904453,
+      "grad_norm": 3.94921875,
+      "learning_rate": 7.398932621747832e-05,
+      "loss": 1.3604,
+      "step": 3909
+    },
+    {
+      "epoch": 3.6095084237249018,
+      "grad_norm": 4.34765625,
+      "learning_rate": 7.398265510340228e-05,
+      "loss": 1.5054,
+      "step": 3910
+    },
+    {
+      "epoch": 3.6104315716593582,
+      "grad_norm": 3.384765625,
+      "learning_rate": 7.397598398932623e-05,
+      "loss": 1.1063,
+      "step": 3911
+    },
+    {
+      "epoch": 3.611354719593815,
+      "grad_norm": 4.33203125,
+      "learning_rate": 7.396931287525017e-05,
+      "loss": 1.418,
+      "step": 3912
+    },
+    {
+      "epoch": 3.612277867528271,
+      "grad_norm": 3.880859375,
+      "learning_rate": 7.396264176117412e-05,
+      "loss": 1.2163,
+      "step": 3913
+    },
+    {
+      "epoch": 3.613201015462728,
+      "grad_norm": 5.05859375,
+      "learning_rate": 7.395597064709808e-05,
+      "loss": 1.6486,
+      "step": 3914
+    },
+    {
+      "epoch": 3.6141241633971846,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.394929953302202e-05,
+      "loss": 1.4875,
+      "step": 3915
+    },
+    {
+      "epoch": 3.615047311331641,
+      "grad_norm": 4.77734375,
+      "learning_rate": 7.394262841894597e-05,
+      "loss": 1.8193,
+      "step": 3916
+    },
+    {
+      "epoch": 3.6159704592660975,
+      "grad_norm": 4.01171875,
+      "learning_rate": 7.393595730486991e-05,
+      "loss": 1.5662,
+      "step": 3917
+    },
+    {
+      "epoch": 3.616893607200554,
+      "grad_norm": 3.927734375,
+      "learning_rate": 7.392928619079387e-05,
+      "loss": 1.3165,
+      "step": 3918
+    },
+    {
+      "epoch": 3.6178167551350104,
+      "grad_norm": 4.1796875,
+      "learning_rate": 7.392261507671782e-05,
+      "loss": 1.6978,
+      "step": 3919
+    },
+    {
+      "epoch": 3.618739903069467,
+      "grad_norm": 4.171875,
+      "learning_rate": 7.391594396264176e-05,
+      "loss": 1.4409,
+      "step": 3920
+    },
+    {
+      "epoch": 3.6196630510039234,
+      "grad_norm": 3.49609375,
+      "learning_rate": 7.390927284856571e-05,
+      "loss": 1.7041,
+      "step": 3921
+    },
+    {
+      "epoch": 3.62058619893838,
+      "grad_norm": 4.22265625,
+      "learning_rate": 7.390260173448967e-05,
+      "loss": 1.3435,
+      "step": 3922
+    },
+    {
+      "epoch": 3.6215093468728363,
+      "grad_norm": 3.6171875,
+      "learning_rate": 7.389593062041361e-05,
+      "loss": 1.3408,
+      "step": 3923
+    },
+    {
+      "epoch": 3.622432494807293,
+      "grad_norm": 3.98828125,
+      "learning_rate": 7.388925950633756e-05,
+      "loss": 1.4993,
+      "step": 3924
+    },
+    {
+      "epoch": 3.6233556427417493,
+      "grad_norm": 4.203125,
+      "learning_rate": 7.38825883922615e-05,
+      "loss": 1.8362,
+      "step": 3925
+    },
+    {
+      "epoch": 3.6242787906762057,
+      "grad_norm": 3.611328125,
+      "learning_rate": 7.387591727818546e-05,
+      "loss": 1.6057,
+      "step": 3926
+    },
+    {
+      "epoch": 3.625201938610662,
+      "grad_norm": 4.4296875,
+      "learning_rate": 7.386924616410941e-05,
+      "loss": 1.4331,
+      "step": 3927
+    },
+    {
+      "epoch": 3.6261250865451187,
+      "grad_norm": 5.24609375,
+      "learning_rate": 7.386257505003335e-05,
+      "loss": 1.4978,
+      "step": 3928
+    },
+    {
+      "epoch": 3.6270482344795756,
+      "grad_norm": 3.982421875,
+      "learning_rate": 7.38559039359573e-05,
+      "loss": 1.25,
+      "step": 3929
+    },
+    {
+      "epoch": 3.6279713824140316,
+      "grad_norm": 4.4375,
+      "learning_rate": 7.384923282188126e-05,
+      "loss": 1.5256,
+      "step": 3930
+    },
+    {
+      "epoch": 3.6288945303484885,
+      "grad_norm": 3.923828125,
+      "learning_rate": 7.38425617078052e-05,
+      "loss": 1.5503,
+      "step": 3931
+    },
+    {
+      "epoch": 3.629817678282945,
+      "grad_norm": 4.51953125,
+      "learning_rate": 7.383589059372915e-05,
+      "loss": 1.5098,
+      "step": 3932
+    },
+    {
+      "epoch": 3.6307408262174015,
+      "grad_norm": 3.9296875,
+      "learning_rate": 7.382921947965311e-05,
+      "loss": 1.5786,
+      "step": 3933
+    },
+    {
+      "epoch": 3.631663974151858,
+      "grad_norm": 4.8203125,
+      "learning_rate": 7.382254836557707e-05,
+      "loss": 1.3492,
+      "step": 3934
+    },
+    {
+      "epoch": 3.6325871220863144,
+      "grad_norm": 5.3046875,
+      "learning_rate": 7.381587725150101e-05,
+      "loss": 1.5618,
+      "step": 3935
+    },
+    {
+      "epoch": 3.633510270020771,
+      "grad_norm": 4.64453125,
+      "learning_rate": 7.380920613742496e-05,
+      "loss": 1.2167,
+      "step": 3936
+    },
+    {
+      "epoch": 3.6344334179552273,
+      "grad_norm": 4.8515625,
+      "learning_rate": 7.38025350233489e-05,
+      "loss": 1.3102,
+      "step": 3937
+    },
+    {
+      "epoch": 3.635356565889684,
+      "grad_norm": 3.892578125,
+      "learning_rate": 7.379586390927286e-05,
+      "loss": 1.2761,
+      "step": 3938
+    },
+    {
+      "epoch": 3.6362797138241403,
+      "grad_norm": 3.44140625,
+      "learning_rate": 7.378919279519681e-05,
+      "loss": 1.4121,
+      "step": 3939
+    },
+    {
+      "epoch": 3.6372028617585968,
+      "grad_norm": 3.94140625,
+      "learning_rate": 7.378252168112075e-05,
+      "loss": 1.3027,
+      "step": 3940
+    },
+    {
+      "epoch": 3.638126009693053,
+      "grad_norm": 5.7578125,
+      "learning_rate": 7.37758505670447e-05,
+      "loss": 1.6587,
+      "step": 3941
+    },
+    {
+      "epoch": 3.6390491576275097,
+      "grad_norm": 4.26171875,
+      "learning_rate": 7.376917945296866e-05,
+      "loss": 1.2522,
+      "step": 3942
+    },
+    {
+      "epoch": 3.639972305561966,
+      "grad_norm": 4.31640625,
+      "learning_rate": 7.37625083388926e-05,
+      "loss": 1.4983,
+      "step": 3943
+    },
+    {
+      "epoch": 3.640895453496423,
+      "grad_norm": 3.748046875,
+      "learning_rate": 7.375583722481655e-05,
+      "loss": 1.3207,
+      "step": 3944
+    },
+    {
+      "epoch": 3.641818601430879,
+      "grad_norm": 4.48046875,
+      "learning_rate": 7.37491661107405e-05,
+      "loss": 1.6343,
+      "step": 3945
+    },
+    {
+      "epoch": 3.642741749365336,
+      "grad_norm": 3.646484375,
+      "learning_rate": 7.374249499666445e-05,
+      "loss": 1.3533,
+      "step": 3946
+    },
+    {
+      "epoch": 3.643664897299792,
+      "grad_norm": 4.984375,
+      "learning_rate": 7.37358238825884e-05,
+      "loss": 1.3987,
+      "step": 3947
+    },
+    {
+      "epoch": 3.644588045234249,
+      "grad_norm": 3.97265625,
+      "learning_rate": 7.372915276851234e-05,
+      "loss": 1.3534,
+      "step": 3948
+    },
+    {
+      "epoch": 3.6455111931687054,
+      "grad_norm": 3.845703125,
+      "learning_rate": 7.372248165443629e-05,
+      "loss": 1.4253,
+      "step": 3949
+    },
+    {
+      "epoch": 3.646434341103162,
+      "grad_norm": 4.1796875,
+      "learning_rate": 7.371581054036025e-05,
+      "loss": 1.572,
+      "step": 3950
+    },
+    {
+      "epoch": 3.6473574890376184,
+      "grad_norm": 4.17578125,
+      "learning_rate": 7.370913942628419e-05,
+      "loss": 1.543,
+      "step": 3951
+    },
+    {
+      "epoch": 3.648280636972075,
+      "grad_norm": 3.68359375,
+      "learning_rate": 7.370246831220814e-05,
+      "loss": 1.5028,
+      "step": 3952
+    },
+    {
+      "epoch": 3.6492037849065313,
+      "grad_norm": 3.494140625,
+      "learning_rate": 7.369579719813208e-05,
+      "loss": 1.3499,
+      "step": 3953
+    },
+    {
+      "epoch": 3.6501269328409878,
+      "grad_norm": 4.515625,
+      "learning_rate": 7.368912608405604e-05,
+      "loss": 1.2842,
+      "step": 3954
+    },
+    {
+      "epoch": 3.6510500807754442,
+      "grad_norm": 5.640625,
+      "learning_rate": 7.368245496997999e-05,
+      "loss": 1.4507,
+      "step": 3955
+    },
+    {
+      "epoch": 3.6519732287099007,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.367578385590393e-05,
+      "loss": 1.6375,
+      "step": 3956
+    },
+    {
+      "epoch": 3.652896376644357,
+      "grad_norm": 3.7109375,
+      "learning_rate": 7.366911274182788e-05,
+      "loss": 1.4438,
+      "step": 3957
+    },
+    {
+      "epoch": 3.6538195245788136,
+      "grad_norm": 3.845703125,
+      "learning_rate": 7.366244162775184e-05,
+      "loss": 1.4524,
+      "step": 3958
+    },
+    {
+      "epoch": 3.65474267251327,
+      "grad_norm": 3.55078125,
+      "learning_rate": 7.36557705136758e-05,
+      "loss": 1.4648,
+      "step": 3959
+    },
+    {
+      "epoch": 3.6556658204477266,
+      "grad_norm": 3.779296875,
+      "learning_rate": 7.364909939959974e-05,
+      "loss": 1.7002,
+      "step": 3960
+    },
+    {
+      "epoch": 3.6565889683821835,
+      "grad_norm": 3.595703125,
+      "learning_rate": 7.364242828552369e-05,
+      "loss": 1.4407,
+      "step": 3961
+    },
+    {
+      "epoch": 3.6575121163166395,
+      "grad_norm": 4.234375,
+      "learning_rate": 7.363575717144763e-05,
+      "loss": 1.2601,
+      "step": 3962
+    },
+    {
+      "epoch": 3.6584352642510964,
+      "grad_norm": 5.0625,
+      "learning_rate": 7.362908605737159e-05,
+      "loss": 1.2557,
+      "step": 3963
+    },
+    {
+      "epoch": 3.6593584121855525,
+      "grad_norm": 4.0703125,
+      "learning_rate": 7.362241494329554e-05,
+      "loss": 1.5769,
+      "step": 3964
+    },
+    {
+      "epoch": 3.6602815601200094,
+      "grad_norm": 3.61328125,
+      "learning_rate": 7.361574382921948e-05,
+      "loss": 1.7761,
+      "step": 3965
+    },
+    {
+      "epoch": 3.661204708054466,
+      "grad_norm": 3.775390625,
+      "learning_rate": 7.360907271514343e-05,
+      "loss": 1.5942,
+      "step": 3966
+    },
+    {
+      "epoch": 3.6621278559889223,
+      "grad_norm": 3.82421875,
+      "learning_rate": 7.360240160106739e-05,
+      "loss": 1.4983,
+      "step": 3967
+    },
+    {
+      "epoch": 3.663051003923379,
+      "grad_norm": 4.78515625,
+      "learning_rate": 7.359573048699133e-05,
+      "loss": 1.5222,
+      "step": 3968
+    },
+    {
+      "epoch": 3.6639741518578353,
+      "grad_norm": 4.203125,
+      "learning_rate": 7.358905937291528e-05,
+      "loss": 1.2444,
+      "step": 3969
+    },
+    {
+      "epoch": 3.6648972997922917,
+      "grad_norm": 3.7890625,
+      "learning_rate": 7.358238825883922e-05,
+      "loss": 1.4177,
+      "step": 3970
+    },
+    {
+      "epoch": 3.665820447726748,
+      "grad_norm": 3.748046875,
+      "learning_rate": 7.357571714476318e-05,
+      "loss": 1.4382,
+      "step": 3971
+    },
+    {
+      "epoch": 3.6667435956612047,
+      "grad_norm": 3.509765625,
+      "learning_rate": 7.356904603068713e-05,
+      "loss": 1.45,
+      "step": 3972
+    },
+    {
+      "epoch": 3.667666743595661,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.356237491661107e-05,
+      "loss": 1.5,
+      "step": 3973
+    },
+    {
+      "epoch": 3.6685898915301176,
+      "grad_norm": 4.26171875,
+      "learning_rate": 7.355570380253502e-05,
+      "loss": 1.2715,
+      "step": 3974
+    },
+    {
+      "epoch": 3.669513039464574,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.354903268845898e-05,
+      "loss": 1.3,
+      "step": 3975
+    },
+    {
+      "epoch": 3.6704361873990305,
+      "grad_norm": 3.150390625,
+      "learning_rate": 7.354236157438292e-05,
+      "loss": 1.5732,
+      "step": 3976
+    },
+    {
+      "epoch": 3.671359335333487,
+      "grad_norm": 3.4296875,
+      "learning_rate": 7.353569046030687e-05,
+      "loss": 1.3693,
+      "step": 3977
+    },
+    {
+      "epoch": 3.672282483267944,
+      "grad_norm": 4.0234375,
+      "learning_rate": 7.352901934623081e-05,
+      "loss": 1.6545,
+      "step": 3978
+    },
+    {
+      "epoch": 3.6732056312024,
+      "grad_norm": 4.390625,
+      "learning_rate": 7.352234823215477e-05,
+      "loss": 1.7808,
+      "step": 3979
+    },
+    {
+      "epoch": 3.674128779136857,
+      "grad_norm": 3.7734375,
+      "learning_rate": 7.351567711807872e-05,
+      "loss": 1.2375,
+      "step": 3980
+    },
+    {
+      "epoch": 3.6750519270713133,
+      "grad_norm": 3.72265625,
+      "learning_rate": 7.350900600400266e-05,
+      "loss": 1.1613,
+      "step": 3981
+    },
+    {
+      "epoch": 3.67597507500577,
+      "grad_norm": 4.3671875,
+      "learning_rate": 7.350233488992661e-05,
+      "loss": 1.5042,
+      "step": 3982
+    },
+    {
+      "epoch": 3.6768982229402263,
+      "grad_norm": 5.2578125,
+      "learning_rate": 7.349566377585057e-05,
+      "loss": 1.267,
+      "step": 3983
+    },
+    {
+      "epoch": 3.6778213708746827,
+      "grad_norm": 4.4609375,
+      "learning_rate": 7.348899266177453e-05,
+      "loss": 1.4303,
+      "step": 3984
+    },
+    {
+      "epoch": 3.678744518809139,
+      "grad_norm": 4.7734375,
+      "learning_rate": 7.348232154769847e-05,
+      "loss": 1.6204,
+      "step": 3985
+    },
+    {
+      "epoch": 3.6796676667435957,
+      "grad_norm": 3.94140625,
+      "learning_rate": 7.347565043362242e-05,
+      "loss": 1.2659,
+      "step": 3986
+    },
+    {
+      "epoch": 3.680590814678052,
+      "grad_norm": 4.43359375,
+      "learning_rate": 7.346897931954638e-05,
+      "loss": 1.3877,
+      "step": 3987
+    },
+    {
+      "epoch": 3.6815139626125086,
+      "grad_norm": 3.48046875,
+      "learning_rate": 7.346230820547032e-05,
+      "loss": 1.3992,
+      "step": 3988
+    },
+    {
+      "epoch": 3.682437110546965,
+      "grad_norm": 3.755859375,
+      "learning_rate": 7.345563709139427e-05,
+      "loss": 1.5186,
+      "step": 3989
+    },
+    {
+      "epoch": 3.6833602584814216,
+      "grad_norm": 4.171875,
+      "learning_rate": 7.344896597731821e-05,
+      "loss": 1.4189,
+      "step": 3990
+    },
+    {
+      "epoch": 3.684283406415878,
+      "grad_norm": 3.9609375,
+      "learning_rate": 7.344229486324217e-05,
+      "loss": 1.3938,
+      "step": 3991
+    },
+    {
+      "epoch": 3.6852065543503345,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.343562374916612e-05,
+      "loss": 1.3658,
+      "step": 3992
+    },
+    {
+      "epoch": 3.686129702284791,
+      "grad_norm": 4.74609375,
+      "learning_rate": 7.342895263509006e-05,
+      "loss": 1.3696,
+      "step": 3993
+    },
+    {
+      "epoch": 3.6870528502192474,
+      "grad_norm": 5.38671875,
+      "learning_rate": 7.342228152101401e-05,
+      "loss": 1.2817,
+      "step": 3994
+    },
+    {
+      "epoch": 3.6879759981537044,
+      "grad_norm": 4.05078125,
+      "learning_rate": 7.341561040693797e-05,
+      "loss": 1.6108,
+      "step": 3995
+    },
+    {
+      "epoch": 3.6888991460881604,
+      "grad_norm": 4.08984375,
+      "learning_rate": 7.340893929286191e-05,
+      "loss": 1.2817,
+      "step": 3996
+    },
+    {
+      "epoch": 3.6898222940226173,
+      "grad_norm": 4.0859375,
+      "learning_rate": 7.340226817878586e-05,
+      "loss": 1.7747,
+      "step": 3997
+    },
+    {
+      "epoch": 3.6907454419570738,
+      "grad_norm": 3.908203125,
+      "learning_rate": 7.33955970647098e-05,
+      "loss": 1.6506,
+      "step": 3998
+    },
+    {
+      "epoch": 3.6916685898915302,
+      "grad_norm": 4.45703125,
+      "learning_rate": 7.338892595063376e-05,
+      "loss": 1.4958,
+      "step": 3999
+    },
+    {
+      "epoch": 3.6925917378259867,
+      "grad_norm": 4.9921875,
+      "learning_rate": 7.338225483655771e-05,
+      "loss": 1.2798,
+      "step": 4000
+    },
+    {
+      "epoch": 3.6925917378259867,
+      "eval_loss": 0.88232421875,
+      "eval_runtime": 23.3639,
+      "eval_samples_per_second": 975.906,
+      "eval_steps_per_second": 61.034,
+      "step": 4000
+    },
+    {
+      "epoch": 3.693514885760443,
+      "grad_norm": 4.66796875,
+      "learning_rate": 7.337558372248165e-05,
+      "loss": 1.3423,
+      "step": 4001
+    },
+    {
+      "epoch": 3.6944380336948996,
+      "grad_norm": 5.92578125,
+      "learning_rate": 7.33689126084056e-05,
+      "loss": 1.2373,
+      "step": 4002
+    },
+    {
+      "epoch": 3.695361181629356,
+      "grad_norm": 5.06640625,
+      "learning_rate": 7.336224149432956e-05,
+      "loss": 1.6565,
+      "step": 4003
+    },
+    {
+      "epoch": 3.6962843295638126,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.33555703802535e-05,
+      "loss": 1.5825,
+      "step": 4004
+    },
+    {
+      "epoch": 3.697207477498269,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.334889926617745e-05,
+      "loss": 1.5171,
+      "step": 4005
+    },
+    {
+      "epoch": 3.6981306254327255,
+      "grad_norm": 4.39453125,
+      "learning_rate": 7.33422281521014e-05,
+      "loss": 1.3992,
+      "step": 4006
+    },
+    {
+      "epoch": 3.699053773367182,
+      "grad_norm": 4.10546875,
+      "learning_rate": 7.333555703802535e-05,
+      "loss": 1.4028,
+      "step": 4007
+    },
+    {
+      "epoch": 3.6999769213016385,
+      "grad_norm": 3.95703125,
+      "learning_rate": 7.33288859239493e-05,
+      "loss": 1.3567,
+      "step": 4008
+    },
+    {
+      "epoch": 3.700900069236095,
+      "grad_norm": 3.42578125,
+      "learning_rate": 7.332221480987326e-05,
+      "loss": 1.3892,
+      "step": 4009
+    },
+    {
+      "epoch": 3.701823217170552,
+      "grad_norm": 4.00390625,
+      "learning_rate": 7.33155436957972e-05,
+      "loss": 1.3942,
+      "step": 4010
+    },
+    {
+      "epoch": 3.702746365105008,
+      "grad_norm": 4.16796875,
+      "learning_rate": 7.330887258172115e-05,
+      "loss": 1.1748,
+      "step": 4011
+    },
+    {
+      "epoch": 3.703669513039465,
+      "grad_norm": 4.64453125,
+      "learning_rate": 7.330220146764511e-05,
+      "loss": 1.6213,
+      "step": 4012
+    },
+    {
+      "epoch": 3.704592660973921,
+      "grad_norm": 3.966796875,
+      "learning_rate": 7.329553035356905e-05,
+      "loss": 1.3823,
+      "step": 4013
+    },
+    {
+      "epoch": 3.7055158089083777,
+      "grad_norm": 4.359375,
+      "learning_rate": 7.3288859239493e-05,
+      "loss": 1.415,
+      "step": 4014
+    },
+    {
+      "epoch": 3.706438956842834,
+      "grad_norm": 4.49609375,
+      "learning_rate": 7.328218812541694e-05,
+      "loss": 1.5764,
+      "step": 4015
+    },
+    {
+      "epoch": 3.7073621047772907,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.32755170113409e-05,
+      "loss": 0.9901,
+      "step": 4016
+    },
+    {
+      "epoch": 3.708285252711747,
+      "grad_norm": 5.0546875,
+      "learning_rate": 7.326884589726485e-05,
+      "loss": 1.4832,
+      "step": 4017
+    },
+    {
+      "epoch": 3.7092084006462036,
+      "grad_norm": 3.404296875,
+      "learning_rate": 7.32621747831888e-05,
+      "loss": 1.2135,
+      "step": 4018
+    },
+    {
+      "epoch": 3.71013154858066,
+      "grad_norm": 4.12890625,
+      "learning_rate": 7.325550366911274e-05,
+      "loss": 1.439,
+      "step": 4019
+    },
+    {
+      "epoch": 3.7110546965151165,
+      "grad_norm": 3.4765625,
+      "learning_rate": 7.32488325550367e-05,
+      "loss": 1.6277,
+      "step": 4020
+    },
+    {
+      "epoch": 3.711977844449573,
+      "grad_norm": 3.7890625,
+      "learning_rate": 7.324216144096064e-05,
+      "loss": 1.3953,
+      "step": 4021
+    },
+    {
+      "epoch": 3.7129009923840295,
+      "grad_norm": 3.611328125,
+      "learning_rate": 7.323549032688459e-05,
+      "loss": 1.3324,
+      "step": 4022
+    },
+    {
+      "epoch": 3.713824140318486,
+      "grad_norm": 4.5546875,
+      "learning_rate": 7.322881921280854e-05,
+      "loss": 1.5159,
+      "step": 4023
+    },
+    {
+      "epoch": 3.7147472882529424,
+      "grad_norm": 4.97265625,
+      "learning_rate": 7.32221480987325e-05,
+      "loss": 1.3381,
+      "step": 4024
+    },
+    {
+      "epoch": 3.715670436187399,
+      "grad_norm": 4.1015625,
+      "learning_rate": 7.321547698465644e-05,
+      "loss": 1.2114,
+      "step": 4025
+    },
+    {
+      "epoch": 3.7165935841218554,
+      "grad_norm": 4.98828125,
+      "learning_rate": 7.320880587058039e-05,
+      "loss": 1.2202,
+      "step": 4026
+    },
+    {
+      "epoch": 3.7175167320563123,
+      "grad_norm": 4.734375,
+      "learning_rate": 7.320213475650433e-05,
+      "loss": 1.3936,
+      "step": 4027
+    },
+    {
+      "epoch": 3.7184398799907683,
+      "grad_norm": 4.61328125,
+      "learning_rate": 7.319546364242829e-05,
+      "loss": 1.6724,
+      "step": 4028
+    },
+    {
+      "epoch": 3.719363027925225,
+      "grad_norm": 3.91015625,
+      "learning_rate": 7.318879252835223e-05,
+      "loss": 1.4976,
+      "step": 4029
+    },
+    {
+      "epoch": 3.7202861758596812,
+      "grad_norm": 4.49609375,
+      "learning_rate": 7.318212141427618e-05,
+      "loss": 1.5776,
+      "step": 4030
+    },
+    {
+      "epoch": 3.721209323794138,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.317545030020013e-05,
+      "loss": 1.4608,
+      "step": 4031
+    },
+    {
+      "epoch": 3.7221324717285946,
+      "grad_norm": 3.966796875,
+      "learning_rate": 7.316877918612408e-05,
+      "loss": 1.4872,
+      "step": 4032
+    },
+    {
+      "epoch": 3.723055619663051,
+      "grad_norm": 3.783203125,
+      "learning_rate": 7.316210807204803e-05,
+      "loss": 1.1968,
+      "step": 4033
+    },
+    {
+      "epoch": 3.7239787675975076,
+      "grad_norm": 4.125,
+      "learning_rate": 7.315543695797199e-05,
+      "loss": 1.5959,
+      "step": 4034
+    },
+    {
+      "epoch": 3.724901915531964,
+      "grad_norm": 3.724609375,
+      "learning_rate": 7.314876584389593e-05,
+      "loss": 1.5881,
+      "step": 4035
+    },
+    {
+      "epoch": 3.7258250634664205,
+      "grad_norm": 3.619140625,
+      "learning_rate": 7.31420947298199e-05,
+      "loss": 1.4331,
+      "step": 4036
+    },
+    {
+      "epoch": 3.726748211400877,
+      "grad_norm": 4.48046875,
+      "learning_rate": 7.313542361574384e-05,
+      "loss": 1.448,
+      "step": 4037
+    },
+    {
+      "epoch": 3.7276713593353334,
+      "grad_norm": 3.650390625,
+      "learning_rate": 7.312875250166778e-05,
+      "loss": 1.2615,
+      "step": 4038
+    },
+    {
+      "epoch": 3.72859450726979,
+      "grad_norm": 4.55078125,
+      "learning_rate": 7.312208138759173e-05,
+      "loss": 1.6509,
+      "step": 4039
+    },
+    {
+      "epoch": 3.7295176552042464,
+      "grad_norm": 3.880859375,
+      "learning_rate": 7.311541027351569e-05,
+      "loss": 1.646,
+      "step": 4040
+    },
+    {
+      "epoch": 3.730440803138703,
+      "grad_norm": 3.232421875,
+      "learning_rate": 7.310873915943963e-05,
+      "loss": 1.199,
+      "step": 4041
+    },
+    {
+      "epoch": 3.7313639510731593,
+      "grad_norm": 3.72265625,
+      "learning_rate": 7.310206804536358e-05,
+      "loss": 1.499,
+      "step": 4042
+    },
+    {
+      "epoch": 3.732287099007616,
+      "grad_norm": 3.2734375,
+      "learning_rate": 7.309539693128753e-05,
+      "loss": 1.5752,
+      "step": 4043
+    },
+    {
+      "epoch": 3.7332102469420727,
+      "grad_norm": 4.5234375,
+      "learning_rate": 7.308872581721148e-05,
+      "loss": 1.5745,
+      "step": 4044
+    },
+    {
+      "epoch": 3.7341333948765287,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.308205470313543e-05,
+      "loss": 1.4673,
+      "step": 4045
+    },
+    {
+      "epoch": 3.7350565428109856,
+      "grad_norm": 4.98046875,
+      "learning_rate": 7.307538358905938e-05,
+      "loss": 1.4631,
+      "step": 4046
+    },
+    {
+      "epoch": 3.735979690745442,
+      "grad_norm": 4.77734375,
+      "learning_rate": 7.306871247498332e-05,
+      "loss": 1.3909,
+      "step": 4047
+    },
+    {
+      "epoch": 3.7369028386798986,
+      "grad_norm": 3.8671875,
+      "learning_rate": 7.306204136090728e-05,
+      "loss": 1.192,
+      "step": 4048
+    },
+    {
+      "epoch": 3.737825986614355,
+      "grad_norm": 3.91015625,
+      "learning_rate": 7.305537024683122e-05,
+      "loss": 1.3164,
+      "step": 4049
+    },
+    {
+      "epoch": 3.7387491345488115,
+      "grad_norm": 3.396484375,
+      "learning_rate": 7.304869913275517e-05,
+      "loss": 1.3823,
+      "step": 4050
+    },
+    {
+      "epoch": 3.739672282483268,
+      "grad_norm": 3.822265625,
+      "learning_rate": 7.304202801867912e-05,
+      "loss": 1.1328,
+      "step": 4051
+    },
+    {
+      "epoch": 3.7405954304177245,
+      "grad_norm": 4.109375,
+      "learning_rate": 7.303535690460307e-05,
+      "loss": 1.4344,
+      "step": 4052
+    },
+    {
+      "epoch": 3.741518578352181,
+      "grad_norm": 4.62109375,
+      "learning_rate": 7.302868579052702e-05,
+      "loss": 1.6299,
+      "step": 4053
+    },
+    {
+      "epoch": 3.7424417262866374,
+      "grad_norm": 4.58984375,
+      "learning_rate": 7.302201467645097e-05,
+      "loss": 1.3594,
+      "step": 4054
+    },
+    {
+      "epoch": 3.743364874221094,
+      "grad_norm": 4.46484375,
+      "learning_rate": 7.301534356237491e-05,
+      "loss": 1.5627,
+      "step": 4055
+    },
+    {
+      "epoch": 3.7442880221555503,
+      "grad_norm": 3.7890625,
+      "learning_rate": 7.300867244829887e-05,
+      "loss": 1.6167,
+      "step": 4056
+    },
+    {
+      "epoch": 3.745211170090007,
+      "grad_norm": 4.15234375,
+      "learning_rate": 7.300200133422282e-05,
+      "loss": 1.2009,
+      "step": 4057
+    },
+    {
+      "epoch": 3.7461343180244633,
+      "grad_norm": 3.880859375,
+      "learning_rate": 7.299533022014676e-05,
+      "loss": 1.252,
+      "step": 4058
+    },
+    {
+      "epoch": 3.7470574659589198,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.298865910607072e-05,
+      "loss": 1.2222,
+      "step": 4059
+    },
+    {
+      "epoch": 3.7479806138933762,
+      "grad_norm": 4.1328125,
+      "learning_rate": 7.298198799199467e-05,
+      "loss": 1.72,
+      "step": 4060
+    },
+    {
+      "epoch": 3.748903761827833,
+      "grad_norm": 4.94140625,
+      "learning_rate": 7.297531687791862e-05,
+      "loss": 1.3345,
+      "step": 4061
+    },
+    {
+      "epoch": 3.749826909762289,
+      "grad_norm": 3.17578125,
+      "learning_rate": 7.296864576384257e-05,
+      "loss": 1.1092,
+      "step": 4062
+    },
+    {
+      "epoch": 3.750750057696746,
+      "grad_norm": 3.5390625,
+      "learning_rate": 7.296197464976652e-05,
+      "loss": 1.1783,
+      "step": 4063
+    },
+    {
+      "epoch": 3.7516732056312025,
+      "grad_norm": 4.05859375,
+      "learning_rate": 7.295530353569046e-05,
+      "loss": 1.5542,
+      "step": 4064
+    },
+    {
+      "epoch": 3.752596353565659,
+      "grad_norm": 4.48828125,
+      "learning_rate": 7.294863242161442e-05,
+      "loss": 1.4077,
+      "step": 4065
+    },
+    {
+      "epoch": 3.7535195015001155,
+      "grad_norm": 3.470703125,
+      "learning_rate": 7.294196130753836e-05,
+      "loss": 1.4849,
+      "step": 4066
+    },
+    {
+      "epoch": 3.754442649434572,
+      "grad_norm": 3.875,
+      "learning_rate": 7.293529019346231e-05,
+      "loss": 1.2485,
+      "step": 4067
+    },
+    {
+      "epoch": 3.7553657973690284,
+      "grad_norm": 3.646484375,
+      "learning_rate": 7.292861907938626e-05,
+      "loss": 1.3733,
+      "step": 4068
+    },
+    {
+      "epoch": 3.756288945303485,
+      "grad_norm": 3.89453125,
+      "learning_rate": 7.292194796531021e-05,
+      "loss": 1.3939,
+      "step": 4069
+    },
+    {
+      "epoch": 3.7572120932379414,
+      "grad_norm": 4.265625,
+      "learning_rate": 7.291527685123416e-05,
+      "loss": 1.8645,
+      "step": 4070
+    },
+    {
+      "epoch": 3.758135241172398,
+      "grad_norm": 3.697265625,
+      "learning_rate": 7.29086057371581e-05,
+      "loss": 1.3572,
+      "step": 4071
+    },
+    {
+      "epoch": 3.7590583891068543,
+      "grad_norm": 4.13671875,
+      "learning_rate": 7.290193462308205e-05,
+      "loss": 1.5,
+      "step": 4072
+    },
+    {
+      "epoch": 3.7599815370413108,
+      "grad_norm": 4.13671875,
+      "learning_rate": 7.289526350900601e-05,
+      "loss": 1.498,
+      "step": 4073
+    },
+    {
+      "epoch": 3.7609046849757672,
+      "grad_norm": 4.98046875,
+      "learning_rate": 7.288859239492996e-05,
+      "loss": 1.364,
+      "step": 4074
+    },
+    {
+      "epoch": 3.7618278329102237,
+      "grad_norm": 4.84765625,
+      "learning_rate": 7.28819212808539e-05,
+      "loss": 1.5002,
+      "step": 4075
+    },
+    {
+      "epoch": 3.7627509808446806,
+      "grad_norm": 4.26171875,
+      "learning_rate": 7.287525016677785e-05,
+      "loss": 1.5759,
+      "step": 4076
+    },
+    {
+      "epoch": 3.7636741287791367,
+      "grad_norm": 3.939453125,
+      "learning_rate": 7.28685790527018e-05,
+      "loss": 1.4725,
+      "step": 4077
+    },
+    {
+      "epoch": 3.7645972767135936,
+      "grad_norm": 3.947265625,
+      "learning_rate": 7.286190793862575e-05,
+      "loss": 1.1888,
+      "step": 4078
+    },
+    {
+      "epoch": 3.7655204246480496,
+      "grad_norm": 3.853515625,
+      "learning_rate": 7.28552368245497e-05,
+      "loss": 1.5991,
+      "step": 4079
+    },
+    {
+      "epoch": 3.7664435725825065,
+      "grad_norm": 4.0078125,
+      "learning_rate": 7.284856571047366e-05,
+      "loss": 1.3297,
+      "step": 4080
+    },
+    {
+      "epoch": 3.767366720516963,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.28418945963976e-05,
+      "loss": 1.2627,
+      "step": 4081
+    },
+    {
+      "epoch": 3.7682898684514194,
+      "grad_norm": 3.193359375,
+      "learning_rate": 7.283522348232155e-05,
+      "loss": 1.7532,
+      "step": 4082
+    },
+    {
+      "epoch": 3.769213016385876,
+      "grad_norm": 3.373046875,
+      "learning_rate": 7.282855236824549e-05,
+      "loss": 1.4016,
+      "step": 4083
+    },
+    {
+      "epoch": 3.7701361643203324,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.282188125416945e-05,
+      "loss": 1.4204,
+      "step": 4084
+    },
+    {
+      "epoch": 3.771059312254789,
+      "grad_norm": 3.990234375,
+      "learning_rate": 7.281521014009341e-05,
+      "loss": 1.6021,
+      "step": 4085
+    },
+    {
+      "epoch": 3.7719824601892453,
+      "grad_norm": 4.6015625,
+      "learning_rate": 7.280853902601735e-05,
+      "loss": 1.5037,
+      "step": 4086
+    },
+    {
+      "epoch": 3.772905608123702,
+      "grad_norm": 5.07421875,
+      "learning_rate": 7.28018679119413e-05,
+      "loss": 1.2238,
+      "step": 4087
+    },
+    {
+      "epoch": 3.7738287560581583,
+      "grad_norm": 4.89453125,
+      "learning_rate": 7.279519679786525e-05,
+      "loss": 1.5217,
+      "step": 4088
+    },
+    {
+      "epoch": 3.7747519039926147,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.27885256837892e-05,
+      "loss": 1.2712,
+      "step": 4089
+    },
+    {
+      "epoch": 3.775675051927071,
+      "grad_norm": 4.75,
+      "learning_rate": 7.278185456971315e-05,
+      "loss": 1.245,
+      "step": 4090
+    },
+    {
+      "epoch": 3.7765981998615277,
+      "grad_norm": 4.046875,
+      "learning_rate": 7.27751834556371e-05,
+      "loss": 1.3823,
+      "step": 4091
+    },
+    {
+      "epoch": 3.777521347795984,
+      "grad_norm": 4.15234375,
+      "learning_rate": 7.276851234156104e-05,
+      "loss": 1.4966,
+      "step": 4092
+    },
+    {
+      "epoch": 3.778444495730441,
+      "grad_norm": 3.720703125,
+      "learning_rate": 7.2761841227485e-05,
+      "loss": 1.5398,
+      "step": 4093
+    },
+    {
+      "epoch": 3.779367643664897,
+      "grad_norm": 3.9765625,
+      "learning_rate": 7.275517011340895e-05,
+      "loss": 1.2153,
+      "step": 4094
+    },
+    {
+      "epoch": 3.780290791599354,
+      "grad_norm": 4.0078125,
+      "learning_rate": 7.274849899933289e-05,
+      "loss": 1.3674,
+      "step": 4095
+    },
+    {
+      "epoch": 3.78121393953381,
+      "grad_norm": 3.482421875,
+      "learning_rate": 7.274182788525684e-05,
+      "loss": 1.2854,
+      "step": 4096
+    },
+    {
+      "epoch": 3.782137087468267,
+      "grad_norm": 3.603515625,
+      "learning_rate": 7.27351567711808e-05,
+      "loss": 1.4524,
+      "step": 4097
+    },
+    {
+      "epoch": 3.7830602354027234,
+      "grad_norm": 4.171875,
+      "learning_rate": 7.272848565710474e-05,
+      "loss": 1.3523,
+      "step": 4098
+    },
+    {
+      "epoch": 3.78398338333718,
+      "grad_norm": 3.751953125,
+      "learning_rate": 7.272181454302869e-05,
+      "loss": 1.2578,
+      "step": 4099
+    },
+    {
+      "epoch": 3.7849065312716363,
+      "grad_norm": 3.72265625,
+      "learning_rate": 7.271514342895263e-05,
+      "loss": 1.2297,
+      "step": 4100
+    },
+    {
+      "epoch": 3.785829679206093,
+      "grad_norm": 3.173828125,
+      "learning_rate": 7.270847231487659e-05,
+      "loss": 1.0003,
+      "step": 4101
+    },
+    {
+      "epoch": 3.7867528271405493,
+      "grad_norm": 3.5078125,
+      "learning_rate": 7.270180120080054e-05,
+      "loss": 1.5105,
+      "step": 4102
+    },
+    {
+      "epoch": 3.7876759750750058,
+      "grad_norm": 3.63671875,
+      "learning_rate": 7.269513008672448e-05,
+      "loss": 1.5125,
+      "step": 4103
+    },
+    {
+      "epoch": 3.788599123009462,
+      "grad_norm": 4.26171875,
+      "learning_rate": 7.268845897264843e-05,
+      "loss": 1.3198,
+      "step": 4104
+    },
+    {
+      "epoch": 3.7895222709439187,
+      "grad_norm": 4.1953125,
+      "learning_rate": 7.268178785857239e-05,
+      "loss": 1.7751,
+      "step": 4105
+    },
+    {
+      "epoch": 3.790445418878375,
+      "grad_norm": 3.708984375,
+      "learning_rate": 7.267511674449633e-05,
+      "loss": 1.2925,
+      "step": 4106
+    },
+    {
+      "epoch": 3.7913685668128316,
+      "grad_norm": 4.4296875,
+      "learning_rate": 7.266844563042028e-05,
+      "loss": 1.575,
+      "step": 4107
+    },
+    {
+      "epoch": 3.792291714747288,
+      "grad_norm": 3.583984375,
+      "learning_rate": 7.266177451634422e-05,
+      "loss": 1.45,
+      "step": 4108
+    },
+    {
+      "epoch": 3.7932148626817446,
+      "grad_norm": 3.986328125,
+      "learning_rate": 7.265510340226818e-05,
+      "loss": 1.2972,
+      "step": 4109
+    },
+    {
+      "epoch": 3.7941380106162015,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.264843228819214e-05,
+      "loss": 1.1552,
+      "step": 4110
+    },
+    {
+      "epoch": 3.7950611585506575,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.264176117411609e-05,
+      "loss": 1.0936,
+      "step": 4111
+    },
+    {
+      "epoch": 3.7959843064851144,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.263509006004003e-05,
+      "loss": 1.2024,
+      "step": 4112
+    },
+    {
+      "epoch": 3.796907454419571,
+      "grad_norm": 4.21875,
+      "learning_rate": 7.262841894596398e-05,
+      "loss": 1.6282,
+      "step": 4113
+    },
+    {
+      "epoch": 3.7978306023540274,
+      "grad_norm": 3.716796875,
+      "learning_rate": 7.262174783188794e-05,
+      "loss": 0.9522,
+      "step": 4114
+    },
+    {
+      "epoch": 3.798753750288484,
+      "grad_norm": 4.1328125,
+      "learning_rate": 7.261507671781188e-05,
+      "loss": 1.5251,
+      "step": 4115
+    },
+    {
+      "epoch": 3.7996768982229403,
+      "grad_norm": 4.78515625,
+      "learning_rate": 7.260840560373583e-05,
+      "loss": 1.3909,
+      "step": 4116
+    },
+    {
+      "epoch": 3.8006000461573968,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.260173448965979e-05,
+      "loss": 1.3342,
+      "step": 4117
+    },
+    {
+      "epoch": 3.8015231940918532,
+      "grad_norm": 5.578125,
+      "learning_rate": 7.259506337558373e-05,
+      "loss": 1.3813,
+      "step": 4118
+    },
+    {
+      "epoch": 3.8024463420263097,
+      "grad_norm": 4.33984375,
+      "learning_rate": 7.258839226150768e-05,
+      "loss": 1.4996,
+      "step": 4119
+    },
+    {
+      "epoch": 3.803369489960766,
+      "grad_norm": 4.0859375,
+      "learning_rate": 7.258172114743162e-05,
+      "loss": 1.3525,
+      "step": 4120
+    },
+    {
+      "epoch": 3.8042926378952227,
+      "grad_norm": 3.77734375,
+      "learning_rate": 7.257505003335558e-05,
+      "loss": 1.2949,
+      "step": 4121
+    },
+    {
+      "epoch": 3.805215785829679,
+      "grad_norm": 3.771484375,
+      "learning_rate": 7.256837891927953e-05,
+      "loss": 1.4915,
+      "step": 4122
+    },
+    {
+      "epoch": 3.8061389337641356,
+      "grad_norm": 3.541015625,
+      "learning_rate": 7.256170780520347e-05,
+      "loss": 1.3882,
+      "step": 4123
+    },
+    {
+      "epoch": 3.807062081698592,
+      "grad_norm": 4.37109375,
+      "learning_rate": 7.255503669112742e-05,
+      "loss": 1.4707,
+      "step": 4124
+    },
+    {
+      "epoch": 3.807985229633049,
+      "grad_norm": 3.876953125,
+      "learning_rate": 7.254836557705138e-05,
+      "loss": 1.548,
+      "step": 4125
+    },
+    {
+      "epoch": 3.808908377567505,
+      "grad_norm": 4.1484375,
+      "learning_rate": 7.254169446297532e-05,
+      "loss": 1.3762,
+      "step": 4126
+    },
+    {
+      "epoch": 3.809831525501962,
+      "grad_norm": 3.689453125,
+      "learning_rate": 7.253502334889927e-05,
+      "loss": 1.0136,
+      "step": 4127
+    },
+    {
+      "epoch": 3.810754673436418,
+      "grad_norm": 4.765625,
+      "learning_rate": 7.252835223482321e-05,
+      "loss": 1.3309,
+      "step": 4128
+    },
+    {
+      "epoch": 3.811677821370875,
+      "grad_norm": 4.27734375,
+      "learning_rate": 7.252168112074717e-05,
+      "loss": 1.4622,
+      "step": 4129
+    },
+    {
+      "epoch": 3.8126009693053313,
+      "grad_norm": 3.5546875,
+      "learning_rate": 7.251501000667112e-05,
+      "loss": 1.7766,
+      "step": 4130
+    },
+    {
+      "epoch": 3.813524117239788,
+      "grad_norm": 4.1640625,
+      "learning_rate": 7.250833889259506e-05,
+      "loss": 1.5862,
+      "step": 4131
+    },
+    {
+      "epoch": 3.8144472651742443,
+      "grad_norm": 3.98828125,
+      "learning_rate": 7.250166777851901e-05,
+      "loss": 1.4922,
+      "step": 4132
+    },
+    {
+      "epoch": 3.8153704131087007,
+      "grad_norm": 3.94140625,
+      "learning_rate": 7.249499666444297e-05,
+      "loss": 1.1907,
+      "step": 4133
+    },
+    {
+      "epoch": 3.816293561043157,
+      "grad_norm": 4.625,
+      "learning_rate": 7.248832555036693e-05,
+      "loss": 1.4216,
+      "step": 4134
+    },
+    {
+      "epoch": 3.8172167089776137,
+      "grad_norm": 3.744140625,
+      "learning_rate": 7.248165443629087e-05,
+      "loss": 1.4292,
+      "step": 4135
+    },
+    {
+      "epoch": 3.81813985691207,
+      "grad_norm": 4.109375,
+      "learning_rate": 7.247498332221482e-05,
+      "loss": 1.4204,
+      "step": 4136
+    },
+    {
+      "epoch": 3.8190630048465266,
+      "grad_norm": 4.41015625,
+      "learning_rate": 7.246831220813876e-05,
+      "loss": 1.7747,
+      "step": 4137
+    },
+    {
+      "epoch": 3.819986152780983,
+      "grad_norm": 4.265625,
+      "learning_rate": 7.246164109406272e-05,
+      "loss": 1.3657,
+      "step": 4138
+    },
+    {
+      "epoch": 3.8209093007154395,
+      "grad_norm": 3.3984375,
+      "learning_rate": 7.245496997998667e-05,
+      "loss": 1.3066,
+      "step": 4139
+    },
+    {
+      "epoch": 3.821832448649896,
+      "grad_norm": 3.919921875,
+      "learning_rate": 7.244829886591061e-05,
+      "loss": 1.2891,
+      "step": 4140
+    },
+    {
+      "epoch": 3.8227555965843525,
+      "grad_norm": 3.490234375,
+      "learning_rate": 7.244162775183456e-05,
+      "loss": 1.4407,
+      "step": 4141
+    },
+    {
+      "epoch": 3.8236787445188094,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.243495663775852e-05,
+      "loss": 1.489,
+      "step": 4142
+    },
+    {
+      "epoch": 3.8246018924532654,
+      "grad_norm": 3.4765625,
+      "learning_rate": 7.242828552368246e-05,
+      "loss": 1.4265,
+      "step": 4143
+    },
+    {
+      "epoch": 3.8255250403877223,
+      "grad_norm": 3.62890625,
+      "learning_rate": 7.24216144096064e-05,
+      "loss": 1.4825,
+      "step": 4144
+    },
+    {
+      "epoch": 3.8264481883221784,
+      "grad_norm": 3.31640625,
+      "learning_rate": 7.241494329553035e-05,
+      "loss": 1.4204,
+      "step": 4145
+    },
+    {
+      "epoch": 3.8273713362566353,
+      "grad_norm": 3.77734375,
+      "learning_rate": 7.240827218145431e-05,
+      "loss": 1.5735,
+      "step": 4146
+    },
+    {
+      "epoch": 3.8282944841910918,
+      "grad_norm": 4.84375,
+      "learning_rate": 7.240160106737826e-05,
+      "loss": 1.2878,
+      "step": 4147
+    },
+    {
+      "epoch": 3.829217632125548,
+      "grad_norm": 4.85546875,
+      "learning_rate": 7.23949299533022e-05,
+      "loss": 1.0472,
+      "step": 4148
+    },
+    {
+      "epoch": 3.8301407800600047,
+      "grad_norm": 5.3359375,
+      "learning_rate": 7.238825883922615e-05,
+      "loss": 1.3346,
+      "step": 4149
+    },
+    {
+      "epoch": 3.831063927994461,
+      "grad_norm": 4.09375,
+      "learning_rate": 7.23815877251501e-05,
+      "loss": 1.6179,
+      "step": 4150
+    },
+    {
+      "epoch": 3.8319870759289176,
+      "grad_norm": 4.61328125,
+      "learning_rate": 7.237491661107405e-05,
+      "loss": 1.6313,
+      "step": 4151
+    },
+    {
+      "epoch": 3.832910223863374,
+      "grad_norm": 4.796875,
+      "learning_rate": 7.2368245496998e-05,
+      "loss": 1.2767,
+      "step": 4152
+    },
+    {
+      "epoch": 3.8338333717978306,
+      "grad_norm": 3.908203125,
+      "learning_rate": 7.236157438292194e-05,
+      "loss": 1.4883,
+      "step": 4153
+    },
+    {
+      "epoch": 3.834756519732287,
+      "grad_norm": 4.83984375,
+      "learning_rate": 7.23549032688459e-05,
+      "loss": 1.4517,
+      "step": 4154
+    },
+    {
+      "epoch": 3.8356796676667435,
+      "grad_norm": 3.970703125,
+      "learning_rate": 7.234823215476985e-05,
+      "loss": 1.5068,
+      "step": 4155
+    },
+    {
+      "epoch": 3.8366028156012,
+      "grad_norm": 4.109375,
+      "learning_rate": 7.234156104069379e-05,
+      "loss": 1.7576,
+      "step": 4156
+    },
+    {
+      "epoch": 3.8375259635356564,
+      "grad_norm": 3.740234375,
+      "learning_rate": 7.233488992661774e-05,
+      "loss": 1.3386,
+      "step": 4157
+    },
+    {
+      "epoch": 3.838449111470113,
+      "grad_norm": 4.61328125,
+      "learning_rate": 7.23282188125417e-05,
+      "loss": 1.2059,
+      "step": 4158
+    },
+    {
+      "epoch": 3.83937225940457,
+      "grad_norm": 3.794921875,
+      "learning_rate": 7.232154769846566e-05,
+      "loss": 1.4336,
+      "step": 4159
+    },
+    {
+      "epoch": 3.840295407339026,
+      "grad_norm": 3.857421875,
+      "learning_rate": 7.23148765843896e-05,
+      "loss": 1.3174,
+      "step": 4160
+    },
+    {
+      "epoch": 3.8412185552734828,
+      "grad_norm": 4.58984375,
+      "learning_rate": 7.230820547031355e-05,
+      "loss": 1.6394,
+      "step": 4161
+    },
+    {
+      "epoch": 3.842141703207939,
+      "grad_norm": 3.865234375,
+      "learning_rate": 7.23015343562375e-05,
+      "loss": 1.5315,
+      "step": 4162
+    },
+    {
+      "epoch": 3.8430648511423957,
+      "grad_norm": 4.83203125,
+      "learning_rate": 7.229486324216145e-05,
+      "loss": 1.364,
+      "step": 4163
+    },
+    {
+      "epoch": 3.843987999076852,
+      "grad_norm": 5.0625,
+      "learning_rate": 7.22881921280854e-05,
+      "loss": 1.2369,
+      "step": 4164
+    },
+    {
+      "epoch": 3.8449111470113086,
+      "grad_norm": 4.8984375,
+      "learning_rate": 7.228152101400934e-05,
+      "loss": 1.2029,
+      "step": 4165
+    },
+    {
+      "epoch": 3.845834294945765,
+      "grad_norm": 4.50390625,
+      "learning_rate": 7.22748498999333e-05,
+      "loss": 1.4763,
+      "step": 4166
+    },
+    {
+      "epoch": 3.8467574428802216,
+      "grad_norm": 4.3359375,
+      "learning_rate": 7.226817878585725e-05,
+      "loss": 1.3574,
+      "step": 4167
+    },
+    {
+      "epoch": 3.847680590814678,
+      "grad_norm": 3.66015625,
+      "learning_rate": 7.226150767178119e-05,
+      "loss": 1.3538,
+      "step": 4168
+    },
+    {
+      "epoch": 3.8486037387491345,
+      "grad_norm": 3.5859375,
+      "learning_rate": 7.225483655770514e-05,
+      "loss": 1.2339,
+      "step": 4169
+    },
+    {
+      "epoch": 3.849526886683591,
+      "grad_norm": 4.734375,
+      "learning_rate": 7.22481654436291e-05,
+      "loss": 1.3914,
+      "step": 4170
+    },
+    {
+      "epoch": 3.8504500346180475,
+      "grad_norm": 4.22265625,
+      "learning_rate": 7.224149432955304e-05,
+      "loss": 1.5361,
+      "step": 4171
+    },
+    {
+      "epoch": 3.851373182552504,
+      "grad_norm": 3.876953125,
+      "learning_rate": 7.223482321547699e-05,
+      "loss": 1.1635,
+      "step": 4172
+    },
+    {
+      "epoch": 3.8522963304869604,
+      "grad_norm": 3.853515625,
+      "learning_rate": 7.222815210140093e-05,
+      "loss": 1.5975,
+      "step": 4173
+    },
+    {
+      "epoch": 3.853219478421417,
+      "grad_norm": 4.1015625,
+      "learning_rate": 7.222148098732489e-05,
+      "loss": 1.4519,
+      "step": 4174
+    },
+    {
+      "epoch": 3.8541426263558733,
+      "grad_norm": 3.91796875,
+      "learning_rate": 7.221480987324884e-05,
+      "loss": 1.2281,
+      "step": 4175
+    },
+    {
+      "epoch": 3.8550657742903303,
+      "grad_norm": 4.46484375,
+      "learning_rate": 7.220813875917278e-05,
+      "loss": 1.454,
+      "step": 4176
+    },
+    {
+      "epoch": 3.8559889222247863,
+      "grad_norm": 4.125,
+      "learning_rate": 7.220146764509673e-05,
+      "loss": 1.4468,
+      "step": 4177
+    },
+    {
+      "epoch": 3.856912070159243,
+      "grad_norm": 3.751953125,
+      "learning_rate": 7.219479653102069e-05,
+      "loss": 1.0762,
+      "step": 4178
+    },
+    {
+      "epoch": 3.8578352180936997,
+      "grad_norm": 4.51953125,
+      "learning_rate": 7.218812541694463e-05,
+      "loss": 1.5212,
+      "step": 4179
+    },
+    {
+      "epoch": 3.858758366028156,
+      "grad_norm": 3.61328125,
+      "learning_rate": 7.218145430286858e-05,
+      "loss": 1.5344,
+      "step": 4180
+    },
+    {
+      "epoch": 3.8596815139626126,
+      "grad_norm": 4.7421875,
+      "learning_rate": 7.217478318879252e-05,
+      "loss": 1.2734,
+      "step": 4181
+    },
+    {
+      "epoch": 3.860604661897069,
+      "grad_norm": 4.14453125,
+      "learning_rate": 7.216811207471648e-05,
+      "loss": 1.5632,
+      "step": 4182
+    },
+    {
+      "epoch": 3.8615278098315255,
+      "grad_norm": 4.19921875,
+      "learning_rate": 7.216144096064043e-05,
+      "loss": 1.5427,
+      "step": 4183
+    },
+    {
+      "epoch": 3.862450957765982,
+      "grad_norm": 4.8828125,
+      "learning_rate": 7.215476984656439e-05,
+      "loss": 1.4343,
+      "step": 4184
+    },
+    {
+      "epoch": 3.8633741057004385,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.214809873248833e-05,
+      "loss": 1.5315,
+      "step": 4185
+    },
+    {
+      "epoch": 3.864297253634895,
+      "grad_norm": 5.05078125,
+      "learning_rate": 7.214142761841228e-05,
+      "loss": 1.2651,
+      "step": 4186
+    },
+    {
+      "epoch": 3.8652204015693514,
+      "grad_norm": 4.609375,
+      "learning_rate": 7.213475650433624e-05,
+      "loss": 1.2783,
+      "step": 4187
+    },
+    {
+      "epoch": 3.866143549503808,
+      "grad_norm": 4.04296875,
+      "learning_rate": 7.212808539026018e-05,
+      "loss": 1.7463,
+      "step": 4188
+    },
+    {
+      "epoch": 3.8670666974382644,
+      "grad_norm": 3.869140625,
+      "learning_rate": 7.212141427618413e-05,
+      "loss": 1.2463,
+      "step": 4189
+    },
+    {
+      "epoch": 3.867989845372721,
+      "grad_norm": 4.828125,
+      "learning_rate": 7.211474316210807e-05,
+      "loss": 1.3212,
+      "step": 4190
+    },
+    {
+      "epoch": 3.8689129933071778,
+      "grad_norm": 4.5859375,
+      "learning_rate": 7.210807204803203e-05,
+      "loss": 1.4318,
+      "step": 4191
+    },
+    {
+      "epoch": 3.8698361412416338,
+      "grad_norm": 3.9453125,
+      "learning_rate": 7.210140093395598e-05,
+      "loss": 1.3606,
+      "step": 4192
+    },
+    {
+      "epoch": 3.8707592891760907,
+      "grad_norm": 3.619140625,
+      "learning_rate": 7.209472981987992e-05,
+      "loss": 1.0306,
+      "step": 4193
+    },
+    {
+      "epoch": 3.8716824371105467,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.208805870580387e-05,
+      "loss": 0.9644,
+      "step": 4194
+    },
+    {
+      "epoch": 3.8726055850450036,
+      "grad_norm": 3.884765625,
+      "learning_rate": 7.208138759172783e-05,
+      "loss": 1.3911,
+      "step": 4195
+    },
+    {
+      "epoch": 3.87352873297946,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.207471647765177e-05,
+      "loss": 1.384,
+      "step": 4196
+    },
+    {
+      "epoch": 3.8744518809139166,
+      "grad_norm": 3.70703125,
+      "learning_rate": 7.206804536357572e-05,
+      "loss": 1.0753,
+      "step": 4197
+    },
+    {
+      "epoch": 3.875375028848373,
+      "grad_norm": 3.9140625,
+      "learning_rate": 7.206137424949966e-05,
+      "loss": 1.4309,
+      "step": 4198
+    },
+    {
+      "epoch": 3.8762981767828295,
+      "grad_norm": 3.583984375,
+      "learning_rate": 7.205470313542362e-05,
+      "loss": 1.355,
+      "step": 4199
+    },
+    {
+      "epoch": 3.877221324717286,
+      "grad_norm": 4.51171875,
+      "learning_rate": 7.204803202134757e-05,
+      "loss": 1.4275,
+      "step": 4200
+    },
+    {
+      "epoch": 3.8781444726517424,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.204136090727151e-05,
+      "loss": 1.5085,
+      "step": 4201
+    },
+    {
+      "epoch": 3.879067620586199,
+      "grad_norm": 4.38671875,
+      "learning_rate": 7.203468979319546e-05,
+      "loss": 1.4873,
+      "step": 4202
+    },
+    {
+      "epoch": 3.8799907685206554,
+      "grad_norm": 4.26171875,
+      "learning_rate": 7.202801867911942e-05,
+      "loss": 1.4998,
+      "step": 4203
+    },
+    {
+      "epoch": 3.880913916455112,
+      "grad_norm": 4.15234375,
+      "learning_rate": 7.202134756504336e-05,
+      "loss": 1.1934,
+      "step": 4204
+    },
+    {
+      "epoch": 3.8818370643895683,
+      "grad_norm": 4.36328125,
+      "learning_rate": 7.201467645096731e-05,
+      "loss": 1.7886,
+      "step": 4205
+    },
+    {
+      "epoch": 3.882760212324025,
+      "grad_norm": 4.46484375,
+      "learning_rate": 7.200800533689125e-05,
+      "loss": 1.4888,
+      "step": 4206
+    },
+    {
+      "epoch": 3.8836833602584813,
+      "grad_norm": 3.60546875,
+      "learning_rate": 7.200133422281521e-05,
+      "loss": 1.5403,
+      "step": 4207
+    },
+    {
+      "epoch": 3.884606508192938,
+      "grad_norm": 3.90234375,
+      "learning_rate": 7.199466310873916e-05,
+      "loss": 1.5129,
+      "step": 4208
+    },
+    {
+      "epoch": 3.885529656127394,
+      "grad_norm": 4.3671875,
+      "learning_rate": 7.198799199466312e-05,
+      "loss": 1.5122,
+      "step": 4209
+    },
+    {
+      "epoch": 3.886452804061851,
+      "grad_norm": 4.74609375,
+      "learning_rate": 7.198132088058706e-05,
+      "loss": 1.5286,
+      "step": 4210
+    },
+    {
+      "epoch": 3.887375951996307,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.197464976651102e-05,
+      "loss": 1.3713,
+      "step": 4211
+    },
+    {
+      "epoch": 3.888299099930764,
+      "grad_norm": 3.53515625,
+      "learning_rate": 7.196797865243497e-05,
+      "loss": 1.3943,
+      "step": 4212
+    },
+    {
+      "epoch": 3.8892222478652205,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.196130753835891e-05,
+      "loss": 1.2925,
+      "step": 4213
+    },
+    {
+      "epoch": 3.890145395799677,
+      "grad_norm": 4.7890625,
+      "learning_rate": 7.195463642428286e-05,
+      "loss": 1.5332,
+      "step": 4214
+    },
+    {
+      "epoch": 3.8910685437341335,
+      "grad_norm": 3.8984375,
+      "learning_rate": 7.194796531020682e-05,
+      "loss": 1.2119,
+      "step": 4215
+    },
+    {
+      "epoch": 3.89199169166859,
+      "grad_norm": 3.66015625,
+      "learning_rate": 7.194129419613076e-05,
+      "loss": 1.6812,
+      "step": 4216
+    },
+    {
+      "epoch": 3.8929148396030464,
+      "grad_norm": 3.888671875,
+      "learning_rate": 7.193462308205471e-05,
+      "loss": 1.405,
+      "step": 4217
+    },
+    {
+      "epoch": 3.893837987537503,
+      "grad_norm": 4.09765625,
+      "learning_rate": 7.192795196797865e-05,
+      "loss": 1.4949,
+      "step": 4218
+    },
+    {
+      "epoch": 3.8947611354719593,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.192128085390261e-05,
+      "loss": 1.0836,
+      "step": 4219
+    },
+    {
+      "epoch": 3.895684283406416,
+      "grad_norm": 4.71484375,
+      "learning_rate": 7.191460973982656e-05,
+      "loss": 1.3926,
+      "step": 4220
+    },
+    {
+      "epoch": 3.8966074313408723,
+      "grad_norm": 4.1171875,
+      "learning_rate": 7.19079386257505e-05,
+      "loss": 1.5271,
+      "step": 4221
+    },
+    {
+      "epoch": 3.8975305792753288,
+      "grad_norm": 3.044921875,
+      "learning_rate": 7.190126751167445e-05,
+      "loss": 1.1248,
+      "step": 4222
+    },
+    {
+      "epoch": 3.8984537272097852,
+      "grad_norm": 3.380859375,
+      "learning_rate": 7.189459639759841e-05,
+      "loss": 1.292,
+      "step": 4223
+    },
+    {
+      "epoch": 3.8993768751442417,
+      "grad_norm": 2.994140625,
+      "learning_rate": 7.188792528352235e-05,
+      "loss": 1.6763,
+      "step": 4224
+    },
+    {
+      "epoch": 3.9003000230786986,
+      "grad_norm": 3.26953125,
+      "learning_rate": 7.18812541694463e-05,
+      "loss": 1.2778,
+      "step": 4225
+    },
+    {
+      "epoch": 3.9012231710131546,
+      "grad_norm": 4.21484375,
+      "learning_rate": 7.187458305537024e-05,
+      "loss": 1.7278,
+      "step": 4226
+    },
+    {
+      "epoch": 3.9021463189476115,
+      "grad_norm": 4.1328125,
+      "learning_rate": 7.18679119412942e-05,
+      "loss": 1.1414,
+      "step": 4227
+    },
+    {
+      "epoch": 3.903069466882068,
+      "grad_norm": 4.35546875,
+      "learning_rate": 7.186124082721815e-05,
+      "loss": 1.5962,
+      "step": 4228
+    },
+    {
+      "epoch": 3.9039926148165245,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.18545697131421e-05,
+      "loss": 1.3357,
+      "step": 4229
+    },
+    {
+      "epoch": 3.904915762750981,
+      "grad_norm": 3.669921875,
+      "learning_rate": 7.184789859906604e-05,
+      "loss": 1.408,
+      "step": 4230
+    },
+    {
+      "epoch": 3.9058389106854374,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.184122748499e-05,
+      "loss": 1.2134,
+      "step": 4231
+    },
+    {
+      "epoch": 3.906762058619894,
+      "grad_norm": 3.806640625,
+      "learning_rate": 7.183455637091394e-05,
+      "loss": 1.5227,
+      "step": 4232
+    },
+    {
+      "epoch": 3.9076852065543504,
+      "grad_norm": 4.4296875,
+      "learning_rate": 7.182788525683789e-05,
+      "loss": 1.3339,
+      "step": 4233
+    },
+    {
+      "epoch": 3.908608354488807,
+      "grad_norm": 4.12890625,
+      "learning_rate": 7.182121414276185e-05,
+      "loss": 1.5225,
+      "step": 4234
+    },
+    {
+      "epoch": 3.9095315024232633,
+      "grad_norm": 4.55078125,
+      "learning_rate": 7.181454302868579e-05,
+      "loss": 1.427,
+      "step": 4235
+    },
+    {
+      "epoch": 3.9104546503577198,
+      "grad_norm": 3.865234375,
+      "learning_rate": 7.180787191460975e-05,
+      "loss": 1.3394,
+      "step": 4236
+    },
+    {
+      "epoch": 3.9113777982921762,
+      "grad_norm": 4.21875,
+      "learning_rate": 7.18012008005337e-05,
+      "loss": 1.6443,
+      "step": 4237
+    },
+    {
+      "epoch": 3.9123009462266327,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.179452968645764e-05,
+      "loss": 1.5574,
+      "step": 4238
+    },
+    {
+      "epoch": 3.913224094161089,
+      "grad_norm": 4.7890625,
+      "learning_rate": 7.178785857238159e-05,
+      "loss": 1.4788,
+      "step": 4239
+    },
+    {
+      "epoch": 3.9141472420955457,
+      "grad_norm": 5.09765625,
+      "learning_rate": 7.178118745830555e-05,
+      "loss": 1.5911,
+      "step": 4240
+    },
+    {
+      "epoch": 3.915070390030002,
+      "grad_norm": 3.625,
+      "learning_rate": 7.177451634422949e-05,
+      "loss": 1.1825,
+      "step": 4241
+    },
+    {
+      "epoch": 3.915993537964459,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.176784523015344e-05,
+      "loss": 1.5112,
+      "step": 4242
+    },
+    {
+      "epoch": 3.916916685898915,
+      "grad_norm": 3.11328125,
+      "learning_rate": 7.176117411607738e-05,
+      "loss": 1.0468,
+      "step": 4243
+    },
+    {
+      "epoch": 3.917839833833372,
+      "grad_norm": 3.794921875,
+      "learning_rate": 7.175450300200134e-05,
+      "loss": 1.3602,
+      "step": 4244
+    },
+    {
+      "epoch": 3.9187629817678284,
+      "grad_norm": 4.37890625,
+      "learning_rate": 7.174783188792529e-05,
+      "loss": 1.4324,
+      "step": 4245
+    },
+    {
+      "epoch": 3.919686129702285,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.174116077384923e-05,
+      "loss": 1.6123,
+      "step": 4246
+    },
+    {
+      "epoch": 3.9206092776367414,
+      "grad_norm": 4.21875,
+      "learning_rate": 7.173448965977318e-05,
+      "loss": 1.624,
+      "step": 4247
+    },
+    {
+      "epoch": 3.921532425571198,
+      "grad_norm": 3.771484375,
+      "learning_rate": 7.172781854569714e-05,
+      "loss": 1.3004,
+      "step": 4248
+    },
+    {
+      "epoch": 3.9224555735056543,
+      "grad_norm": 4.29296875,
+      "learning_rate": 7.172114743162108e-05,
+      "loss": 1.4866,
+      "step": 4249
+    },
+    {
+      "epoch": 3.923378721440111,
+      "grad_norm": 3.314453125,
+      "learning_rate": 7.171447631754503e-05,
+      "loss": 1.4346,
+      "step": 4250
+    },
+    {
+      "epoch": 3.9243018693745673,
+      "grad_norm": 3.865234375,
+      "learning_rate": 7.170780520346897e-05,
+      "loss": 1.2688,
+      "step": 4251
+    },
+    {
+      "epoch": 3.9252250173090237,
+      "grad_norm": 5.35546875,
+      "learning_rate": 7.170113408939293e-05,
+      "loss": 1.3809,
+      "step": 4252
+    },
+    {
+      "epoch": 3.92614816524348,
+      "grad_norm": 4.26953125,
+      "learning_rate": 7.169446297531688e-05,
+      "loss": 1.3564,
+      "step": 4253
+    },
+    {
+      "epoch": 3.9270713131779367,
+      "grad_norm": 4.69140625,
+      "learning_rate": 7.168779186124082e-05,
+      "loss": 1.3594,
+      "step": 4254
+    },
+    {
+      "epoch": 3.927994461112393,
+      "grad_norm": 3.662109375,
+      "learning_rate": 7.168112074716477e-05,
+      "loss": 1.3608,
+      "step": 4255
+    },
+    {
+      "epoch": 3.9289176090468496,
+      "grad_norm": 3.62109375,
+      "learning_rate": 7.167444963308873e-05,
+      "loss": 1.4695,
+      "step": 4256
+    },
+    {
+      "epoch": 3.9298407569813065,
+      "grad_norm": 3.990234375,
+      "learning_rate": 7.166777851901267e-05,
+      "loss": 1.4634,
+      "step": 4257
+    },
+    {
+      "epoch": 3.9307639049157626,
+      "grad_norm": 3.380859375,
+      "learning_rate": 7.166110740493662e-05,
+      "loss": 1.2822,
+      "step": 4258
+    },
+    {
+      "epoch": 3.9316870528502195,
+      "grad_norm": 4.1015625,
+      "learning_rate": 7.165443629086058e-05,
+      "loss": 1.6211,
+      "step": 4259
+    },
+    {
+      "epoch": 3.9326102007846755,
+      "grad_norm": 3.78515625,
+      "learning_rate": 7.164776517678454e-05,
+      "loss": 1.5125,
+      "step": 4260
+    },
+    {
+      "epoch": 3.9335333487191324,
+      "grad_norm": 3.576171875,
+      "learning_rate": 7.164109406270848e-05,
+      "loss": 1.5443,
+      "step": 4261
+    },
+    {
+      "epoch": 3.934456496653589,
+      "grad_norm": 3.478515625,
+      "learning_rate": 7.163442294863243e-05,
+      "loss": 1.318,
+      "step": 4262
+    },
+    {
+      "epoch": 3.9353796445880453,
+      "grad_norm": 3.75390625,
+      "learning_rate": 7.162775183455637e-05,
+      "loss": 1.1721,
+      "step": 4263
+    },
+    {
+      "epoch": 3.936302792522502,
+      "grad_norm": 3.869140625,
+      "learning_rate": 7.162108072048033e-05,
+      "loss": 1.3636,
+      "step": 4264
+    },
+    {
+      "epoch": 3.9372259404569583,
+      "grad_norm": 3.314453125,
+      "learning_rate": 7.161440960640428e-05,
+      "loss": 1.3384,
+      "step": 4265
+    },
+    {
+      "epoch": 3.9381490883914148,
+      "grad_norm": 4.76171875,
+      "learning_rate": 7.160773849232822e-05,
+      "loss": 1.1799,
+      "step": 4266
+    },
+    {
+      "epoch": 3.9390722363258712,
+      "grad_norm": 3.935546875,
+      "learning_rate": 7.160106737825217e-05,
+      "loss": 1.459,
+      "step": 4267
+    },
+    {
+      "epoch": 3.9399953842603277,
+      "grad_norm": 4.6328125,
+      "learning_rate": 7.159439626417613e-05,
+      "loss": 0.9699,
+      "step": 4268
+    },
+    {
+      "epoch": 3.940918532194784,
+      "grad_norm": 4.52734375,
+      "learning_rate": 7.158772515010007e-05,
+      "loss": 1.3894,
+      "step": 4269
+    },
+    {
+      "epoch": 3.9418416801292406,
+      "grad_norm": 4.5390625,
+      "learning_rate": 7.158105403602402e-05,
+      "loss": 1.5723,
+      "step": 4270
+    },
+    {
+      "epoch": 3.942764828063697,
+      "grad_norm": 3.470703125,
+      "learning_rate": 7.157438292194796e-05,
+      "loss": 1.6738,
+      "step": 4271
+    },
+    {
+      "epoch": 3.9436879759981536,
+      "grad_norm": 3.353515625,
+      "learning_rate": 7.156771180787192e-05,
+      "loss": 1.5808,
+      "step": 4272
+    },
+    {
+      "epoch": 3.94461112393261,
+      "grad_norm": 3.587890625,
+      "learning_rate": 7.156104069379587e-05,
+      "loss": 1.5186,
+      "step": 4273
+    },
+    {
+      "epoch": 3.945534271867067,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.155436957971981e-05,
+      "loss": 1.4048,
+      "step": 4274
+    },
+    {
+      "epoch": 3.946457419801523,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.154769846564376e-05,
+      "loss": 1.452,
+      "step": 4275
+    },
+    {
+      "epoch": 3.94738056773598,
+      "grad_norm": 3.81640625,
+      "learning_rate": 7.154102735156772e-05,
+      "loss": 1.3662,
+      "step": 4276
+    },
+    {
+      "epoch": 3.948303715670436,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.153435623749166e-05,
+      "loss": 1.6003,
+      "step": 4277
+    },
+    {
+      "epoch": 3.949226863604893,
+      "grad_norm": 3.4453125,
+      "learning_rate": 7.152768512341561e-05,
+      "loss": 1.1211,
+      "step": 4278
+    },
+    {
+      "epoch": 3.9501500115393493,
+      "grad_norm": 3.958984375,
+      "learning_rate": 7.152101400933955e-05,
+      "loss": 1.4197,
+      "step": 4279
+    },
+    {
+      "epoch": 3.9510731594738058,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.151434289526351e-05,
+      "loss": 1.4187,
+      "step": 4280
+    },
+    {
+      "epoch": 3.9519963074082622,
+      "grad_norm": 4.453125,
+      "learning_rate": 7.150767178118746e-05,
+      "loss": 1.7178,
+      "step": 4281
+    },
+    {
+      "epoch": 3.9529194553427187,
+      "grad_norm": 3.755859375,
+      "learning_rate": 7.15010006671114e-05,
+      "loss": 1.6758,
+      "step": 4282
+    },
+    {
+      "epoch": 3.953842603277175,
+      "grad_norm": 4.265625,
+      "learning_rate": 7.149432955303535e-05,
+      "loss": 1.015,
+      "step": 4283
+    },
+    {
+      "epoch": 3.9547657512116317,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.148765843895931e-05,
+      "loss": 1.3979,
+      "step": 4284
+    },
+    {
+      "epoch": 3.955688899146088,
+      "grad_norm": 3.76953125,
+      "learning_rate": 7.148098732488327e-05,
+      "loss": 1.2168,
+      "step": 4285
+    },
+    {
+      "epoch": 3.9566120470805446,
+      "grad_norm": 5.0234375,
+      "learning_rate": 7.147431621080721e-05,
+      "loss": 1.604,
+      "step": 4286
+    },
+    {
+      "epoch": 3.957535195015001,
+      "grad_norm": 4.69921875,
+      "learning_rate": 7.146764509673116e-05,
+      "loss": 1.4026,
+      "step": 4287
+    },
+    {
+      "epoch": 3.9584583429494575,
+      "grad_norm": 5.390625,
+      "learning_rate": 7.14609739826551e-05,
+      "loss": 1.4768,
+      "step": 4288
+    },
+    {
+      "epoch": 3.959381490883914,
+      "grad_norm": 4.453125,
+      "learning_rate": 7.145430286857906e-05,
+      "loss": 1.6768,
+      "step": 4289
+    },
+    {
+      "epoch": 3.9603046388183705,
+      "grad_norm": 5.609375,
+      "learning_rate": 7.144763175450301e-05,
+      "loss": 1.4929,
+      "step": 4290
+    },
+    {
+      "epoch": 3.9612277867528274,
+      "grad_norm": 3.724609375,
+      "learning_rate": 7.144096064042695e-05,
+      "loss": 1.6252,
+      "step": 4291
+    },
+    {
+      "epoch": 3.9621509346872834,
+      "grad_norm": 3.90234375,
+      "learning_rate": 7.14342895263509e-05,
+      "loss": 1.3477,
+      "step": 4292
+    },
+    {
+      "epoch": 3.9630740826217403,
+      "grad_norm": 3.869140625,
+      "learning_rate": 7.142761841227486e-05,
+      "loss": 1.5566,
+      "step": 4293
+    },
+    {
+      "epoch": 3.963997230556197,
+      "grad_norm": 4.6640625,
+      "learning_rate": 7.14209472981988e-05,
+      "loss": 1.2664,
+      "step": 4294
+    },
+    {
+      "epoch": 3.9649203784906533,
+      "grad_norm": 5.08203125,
+      "learning_rate": 7.141427618412275e-05,
+      "loss": 1.3828,
+      "step": 4295
+    },
+    {
+      "epoch": 3.9658435264251097,
+      "grad_norm": 3.154296875,
+      "learning_rate": 7.14076050700467e-05,
+      "loss": 1.4834,
+      "step": 4296
+    },
+    {
+      "epoch": 3.966766674359566,
+      "grad_norm": 4.46484375,
+      "learning_rate": 7.140093395597065e-05,
+      "loss": 1.5479,
+      "step": 4297
+    },
+    {
+      "epoch": 3.9676898222940227,
+      "grad_norm": 4.35546875,
+      "learning_rate": 7.13942628418946e-05,
+      "loss": 1.4832,
+      "step": 4298
+    },
+    {
+      "epoch": 3.968612970228479,
+      "grad_norm": 3.962890625,
+      "learning_rate": 7.138759172781854e-05,
+      "loss": 1.5173,
+      "step": 4299
+    },
+    {
+      "epoch": 3.9695361181629356,
+      "grad_norm": 4.125,
+      "learning_rate": 7.138092061374249e-05,
+      "loss": 1.4612,
+      "step": 4300
+    },
+    {
+      "epoch": 3.970459266097392,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.137424949966645e-05,
+      "loss": 1.5576,
+      "step": 4301
+    },
+    {
+      "epoch": 3.9713824140318486,
+      "grad_norm": 3.9296875,
+      "learning_rate": 7.13675783855904e-05,
+      "loss": 1.4058,
+      "step": 4302
+    },
+    {
+      "epoch": 3.972305561966305,
+      "grad_norm": 4.19921875,
+      "learning_rate": 7.136090727151434e-05,
+      "loss": 1.344,
+      "step": 4303
+    },
+    {
+      "epoch": 3.9732287099007615,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.13542361574383e-05,
+      "loss": 1.4922,
+      "step": 4304
+    },
+    {
+      "epoch": 3.974151857835218,
+      "grad_norm": 4.19921875,
+      "learning_rate": 7.134756504336224e-05,
+      "loss": 1.4083,
+      "step": 4305
+    },
+    {
+      "epoch": 3.9750750057696744,
+      "grad_norm": 3.943359375,
+      "learning_rate": 7.134089392928619e-05,
+      "loss": 1.47,
+      "step": 4306
+    },
+    {
+      "epoch": 3.975998153704131,
+      "grad_norm": 4.76171875,
+      "learning_rate": 7.133422281521014e-05,
+      "loss": 1.6077,
+      "step": 4307
+    },
+    {
+      "epoch": 3.976921301638588,
+      "grad_norm": 3.986328125,
+      "learning_rate": 7.13275517011341e-05,
+      "loss": 1.3335,
+      "step": 4308
+    },
+    {
+      "epoch": 3.977844449573044,
+      "grad_norm": 4.78515625,
+      "learning_rate": 7.132088058705805e-05,
+      "loss": 1.3665,
+      "step": 4309
+    },
+    {
+      "epoch": 3.9787675975075008,
+      "grad_norm": 3.2109375,
+      "learning_rate": 7.1314209472982e-05,
+      "loss": 1.3228,
+      "step": 4310
+    },
+    {
+      "epoch": 3.9796907454419572,
+      "grad_norm": 3.869140625,
+      "learning_rate": 7.130753835890594e-05,
+      "loss": 1.4177,
+      "step": 4311
+    },
+    {
+      "epoch": 3.9806138933764137,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.130086724482989e-05,
+      "loss": 1.5881,
+      "step": 4312
+    },
+    {
+      "epoch": 3.98153704131087,
+      "grad_norm": 4.13671875,
+      "learning_rate": 7.129419613075385e-05,
+      "loss": 1.4951,
+      "step": 4313
+    },
+    {
+      "epoch": 3.9824601892453266,
+      "grad_norm": 4.1640625,
+      "learning_rate": 7.12875250166778e-05,
+      "loss": 1.2603,
+      "step": 4314
+    },
+    {
+      "epoch": 3.983383337179783,
+      "grad_norm": 3.884765625,
+      "learning_rate": 7.128085390260174e-05,
+      "loss": 1.3533,
+      "step": 4315
+    },
+    {
+      "epoch": 3.9843064851142396,
+      "grad_norm": 3.74609375,
+      "learning_rate": 7.127418278852568e-05,
+      "loss": 1.6982,
+      "step": 4316
+    },
+    {
+      "epoch": 3.985229633048696,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.126751167444964e-05,
+      "loss": 1.6199,
+      "step": 4317
+    },
+    {
+      "epoch": 3.9861527809831525,
+      "grad_norm": 3.94140625,
+      "learning_rate": 7.126084056037359e-05,
+      "loss": 1.5552,
+      "step": 4318
+    },
+    {
+      "epoch": 3.987075928917609,
+      "grad_norm": 3.787109375,
+      "learning_rate": 7.125416944629753e-05,
+      "loss": 1.6353,
+      "step": 4319
+    },
+    {
+      "epoch": 3.9879990768520654,
+      "grad_norm": 4.4609375,
+      "learning_rate": 7.124749833222148e-05,
+      "loss": 1.5457,
+      "step": 4320
+    },
+    {
+      "epoch": 3.988922224786522,
+      "grad_norm": 4.57421875,
+      "learning_rate": 7.124082721814544e-05,
+      "loss": 1.3347,
+      "step": 4321
+    },
+    {
+      "epoch": 3.9898453727209784,
+      "grad_norm": 4.21875,
+      "learning_rate": 7.123415610406938e-05,
+      "loss": 1.666,
+      "step": 4322
+    },
+    {
+      "epoch": 3.9907685206554353,
+      "grad_norm": 4.20703125,
+      "learning_rate": 7.122748498999333e-05,
+      "loss": 1.5586,
+      "step": 4323
+    },
+    {
+      "epoch": 3.9916916685898913,
+      "grad_norm": 4.10546875,
+      "learning_rate": 7.122081387591728e-05,
+      "loss": 1.4485,
+      "step": 4324
+    },
+    {
+      "epoch": 3.9926148165243482,
+      "grad_norm": 3.87109375,
+      "learning_rate": 7.121414276184123e-05,
+      "loss": 1.4111,
+      "step": 4325
+    },
+    {
+      "epoch": 3.9935379644588043,
+      "grad_norm": 4.45703125,
+      "learning_rate": 7.120747164776518e-05,
+      "loss": 1.5305,
+      "step": 4326
+    },
+    {
+      "epoch": 3.994461112393261,
+      "grad_norm": 4.21484375,
+      "learning_rate": 7.120080053368913e-05,
+      "loss": 1.5854,
+      "step": 4327
+    },
+    {
+      "epoch": 3.9953842603277177,
+      "grad_norm": 4.26953125,
+      "learning_rate": 7.119412941961307e-05,
+      "loss": 1.2834,
+      "step": 4328
+    },
+    {
+      "epoch": 3.996307408262174,
+      "grad_norm": 4.5546875,
+      "learning_rate": 7.118745830553703e-05,
+      "loss": 1.4128,
+      "step": 4329
+    },
+    {
+      "epoch": 3.9972305561966306,
+      "grad_norm": 4.625,
+      "learning_rate": 7.118078719146098e-05,
+      "loss": 1.5762,
+      "step": 4330
+    },
+    {
+      "epoch": 3.998153704131087,
+      "grad_norm": 3.255859375,
+      "learning_rate": 7.117411607738492e-05,
+      "loss": 1.1023,
+      "step": 4331
+    },
+    {
+      "epoch": 3.9990768520655435,
+      "grad_norm": 3.68359375,
+      "learning_rate": 7.116744496330887e-05,
+      "loss": 1.4111,
+      "step": 4332
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.1328125,
+      "learning_rate": 7.116077384923282e-05,
+      "loss": 1.0882,
+      "step": 4333
+    },
+    {
+      "epoch": 4.000923147934457,
+      "grad_norm": 4.171875,
+      "learning_rate": 7.115410273515677e-05,
+      "loss": 1.6987,
+      "step": 4334
+    },
+    {
+      "epoch": 4.001846295868913,
+      "grad_norm": 4.3984375,
+      "learning_rate": 7.114743162108073e-05,
+      "loss": 1.5806,
+      "step": 4335
+    },
+    {
+      "epoch": 4.00276944380337,
+      "grad_norm": 4.109375,
+      "learning_rate": 7.114076050700467e-05,
+      "loss": 1.3384,
+      "step": 4336
+    },
+    {
+      "epoch": 4.003692591737826,
+      "grad_norm": 3.759765625,
+      "learning_rate": 7.113408939292862e-05,
+      "loss": 1.3669,
+      "step": 4337
+    },
+    {
+      "epoch": 4.004615739672283,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.112741827885258e-05,
+      "loss": 1.4187,
+      "step": 4338
+    },
+    {
+      "epoch": 4.005538887606739,
+      "grad_norm": 3.8515625,
+      "learning_rate": 7.112074716477652e-05,
+      "loss": 1.2968,
+      "step": 4339
+    },
+    {
+      "epoch": 4.006462035541196,
+      "grad_norm": 4.21875,
+      "learning_rate": 7.111407605070047e-05,
+      "loss": 1.4324,
+      "step": 4340
+    },
+    {
+      "epoch": 4.007385183475652,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.110740493662442e-05,
+      "loss": 1.2344,
+      "step": 4341
+    },
+    {
+      "epoch": 4.008308331410109,
+      "grad_norm": 4.20703125,
+      "learning_rate": 7.110073382254837e-05,
+      "loss": 1.3199,
+      "step": 4342
+    },
+    {
+      "epoch": 4.009231479344565,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.109406270847232e-05,
+      "loss": 1.436,
+      "step": 4343
+    },
+    {
+      "epoch": 4.010154627279022,
+      "grad_norm": 4.25,
+      "learning_rate": 7.108739159439627e-05,
+      "loss": 1.3967,
+      "step": 4344
+    },
+    {
+      "epoch": 4.011077775213478,
+      "grad_norm": 4.171875,
+      "learning_rate": 7.108072048032022e-05,
+      "loss": 1.353,
+      "step": 4345
+    },
+    {
+      "epoch": 4.0120009231479346,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.107404936624417e-05,
+      "loss": 1.3176,
+      "step": 4346
+    },
+    {
+      "epoch": 4.012924071082391,
+      "grad_norm": 4.30078125,
+      "learning_rate": 7.106737825216812e-05,
+      "loss": 1.2083,
+      "step": 4347
+    },
+    {
+      "epoch": 4.0138472190168475,
+      "grad_norm": 4.49609375,
+      "learning_rate": 7.106070713809206e-05,
+      "loss": 1.4047,
+      "step": 4348
+    },
+    {
+      "epoch": 4.0147703669513035,
+      "grad_norm": 4.20703125,
+      "learning_rate": 7.105403602401602e-05,
+      "loss": 1.5366,
+      "step": 4349
+    },
+    {
+      "epoch": 4.01569351488576,
+      "grad_norm": 3.9140625,
+      "learning_rate": 7.104736490993996e-05,
+      "loss": 1.5852,
+      "step": 4350
+    },
+    {
+      "epoch": 4.016616662820217,
+      "grad_norm": 4.140625,
+      "learning_rate": 7.104069379586391e-05,
+      "loss": 1.6792,
+      "step": 4351
+    },
+    {
+      "epoch": 4.017539810754673,
+      "grad_norm": 4.10546875,
+      "learning_rate": 7.103402268178786e-05,
+      "loss": 1.4668,
+      "step": 4352
+    },
+    {
+      "epoch": 4.01846295868913,
+      "grad_norm": 3.98828125,
+      "learning_rate": 7.102735156771181e-05,
+      "loss": 1.5787,
+      "step": 4353
+    },
+    {
+      "epoch": 4.019386106623586,
+      "grad_norm": 4.390625,
+      "learning_rate": 7.102068045363576e-05,
+      "loss": 1.4341,
+      "step": 4354
+    },
+    {
+      "epoch": 4.020309254558043,
+      "grad_norm": 4.41796875,
+      "learning_rate": 7.10140093395597e-05,
+      "loss": 1.3802,
+      "step": 4355
+    },
+    {
+      "epoch": 4.021232402492499,
+      "grad_norm": 3.55859375,
+      "learning_rate": 7.100733822548365e-05,
+      "loss": 1.0208,
+      "step": 4356
+    },
+    {
+      "epoch": 4.022155550426956,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.100066711140761e-05,
+      "loss": 1.1699,
+      "step": 4357
+    },
+    {
+      "epoch": 4.023078698361412,
+      "grad_norm": 4.05078125,
+      "learning_rate": 7.099399599733156e-05,
+      "loss": 1.3215,
+      "step": 4358
+    },
+    {
+      "epoch": 4.024001846295869,
+      "grad_norm": 4.4140625,
+      "learning_rate": 7.09873248832555e-05,
+      "loss": 1.3708,
+      "step": 4359
+    },
+    {
+      "epoch": 4.024924994230325,
+      "grad_norm": 4.46484375,
+      "learning_rate": 7.098065376917946e-05,
+      "loss": 1.3529,
+      "step": 4360
+    },
+    {
+      "epoch": 4.025848142164782,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.09739826551034e-05,
+      "loss": 1.4938,
+      "step": 4361
+    },
+    {
+      "epoch": 4.026771290099238,
+      "grad_norm": 3.85546875,
+      "learning_rate": 7.096731154102736e-05,
+      "loss": 1.0619,
+      "step": 4362
+    },
+    {
+      "epoch": 4.027694438033695,
+      "grad_norm": 3.611328125,
+      "learning_rate": 7.096064042695131e-05,
+      "loss": 1.2786,
+      "step": 4363
+    },
+    {
+      "epoch": 4.028617585968151,
+      "grad_norm": 3.810546875,
+      "learning_rate": 7.095396931287526e-05,
+      "loss": 1.1272,
+      "step": 4364
+    },
+    {
+      "epoch": 4.029540733902608,
+      "grad_norm": 4.06640625,
+      "learning_rate": 7.09472981987992e-05,
+      "loss": 1.467,
+      "step": 4365
+    },
+    {
+      "epoch": 4.030463881837065,
+      "grad_norm": 4.68359375,
+      "learning_rate": 7.094062708472316e-05,
+      "loss": 1.2676,
+      "step": 4366
+    },
+    {
+      "epoch": 4.031387029771521,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.09339559706471e-05,
+      "loss": 1.4531,
+      "step": 4367
+    },
+    {
+      "epoch": 4.032310177705978,
+      "grad_norm": 4.37109375,
+      "learning_rate": 7.092728485657105e-05,
+      "loss": 1.4666,
+      "step": 4368
+    },
+    {
+      "epoch": 4.033233325640434,
+      "grad_norm": 4.578125,
+      "learning_rate": 7.0920613742495e-05,
+      "loss": 1.0421,
+      "step": 4369
+    },
+    {
+      "epoch": 4.034156473574891,
+      "grad_norm": 3.50390625,
+      "learning_rate": 7.091394262841895e-05,
+      "loss": 1.4094,
+      "step": 4370
+    },
+    {
+      "epoch": 4.035079621509347,
+      "grad_norm": 4.25,
+      "learning_rate": 7.09072715143429e-05,
+      "loss": 1.4155,
+      "step": 4371
+    },
+    {
+      "epoch": 4.036002769443804,
+      "grad_norm": 4.9375,
+      "learning_rate": 7.090060040026685e-05,
+      "loss": 1.2385,
+      "step": 4372
+    },
+    {
+      "epoch": 4.03692591737826,
+      "grad_norm": 4.125,
+      "learning_rate": 7.089392928619079e-05,
+      "loss": 1.2683,
+      "step": 4373
+    },
+    {
+      "epoch": 4.037849065312717,
+      "grad_norm": 3.798828125,
+      "learning_rate": 7.088725817211475e-05,
+      "loss": 1.2897,
+      "step": 4374
+    },
+    {
+      "epoch": 4.038772213247173,
+      "grad_norm": 3.599609375,
+      "learning_rate": 7.08805870580387e-05,
+      "loss": 1.2294,
+      "step": 4375
+    },
+    {
+      "epoch": 4.0396953611816295,
+      "grad_norm": 4.32421875,
+      "learning_rate": 7.087391594396264e-05,
+      "loss": 1.4353,
+      "step": 4376
+    },
+    {
+      "epoch": 4.0406185091160856,
+      "grad_norm": 4.0390625,
+      "learning_rate": 7.086724482988659e-05,
+      "loss": 1.6721,
+      "step": 4377
+    },
+    {
+      "epoch": 4.0415416570505425,
+      "grad_norm": 3.634765625,
+      "learning_rate": 7.086057371581055e-05,
+      "loss": 1.5776,
+      "step": 4378
+    },
+    {
+      "epoch": 4.0424648049849985,
+      "grad_norm": 3.623046875,
+      "learning_rate": 7.085390260173449e-05,
+      "loss": 1.5938,
+      "step": 4379
+    },
+    {
+      "epoch": 4.043387952919455,
+      "grad_norm": 4.20703125,
+      "learning_rate": 7.084723148765844e-05,
+      "loss": 1.4263,
+      "step": 4380
+    },
+    {
+      "epoch": 4.044311100853911,
+      "grad_norm": 3.708984375,
+      "learning_rate": 7.084056037358238e-05,
+      "loss": 1.2809,
+      "step": 4381
+    },
+    {
+      "epoch": 4.045234248788368,
+      "grad_norm": 3.103515625,
+      "learning_rate": 7.083388925950634e-05,
+      "loss": 1.3675,
+      "step": 4382
+    },
+    {
+      "epoch": 4.046157396722825,
+      "grad_norm": 3.990234375,
+      "learning_rate": 7.082721814543029e-05,
+      "loss": 1.4819,
+      "step": 4383
+    },
+    {
+      "epoch": 4.047080544657281,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.082054703135423e-05,
+      "loss": 1.2927,
+      "step": 4384
+    },
+    {
+      "epoch": 4.048003692591738,
+      "grad_norm": 3.994140625,
+      "learning_rate": 7.081387591727819e-05,
+      "loss": 1.3223,
+      "step": 4385
+    },
+    {
+      "epoch": 4.048926840526194,
+      "grad_norm": 4.27734375,
+      "learning_rate": 7.080720480320215e-05,
+      "loss": 1.3994,
+      "step": 4386
+    },
+    {
+      "epoch": 4.049849988460651,
+      "grad_norm": 3.83203125,
+      "learning_rate": 7.08005336891261e-05,
+      "loss": 1.5002,
+      "step": 4387
+    },
+    {
+      "epoch": 4.050773136395107,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.079386257505004e-05,
+      "loss": 1.3628,
+      "step": 4388
+    },
+    {
+      "epoch": 4.051696284329564,
+      "grad_norm": 4.11328125,
+      "learning_rate": 7.078719146097399e-05,
+      "loss": 1.46,
+      "step": 4389
+    },
+    {
+      "epoch": 4.05261943226402,
+      "grad_norm": 4.42578125,
+      "learning_rate": 7.078052034689794e-05,
+      "loss": 1.4639,
+      "step": 4390
+    },
+    {
+      "epoch": 4.053542580198477,
+      "grad_norm": 4.14453125,
+      "learning_rate": 7.077384923282189e-05,
+      "loss": 1.5649,
+      "step": 4391
+    },
+    {
+      "epoch": 4.054465728132933,
+      "grad_norm": 4.62890625,
+      "learning_rate": 7.076717811874584e-05,
+      "loss": 1.728,
+      "step": 4392
+    },
+    {
+      "epoch": 4.05538887606739,
+      "grad_norm": 3.478515625,
+      "learning_rate": 7.076050700466978e-05,
+      "loss": 1.5206,
+      "step": 4393
+    },
+    {
+      "epoch": 4.056312024001846,
+      "grad_norm": 3.90234375,
+      "learning_rate": 7.075383589059374e-05,
+      "loss": 1.5247,
+      "step": 4394
+    },
+    {
+      "epoch": 4.057235171936303,
+      "grad_norm": 3.349609375,
+      "learning_rate": 7.074716477651769e-05,
+      "loss": 1.4529,
+      "step": 4395
+    },
+    {
+      "epoch": 4.058158319870759,
+      "grad_norm": 3.54296875,
+      "learning_rate": 7.074049366244163e-05,
+      "loss": 1.6375,
+      "step": 4396
+    },
+    {
+      "epoch": 4.059081467805216,
+      "grad_norm": 3.681640625,
+      "learning_rate": 7.073382254836558e-05,
+      "loss": 1.7805,
+      "step": 4397
+    },
+    {
+      "epoch": 4.060004615739672,
+      "grad_norm": 3.958984375,
+      "learning_rate": 7.072715143428954e-05,
+      "loss": 1.707,
+      "step": 4398
+    },
+    {
+      "epoch": 4.060927763674129,
+      "grad_norm": 3.794921875,
+      "learning_rate": 7.072048032021348e-05,
+      "loss": 1.3962,
+      "step": 4399
+    },
+    {
+      "epoch": 4.061850911608586,
+      "grad_norm": 3.814453125,
+      "learning_rate": 7.071380920613743e-05,
+      "loss": 1.5015,
+      "step": 4400
+    },
+    {
+      "epoch": 4.062774059543042,
+      "grad_norm": 3.63671875,
+      "learning_rate": 7.070713809206137e-05,
+      "loss": 1.4275,
+      "step": 4401
+    },
+    {
+      "epoch": 4.063697207477499,
+      "grad_norm": 3.439453125,
+      "learning_rate": 7.070046697798533e-05,
+      "loss": 1.3167,
+      "step": 4402
+    },
+    {
+      "epoch": 4.064620355411955,
+      "grad_norm": 3.68359375,
+      "learning_rate": 7.069379586390928e-05,
+      "loss": 1.2681,
+      "step": 4403
+    },
+    {
+      "epoch": 4.065543503346412,
+      "grad_norm": 4.0703125,
+      "learning_rate": 7.068712474983322e-05,
+      "loss": 1.4182,
+      "step": 4404
+    },
+    {
+      "epoch": 4.066466651280868,
+      "grad_norm": 4.05078125,
+      "learning_rate": 7.068045363575717e-05,
+      "loss": 1.5557,
+      "step": 4405
+    },
+    {
+      "epoch": 4.0673897992153245,
+      "grad_norm": 3.00390625,
+      "learning_rate": 7.067378252168113e-05,
+      "loss": 1.1266,
+      "step": 4406
+    },
+    {
+      "epoch": 4.0683129471497805,
+      "grad_norm": 3.78515625,
+      "learning_rate": 7.066711140760507e-05,
+      "loss": 1.4314,
+      "step": 4407
+    },
+    {
+      "epoch": 4.0692360950842374,
+      "grad_norm": 3.537109375,
+      "learning_rate": 7.066044029352902e-05,
+      "loss": 1.4902,
+      "step": 4408
+    },
+    {
+      "epoch": 4.0701592430186935,
+      "grad_norm": 4.0234375,
+      "learning_rate": 7.065376917945296e-05,
+      "loss": 1.4944,
+      "step": 4409
+    },
+    {
+      "epoch": 4.07108239095315,
+      "grad_norm": 3.736328125,
+      "learning_rate": 7.064709806537692e-05,
+      "loss": 1.2592,
+      "step": 4410
+    },
+    {
+      "epoch": 4.072005538887606,
+      "grad_norm": 4.26953125,
+      "learning_rate": 7.064042695130088e-05,
+      "loss": 1.2227,
+      "step": 4411
+    },
+    {
+      "epoch": 4.072928686822063,
+      "grad_norm": 4.7109375,
+      "learning_rate": 7.063375583722483e-05,
+      "loss": 1.3518,
+      "step": 4412
+    },
+    {
+      "epoch": 4.073851834756519,
+      "grad_norm": 3.275390625,
+      "learning_rate": 7.062708472314877e-05,
+      "loss": 1.4414,
+      "step": 4413
+    },
+    {
+      "epoch": 4.074774982690976,
+      "grad_norm": 3.962890625,
+      "learning_rate": 7.062041360907272e-05,
+      "loss": 1.3447,
+      "step": 4414
+    },
+    {
+      "epoch": 4.075698130625432,
+      "grad_norm": 3.95703125,
+      "learning_rate": 7.061374249499668e-05,
+      "loss": 1.2628,
+      "step": 4415
+    },
+    {
+      "epoch": 4.076621278559889,
+      "grad_norm": 3.6328125,
+      "learning_rate": 7.060707138092062e-05,
+      "loss": 1.5684,
+      "step": 4416
+    },
+    {
+      "epoch": 4.077544426494346,
+      "grad_norm": 4.265625,
+      "learning_rate": 7.060040026684457e-05,
+      "loss": 1.4495,
+      "step": 4417
+    },
+    {
+      "epoch": 4.078467574428802,
+      "grad_norm": 3.744140625,
+      "learning_rate": 7.059372915276851e-05,
+      "loss": 1.1146,
+      "step": 4418
+    },
+    {
+      "epoch": 4.079390722363259,
+      "grad_norm": 3.94140625,
+      "learning_rate": 7.058705803869247e-05,
+      "loss": 1.4788,
+      "step": 4419
+    },
+    {
+      "epoch": 4.080313870297715,
+      "grad_norm": 3.392578125,
+      "learning_rate": 7.058038692461642e-05,
+      "loss": 1.3867,
+      "step": 4420
+    },
+    {
+      "epoch": 4.081237018232172,
+      "grad_norm": 3.6640625,
+      "learning_rate": 7.057371581054036e-05,
+      "loss": 1.4656,
+      "step": 4421
+    },
+    {
+      "epoch": 4.082160166166628,
+      "grad_norm": 3.677734375,
+      "learning_rate": 7.056704469646431e-05,
+      "loss": 1.4164,
+      "step": 4422
+    },
+    {
+      "epoch": 4.083083314101085,
+      "grad_norm": 3.41015625,
+      "learning_rate": 7.056037358238827e-05,
+      "loss": 1.4392,
+      "step": 4423
+    },
+    {
+      "epoch": 4.084006462035541,
+      "grad_norm": 3.673828125,
+      "learning_rate": 7.055370246831221e-05,
+      "loss": 1.4231,
+      "step": 4424
+    },
+    {
+      "epoch": 4.084929609969998,
+      "grad_norm": 3.48046875,
+      "learning_rate": 7.054703135423616e-05,
+      "loss": 1.5242,
+      "step": 4425
+    },
+    {
+      "epoch": 4.085852757904454,
+      "grad_norm": 3.87890625,
+      "learning_rate": 7.05403602401601e-05,
+      "loss": 1.4166,
+      "step": 4426
+    },
+    {
+      "epoch": 4.086775905838911,
+      "grad_norm": 4.5546875,
+      "learning_rate": 7.053368912608406e-05,
+      "loss": 1.5505,
+      "step": 4427
+    },
+    {
+      "epoch": 4.087699053773367,
+      "grad_norm": 3.458984375,
+      "learning_rate": 7.0527018012008e-05,
+      "loss": 1.2839,
+      "step": 4428
+    },
+    {
+      "epoch": 4.088622201707824,
+      "grad_norm": 3.779296875,
+      "learning_rate": 7.052034689793195e-05,
+      "loss": 1.2551,
+      "step": 4429
+    },
+    {
+      "epoch": 4.08954534964228,
+      "grad_norm": 4.0703125,
+      "learning_rate": 7.05136757838559e-05,
+      "loss": 1.4194,
+      "step": 4430
+    },
+    {
+      "epoch": 4.090468497576737,
+      "grad_norm": 3.68359375,
+      "learning_rate": 7.050700466977986e-05,
+      "loss": 1.3257,
+      "step": 4431
+    },
+    {
+      "epoch": 4.091391645511193,
+      "grad_norm": 3.9921875,
+      "learning_rate": 7.05003335557038e-05,
+      "loss": 1.4019,
+      "step": 4432
+    },
+    {
+      "epoch": 4.09231479344565,
+      "grad_norm": 3.29296875,
+      "learning_rate": 7.049366244162775e-05,
+      "loss": 1.1664,
+      "step": 4433
+    },
+    {
+      "epoch": 4.0932379413801065,
+      "grad_norm": 4.66796875,
+      "learning_rate": 7.048699132755169e-05,
+      "loss": 1.1936,
+      "step": 4434
+    },
+    {
+      "epoch": 4.094161089314563,
+      "grad_norm": 4.36328125,
+      "learning_rate": 7.048032021347567e-05,
+      "loss": 1.6213,
+      "step": 4435
+    },
+    {
+      "epoch": 4.0950842372490195,
+      "grad_norm": 3.763671875,
+      "learning_rate": 7.047364909939961e-05,
+      "loss": 1.4519,
+      "step": 4436
+    },
+    {
+      "epoch": 4.0960073851834755,
+      "grad_norm": 4.109375,
+      "learning_rate": 7.046697798532356e-05,
+      "loss": 1.2126,
+      "step": 4437
+    },
+    {
+      "epoch": 4.096930533117932,
+      "grad_norm": 5.1796875,
+      "learning_rate": 7.04603068712475e-05,
+      "loss": 1.2002,
+      "step": 4438
+    },
+    {
+      "epoch": 4.0978536810523885,
+      "grad_norm": 3.685546875,
+      "learning_rate": 7.045363575717146e-05,
+      "loss": 1.3356,
+      "step": 4439
+    },
+    {
+      "epoch": 4.098776828986845,
+      "grad_norm": 3.1953125,
+      "learning_rate": 7.04469646430954e-05,
+      "loss": 1.156,
+      "step": 4440
+    },
+    {
+      "epoch": 4.099699976921301,
+      "grad_norm": 3.966796875,
+      "learning_rate": 7.044029352901935e-05,
+      "loss": 1.286,
+      "step": 4441
+    },
+    {
+      "epoch": 4.100623124855758,
+      "grad_norm": 3.759765625,
+      "learning_rate": 7.04336224149433e-05,
+      "loss": 1.5649,
+      "step": 4442
+    },
+    {
+      "epoch": 4.101546272790214,
+      "grad_norm": 3.34765625,
+      "learning_rate": 7.042695130086726e-05,
+      "loss": 1.3085,
+      "step": 4443
+    },
+    {
+      "epoch": 4.102469420724671,
+      "grad_norm": 3.3515625,
+      "learning_rate": 7.04202801867912e-05,
+      "loss": 1.208,
+      "step": 4444
+    },
+    {
+      "epoch": 4.103392568659127,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.041360907271515e-05,
+      "loss": 1.1809,
+      "step": 4445
+    },
+    {
+      "epoch": 4.104315716593584,
+      "grad_norm": 3.970703125,
+      "learning_rate": 7.040693795863909e-05,
+      "loss": 1.4958,
+      "step": 4446
+    },
+    {
+      "epoch": 4.10523886452804,
+      "grad_norm": 4.3046875,
+      "learning_rate": 7.040026684456305e-05,
+      "loss": 1.5563,
+      "step": 4447
+    },
+    {
+      "epoch": 4.106162012462497,
+      "grad_norm": 5.0,
+      "learning_rate": 7.0393595730487e-05,
+      "loss": 1.3718,
+      "step": 4448
+    },
+    {
+      "epoch": 4.107085160396954,
+      "grad_norm": 5.515625,
+      "learning_rate": 7.038692461641094e-05,
+      "loss": 1.4355,
+      "step": 4449
+    },
+    {
+      "epoch": 4.10800830833141,
+      "grad_norm": 4.93359375,
+      "learning_rate": 7.038025350233489e-05,
+      "loss": 1.5701,
+      "step": 4450
+    },
+    {
+      "epoch": 4.108931456265867,
+      "grad_norm": 3.619140625,
+      "learning_rate": 7.037358238825885e-05,
+      "loss": 1.4768,
+      "step": 4451
+    },
+    {
+      "epoch": 4.109854604200323,
+      "grad_norm": 4.16015625,
+      "learning_rate": 7.036691127418279e-05,
+      "loss": 1.6934,
+      "step": 4452
+    },
+    {
+      "epoch": 4.11077775213478,
+      "grad_norm": 4.265625,
+      "learning_rate": 7.036024016010674e-05,
+      "loss": 1.3308,
+      "step": 4453
+    },
+    {
+      "epoch": 4.111700900069236,
+      "grad_norm": 3.607421875,
+      "learning_rate": 7.035356904603068e-05,
+      "loss": 1.4067,
+      "step": 4454
+    },
+    {
+      "epoch": 4.112624048003693,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.034689793195464e-05,
+      "loss": 1.2488,
+      "step": 4455
+    },
+    {
+      "epoch": 4.113547195938149,
+      "grad_norm": 3.603515625,
+      "learning_rate": 7.034022681787859e-05,
+      "loss": 1.4788,
+      "step": 4456
+    },
+    {
+      "epoch": 4.114470343872606,
+      "grad_norm": 4.46484375,
+      "learning_rate": 7.033355570380253e-05,
+      "loss": 1.3838,
+      "step": 4457
+    },
+    {
+      "epoch": 4.115393491807062,
+      "grad_norm": 3.67578125,
+      "learning_rate": 7.032688458972648e-05,
+      "loss": 1.4141,
+      "step": 4458
+    },
+    {
+      "epoch": 4.116316639741519,
+      "grad_norm": 3.576171875,
+      "learning_rate": 7.032021347565044e-05,
+      "loss": 1.3726,
+      "step": 4459
+    },
+    {
+      "epoch": 4.117239787675975,
+      "grad_norm": 3.9453125,
+      "learning_rate": 7.03135423615744e-05,
+      "loss": 1.6528,
+      "step": 4460
+    },
+    {
+      "epoch": 4.118162935610432,
+      "grad_norm": 3.96484375,
+      "learning_rate": 7.030687124749834e-05,
+      "loss": 1.3674,
+      "step": 4461
+    },
+    {
+      "epoch": 4.119086083544888,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.030020013342229e-05,
+      "loss": 1.5156,
+      "step": 4462
+    },
+    {
+      "epoch": 4.120009231479345,
+      "grad_norm": 4.70703125,
+      "learning_rate": 7.029352901934623e-05,
+      "loss": 1.5671,
+      "step": 4463
+    },
+    {
+      "epoch": 4.120932379413801,
+      "grad_norm": 3.857421875,
+      "learning_rate": 7.028685790527019e-05,
+      "loss": 1.5835,
+      "step": 4464
+    },
+    {
+      "epoch": 4.1218555273482576,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.028018679119414e-05,
+      "loss": 1.6758,
+      "step": 4465
+    },
+    {
+      "epoch": 4.1227786752827145,
+      "grad_norm": 4.08203125,
+      "learning_rate": 7.027351567711808e-05,
+      "loss": 1.4333,
+      "step": 4466
+    },
+    {
+      "epoch": 4.1237018232171705,
+      "grad_norm": 4.8828125,
+      "learning_rate": 7.026684456304203e-05,
+      "loss": 1.2448,
+      "step": 4467
+    },
+    {
+      "epoch": 4.124624971151627,
+      "grad_norm": 3.951171875,
+      "learning_rate": 7.026017344896599e-05,
+      "loss": 1.3137,
+      "step": 4468
+    },
+    {
+      "epoch": 4.125548119086083,
+      "grad_norm": 4.6328125,
+      "learning_rate": 7.025350233488993e-05,
+      "loss": 1.4172,
+      "step": 4469
+    },
+    {
+      "epoch": 4.12647126702054,
+      "grad_norm": 4.0859375,
+      "learning_rate": 7.024683122081388e-05,
+      "loss": 1.5779,
+      "step": 4470
+    },
+    {
+      "epoch": 4.127394414954996,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.024016010673782e-05,
+      "loss": 1.4795,
+      "step": 4471
+    },
+    {
+      "epoch": 4.128317562889453,
+      "grad_norm": 3.87109375,
+      "learning_rate": 7.023348899266178e-05,
+      "loss": 1.4944,
+      "step": 4472
+    },
+    {
+      "epoch": 4.129240710823909,
+      "grad_norm": 3.697265625,
+      "learning_rate": 7.022681787858573e-05,
+      "loss": 1.3414,
+      "step": 4473
+    },
+    {
+      "epoch": 4.130163858758366,
+      "grad_norm": 3.716796875,
+      "learning_rate": 7.022014676450967e-05,
+      "loss": 1.4524,
+      "step": 4474
+    },
+    {
+      "epoch": 4.131087006692822,
+      "grad_norm": 3.990234375,
+      "learning_rate": 7.021347565043362e-05,
+      "loss": 1.3413,
+      "step": 4475
+    },
+    {
+      "epoch": 4.132010154627279,
+      "grad_norm": 3.966796875,
+      "learning_rate": 7.020680453635758e-05,
+      "loss": 1.3302,
+      "step": 4476
+    },
+    {
+      "epoch": 4.132933302561735,
+      "grad_norm": 3.912109375,
+      "learning_rate": 7.020013342228152e-05,
+      "loss": 1.1935,
+      "step": 4477
+    },
+    {
+      "epoch": 4.133856450496192,
+      "grad_norm": 4.58203125,
+      "learning_rate": 7.019346230820547e-05,
+      "loss": 1.3728,
+      "step": 4478
+    },
+    {
+      "epoch": 4.134779598430648,
+      "grad_norm": 5.09765625,
+      "learning_rate": 7.018679119412941e-05,
+      "loss": 1.1763,
+      "step": 4479
+    },
+    {
+      "epoch": 4.135702746365105,
+      "grad_norm": 3.6015625,
+      "learning_rate": 7.018012008005337e-05,
+      "loss": 1.439,
+      "step": 4480
+    },
+    {
+      "epoch": 4.136625894299561,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.017344896597732e-05,
+      "loss": 1.1082,
+      "step": 4481
+    },
+    {
+      "epoch": 4.137549042234018,
+      "grad_norm": 3.666015625,
+      "learning_rate": 7.016677785190126e-05,
+      "loss": 0.9514,
+      "step": 4482
+    },
+    {
+      "epoch": 4.138472190168475,
+      "grad_norm": 4.37109375,
+      "learning_rate": 7.016010673782521e-05,
+      "loss": 1.7983,
+      "step": 4483
+    },
+    {
+      "epoch": 4.139395338102931,
+      "grad_norm": 3.72265625,
+      "learning_rate": 7.015343562374917e-05,
+      "loss": 1.0682,
+      "step": 4484
+    },
+    {
+      "epoch": 4.140318486037388,
+      "grad_norm": 4.109375,
+      "learning_rate": 7.014676450967313e-05,
+      "loss": 1.3462,
+      "step": 4485
+    },
+    {
+      "epoch": 4.141241633971844,
+      "grad_norm": 4.00390625,
+      "learning_rate": 7.014009339559707e-05,
+      "loss": 1.3618,
+      "step": 4486
+    },
+    {
+      "epoch": 4.142164781906301,
+      "grad_norm": 5.1015625,
+      "learning_rate": 7.013342228152102e-05,
+      "loss": 1.616,
+      "step": 4487
+    },
+    {
+      "epoch": 4.143087929840757,
+      "grad_norm": 4.125,
+      "learning_rate": 7.012675116744498e-05,
+      "loss": 1.6111,
+      "step": 4488
+    },
+    {
+      "epoch": 4.144011077775214,
+      "grad_norm": 4.2890625,
+      "learning_rate": 7.012008005336892e-05,
+      "loss": 1.2018,
+      "step": 4489
+    },
+    {
+      "epoch": 4.14493422570967,
+      "grad_norm": 4.3046875,
+      "learning_rate": 7.011340893929287e-05,
+      "loss": 1.2668,
+      "step": 4490
+    },
+    {
+      "epoch": 4.145857373644127,
+      "grad_norm": 3.275390625,
+      "learning_rate": 7.010673782521681e-05,
+      "loss": 1.5547,
+      "step": 4491
+    },
+    {
+      "epoch": 4.146780521578583,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.010006671114077e-05,
+      "loss": 1.5002,
+      "step": 4492
+    },
+    {
+      "epoch": 4.14770366951304,
+      "grad_norm": 3.779296875,
+      "learning_rate": 7.009339559706472e-05,
+      "loss": 1.5806,
+      "step": 4493
+    },
+    {
+      "epoch": 4.148626817447496,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.008672448298866e-05,
+      "loss": 1.1924,
+      "step": 4494
+    },
+    {
+      "epoch": 4.1495499653819525,
+      "grad_norm": 4.109375,
+      "learning_rate": 7.008005336891261e-05,
+      "loss": 1.4829,
+      "step": 4495
+    },
+    {
+      "epoch": 4.150473113316409,
+      "grad_norm": 4.58203125,
+      "learning_rate": 7.007338225483657e-05,
+      "loss": 1.5239,
+      "step": 4496
+    },
+    {
+      "epoch": 4.1513962612508655,
+      "grad_norm": 3.841796875,
+      "learning_rate": 7.006671114076051e-05,
+      "loss": 1.2061,
+      "step": 4497
+    },
+    {
+      "epoch": 4.152319409185322,
+      "grad_norm": 3.580078125,
+      "learning_rate": 7.006004002668446e-05,
+      "loss": 1.7239,
+      "step": 4498
+    },
+    {
+      "epoch": 4.153242557119778,
+      "grad_norm": 3.8203125,
+      "learning_rate": 7.00533689126084e-05,
+      "loss": 1.1971,
+      "step": 4499
+    },
+    {
+      "epoch": 4.154165705054235,
+      "grad_norm": 4.1640625,
+      "learning_rate": 7.004669779853236e-05,
+      "loss": 1.3477,
+      "step": 4500
+    },
+    {
+      "epoch": 4.155088852988691,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.004002668445631e-05,
+      "loss": 1.355,
+      "step": 4501
+    },
+    {
+      "epoch": 4.156012000923148,
+      "grad_norm": 3.568359375,
+      "learning_rate": 7.003335557038025e-05,
+      "loss": 1.2056,
+      "step": 4502
+    },
+    {
+      "epoch": 4.156935148857604,
+      "grad_norm": 3.498046875,
+      "learning_rate": 7.00266844563042e-05,
+      "loss": 1.1518,
+      "step": 4503
+    },
+    {
+      "epoch": 4.157858296792061,
+      "grad_norm": 3.5390625,
+      "learning_rate": 7.002001334222816e-05,
+      "loss": 1.1841,
+      "step": 4504
+    },
+    {
+      "epoch": 4.158781444726517,
+      "grad_norm": 3.712890625,
+      "learning_rate": 7.00133422281521e-05,
+      "loss": 1.4915,
+      "step": 4505
+    },
+    {
+      "epoch": 4.159704592660974,
+      "grad_norm": 3.7734375,
+      "learning_rate": 7.000667111407605e-05,
+      "loss": 1.6882,
+      "step": 4506
+    },
+    {
+      "epoch": 4.16062774059543,
+      "grad_norm": 4.14453125,
+      "learning_rate": 7e-05,
+      "loss": 1.7148,
+      "step": 4507
+    },
+    {
+      "epoch": 4.161550888529887,
+      "grad_norm": 3.24609375,
+      "learning_rate": 6.999332888592395e-05,
+      "loss": 1.446,
+      "step": 4508
+    },
+    {
+      "epoch": 4.162474036464343,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.99866577718479e-05,
+      "loss": 1.6897,
+      "step": 4509
+    },
+    {
+      "epoch": 4.1633971843988,
+      "grad_norm": 4.3828125,
+      "learning_rate": 6.997998665777186e-05,
+      "loss": 1.3191,
+      "step": 4510
+    },
+    {
+      "epoch": 4.164320332333256,
+      "grad_norm": 4.45703125,
+      "learning_rate": 6.99733155436958e-05,
+      "loss": 1.6875,
+      "step": 4511
+    },
+    {
+      "epoch": 4.165243480267713,
+      "grad_norm": 4.0390625,
+      "learning_rate": 6.996664442961975e-05,
+      "loss": 1.3899,
+      "step": 4512
+    },
+    {
+      "epoch": 4.166166628202169,
+      "grad_norm": 4.6640625,
+      "learning_rate": 6.995997331554371e-05,
+      "loss": 0.6849,
+      "step": 4513
+    },
+    {
+      "epoch": 4.167089776136626,
+      "grad_norm": 4.328125,
+      "learning_rate": 6.995330220146765e-05,
+      "loss": 1.1295,
+      "step": 4514
+    },
+    {
+      "epoch": 4.168012924071083,
+      "grad_norm": 4.703125,
+      "learning_rate": 6.99466310873916e-05,
+      "loss": 1.4031,
+      "step": 4515
+    },
+    {
+      "epoch": 4.168936072005539,
+      "grad_norm": 4.40234375,
+      "learning_rate": 6.993995997331554e-05,
+      "loss": 1.533,
+      "step": 4516
+    },
+    {
+      "epoch": 4.169859219939996,
+      "grad_norm": 5.07421875,
+      "learning_rate": 6.99332888592395e-05,
+      "loss": 1.5327,
+      "step": 4517
+    },
+    {
+      "epoch": 4.170782367874452,
+      "grad_norm": 6.06640625,
+      "learning_rate": 6.992661774516345e-05,
+      "loss": 1.4792,
+      "step": 4518
+    },
+    {
+      "epoch": 4.171705515808909,
+      "grad_norm": 4.83984375,
+      "learning_rate": 6.99199466310874e-05,
+      "loss": 1.1526,
+      "step": 4519
+    },
+    {
+      "epoch": 4.172628663743365,
+      "grad_norm": 4.828125,
+      "learning_rate": 6.991327551701134e-05,
+      "loss": 1.4961,
+      "step": 4520
+    },
+    {
+      "epoch": 4.173551811677822,
+      "grad_norm": 4.3984375,
+      "learning_rate": 6.99066044029353e-05,
+      "loss": 1.6045,
+      "step": 4521
+    },
+    {
+      "epoch": 4.174474959612278,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.989993328885924e-05,
+      "loss": 1.3787,
+      "step": 4522
+    },
+    {
+      "epoch": 4.175398107546735,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.989326217478319e-05,
+      "loss": 1.5222,
+      "step": 4523
+    },
+    {
+      "epoch": 4.176321255481191,
+      "grad_norm": 3.9765625,
+      "learning_rate": 6.988659106070713e-05,
+      "loss": 1.3435,
+      "step": 4524
+    },
+    {
+      "epoch": 4.1772444034156475,
+      "grad_norm": 4.90625,
+      "learning_rate": 6.987991994663109e-05,
+      "loss": 1.562,
+      "step": 4525
+    },
+    {
+      "epoch": 4.1781675513501035,
+      "grad_norm": 3.86328125,
+      "learning_rate": 6.987324883255504e-05,
+      "loss": 1.6162,
+      "step": 4526
+    },
+    {
+      "epoch": 4.1790906992845605,
+      "grad_norm": 3.623046875,
+      "learning_rate": 6.986657771847898e-05,
+      "loss": 1.3816,
+      "step": 4527
+    },
+    {
+      "epoch": 4.1800138472190165,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.985990660440293e-05,
+      "loss": 1.2684,
+      "step": 4528
+    },
+    {
+      "epoch": 4.180936995153473,
+      "grad_norm": 4.51171875,
+      "learning_rate": 6.985323549032689e-05,
+      "loss": 1.395,
+      "step": 4529
+    },
+    {
+      "epoch": 4.181860143087929,
+      "grad_norm": 4.328125,
+      "learning_rate": 6.984656437625083e-05,
+      "loss": 1.4365,
+      "step": 4530
+    },
+    {
+      "epoch": 4.182783291022386,
+      "grad_norm": 3.76953125,
+      "learning_rate": 6.983989326217478e-05,
+      "loss": 1.2635,
+      "step": 4531
+    },
+    {
+      "epoch": 4.183706438956843,
+      "grad_norm": 4.03515625,
+      "learning_rate": 6.983322214809874e-05,
+      "loss": 1.5276,
+      "step": 4532
+    },
+    {
+      "epoch": 4.184629586891299,
+      "grad_norm": 4.0,
+      "learning_rate": 6.982655103402268e-05,
+      "loss": 1.5254,
+      "step": 4533
+    },
+    {
+      "epoch": 4.185552734825756,
+      "grad_norm": 3.1328125,
+      "learning_rate": 6.981987991994663e-05,
+      "loss": 1.2423,
+      "step": 4534
+    },
+    {
+      "epoch": 4.186475882760212,
+      "grad_norm": 3.763671875,
+      "learning_rate": 6.981320880587059e-05,
+      "loss": 1.051,
+      "step": 4535
+    },
+    {
+      "epoch": 4.187399030694669,
+      "grad_norm": 5.13671875,
+      "learning_rate": 6.980653769179453e-05,
+      "loss": 1.3501,
+      "step": 4536
+    },
+    {
+      "epoch": 4.188322178629125,
+      "grad_norm": 4.3984375,
+      "learning_rate": 6.979986657771849e-05,
+      "loss": 1.2451,
+      "step": 4537
+    },
+    {
+      "epoch": 4.189245326563582,
+      "grad_norm": 4.17578125,
+      "learning_rate": 6.979319546364244e-05,
+      "loss": 1.5295,
+      "step": 4538
+    },
+    {
+      "epoch": 4.190168474498038,
+      "grad_norm": 4.0859375,
+      "learning_rate": 6.978652434956638e-05,
+      "loss": 1.2664,
+      "step": 4539
+    },
+    {
+      "epoch": 4.191091622432495,
+      "grad_norm": 3.962890625,
+      "learning_rate": 6.977985323549033e-05,
+      "loss": 1.184,
+      "step": 4540
+    },
+    {
+      "epoch": 4.192014770366951,
+      "grad_norm": 3.748046875,
+      "learning_rate": 6.977318212141429e-05,
+      "loss": 1.5212,
+      "step": 4541
+    },
+    {
+      "epoch": 4.192937918301408,
+      "grad_norm": 3.361328125,
+      "learning_rate": 6.976651100733823e-05,
+      "loss": 1.3391,
+      "step": 4542
+    },
+    {
+      "epoch": 4.193861066235864,
+      "grad_norm": 3.826171875,
+      "learning_rate": 6.975983989326218e-05,
+      "loss": 1.7493,
+      "step": 4543
+    },
+    {
+      "epoch": 4.194784214170321,
+      "grad_norm": 4.3046875,
+      "learning_rate": 6.975316877918612e-05,
+      "loss": 1.2778,
+      "step": 4544
+    },
+    {
+      "epoch": 4.195707362104777,
+      "grad_norm": 4.28515625,
+      "learning_rate": 6.974649766511008e-05,
+      "loss": 1.2086,
+      "step": 4545
+    },
+    {
+      "epoch": 4.196630510039234,
+      "grad_norm": 4.484375,
+      "learning_rate": 6.973982655103403e-05,
+      "loss": 1.4719,
+      "step": 4546
+    },
+    {
+      "epoch": 4.197553657973691,
+      "grad_norm": 3.458984375,
+      "learning_rate": 6.973315543695797e-05,
+      "loss": 1.3884,
+      "step": 4547
+    },
+    {
+      "epoch": 4.198476805908147,
+      "grad_norm": 3.615234375,
+      "learning_rate": 6.972648432288192e-05,
+      "loss": 1.3352,
+      "step": 4548
+    },
+    {
+      "epoch": 4.199399953842604,
+      "grad_norm": 5.046875,
+      "learning_rate": 6.971981320880588e-05,
+      "loss": 1.407,
+      "step": 4549
+    },
+    {
+      "epoch": 4.20032310177706,
+      "grad_norm": 4.25390625,
+      "learning_rate": 6.971314209472982e-05,
+      "loss": 1.4341,
+      "step": 4550
+    },
+    {
+      "epoch": 4.201246249711517,
+      "grad_norm": 4.2421875,
+      "learning_rate": 6.970647098065377e-05,
+      "loss": 1.3647,
+      "step": 4551
+    },
+    {
+      "epoch": 4.202169397645973,
+      "grad_norm": 4.07421875,
+      "learning_rate": 6.969979986657771e-05,
+      "loss": 1.6519,
+      "step": 4552
+    },
+    {
+      "epoch": 4.2030925455804296,
+      "grad_norm": 3.71484375,
+      "learning_rate": 6.969312875250167e-05,
+      "loss": 0.9843,
+      "step": 4553
+    },
+    {
+      "epoch": 4.204015693514886,
+      "grad_norm": 4.64453125,
+      "learning_rate": 6.968645763842562e-05,
+      "loss": 1.4714,
+      "step": 4554
+    },
+    {
+      "epoch": 4.2049388414493425,
+      "grad_norm": 4.0390625,
+      "learning_rate": 6.967978652434956e-05,
+      "loss": 1.3357,
+      "step": 4555
+    },
+    {
+      "epoch": 4.2058619893837985,
+      "grad_norm": 3.712890625,
+      "learning_rate": 6.967311541027351e-05,
+      "loss": 1.5479,
+      "step": 4556
+    },
+    {
+      "epoch": 4.206785137318255,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.966644429619747e-05,
+      "loss": 0.9452,
+      "step": 4557
+    },
+    {
+      "epoch": 4.2077082852527115,
+      "grad_norm": 4.22265625,
+      "learning_rate": 6.965977318212141e-05,
+      "loss": 1.3993,
+      "step": 4558
+    },
+    {
+      "epoch": 4.208631433187168,
+      "grad_norm": 3.263671875,
+      "learning_rate": 6.965310206804536e-05,
+      "loss": 1.4919,
+      "step": 4559
+    },
+    {
+      "epoch": 4.209554581121624,
+      "grad_norm": 4.0859375,
+      "learning_rate": 6.964643095396932e-05,
+      "loss": 1.0605,
+      "step": 4560
+    },
+    {
+      "epoch": 4.210477729056081,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.963975983989326e-05,
+      "loss": 1.564,
+      "step": 4561
+    },
+    {
+      "epoch": 4.211400876990537,
+      "grad_norm": 4.17578125,
+      "learning_rate": 6.963308872581722e-05,
+      "loss": 1.4293,
+      "step": 4562
+    },
+    {
+      "epoch": 4.212324024924994,
+      "grad_norm": 3.482421875,
+      "learning_rate": 6.962641761174117e-05,
+      "loss": 1.5588,
+      "step": 4563
+    },
+    {
+      "epoch": 4.21324717285945,
+      "grad_norm": 4.70703125,
+      "learning_rate": 6.961974649766511e-05,
+      "loss": 1.469,
+      "step": 4564
+    },
+    {
+      "epoch": 4.214170320793907,
+      "grad_norm": 3.314453125,
+      "learning_rate": 6.961307538358906e-05,
+      "loss": 1.1264,
+      "step": 4565
+    },
+    {
+      "epoch": 4.215093468728364,
+      "grad_norm": 4.609375,
+      "learning_rate": 6.960640426951302e-05,
+      "loss": 1.2271,
+      "step": 4566
+    },
+    {
+      "epoch": 4.21601661666282,
+      "grad_norm": 3.779296875,
+      "learning_rate": 6.959973315543696e-05,
+      "loss": 1.344,
+      "step": 4567
+    },
+    {
+      "epoch": 4.216939764597277,
+      "grad_norm": 3.42578125,
+      "learning_rate": 6.959306204136091e-05,
+      "loss": 1.2126,
+      "step": 4568
+    },
+    {
+      "epoch": 4.217862912531733,
+      "grad_norm": 3.78515625,
+      "learning_rate": 6.958639092728487e-05,
+      "loss": 1.4802,
+      "step": 4569
+    },
+    {
+      "epoch": 4.21878606046619,
+      "grad_norm": 4.07421875,
+      "learning_rate": 6.957971981320881e-05,
+      "loss": 1.6453,
+      "step": 4570
+    },
+    {
+      "epoch": 4.219709208400646,
+      "grad_norm": 3.8828125,
+      "learning_rate": 6.957304869913276e-05,
+      "loss": 1.5076,
+      "step": 4571
+    },
+    {
+      "epoch": 4.220632356335103,
+      "grad_norm": 3.66015625,
+      "learning_rate": 6.95663775850567e-05,
+      "loss": 1.3452,
+      "step": 4572
+    },
+    {
+      "epoch": 4.221555504269559,
+      "grad_norm": 4.0390625,
+      "learning_rate": 6.955970647098066e-05,
+      "loss": 1.2119,
+      "step": 4573
+    },
+    {
+      "epoch": 4.222478652204016,
+      "grad_norm": 4.640625,
+      "learning_rate": 6.955303535690461e-05,
+      "loss": 1.1686,
+      "step": 4574
+    },
+    {
+      "epoch": 4.223401800138472,
+      "grad_norm": 4.09765625,
+      "learning_rate": 6.954636424282855e-05,
+      "loss": 1.4656,
+      "step": 4575
+    },
+    {
+      "epoch": 4.224324948072929,
+      "grad_norm": 3.744140625,
+      "learning_rate": 6.95396931287525e-05,
+      "loss": 1.3306,
+      "step": 4576
+    },
+    {
+      "epoch": 4.225248096007385,
+      "grad_norm": 4.07421875,
+      "learning_rate": 6.953302201467646e-05,
+      "loss": 1.5459,
+      "step": 4577
+    },
+    {
+      "epoch": 4.226171243941842,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.95263509006004e-05,
+      "loss": 1.2758,
+      "step": 4578
+    },
+    {
+      "epoch": 4.227094391876298,
+      "grad_norm": 3.93359375,
+      "learning_rate": 6.951967978652435e-05,
+      "loss": 1.4753,
+      "step": 4579
+    },
+    {
+      "epoch": 4.228017539810755,
+      "grad_norm": 4.1171875,
+      "learning_rate": 6.95130086724483e-05,
+      "loss": 1.3035,
+      "step": 4580
+    },
+    {
+      "epoch": 4.228940687745212,
+      "grad_norm": 3.880859375,
+      "learning_rate": 6.950633755837225e-05,
+      "loss": 1.2279,
+      "step": 4581
+    },
+    {
+      "epoch": 4.229863835679668,
+      "grad_norm": 4.49609375,
+      "learning_rate": 6.94996664442962e-05,
+      "loss": 1.3401,
+      "step": 4582
+    },
+    {
+      "epoch": 4.2307869836141245,
+      "grad_norm": 4.36328125,
+      "learning_rate": 6.949299533022014e-05,
+      "loss": 1.4622,
+      "step": 4583
+    },
+    {
+      "epoch": 4.2317101315485806,
+      "grad_norm": 4.17578125,
+      "learning_rate": 6.948632421614409e-05,
+      "loss": 1.0599,
+      "step": 4584
+    },
+    {
+      "epoch": 4.2326332794830375,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.947965310206805e-05,
+      "loss": 1.3617,
+      "step": 4585
+    },
+    {
+      "epoch": 4.2335564274174935,
+      "grad_norm": 3.962890625,
+      "learning_rate": 6.947298198799201e-05,
+      "loss": 1.3123,
+      "step": 4586
+    },
+    {
+      "epoch": 4.23447957535195,
+      "grad_norm": 4.08984375,
+      "learning_rate": 6.946631087391595e-05,
+      "loss": 1.217,
+      "step": 4587
+    },
+    {
+      "epoch": 4.235402723286406,
+      "grad_norm": 3.65234375,
+      "learning_rate": 6.94596397598399e-05,
+      "loss": 1.2185,
+      "step": 4588
+    },
+    {
+      "epoch": 4.236325871220863,
+      "grad_norm": 4.73046875,
+      "learning_rate": 6.945296864576384e-05,
+      "loss": 1.407,
+      "step": 4589
+    },
+    {
+      "epoch": 4.237249019155319,
+      "grad_norm": 4.171875,
+      "learning_rate": 6.94462975316878e-05,
+      "loss": 1.6831,
+      "step": 4590
+    },
+    {
+      "epoch": 4.238172167089776,
+      "grad_norm": 3.5078125,
+      "learning_rate": 6.943962641761175e-05,
+      "loss": 1.2494,
+      "step": 4591
+    },
+    {
+      "epoch": 4.239095315024232,
+      "grad_norm": 3.892578125,
+      "learning_rate": 6.94329553035357e-05,
+      "loss": 1.5471,
+      "step": 4592
+    },
+    {
+      "epoch": 4.240018462958689,
+      "grad_norm": 3.9453125,
+      "learning_rate": 6.942628418945964e-05,
+      "loss": 1.4265,
+      "step": 4593
+    },
+    {
+      "epoch": 4.240941610893145,
+      "grad_norm": 5.3046875,
+      "learning_rate": 6.94196130753836e-05,
+      "loss": 1.1759,
+      "step": 4594
+    },
+    {
+      "epoch": 4.241864758827602,
+      "grad_norm": 4.4921875,
+      "learning_rate": 6.941294196130754e-05,
+      "loss": 1.2316,
+      "step": 4595
+    },
+    {
+      "epoch": 4.242787906762058,
+      "grad_norm": 4.203125,
+      "learning_rate": 6.940627084723149e-05,
+      "loss": 1.6262,
+      "step": 4596
+    },
+    {
+      "epoch": 4.243711054696515,
+      "grad_norm": 4.05859375,
+      "learning_rate": 6.939959973315543e-05,
+      "loss": 1.4359,
+      "step": 4597
+    },
+    {
+      "epoch": 4.244634202630972,
+      "grad_norm": 3.548828125,
+      "learning_rate": 6.93929286190794e-05,
+      "loss": 1.5029,
+      "step": 4598
+    },
+    {
+      "epoch": 4.245557350565428,
+      "grad_norm": 3.83203125,
+      "learning_rate": 6.938625750500334e-05,
+      "loss": 1.1674,
+      "step": 4599
+    },
+    {
+      "epoch": 4.246480498499885,
+      "grad_norm": 4.23828125,
+      "learning_rate": 6.937958639092728e-05,
+      "loss": 1.3196,
+      "step": 4600
+    },
+    {
+      "epoch": 4.247403646434341,
+      "grad_norm": 3.654296875,
+      "learning_rate": 6.937291527685123e-05,
+      "loss": 1.4607,
+      "step": 4601
+    },
+    {
+      "epoch": 4.248326794368798,
+      "grad_norm": 3.9765625,
+      "learning_rate": 6.936624416277519e-05,
+      "loss": 1.1858,
+      "step": 4602
+    },
+    {
+      "epoch": 4.249249942303254,
+      "grad_norm": 3.935546875,
+      "learning_rate": 6.935957304869913e-05,
+      "loss": 1.4124,
+      "step": 4603
+    },
+    {
+      "epoch": 4.250173090237711,
+      "grad_norm": 4.05078125,
+      "learning_rate": 6.935290193462308e-05,
+      "loss": 1.3567,
+      "step": 4604
+    },
+    {
+      "epoch": 4.251096238172167,
+      "grad_norm": 4.40234375,
+      "learning_rate": 6.934623082054703e-05,
+      "loss": 1.5217,
+      "step": 4605
+    },
+    {
+      "epoch": 4.252019386106624,
+      "grad_norm": 5.3359375,
+      "learning_rate": 6.933955970647098e-05,
+      "loss": 1.4458,
+      "step": 4606
+    },
+    {
+      "epoch": 4.25294253404108,
+      "grad_norm": 3.759765625,
+      "learning_rate": 6.933288859239493e-05,
+      "loss": 1.1671,
+      "step": 4607
+    },
+    {
+      "epoch": 4.253865681975537,
+      "grad_norm": 3.974609375,
+      "learning_rate": 6.932621747831888e-05,
+      "loss": 1.4551,
+      "step": 4608
+    },
+    {
+      "epoch": 4.254788829909993,
+      "grad_norm": 4.29296875,
+      "learning_rate": 6.931954636424282e-05,
+      "loss": 0.9715,
+      "step": 4609
+    },
+    {
+      "epoch": 4.25571197784445,
+      "grad_norm": 4.6015625,
+      "learning_rate": 6.93128752501668e-05,
+      "loss": 1.2612,
+      "step": 4610
+    },
+    {
+      "epoch": 4.256635125778906,
+      "grad_norm": 3.841796875,
+      "learning_rate": 6.930620413609074e-05,
+      "loss": 1.3867,
+      "step": 4611
+    },
+    {
+      "epoch": 4.257558273713363,
+      "grad_norm": 3.31640625,
+      "learning_rate": 6.929953302201468e-05,
+      "loss": 1.646,
+      "step": 4612
+    },
+    {
+      "epoch": 4.258481421647819,
+      "grad_norm": 3.556640625,
+      "learning_rate": 6.929286190793863e-05,
+      "loss": 1.2744,
+      "step": 4613
+    },
+    {
+      "epoch": 4.2594045695822755,
+      "grad_norm": 3.66796875,
+      "learning_rate": 6.928619079386259e-05,
+      "loss": 1.6409,
+      "step": 4614
+    },
+    {
+      "epoch": 4.2603277175167324,
+      "grad_norm": 3.580078125,
+      "learning_rate": 6.927951967978653e-05,
+      "loss": 1.6289,
+      "step": 4615
+    },
+    {
+      "epoch": 4.2612508654511885,
+      "grad_norm": 3.583984375,
+      "learning_rate": 6.927284856571048e-05,
+      "loss": 1.5054,
+      "step": 4616
+    },
+    {
+      "epoch": 4.262174013385645,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.926617745163442e-05,
+      "loss": 1.1571,
+      "step": 4617
+    },
+    {
+      "epoch": 4.263097161320101,
+      "grad_norm": 4.078125,
+      "learning_rate": 6.925950633755838e-05,
+      "loss": 1.3887,
+      "step": 4618
+    },
+    {
+      "epoch": 4.264020309254558,
+      "grad_norm": 3.888671875,
+      "learning_rate": 6.925283522348233e-05,
+      "loss": 1.5488,
+      "step": 4619
+    },
+    {
+      "epoch": 4.264943457189014,
+      "grad_norm": 3.595703125,
+      "learning_rate": 6.924616410940627e-05,
+      "loss": 1.1548,
+      "step": 4620
+    },
+    {
+      "epoch": 4.265866605123471,
+      "grad_norm": 4.85546875,
+      "learning_rate": 6.923949299533022e-05,
+      "loss": 1.3704,
+      "step": 4621
+    },
+    {
+      "epoch": 4.266789753057927,
+      "grad_norm": 4.26171875,
+      "learning_rate": 6.923282188125418e-05,
+      "loss": 1.3956,
+      "step": 4622
+    },
+    {
+      "epoch": 4.267712900992384,
+      "grad_norm": 4.02734375,
+      "learning_rate": 6.922615076717812e-05,
+      "loss": 1.4026,
+      "step": 4623
+    },
+    {
+      "epoch": 4.26863604892684,
+      "grad_norm": 3.33203125,
+      "learning_rate": 6.921947965310207e-05,
+      "loss": 1.4255,
+      "step": 4624
+    },
+    {
+      "epoch": 4.269559196861297,
+      "grad_norm": 5.02734375,
+      "learning_rate": 6.921280853902602e-05,
+      "loss": 1.4325,
+      "step": 4625
+    },
+    {
+      "epoch": 4.270482344795753,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.920613742494997e-05,
+      "loss": 1.6746,
+      "step": 4626
+    },
+    {
+      "epoch": 4.27140549273021,
+      "grad_norm": 3.673828125,
+      "learning_rate": 6.919946631087392e-05,
+      "loss": 1.3641,
+      "step": 4627
+    },
+    {
+      "epoch": 4.272328640664666,
+      "grad_norm": 3.72265625,
+      "learning_rate": 6.919279519679787e-05,
+      "loss": 1.3782,
+      "step": 4628
+    },
+    {
+      "epoch": 4.273251788599123,
+      "grad_norm": 3.7890625,
+      "learning_rate": 6.918612408272181e-05,
+      "loss": 1.594,
+      "step": 4629
+    },
+    {
+      "epoch": 4.27417493653358,
+      "grad_norm": 3.525390625,
+      "learning_rate": 6.917945296864577e-05,
+      "loss": 1.5566,
+      "step": 4630
+    },
+    {
+      "epoch": 4.275098084468036,
+      "grad_norm": 4.6015625,
+      "learning_rate": 6.917278185456972e-05,
+      "loss": 1.4314,
+      "step": 4631
+    },
+    {
+      "epoch": 4.276021232402493,
+      "grad_norm": 5.85546875,
+      "learning_rate": 6.916611074049366e-05,
+      "loss": 1.3623,
+      "step": 4632
+    },
+    {
+      "epoch": 4.276944380336949,
+      "grad_norm": 3.95703125,
+      "learning_rate": 6.91594396264176e-05,
+      "loss": 1.498,
+      "step": 4633
+    },
+    {
+      "epoch": 4.277867528271406,
+      "grad_norm": 4.50390625,
+      "learning_rate": 6.915276851234156e-05,
+      "loss": 1.5647,
+      "step": 4634
+    },
+    {
+      "epoch": 4.278790676205862,
+      "grad_norm": 2.955078125,
+      "learning_rate": 6.914609739826552e-05,
+      "loss": 1.1209,
+      "step": 4635
+    },
+    {
+      "epoch": 4.279713824140319,
+      "grad_norm": 3.599609375,
+      "learning_rate": 6.913942628418947e-05,
+      "loss": 1.5312,
+      "step": 4636
+    },
+    {
+      "epoch": 4.280636972074775,
+      "grad_norm": 4.0703125,
+      "learning_rate": 6.913275517011341e-05,
+      "loss": 1.2921,
+      "step": 4637
+    },
+    {
+      "epoch": 4.281560120009232,
+      "grad_norm": 4.4765625,
+      "learning_rate": 6.912608405603736e-05,
+      "loss": 1.5015,
+      "step": 4638
+    },
+    {
+      "epoch": 4.282483267943688,
+      "grad_norm": 3.716796875,
+      "learning_rate": 6.911941294196132e-05,
+      "loss": 1.407,
+      "step": 4639
+    },
+    {
+      "epoch": 4.283406415878145,
+      "grad_norm": 3.75390625,
+      "learning_rate": 6.911274182788526e-05,
+      "loss": 1.5352,
+      "step": 4640
+    },
+    {
+      "epoch": 4.284329563812601,
+      "grad_norm": 3.1015625,
+      "learning_rate": 6.910607071380921e-05,
+      "loss": 1.502,
+      "step": 4641
+    },
+    {
+      "epoch": 4.285252711747058,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.909939959973316e-05,
+      "loss": 1.4783,
+      "step": 4642
+    },
+    {
+      "epoch": 4.286175859681514,
+      "grad_norm": 3.8359375,
+      "learning_rate": 6.909272848565711e-05,
+      "loss": 1.479,
+      "step": 4643
+    },
+    {
+      "epoch": 4.2870990076159705,
+      "grad_norm": 4.66015625,
+      "learning_rate": 6.908605737158106e-05,
+      "loss": 1.478,
+      "step": 4644
+    },
+    {
+      "epoch": 4.2880221555504265,
+      "grad_norm": 3.36328125,
+      "learning_rate": 6.9079386257505e-05,
+      "loss": 1.1501,
+      "step": 4645
+    },
+    {
+      "epoch": 4.2889453034848835,
+      "grad_norm": 3.8515625,
+      "learning_rate": 6.907271514342895e-05,
+      "loss": 1.5209,
+      "step": 4646
+    },
+    {
+      "epoch": 4.2898684514193395,
+      "grad_norm": 3.755859375,
+      "learning_rate": 6.906604402935291e-05,
+      "loss": 1.4971,
+      "step": 4647
+    },
+    {
+      "epoch": 4.290791599353796,
+      "grad_norm": 3.552734375,
+      "learning_rate": 6.905937291527686e-05,
+      "loss": 1.4111,
+      "step": 4648
+    },
+    {
+      "epoch": 4.291714747288253,
+      "grad_norm": 3.482421875,
+      "learning_rate": 6.90527018012008e-05,
+      "loss": 1.5271,
+      "step": 4649
+    },
+    {
+      "epoch": 4.292637895222709,
+      "grad_norm": 4.14453125,
+      "learning_rate": 6.904603068712475e-05,
+      "loss": 1.2473,
+      "step": 4650
+    },
+    {
+      "epoch": 4.293561043157166,
+      "grad_norm": 4.19921875,
+      "learning_rate": 6.90393595730487e-05,
+      "loss": 1.4567,
+      "step": 4651
+    },
+    {
+      "epoch": 4.294484191091622,
+      "grad_norm": 5.82421875,
+      "learning_rate": 6.903268845897265e-05,
+      "loss": 1.4631,
+      "step": 4652
+    },
+    {
+      "epoch": 4.295407339026079,
+      "grad_norm": 5.27734375,
+      "learning_rate": 6.90260173448966e-05,
+      "loss": 1.1178,
+      "step": 4653
+    },
+    {
+      "epoch": 4.296330486960535,
+      "grad_norm": 3.552734375,
+      "learning_rate": 6.901934623082054e-05,
+      "loss": 1.3091,
+      "step": 4654
+    },
+    {
+      "epoch": 4.297253634894992,
+      "grad_norm": 4.578125,
+      "learning_rate": 6.90126751167445e-05,
+      "loss": 1.6904,
+      "step": 4655
+    },
+    {
+      "epoch": 4.298176782829448,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.900600400266845e-05,
+      "loss": 1.3376,
+      "step": 4656
+    },
+    {
+      "epoch": 4.299099930763905,
+      "grad_norm": 3.912109375,
+      "learning_rate": 6.899933288859239e-05,
+      "loss": 1.4177,
+      "step": 4657
+    },
+    {
+      "epoch": 4.300023078698361,
+      "grad_norm": 3.791015625,
+      "learning_rate": 6.899266177451634e-05,
+      "loss": 1.3987,
+      "step": 4658
+    },
+    {
+      "epoch": 4.300946226632818,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.89859906604403e-05,
+      "loss": 1.3228,
+      "step": 4659
+    },
+    {
+      "epoch": 4.301869374567274,
+      "grad_norm": 3.595703125,
+      "learning_rate": 6.897931954636425e-05,
+      "loss": 1.1379,
+      "step": 4660
+    },
+    {
+      "epoch": 4.302792522501731,
+      "grad_norm": 3.72265625,
+      "learning_rate": 6.89726484322882e-05,
+      "loss": 1.3318,
+      "step": 4661
+    },
+    {
+      "epoch": 4.303715670436187,
+      "grad_norm": 3.283203125,
+      "learning_rate": 6.896597731821215e-05,
+      "loss": 1.3513,
+      "step": 4662
+    },
+    {
+      "epoch": 4.304638818370644,
+      "grad_norm": 3.5,
+      "learning_rate": 6.89593062041361e-05,
+      "loss": 1.4353,
+      "step": 4663
+    },
+    {
+      "epoch": 4.305561966305101,
+      "grad_norm": 4.5078125,
+      "learning_rate": 6.895263509006005e-05,
+      "loss": 1.6174,
+      "step": 4664
+    },
+    {
+      "epoch": 4.306485114239557,
+      "grad_norm": 4.23046875,
+      "learning_rate": 6.8945963975984e-05,
+      "loss": 1.5486,
+      "step": 4665
+    },
+    {
+      "epoch": 4.307408262174014,
+      "grad_norm": 4.82421875,
+      "learning_rate": 6.893929286190794e-05,
+      "loss": 1.3933,
+      "step": 4666
+    },
+    {
+      "epoch": 4.30833141010847,
+      "grad_norm": 4.17578125,
+      "learning_rate": 6.89326217478319e-05,
+      "loss": 1.1659,
+      "step": 4667
+    },
+    {
+      "epoch": 4.309254558042927,
+      "grad_norm": 3.98046875,
+      "learning_rate": 6.892595063375585e-05,
+      "loss": 1.3671,
+      "step": 4668
+    },
+    {
+      "epoch": 4.310177705977383,
+      "grad_norm": 4.38671875,
+      "learning_rate": 6.891927951967979e-05,
+      "loss": 1.4711,
+      "step": 4669
+    },
+    {
+      "epoch": 4.31110085391184,
+      "grad_norm": 3.990234375,
+      "learning_rate": 6.891260840560374e-05,
+      "loss": 1.3309,
+      "step": 4670
+    },
+    {
+      "epoch": 4.312024001846296,
+      "grad_norm": 3.615234375,
+      "learning_rate": 6.89059372915277e-05,
+      "loss": 1.42,
+      "step": 4671
+    },
+    {
+      "epoch": 4.3129471497807526,
+      "grad_norm": 3.326171875,
+      "learning_rate": 6.889926617745164e-05,
+      "loss": 1.3518,
+      "step": 4672
+    },
+    {
+      "epoch": 4.313870297715209,
+      "grad_norm": 3.69921875,
+      "learning_rate": 6.889259506337559e-05,
+      "loss": 1.6006,
+      "step": 4673
+    },
+    {
+      "epoch": 4.3147934456496655,
+      "grad_norm": 4.14453125,
+      "learning_rate": 6.888592394929953e-05,
+      "loss": 1.3689,
+      "step": 4674
+    },
+    {
+      "epoch": 4.3157165935841215,
+      "grad_norm": 4.140625,
+      "learning_rate": 6.887925283522349e-05,
+      "loss": 1.5784,
+      "step": 4675
+    },
+    {
+      "epoch": 4.316639741518578,
+      "grad_norm": 3.681640625,
+      "learning_rate": 6.887258172114744e-05,
+      "loss": 1.061,
+      "step": 4676
+    },
+    {
+      "epoch": 4.3175628894530345,
+      "grad_norm": 4.0390625,
+      "learning_rate": 6.886591060707138e-05,
+      "loss": 1.3793,
+      "step": 4677
+    },
+    {
+      "epoch": 4.318486037387491,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.885923949299533e-05,
+      "loss": 1.384,
+      "step": 4678
+    },
+    {
+      "epoch": 4.319409185321948,
+      "grad_norm": 3.43359375,
+      "learning_rate": 6.885256837891929e-05,
+      "loss": 1.3403,
+      "step": 4679
+    },
+    {
+      "epoch": 4.320332333256404,
+      "grad_norm": 3.7109375,
+      "learning_rate": 6.884589726484323e-05,
+      "loss": 1.3186,
+      "step": 4680
+    },
+    {
+      "epoch": 4.321255481190861,
+      "grad_norm": 3.53515625,
+      "learning_rate": 6.883922615076718e-05,
+      "loss": 1.3284,
+      "step": 4681
+    },
+    {
+      "epoch": 4.322178629125317,
+      "grad_norm": 4.23046875,
+      "learning_rate": 6.883255503669112e-05,
+      "loss": 1.6379,
+      "step": 4682
+    },
+    {
+      "epoch": 4.323101777059774,
+      "grad_norm": 3.603515625,
+      "learning_rate": 6.882588392261508e-05,
+      "loss": 1.4961,
+      "step": 4683
+    },
+    {
+      "epoch": 4.32402492499423,
+      "grad_norm": 3.927734375,
+      "learning_rate": 6.881921280853903e-05,
+      "loss": 1.2812,
+      "step": 4684
+    },
+    {
+      "epoch": 4.324948072928687,
+      "grad_norm": 4.3046875,
+      "learning_rate": 6.881254169446299e-05,
+      "loss": 1.686,
+      "step": 4685
+    },
+    {
+      "epoch": 4.325871220863143,
+      "grad_norm": 3.822265625,
+      "learning_rate": 6.880587058038693e-05,
+      "loss": 1.3676,
+      "step": 4686
+    },
+    {
+      "epoch": 4.3267943687976,
+      "grad_norm": 4.16796875,
+      "learning_rate": 6.879919946631088e-05,
+      "loss": 1.1243,
+      "step": 4687
+    },
+    {
+      "epoch": 4.327717516732056,
+      "grad_norm": 4.63671875,
+      "learning_rate": 6.879252835223484e-05,
+      "loss": 1.4258,
+      "step": 4688
+    },
+    {
+      "epoch": 4.328640664666513,
+      "grad_norm": 4.890625,
+      "learning_rate": 6.878585723815878e-05,
+      "loss": 1.396,
+      "step": 4689
+    },
+    {
+      "epoch": 4.329563812600969,
+      "grad_norm": 3.978515625,
+      "learning_rate": 6.877918612408273e-05,
+      "loss": 1.4109,
+      "step": 4690
+    },
+    {
+      "epoch": 4.330486960535426,
+      "grad_norm": 3.431640625,
+      "learning_rate": 6.877251501000667e-05,
+      "loss": 1.3567,
+      "step": 4691
+    },
+    {
+      "epoch": 4.331410108469882,
+      "grad_norm": 3.55859375,
+      "learning_rate": 6.876584389593063e-05,
+      "loss": 1.4844,
+      "step": 4692
+    },
+    {
+      "epoch": 4.332333256404339,
+      "grad_norm": 4.1015625,
+      "learning_rate": 6.875917278185458e-05,
+      "loss": 1.5144,
+      "step": 4693
+    },
+    {
+      "epoch": 4.333256404338795,
+      "grad_norm": 4.05078125,
+      "learning_rate": 6.875250166777852e-05,
+      "loss": 1.4766,
+      "step": 4694
+    },
+    {
+      "epoch": 4.334179552273252,
+      "grad_norm": 3.9921875,
+      "learning_rate": 6.874583055370247e-05,
+      "loss": 1.1693,
+      "step": 4695
+    },
+    {
+      "epoch": 4.335102700207708,
+      "grad_norm": 3.9609375,
+      "learning_rate": 6.873915943962643e-05,
+      "loss": 1.4727,
+      "step": 4696
+    },
+    {
+      "epoch": 4.336025848142165,
+      "grad_norm": 4.61328125,
+      "learning_rate": 6.873248832555037e-05,
+      "loss": 1.6099,
+      "step": 4697
+    },
+    {
+      "epoch": 4.336948996076622,
+      "grad_norm": 3.51953125,
+      "learning_rate": 6.872581721147432e-05,
+      "loss": 1.437,
+      "step": 4698
+    },
+    {
+      "epoch": 4.337872144011078,
+      "grad_norm": 3.978515625,
+      "learning_rate": 6.871914609739826e-05,
+      "loss": 1.1973,
+      "step": 4699
+    },
+    {
+      "epoch": 4.338795291945535,
+      "grad_norm": 4.3359375,
+      "learning_rate": 6.871247498332222e-05,
+      "loss": 1.498,
+      "step": 4700
+    },
+    {
+      "epoch": 4.339718439879991,
+      "grad_norm": 4.88671875,
+      "learning_rate": 6.870580386924617e-05,
+      "loss": 1.0847,
+      "step": 4701
+    },
+    {
+      "epoch": 4.3406415878144475,
+      "grad_norm": 3.3203125,
+      "learning_rate": 6.869913275517011e-05,
+      "loss": 0.9388,
+      "step": 4702
+    },
+    {
+      "epoch": 4.341564735748904,
+      "grad_norm": 4.23828125,
+      "learning_rate": 6.869246164109406e-05,
+      "loss": 1.373,
+      "step": 4703
+    },
+    {
+      "epoch": 4.3424878836833605,
+      "grad_norm": 4.43359375,
+      "learning_rate": 6.868579052701802e-05,
+      "loss": 1.4736,
+      "step": 4704
+    },
+    {
+      "epoch": 4.3434110316178165,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.867911941294196e-05,
+      "loss": 1.3176,
+      "step": 4705
+    },
+    {
+      "epoch": 4.344334179552273,
+      "grad_norm": 4.4453125,
+      "learning_rate": 6.867244829886591e-05,
+      "loss": 1.3149,
+      "step": 4706
+    },
+    {
+      "epoch": 4.345257327486729,
+      "grad_norm": 4.05078125,
+      "learning_rate": 6.866577718478985e-05,
+      "loss": 1.3491,
+      "step": 4707
+    },
+    {
+      "epoch": 4.346180475421186,
+      "grad_norm": 4.9921875,
+      "learning_rate": 6.865910607071381e-05,
+      "loss": 1.489,
+      "step": 4708
+    },
+    {
+      "epoch": 4.347103623355642,
+      "grad_norm": 4.69921875,
+      "learning_rate": 6.865243495663776e-05,
+      "loss": 1.302,
+      "step": 4709
+    },
+    {
+      "epoch": 4.348026771290099,
+      "grad_norm": 5.65234375,
+      "learning_rate": 6.86457638425617e-05,
+      "loss": 1.5107,
+      "step": 4710
+    },
+    {
+      "epoch": 4.348949919224555,
+      "grad_norm": 3.6328125,
+      "learning_rate": 6.863909272848566e-05,
+      "loss": 1.6169,
+      "step": 4711
+    },
+    {
+      "epoch": 4.349873067159012,
+      "grad_norm": 5.078125,
+      "learning_rate": 6.863242161440962e-05,
+      "loss": 1.0339,
+      "step": 4712
+    },
+    {
+      "epoch": 4.350796215093469,
+      "grad_norm": 4.296875,
+      "learning_rate": 6.862575050033357e-05,
+      "loss": 1.4495,
+      "step": 4713
+    },
+    {
+      "epoch": 4.351719363027925,
+      "grad_norm": 3.908203125,
+      "learning_rate": 6.861907938625751e-05,
+      "loss": 1.2816,
+      "step": 4714
+    },
+    {
+      "epoch": 4.352642510962382,
+      "grad_norm": 3.580078125,
+      "learning_rate": 6.861240827218146e-05,
+      "loss": 1.6565,
+      "step": 4715
+    },
+    {
+      "epoch": 4.353565658896838,
+      "grad_norm": 4.08203125,
+      "learning_rate": 6.860573715810542e-05,
+      "loss": 1.2034,
+      "step": 4716
+    },
+    {
+      "epoch": 4.354488806831295,
+      "grad_norm": 3.875,
+      "learning_rate": 6.859906604402936e-05,
+      "loss": 1.4443,
+      "step": 4717
+    },
+    {
+      "epoch": 4.355411954765751,
+      "grad_norm": 3.80078125,
+      "learning_rate": 6.85923949299533e-05,
+      "loss": 1.114,
+      "step": 4718
+    },
+    {
+      "epoch": 4.356335102700208,
+      "grad_norm": 3.626953125,
+      "learning_rate": 6.858572381587725e-05,
+      "loss": 1.5549,
+      "step": 4719
+    },
+    {
+      "epoch": 4.357258250634664,
+      "grad_norm": 3.6328125,
+      "learning_rate": 6.857905270180121e-05,
+      "loss": 1.6316,
+      "step": 4720
+    },
+    {
+      "epoch": 4.358181398569121,
+      "grad_norm": 3.541015625,
+      "learning_rate": 6.857238158772516e-05,
+      "loss": 1.2681,
+      "step": 4721
+    },
+    {
+      "epoch": 4.359104546503577,
+      "grad_norm": 5.2265625,
+      "learning_rate": 6.85657104736491e-05,
+      "loss": 1.3542,
+      "step": 4722
+    },
+    {
+      "epoch": 4.360027694438034,
+      "grad_norm": 3.658203125,
+      "learning_rate": 6.855903935957305e-05,
+      "loss": 1.4753,
+      "step": 4723
+    },
+    {
+      "epoch": 4.36095084237249,
+      "grad_norm": 3.43359375,
+      "learning_rate": 6.8552368245497e-05,
+      "loss": 1.0732,
+      "step": 4724
+    },
+    {
+      "epoch": 4.361873990306947,
+      "grad_norm": 3.388671875,
+      "learning_rate": 6.854569713142095e-05,
+      "loss": 1.3667,
+      "step": 4725
+    },
+    {
+      "epoch": 4.362797138241403,
+      "grad_norm": 3.6171875,
+      "learning_rate": 6.85390260173449e-05,
+      "loss": 1.1188,
+      "step": 4726
+    },
+    {
+      "epoch": 4.36372028617586,
+      "grad_norm": 3.224609375,
+      "learning_rate": 6.853235490326884e-05,
+      "loss": 1.3992,
+      "step": 4727
+    },
+    {
+      "epoch": 4.364643434110317,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.85256837891928e-05,
+      "loss": 1.4265,
+      "step": 4728
+    },
+    {
+      "epoch": 4.365566582044773,
+      "grad_norm": 3.962890625,
+      "learning_rate": 6.851901267511675e-05,
+      "loss": 1.3691,
+      "step": 4729
+    },
+    {
+      "epoch": 4.36648972997923,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.851234156104069e-05,
+      "loss": 1.0373,
+      "step": 4730
+    },
+    {
+      "epoch": 4.367412877913686,
+      "grad_norm": 3.87109375,
+      "learning_rate": 6.850567044696464e-05,
+      "loss": 1.0984,
+      "step": 4731
+    },
+    {
+      "epoch": 4.3683360258481425,
+      "grad_norm": 4.70703125,
+      "learning_rate": 6.84989993328886e-05,
+      "loss": 1.4001,
+      "step": 4732
+    },
+    {
+      "epoch": 4.3692591737825985,
+      "grad_norm": 5.55859375,
+      "learning_rate": 6.849232821881254e-05,
+      "loss": 1.3809,
+      "step": 4733
+    },
+    {
+      "epoch": 4.3701823217170555,
+      "grad_norm": 3.775390625,
+      "learning_rate": 6.848565710473649e-05,
+      "loss": 1.6301,
+      "step": 4734
+    },
+    {
+      "epoch": 4.3711054696515115,
+      "grad_norm": 3.974609375,
+      "learning_rate": 6.847898599066043e-05,
+      "loss": 1.3452,
+      "step": 4735
+    },
+    {
+      "epoch": 4.372028617585968,
+      "grad_norm": 4.51953125,
+      "learning_rate": 6.847231487658439e-05,
+      "loss": 1.5242,
+      "step": 4736
+    },
+    {
+      "epoch": 4.372951765520424,
+      "grad_norm": 5.85546875,
+      "learning_rate": 6.846564376250835e-05,
+      "loss": 1.5903,
+      "step": 4737
+    },
+    {
+      "epoch": 4.373874913454881,
+      "grad_norm": 3.39453125,
+      "learning_rate": 6.84589726484323e-05,
+      "loss": 1.3397,
+      "step": 4738
+    },
+    {
+      "epoch": 4.374798061389337,
+      "grad_norm": 3.61328125,
+      "learning_rate": 6.845230153435624e-05,
+      "loss": 1.3682,
+      "step": 4739
+    },
+    {
+      "epoch": 4.375721209323794,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.844563042028019e-05,
+      "loss": 1.3191,
+      "step": 4740
+    },
+    {
+      "epoch": 4.37664435725825,
+      "grad_norm": 3.916015625,
+      "learning_rate": 6.843895930620415e-05,
+      "loss": 1.2308,
+      "step": 4741
+    },
+    {
+      "epoch": 4.377567505192707,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.843228819212809e-05,
+      "loss": 1.4139,
+      "step": 4742
+    },
+    {
+      "epoch": 4.378490653127163,
+      "grad_norm": 3.875,
+      "learning_rate": 6.842561707805204e-05,
+      "loss": 1.6252,
+      "step": 4743
+    },
+    {
+      "epoch": 4.37941380106162,
+      "grad_norm": 3.2109375,
+      "learning_rate": 6.841894596397598e-05,
+      "loss": 1.4744,
+      "step": 4744
+    },
+    {
+      "epoch": 4.380336948996076,
+      "grad_norm": 4.265625,
+      "learning_rate": 6.841227484989994e-05,
+      "loss": 1.5667,
+      "step": 4745
+    },
+    {
+      "epoch": 4.381260096930533,
+      "grad_norm": 3.677734375,
+      "learning_rate": 6.840560373582389e-05,
+      "loss": 1.5278,
+      "step": 4746
+    },
+    {
+      "epoch": 4.38218324486499,
+      "grad_norm": 4.45703125,
+      "learning_rate": 6.839893262174783e-05,
+      "loss": 1.46,
+      "step": 4747
+    },
+    {
+      "epoch": 4.383106392799446,
+      "grad_norm": 3.888671875,
+      "learning_rate": 6.839226150767178e-05,
+      "loss": 1.5042,
+      "step": 4748
+    },
+    {
+      "epoch": 4.384029540733903,
+      "grad_norm": 4.1796875,
+      "learning_rate": 6.838559039359574e-05,
+      "loss": 1.3914,
+      "step": 4749
+    },
+    {
+      "epoch": 4.384952688668359,
+      "grad_norm": 4.328125,
+      "learning_rate": 6.837891927951968e-05,
+      "loss": 1.2361,
+      "step": 4750
+    },
+    {
+      "epoch": 4.385875836602816,
+      "grad_norm": 4.09765625,
+      "learning_rate": 6.837224816544363e-05,
+      "loss": 1.3962,
+      "step": 4751
+    },
+    {
+      "epoch": 4.386798984537272,
+      "grad_norm": 3.783203125,
+      "learning_rate": 6.836557705136757e-05,
+      "loss": 1.2543,
+      "step": 4752
+    },
+    {
+      "epoch": 4.387722132471729,
+      "grad_norm": 3.94140625,
+      "learning_rate": 6.835890593729153e-05,
+      "loss": 1.6306,
+      "step": 4753
+    },
+    {
+      "epoch": 4.388645280406185,
+      "grad_norm": 3.466796875,
+      "learning_rate": 6.835223482321548e-05,
+      "loss": 1.3257,
+      "step": 4754
+    },
+    {
+      "epoch": 4.389568428340642,
+      "grad_norm": 4.6484375,
+      "learning_rate": 6.834556370913942e-05,
+      "loss": 1.5972,
+      "step": 4755
+    },
+    {
+      "epoch": 4.390491576275098,
+      "grad_norm": 4.03515625,
+      "learning_rate": 6.833889259506338e-05,
+      "loss": 1.6599,
+      "step": 4756
+    },
+    {
+      "epoch": 4.391414724209555,
+      "grad_norm": 3.80078125,
+      "learning_rate": 6.833222148098733e-05,
+      "loss": 1.4902,
+      "step": 4757
+    },
+    {
+      "epoch": 4.392337872144011,
+      "grad_norm": 4.6796875,
+      "learning_rate": 6.832555036691127e-05,
+      "loss": 1.5562,
+      "step": 4758
+    },
+    {
+      "epoch": 4.393261020078468,
+      "grad_norm": 4.81640625,
+      "learning_rate": 6.831887925283522e-05,
+      "loss": 1.1027,
+      "step": 4759
+    },
+    {
+      "epoch": 4.394184168012924,
+      "grad_norm": 5.12890625,
+      "learning_rate": 6.831220813875918e-05,
+      "loss": 1.3245,
+      "step": 4760
+    },
+    {
+      "epoch": 4.395107315947381,
+      "grad_norm": 4.3359375,
+      "learning_rate": 6.830553702468314e-05,
+      "loss": 1.5212,
+      "step": 4761
+    },
+    {
+      "epoch": 4.3960304638818375,
+      "grad_norm": 3.44140625,
+      "learning_rate": 6.829886591060708e-05,
+      "loss": 1.3074,
+      "step": 4762
+    },
+    {
+      "epoch": 4.3969536118162935,
+      "grad_norm": 3.435546875,
+      "learning_rate": 6.829219479653103e-05,
+      "loss": 1.4485,
+      "step": 4763
+    },
+    {
+      "epoch": 4.39787675975075,
+      "grad_norm": 4.37890625,
+      "learning_rate": 6.828552368245497e-05,
+      "loss": 1.2482,
+      "step": 4764
+    },
+    {
+      "epoch": 4.3987999076852065,
+      "grad_norm": 4.40625,
+      "learning_rate": 6.827885256837893e-05,
+      "loss": 1.4136,
+      "step": 4765
+    },
+    {
+      "epoch": 4.399723055619663,
+      "grad_norm": 4.046875,
+      "learning_rate": 6.827218145430288e-05,
+      "loss": 1.5193,
+      "step": 4766
+    },
+    {
+      "epoch": 4.400646203554119,
+      "grad_norm": 3.638671875,
+      "learning_rate": 6.826551034022682e-05,
+      "loss": 1.491,
+      "step": 4767
+    },
+    {
+      "epoch": 4.401569351488576,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.825883922615077e-05,
+      "loss": 1.4685,
+      "step": 4768
+    },
+    {
+      "epoch": 4.402492499423032,
+      "grad_norm": 4.12109375,
+      "learning_rate": 6.825216811207473e-05,
+      "loss": 1.4164,
+      "step": 4769
+    },
+    {
+      "epoch": 4.403415647357489,
+      "grad_norm": 3.81640625,
+      "learning_rate": 6.824549699799867e-05,
+      "loss": 1.4785,
+      "step": 4770
+    },
+    {
+      "epoch": 4.404338795291945,
+      "grad_norm": 4.34765625,
+      "learning_rate": 6.823882588392262e-05,
+      "loss": 1.5007,
+      "step": 4771
+    },
+    {
+      "epoch": 4.405261943226402,
+      "grad_norm": 3.83984375,
+      "learning_rate": 6.823215476984656e-05,
+      "loss": 1.4583,
+      "step": 4772
+    },
+    {
+      "epoch": 4.406185091160858,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.822548365577052e-05,
+      "loss": 1.5703,
+      "step": 4773
+    },
+    {
+      "epoch": 4.407108239095315,
+      "grad_norm": 3.990234375,
+      "learning_rate": 6.821881254169447e-05,
+      "loss": 1.509,
+      "step": 4774
+    },
+    {
+      "epoch": 4.408031387029771,
+      "grad_norm": 4.015625,
+      "learning_rate": 6.821214142761841e-05,
+      "loss": 1.499,
+      "step": 4775
+    },
+    {
+      "epoch": 4.408954534964228,
+      "grad_norm": 4.02734375,
+      "learning_rate": 6.820547031354236e-05,
+      "loss": 1.3765,
+      "step": 4776
+    },
+    {
+      "epoch": 4.409877682898685,
+      "grad_norm": 3.583984375,
+      "learning_rate": 6.819879919946632e-05,
+      "loss": 1.2444,
+      "step": 4777
+    },
+    {
+      "epoch": 4.410800830833141,
+      "grad_norm": 4.0390625,
+      "learning_rate": 6.819212808539026e-05,
+      "loss": 1.3606,
+      "step": 4778
+    },
+    {
+      "epoch": 4.411723978767598,
+      "grad_norm": 3.408203125,
+      "learning_rate": 6.818545697131421e-05,
+      "loss": 1.5647,
+      "step": 4779
+    },
+    {
+      "epoch": 4.412647126702054,
+      "grad_norm": 3.880859375,
+      "learning_rate": 6.817878585723815e-05,
+      "loss": 1.5149,
+      "step": 4780
+    },
+    {
+      "epoch": 4.413570274636511,
+      "grad_norm": 3.248046875,
+      "learning_rate": 6.817211474316211e-05,
+      "loss": 1.708,
+      "step": 4781
+    },
+    {
+      "epoch": 4.414493422570967,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.816544362908606e-05,
+      "loss": 1.5173,
+      "step": 4782
+    },
+    {
+      "epoch": 4.415416570505424,
+      "grad_norm": 4.4140625,
+      "learning_rate": 6.815877251501e-05,
+      "loss": 1.3516,
+      "step": 4783
+    },
+    {
+      "epoch": 4.41633971843988,
+      "grad_norm": 4.92578125,
+      "learning_rate": 6.815210140093395e-05,
+      "loss": 1.3528,
+      "step": 4784
+    },
+    {
+      "epoch": 4.417262866374337,
+      "grad_norm": 3.373046875,
+      "learning_rate": 6.814543028685791e-05,
+      "loss": 1.3342,
+      "step": 4785
+    },
+    {
+      "epoch": 4.418186014308793,
+      "grad_norm": 3.6484375,
+      "learning_rate": 6.813875917278187e-05,
+      "loss": 1.3953,
+      "step": 4786
+    },
+    {
+      "epoch": 4.41910916224325,
+      "grad_norm": 4.1015625,
+      "learning_rate": 6.813208805870581e-05,
+      "loss": 1.1208,
+      "step": 4787
+    },
+    {
+      "epoch": 4.420032310177706,
+      "grad_norm": 3.759765625,
+      "learning_rate": 6.812541694462976e-05,
+      "loss": 1.2881,
+      "step": 4788
+    },
+    {
+      "epoch": 4.420955458112163,
+      "grad_norm": 3.55078125,
+      "learning_rate": 6.81187458305537e-05,
+      "loss": 1.2793,
+      "step": 4789
+    },
+    {
+      "epoch": 4.421878606046619,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.811207471647766e-05,
+      "loss": 1.2754,
+      "step": 4790
+    },
+    {
+      "epoch": 4.422801753981076,
+      "grad_norm": 3.689453125,
+      "learning_rate": 6.810540360240161e-05,
+      "loss": 1.6052,
+      "step": 4791
+    },
+    {
+      "epoch": 4.423724901915532,
+      "grad_norm": 4.609375,
+      "learning_rate": 6.809873248832555e-05,
+      "loss": 1.3843,
+      "step": 4792
+    },
+    {
+      "epoch": 4.4246480498499885,
+      "grad_norm": 4.3203125,
+      "learning_rate": 6.80920613742495e-05,
+      "loss": 1.5105,
+      "step": 4793
+    },
+    {
+      "epoch": 4.4255711977844445,
+      "grad_norm": 3.49609375,
+      "learning_rate": 6.808539026017346e-05,
+      "loss": 1.6147,
+      "step": 4794
+    },
+    {
+      "epoch": 4.426494345718901,
+      "grad_norm": 4.4765625,
+      "learning_rate": 6.80787191460974e-05,
+      "loss": 1.613,
+      "step": 4795
+    },
+    {
+      "epoch": 4.427417493653358,
+      "grad_norm": 3.79296875,
+      "learning_rate": 6.807204803202135e-05,
+      "loss": 1.718,
+      "step": 4796
+    },
+    {
+      "epoch": 4.428340641587814,
+      "grad_norm": 4.4921875,
+      "learning_rate": 6.806537691794531e-05,
+      "loss": 1.5632,
+      "step": 4797
+    },
+    {
+      "epoch": 4.429263789522271,
+      "grad_norm": 3.576171875,
+      "learning_rate": 6.805870580386925e-05,
+      "loss": 1.1628,
+      "step": 4798
+    },
+    {
+      "epoch": 4.430186937456727,
+      "grad_norm": 4.51171875,
+      "learning_rate": 6.80520346897932e-05,
+      "loss": 1.3047,
+      "step": 4799
+    },
+    {
+      "epoch": 4.431110085391184,
+      "grad_norm": 4.0859375,
+      "learning_rate": 6.804536357571714e-05,
+      "loss": 1.3608,
+      "step": 4800
+    },
+    {
+      "epoch": 4.43203323332564,
+      "grad_norm": 4.3984375,
+      "learning_rate": 6.80386924616411e-05,
+      "loss": 1.4072,
+      "step": 4801
+    },
+    {
+      "epoch": 4.432956381260097,
+      "grad_norm": 3.755859375,
+      "learning_rate": 6.803202134756505e-05,
+      "loss": 1.5798,
+      "step": 4802
+    },
+    {
+      "epoch": 4.433879529194553,
+      "grad_norm": 3.556640625,
+      "learning_rate": 6.8025350233489e-05,
+      "loss": 1.1198,
+      "step": 4803
+    },
+    {
+      "epoch": 4.43480267712901,
+      "grad_norm": 4.30859375,
+      "learning_rate": 6.801867911941294e-05,
+      "loss": 1.1475,
+      "step": 4804
+    },
+    {
+      "epoch": 4.435725825063466,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.80120080053369e-05,
+      "loss": 1.4868,
+      "step": 4805
+    },
+    {
+      "epoch": 4.436648972997923,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.800533689126084e-05,
+      "loss": 1.4287,
+      "step": 4806
+    },
+    {
+      "epoch": 4.437572120932379,
+      "grad_norm": 3.994140625,
+      "learning_rate": 6.799866577718479e-05,
+      "loss": 1.3923,
+      "step": 4807
+    },
+    {
+      "epoch": 4.438495268866836,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.799199466310873e-05,
+      "loss": 1.3124,
+      "step": 4808
+    },
+    {
+      "epoch": 4.439418416801292,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.798532354903269e-05,
+      "loss": 1.0332,
+      "step": 4809
+    },
+    {
+      "epoch": 4.440341564735749,
+      "grad_norm": 3.64453125,
+      "learning_rate": 6.797865243495664e-05,
+      "loss": 1.2605,
+      "step": 4810
+    },
+    {
+      "epoch": 4.441264712670206,
+      "grad_norm": 4.1484375,
+      "learning_rate": 6.79719813208806e-05,
+      "loss": 1.3938,
+      "step": 4811
+    },
+    {
+      "epoch": 4.442187860604662,
+      "grad_norm": 4.3203125,
+      "learning_rate": 6.796531020680454e-05,
+      "loss": 1.4082,
+      "step": 4812
+    },
+    {
+      "epoch": 4.443111008539119,
+      "grad_norm": 3.611328125,
+      "learning_rate": 6.795863909272849e-05,
+      "loss": 1.3098,
+      "step": 4813
+    },
+    {
+      "epoch": 4.444034156473575,
+      "grad_norm": 4.19140625,
+      "learning_rate": 6.795196797865245e-05,
+      "loss": 1.2867,
+      "step": 4814
+    },
+    {
+      "epoch": 4.444957304408032,
+      "grad_norm": 4.625,
+      "learning_rate": 6.794529686457639e-05,
+      "loss": 1.4955,
+      "step": 4815
+    },
+    {
+      "epoch": 4.445880452342488,
+      "grad_norm": 4.2578125,
+      "learning_rate": 6.793862575050034e-05,
+      "loss": 1.2996,
+      "step": 4816
+    },
+    {
+      "epoch": 4.446803600276945,
+      "grad_norm": 4.3984375,
+      "learning_rate": 6.793195463642428e-05,
+      "loss": 1.6614,
+      "step": 4817
+    },
+    {
+      "epoch": 4.447726748211401,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.792528352234824e-05,
+      "loss": 1.5015,
+      "step": 4818
+    },
+    {
+      "epoch": 4.448649896145858,
+      "grad_norm": 4.84765625,
+      "learning_rate": 6.791861240827219e-05,
+      "loss": 1.6992,
+      "step": 4819
+    },
+    {
+      "epoch": 4.449573044080314,
+      "grad_norm": 4.2578125,
+      "learning_rate": 6.791194129419613e-05,
+      "loss": 1.3203,
+      "step": 4820
+    },
+    {
+      "epoch": 4.4504961920147705,
+      "grad_norm": 3.916015625,
+      "learning_rate": 6.790527018012008e-05,
+      "loss": 1.52,
+      "step": 4821
+    },
+    {
+      "epoch": 4.451419339949227,
+      "grad_norm": 3.90234375,
+      "learning_rate": 6.789859906604404e-05,
+      "loss": 1.3838,
+      "step": 4822
+    },
+    {
+      "epoch": 4.4523424878836835,
+      "grad_norm": 3.953125,
+      "learning_rate": 6.789192795196798e-05,
+      "loss": 1.231,
+      "step": 4823
+    },
+    {
+      "epoch": 4.4532656358181395,
+      "grad_norm": 3.201171875,
+      "learning_rate": 6.788525683789193e-05,
+      "loss": 1.2937,
+      "step": 4824
+    },
+    {
+      "epoch": 4.454188783752596,
+      "grad_norm": 3.017578125,
+      "learning_rate": 6.787858572381587e-05,
+      "loss": 1.0839,
+      "step": 4825
+    },
+    {
+      "epoch": 4.455111931687052,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.787191460973983e-05,
+      "loss": 1.4246,
+      "step": 4826
+    },
+    {
+      "epoch": 4.456035079621509,
+      "grad_norm": 3.912109375,
+      "learning_rate": 6.786524349566378e-05,
+      "loss": 1.3704,
+      "step": 4827
+    },
+    {
+      "epoch": 4.456958227555965,
+      "grad_norm": 3.9296875,
+      "learning_rate": 6.785857238158772e-05,
+      "loss": 1.3484,
+      "step": 4828
+    },
+    {
+      "epoch": 4.457881375490422,
+      "grad_norm": 4.33984375,
+      "learning_rate": 6.785190126751167e-05,
+      "loss": 1.4351,
+      "step": 4829
+    },
+    {
+      "epoch": 4.458804523424879,
+      "grad_norm": 3.931640625,
+      "learning_rate": 6.784523015343563e-05,
+      "loss": 1.5745,
+      "step": 4830
+    },
+    {
+      "epoch": 4.459727671359335,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.783855903935957e-05,
+      "loss": 1.5969,
+      "step": 4831
+    },
+    {
+      "epoch": 4.460650819293792,
+      "grad_norm": 4.41015625,
+      "learning_rate": 6.783188792528352e-05,
+      "loss": 1.7014,
+      "step": 4832
+    },
+    {
+      "epoch": 4.461573967228248,
+      "grad_norm": 4.06640625,
+      "learning_rate": 6.782521681120746e-05,
+      "loss": 1.5259,
+      "step": 4833
+    },
+    {
+      "epoch": 4.462497115162705,
+      "grad_norm": 3.904296875,
+      "learning_rate": 6.781854569713142e-05,
+      "loss": 1.3811,
+      "step": 4834
+    },
+    {
+      "epoch": 4.463420263097161,
+      "grad_norm": 3.580078125,
+      "learning_rate": 6.781187458305537e-05,
+      "loss": 1.4199,
+      "step": 4835
+    },
+    {
+      "epoch": 4.464343411031618,
+      "grad_norm": 3.7421875,
+      "learning_rate": 6.780520346897933e-05,
+      "loss": 1.3677,
+      "step": 4836
+    },
+    {
+      "epoch": 4.465266558966074,
+      "grad_norm": 3.716796875,
+      "learning_rate": 6.779853235490327e-05,
+      "loss": 1.5276,
+      "step": 4837
+    },
+    {
+      "epoch": 4.466189706900531,
+      "grad_norm": 3.822265625,
+      "learning_rate": 6.779186124082723e-05,
+      "loss": 1.3066,
+      "step": 4838
+    },
+    {
+      "epoch": 4.467112854834987,
+      "grad_norm": 3.896484375,
+      "learning_rate": 6.778519012675118e-05,
+      "loss": 1.4033,
+      "step": 4839
+    },
+    {
+      "epoch": 4.468036002769444,
+      "grad_norm": 5.1328125,
+      "learning_rate": 6.777851901267512e-05,
+      "loss": 1.4132,
+      "step": 4840
+    },
+    {
+      "epoch": 4.4689591507039,
+      "grad_norm": 3.73046875,
+      "learning_rate": 6.777184789859907e-05,
+      "loss": 1.5164,
+      "step": 4841
+    },
+    {
+      "epoch": 4.469882298638357,
+      "grad_norm": 3.60546875,
+      "learning_rate": 6.776517678452303e-05,
+      "loss": 1.4519,
+      "step": 4842
+    },
+    {
+      "epoch": 4.470805446572813,
+      "grad_norm": 3.892578125,
+      "learning_rate": 6.775850567044697e-05,
+      "loss": 1.3669,
+      "step": 4843
+    },
+    {
+      "epoch": 4.47172859450727,
+      "grad_norm": 3.392578125,
+      "learning_rate": 6.775183455637092e-05,
+      "loss": 1.569,
+      "step": 4844
+    },
+    {
+      "epoch": 4.472651742441727,
+      "grad_norm": 3.58984375,
+      "learning_rate": 6.774516344229486e-05,
+      "loss": 1.3137,
+      "step": 4845
+    },
+    {
+      "epoch": 4.473574890376183,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.773849232821882e-05,
+      "loss": 1.0944,
+      "step": 4846
+    },
+    {
+      "epoch": 4.47449803831064,
+      "grad_norm": 3.619140625,
+      "learning_rate": 6.773182121414277e-05,
+      "loss": 1.3591,
+      "step": 4847
+    },
+    {
+      "epoch": 4.475421186245096,
+      "grad_norm": 3.673828125,
+      "learning_rate": 6.772515010006671e-05,
+      "loss": 1.2576,
+      "step": 4848
+    },
+    {
+      "epoch": 4.476344334179553,
+      "grad_norm": 4.05078125,
+      "learning_rate": 6.771847898599066e-05,
+      "loss": 1.5146,
+      "step": 4849
+    },
+    {
+      "epoch": 4.477267482114009,
+      "grad_norm": 4.62890625,
+      "learning_rate": 6.771180787191462e-05,
+      "loss": 1.4458,
+      "step": 4850
+    },
+    {
+      "epoch": 4.4781906300484655,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.770513675783856e-05,
+      "loss": 1.2496,
+      "step": 4851
+    },
+    {
+      "epoch": 4.4791137779829215,
+      "grad_norm": 3.884765625,
+      "learning_rate": 6.769846564376251e-05,
+      "loss": 1.5015,
+      "step": 4852
+    },
+    {
+      "epoch": 4.4800369259173785,
+      "grad_norm": 4.375,
+      "learning_rate": 6.769179452968645e-05,
+      "loss": 1.2047,
+      "step": 4853
+    },
+    {
+      "epoch": 4.4809600738518345,
+      "grad_norm": 3.83203125,
+      "learning_rate": 6.768512341561041e-05,
+      "loss": 1.7515,
+      "step": 4854
+    },
+    {
+      "epoch": 4.481883221786291,
+      "grad_norm": 2.904296875,
+      "learning_rate": 6.767845230153436e-05,
+      "loss": 1.2954,
+      "step": 4855
+    },
+    {
+      "epoch": 4.482806369720747,
+      "grad_norm": 3.794921875,
+      "learning_rate": 6.76717811874583e-05,
+      "loss": 0.9907,
+      "step": 4856
+    },
+    {
+      "epoch": 4.483729517655204,
+      "grad_norm": 4.5703125,
+      "learning_rate": 6.766511007338225e-05,
+      "loss": 1.5735,
+      "step": 4857
+    },
+    {
+      "epoch": 4.48465266558966,
+      "grad_norm": 3.751953125,
+      "learning_rate": 6.765843895930621e-05,
+      "loss": 1.4563,
+      "step": 4858
+    },
+    {
+      "epoch": 4.485575813524117,
+      "grad_norm": 4.46484375,
+      "learning_rate": 6.765176784523015e-05,
+      "loss": 1.2695,
+      "step": 4859
+    },
+    {
+      "epoch": 4.486498961458574,
+      "grad_norm": 3.513671875,
+      "learning_rate": 6.76450967311541e-05,
+      "loss": 1.3344,
+      "step": 4860
+    },
+    {
+      "epoch": 4.48742210939303,
+      "grad_norm": 4.92578125,
+      "learning_rate": 6.763842561707806e-05,
+      "loss": 1.4973,
+      "step": 4861
+    },
+    {
+      "epoch": 4.488345257327487,
+      "grad_norm": 3.66015625,
+      "learning_rate": 6.7631754503002e-05,
+      "loss": 1.2996,
+      "step": 4862
+    },
+    {
+      "epoch": 4.489268405261943,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.762508338892596e-05,
+      "loss": 1.0724,
+      "step": 4863
+    },
+    {
+      "epoch": 4.4901915531964,
+      "grad_norm": 3.57421875,
+      "learning_rate": 6.761841227484991e-05,
+      "loss": 1.4581,
+      "step": 4864
+    },
+    {
+      "epoch": 4.491114701130856,
+      "grad_norm": 3.8359375,
+      "learning_rate": 6.761174116077385e-05,
+      "loss": 1.3196,
+      "step": 4865
+    },
+    {
+      "epoch": 4.492037849065313,
+      "grad_norm": 3.7109375,
+      "learning_rate": 6.76050700466978e-05,
+      "loss": 1.2095,
+      "step": 4866
+    },
+    {
+      "epoch": 4.492960996999769,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.759839893262176e-05,
+      "loss": 1.5391,
+      "step": 4867
+    },
+    {
+      "epoch": 4.493884144934226,
+      "grad_norm": 3.892578125,
+      "learning_rate": 6.75917278185457e-05,
+      "loss": 1.5947,
+      "step": 4868
+    },
+    {
+      "epoch": 4.494807292868682,
+      "grad_norm": 3.74609375,
+      "learning_rate": 6.758505670446965e-05,
+      "loss": 1.4675,
+      "step": 4869
+    },
+    {
+      "epoch": 4.495730440803139,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.75783855903936e-05,
+      "loss": 1.5071,
+      "step": 4870
+    },
+    {
+      "epoch": 4.496653588737595,
+      "grad_norm": 3.837890625,
+      "learning_rate": 6.757171447631755e-05,
+      "loss": 1.3689,
+      "step": 4871
+    },
+    {
+      "epoch": 4.497576736672052,
+      "grad_norm": 3.619140625,
+      "learning_rate": 6.75650433622415e-05,
+      "loss": 1.2808,
+      "step": 4872
+    },
+    {
+      "epoch": 4.498499884606508,
+      "grad_norm": 3.5078125,
+      "learning_rate": 6.755837224816544e-05,
+      "loss": 1.3767,
+      "step": 4873
+    },
+    {
+      "epoch": 4.499423032540965,
+      "grad_norm": 3.626953125,
+      "learning_rate": 6.755170113408939e-05,
+      "loss": 1.2688,
+      "step": 4874
+    },
+    {
+      "epoch": 4.500346180475422,
+      "grad_norm": 3.85546875,
+      "learning_rate": 6.754503002001335e-05,
+      "loss": 1.1259,
+      "step": 4875
+    },
+    {
+      "epoch": 4.501269328409878,
+      "grad_norm": 4.4765625,
+      "learning_rate": 6.75383589059373e-05,
+      "loss": 1.4552,
+      "step": 4876
+    },
+    {
+      "epoch": 4.502192476344334,
+      "grad_norm": 4.58203125,
+      "learning_rate": 6.753168779186124e-05,
+      "loss": 1.6125,
+      "step": 4877
+    },
+    {
+      "epoch": 4.503115624278791,
+      "grad_norm": 4.94921875,
+      "learning_rate": 6.752501667778519e-05,
+      "loss": 1.7544,
+      "step": 4878
+    },
+    {
+      "epoch": 4.5040387722132476,
+      "grad_norm": 3.59765625,
+      "learning_rate": 6.751834556370914e-05,
+      "loss": 1.5205,
+      "step": 4879
+    },
+    {
+      "epoch": 4.504961920147704,
+      "grad_norm": 4.5234375,
+      "learning_rate": 6.751167444963309e-05,
+      "loss": 1.4492,
+      "step": 4880
+    },
+    {
+      "epoch": 4.5058850680821605,
+      "grad_norm": 3.966796875,
+      "learning_rate": 6.750500333555704e-05,
+      "loss": 1.4749,
+      "step": 4881
+    },
+    {
+      "epoch": 4.5068082160166165,
+      "grad_norm": 4.5,
+      "learning_rate": 6.749833222148098e-05,
+      "loss": 1.4841,
+      "step": 4882
+    },
+    {
+      "epoch": 4.507731363951073,
+      "grad_norm": 4.2890625,
+      "learning_rate": 6.749166110740494e-05,
+      "loss": 1.2954,
+      "step": 4883
+    },
+    {
+      "epoch": 4.5086545118855295,
+      "grad_norm": 4.06640625,
+      "learning_rate": 6.748498999332888e-05,
+      "loss": 1.4609,
+      "step": 4884
+    },
+    {
+      "epoch": 4.509577659819986,
+      "grad_norm": 3.498046875,
+      "learning_rate": 6.747831887925283e-05,
+      "loss": 1.6528,
+      "step": 4885
+    },
+    {
+      "epoch": 4.510500807754442,
+      "grad_norm": 3.103515625,
+      "learning_rate": 6.747164776517679e-05,
+      "loss": 1.3975,
+      "step": 4886
+    },
+    {
+      "epoch": 4.511423955688899,
+      "grad_norm": 4.36328125,
+      "learning_rate": 6.746497665110075e-05,
+      "loss": 1.2919,
+      "step": 4887
+    },
+    {
+      "epoch": 4.512347103623355,
+      "grad_norm": 4.08984375,
+      "learning_rate": 6.74583055370247e-05,
+      "loss": 1.4567,
+      "step": 4888
+    },
+    {
+      "epoch": 4.513270251557812,
+      "grad_norm": 3.615234375,
+      "learning_rate": 6.745163442294864e-05,
+      "loss": 1.6987,
+      "step": 4889
+    },
+    {
+      "epoch": 4.514193399492268,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.744496330887258e-05,
+      "loss": 1.4714,
+      "step": 4890
+    },
+    {
+      "epoch": 4.515116547426725,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.743829219479654e-05,
+      "loss": 1.1357,
+      "step": 4891
+    },
+    {
+      "epoch": 4.516039695361181,
+      "grad_norm": 4.0546875,
+      "learning_rate": 6.743162108072049e-05,
+      "loss": 1.4246,
+      "step": 4892
+    },
+    {
+      "epoch": 4.516962843295638,
+      "grad_norm": 3.958984375,
+      "learning_rate": 6.742494996664443e-05,
+      "loss": 1.3745,
+      "step": 4893
+    },
+    {
+      "epoch": 4.517885991230095,
+      "grad_norm": 4.0859375,
+      "learning_rate": 6.741827885256838e-05,
+      "loss": 1.278,
+      "step": 4894
+    },
+    {
+      "epoch": 4.518809139164551,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.741160773849234e-05,
+      "loss": 1.3528,
+      "step": 4895
+    },
+    {
+      "epoch": 4.519732287099008,
+      "grad_norm": 4.45703125,
+      "learning_rate": 6.740493662441628e-05,
+      "loss": 1.521,
+      "step": 4896
+    },
+    {
+      "epoch": 4.520655435033464,
+      "grad_norm": 3.58203125,
+      "learning_rate": 6.739826551034023e-05,
+      "loss": 1.5742,
+      "step": 4897
+    },
+    {
+      "epoch": 4.521578582967921,
+      "grad_norm": 3.27734375,
+      "learning_rate": 6.739159439626418e-05,
+      "loss": 1.4441,
+      "step": 4898
+    },
+    {
+      "epoch": 4.522501730902377,
+      "grad_norm": 4.9921875,
+      "learning_rate": 6.738492328218813e-05,
+      "loss": 1.146,
+      "step": 4899
+    },
+    {
+      "epoch": 4.523424878836834,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.737825216811208e-05,
+      "loss": 1.5056,
+      "step": 4900
+    },
+    {
+      "epoch": 4.52434802677129,
+      "grad_norm": 3.96484375,
+      "learning_rate": 6.737158105403602e-05,
+      "loss": 1.3987,
+      "step": 4901
+    },
+    {
+      "epoch": 4.525271174705747,
+      "grad_norm": 4.16015625,
+      "learning_rate": 6.736490993995997e-05,
+      "loss": 0.9532,
+      "step": 4902
+    },
+    {
+      "epoch": 4.526194322640203,
+      "grad_norm": 4.7421875,
+      "learning_rate": 6.735823882588393e-05,
+      "loss": 1.48,
+      "step": 4903
+    },
+    {
+      "epoch": 4.52711747057466,
+      "grad_norm": 4.0390625,
+      "learning_rate": 6.735156771180787e-05,
+      "loss": 1.4912,
+      "step": 4904
+    },
+    {
+      "epoch": 4.528040618509116,
+      "grad_norm": 3.78515625,
+      "learning_rate": 6.734489659773182e-05,
+      "loss": 0.9042,
+      "step": 4905
+    },
+    {
+      "epoch": 4.528963766443573,
+      "grad_norm": 4.52734375,
+      "learning_rate": 6.733822548365577e-05,
+      "loss": 1.3245,
+      "step": 4906
+    },
+    {
+      "epoch": 4.529886914378029,
+      "grad_norm": 3.767578125,
+      "learning_rate": 6.733155436957972e-05,
+      "loss": 1.4875,
+      "step": 4907
+    },
+    {
+      "epoch": 4.530810062312486,
+      "grad_norm": 3.65234375,
+      "learning_rate": 6.732488325550367e-05,
+      "loss": 1.6323,
+      "step": 4908
+    },
+    {
+      "epoch": 4.5317332102469425,
+      "grad_norm": 4.44140625,
+      "learning_rate": 6.731821214142762e-05,
+      "loss": 1.4463,
+      "step": 4909
+    },
+    {
+      "epoch": 4.532656358181399,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.731154102735156e-05,
+      "loss": 1.5898,
+      "step": 4910
+    },
+    {
+      "epoch": 4.533579506115855,
+      "grad_norm": 4.5703125,
+      "learning_rate": 6.730486991327552e-05,
+      "loss": 1.4463,
+      "step": 4911
+    },
+    {
+      "epoch": 4.5345026540503115,
+      "grad_norm": 3.564453125,
+      "learning_rate": 6.729819879919948e-05,
+      "loss": 1.4329,
+      "step": 4912
+    },
+    {
+      "epoch": 4.535425801984768,
+      "grad_norm": 4.69140625,
+      "learning_rate": 6.729152768512342e-05,
+      "loss": 1.3511,
+      "step": 4913
+    },
+    {
+      "epoch": 4.536348949919224,
+      "grad_norm": 3.8359375,
+      "learning_rate": 6.728485657104737e-05,
+      "loss": 1.552,
+      "step": 4914
+    },
+    {
+      "epoch": 4.537272097853681,
+      "grad_norm": 4.32421875,
+      "learning_rate": 6.727818545697132e-05,
+      "loss": 1.4885,
+      "step": 4915
+    },
+    {
+      "epoch": 4.538195245788137,
+      "grad_norm": 4.01171875,
+      "learning_rate": 6.727151434289527e-05,
+      "loss": 1.4521,
+      "step": 4916
+    },
+    {
+      "epoch": 4.539118393722594,
+      "grad_norm": 3.7265625,
+      "learning_rate": 6.726484322881922e-05,
+      "loss": 1.415,
+      "step": 4917
+    },
+    {
+      "epoch": 4.54004154165705,
+      "grad_norm": 3.755859375,
+      "learning_rate": 6.725817211474317e-05,
+      "loss": 1.5547,
+      "step": 4918
+    },
+    {
+      "epoch": 4.540964689591507,
+      "grad_norm": 4.80078125,
+      "learning_rate": 6.725150100066711e-05,
+      "loss": 1.2991,
+      "step": 4919
+    },
+    {
+      "epoch": 4.541887837525963,
+      "grad_norm": 3.232421875,
+      "learning_rate": 6.724482988659107e-05,
+      "loss": 1.1682,
+      "step": 4920
+    },
+    {
+      "epoch": 4.54281098546042,
+      "grad_norm": 4.12109375,
+      "learning_rate": 6.723815877251501e-05,
+      "loss": 1.2534,
+      "step": 4921
+    },
+    {
+      "epoch": 4.543734133394876,
+      "grad_norm": 4.06640625,
+      "learning_rate": 6.723148765843896e-05,
+      "loss": 1.3254,
+      "step": 4922
+    },
+    {
+      "epoch": 4.544657281329333,
+      "grad_norm": 3.876953125,
+      "learning_rate": 6.72248165443629e-05,
+      "loss": 1.3104,
+      "step": 4923
+    },
+    {
+      "epoch": 4.545580429263789,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.721814543028686e-05,
+      "loss": 1.5227,
+      "step": 4924
+    },
+    {
+      "epoch": 4.546503577198246,
+      "grad_norm": 4.234375,
+      "learning_rate": 6.721147431621081e-05,
+      "loss": 1.3931,
+      "step": 4925
+    },
+    {
+      "epoch": 4.547426725132702,
+      "grad_norm": 3.482421875,
+      "learning_rate": 6.720480320213476e-05,
+      "loss": 1.5269,
+      "step": 4926
+    },
+    {
+      "epoch": 4.548349873067159,
+      "grad_norm": 4.3984375,
+      "learning_rate": 6.71981320880587e-05,
+      "loss": 1.2045,
+      "step": 4927
+    },
+    {
+      "epoch": 4.549273021001616,
+      "grad_norm": 4.50390625,
+      "learning_rate": 6.719146097398266e-05,
+      "loss": 1.4607,
+      "step": 4928
+    },
+    {
+      "epoch": 4.550196168936072,
+      "grad_norm": 4.34375,
+      "learning_rate": 6.71847898599066e-05,
+      "loss": 1.5784,
+      "step": 4929
+    },
+    {
+      "epoch": 4.551119316870529,
+      "grad_norm": 4.2578125,
+      "learning_rate": 6.717811874583055e-05,
+      "loss": 1.6157,
+      "step": 4930
+    },
+    {
+      "epoch": 4.552042464804985,
+      "grad_norm": 3.8203125,
+      "learning_rate": 6.71714476317545e-05,
+      "loss": 1.386,
+      "step": 4931
+    },
+    {
+      "epoch": 4.552965612739442,
+      "grad_norm": 5.37890625,
+      "learning_rate": 6.716477651767846e-05,
+      "loss": 1.2327,
+      "step": 4932
+    },
+    {
+      "epoch": 4.553888760673898,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.71581054036024e-05,
+      "loss": 1.5947,
+      "step": 4933
+    },
+    {
+      "epoch": 4.554811908608355,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.715143428952635e-05,
+      "loss": 1.4231,
+      "step": 4934
+    },
+    {
+      "epoch": 4.555735056542811,
+      "grad_norm": 3.564453125,
+      "learning_rate": 6.714476317545029e-05,
+      "loss": 1.3176,
+      "step": 4935
+    },
+    {
+      "epoch": 4.556658204477268,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.713809206137426e-05,
+      "loss": 1.4153,
+      "step": 4936
+    },
+    {
+      "epoch": 4.557581352411724,
+      "grad_norm": 3.681640625,
+      "learning_rate": 6.713142094729821e-05,
+      "loss": 0.8855,
+      "step": 4937
+    },
+    {
+      "epoch": 4.558504500346181,
+      "grad_norm": 3.158203125,
+      "learning_rate": 6.712474983322215e-05,
+      "loss": 1.1592,
+      "step": 4938
+    },
+    {
+      "epoch": 4.559427648280637,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.71180787191461e-05,
+      "loss": 1.4536,
+      "step": 4939
+    },
+    {
+      "epoch": 4.5603507962150935,
+      "grad_norm": 3.7421875,
+      "learning_rate": 6.711140760507006e-05,
+      "loss": 1.3071,
+      "step": 4940
+    },
+    {
+      "epoch": 4.56127394414955,
+      "grad_norm": 3.51171875,
+      "learning_rate": 6.7104736490994e-05,
+      "loss": 1.2004,
+      "step": 4941
+    },
+    {
+      "epoch": 4.5621970920840065,
+      "grad_norm": 3.623046875,
+      "learning_rate": 6.709806537691795e-05,
+      "loss": 1.2051,
+      "step": 4942
+    },
+    {
+      "epoch": 4.563120240018463,
+      "grad_norm": 3.447265625,
+      "learning_rate": 6.70913942628419e-05,
+      "loss": 1.4456,
+      "step": 4943
+    },
+    {
+      "epoch": 4.564043387952919,
+      "grad_norm": 4.19921875,
+      "learning_rate": 6.708472314876585e-05,
+      "loss": 1.2582,
+      "step": 4944
+    },
+    {
+      "epoch": 4.564966535887376,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.70780520346898e-05,
+      "loss": 1.1642,
+      "step": 4945
+    },
+    {
+      "epoch": 4.565889683821832,
+      "grad_norm": 4.47265625,
+      "learning_rate": 6.707138092061375e-05,
+      "loss": 1.4104,
+      "step": 4946
+    },
+    {
+      "epoch": 4.566812831756289,
+      "grad_norm": 3.736328125,
+      "learning_rate": 6.706470980653769e-05,
+      "loss": 1.5085,
+      "step": 4947
+    },
+    {
+      "epoch": 4.567735979690745,
+      "grad_norm": 4.0546875,
+      "learning_rate": 6.705803869246165e-05,
+      "loss": 1.7749,
+      "step": 4948
+    },
+    {
+      "epoch": 4.568659127625202,
+      "grad_norm": 4.0625,
+      "learning_rate": 6.70513675783856e-05,
+      "loss": 1.3274,
+      "step": 4949
+    },
+    {
+      "epoch": 4.569582275559658,
+      "grad_norm": 3.685546875,
+      "learning_rate": 6.704469646430954e-05,
+      "loss": 1.0981,
+      "step": 4950
+    },
+    {
+      "epoch": 4.570505423494115,
+      "grad_norm": 4.4921875,
+      "learning_rate": 6.703802535023349e-05,
+      "loss": 1.1569,
+      "step": 4951
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 5.2109375,
+      "learning_rate": 6.703135423615745e-05,
+      "loss": 1.3843,
+      "step": 4952
+    },
+    {
+      "epoch": 4.572351719363028,
+      "grad_norm": 3.681640625,
+      "learning_rate": 6.702468312208139e-05,
+      "loss": 1.2262,
+      "step": 4953
+    },
+    {
+      "epoch": 4.573274867297484,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.701801200800534e-05,
+      "loss": 1.6375,
+      "step": 4954
+    },
+    {
+      "epoch": 4.574198015231941,
+      "grad_norm": 3.349609375,
+      "learning_rate": 6.701134089392928e-05,
+      "loss": 1.057,
+      "step": 4955
+    },
+    {
+      "epoch": 4.575121163166397,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.700466977985324e-05,
+      "loss": 1.4102,
+      "step": 4956
+    },
+    {
+      "epoch": 4.576044311100854,
+      "grad_norm": 3.12890625,
+      "learning_rate": 6.699799866577719e-05,
+      "loss": 1.0286,
+      "step": 4957
+    },
+    {
+      "epoch": 4.576967459035311,
+      "grad_norm": 4.5234375,
+      "learning_rate": 6.699132755170113e-05,
+      "loss": 1.4988,
+      "step": 4958
+    },
+    {
+      "epoch": 4.577890606969767,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.698465643762508e-05,
+      "loss": 1.4224,
+      "step": 4959
+    },
+    {
+      "epoch": 4.578813754904223,
+      "grad_norm": 3.583984375,
+      "learning_rate": 6.697798532354904e-05,
+      "loss": 1.514,
+      "step": 4960
+    },
+    {
+      "epoch": 4.57973690283868,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.6971314209473e-05,
+      "loss": 1.6619,
+      "step": 4961
+    },
+    {
+      "epoch": 4.580660050773137,
+      "grad_norm": 4.0625,
+      "learning_rate": 6.696464309539694e-05,
+      "loss": 1.015,
+      "step": 4962
+    },
+    {
+      "epoch": 4.581583198707593,
+      "grad_norm": 3.822265625,
+      "learning_rate": 6.695797198132089e-05,
+      "loss": 1.4688,
+      "step": 4963
+    },
+    {
+      "epoch": 4.58250634664205,
+      "grad_norm": 3.029296875,
+      "learning_rate": 6.695130086724483e-05,
+      "loss": 1.306,
+      "step": 4964
+    },
+    {
+      "epoch": 4.583429494576506,
+      "grad_norm": 3.69140625,
+      "learning_rate": 6.694462975316879e-05,
+      "loss": 1.6355,
+      "step": 4965
+    },
+    {
+      "epoch": 4.584352642510963,
+      "grad_norm": 3.7421875,
+      "learning_rate": 6.693795863909274e-05,
+      "loss": 1.4358,
+      "step": 4966
+    },
+    {
+      "epoch": 4.585275790445419,
+      "grad_norm": 4.28125,
+      "learning_rate": 6.693128752501668e-05,
+      "loss": 1.3049,
+      "step": 4967
+    },
+    {
+      "epoch": 4.586198938379876,
+      "grad_norm": 3.427734375,
+      "learning_rate": 6.692461641094063e-05,
+      "loss": 1.3157,
+      "step": 4968
+    },
+    {
+      "epoch": 4.587122086314332,
+      "grad_norm": 4.18359375,
+      "learning_rate": 6.691794529686459e-05,
+      "loss": 1.1996,
+      "step": 4969
+    },
+    {
+      "epoch": 4.5880452342487885,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.691127418278853e-05,
+      "loss": 1.0302,
+      "step": 4970
+    },
+    {
+      "epoch": 4.5889683821832445,
+      "grad_norm": 3.677734375,
+      "learning_rate": 6.690460306871248e-05,
+      "loss": 1.3411,
+      "step": 4971
+    },
+    {
+      "epoch": 4.5898915301177015,
+      "grad_norm": 3.505859375,
+      "learning_rate": 6.689793195463642e-05,
+      "loss": 1.5498,
+      "step": 4972
+    },
+    {
+      "epoch": 4.5908146780521575,
+      "grad_norm": 3.787109375,
+      "learning_rate": 6.689126084056038e-05,
+      "loss": 1.6724,
+      "step": 4973
+    },
+    {
+      "epoch": 4.591737825986614,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.688458972648433e-05,
+      "loss": 1.5737,
+      "step": 4974
+    },
+    {
+      "epoch": 4.59266097392107,
+      "grad_norm": 3.45703125,
+      "learning_rate": 6.687791861240827e-05,
+      "loss": 1.2644,
+      "step": 4975
+    },
+    {
+      "epoch": 4.593584121855527,
+      "grad_norm": 4.1015625,
+      "learning_rate": 6.687124749833222e-05,
+      "loss": 1.4326,
+      "step": 4976
+    },
+    {
+      "epoch": 4.594507269789984,
+      "grad_norm": 4.0234375,
+      "learning_rate": 6.686457638425618e-05,
+      "loss": 1.337,
+      "step": 4977
+    },
+    {
+      "epoch": 4.59543041772444,
+      "grad_norm": 3.728515625,
+      "learning_rate": 6.685790527018012e-05,
+      "loss": 1.3625,
+      "step": 4978
+    },
+    {
+      "epoch": 4.596353565658897,
+      "grad_norm": 3.66796875,
+      "learning_rate": 6.685123415610407e-05,
+      "loss": 1.4534,
+      "step": 4979
+    },
+    {
+      "epoch": 4.597276713593353,
+      "grad_norm": 3.63671875,
+      "learning_rate": 6.684456304202803e-05,
+      "loss": 0.9515,
+      "step": 4980
+    },
+    {
+      "epoch": 4.59819986152781,
+      "grad_norm": 3.9765625,
+      "learning_rate": 6.683789192795197e-05,
+      "loss": 1.3777,
+      "step": 4981
+    },
+    {
+      "epoch": 4.599123009462266,
+      "grad_norm": 4.39453125,
+      "learning_rate": 6.683122081387592e-05,
+      "loss": 1.3284,
+      "step": 4982
+    },
+    {
+      "epoch": 4.600046157396723,
+      "grad_norm": 3.783203125,
+      "learning_rate": 6.682454969979986e-05,
+      "loss": 1.2159,
+      "step": 4983
+    },
+    {
+      "epoch": 4.600969305331179,
+      "grad_norm": 3.517578125,
+      "learning_rate": 6.681787858572382e-05,
+      "loss": 1.4587,
+      "step": 4984
+    },
+    {
+      "epoch": 4.601892453265636,
+      "grad_norm": 3.685546875,
+      "learning_rate": 6.681120747164777e-05,
+      "loss": 1.198,
+      "step": 4985
+    },
+    {
+      "epoch": 4.602815601200092,
+      "grad_norm": 3.771484375,
+      "learning_rate": 6.680453635757173e-05,
+      "loss": 1.5962,
+      "step": 4986
+    },
+    {
+      "epoch": 4.603738749134549,
+      "grad_norm": 3.544921875,
+      "learning_rate": 6.679786524349567e-05,
+      "loss": 1.3367,
+      "step": 4987
+    },
+    {
+      "epoch": 4.604661897069005,
+      "grad_norm": 3.900390625,
+      "learning_rate": 6.679119412941962e-05,
+      "loss": 1.1185,
+      "step": 4988
+    },
+    {
+      "epoch": 4.605585045003462,
+      "grad_norm": 5.33984375,
+      "learning_rate": 6.678452301534358e-05,
+      "loss": 1.4875,
+      "step": 4989
+    },
+    {
+      "epoch": 4.606508192937918,
+      "grad_norm": 4.4609375,
+      "learning_rate": 6.677785190126752e-05,
+      "loss": 1.3386,
+      "step": 4990
+    },
+    {
+      "epoch": 4.607431340872375,
+      "grad_norm": 3.4140625,
+      "learning_rate": 6.677118078719147e-05,
+      "loss": 1.5112,
+      "step": 4991
+    },
+    {
+      "epoch": 4.608354488806832,
+      "grad_norm": 4.00390625,
+      "learning_rate": 6.676450967311541e-05,
+      "loss": 1.5935,
+      "step": 4992
+    },
+    {
+      "epoch": 4.609277636741288,
+      "grad_norm": 3.900390625,
+      "learning_rate": 6.675783855903937e-05,
+      "loss": 1.4858,
+      "step": 4993
+    },
+    {
+      "epoch": 4.610200784675744,
+      "grad_norm": 2.96484375,
+      "learning_rate": 6.675116744496332e-05,
+      "loss": 1.257,
+      "step": 4994
+    },
+    {
+      "epoch": 4.611123932610201,
+      "grad_norm": 4.3203125,
+      "learning_rate": 6.674449633088726e-05,
+      "loss": 1.7297,
+      "step": 4995
+    },
+    {
+      "epoch": 4.612047080544658,
+      "grad_norm": 3.6796875,
+      "learning_rate": 6.67378252168112e-05,
+      "loss": 1.3743,
+      "step": 4996
+    },
+    {
+      "epoch": 4.612970228479114,
+      "grad_norm": 3.67578125,
+      "learning_rate": 6.673115410273517e-05,
+      "loss": 1.2465,
+      "step": 4997
+    },
+    {
+      "epoch": 4.613893376413571,
+      "grad_norm": 4.1484375,
+      "learning_rate": 6.672448298865911e-05,
+      "loss": 1.5518,
+      "step": 4998
+    },
+    {
+      "epoch": 4.614816524348027,
+      "grad_norm": 3.833984375,
+      "learning_rate": 6.671781187458306e-05,
+      "loss": 1.3496,
+      "step": 4999
+    },
+    {
+      "epoch": 4.6157396722824835,
+      "grad_norm": 3.93359375,
+      "learning_rate": 6.6711140760507e-05,
+      "loss": 1.3582,
+      "step": 5000
+    },
+    {
+      "epoch": 4.6157396722824835,
+      "eval_loss": 0.88427734375,
+      "eval_runtime": 23.3233,
+      "eval_samples_per_second": 977.606,
+      "eval_steps_per_second": 61.141,
+      "step": 5000
+    },
+    {
+      "epoch": 4.6166628202169395,
+      "grad_norm": 3.923828125,
+      "learning_rate": 6.670446964643096e-05,
+      "loss": 1.5759,
+      "step": 5001
+    },
+    {
+      "epoch": 4.617585968151396,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.66977985323549e-05,
+      "loss": 1.2448,
+      "step": 5002
+    },
+    {
+      "epoch": 4.6185091160858525,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.669112741827885e-05,
+      "loss": 1.4297,
+      "step": 5003
+    },
+    {
+      "epoch": 4.619432264020309,
+      "grad_norm": 3.966796875,
+      "learning_rate": 6.66844563042028e-05,
+      "loss": 1.5038,
+      "step": 5004
+    },
+    {
+      "epoch": 4.620355411954765,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.667778519012676e-05,
+      "loss": 1.6147,
+      "step": 5005
+    },
+    {
+      "epoch": 4.621278559889222,
+      "grad_norm": 3.830078125,
+      "learning_rate": 6.66711140760507e-05,
+      "loss": 1.3029,
+      "step": 5006
+    },
+    {
+      "epoch": 4.622201707823679,
+      "grad_norm": 3.93359375,
+      "learning_rate": 6.666444296197465e-05,
+      "loss": 1.6135,
+      "step": 5007
+    },
+    {
+      "epoch": 4.623124855758135,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.665777184789859e-05,
+      "loss": 1.5034,
+      "step": 5008
+    },
+    {
+      "epoch": 4.624048003692591,
+      "grad_norm": 4.47265625,
+      "learning_rate": 6.665110073382255e-05,
+      "loss": 1.2964,
+      "step": 5009
+    },
+    {
+      "epoch": 4.624971151627048,
+      "grad_norm": 3.583984375,
+      "learning_rate": 6.66444296197465e-05,
+      "loss": 1.449,
+      "step": 5010
+    },
+    {
+      "epoch": 4.625894299561505,
+      "grad_norm": 4.296875,
+      "learning_rate": 6.663775850567046e-05,
+      "loss": 1.4102,
+      "step": 5011
+    },
+    {
+      "epoch": 4.626817447495961,
+      "grad_norm": 4.265625,
+      "learning_rate": 6.66310873915944e-05,
+      "loss": 1.5979,
+      "step": 5012
+    },
+    {
+      "epoch": 4.627740595430418,
+      "grad_norm": 4.1171875,
+      "learning_rate": 6.662441627751835e-05,
+      "loss": 1.3971,
+      "step": 5013
+    },
+    {
+      "epoch": 4.628663743364874,
+      "grad_norm": 3.482421875,
+      "learning_rate": 6.66177451634423e-05,
+      "loss": 1.3905,
+      "step": 5014
+    },
+    {
+      "epoch": 4.629586891299331,
+      "grad_norm": 4.9296875,
+      "learning_rate": 6.661107404936625e-05,
+      "loss": 1.3297,
+      "step": 5015
+    },
+    {
+      "epoch": 4.630510039233787,
+      "grad_norm": 5.1796875,
+      "learning_rate": 6.66044029352902e-05,
+      "loss": 1.437,
+      "step": 5016
+    },
+    {
+      "epoch": 4.631433187168244,
+      "grad_norm": 4.30078125,
+      "learning_rate": 6.659773182121414e-05,
+      "loss": 1.2388,
+      "step": 5017
+    },
+    {
+      "epoch": 4.6323563351027,
+      "grad_norm": 5.05078125,
+      "learning_rate": 6.65910607071381e-05,
+      "loss": 1.5845,
+      "step": 5018
+    },
+    {
+      "epoch": 4.633279483037157,
+      "grad_norm": 4.609375,
+      "learning_rate": 6.658438959306205e-05,
+      "loss": 1.0611,
+      "step": 5019
+    },
+    {
+      "epoch": 4.634202630971613,
+      "grad_norm": 3.904296875,
+      "learning_rate": 6.657771847898599e-05,
+      "loss": 1.3364,
+      "step": 5020
+    },
+    {
+      "epoch": 4.63512577890607,
+      "grad_norm": 4.3359375,
+      "learning_rate": 6.657104736490995e-05,
+      "loss": 1.3538,
+      "step": 5021
+    },
+    {
+      "epoch": 4.636048926840526,
+      "grad_norm": 3.896484375,
+      "learning_rate": 6.65643762508339e-05,
+      "loss": 1.6531,
+      "step": 5022
+    },
+    {
+      "epoch": 4.636972074774983,
+      "grad_norm": 3.41796875,
+      "learning_rate": 6.655770513675784e-05,
+      "loss": 1.5571,
+      "step": 5023
+    },
+    {
+      "epoch": 4.637895222709439,
+      "grad_norm": 4.9921875,
+      "learning_rate": 6.655103402268179e-05,
+      "loss": 1.5458,
+      "step": 5024
+    },
+    {
+      "epoch": 4.638818370643896,
+      "grad_norm": 4.48046875,
+      "learning_rate": 6.654436290860575e-05,
+      "loss": 1.5273,
+      "step": 5025
+    },
+    {
+      "epoch": 4.639741518578353,
+      "grad_norm": 4.79296875,
+      "learning_rate": 6.653769179452969e-05,
+      "loss": 1.4529,
+      "step": 5026
+    },
+    {
+      "epoch": 4.640664666512809,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.653102068045364e-05,
+      "loss": 1.4272,
+      "step": 5027
+    },
+    {
+      "epoch": 4.6415878144472655,
+      "grad_norm": 4.35546875,
+      "learning_rate": 6.652434956637758e-05,
+      "loss": 1.3303,
+      "step": 5028
+    },
+    {
+      "epoch": 4.642510962381722,
+      "grad_norm": 3.244140625,
+      "learning_rate": 6.651767845230154e-05,
+      "loss": 1.4409,
+      "step": 5029
+    },
+    {
+      "epoch": 4.6434341103161785,
+      "grad_norm": 3.478515625,
+      "learning_rate": 6.651100733822549e-05,
+      "loss": 1.4072,
+      "step": 5030
+    },
+    {
+      "epoch": 4.6443572582506345,
+      "grad_norm": 3.541015625,
+      "learning_rate": 6.650433622414943e-05,
+      "loss": 1.3152,
+      "step": 5031
+    },
+    {
+      "epoch": 4.645280406185091,
+      "grad_norm": 3.9296875,
+      "learning_rate": 6.649766511007338e-05,
+      "loss": 1.4534,
+      "step": 5032
+    },
+    {
+      "epoch": 4.646203554119547,
+      "grad_norm": 3.408203125,
+      "learning_rate": 6.649099399599734e-05,
+      "loss": 1.3783,
+      "step": 5033
+    },
+    {
+      "epoch": 4.647126702054004,
+      "grad_norm": 4.72265625,
+      "learning_rate": 6.648432288192128e-05,
+      "loss": 1.1599,
+      "step": 5034
+    },
+    {
+      "epoch": 4.64804984998846,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.647765176784523e-05,
+      "loss": 1.1578,
+      "step": 5035
+    },
+    {
+      "epoch": 4.648972997922917,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.647098065376919e-05,
+      "loss": 1.2491,
+      "step": 5036
+    },
+    {
+      "epoch": 4.649896145857373,
+      "grad_norm": 3.544921875,
+      "learning_rate": 6.646430953969313e-05,
+      "loss": 1.5378,
+      "step": 5037
+    },
+    {
+      "epoch": 4.65081929379183,
+      "grad_norm": 3.2109375,
+      "learning_rate": 6.645763842561709e-05,
+      "loss": 1.668,
+      "step": 5038
+    },
+    {
+      "epoch": 4.651742441726286,
+      "grad_norm": 3.517578125,
+      "learning_rate": 6.645096731154104e-05,
+      "loss": 1.3824,
+      "step": 5039
+    },
+    {
+      "epoch": 4.652665589660743,
+      "grad_norm": 3.25390625,
+      "learning_rate": 6.644429619746498e-05,
+      "loss": 1.2317,
+      "step": 5040
+    },
+    {
+      "epoch": 4.6535887375952,
+      "grad_norm": 3.9921875,
+      "learning_rate": 6.643762508338893e-05,
+      "loss": 1.2878,
+      "step": 5041
+    },
+    {
+      "epoch": 4.654511885529656,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.643095396931289e-05,
+      "loss": 1.4042,
+      "step": 5042
+    },
+    {
+      "epoch": 4.655435033464112,
+      "grad_norm": 3.849609375,
+      "learning_rate": 6.642428285523683e-05,
+      "loss": 1.3879,
+      "step": 5043
+    },
+    {
+      "epoch": 4.656358181398569,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.641761174116078e-05,
+      "loss": 1.7017,
+      "step": 5044
+    },
+    {
+      "epoch": 4.657281329333026,
+      "grad_norm": 3.9296875,
+      "learning_rate": 6.641094062708472e-05,
+      "loss": 1.4387,
+      "step": 5045
+    },
+    {
+      "epoch": 4.658204477267482,
+      "grad_norm": 3.591796875,
+      "learning_rate": 6.640426951300868e-05,
+      "loss": 1.6008,
+      "step": 5046
+    },
+    {
+      "epoch": 4.659127625201939,
+      "grad_norm": 4.2109375,
+      "learning_rate": 6.639759839893263e-05,
+      "loss": 1.2881,
+      "step": 5047
+    },
+    {
+      "epoch": 4.660050773136395,
+      "grad_norm": 3.955078125,
+      "learning_rate": 6.639092728485657e-05,
+      "loss": 1.0269,
+      "step": 5048
+    },
+    {
+      "epoch": 4.660973921070852,
+      "grad_norm": 4.47265625,
+      "learning_rate": 6.638425617078052e-05,
+      "loss": 1.4482,
+      "step": 5049
+    },
+    {
+      "epoch": 4.661897069005308,
+      "grad_norm": 3.98046875,
+      "learning_rate": 6.637758505670448e-05,
+      "loss": 1.3441,
+      "step": 5050
+    },
+    {
+      "epoch": 4.662820216939765,
+      "grad_norm": 4.140625,
+      "learning_rate": 6.637091394262842e-05,
+      "loss": 1.2334,
+      "step": 5051
+    },
+    {
+      "epoch": 4.663743364874221,
+      "grad_norm": 5.0859375,
+      "learning_rate": 6.636424282855237e-05,
+      "loss": 1.3098,
+      "step": 5052
+    },
+    {
+      "epoch": 4.664666512808678,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.635757171447631e-05,
+      "loss": 1.192,
+      "step": 5053
+    },
+    {
+      "epoch": 4.665589660743134,
+      "grad_norm": 3.70703125,
+      "learning_rate": 6.635090060040027e-05,
+      "loss": 1.7473,
+      "step": 5054
+    },
+    {
+      "epoch": 4.666512808677591,
+      "grad_norm": 3.44921875,
+      "learning_rate": 6.634422948632422e-05,
+      "loss": 1.5867,
+      "step": 5055
+    },
+    {
+      "epoch": 4.667435956612048,
+      "grad_norm": 3.298828125,
+      "learning_rate": 6.633755837224816e-05,
+      "loss": 1.3561,
+      "step": 5056
+    },
+    {
+      "epoch": 4.668359104546504,
+      "grad_norm": 3.95703125,
+      "learning_rate": 6.633088725817211e-05,
+      "loss": 1.3667,
+      "step": 5057
+    },
+    {
+      "epoch": 4.66928225248096,
+      "grad_norm": 3.392578125,
+      "learning_rate": 6.632421614409607e-05,
+      "loss": 1.1263,
+      "step": 5058
+    },
+    {
+      "epoch": 4.6702054004154165,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.631754503002001e-05,
+      "loss": 1.3828,
+      "step": 5059
+    },
+    {
+      "epoch": 4.6711285483498735,
+      "grad_norm": 3.509765625,
+      "learning_rate": 6.631087391594396e-05,
+      "loss": 1.4319,
+      "step": 5060
+    },
+    {
+      "epoch": 4.6720516962843295,
+      "grad_norm": 4.52734375,
+      "learning_rate": 6.630420280186792e-05,
+      "loss": 1.2546,
+      "step": 5061
+    },
+    {
+      "epoch": 4.672974844218786,
+      "grad_norm": 3.861328125,
+      "learning_rate": 6.629753168779188e-05,
+      "loss": 1.4885,
+      "step": 5062
+    },
+    {
+      "epoch": 4.673897992153242,
+      "grad_norm": 3.9609375,
+      "learning_rate": 6.629086057371582e-05,
+      "loss": 1.5112,
+      "step": 5063
+    },
+    {
+      "epoch": 4.674821140087699,
+      "grad_norm": 3.0859375,
+      "learning_rate": 6.628418945963977e-05,
+      "loss": 1.418,
+      "step": 5064
+    },
+    {
+      "epoch": 4.675744288022155,
+      "grad_norm": 3.732421875,
+      "learning_rate": 6.627751834556371e-05,
+      "loss": 1.3677,
+      "step": 5065
+    },
+    {
+      "epoch": 4.676667435956612,
+      "grad_norm": 4.20703125,
+      "learning_rate": 6.627084723148767e-05,
+      "loss": 1.5601,
+      "step": 5066
+    },
+    {
+      "epoch": 4.677590583891068,
+      "grad_norm": 4.5546875,
+      "learning_rate": 6.626417611741162e-05,
+      "loss": 1.4854,
+      "step": 5067
+    },
+    {
+      "epoch": 4.678513731825525,
+      "grad_norm": 4.13671875,
+      "learning_rate": 6.625750500333556e-05,
+      "loss": 1.5469,
+      "step": 5068
+    },
+    {
+      "epoch": 4.679436879759981,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.625083388925951e-05,
+      "loss": 1.3938,
+      "step": 5069
+    },
+    {
+      "epoch": 4.680360027694438,
+      "grad_norm": 3.7578125,
+      "learning_rate": 6.624416277518347e-05,
+      "loss": 1.0836,
+      "step": 5070
+    },
+    {
+      "epoch": 4.681283175628894,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.623749166110741e-05,
+      "loss": 1.6672,
+      "step": 5071
+    },
+    {
+      "epoch": 4.682206323563351,
+      "grad_norm": 4.20703125,
+      "learning_rate": 6.623082054703136e-05,
+      "loss": 1.3141,
+      "step": 5072
+    },
+    {
+      "epoch": 4.683129471497807,
+      "grad_norm": 3.97265625,
+      "learning_rate": 6.62241494329553e-05,
+      "loss": 1.478,
+      "step": 5073
+    },
+    {
+      "epoch": 4.684052619432264,
+      "grad_norm": 3.423828125,
+      "learning_rate": 6.621747831887926e-05,
+      "loss": 1.4839,
+      "step": 5074
+    },
+    {
+      "epoch": 4.684975767366721,
+      "grad_norm": 4.1328125,
+      "learning_rate": 6.621080720480321e-05,
+      "loss": 1.3445,
+      "step": 5075
+    },
+    {
+      "epoch": 4.685898915301177,
+      "grad_norm": 3.720703125,
+      "learning_rate": 6.620413609072715e-05,
+      "loss": 1.4053,
+      "step": 5076
+    },
+    {
+      "epoch": 4.686822063235634,
+      "grad_norm": 4.171875,
+      "learning_rate": 6.61974649766511e-05,
+      "loss": 1.1813,
+      "step": 5077
+    },
+    {
+      "epoch": 4.68774521117009,
+      "grad_norm": 3.724609375,
+      "learning_rate": 6.619079386257506e-05,
+      "loss": 1.436,
+      "step": 5078
+    },
+    {
+      "epoch": 4.688668359104547,
+      "grad_norm": 3.919921875,
+      "learning_rate": 6.6184122748499e-05,
+      "loss": 1.4705,
+      "step": 5079
+    },
+    {
+      "epoch": 4.689591507039003,
+      "grad_norm": 3.5859375,
+      "learning_rate": 6.617745163442295e-05,
+      "loss": 1.4802,
+      "step": 5080
+    },
+    {
+      "epoch": 4.69051465497346,
+      "grad_norm": 3.822265625,
+      "learning_rate": 6.61707805203469e-05,
+      "loss": 1.3589,
+      "step": 5081
+    },
+    {
+      "epoch": 4.691437802907916,
+      "grad_norm": 3.232421875,
+      "learning_rate": 6.616410940627085e-05,
+      "loss": 1.5044,
+      "step": 5082
+    },
+    {
+      "epoch": 4.692360950842373,
+      "grad_norm": 3.55078125,
+      "learning_rate": 6.61574382921948e-05,
+      "loss": 1.4023,
+      "step": 5083
+    },
+    {
+      "epoch": 4.693284098776829,
+      "grad_norm": 3.787109375,
+      "learning_rate": 6.615076717811874e-05,
+      "loss": 1.5957,
+      "step": 5084
+    },
+    {
+      "epoch": 4.694207246711286,
+      "grad_norm": 3.8203125,
+      "learning_rate": 6.614409606404269e-05,
+      "loss": 1.5195,
+      "step": 5085
+    },
+    {
+      "epoch": 4.695130394645742,
+      "grad_norm": 3.833984375,
+      "learning_rate": 6.613742494996665e-05,
+      "loss": 1.4131,
+      "step": 5086
+    },
+    {
+      "epoch": 4.696053542580199,
+      "grad_norm": 3.955078125,
+      "learning_rate": 6.613075383589061e-05,
+      "loss": 1.512,
+      "step": 5087
+    },
+    {
+      "epoch": 4.696976690514655,
+      "grad_norm": 4.515625,
+      "learning_rate": 6.612408272181455e-05,
+      "loss": 1.1787,
+      "step": 5088
+    },
+    {
+      "epoch": 4.6978998384491115,
+      "grad_norm": 3.37890625,
+      "learning_rate": 6.61174116077385e-05,
+      "loss": 1.234,
+      "step": 5089
+    },
+    {
+      "epoch": 4.698822986383568,
+      "grad_norm": 3.69921875,
+      "learning_rate": 6.611074049366244e-05,
+      "loss": 1.438,
+      "step": 5090
+    },
+    {
+      "epoch": 4.6997461343180245,
+      "grad_norm": 3.17578125,
+      "learning_rate": 6.61040693795864e-05,
+      "loss": 1.4375,
+      "step": 5091
+    },
+    {
+      "epoch": 4.7006692822524805,
+      "grad_norm": 3.4921875,
+      "learning_rate": 6.609739826551035e-05,
+      "loss": 1.55,
+      "step": 5092
+    },
+    {
+      "epoch": 4.701592430186937,
+      "grad_norm": 4.09765625,
+      "learning_rate": 6.609072715143429e-05,
+      "loss": 1.3599,
+      "step": 5093
+    },
+    {
+      "epoch": 4.702515578121394,
+      "grad_norm": 3.7265625,
+      "learning_rate": 6.608405603735824e-05,
+      "loss": 0.7994,
+      "step": 5094
+    },
+    {
+      "epoch": 4.70343872605585,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.60773849232822e-05,
+      "loss": 1.4269,
+      "step": 5095
+    },
+    {
+      "epoch": 4.704361873990307,
+      "grad_norm": 3.42578125,
+      "learning_rate": 6.607071380920614e-05,
+      "loss": 1.1761,
+      "step": 5096
+    },
+    {
+      "epoch": 4.705285021924763,
+      "grad_norm": 3.423828125,
+      "learning_rate": 6.606404269513009e-05,
+      "loss": 1.6003,
+      "step": 5097
+    },
+    {
+      "epoch": 4.70620816985922,
+      "grad_norm": 3.095703125,
+      "learning_rate": 6.605737158105403e-05,
+      "loss": 1.5408,
+      "step": 5098
+    },
+    {
+      "epoch": 4.707131317793676,
+      "grad_norm": 3.658203125,
+      "learning_rate": 6.605070046697799e-05,
+      "loss": 1.4883,
+      "step": 5099
+    },
+    {
+      "epoch": 4.708054465728133,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.604402935290194e-05,
+      "loss": 1.2266,
+      "step": 5100
+    },
+    {
+      "epoch": 4.708977613662589,
+      "grad_norm": 3.693359375,
+      "learning_rate": 6.603735823882588e-05,
+      "loss": 1.1771,
+      "step": 5101
+    },
+    {
+      "epoch": 4.709900761597046,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.603068712474983e-05,
+      "loss": 1.2074,
+      "step": 5102
+    },
+    {
+      "epoch": 4.710823909531502,
+      "grad_norm": 4.484375,
+      "learning_rate": 6.602401601067379e-05,
+      "loss": 1.2205,
+      "step": 5103
+    },
+    {
+      "epoch": 4.711747057465959,
+      "grad_norm": 4.58984375,
+      "learning_rate": 6.601734489659773e-05,
+      "loss": 1.3911,
+      "step": 5104
+    },
+    {
+      "epoch": 4.712670205400415,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.601067378252168e-05,
+      "loss": 1.4412,
+      "step": 5105
+    },
+    {
+      "epoch": 4.713593353334872,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.600400266844562e-05,
+      "loss": 1.3635,
+      "step": 5106
+    },
+    {
+      "epoch": 4.714516501269328,
+      "grad_norm": 4.53515625,
+      "learning_rate": 6.599733155436958e-05,
+      "loss": 1.1187,
+      "step": 5107
+    },
+    {
+      "epoch": 4.715439649203785,
+      "grad_norm": 3.8046875,
+      "learning_rate": 6.599066044029353e-05,
+      "loss": 1.2583,
+      "step": 5108
+    },
+    {
+      "epoch": 4.716362797138242,
+      "grad_norm": 3.498046875,
+      "learning_rate": 6.598398932621747e-05,
+      "loss": 1.4694,
+      "step": 5109
+    },
+    {
+      "epoch": 4.717285945072698,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.597731821214142e-05,
+      "loss": 1.4675,
+      "step": 5110
+    },
+    {
+      "epoch": 4.718209093007155,
+      "grad_norm": 4.35546875,
+      "learning_rate": 6.597064709806538e-05,
+      "loss": 1.1328,
+      "step": 5111
+    },
+    {
+      "epoch": 4.719132240941611,
+      "grad_norm": 3.50390625,
+      "learning_rate": 6.596397598398934e-05,
+      "loss": 1.1682,
+      "step": 5112
+    },
+    {
+      "epoch": 4.720055388876068,
+      "grad_norm": 6.32421875,
+      "learning_rate": 6.595730486991328e-05,
+      "loss": 1.4634,
+      "step": 5113
+    },
+    {
+      "epoch": 4.720978536810524,
+      "grad_norm": 3.66796875,
+      "learning_rate": 6.595063375583723e-05,
+      "loss": 1.4653,
+      "step": 5114
+    },
+    {
+      "epoch": 4.721901684744981,
+      "grad_norm": 4.09765625,
+      "learning_rate": 6.594396264176119e-05,
+      "loss": 1.2377,
+      "step": 5115
+    },
+    {
+      "epoch": 4.722824832679437,
+      "grad_norm": 4.40234375,
+      "learning_rate": 6.593729152768513e-05,
+      "loss": 1.4695,
+      "step": 5116
+    },
+    {
+      "epoch": 4.723747980613894,
+      "grad_norm": 4.7421875,
+      "learning_rate": 6.593062041360908e-05,
+      "loss": 1.5066,
+      "step": 5117
+    },
+    {
+      "epoch": 4.72467112854835,
+      "grad_norm": 4.421875,
+      "learning_rate": 6.592394929953302e-05,
+      "loss": 1.3896,
+      "step": 5118
+    },
+    {
+      "epoch": 4.7255942764828065,
+      "grad_norm": 3.802734375,
+      "learning_rate": 6.591727818545698e-05,
+      "loss": 1.4268,
+      "step": 5119
+    },
+    {
+      "epoch": 4.7265174244172625,
+      "grad_norm": 3.75,
+      "learning_rate": 6.591060707138093e-05,
+      "loss": 1.1837,
+      "step": 5120
+    },
+    {
+      "epoch": 4.727440572351719,
+      "grad_norm": 3.529296875,
+      "learning_rate": 6.590393595730487e-05,
+      "loss": 1.4097,
+      "step": 5121
+    },
+    {
+      "epoch": 4.7283637202861755,
+      "grad_norm": 4.01171875,
+      "learning_rate": 6.589726484322882e-05,
+      "loss": 1.1263,
+      "step": 5122
+    },
+    {
+      "epoch": 4.729286868220632,
+      "grad_norm": 4.29296875,
+      "learning_rate": 6.589059372915278e-05,
+      "loss": 1.2268,
+      "step": 5123
+    },
+    {
+      "epoch": 4.730210016155089,
+      "grad_norm": 4.5390625,
+      "learning_rate": 6.588392261507672e-05,
+      "loss": 1.4297,
+      "step": 5124
+    },
+    {
+      "epoch": 4.731133164089545,
+      "grad_norm": 4.03515625,
+      "learning_rate": 6.587725150100067e-05,
+      "loss": 1.293,
+      "step": 5125
+    },
+    {
+      "epoch": 4.732056312024002,
+      "grad_norm": 4.08984375,
+      "learning_rate": 6.587058038692461e-05,
+      "loss": 1.0388,
+      "step": 5126
+    },
+    {
+      "epoch": 4.732979459958458,
+      "grad_norm": 4.46484375,
+      "learning_rate": 6.586390927284857e-05,
+      "loss": 1.202,
+      "step": 5127
+    },
+    {
+      "epoch": 4.733902607892915,
+      "grad_norm": 4.1328125,
+      "learning_rate": 6.585723815877252e-05,
+      "loss": 1.3145,
+      "step": 5128
+    },
+    {
+      "epoch": 4.734825755827371,
+      "grad_norm": 3.94140625,
+      "learning_rate": 6.585056704469646e-05,
+      "loss": 1.6272,
+      "step": 5129
+    },
+    {
+      "epoch": 4.735748903761828,
+      "grad_norm": 3.490234375,
+      "learning_rate": 6.584389593062041e-05,
+      "loss": 1.4756,
+      "step": 5130
+    },
+    {
+      "epoch": 4.736672051696284,
+      "grad_norm": 4.078125,
+      "learning_rate": 6.583722481654437e-05,
+      "loss": 1.5012,
+      "step": 5131
+    },
+    {
+      "epoch": 4.737595199630741,
+      "grad_norm": 4.640625,
+      "learning_rate": 6.583055370246831e-05,
+      "loss": 1.0079,
+      "step": 5132
+    },
+    {
+      "epoch": 4.738518347565197,
+      "grad_norm": 3.904296875,
+      "learning_rate": 6.582388258839226e-05,
+      "loss": 1.4143,
+      "step": 5133
+    },
+    {
+      "epoch": 4.739441495499654,
+      "grad_norm": 3.330078125,
+      "learning_rate": 6.58172114743162e-05,
+      "loss": 1.4958,
+      "step": 5134
+    },
+    {
+      "epoch": 4.74036464343411,
+      "grad_norm": 4.3046875,
+      "learning_rate": 6.581054036024016e-05,
+      "loss": 1.3669,
+      "step": 5135
+    },
+    {
+      "epoch": 4.741287791368567,
+      "grad_norm": 3.642578125,
+      "learning_rate": 6.580386924616411e-05,
+      "loss": 1.6335,
+      "step": 5136
+    },
+    {
+      "epoch": 4.742210939303023,
+      "grad_norm": 3.689453125,
+      "learning_rate": 6.579719813208807e-05,
+      "loss": 1.3228,
+      "step": 5137
+    },
+    {
+      "epoch": 4.74313408723748,
+      "grad_norm": 4.046875,
+      "learning_rate": 6.579052701801201e-05,
+      "loss": 1.3428,
+      "step": 5138
+    },
+    {
+      "epoch": 4.744057235171937,
+      "grad_norm": 4.49609375,
+      "learning_rate": 6.578385590393596e-05,
+      "loss": 1.3038,
+      "step": 5139
+    },
+    {
+      "epoch": 4.744980383106393,
+      "grad_norm": 3.875,
+      "learning_rate": 6.577718478985992e-05,
+      "loss": 1.4751,
+      "step": 5140
+    },
+    {
+      "epoch": 4.745903531040849,
+      "grad_norm": 3.2109375,
+      "learning_rate": 6.577051367578386e-05,
+      "loss": 1.3828,
+      "step": 5141
+    },
+    {
+      "epoch": 4.746826678975306,
+      "grad_norm": 3.349609375,
+      "learning_rate": 6.576384256170781e-05,
+      "loss": 1.6299,
+      "step": 5142
+    },
+    {
+      "epoch": 4.747749826909763,
+      "grad_norm": 3.439453125,
+      "learning_rate": 6.575717144763175e-05,
+      "loss": 1.446,
+      "step": 5143
+    },
+    {
+      "epoch": 4.748672974844219,
+      "grad_norm": 3.380859375,
+      "learning_rate": 6.575050033355571e-05,
+      "loss": 1.6213,
+      "step": 5144
+    },
+    {
+      "epoch": 4.749596122778676,
+      "grad_norm": 3.4453125,
+      "learning_rate": 6.574382921947966e-05,
+      "loss": 1.3928,
+      "step": 5145
+    },
+    {
+      "epoch": 4.750519270713132,
+      "grad_norm": 3.5546875,
+      "learning_rate": 6.57371581054036e-05,
+      "loss": 1.4302,
+      "step": 5146
+    },
+    {
+      "epoch": 4.7514424186475885,
+      "grad_norm": 3.5234375,
+      "learning_rate": 6.573048699132755e-05,
+      "loss": 1.2858,
+      "step": 5147
+    },
+    {
+      "epoch": 4.752365566582045,
+      "grad_norm": 3.24609375,
+      "learning_rate": 6.572381587725151e-05,
+      "loss": 1.4608,
+      "step": 5148
+    },
+    {
+      "epoch": 4.7532887145165015,
+      "grad_norm": 3.208984375,
+      "learning_rate": 6.571714476317545e-05,
+      "loss": 1.5056,
+      "step": 5149
+    },
+    {
+      "epoch": 4.7542118624509575,
+      "grad_norm": 4.015625,
+      "learning_rate": 6.57104736490994e-05,
+      "loss": 1.2789,
+      "step": 5150
+    },
+    {
+      "epoch": 4.755135010385414,
+      "grad_norm": 4.9765625,
+      "learning_rate": 6.570380253502334e-05,
+      "loss": 1.4395,
+      "step": 5151
+    },
+    {
+      "epoch": 4.75605815831987,
+      "grad_norm": 3.44921875,
+      "learning_rate": 6.56971314209473e-05,
+      "loss": 1.3108,
+      "step": 5152
+    },
+    {
+      "epoch": 4.756981306254327,
+      "grad_norm": 3.9296875,
+      "learning_rate": 6.569046030687125e-05,
+      "loss": 1.5242,
+      "step": 5153
+    },
+    {
+      "epoch": 4.757904454188783,
+      "grad_norm": 4.078125,
+      "learning_rate": 6.56837891927952e-05,
+      "loss": 1.4792,
+      "step": 5154
+    },
+    {
+      "epoch": 4.75882760212324,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.567711807871914e-05,
+      "loss": 1.2349,
+      "step": 5155
+    },
+    {
+      "epoch": 4.759750750057696,
+      "grad_norm": 4.62109375,
+      "learning_rate": 6.56704469646431e-05,
+      "loss": 1.4153,
+      "step": 5156
+    },
+    {
+      "epoch": 4.760673897992153,
+      "grad_norm": 3.892578125,
+      "learning_rate": 6.566377585056704e-05,
+      "loss": 1.339,
+      "step": 5157
+    },
+    {
+      "epoch": 4.76159704592661,
+      "grad_norm": 4.04296875,
+      "learning_rate": 6.565710473649099e-05,
+      "loss": 1.1205,
+      "step": 5158
+    },
+    {
+      "epoch": 4.762520193861066,
+      "grad_norm": 3.603515625,
+      "learning_rate": 6.565043362241494e-05,
+      "loss": 1.2808,
+      "step": 5159
+    },
+    {
+      "epoch": 4.763443341795523,
+      "grad_norm": 3.740234375,
+      "learning_rate": 6.56437625083389e-05,
+      "loss": 1.3428,
+      "step": 5160
+    },
+    {
+      "epoch": 4.764366489729979,
+      "grad_norm": 3.337890625,
+      "learning_rate": 6.563709139426284e-05,
+      "loss": 1.2842,
+      "step": 5161
+    },
+    {
+      "epoch": 4.765289637664436,
+      "grad_norm": 4.0,
+      "learning_rate": 6.56304202801868e-05,
+      "loss": 1.3466,
+      "step": 5162
+    },
+    {
+      "epoch": 4.766212785598892,
+      "grad_norm": 4.015625,
+      "learning_rate": 6.562374916611074e-05,
+      "loss": 1.3667,
+      "step": 5163
+    },
+    {
+      "epoch": 4.767135933533349,
+      "grad_norm": 3.505859375,
+      "learning_rate": 6.56170780520347e-05,
+      "loss": 1.2866,
+      "step": 5164
+    },
+    {
+      "epoch": 4.768059081467805,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.561040693795865e-05,
+      "loss": 1.0444,
+      "step": 5165
+    },
+    {
+      "epoch": 4.768982229402262,
+      "grad_norm": 3.732421875,
+      "learning_rate": 6.56037358238826e-05,
+      "loss": 1.3368,
+      "step": 5166
+    },
+    {
+      "epoch": 4.769905377336718,
+      "grad_norm": 4.3671875,
+      "learning_rate": 6.559706470980654e-05,
+      "loss": 1.4709,
+      "step": 5167
+    },
+    {
+      "epoch": 4.770828525271175,
+      "grad_norm": 3.529296875,
+      "learning_rate": 6.55903935957305e-05,
+      "loss": 1.4817,
+      "step": 5168
+    },
+    {
+      "epoch": 4.771751673205631,
+      "grad_norm": 3.787109375,
+      "learning_rate": 6.558372248165444e-05,
+      "loss": 1.3193,
+      "step": 5169
+    },
+    {
+      "epoch": 4.772674821140088,
+      "grad_norm": 3.54296875,
+      "learning_rate": 6.557705136757839e-05,
+      "loss": 1.4236,
+      "step": 5170
+    },
+    {
+      "epoch": 4.773597969074544,
+      "grad_norm": 4.37109375,
+      "learning_rate": 6.557038025350233e-05,
+      "loss": 1.5442,
+      "step": 5171
+    },
+    {
+      "epoch": 4.774521117009001,
+      "grad_norm": 3.1171875,
+      "learning_rate": 6.55637091394263e-05,
+      "loss": 1.5178,
+      "step": 5172
+    },
+    {
+      "epoch": 4.775444264943458,
+      "grad_norm": 3.54296875,
+      "learning_rate": 6.555703802535024e-05,
+      "loss": 1.3025,
+      "step": 5173
+    },
+    {
+      "epoch": 4.776367412877914,
+      "grad_norm": 3.970703125,
+      "learning_rate": 6.555036691127418e-05,
+      "loss": 1.3868,
+      "step": 5174
+    },
+    {
+      "epoch": 4.77729056081237,
+      "grad_norm": 3.68359375,
+      "learning_rate": 6.554369579719813e-05,
+      "loss": 1.3835,
+      "step": 5175
+    },
+    {
+      "epoch": 4.778213708746827,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.553702468312209e-05,
+      "loss": 1.3669,
+      "step": 5176
+    },
+    {
+      "epoch": 4.7791368566812835,
+      "grad_norm": 4.1953125,
+      "learning_rate": 6.553035356904603e-05,
+      "loss": 1.5706,
+      "step": 5177
+    },
+    {
+      "epoch": 4.7800600046157395,
+      "grad_norm": 3.201171875,
+      "learning_rate": 6.552368245496998e-05,
+      "loss": 1.519,
+      "step": 5178
+    },
+    {
+      "epoch": 4.7809831525501965,
+      "grad_norm": 4.1171875,
+      "learning_rate": 6.551701134089393e-05,
+      "loss": 1.4565,
+      "step": 5179
+    },
+    {
+      "epoch": 4.7819063004846525,
+      "grad_norm": 3.462890625,
+      "learning_rate": 6.551034022681788e-05,
+      "loss": 1.4185,
+      "step": 5180
+    },
+    {
+      "epoch": 4.782829448419109,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.550366911274183e-05,
+      "loss": 1.5178,
+      "step": 5181
+    },
+    {
+      "epoch": 4.783752596353565,
+      "grad_norm": 3.884765625,
+      "learning_rate": 6.549699799866578e-05,
+      "loss": 1.2971,
+      "step": 5182
+    },
+    {
+      "epoch": 4.784675744288022,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.549032688458972e-05,
+      "loss": 1.2478,
+      "step": 5183
+    },
+    {
+      "epoch": 4.785598892222478,
+      "grad_norm": 4.45703125,
+      "learning_rate": 6.548365577051368e-05,
+      "loss": 1.2578,
+      "step": 5184
+    },
+    {
+      "epoch": 4.786522040156935,
+      "grad_norm": 4.20703125,
+      "learning_rate": 6.547698465643762e-05,
+      "loss": 1.3574,
+      "step": 5185
+    },
+    {
+      "epoch": 4.787445188091391,
+      "grad_norm": 3.80078125,
+      "learning_rate": 6.547031354236157e-05,
+      "loss": 1.3159,
+      "step": 5186
+    },
+    {
+      "epoch": 4.788368336025848,
+      "grad_norm": 4.39453125,
+      "learning_rate": 6.546364242828553e-05,
+      "loss": 1.4253,
+      "step": 5187
+    },
+    {
+      "epoch": 4.789291483960305,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.545697131420947e-05,
+      "loss": 1.2338,
+      "step": 5188
+    },
+    {
+      "epoch": 4.790214631894761,
+      "grad_norm": 3.8671875,
+      "learning_rate": 6.545030020013343e-05,
+      "loss": 1.4297,
+      "step": 5189
+    },
+    {
+      "epoch": 4.791137779829217,
+      "grad_norm": 3.74609375,
+      "learning_rate": 6.544362908605738e-05,
+      "loss": 1.3872,
+      "step": 5190
+    },
+    {
+      "epoch": 4.792060927763674,
+      "grad_norm": 4.18359375,
+      "learning_rate": 6.543695797198132e-05,
+      "loss": 1.0447,
+      "step": 5191
+    },
+    {
+      "epoch": 4.792984075698131,
+      "grad_norm": 5.26953125,
+      "learning_rate": 6.543028685790527e-05,
+      "loss": 1.5767,
+      "step": 5192
+    },
+    {
+      "epoch": 4.793907223632587,
+      "grad_norm": 4.1875,
+      "learning_rate": 6.542361574382923e-05,
+      "loss": 1.3159,
+      "step": 5193
+    },
+    {
+      "epoch": 4.794830371567044,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.541694462975317e-05,
+      "loss": 1.4971,
+      "step": 5194
+    },
+    {
+      "epoch": 4.7957535195015,
+      "grad_norm": 4.26171875,
+      "learning_rate": 6.541027351567712e-05,
+      "loss": 1.4049,
+      "step": 5195
+    },
+    {
+      "epoch": 4.796676667435957,
+      "grad_norm": 4.28515625,
+      "learning_rate": 6.540360240160107e-05,
+      "loss": 1.4114,
+      "step": 5196
+    },
+    {
+      "epoch": 4.797599815370413,
+      "grad_norm": 3.873046875,
+      "learning_rate": 6.539693128752502e-05,
+      "loss": 1.4263,
+      "step": 5197
+    },
+    {
+      "epoch": 4.79852296330487,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.539026017344897e-05,
+      "loss": 1.6462,
+      "step": 5198
+    },
+    {
+      "epoch": 4.799446111239326,
+      "grad_norm": 3.654296875,
+      "learning_rate": 6.538358905937292e-05,
+      "loss": 1.3887,
+      "step": 5199
+    },
+    {
+      "epoch": 4.800369259173783,
+      "grad_norm": 3.66015625,
+      "learning_rate": 6.537691794529686e-05,
+      "loss": 1.3357,
+      "step": 5200
+    },
+    {
+      "epoch": 4.801292407108239,
+      "grad_norm": 3.404296875,
+      "learning_rate": 6.537024683122082e-05,
+      "loss": 1.4497,
+      "step": 5201
+    },
+    {
+      "epoch": 4.802215555042696,
+      "grad_norm": 3.4609375,
+      "learning_rate": 6.536357571714477e-05,
+      "loss": 1.5129,
+      "step": 5202
+    },
+    {
+      "epoch": 4.803138702977152,
+      "grad_norm": 3.91796875,
+      "learning_rate": 6.535690460306871e-05,
+      "loss": 1.3608,
+      "step": 5203
+    },
+    {
+      "epoch": 4.804061850911609,
+      "grad_norm": 3.48046875,
+      "learning_rate": 6.535023348899266e-05,
+      "loss": 1.6421,
+      "step": 5204
+    },
+    {
+      "epoch": 4.804984998846065,
+      "grad_norm": 4.35546875,
+      "learning_rate": 6.534356237491661e-05,
+      "loss": 1.5442,
+      "step": 5205
+    },
+    {
+      "epoch": 4.805908146780522,
+      "grad_norm": 3.619140625,
+      "learning_rate": 6.533689126084056e-05,
+      "loss": 1.6172,
+      "step": 5206
+    },
+    {
+      "epoch": 4.8068312947149785,
+      "grad_norm": 4.98046875,
+      "learning_rate": 6.53302201467645e-05,
+      "loss": 1.3982,
+      "step": 5207
+    },
+    {
+      "epoch": 4.8077544426494345,
+      "grad_norm": 3.71875,
+      "learning_rate": 6.532354903268846e-05,
+      "loss": 1.3109,
+      "step": 5208
+    },
+    {
+      "epoch": 4.808677590583891,
+      "grad_norm": 4.07421875,
+      "learning_rate": 6.531687791861241e-05,
+      "loss": 1.098,
+      "step": 5209
+    },
+    {
+      "epoch": 4.8096007385183475,
+      "grad_norm": 4.515625,
+      "learning_rate": 6.531020680453636e-05,
+      "loss": 1.2993,
+      "step": 5210
+    },
+    {
+      "epoch": 4.810523886452804,
+      "grad_norm": 3.634765625,
+      "learning_rate": 6.53035356904603e-05,
+      "loss": 1.604,
+      "step": 5211
+    },
+    {
+      "epoch": 4.81144703438726,
+      "grad_norm": 4.796875,
+      "learning_rate": 6.529686457638426e-05,
+      "loss": 1.3809,
+      "step": 5212
+    },
+    {
+      "epoch": 4.812370182321717,
+      "grad_norm": 3.79296875,
+      "learning_rate": 6.529019346230822e-05,
+      "loss": 1.3071,
+      "step": 5213
+    },
+    {
+      "epoch": 4.813293330256173,
+      "grad_norm": 3.232421875,
+      "learning_rate": 6.528352234823216e-05,
+      "loss": 1.426,
+      "step": 5214
+    },
+    {
+      "epoch": 4.81421647819063,
+      "grad_norm": 3.29296875,
+      "learning_rate": 6.527685123415611e-05,
+      "loss": 1.2971,
+      "step": 5215
+    },
+    {
+      "epoch": 4.815139626125086,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.527018012008006e-05,
+      "loss": 1.4364,
+      "step": 5216
+    },
+    {
+      "epoch": 4.816062774059543,
+      "grad_norm": 4.10546875,
+      "learning_rate": 6.526350900600401e-05,
+      "loss": 1.5857,
+      "step": 5217
+    },
+    {
+      "epoch": 4.816985921993999,
+      "grad_norm": 3.548828125,
+      "learning_rate": 6.525683789192796e-05,
+      "loss": 1.5068,
+      "step": 5218
+    },
+    {
+      "epoch": 4.817909069928456,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.52501667778519e-05,
+      "loss": 1.5583,
+      "step": 5219
+    },
+    {
+      "epoch": 4.818832217862912,
+      "grad_norm": 3.806640625,
+      "learning_rate": 6.524349566377585e-05,
+      "loss": 1.2991,
+      "step": 5220
+    },
+    {
+      "epoch": 4.819755365797369,
+      "grad_norm": 3.89453125,
+      "learning_rate": 6.523682454969981e-05,
+      "loss": 1.2576,
+      "step": 5221
+    },
+    {
+      "epoch": 4.820678513731826,
+      "grad_norm": 3.138671875,
+      "learning_rate": 6.523015343562375e-05,
+      "loss": 1.3486,
+      "step": 5222
+    },
+    {
+      "epoch": 4.821601661666282,
+      "grad_norm": 3.998046875,
+      "learning_rate": 6.52234823215477e-05,
+      "loss": 1.3103,
+      "step": 5223
+    },
+    {
+      "epoch": 4.822524809600738,
+      "grad_norm": 4.109375,
+      "learning_rate": 6.521681120747165e-05,
+      "loss": 1.595,
+      "step": 5224
+    },
+    {
+      "epoch": 4.823447957535195,
+      "grad_norm": 3.45703125,
+      "learning_rate": 6.52101400933956e-05,
+      "loss": 1.4939,
+      "step": 5225
+    },
+    {
+      "epoch": 4.824371105469652,
+      "grad_norm": 3.841796875,
+      "learning_rate": 6.520346897931955e-05,
+      "loss": 1.1176,
+      "step": 5226
+    },
+    {
+      "epoch": 4.825294253404108,
+      "grad_norm": 3.416015625,
+      "learning_rate": 6.51967978652435e-05,
+      "loss": 1.624,
+      "step": 5227
+    },
+    {
+      "epoch": 4.826217401338565,
+      "grad_norm": 3.498046875,
+      "learning_rate": 6.519012675116744e-05,
+      "loss": 1.626,
+      "step": 5228
+    },
+    {
+      "epoch": 4.827140549273021,
+      "grad_norm": 3.763671875,
+      "learning_rate": 6.51834556370914e-05,
+      "loss": 1.7644,
+      "step": 5229
+    },
+    {
+      "epoch": 4.828063697207478,
+      "grad_norm": 5.40625,
+      "learning_rate": 6.517678452301535e-05,
+      "loss": 0.9122,
+      "step": 5230
+    },
+    {
+      "epoch": 4.828986845141934,
+      "grad_norm": 3.505859375,
+      "learning_rate": 6.517011340893929e-05,
+      "loss": 1.21,
+      "step": 5231
+    },
+    {
+      "epoch": 4.829909993076391,
+      "grad_norm": 3.12109375,
+      "learning_rate": 6.516344229486324e-05,
+      "loss": 1.0279,
+      "step": 5232
+    },
+    {
+      "epoch": 4.830833141010847,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.51567711807872e-05,
+      "loss": 1.4436,
+      "step": 5233
+    },
+    {
+      "epoch": 4.831756288945304,
+      "grad_norm": 4.1953125,
+      "learning_rate": 6.515010006671114e-05,
+      "loss": 1.2928,
+      "step": 5234
+    },
+    {
+      "epoch": 4.83267943687976,
+      "grad_norm": 4.28515625,
+      "learning_rate": 6.514342895263509e-05,
+      "loss": 1.1614,
+      "step": 5235
+    },
+    {
+      "epoch": 4.833602584814217,
+      "grad_norm": 4.05859375,
+      "learning_rate": 6.513675783855903e-05,
+      "loss": 1.5991,
+      "step": 5236
+    },
+    {
+      "epoch": 4.834525732748673,
+      "grad_norm": 3.6796875,
+      "learning_rate": 6.513008672448299e-05,
+      "loss": 1.4255,
+      "step": 5237
+    },
+    {
+      "epoch": 4.8354488806831295,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.512341561040695e-05,
+      "loss": 1.5054,
+      "step": 5238
+    },
+    {
+      "epoch": 4.8363720286175855,
+      "grad_norm": 4.390625,
+      "learning_rate": 6.51167444963309e-05,
+      "loss": 1.5557,
+      "step": 5239
+    },
+    {
+      "epoch": 4.837295176552042,
+      "grad_norm": 4.14453125,
+      "learning_rate": 6.511007338225484e-05,
+      "loss": 1.6602,
+      "step": 5240
+    },
+    {
+      "epoch": 4.838218324486499,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.510340226817879e-05,
+      "loss": 1.1817,
+      "step": 5241
+    },
+    {
+      "epoch": 4.839141472420955,
+      "grad_norm": 3.822265625,
+      "learning_rate": 6.509673115410274e-05,
+      "loss": 1.4883,
+      "step": 5242
+    },
+    {
+      "epoch": 4.840064620355412,
+      "grad_norm": 4.3671875,
+      "learning_rate": 6.509006004002669e-05,
+      "loss": 1.5645,
+      "step": 5243
+    },
+    {
+      "epoch": 4.840987768289868,
+      "grad_norm": 3.6796875,
+      "learning_rate": 6.508338892595064e-05,
+      "loss": 1.6858,
+      "step": 5244
+    },
+    {
+      "epoch": 4.841910916224325,
+      "grad_norm": 3.638671875,
+      "learning_rate": 6.507671781187458e-05,
+      "loss": 1.5229,
+      "step": 5245
+    },
+    {
+      "epoch": 4.842834064158781,
+      "grad_norm": 3.791015625,
+      "learning_rate": 6.507004669779854e-05,
+      "loss": 1.5981,
+      "step": 5246
+    },
+    {
+      "epoch": 4.843757212093238,
+      "grad_norm": 4.45703125,
+      "learning_rate": 6.506337558372249e-05,
+      "loss": 1.5515,
+      "step": 5247
+    },
+    {
+      "epoch": 4.844680360027694,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.505670446964643e-05,
+      "loss": 1.3481,
+      "step": 5248
+    },
+    {
+      "epoch": 4.845603507962151,
+      "grad_norm": 3.6484375,
+      "learning_rate": 6.505003335557039e-05,
+      "loss": 1.3394,
+      "step": 5249
+    },
+    {
+      "epoch": 4.846526655896607,
+      "grad_norm": 4.08203125,
+      "learning_rate": 6.504336224149434e-05,
+      "loss": 1.4451,
+      "step": 5250
+    },
+    {
+      "epoch": 4.847449803831064,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.503669112741828e-05,
+      "loss": 1.4365,
+      "step": 5251
+    },
+    {
+      "epoch": 4.84837295176552,
+      "grad_norm": 4.00390625,
+      "learning_rate": 6.503002001334223e-05,
+      "loss": 1.4741,
+      "step": 5252
+    },
+    {
+      "epoch": 4.849296099699977,
+      "grad_norm": 4.14453125,
+      "learning_rate": 6.502334889926619e-05,
+      "loss": 1.4714,
+      "step": 5253
+    },
+    {
+      "epoch": 4.850219247634433,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.501667778519013e-05,
+      "loss": 1.3096,
+      "step": 5254
+    },
+    {
+      "epoch": 4.85114239556889,
+      "grad_norm": 4.0,
+      "learning_rate": 6.501000667111408e-05,
+      "loss": 1.4692,
+      "step": 5255
+    },
+    {
+      "epoch": 4.852065543503347,
+      "grad_norm": 3.634765625,
+      "learning_rate": 6.500333555703802e-05,
+      "loss": 1.5537,
+      "step": 5256
+    },
+    {
+      "epoch": 4.852988691437803,
+      "grad_norm": 4.23828125,
+      "learning_rate": 6.499666444296198e-05,
+      "loss": 1.5867,
+      "step": 5257
+    },
+    {
+      "epoch": 4.85391183937226,
+      "grad_norm": 3.33984375,
+      "learning_rate": 6.498999332888593e-05,
+      "loss": 1.3167,
+      "step": 5258
+    },
+    {
+      "epoch": 4.854834987306716,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.498332221480987e-05,
+      "loss": 1.6763,
+      "step": 5259
+    },
+    {
+      "epoch": 4.855758135241173,
+      "grad_norm": 3.740234375,
+      "learning_rate": 6.497665110073382e-05,
+      "loss": 1.3416,
+      "step": 5260
+    },
+    {
+      "epoch": 4.856681283175629,
+      "grad_norm": 3.392578125,
+      "learning_rate": 6.496997998665778e-05,
+      "loss": 1.4146,
+      "step": 5261
+    },
+    {
+      "epoch": 4.857604431110086,
+      "grad_norm": 5.71484375,
+      "learning_rate": 6.496330887258173e-05,
+      "loss": 1.4111,
+      "step": 5262
+    },
+    {
+      "epoch": 4.858527579044542,
+      "grad_norm": 3.951171875,
+      "learning_rate": 6.495663775850568e-05,
+      "loss": 1.4854,
+      "step": 5263
+    },
+    {
+      "epoch": 4.859450726978999,
+      "grad_norm": 4.08203125,
+      "learning_rate": 6.494996664442963e-05,
+      "loss": 1.2645,
+      "step": 5264
+    },
+    {
+      "epoch": 4.860373874913455,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.494329553035357e-05,
+      "loss": 1.3633,
+      "step": 5265
+    },
+    {
+      "epoch": 4.8612970228479115,
+      "grad_norm": 3.4296875,
+      "learning_rate": 6.493662441627753e-05,
+      "loss": 1.2552,
+      "step": 5266
+    },
+    {
+      "epoch": 4.862220170782368,
+      "grad_norm": 3.41796875,
+      "learning_rate": 6.492995330220148e-05,
+      "loss": 1.2521,
+      "step": 5267
+    },
+    {
+      "epoch": 4.8631433187168245,
+      "grad_norm": 4.05078125,
+      "learning_rate": 6.492328218812542e-05,
+      "loss": 1.6282,
+      "step": 5268
+    },
+    {
+      "epoch": 4.8640664666512805,
+      "grad_norm": 4.1796875,
+      "learning_rate": 6.491661107404937e-05,
+      "loss": 1.3579,
+      "step": 5269
+    },
+    {
+      "epoch": 4.864989614585737,
+      "grad_norm": 3.9296875,
+      "learning_rate": 6.490993995997333e-05,
+      "loss": 1.2773,
+      "step": 5270
+    },
+    {
+      "epoch": 4.865912762520194,
+      "grad_norm": 4.39453125,
+      "learning_rate": 6.490326884589727e-05,
+      "loss": 1.564,
+      "step": 5271
+    },
+    {
+      "epoch": 4.86683591045465,
+      "grad_norm": 4.2265625,
+      "learning_rate": 6.489659773182122e-05,
+      "loss": 1.4604,
+      "step": 5272
+    },
+    {
+      "epoch": 4.867759058389106,
+      "grad_norm": 4.3046875,
+      "learning_rate": 6.488992661774516e-05,
+      "loss": 1.4695,
+      "step": 5273
+    },
+    {
+      "epoch": 4.868682206323563,
+      "grad_norm": 4.2734375,
+      "learning_rate": 6.488325550366912e-05,
+      "loss": 1.6357,
+      "step": 5274
+    },
+    {
+      "epoch": 4.86960535425802,
+      "grad_norm": 4.1328125,
+      "learning_rate": 6.487658438959307e-05,
+      "loss": 1.6423,
+      "step": 5275
+    },
+    {
+      "epoch": 4.870528502192476,
+      "grad_norm": 4.2109375,
+      "learning_rate": 6.486991327551701e-05,
+      "loss": 1.4023,
+      "step": 5276
+    },
+    {
+      "epoch": 4.871451650126933,
+      "grad_norm": 4.265625,
+      "learning_rate": 6.486324216144096e-05,
+      "loss": 1.5647,
+      "step": 5277
+    },
+    {
+      "epoch": 4.872374798061389,
+      "grad_norm": 4.54296875,
+      "learning_rate": 6.485657104736492e-05,
+      "loss": 1.342,
+      "step": 5278
+    },
+    {
+      "epoch": 4.873297945995846,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.484989993328886e-05,
+      "loss": 1.1936,
+      "step": 5279
+    },
+    {
+      "epoch": 4.874221093930302,
+      "grad_norm": 3.365234375,
+      "learning_rate": 6.484322881921281e-05,
+      "loss": 1.4067,
+      "step": 5280
+    },
+    {
+      "epoch": 4.875144241864759,
+      "grad_norm": 3.439453125,
+      "learning_rate": 6.483655770513675e-05,
+      "loss": 1.2238,
+      "step": 5281
+    },
+    {
+      "epoch": 4.876067389799215,
+      "grad_norm": 3.556640625,
+      "learning_rate": 6.482988659106071e-05,
+      "loss": 1.5151,
+      "step": 5282
+    },
+    {
+      "epoch": 4.876990537733672,
+      "grad_norm": 3.357421875,
+      "learning_rate": 6.482321547698466e-05,
+      "loss": 1.403,
+      "step": 5283
+    },
+    {
+      "epoch": 4.877913685668128,
+      "grad_norm": 3.044921875,
+      "learning_rate": 6.48165443629086e-05,
+      "loss": 1.5442,
+      "step": 5284
+    },
+    {
+      "epoch": 4.878836833602585,
+      "grad_norm": 3.248046875,
+      "learning_rate": 6.480987324883255e-05,
+      "loss": 1.2584,
+      "step": 5285
+    },
+    {
+      "epoch": 4.879759981537041,
+      "grad_norm": 3.412109375,
+      "learning_rate": 6.48032021347565e-05,
+      "loss": 1.2941,
+      "step": 5286
+    },
+    {
+      "epoch": 4.880683129471498,
+      "grad_norm": 4.10546875,
+      "learning_rate": 6.479653102068047e-05,
+      "loss": 1.3508,
+      "step": 5287
+    },
+    {
+      "epoch": 4.881606277405954,
+      "grad_norm": 3.529296875,
+      "learning_rate": 6.478985990660441e-05,
+      "loss": 1.3408,
+      "step": 5288
+    },
+    {
+      "epoch": 4.882529425340411,
+      "grad_norm": 3.482421875,
+      "learning_rate": 6.478318879252836e-05,
+      "loss": 1.2654,
+      "step": 5289
+    },
+    {
+      "epoch": 4.883452573274868,
+      "grad_norm": 3.56640625,
+      "learning_rate": 6.477651767845232e-05,
+      "loss": 1.2898,
+      "step": 5290
+    },
+    {
+      "epoch": 4.884375721209324,
+      "grad_norm": 3.599609375,
+      "learning_rate": 6.476984656437626e-05,
+      "loss": 1.4194,
+      "step": 5291
+    },
+    {
+      "epoch": 4.885298869143781,
+      "grad_norm": 3.33984375,
+      "learning_rate": 6.47631754503002e-05,
+      "loss": 1.2251,
+      "step": 5292
+    },
+    {
+      "epoch": 4.886222017078237,
+      "grad_norm": 3.732421875,
+      "learning_rate": 6.475650433622415e-05,
+      "loss": 1.4451,
+      "step": 5293
+    },
+    {
+      "epoch": 4.887145165012694,
+      "grad_norm": 4.03515625,
+      "learning_rate": 6.474983322214811e-05,
+      "loss": 1.1792,
+      "step": 5294
+    },
+    {
+      "epoch": 4.88806831294715,
+      "grad_norm": 3.541015625,
+      "learning_rate": 6.474316210807206e-05,
+      "loss": 1.2574,
+      "step": 5295
+    },
+    {
+      "epoch": 4.8889914608816065,
+      "grad_norm": 3.294921875,
+      "learning_rate": 6.4736490993996e-05,
+      "loss": 1.2994,
+      "step": 5296
+    },
+    {
+      "epoch": 4.8899146088160625,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.472981987991995e-05,
+      "loss": 1.6162,
+      "step": 5297
+    },
+    {
+      "epoch": 4.8908377567505195,
+      "grad_norm": 4.15234375,
+      "learning_rate": 6.47231487658439e-05,
+      "loss": 1.2804,
+      "step": 5298
+    },
+    {
+      "epoch": 4.8917609046849755,
+      "grad_norm": 3.7578125,
+      "learning_rate": 6.471647765176785e-05,
+      "loss": 1.5884,
+      "step": 5299
+    },
+    {
+      "epoch": 4.892684052619432,
+      "grad_norm": 3.97265625,
+      "learning_rate": 6.47098065376918e-05,
+      "loss": 1.5991,
+      "step": 5300
+    },
+    {
+      "epoch": 4.893607200553888,
+      "grad_norm": 4.0546875,
+      "learning_rate": 6.470313542361574e-05,
+      "loss": 1.2216,
+      "step": 5301
+    },
+    {
+      "epoch": 4.894530348488345,
+      "grad_norm": 3.275390625,
+      "learning_rate": 6.46964643095397e-05,
+      "loss": 1.4255,
+      "step": 5302
+    },
+    {
+      "epoch": 4.895453496422801,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.468979319546365e-05,
+      "loss": 1.2483,
+      "step": 5303
+    },
+    {
+      "epoch": 4.896376644357258,
+      "grad_norm": 3.732421875,
+      "learning_rate": 6.468312208138759e-05,
+      "loss": 1.4102,
+      "step": 5304
+    },
+    {
+      "epoch": 4.897299792291715,
+      "grad_norm": 3.755859375,
+      "learning_rate": 6.467645096731154e-05,
+      "loss": 1.4673,
+      "step": 5305
+    },
+    {
+      "epoch": 4.898222940226171,
+      "grad_norm": 4.0546875,
+      "learning_rate": 6.46697798532355e-05,
+      "loss": 1.418,
+      "step": 5306
+    },
+    {
+      "epoch": 4.899146088160627,
+      "grad_norm": 3.509765625,
+      "learning_rate": 6.466310873915944e-05,
+      "loss": 1.3674,
+      "step": 5307
+    },
+    {
+      "epoch": 4.900069236095084,
+      "grad_norm": 3.775390625,
+      "learning_rate": 6.465643762508339e-05,
+      "loss": 1.4761,
+      "step": 5308
+    },
+    {
+      "epoch": 4.900992384029541,
+      "grad_norm": 4.56640625,
+      "learning_rate": 6.464976651100733e-05,
+      "loss": 1.6775,
+      "step": 5309
+    },
+    {
+      "epoch": 4.901915531963997,
+      "grad_norm": 3.66796875,
+      "learning_rate": 6.464309539693129e-05,
+      "loss": 1.7683,
+      "step": 5310
+    },
+    {
+      "epoch": 4.902838679898454,
+      "grad_norm": 4.0859375,
+      "learning_rate": 6.463642428285524e-05,
+      "loss": 1.4368,
+      "step": 5311
+    },
+    {
+      "epoch": 4.90376182783291,
+      "grad_norm": 3.337890625,
+      "learning_rate": 6.46297531687792e-05,
+      "loss": 1.1361,
+      "step": 5312
+    },
+    {
+      "epoch": 4.904684975767367,
+      "grad_norm": 4.6796875,
+      "learning_rate": 6.462308205470314e-05,
+      "loss": 1.418,
+      "step": 5313
+    },
+    {
+      "epoch": 4.905608123701823,
+      "grad_norm": 3.931640625,
+      "learning_rate": 6.461641094062709e-05,
+      "loss": 1.6111,
+      "step": 5314
+    },
+    {
+      "epoch": 4.90653127163628,
+      "grad_norm": 3.97265625,
+      "learning_rate": 6.460973982655105e-05,
+      "loss": 1.2922,
+      "step": 5315
+    },
+    {
+      "epoch": 4.907454419570736,
+      "grad_norm": 4.30078125,
+      "learning_rate": 6.460306871247499e-05,
+      "loss": 1.3007,
+      "step": 5316
+    },
+    {
+      "epoch": 4.908377567505193,
+      "grad_norm": 3.447265625,
+      "learning_rate": 6.459639759839894e-05,
+      "loss": 0.787,
+      "step": 5317
+    },
+    {
+      "epoch": 4.909300715439649,
+      "grad_norm": 3.72265625,
+      "learning_rate": 6.458972648432288e-05,
+      "loss": 1.4744,
+      "step": 5318
+    },
+    {
+      "epoch": 4.910223863374106,
+      "grad_norm": 4.5546875,
+      "learning_rate": 6.458305537024684e-05,
+      "loss": 1.4377,
+      "step": 5319
+    },
+    {
+      "epoch": 4.911147011308563,
+      "grad_norm": 3.7421875,
+      "learning_rate": 6.457638425617079e-05,
+      "loss": 1.2703,
+      "step": 5320
+    },
+    {
+      "epoch": 4.912070159243019,
+      "grad_norm": 3.82421875,
+      "learning_rate": 6.456971314209473e-05,
+      "loss": 1.5063,
+      "step": 5321
+    },
+    {
+      "epoch": 4.912993307177475,
+      "grad_norm": 3.69140625,
+      "learning_rate": 6.456304202801868e-05,
+      "loss": 1.4197,
+      "step": 5322
+    },
+    {
+      "epoch": 4.913916455111932,
+      "grad_norm": 3.90234375,
+      "learning_rate": 6.455637091394264e-05,
+      "loss": 1.2944,
+      "step": 5323
+    },
+    {
+      "epoch": 4.914839603046389,
+      "grad_norm": 3.056640625,
+      "learning_rate": 6.454969979986658e-05,
+      "loss": 1.3035,
+      "step": 5324
+    },
+    {
+      "epoch": 4.915762750980845,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.454302868579053e-05,
+      "loss": 1.2725,
+      "step": 5325
+    },
+    {
+      "epoch": 4.9166858989153015,
+      "grad_norm": 3.5546875,
+      "learning_rate": 6.453635757171447e-05,
+      "loss": 1.3372,
+      "step": 5326
+    },
+    {
+      "epoch": 4.9176090468497575,
+      "grad_norm": 3.6484375,
+      "learning_rate": 6.452968645763843e-05,
+      "loss": 1.2383,
+      "step": 5327
+    },
+    {
+      "epoch": 4.918532194784214,
+      "grad_norm": 3.31640625,
+      "learning_rate": 6.452301534356238e-05,
+      "loss": 1.2328,
+      "step": 5328
+    },
+    {
+      "epoch": 4.9194553427186705,
+      "grad_norm": 4.26171875,
+      "learning_rate": 6.451634422948632e-05,
+      "loss": 1.4092,
+      "step": 5329
+    },
+    {
+      "epoch": 4.920378490653127,
+      "grad_norm": 3.642578125,
+      "learning_rate": 6.450967311541027e-05,
+      "loss": 1.5583,
+      "step": 5330
+    },
+    {
+      "epoch": 4.921301638587583,
+      "grad_norm": 3.4921875,
+      "learning_rate": 6.450300200133423e-05,
+      "loss": 1.3899,
+      "step": 5331
+    },
+    {
+      "epoch": 4.92222478652204,
+      "grad_norm": 3.212890625,
+      "learning_rate": 6.449633088725817e-05,
+      "loss": 1.521,
+      "step": 5332
+    },
+    {
+      "epoch": 4.923147934456496,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.448965977318212e-05,
+      "loss": 1.3389,
+      "step": 5333
+    },
+    {
+      "epoch": 4.924071082390953,
+      "grad_norm": 3.544921875,
+      "learning_rate": 6.448298865910606e-05,
+      "loss": 1.1844,
+      "step": 5334
+    },
+    {
+      "epoch": 4.924994230325409,
+      "grad_norm": 2.875,
+      "learning_rate": 6.447631754503002e-05,
+      "loss": 1.0697,
+      "step": 5335
+    },
+    {
+      "epoch": 4.925917378259866,
+      "grad_norm": 3.08984375,
+      "learning_rate": 6.446964643095397e-05,
+      "loss": 1.5208,
+      "step": 5336
+    },
+    {
+      "epoch": 4.926840526194322,
+      "grad_norm": 4.13671875,
+      "learning_rate": 6.446297531687793e-05,
+      "loss": 1.2509,
+      "step": 5337
+    },
+    {
+      "epoch": 4.927763674128779,
+      "grad_norm": 4.40625,
+      "learning_rate": 6.445630420280187e-05,
+      "loss": 1.7041,
+      "step": 5338
+    },
+    {
+      "epoch": 4.928686822063236,
+      "grad_norm": 3.951171875,
+      "learning_rate": 6.444963308872583e-05,
+      "loss": 1.519,
+      "step": 5339
+    },
+    {
+      "epoch": 4.929609969997692,
+      "grad_norm": 3.82421875,
+      "learning_rate": 6.444296197464978e-05,
+      "loss": 1.3523,
+      "step": 5340
+    },
+    {
+      "epoch": 4.930533117932149,
+      "grad_norm": 4.22265625,
+      "learning_rate": 6.443629086057372e-05,
+      "loss": 1.4998,
+      "step": 5341
+    },
+    {
+      "epoch": 4.931456265866605,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.442961974649767e-05,
+      "loss": 1.6423,
+      "step": 5342
+    },
+    {
+      "epoch": 4.932379413801062,
+      "grad_norm": 3.8203125,
+      "learning_rate": 6.442294863242163e-05,
+      "loss": 1.5029,
+      "step": 5343
+    },
+    {
+      "epoch": 4.933302561735518,
+      "grad_norm": 3.3984375,
+      "learning_rate": 6.441627751834557e-05,
+      "loss": 1.3105,
+      "step": 5344
+    },
+    {
+      "epoch": 4.934225709669975,
+      "grad_norm": 4.140625,
+      "learning_rate": 6.440960640426952e-05,
+      "loss": 1.2991,
+      "step": 5345
+    },
+    {
+      "epoch": 4.935148857604431,
+      "grad_norm": 4.1171875,
+      "learning_rate": 6.440293529019346e-05,
+      "loss": 1.5193,
+      "step": 5346
+    },
+    {
+      "epoch": 4.936072005538888,
+      "grad_norm": 3.498046875,
+      "learning_rate": 6.439626417611742e-05,
+      "loss": 1.4868,
+      "step": 5347
+    },
+    {
+      "epoch": 4.936995153473344,
+      "grad_norm": 3.775390625,
+      "learning_rate": 6.438959306204137e-05,
+      "loss": 1.3879,
+      "step": 5348
+    },
+    {
+      "epoch": 4.937918301407801,
+      "grad_norm": 4.25,
+      "learning_rate": 6.438292194796531e-05,
+      "loss": 1.4426,
+      "step": 5349
+    },
+    {
+      "epoch": 4.938841449342257,
+      "grad_norm": 5.1796875,
+      "learning_rate": 6.437625083388926e-05,
+      "loss": 1.2609,
+      "step": 5350
+    },
+    {
+      "epoch": 4.939764597276714,
+      "grad_norm": 3.767578125,
+      "learning_rate": 6.436957971981322e-05,
+      "loss": 1.4775,
+      "step": 5351
+    },
+    {
+      "epoch": 4.94068774521117,
+      "grad_norm": 3.92578125,
+      "learning_rate": 6.436290860573716e-05,
+      "loss": 1.4595,
+      "step": 5352
+    },
+    {
+      "epoch": 4.941610893145627,
+      "grad_norm": 4.33984375,
+      "learning_rate": 6.435623749166111e-05,
+      "loss": 1.47,
+      "step": 5353
+    },
+    {
+      "epoch": 4.9425340410800835,
+      "grad_norm": 3.955078125,
+      "learning_rate": 6.434956637758505e-05,
+      "loss": 1.51,
+      "step": 5354
+    },
+    {
+      "epoch": 4.94345718901454,
+      "grad_norm": 4.70703125,
+      "learning_rate": 6.434289526350901e-05,
+      "loss": 1.4161,
+      "step": 5355
+    },
+    {
+      "epoch": 4.944380336948996,
+      "grad_norm": 4.3671875,
+      "learning_rate": 6.433622414943296e-05,
+      "loss": 1.1753,
+      "step": 5356
+    },
+    {
+      "epoch": 4.9453034848834525,
+      "grad_norm": 4.56640625,
+      "learning_rate": 6.43295530353569e-05,
+      "loss": 1.2074,
+      "step": 5357
+    },
+    {
+      "epoch": 4.946226632817909,
+      "grad_norm": 3.517578125,
+      "learning_rate": 6.432288192128085e-05,
+      "loss": 1.3167,
+      "step": 5358
+    },
+    {
+      "epoch": 4.9471497807523654,
+      "grad_norm": 3.5703125,
+      "learning_rate": 6.431621080720481e-05,
+      "loss": 1.3188,
+      "step": 5359
+    },
+    {
+      "epoch": 4.948072928686822,
+      "grad_norm": 3.689453125,
+      "learning_rate": 6.430953969312875e-05,
+      "loss": 1.2661,
+      "step": 5360
+    },
+    {
+      "epoch": 4.948996076621278,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.43028685790527e-05,
+      "loss": 0.8624,
+      "step": 5361
+    },
+    {
+      "epoch": 4.949919224555735,
+      "grad_norm": 4.52734375,
+      "learning_rate": 6.429619746497666e-05,
+      "loss": 1.5603,
+      "step": 5362
+    },
+    {
+      "epoch": 4.950842372490191,
+      "grad_norm": 3.611328125,
+      "learning_rate": 6.42895263509006e-05,
+      "loss": 1.2799,
+      "step": 5363
+    },
+    {
+      "epoch": 4.951765520424648,
+      "grad_norm": 3.505859375,
+      "learning_rate": 6.428285523682456e-05,
+      "loss": 1.5122,
+      "step": 5364
+    },
+    {
+      "epoch": 4.952688668359104,
+      "grad_norm": 2.74609375,
+      "learning_rate": 6.427618412274851e-05,
+      "loss": 1.0675,
+      "step": 5365
+    },
+    {
+      "epoch": 4.953611816293561,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.426951300867245e-05,
+      "loss": 1.5217,
+      "step": 5366
+    },
+    {
+      "epoch": 4.954534964228017,
+      "grad_norm": 3.52734375,
+      "learning_rate": 6.42628418945964e-05,
+      "loss": 1.4111,
+      "step": 5367
+    },
+    {
+      "epoch": 4.955458112162474,
+      "grad_norm": 3.626953125,
+      "learning_rate": 6.425617078052036e-05,
+      "loss": 1.2648,
+      "step": 5368
+    },
+    {
+      "epoch": 4.956381260096931,
+      "grad_norm": 4.49609375,
+      "learning_rate": 6.42494996664443e-05,
+      "loss": 1.3586,
+      "step": 5369
+    },
+    {
+      "epoch": 4.957304408031387,
+      "grad_norm": 4.5,
+      "learning_rate": 6.424282855236825e-05,
+      "loss": 1.6206,
+      "step": 5370
+    },
+    {
+      "epoch": 4.958227555965843,
+      "grad_norm": 3.8671875,
+      "learning_rate": 6.42361574382922e-05,
+      "loss": 1.6682,
+      "step": 5371
+    },
+    {
+      "epoch": 4.9591507039003,
+      "grad_norm": 3.900390625,
+      "learning_rate": 6.422948632421615e-05,
+      "loss": 1.3838,
+      "step": 5372
+    },
+    {
+      "epoch": 4.960073851834757,
+      "grad_norm": 3.544921875,
+      "learning_rate": 6.42228152101401e-05,
+      "loss": 1.5007,
+      "step": 5373
+    },
+    {
+      "epoch": 4.960996999769213,
+      "grad_norm": 4.15234375,
+      "learning_rate": 6.421614409606404e-05,
+      "loss": 1.5217,
+      "step": 5374
+    },
+    {
+      "epoch": 4.96192014770367,
+      "grad_norm": 3.81640625,
+      "learning_rate": 6.420947298198799e-05,
+      "loss": 1.54,
+      "step": 5375
+    },
+    {
+      "epoch": 4.962843295638126,
+      "grad_norm": 3.216796875,
+      "learning_rate": 6.420280186791195e-05,
+      "loss": 1.5686,
+      "step": 5376
+    },
+    {
+      "epoch": 4.963766443572583,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.419613075383589e-05,
+      "loss": 1.1442,
+      "step": 5377
+    },
+    {
+      "epoch": 4.964689591507039,
+      "grad_norm": 3.81640625,
+      "learning_rate": 6.418945963975984e-05,
+      "loss": 1.4038,
+      "step": 5378
+    },
+    {
+      "epoch": 4.965612739441496,
+      "grad_norm": 3.146484375,
+      "learning_rate": 6.418278852568378e-05,
+      "loss": 1.5466,
+      "step": 5379
+    },
+    {
+      "epoch": 4.966535887375952,
+      "grad_norm": 3.5078125,
+      "learning_rate": 6.417611741160774e-05,
+      "loss": 1.604,
+      "step": 5380
+    },
+    {
+      "epoch": 4.967459035310409,
+      "grad_norm": 4.00390625,
+      "learning_rate": 6.416944629753169e-05,
+      "loss": 1.1152,
+      "step": 5381
+    },
+    {
+      "epoch": 4.968382183244865,
+      "grad_norm": 3.662109375,
+      "learning_rate": 6.416277518345563e-05,
+      "loss": 1.2864,
+      "step": 5382
+    },
+    {
+      "epoch": 4.969305331179322,
+      "grad_norm": 3.77734375,
+      "learning_rate": 6.415610406937958e-05,
+      "loss": 1.574,
+      "step": 5383
+    },
+    {
+      "epoch": 4.970228479113778,
+      "grad_norm": 4.22265625,
+      "learning_rate": 6.414943295530354e-05,
+      "loss": 1.408,
+      "step": 5384
+    },
+    {
+      "epoch": 4.9711516270482345,
+      "grad_norm": 3.552734375,
+      "learning_rate": 6.414276184122748e-05,
+      "loss": 1.0748,
+      "step": 5385
+    },
+    {
+      "epoch": 4.972074774982691,
+      "grad_norm": 3.494140625,
+      "learning_rate": 6.413609072715143e-05,
+      "loss": 1.2804,
+      "step": 5386
+    },
+    {
+      "epoch": 4.9729979229171475,
+      "grad_norm": 3.771484375,
+      "learning_rate": 6.412941961307539e-05,
+      "loss": 1.6704,
+      "step": 5387
+    },
+    {
+      "epoch": 4.973921070851604,
+      "grad_norm": 3.837890625,
+      "learning_rate": 6.412274849899935e-05,
+      "loss": 1.5627,
+      "step": 5388
+    },
+    {
+      "epoch": 4.97484421878606,
+      "grad_norm": 4.37890625,
+      "learning_rate": 6.411607738492329e-05,
+      "loss": 1.4509,
+      "step": 5389
+    },
+    {
+      "epoch": 4.975767366720517,
+      "grad_norm": 3.66015625,
+      "learning_rate": 6.410940627084724e-05,
+      "loss": 1.297,
+      "step": 5390
+    },
+    {
+      "epoch": 4.976690514654973,
+      "grad_norm": 3.72265625,
+      "learning_rate": 6.410273515677118e-05,
+      "loss": 1.6013,
+      "step": 5391
+    },
+    {
+      "epoch": 4.97761366258943,
+      "grad_norm": 3.876953125,
+      "learning_rate": 6.409606404269514e-05,
+      "loss": 1.3726,
+      "step": 5392
+    },
+    {
+      "epoch": 4.978536810523886,
+      "grad_norm": 3.8359375,
+      "learning_rate": 6.408939292861909e-05,
+      "loss": 1.5962,
+      "step": 5393
+    },
+    {
+      "epoch": 4.979459958458343,
+      "grad_norm": 4.6171875,
+      "learning_rate": 6.408272181454303e-05,
+      "loss": 1.4822,
+      "step": 5394
+    },
+    {
+      "epoch": 4.980383106392799,
+      "grad_norm": 4.55078125,
+      "learning_rate": 6.407605070046698e-05,
+      "loss": 1.561,
+      "step": 5395
+    },
+    {
+      "epoch": 4.981306254327256,
+      "grad_norm": 3.740234375,
+      "learning_rate": 6.406937958639094e-05,
+      "loss": 1.6318,
+      "step": 5396
+    },
+    {
+      "epoch": 4.982229402261712,
+      "grad_norm": 3.455078125,
+      "learning_rate": 6.406270847231488e-05,
+      "loss": 1.4058,
+      "step": 5397
+    },
+    {
+      "epoch": 4.983152550196169,
+      "grad_norm": 3.5078125,
+      "learning_rate": 6.405603735823883e-05,
+      "loss": 1.1974,
+      "step": 5398
+    },
+    {
+      "epoch": 4.984075698130625,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.404936624416277e-05,
+      "loss": 1.4558,
+      "step": 5399
+    },
+    {
+      "epoch": 4.984998846065082,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.404269513008673e-05,
+      "loss": 1.5984,
+      "step": 5400
+    },
+    {
+      "epoch": 4.985921993999538,
+      "grad_norm": 3.41015625,
+      "learning_rate": 6.403602401601068e-05,
+      "loss": 1.4214,
+      "step": 5401
+    },
+    {
+      "epoch": 4.986845141933995,
+      "grad_norm": 4.64453125,
+      "learning_rate": 6.402935290193462e-05,
+      "loss": 1.0731,
+      "step": 5402
+    },
+    {
+      "epoch": 4.987768289868452,
+      "grad_norm": 4.25,
+      "learning_rate": 6.402268178785857e-05,
+      "loss": 1.4783,
+      "step": 5403
+    },
+    {
+      "epoch": 4.988691437802908,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.401601067378253e-05,
+      "loss": 1.3619,
+      "step": 5404
+    },
+    {
+      "epoch": 4.989614585737364,
+      "grad_norm": 4.2265625,
+      "learning_rate": 6.400933955970647e-05,
+      "loss": 1.1833,
+      "step": 5405
+    },
+    {
+      "epoch": 4.990537733671821,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.400266844563042e-05,
+      "loss": 1.3486,
+      "step": 5406
+    },
+    {
+      "epoch": 4.991460881606278,
+      "grad_norm": 4.1953125,
+      "learning_rate": 6.399599733155436e-05,
+      "loss": 1.2294,
+      "step": 5407
+    },
+    {
+      "epoch": 4.992384029540734,
+      "grad_norm": 3.447265625,
+      "learning_rate": 6.398932621747832e-05,
+      "loss": 1.1851,
+      "step": 5408
+    },
+    {
+      "epoch": 4.993307177475191,
+      "grad_norm": 5.14453125,
+      "learning_rate": 6.398265510340227e-05,
+      "loss": 1.1038,
+      "step": 5409
+    },
+    {
+      "epoch": 4.994230325409647,
+      "grad_norm": 3.712890625,
+      "learning_rate": 6.397598398932621e-05,
+      "loss": 1.3286,
+      "step": 5410
+    },
+    {
+      "epoch": 4.995153473344104,
+      "grad_norm": 3.615234375,
+      "learning_rate": 6.396931287525016e-05,
+      "loss": 1.226,
+      "step": 5411
+    },
+    {
+      "epoch": 4.99607662127856,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.396264176117412e-05,
+      "loss": 1.0714,
+      "step": 5412
+    },
+    {
+      "epoch": 4.996999769213017,
+      "grad_norm": 4.29296875,
+      "learning_rate": 6.395597064709808e-05,
+      "loss": 1.5178,
+      "step": 5413
+    },
+    {
+      "epoch": 4.997922917147473,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.394929953302202e-05,
+      "loss": 1.6089,
+      "step": 5414
+    },
+    {
+      "epoch": 4.9988460650819295,
+      "grad_norm": 4.1484375,
+      "learning_rate": 6.394262841894597e-05,
+      "loss": 1.5391,
+      "step": 5415
+    },
+    {
+      "epoch": 4.9997692130163855,
+      "grad_norm": 5.00390625,
+      "learning_rate": 6.393595730486991e-05,
+      "loss": 1.4604,
+      "step": 5416
+    },
+    {
+      "epoch": 5.0006923609508425,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.392928619079387e-05,
+      "loss": 1.072,
+      "step": 5417
+    },
+    {
+      "epoch": 5.0016155088852985,
+      "grad_norm": 4.171875,
+      "learning_rate": 6.392261507671782e-05,
+      "loss": 1.645,
+      "step": 5418
+    },
+    {
+      "epoch": 5.002538656819755,
+      "grad_norm": 3.8203125,
+      "learning_rate": 6.391594396264176e-05,
+      "loss": 1.4197,
+      "step": 5419
+    },
+    {
+      "epoch": 5.003461804754211,
+      "grad_norm": 3.748046875,
+      "learning_rate": 6.390927284856571e-05,
+      "loss": 1.4749,
+      "step": 5420
+    },
+    {
+      "epoch": 5.004384952688668,
+      "grad_norm": 3.88671875,
+      "learning_rate": 6.390260173448967e-05,
+      "loss": 0.9814,
+      "step": 5421
+    },
+    {
+      "epoch": 5.005308100623125,
+      "grad_norm": 3.857421875,
+      "learning_rate": 6.389593062041361e-05,
+      "loss": 1.4465,
+      "step": 5422
+    },
+    {
+      "epoch": 5.006231248557581,
+      "grad_norm": 3.728515625,
+      "learning_rate": 6.388925950633756e-05,
+      "loss": 1.2756,
+      "step": 5423
+    },
+    {
+      "epoch": 5.007154396492038,
+      "grad_norm": 3.748046875,
+      "learning_rate": 6.38825883922615e-05,
+      "loss": 1.3843,
+      "step": 5424
+    },
+    {
+      "epoch": 5.008077544426494,
+      "grad_norm": 4.390625,
+      "learning_rate": 6.387591727818546e-05,
+      "loss": 1.2028,
+      "step": 5425
+    },
+    {
+      "epoch": 5.009000692360951,
+      "grad_norm": 4.33203125,
+      "learning_rate": 6.386924616410941e-05,
+      "loss": 1.5972,
+      "step": 5426
+    },
+    {
+      "epoch": 5.009923840295407,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.386257505003335e-05,
+      "loss": 1.2839,
+      "step": 5427
+    },
+    {
+      "epoch": 5.010846988229864,
+      "grad_norm": 4.12109375,
+      "learning_rate": 6.38559039359573e-05,
+      "loss": 1.2722,
+      "step": 5428
+    },
+    {
+      "epoch": 5.01177013616432,
+      "grad_norm": 4.2109375,
+      "learning_rate": 6.384923282188126e-05,
+      "loss": 1.5681,
+      "step": 5429
+    },
+    {
+      "epoch": 5.012693284098777,
+      "grad_norm": 4.33984375,
+      "learning_rate": 6.38425617078052e-05,
+      "loss": 1.3977,
+      "step": 5430
+    },
+    {
+      "epoch": 5.013616432033233,
+      "grad_norm": 3.998046875,
+      "learning_rate": 6.383589059372915e-05,
+      "loss": 1.344,
+      "step": 5431
+    },
+    {
+      "epoch": 5.01453957996769,
+      "grad_norm": 3.29296875,
+      "learning_rate": 6.382921947965311e-05,
+      "loss": 1.3845,
+      "step": 5432
+    },
+    {
+      "epoch": 5.015462727902146,
+      "grad_norm": 2.837890625,
+      "learning_rate": 6.382254836557705e-05,
+      "loss": 1.0085,
+      "step": 5433
+    },
+    {
+      "epoch": 5.016385875836603,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.3815877251501e-05,
+      "loss": 1.2728,
+      "step": 5434
+    },
+    {
+      "epoch": 5.017309023771059,
+      "grad_norm": 3.841796875,
+      "learning_rate": 6.380920613742494e-05,
+      "loss": 1.3153,
+      "step": 5435
+    },
+    {
+      "epoch": 5.018232171705516,
+      "grad_norm": 4.0390625,
+      "learning_rate": 6.38025350233489e-05,
+      "loss": 1.3608,
+      "step": 5436
+    },
+    {
+      "epoch": 5.019155319639973,
+      "grad_norm": 3.970703125,
+      "learning_rate": 6.379586390927286e-05,
+      "loss": 1.2827,
+      "step": 5437
+    },
+    {
+      "epoch": 5.020078467574429,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.378919279519681e-05,
+      "loss": 1.4539,
+      "step": 5438
+    },
+    {
+      "epoch": 5.021001615508886,
+      "grad_norm": 3.84765625,
+      "learning_rate": 6.378252168112075e-05,
+      "loss": 0.9388,
+      "step": 5439
+    },
+    {
+      "epoch": 5.021924763443342,
+      "grad_norm": 3.451171875,
+      "learning_rate": 6.37758505670447e-05,
+      "loss": 1.0022,
+      "step": 5440
+    },
+    {
+      "epoch": 5.022847911377799,
+      "grad_norm": 3.451171875,
+      "learning_rate": 6.376917945296866e-05,
+      "loss": 1.1971,
+      "step": 5441
+    },
+    {
+      "epoch": 5.023771059312255,
+      "grad_norm": 3.93359375,
+      "learning_rate": 6.37625083388926e-05,
+      "loss": 1.3743,
+      "step": 5442
+    },
+    {
+      "epoch": 5.024694207246712,
+      "grad_norm": 3.625,
+      "learning_rate": 6.375583722481655e-05,
+      "loss": 1.0073,
+      "step": 5443
+    },
+    {
+      "epoch": 5.025617355181168,
+      "grad_norm": 4.77734375,
+      "learning_rate": 6.37491661107405e-05,
+      "loss": 1.6519,
+      "step": 5444
+    },
+    {
+      "epoch": 5.0265405031156245,
+      "grad_norm": 3.791015625,
+      "learning_rate": 6.374249499666445e-05,
+      "loss": 1.5474,
+      "step": 5445
+    },
+    {
+      "epoch": 5.0274636510500805,
+      "grad_norm": 3.99609375,
+      "learning_rate": 6.37358238825884e-05,
+      "loss": 1.4255,
+      "step": 5446
+    },
+    {
+      "epoch": 5.028386798984537,
+      "grad_norm": 3.86328125,
+      "learning_rate": 6.372915276851234e-05,
+      "loss": 1.2498,
+      "step": 5447
+    },
+    {
+      "epoch": 5.0293099469189935,
+      "grad_norm": 4.46875,
+      "learning_rate": 6.372248165443629e-05,
+      "loss": 1.3427,
+      "step": 5448
+    },
+    {
+      "epoch": 5.03023309485345,
+      "grad_norm": 3.525390625,
+      "learning_rate": 6.371581054036025e-05,
+      "loss": 1.4209,
+      "step": 5449
+    },
+    {
+      "epoch": 5.031156242787906,
+      "grad_norm": 4.35546875,
+      "learning_rate": 6.37091394262842e-05,
+      "loss": 1.3945,
+      "step": 5450
+    },
+    {
+      "epoch": 5.032079390722363,
+      "grad_norm": 4.0703125,
+      "learning_rate": 6.370246831220814e-05,
+      "loss": 1.217,
+      "step": 5451
+    },
+    {
+      "epoch": 5.033002538656819,
+      "grad_norm": 3.794921875,
+      "learning_rate": 6.369579719813208e-05,
+      "loss": 1.1478,
+      "step": 5452
+    },
+    {
+      "epoch": 5.033925686591276,
+      "grad_norm": 3.955078125,
+      "learning_rate": 6.368912608405604e-05,
+      "loss": 1.1637,
+      "step": 5453
+    },
+    {
+      "epoch": 5.034848834525733,
+      "grad_norm": 3.75390625,
+      "learning_rate": 6.368245496997999e-05,
+      "loss": 1.5928,
+      "step": 5454
+    },
+    {
+      "epoch": 5.035771982460189,
+      "grad_norm": 3.85546875,
+      "learning_rate": 6.367578385590393e-05,
+      "loss": 1.3052,
+      "step": 5455
+    },
+    {
+      "epoch": 5.036695130394646,
+      "grad_norm": 4.6328125,
+      "learning_rate": 6.366911274182788e-05,
+      "loss": 0.9929,
+      "step": 5456
+    },
+    {
+      "epoch": 5.037618278329102,
+      "grad_norm": 4.28125,
+      "learning_rate": 6.366244162775184e-05,
+      "loss": 1.323,
+      "step": 5457
+    },
+    {
+      "epoch": 5.038541426263559,
+      "grad_norm": 3.16796875,
+      "learning_rate": 6.365577051367578e-05,
+      "loss": 1.2999,
+      "step": 5458
+    },
+    {
+      "epoch": 5.039464574198015,
+      "grad_norm": 2.4921875,
+      "learning_rate": 6.364909939959973e-05,
+      "loss": 1.078,
+      "step": 5459
+    },
+    {
+      "epoch": 5.040387722132472,
+      "grad_norm": 4.00390625,
+      "learning_rate": 6.364242828552368e-05,
+      "loss": 1.3647,
+      "step": 5460
+    },
+    {
+      "epoch": 5.041310870066928,
+      "grad_norm": 4.02734375,
+      "learning_rate": 6.363575717144763e-05,
+      "loss": 1.5891,
+      "step": 5461
+    },
+    {
+      "epoch": 5.042234018001385,
+      "grad_norm": 4.0234375,
+      "learning_rate": 6.362908605737158e-05,
+      "loss": 1.4355,
+      "step": 5462
+    },
+    {
+      "epoch": 5.043157165935841,
+      "grad_norm": 3.400390625,
+      "learning_rate": 6.362241494329554e-05,
+      "loss": 1.3169,
+      "step": 5463
+    },
+    {
+      "epoch": 5.044080313870298,
+      "grad_norm": 3.505859375,
+      "learning_rate": 6.361574382921948e-05,
+      "loss": 1.3372,
+      "step": 5464
+    },
+    {
+      "epoch": 5.045003461804754,
+      "grad_norm": 3.61328125,
+      "learning_rate": 6.360907271514343e-05,
+      "loss": 1.5601,
+      "step": 5465
+    },
+    {
+      "epoch": 5.045926609739211,
+      "grad_norm": 3.783203125,
+      "learning_rate": 6.360240160106739e-05,
+      "loss": 1.2903,
+      "step": 5466
+    },
+    {
+      "epoch": 5.046849757673667,
+      "grad_norm": 3.244140625,
+      "learning_rate": 6.359573048699133e-05,
+      "loss": 0.9336,
+      "step": 5467
+    },
+    {
+      "epoch": 5.047772905608124,
+      "grad_norm": 3.3984375,
+      "learning_rate": 6.358905937291528e-05,
+      "loss": 1.1055,
+      "step": 5468
+    },
+    {
+      "epoch": 5.04869605354258,
+      "grad_norm": 3.501953125,
+      "learning_rate": 6.358238825883923e-05,
+      "loss": 1.377,
+      "step": 5469
+    },
+    {
+      "epoch": 5.049619201477037,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.357571714476318e-05,
+      "loss": 1.1732,
+      "step": 5470
+    },
+    {
+      "epoch": 5.050542349411494,
+      "grad_norm": 4.0,
+      "learning_rate": 6.356904603068713e-05,
+      "loss": 1.3726,
+      "step": 5471
+    },
+    {
+      "epoch": 5.05146549734595,
+      "grad_norm": 3.970703125,
+      "learning_rate": 6.356237491661107e-05,
+      "loss": 1.4919,
+      "step": 5472
+    },
+    {
+      "epoch": 5.0523886452804065,
+      "grad_norm": 3.560546875,
+      "learning_rate": 6.355570380253503e-05,
+      "loss": 1.4512,
+      "step": 5473
+    },
+    {
+      "epoch": 5.053311793214863,
+      "grad_norm": 3.705078125,
+      "learning_rate": 6.354903268845898e-05,
+      "loss": 0.8586,
+      "step": 5474
+    },
+    {
+      "epoch": 5.0542349411493195,
+      "grad_norm": 4.33984375,
+      "learning_rate": 6.354236157438292e-05,
+      "loss": 1.7166,
+      "step": 5475
+    },
+    {
+      "epoch": 5.0551580890837755,
+      "grad_norm": 4.578125,
+      "learning_rate": 6.353569046030687e-05,
+      "loss": 1.3032,
+      "step": 5476
+    },
+    {
+      "epoch": 5.056081237018232,
+      "grad_norm": 3.9140625,
+      "learning_rate": 6.352901934623083e-05,
+      "loss": 1.4094,
+      "step": 5477
+    },
+    {
+      "epoch": 5.0570043849526884,
+      "grad_norm": 3.3359375,
+      "learning_rate": 6.352234823215477e-05,
+      "loss": 1.0889,
+      "step": 5478
+    },
+    {
+      "epoch": 5.057927532887145,
+      "grad_norm": 3.435546875,
+      "learning_rate": 6.351567711807872e-05,
+      "loss": 1.5168,
+      "step": 5479
+    },
+    {
+      "epoch": 5.058850680821601,
+      "grad_norm": 3.287109375,
+      "learning_rate": 6.350900600400267e-05,
+      "loss": 1.2717,
+      "step": 5480
+    },
+    {
+      "epoch": 5.059773828756058,
+      "grad_norm": 3.623046875,
+      "learning_rate": 6.350233488992662e-05,
+      "loss": 1.5244,
+      "step": 5481
+    },
+    {
+      "epoch": 5.060696976690514,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.349566377585057e-05,
+      "loss": 1.575,
+      "step": 5482
+    },
+    {
+      "epoch": 5.061620124624971,
+      "grad_norm": 3.224609375,
+      "learning_rate": 6.348899266177452e-05,
+      "loss": 1.5193,
+      "step": 5483
+    },
+    {
+      "epoch": 5.062543272559427,
+      "grad_norm": 3.248046875,
+      "learning_rate": 6.348232154769846e-05,
+      "loss": 1.3479,
+      "step": 5484
+    },
+    {
+      "epoch": 5.063466420493884,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.347565043362242e-05,
+      "loss": 1.3262,
+      "step": 5485
+    },
+    {
+      "epoch": 5.06438956842834,
+      "grad_norm": 3.1328125,
+      "learning_rate": 6.346897931954637e-05,
+      "loss": 1.4919,
+      "step": 5486
+    },
+    {
+      "epoch": 5.065312716362797,
+      "grad_norm": 3.017578125,
+      "learning_rate": 6.346230820547031e-05,
+      "loss": 1.5398,
+      "step": 5487
+    },
+    {
+      "epoch": 5.066235864297254,
+      "grad_norm": 4.125,
+      "learning_rate": 6.345563709139427e-05,
+      "loss": 1.1316,
+      "step": 5488
+    },
+    {
+      "epoch": 5.06715901223171,
+      "grad_norm": 3.275390625,
+      "learning_rate": 6.344896597731821e-05,
+      "loss": 1.4343,
+      "step": 5489
+    },
+    {
+      "epoch": 5.068082160166167,
+      "grad_norm": 3.998046875,
+      "learning_rate": 6.344229486324217e-05,
+      "loss": 1.3281,
+      "step": 5490
+    },
+    {
+      "epoch": 5.069005308100623,
+      "grad_norm": 4.16015625,
+      "learning_rate": 6.343562374916612e-05,
+      "loss": 1.2427,
+      "step": 5491
+    },
+    {
+      "epoch": 5.06992845603508,
+      "grad_norm": 4.21484375,
+      "learning_rate": 6.342895263509006e-05,
+      "loss": 1.2917,
+      "step": 5492
+    },
+    {
+      "epoch": 5.070851603969536,
+      "grad_norm": 4.32421875,
+      "learning_rate": 6.342228152101401e-05,
+      "loss": 1.6611,
+      "step": 5493
+    },
+    {
+      "epoch": 5.071774751903993,
+      "grad_norm": 3.509765625,
+      "learning_rate": 6.341561040693797e-05,
+      "loss": 1.4636,
+      "step": 5494
+    },
+    {
+      "epoch": 5.072697899838449,
+      "grad_norm": 3.73046875,
+      "learning_rate": 6.340893929286191e-05,
+      "loss": 1.2222,
+      "step": 5495
+    },
+    {
+      "epoch": 5.073621047772906,
+      "grad_norm": 3.2578125,
+      "learning_rate": 6.340226817878586e-05,
+      "loss": 1.3691,
+      "step": 5496
+    },
+    {
+      "epoch": 5.074544195707362,
+      "grad_norm": 3.69921875,
+      "learning_rate": 6.33955970647098e-05,
+      "loss": 1.3149,
+      "step": 5497
+    },
+    {
+      "epoch": 5.075467343641819,
+      "grad_norm": 4.296875,
+      "learning_rate": 6.338892595063376e-05,
+      "loss": 1.4937,
+      "step": 5498
+    },
+    {
+      "epoch": 5.076390491576275,
+      "grad_norm": 3.8359375,
+      "learning_rate": 6.338225483655771e-05,
+      "loss": 1.1285,
+      "step": 5499
+    },
+    {
+      "epoch": 5.077313639510732,
+      "grad_norm": 4.21875,
+      "learning_rate": 6.337558372248166e-05,
+      "loss": 1.1895,
+      "step": 5500
+    },
+    {
+      "epoch": 5.078236787445188,
+      "grad_norm": 3.986328125,
+      "learning_rate": 6.33689126084056e-05,
+      "loss": 1.2477,
+      "step": 5501
+    },
+    {
+      "epoch": 5.079159935379645,
+      "grad_norm": 3.89453125,
+      "learning_rate": 6.336224149432956e-05,
+      "loss": 1.1555,
+      "step": 5502
+    },
+    {
+      "epoch": 5.080083083314101,
+      "grad_norm": 3.55859375,
+      "learning_rate": 6.33555703802535e-05,
+      "loss": 1.4719,
+      "step": 5503
+    },
+    {
+      "epoch": 5.0810062312485575,
+      "grad_norm": 3.857421875,
+      "learning_rate": 6.334889926617745e-05,
+      "loss": 1.6641,
+      "step": 5504
+    },
+    {
+      "epoch": 5.0819293791830145,
+      "grad_norm": 3.4453125,
+      "learning_rate": 6.33422281521014e-05,
+      "loss": 1.0893,
+      "step": 5505
+    },
+    {
+      "epoch": 5.0828525271174705,
+      "grad_norm": 4.015625,
+      "learning_rate": 6.333555703802536e-05,
+      "loss": 1.5474,
+      "step": 5506
+    },
+    {
+      "epoch": 5.083775675051927,
+      "grad_norm": 4.00390625,
+      "learning_rate": 6.33288859239493e-05,
+      "loss": 1.2183,
+      "step": 5507
+    },
+    {
+      "epoch": 5.084698822986383,
+      "grad_norm": 4.5703125,
+      "learning_rate": 6.332221480987325e-05,
+      "loss": 1.3342,
+      "step": 5508
+    },
+    {
+      "epoch": 5.08562197092084,
+      "grad_norm": 4.5546875,
+      "learning_rate": 6.331554369579719e-05,
+      "loss": 1.5918,
+      "step": 5509
+    },
+    {
+      "epoch": 5.086545118855296,
+      "grad_norm": 3.93359375,
+      "learning_rate": 6.330887258172115e-05,
+      "loss": 1.0973,
+      "step": 5510
+    },
+    {
+      "epoch": 5.087468266789753,
+      "grad_norm": 3.486328125,
+      "learning_rate": 6.33022014676451e-05,
+      "loss": 1.2322,
+      "step": 5511
+    },
+    {
+      "epoch": 5.088391414724209,
+      "grad_norm": 3.86328125,
+      "learning_rate": 6.329553035356904e-05,
+      "loss": 1.0482,
+      "step": 5512
+    },
+    {
+      "epoch": 5.089314562658666,
+      "grad_norm": 3.681640625,
+      "learning_rate": 6.3288859239493e-05,
+      "loss": 1.3782,
+      "step": 5513
+    },
+    {
+      "epoch": 5.090237710593122,
+      "grad_norm": 4.0703125,
+      "learning_rate": 6.328218812541696e-05,
+      "loss": 1.3582,
+      "step": 5514
+    },
+    {
+      "epoch": 5.091160858527579,
+      "grad_norm": 4.546875,
+      "learning_rate": 6.32755170113409e-05,
+      "loss": 1.1136,
+      "step": 5515
+    },
+    {
+      "epoch": 5.092084006462035,
+      "grad_norm": 5.15625,
+      "learning_rate": 6.326884589726485e-05,
+      "loss": 1.4167,
+      "step": 5516
+    },
+    {
+      "epoch": 5.093007154396492,
+      "grad_norm": 3.00390625,
+      "learning_rate": 6.32621747831888e-05,
+      "loss": 1.0996,
+      "step": 5517
+    },
+    {
+      "epoch": 5.093930302330948,
+      "grad_norm": 3.376953125,
+      "learning_rate": 6.325550366911275e-05,
+      "loss": 1.1643,
+      "step": 5518
+    },
+    {
+      "epoch": 5.094853450265405,
+      "grad_norm": 3.3828125,
+      "learning_rate": 6.32488325550367e-05,
+      "loss": 1.4792,
+      "step": 5519
+    },
+    {
+      "epoch": 5.095776598199862,
+      "grad_norm": 3.880859375,
+      "learning_rate": 6.324216144096065e-05,
+      "loss": 1.1216,
+      "step": 5520
+    },
+    {
+      "epoch": 5.096699746134318,
+      "grad_norm": 4.1953125,
+      "learning_rate": 6.323549032688459e-05,
+      "loss": 1.4524,
+      "step": 5521
+    },
+    {
+      "epoch": 5.097622894068775,
+      "grad_norm": 3.625,
+      "learning_rate": 6.322881921280855e-05,
+      "loss": 1.5398,
+      "step": 5522
+    },
+    {
+      "epoch": 5.098546042003231,
+      "grad_norm": 3.74609375,
+      "learning_rate": 6.32221480987325e-05,
+      "loss": 1.4155,
+      "step": 5523
+    },
+    {
+      "epoch": 5.099469189937688,
+      "grad_norm": 4.0234375,
+      "learning_rate": 6.321547698465644e-05,
+      "loss": 1.0226,
+      "step": 5524
+    },
+    {
+      "epoch": 5.100392337872144,
+      "grad_norm": 4.375,
+      "learning_rate": 6.320880587058039e-05,
+      "loss": 1.2024,
+      "step": 5525
+    },
+    {
+      "epoch": 5.101315485806601,
+      "grad_norm": 3.552734375,
+      "learning_rate": 6.320213475650434e-05,
+      "loss": 1.3862,
+      "step": 5526
+    },
+    {
+      "epoch": 5.102238633741057,
+      "grad_norm": 3.62890625,
+      "learning_rate": 6.319546364242829e-05,
+      "loss": 1.3416,
+      "step": 5527
+    },
+    {
+      "epoch": 5.103161781675514,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.318879252835224e-05,
+      "loss": 1.396,
+      "step": 5528
+    },
+    {
+      "epoch": 5.10408492960997,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.318212141427618e-05,
+      "loss": 1.3657,
+      "step": 5529
+    },
+    {
+      "epoch": 5.105008077544427,
+      "grad_norm": 3.99609375,
+      "learning_rate": 6.317545030020014e-05,
+      "loss": 1.5544,
+      "step": 5530
+    },
+    {
+      "epoch": 5.105931225478883,
+      "grad_norm": 4.25,
+      "learning_rate": 6.316877918612409e-05,
+      "loss": 1.2786,
+      "step": 5531
+    },
+    {
+      "epoch": 5.10685437341334,
+      "grad_norm": 3.4140625,
+      "learning_rate": 6.316210807204803e-05,
+      "loss": 1.2815,
+      "step": 5532
+    },
+    {
+      "epoch": 5.107777521347796,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.315543695797198e-05,
+      "loss": 1.665,
+      "step": 5533
+    },
+    {
+      "epoch": 5.1087006692822525,
+      "grad_norm": 3.908203125,
+      "learning_rate": 6.314876584389594e-05,
+      "loss": 1.22,
+      "step": 5534
+    },
+    {
+      "epoch": 5.1096238172167086,
+      "grad_norm": 3.404296875,
+      "learning_rate": 6.314209472981988e-05,
+      "loss": 1.327,
+      "step": 5535
+    },
+    {
+      "epoch": 5.1105469651511655,
+      "grad_norm": 3.537109375,
+      "learning_rate": 6.313542361574383e-05,
+      "loss": 1.5017,
+      "step": 5536
+    },
+    {
+      "epoch": 5.111470113085622,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.312875250166777e-05,
+      "loss": 1.4133,
+      "step": 5537
+    },
+    {
+      "epoch": 5.112393261020078,
+      "grad_norm": 4.0546875,
+      "learning_rate": 6.312208138759173e-05,
+      "loss": 1.2412,
+      "step": 5538
+    },
+    {
+      "epoch": 5.113316408954535,
+      "grad_norm": 4.578125,
+      "learning_rate": 6.311541027351569e-05,
+      "loss": 1.6409,
+      "step": 5539
+    },
+    {
+      "epoch": 5.114239556888991,
+      "grad_norm": 3.517578125,
+      "learning_rate": 6.310873915943964e-05,
+      "loss": 1.325,
+      "step": 5540
+    },
+    {
+      "epoch": 5.115162704823448,
+      "grad_norm": 4.7890625,
+      "learning_rate": 6.310206804536358e-05,
+      "loss": 1.5973,
+      "step": 5541
+    },
+    {
+      "epoch": 5.116085852757904,
+      "grad_norm": 3.552734375,
+      "learning_rate": 6.309539693128753e-05,
+      "loss": 1.4231,
+      "step": 5542
+    },
+    {
+      "epoch": 5.117009000692361,
+      "grad_norm": 3.732421875,
+      "learning_rate": 6.308872581721149e-05,
+      "loss": 1.5427,
+      "step": 5543
+    },
+    {
+      "epoch": 5.117932148626817,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.308205470313543e-05,
+      "loss": 1.4324,
+      "step": 5544
+    },
+    {
+      "epoch": 5.118855296561274,
+      "grad_norm": 4.671875,
+      "learning_rate": 6.307538358905938e-05,
+      "loss": 1.3158,
+      "step": 5545
+    },
+    {
+      "epoch": 5.11977844449573,
+      "grad_norm": 3.474609375,
+      "learning_rate": 6.306871247498332e-05,
+      "loss": 1.01,
+      "step": 5546
+    },
+    {
+      "epoch": 5.120701592430187,
+      "grad_norm": 4.359375,
+      "learning_rate": 6.306204136090728e-05,
+      "loss": 1.1139,
+      "step": 5547
+    },
+    {
+      "epoch": 5.121624740364643,
+      "grad_norm": 3.810546875,
+      "learning_rate": 6.305537024683123e-05,
+      "loss": 1.3821,
+      "step": 5548
+    },
+    {
+      "epoch": 5.1225478882991,
+      "grad_norm": 4.234375,
+      "learning_rate": 6.304869913275517e-05,
+      "loss": 1.4351,
+      "step": 5549
+    },
+    {
+      "epoch": 5.123471036233556,
+      "grad_norm": 3.41796875,
+      "learning_rate": 6.304202801867912e-05,
+      "loss": 1.2723,
+      "step": 5550
+    },
+    {
+      "epoch": 5.124394184168013,
+      "grad_norm": 3.5390625,
+      "learning_rate": 6.303535690460308e-05,
+      "loss": 1.2162,
+      "step": 5551
+    },
+    {
+      "epoch": 5.125317332102469,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.302868579052702e-05,
+      "loss": 1.5967,
+      "step": 5552
+    },
+    {
+      "epoch": 5.126240480036926,
+      "grad_norm": 4.5546875,
+      "learning_rate": 6.302201467645097e-05,
+      "loss": 1.4924,
+      "step": 5553
+    },
+    {
+      "epoch": 5.127163627971383,
+      "grad_norm": 4.3125,
+      "learning_rate": 6.301534356237491e-05,
+      "loss": 1.4846,
+      "step": 5554
+    },
+    {
+      "epoch": 5.128086775905839,
+      "grad_norm": 4.07421875,
+      "learning_rate": 6.300867244829887e-05,
+      "loss": 1.3655,
+      "step": 5555
+    },
+    {
+      "epoch": 5.129009923840296,
+      "grad_norm": 4.2734375,
+      "learning_rate": 6.300200133422282e-05,
+      "loss": 1.4612,
+      "step": 5556
+    },
+    {
+      "epoch": 5.129933071774752,
+      "grad_norm": 3.69140625,
+      "learning_rate": 6.299533022014676e-05,
+      "loss": 1.446,
+      "step": 5557
+    },
+    {
+      "epoch": 5.130856219709209,
+      "grad_norm": 3.263671875,
+      "learning_rate": 6.298865910607071e-05,
+      "loss": 1.6687,
+      "step": 5558
+    },
+    {
+      "epoch": 5.131779367643665,
+      "grad_norm": 4.58203125,
+      "learning_rate": 6.298198799199467e-05,
+      "loss": 1.5144,
+      "step": 5559
+    },
+    {
+      "epoch": 5.132702515578122,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.297531687791861e-05,
+      "loss": 1.0474,
+      "step": 5560
+    },
+    {
+      "epoch": 5.133625663512578,
+      "grad_norm": 4.0234375,
+      "learning_rate": 6.296864576384256e-05,
+      "loss": 1.4907,
+      "step": 5561
+    },
+    {
+      "epoch": 5.134548811447035,
+      "grad_norm": 3.642578125,
+      "learning_rate": 6.29619746497665e-05,
+      "loss": 1.4097,
+      "step": 5562
+    },
+    {
+      "epoch": 5.135471959381491,
+      "grad_norm": 3.494140625,
+      "learning_rate": 6.295530353569047e-05,
+      "loss": 1.449,
+      "step": 5563
+    },
+    {
+      "epoch": 5.1363951073159475,
+      "grad_norm": 3.509765625,
+      "learning_rate": 6.294863242161442e-05,
+      "loss": 1.4407,
+      "step": 5564
+    },
+    {
+      "epoch": 5.1373182552504035,
+      "grad_norm": 3.3671875,
+      "learning_rate": 6.294196130753837e-05,
+      "loss": 1.4287,
+      "step": 5565
+    },
+    {
+      "epoch": 5.1382414031848604,
+      "grad_norm": 3.513671875,
+      "learning_rate": 6.293529019346231e-05,
+      "loss": 1.2234,
+      "step": 5566
+    },
+    {
+      "epoch": 5.1391645511193165,
+      "grad_norm": 3.595703125,
+      "learning_rate": 6.292861907938627e-05,
+      "loss": 1.5791,
+      "step": 5567
+    },
+    {
+      "epoch": 5.140087699053773,
+      "grad_norm": 3.705078125,
+      "learning_rate": 6.292194796531022e-05,
+      "loss": 1.4507,
+      "step": 5568
+    },
+    {
+      "epoch": 5.14101084698823,
+      "grad_norm": 4.828125,
+      "learning_rate": 6.291527685123416e-05,
+      "loss": 1.5635,
+      "step": 5569
+    },
+    {
+      "epoch": 5.141933994922686,
+      "grad_norm": 3.673828125,
+      "learning_rate": 6.29086057371581e-05,
+      "loss": 1.4448,
+      "step": 5570
+    },
+    {
+      "epoch": 5.142857142857143,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.290193462308207e-05,
+      "loss": 1.4666,
+      "step": 5571
+    },
+    {
+      "epoch": 5.143780290791599,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.289526350900601e-05,
+      "loss": 1.2119,
+      "step": 5572
+    },
+    {
+      "epoch": 5.144703438726056,
+      "grad_norm": 4.34765625,
+      "learning_rate": 6.288859239492996e-05,
+      "loss": 1.3486,
+      "step": 5573
+    },
+    {
+      "epoch": 5.145626586660512,
+      "grad_norm": 3.8203125,
+      "learning_rate": 6.28819212808539e-05,
+      "loss": 1.2235,
+      "step": 5574
+    },
+    {
+      "epoch": 5.146549734594969,
+      "grad_norm": 4.5546875,
+      "learning_rate": 6.287525016677786e-05,
+      "loss": 1.3872,
+      "step": 5575
+    },
+    {
+      "epoch": 5.147472882529425,
+      "grad_norm": 3.31640625,
+      "learning_rate": 6.28685790527018e-05,
+      "loss": 1.4978,
+      "step": 5576
+    },
+    {
+      "epoch": 5.148396030463882,
+      "grad_norm": 4.13671875,
+      "learning_rate": 6.286190793862575e-05,
+      "loss": 1.2001,
+      "step": 5577
+    },
+    {
+      "epoch": 5.149319178398338,
+      "grad_norm": 3.490234375,
+      "learning_rate": 6.28552368245497e-05,
+      "loss": 1.489,
+      "step": 5578
+    },
+    {
+      "epoch": 5.150242326332795,
+      "grad_norm": 4.26953125,
+      "learning_rate": 6.284856571047366e-05,
+      "loss": 1.5725,
+      "step": 5579
+    },
+    {
+      "epoch": 5.151165474267251,
+      "grad_norm": 4.48828125,
+      "learning_rate": 6.28418945963976e-05,
+      "loss": 1.2434,
+      "step": 5580
+    },
+    {
+      "epoch": 5.152088622201708,
+      "grad_norm": 3.91796875,
+      "learning_rate": 6.283522348232155e-05,
+      "loss": 1.3091,
+      "step": 5581
+    },
+    {
+      "epoch": 5.153011770136164,
+      "grad_norm": 4.46484375,
+      "learning_rate": 6.282855236824549e-05,
+      "loss": 1.2959,
+      "step": 5582
+    },
+    {
+      "epoch": 5.153934918070621,
+      "grad_norm": 4.1015625,
+      "learning_rate": 6.282188125416945e-05,
+      "loss": 1.2751,
+      "step": 5583
+    },
+    {
+      "epoch": 5.154858066005077,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.28152101400934e-05,
+      "loss": 1.3928,
+      "step": 5584
+    },
+    {
+      "epoch": 5.155781213939534,
+      "grad_norm": 3.89453125,
+      "learning_rate": 6.280853902601734e-05,
+      "loss": 1.4604,
+      "step": 5585
+    },
+    {
+      "epoch": 5.156704361873991,
+      "grad_norm": 3.83203125,
+      "learning_rate": 6.280186791194129e-05,
+      "loss": 1.3875,
+      "step": 5586
+    },
+    {
+      "epoch": 5.157627509808447,
+      "grad_norm": 4.421875,
+      "learning_rate": 6.279519679786525e-05,
+      "loss": 1.061,
+      "step": 5587
+    },
+    {
+      "epoch": 5.158550657742904,
+      "grad_norm": 4.49609375,
+      "learning_rate": 6.27885256837892e-05,
+      "loss": 1.5654,
+      "step": 5588
+    },
+    {
+      "epoch": 5.15947380567736,
+      "grad_norm": 3.470703125,
+      "learning_rate": 6.278185456971315e-05,
+      "loss": 1.4098,
+      "step": 5589
+    },
+    {
+      "epoch": 5.160396953611817,
+      "grad_norm": 4.6171875,
+      "learning_rate": 6.27751834556371e-05,
+      "loss": 1.2455,
+      "step": 5590
+    },
+    {
+      "epoch": 5.161320101546273,
+      "grad_norm": 2.865234375,
+      "learning_rate": 6.276851234156104e-05,
+      "loss": 0.7967,
+      "step": 5591
+    },
+    {
+      "epoch": 5.1622432494807295,
+      "grad_norm": 4.29296875,
+      "learning_rate": 6.2761841227485e-05,
+      "loss": 1.3284,
+      "step": 5592
+    },
+    {
+      "epoch": 5.163166397415186,
+      "grad_norm": 3.82421875,
+      "learning_rate": 6.275517011340895e-05,
+      "loss": 1.3945,
+      "step": 5593
+    },
+    {
+      "epoch": 5.1640895453496425,
+      "grad_norm": 4.1015625,
+      "learning_rate": 6.274849899933289e-05,
+      "loss": 1.3408,
+      "step": 5594
+    },
+    {
+      "epoch": 5.1650126932840985,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.274182788525684e-05,
+      "loss": 0.9797,
+      "step": 5595
+    },
+    {
+      "epoch": 5.165935841218555,
+      "grad_norm": 3.38671875,
+      "learning_rate": 6.27351567711808e-05,
+      "loss": 1.6418,
+      "step": 5596
+    },
+    {
+      "epoch": 5.1668589891530114,
+      "grad_norm": 3.94921875,
+      "learning_rate": 6.272848565710474e-05,
+      "loss": 1.4268,
+      "step": 5597
+    },
+    {
+      "epoch": 5.167782137087468,
+      "grad_norm": 3.607421875,
+      "learning_rate": 6.272181454302869e-05,
+      "loss": 1.4631,
+      "step": 5598
+    },
+    {
+      "epoch": 5.168705285021924,
+      "grad_norm": 3.97265625,
+      "learning_rate": 6.271514342895263e-05,
+      "loss": 1.3547,
+      "step": 5599
+    },
+    {
+      "epoch": 5.169628432956381,
+      "grad_norm": 3.001953125,
+      "learning_rate": 6.270847231487659e-05,
+      "loss": 1.4912,
+      "step": 5600
+    },
+    {
+      "epoch": 5.170551580890837,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.270180120080054e-05,
+      "loss": 1.498,
+      "step": 5601
+    },
+    {
+      "epoch": 5.171474728825294,
+      "grad_norm": 3.318359375,
+      "learning_rate": 6.269513008672448e-05,
+      "loss": 1.4121,
+      "step": 5602
+    },
+    {
+      "epoch": 5.172397876759751,
+      "grad_norm": 4.71484375,
+      "learning_rate": 6.268845897264843e-05,
+      "loss": 1.1459,
+      "step": 5603
+    },
+    {
+      "epoch": 5.173321024694207,
+      "grad_norm": 4.26171875,
+      "learning_rate": 6.268178785857239e-05,
+      "loss": 1.3965,
+      "step": 5604
+    },
+    {
+      "epoch": 5.174244172628664,
+      "grad_norm": 3.76953125,
+      "learning_rate": 6.267511674449633e-05,
+      "loss": 1.379,
+      "step": 5605
+    },
+    {
+      "epoch": 5.17516732056312,
+      "grad_norm": 4.41796875,
+      "learning_rate": 6.266844563042028e-05,
+      "loss": 1.418,
+      "step": 5606
+    },
+    {
+      "epoch": 5.176090468497577,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.266177451634422e-05,
+      "loss": 1.5647,
+      "step": 5607
+    },
+    {
+      "epoch": 5.177013616432033,
+      "grad_norm": 3.4921875,
+      "learning_rate": 6.265510340226818e-05,
+      "loss": 1.4126,
+      "step": 5608
+    },
+    {
+      "epoch": 5.17793676436649,
+      "grad_norm": 3.3515625,
+      "learning_rate": 6.264843228819213e-05,
+      "loss": 1.4189,
+      "step": 5609
+    },
+    {
+      "epoch": 5.178859912300946,
+      "grad_norm": 3.5078125,
+      "learning_rate": 6.264176117411607e-05,
+      "loss": 1.3979,
+      "step": 5610
+    },
+    {
+      "epoch": 5.179783060235403,
+      "grad_norm": 4.3828125,
+      "learning_rate": 6.263509006004002e-05,
+      "loss": 1.3838,
+      "step": 5611
+    },
+    {
+      "epoch": 5.180706208169859,
+      "grad_norm": 4.7890625,
+      "learning_rate": 6.262841894596398e-05,
+      "loss": 1.5286,
+      "step": 5612
+    },
+    {
+      "epoch": 5.181629356104316,
+      "grad_norm": 3.728515625,
+      "learning_rate": 6.262174783188794e-05,
+      "loss": 1.4088,
+      "step": 5613
+    },
+    {
+      "epoch": 5.182552504038772,
+      "grad_norm": 3.85546875,
+      "learning_rate": 6.261507671781188e-05,
+      "loss": 1.5583,
+      "step": 5614
+    },
+    {
+      "epoch": 5.183475651973229,
+      "grad_norm": 4.69921875,
+      "learning_rate": 6.260840560373583e-05,
+      "loss": 1.2847,
+      "step": 5615
+    },
+    {
+      "epoch": 5.184398799907685,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.260173448965979e-05,
+      "loss": 1.1307,
+      "step": 5616
+    },
+    {
+      "epoch": 5.185321947842142,
+      "grad_norm": 3.81640625,
+      "learning_rate": 6.259506337558373e-05,
+      "loss": 1.2198,
+      "step": 5617
+    },
+    {
+      "epoch": 5.186245095776599,
+      "grad_norm": 4.140625,
+      "learning_rate": 6.258839226150768e-05,
+      "loss": 1.2443,
+      "step": 5618
+    },
+    {
+      "epoch": 5.187168243711055,
+      "grad_norm": 3.271484375,
+      "learning_rate": 6.258172114743162e-05,
+      "loss": 1.3231,
+      "step": 5619
+    },
+    {
+      "epoch": 5.188091391645512,
+      "grad_norm": 3.216796875,
+      "learning_rate": 6.257505003335558e-05,
+      "loss": 1.2681,
+      "step": 5620
+    },
+    {
+      "epoch": 5.189014539579968,
+      "grad_norm": 3.509765625,
+      "learning_rate": 6.256837891927953e-05,
+      "loss": 1.4858,
+      "step": 5621
+    },
+    {
+      "epoch": 5.1899376875144245,
+      "grad_norm": 4.30859375,
+      "learning_rate": 6.256170780520347e-05,
+      "loss": 1.321,
+      "step": 5622
+    },
+    {
+      "epoch": 5.1908608354488806,
+      "grad_norm": 3.50390625,
+      "learning_rate": 6.255503669112742e-05,
+      "loss": 1.4453,
+      "step": 5623
+    },
+    {
+      "epoch": 5.1917839833833375,
+      "grad_norm": 4.08203125,
+      "learning_rate": 6.254836557705138e-05,
+      "loss": 1.3518,
+      "step": 5624
+    },
+    {
+      "epoch": 5.1927071313177935,
+      "grad_norm": 3.75390625,
+      "learning_rate": 6.254169446297532e-05,
+      "loss": 1.2329,
+      "step": 5625
+    },
+    {
+      "epoch": 5.19363027925225,
+      "grad_norm": 3.966796875,
+      "learning_rate": 6.253502334889927e-05,
+      "loss": 1.4202,
+      "step": 5626
+    },
+    {
+      "epoch": 5.194553427186706,
+      "grad_norm": 3.654296875,
+      "learning_rate": 6.252835223482321e-05,
+      "loss": 1.0751,
+      "step": 5627
+    },
+    {
+      "epoch": 5.195476575121163,
+      "grad_norm": 3.55078125,
+      "learning_rate": 6.252168112074717e-05,
+      "loss": 1.283,
+      "step": 5628
+    },
+    {
+      "epoch": 5.196399723055619,
+      "grad_norm": 3.833984375,
+      "learning_rate": 6.251501000667112e-05,
+      "loss": 1.3816,
+      "step": 5629
+    },
+    {
+      "epoch": 5.197322870990076,
+      "grad_norm": 4.12109375,
+      "learning_rate": 6.250833889259506e-05,
+      "loss": 1.4436,
+      "step": 5630
+    },
+    {
+      "epoch": 5.198246018924532,
+      "grad_norm": 4.171875,
+      "learning_rate": 6.250166777851901e-05,
+      "loss": 1.3945,
+      "step": 5631
+    },
+    {
+      "epoch": 5.199169166858989,
+      "grad_norm": 5.1328125,
+      "learning_rate": 6.249499666444297e-05,
+      "loss": 1.3003,
+      "step": 5632
+    },
+    {
+      "epoch": 5.200092314793445,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.248832555036691e-05,
+      "loss": 1.5452,
+      "step": 5633
+    },
+    {
+      "epoch": 5.201015462727902,
+      "grad_norm": 4.140625,
+      "learning_rate": 6.248165443629086e-05,
+      "loss": 1.4475,
+      "step": 5634
+    },
+    {
+      "epoch": 5.201938610662358,
+      "grad_norm": 3.447265625,
+      "learning_rate": 6.24749833222148e-05,
+      "loss": 1.4949,
+      "step": 5635
+    },
+    {
+      "epoch": 5.202861758596815,
+      "grad_norm": 4.7890625,
+      "learning_rate": 6.246831220813876e-05,
+      "loss": 1.4958,
+      "step": 5636
+    },
+    {
+      "epoch": 5.203784906531272,
+      "grad_norm": 3.958984375,
+      "learning_rate": 6.246164109406271e-05,
+      "loss": 1.3181,
+      "step": 5637
+    },
+    {
+      "epoch": 5.204708054465728,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.245496997998667e-05,
+      "loss": 1.2957,
+      "step": 5638
+    },
+    {
+      "epoch": 5.205631202400185,
+      "grad_norm": 5.33203125,
+      "learning_rate": 6.244829886591061e-05,
+      "loss": 1.4236,
+      "step": 5639
+    },
+    {
+      "epoch": 5.206554350334641,
+      "grad_norm": 3.255859375,
+      "learning_rate": 6.244162775183456e-05,
+      "loss": 1.4504,
+      "step": 5640
+    },
+    {
+      "epoch": 5.207477498269098,
+      "grad_norm": 3.66796875,
+      "learning_rate": 6.243495663775852e-05,
+      "loss": 1.2855,
+      "step": 5641
+    },
+    {
+      "epoch": 5.208400646203554,
+      "grad_norm": 4.18359375,
+      "learning_rate": 6.242828552368246e-05,
+      "loss": 1.3713,
+      "step": 5642
+    },
+    {
+      "epoch": 5.209323794138011,
+      "grad_norm": 4.03515625,
+      "learning_rate": 6.242161440960641e-05,
+      "loss": 1.1776,
+      "step": 5643
+    },
+    {
+      "epoch": 5.210246942072467,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.241494329553035e-05,
+      "loss": 1.5325,
+      "step": 5644
+    },
+    {
+      "epoch": 5.211170090006924,
+      "grad_norm": 3.923828125,
+      "learning_rate": 6.240827218145431e-05,
+      "loss": 1.2771,
+      "step": 5645
+    },
+    {
+      "epoch": 5.21209323794138,
+      "grad_norm": 3.64453125,
+      "learning_rate": 6.240160106737826e-05,
+      "loss": 1.3033,
+      "step": 5646
+    },
+    {
+      "epoch": 5.213016385875837,
+      "grad_norm": 3.4453125,
+      "learning_rate": 6.23949299533022e-05,
+      "loss": 1.29,
+      "step": 5647
+    },
+    {
+      "epoch": 5.213939533810293,
+      "grad_norm": 3.802734375,
+      "learning_rate": 6.238825883922615e-05,
+      "loss": 1.6523,
+      "step": 5648
+    },
+    {
+      "epoch": 5.21486268174475,
+      "grad_norm": 4.68359375,
+      "learning_rate": 6.238158772515011e-05,
+      "loss": 1.4028,
+      "step": 5649
+    },
+    {
+      "epoch": 5.215785829679206,
+      "grad_norm": 3.81640625,
+      "learning_rate": 6.237491661107405e-05,
+      "loss": 1.3992,
+      "step": 5650
+    },
+    {
+      "epoch": 5.216708977613663,
+      "grad_norm": 4.78515625,
+      "learning_rate": 6.2368245496998e-05,
+      "loss": 1.385,
+      "step": 5651
+    },
+    {
+      "epoch": 5.2176321255481195,
+      "grad_norm": 3.646484375,
+      "learning_rate": 6.236157438292194e-05,
+      "loss": 1.5059,
+      "step": 5652
+    },
+    {
+      "epoch": 5.2185552734825755,
+      "grad_norm": 3.599609375,
+      "learning_rate": 6.23549032688459e-05,
+      "loss": 1.2649,
+      "step": 5653
+    },
+    {
+      "epoch": 5.219478421417032,
+      "grad_norm": 3.263671875,
+      "learning_rate": 6.234823215476985e-05,
+      "loss": 1.5574,
+      "step": 5654
+    },
+    {
+      "epoch": 5.2204015693514885,
+      "grad_norm": 3.439453125,
+      "learning_rate": 6.23415610406938e-05,
+      "loss": 1.5225,
+      "step": 5655
+    },
+    {
+      "epoch": 5.221324717285945,
+      "grad_norm": 3.912109375,
+      "learning_rate": 6.233488992661774e-05,
+      "loss": 1.2734,
+      "step": 5656
+    },
+    {
+      "epoch": 5.222247865220401,
+      "grad_norm": 3.58984375,
+      "learning_rate": 6.23282188125417e-05,
+      "loss": 1.297,
+      "step": 5657
+    },
+    {
+      "epoch": 5.223171013154858,
+      "grad_norm": 4.0078125,
+      "learning_rate": 6.232154769846564e-05,
+      "loss": 1.175,
+      "step": 5658
+    },
+    {
+      "epoch": 5.224094161089314,
+      "grad_norm": 3.23828125,
+      "learning_rate": 6.231487658438959e-05,
+      "loss": 1.429,
+      "step": 5659
+    },
+    {
+      "epoch": 5.225017309023771,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.230820547031355e-05,
+      "loss": 1.4365,
+      "step": 5660
+    },
+    {
+      "epoch": 5.225940456958227,
+      "grad_norm": 3.169921875,
+      "learning_rate": 6.230153435623749e-05,
+      "loss": 1.3048,
+      "step": 5661
+    },
+    {
+      "epoch": 5.226863604892684,
+      "grad_norm": 3.9765625,
+      "learning_rate": 6.229486324216144e-05,
+      "loss": 1.3376,
+      "step": 5662
+    },
+    {
+      "epoch": 5.22778675282714,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.22881921280854e-05,
+      "loss": 1.4032,
+      "step": 5663
+    },
+    {
+      "epoch": 5.228709900761597,
+      "grad_norm": 3.96875,
+      "learning_rate": 6.228152101400934e-05,
+      "loss": 1.511,
+      "step": 5664
+    },
+    {
+      "epoch": 5.229633048696053,
+      "grad_norm": 3.16015625,
+      "learning_rate": 6.22748498999333e-05,
+      "loss": 1.298,
+      "step": 5665
+    },
+    {
+      "epoch": 5.23055619663051,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.226817878585725e-05,
+      "loss": 1.2906,
+      "step": 5666
+    },
+    {
+      "epoch": 5.231479344564967,
+      "grad_norm": 4.44921875,
+      "learning_rate": 6.226150767178119e-05,
+      "loss": 1.1147,
+      "step": 5667
+    },
+    {
+      "epoch": 5.232402492499423,
+      "grad_norm": 4.73828125,
+      "learning_rate": 6.225483655770514e-05,
+      "loss": 1.2307,
+      "step": 5668
+    },
+    {
+      "epoch": 5.23332564043388,
+      "grad_norm": 3.66015625,
+      "learning_rate": 6.22481654436291e-05,
+      "loss": 1.5,
+      "step": 5669
+    },
+    {
+      "epoch": 5.234248788368336,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.224149432955304e-05,
+      "loss": 1.2793,
+      "step": 5670
+    },
+    {
+      "epoch": 5.235171936302793,
+      "grad_norm": 3.654296875,
+      "learning_rate": 6.223482321547699e-05,
+      "loss": 1.2952,
+      "step": 5671
+    },
+    {
+      "epoch": 5.236095084237249,
+      "grad_norm": 3.80078125,
+      "learning_rate": 6.222815210140093e-05,
+      "loss": 1.3699,
+      "step": 5672
+    },
+    {
+      "epoch": 5.237018232171706,
+      "grad_norm": 3.255859375,
+      "learning_rate": 6.222148098732489e-05,
+      "loss": 1.3224,
+      "step": 5673
+    },
+    {
+      "epoch": 5.237941380106162,
+      "grad_norm": 4.41015625,
+      "learning_rate": 6.221480987324884e-05,
+      "loss": 1.1986,
+      "step": 5674
+    },
+    {
+      "epoch": 5.238864528040619,
+      "grad_norm": 4.46484375,
+      "learning_rate": 6.220813875917278e-05,
+      "loss": 1.3695,
+      "step": 5675
+    },
+    {
+      "epoch": 5.239787675975075,
+      "grad_norm": 4.37109375,
+      "learning_rate": 6.220146764509673e-05,
+      "loss": 1.3901,
+      "step": 5676
+    },
+    {
+      "epoch": 5.240710823909532,
+      "grad_norm": 3.3984375,
+      "learning_rate": 6.219479653102069e-05,
+      "loss": 1.3997,
+      "step": 5677
+    },
+    {
+      "epoch": 5.241633971843988,
+      "grad_norm": 3.8671875,
+      "learning_rate": 6.218812541694463e-05,
+      "loss": 1.2089,
+      "step": 5678
+    },
+    {
+      "epoch": 5.242557119778445,
+      "grad_norm": 3.458984375,
+      "learning_rate": 6.218145430286858e-05,
+      "loss": 1.542,
+      "step": 5679
+    },
+    {
+      "epoch": 5.243480267712901,
+      "grad_norm": 3.53515625,
+      "learning_rate": 6.217478318879252e-05,
+      "loss": 1.4998,
+      "step": 5680
+    },
+    {
+      "epoch": 5.244403415647358,
+      "grad_norm": 3.408203125,
+      "learning_rate": 6.216811207471648e-05,
+      "loss": 1.3164,
+      "step": 5681
+    },
+    {
+      "epoch": 5.245326563581814,
+      "grad_norm": 4.46875,
+      "learning_rate": 6.216144096064043e-05,
+      "loss": 1.6165,
+      "step": 5682
+    },
+    {
+      "epoch": 5.2462497115162705,
+      "grad_norm": 4.5859375,
+      "learning_rate": 6.215476984656437e-05,
+      "loss": 1.4094,
+      "step": 5683
+    },
+    {
+      "epoch": 5.2471728594507265,
+      "grad_norm": 3.75390625,
+      "learning_rate": 6.214809873248832e-05,
+      "loss": 1.0649,
+      "step": 5684
+    },
+    {
+      "epoch": 5.2480960073851834,
+      "grad_norm": 3.642578125,
+      "learning_rate": 6.214142761841228e-05,
+      "loss": 1.174,
+      "step": 5685
+    },
+    {
+      "epoch": 5.24901915531964,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.213475650433622e-05,
+      "loss": 1.2271,
+      "step": 5686
+    },
+    {
+      "epoch": 5.249942303254096,
+      "grad_norm": 4.28515625,
+      "learning_rate": 6.212808539026017e-05,
+      "loss": 1.4773,
+      "step": 5687
+    },
+    {
+      "epoch": 5.250865451188553,
+      "grad_norm": 3.791015625,
+      "learning_rate": 6.212141427618413e-05,
+      "loss": 1.3632,
+      "step": 5688
+    },
+    {
+      "epoch": 5.251788599123009,
+      "grad_norm": 4.69921875,
+      "learning_rate": 6.211474316210807e-05,
+      "loss": 1.4475,
+      "step": 5689
+    },
+    {
+      "epoch": 5.252711747057466,
+      "grad_norm": 3.69140625,
+      "learning_rate": 6.210807204803203e-05,
+      "loss": 1.4497,
+      "step": 5690
+    },
+    {
+      "epoch": 5.253634894991922,
+      "grad_norm": 3.7265625,
+      "learning_rate": 6.210140093395598e-05,
+      "loss": 1.624,
+      "step": 5691
+    },
+    {
+      "epoch": 5.254558042926379,
+      "grad_norm": 4.33203125,
+      "learning_rate": 6.209472981987992e-05,
+      "loss": 1.2477,
+      "step": 5692
+    },
+    {
+      "epoch": 5.255481190860835,
+      "grad_norm": 3.357421875,
+      "learning_rate": 6.208805870580387e-05,
+      "loss": 1.3398,
+      "step": 5693
+    },
+    {
+      "epoch": 5.256404338795292,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.208138759172783e-05,
+      "loss": 1.2622,
+      "step": 5694
+    },
+    {
+      "epoch": 5.257327486729748,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.207471647765177e-05,
+      "loss": 1.3994,
+      "step": 5695
+    },
+    {
+      "epoch": 5.258250634664205,
+      "grad_norm": 3.94140625,
+      "learning_rate": 6.206804536357572e-05,
+      "loss": 1.2631,
+      "step": 5696
+    },
+    {
+      "epoch": 5.259173782598661,
+      "grad_norm": 3.5234375,
+      "learning_rate": 6.206137424949966e-05,
+      "loss": 1.458,
+      "step": 5697
+    },
+    {
+      "epoch": 5.260096930533118,
+      "grad_norm": 3.677734375,
+      "learning_rate": 6.205470313542362e-05,
+      "loss": 1.5396,
+      "step": 5698
+    },
+    {
+      "epoch": 5.261020078467574,
+      "grad_norm": 3.69921875,
+      "learning_rate": 6.204803202134757e-05,
+      "loss": 1.5334,
+      "step": 5699
+    },
+    {
+      "epoch": 5.261943226402031,
+      "grad_norm": 4.6171875,
+      "learning_rate": 6.204136090727151e-05,
+      "loss": 1.302,
+      "step": 5700
+    },
+    {
+      "epoch": 5.262866374336488,
+      "grad_norm": 4.94921875,
+      "learning_rate": 6.203468979319547e-05,
+      "loss": 1.2915,
+      "step": 5701
+    },
+    {
+      "epoch": 5.263789522270944,
+      "grad_norm": 3.96484375,
+      "learning_rate": 6.202801867911942e-05,
+      "loss": 1.3684,
+      "step": 5702
+    },
+    {
+      "epoch": 5.264712670205401,
+      "grad_norm": 3.205078125,
+      "learning_rate": 6.202134756504336e-05,
+      "loss": 1.3989,
+      "step": 5703
+    },
+    {
+      "epoch": 5.265635818139857,
+      "grad_norm": 3.4140625,
+      "learning_rate": 6.201467645096731e-05,
+      "loss": 1.4175,
+      "step": 5704
+    },
+    {
+      "epoch": 5.266558966074314,
+      "grad_norm": 3.720703125,
+      "learning_rate": 6.200800533689127e-05,
+      "loss": 1.334,
+      "step": 5705
+    },
+    {
+      "epoch": 5.26748211400877,
+      "grad_norm": 3.830078125,
+      "learning_rate": 6.200133422281521e-05,
+      "loss": 1.6543,
+      "step": 5706
+    },
+    {
+      "epoch": 5.268405261943227,
+      "grad_norm": 3.982421875,
+      "learning_rate": 6.199466310873916e-05,
+      "loss": 1.275,
+      "step": 5707
+    },
+    {
+      "epoch": 5.269328409877683,
+      "grad_norm": 4.0546875,
+      "learning_rate": 6.19879919946631e-05,
+      "loss": 1.4922,
+      "step": 5708
+    },
+    {
+      "epoch": 5.27025155781214,
+      "grad_norm": 3.244140625,
+      "learning_rate": 6.198132088058706e-05,
+      "loss": 1.4619,
+      "step": 5709
+    },
+    {
+      "epoch": 5.271174705746596,
+      "grad_norm": 3.935546875,
+      "learning_rate": 6.197464976651101e-05,
+      "loss": 1.4873,
+      "step": 5710
+    },
+    {
+      "epoch": 5.2720978536810525,
+      "grad_norm": 4.91015625,
+      "learning_rate": 6.196797865243495e-05,
+      "loss": 1.1017,
+      "step": 5711
+    },
+    {
+      "epoch": 5.273021001615509,
+      "grad_norm": 3.373046875,
+      "learning_rate": 6.19613075383589e-05,
+      "loss": 1.0939,
+      "step": 5712
+    },
+    {
+      "epoch": 5.2739441495499655,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.195463642428286e-05,
+      "loss": 1.5469,
+      "step": 5713
+    },
+    {
+      "epoch": 5.2748672974844215,
+      "grad_norm": 3.712890625,
+      "learning_rate": 6.194796531020682e-05,
+      "loss": 1.1958,
+      "step": 5714
+    },
+    {
+      "epoch": 5.275790445418878,
+      "grad_norm": 3.912109375,
+      "learning_rate": 6.194129419613076e-05,
+      "loss": 1.097,
+      "step": 5715
+    },
+    {
+      "epoch": 5.276713593353335,
+      "grad_norm": 3.7109375,
+      "learning_rate": 6.193462308205471e-05,
+      "loss": 1.4932,
+      "step": 5716
+    },
+    {
+      "epoch": 5.277636741287791,
+      "grad_norm": 4.8671875,
+      "learning_rate": 6.192795196797865e-05,
+      "loss": 1.256,
+      "step": 5717
+    },
+    {
+      "epoch": 5.278559889222248,
+      "grad_norm": 4.22265625,
+      "learning_rate": 6.192128085390261e-05,
+      "loss": 1.2421,
+      "step": 5718
+    },
+    {
+      "epoch": 5.279483037156704,
+      "grad_norm": 5.01953125,
+      "learning_rate": 6.191460973982656e-05,
+      "loss": 1.4465,
+      "step": 5719
+    },
+    {
+      "epoch": 5.280406185091161,
+      "grad_norm": 3.431640625,
+      "learning_rate": 6.19079386257505e-05,
+      "loss": 1.4119,
+      "step": 5720
+    },
+    {
+      "epoch": 5.281329333025617,
+      "grad_norm": 3.673828125,
+      "learning_rate": 6.190126751167445e-05,
+      "loss": 1.3423,
+      "step": 5721
+    },
+    {
+      "epoch": 5.282252480960074,
+      "grad_norm": 3.630859375,
+      "learning_rate": 6.189459639759841e-05,
+      "loss": 1.3455,
+      "step": 5722
+    },
+    {
+      "epoch": 5.28317562889453,
+      "grad_norm": 3.396484375,
+      "learning_rate": 6.188792528352235e-05,
+      "loss": 1.1842,
+      "step": 5723
+    },
+    {
+      "epoch": 5.284098776828987,
+      "grad_norm": 3.974609375,
+      "learning_rate": 6.18812541694463e-05,
+      "loss": 1.5168,
+      "step": 5724
+    },
+    {
+      "epoch": 5.285021924763443,
+      "grad_norm": 3.9296875,
+      "learning_rate": 6.187458305537024e-05,
+      "loss": 1.2981,
+      "step": 5725
+    },
+    {
+      "epoch": 5.2859450726979,
+      "grad_norm": 3.939453125,
+      "learning_rate": 6.18679119412942e-05,
+      "loss": 1.4587,
+      "step": 5726
+    },
+    {
+      "epoch": 5.286868220632356,
+      "grad_norm": 4.4296875,
+      "learning_rate": 6.186124082721815e-05,
+      "loss": 1.3716,
+      "step": 5727
+    },
+    {
+      "epoch": 5.287791368566813,
+      "grad_norm": 3.380859375,
+      "learning_rate": 6.18545697131421e-05,
+      "loss": 1.481,
+      "step": 5728
+    },
+    {
+      "epoch": 5.288714516501269,
+      "grad_norm": 4.96484375,
+      "learning_rate": 6.184789859906604e-05,
+      "loss": 1.3186,
+      "step": 5729
+    },
+    {
+      "epoch": 5.289637664435726,
+      "grad_norm": 4.12109375,
+      "learning_rate": 6.184122748499e-05,
+      "loss": 1.457,
+      "step": 5730
+    },
+    {
+      "epoch": 5.290560812370182,
+      "grad_norm": 4.359375,
+      "learning_rate": 6.183455637091394e-05,
+      "loss": 1.3324,
+      "step": 5731
+    },
+    {
+      "epoch": 5.291483960304639,
+      "grad_norm": 3.912109375,
+      "learning_rate": 6.182788525683789e-05,
+      "loss": 1.4285,
+      "step": 5732
+    },
+    {
+      "epoch": 5.292407108239095,
+      "grad_norm": 3.1171875,
+      "learning_rate": 6.182121414276184e-05,
+      "loss": 1.3359,
+      "step": 5733
+    },
+    {
+      "epoch": 5.293330256173552,
+      "grad_norm": 3.6484375,
+      "learning_rate": 6.18145430286858e-05,
+      "loss": 1.374,
+      "step": 5734
+    },
+    {
+      "epoch": 5.294253404108009,
+      "grad_norm": 3.48046875,
+      "learning_rate": 6.180787191460974e-05,
+      "loss": 1.4973,
+      "step": 5735
+    },
+    {
+      "epoch": 5.295176552042465,
+      "grad_norm": 3.83984375,
+      "learning_rate": 6.180120080053368e-05,
+      "loss": 1.3833,
+      "step": 5736
+    },
+    {
+      "epoch": 5.296099699976922,
+      "grad_norm": 3.505859375,
+      "learning_rate": 6.179452968645763e-05,
+      "loss": 1.2402,
+      "step": 5737
+    },
+    {
+      "epoch": 5.297022847911378,
+      "grad_norm": 3.408203125,
+      "learning_rate": 6.17878585723816e-05,
+      "loss": 1.5154,
+      "step": 5738
+    },
+    {
+      "epoch": 5.297945995845835,
+      "grad_norm": 3.2734375,
+      "learning_rate": 6.178118745830555e-05,
+      "loss": 1.5649,
+      "step": 5739
+    },
+    {
+      "epoch": 5.298869143780291,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.17745163442295e-05,
+      "loss": 1.2423,
+      "step": 5740
+    },
+    {
+      "epoch": 5.2997922917147475,
+      "grad_norm": 3.63671875,
+      "learning_rate": 6.176784523015344e-05,
+      "loss": 1.0975,
+      "step": 5741
+    },
+    {
+      "epoch": 5.3007154396492036,
+      "grad_norm": 3.814453125,
+      "learning_rate": 6.17611741160774e-05,
+      "loss": 1.5112,
+      "step": 5742
+    },
+    {
+      "epoch": 5.3016385875836605,
+      "grad_norm": 3.837890625,
+      "learning_rate": 6.175450300200134e-05,
+      "loss": 1.3157,
+      "step": 5743
+    },
+    {
+      "epoch": 5.3025617355181165,
+      "grad_norm": 4.33203125,
+      "learning_rate": 6.174783188792529e-05,
+      "loss": 1.1361,
+      "step": 5744
+    },
+    {
+      "epoch": 5.303484883452573,
+      "grad_norm": 3.408203125,
+      "learning_rate": 6.174116077384923e-05,
+      "loss": 1.3323,
+      "step": 5745
+    },
+    {
+      "epoch": 5.304408031387029,
+      "grad_norm": 4.08984375,
+      "learning_rate": 6.17344896597732e-05,
+      "loss": 1.3215,
+      "step": 5746
+    },
+    {
+      "epoch": 5.305331179321486,
+      "grad_norm": 3.662109375,
+      "learning_rate": 6.172781854569714e-05,
+      "loss": 1.3125,
+      "step": 5747
+    },
+    {
+      "epoch": 5.306254327255942,
+      "grad_norm": 3.216796875,
+      "learning_rate": 6.172114743162108e-05,
+      "loss": 1.3867,
+      "step": 5748
+    },
+    {
+      "epoch": 5.307177475190399,
+      "grad_norm": 3.888671875,
+      "learning_rate": 6.171447631754503e-05,
+      "loss": 1.3848,
+      "step": 5749
+    },
+    {
+      "epoch": 5.308100623124856,
+      "grad_norm": 3.787109375,
+      "learning_rate": 6.170780520346899e-05,
+      "loss": 1.7986,
+      "step": 5750
+    },
+    {
+      "epoch": 5.309023771059312,
+      "grad_norm": 3.099609375,
+      "learning_rate": 6.170113408939293e-05,
+      "loss": 1.4758,
+      "step": 5751
+    },
+    {
+      "epoch": 5.309946918993769,
+      "grad_norm": 3.8671875,
+      "learning_rate": 6.169446297531688e-05,
+      "loss": 1.5078,
+      "step": 5752
+    },
+    {
+      "epoch": 5.310870066928225,
+      "grad_norm": 3.85546875,
+      "learning_rate": 6.168779186124083e-05,
+      "loss": 1.1495,
+      "step": 5753
+    },
+    {
+      "epoch": 5.311793214862682,
+      "grad_norm": 3.767578125,
+      "learning_rate": 6.168112074716478e-05,
+      "loss": 1.7034,
+      "step": 5754
+    },
+    {
+      "epoch": 5.312716362797138,
+      "grad_norm": 4.37109375,
+      "learning_rate": 6.167444963308873e-05,
+      "loss": 1.3464,
+      "step": 5755
+    },
+    {
+      "epoch": 5.313639510731595,
+      "grad_norm": 3.310546875,
+      "learning_rate": 6.166777851901267e-05,
+      "loss": 1.3577,
+      "step": 5756
+    },
+    {
+      "epoch": 5.314562658666051,
+      "grad_norm": 3.771484375,
+      "learning_rate": 6.166110740493662e-05,
+      "loss": 1.5164,
+      "step": 5757
+    },
+    {
+      "epoch": 5.315485806600508,
+      "grad_norm": 3.62890625,
+      "learning_rate": 6.165443629086058e-05,
+      "loss": 1.2134,
+      "step": 5758
+    },
+    {
+      "epoch": 5.316408954534964,
+      "grad_norm": 3.67578125,
+      "learning_rate": 6.164776517678452e-05,
+      "loss": 1.3629,
+      "step": 5759
+    },
+    {
+      "epoch": 5.317332102469421,
+      "grad_norm": 3.46484375,
+      "learning_rate": 6.164109406270847e-05,
+      "loss": 1.4561,
+      "step": 5760
+    },
+    {
+      "epoch": 5.318255250403877,
+      "grad_norm": 3.20703125,
+      "learning_rate": 6.163442294863242e-05,
+      "loss": 1.313,
+      "step": 5761
+    },
+    {
+      "epoch": 5.319178398338334,
+      "grad_norm": 3.759765625,
+      "learning_rate": 6.162775183455637e-05,
+      "loss": 1.6746,
+      "step": 5762
+    },
+    {
+      "epoch": 5.32010154627279,
+      "grad_norm": 3.591796875,
+      "learning_rate": 6.162108072048033e-05,
+      "loss": 1.3257,
+      "step": 5763
+    },
+    {
+      "epoch": 5.321024694207247,
+      "grad_norm": 3.654296875,
+      "learning_rate": 6.161440960640428e-05,
+      "loss": 1.5793,
+      "step": 5764
+    },
+    {
+      "epoch": 5.321947842141703,
+      "grad_norm": 4.6640625,
+      "learning_rate": 6.160773849232822e-05,
+      "loss": 1.6743,
+      "step": 5765
+    },
+    {
+      "epoch": 5.32287099007616,
+      "grad_norm": 3.392578125,
+      "learning_rate": 6.160106737825217e-05,
+      "loss": 1.3623,
+      "step": 5766
+    },
+    {
+      "epoch": 5.323794138010616,
+      "grad_norm": 3.9765625,
+      "learning_rate": 6.159439626417613e-05,
+      "loss": 1.4546,
+      "step": 5767
+    },
+    {
+      "epoch": 5.324717285945073,
+      "grad_norm": 4.03515625,
+      "learning_rate": 6.158772515010007e-05,
+      "loss": 1.2198,
+      "step": 5768
+    },
+    {
+      "epoch": 5.32564043387953,
+      "grad_norm": 3.32421875,
+      "learning_rate": 6.158105403602402e-05,
+      "loss": 1.3132,
+      "step": 5769
+    },
+    {
+      "epoch": 5.326563581813986,
+      "grad_norm": 3.4765625,
+      "learning_rate": 6.157438292194797e-05,
+      "loss": 1.1582,
+      "step": 5770
+    },
+    {
+      "epoch": 5.3274867297484425,
+      "grad_norm": 3.560546875,
+      "learning_rate": 6.156771180787192e-05,
+      "loss": 1.3594,
+      "step": 5771
+    },
+    {
+      "epoch": 5.3284098776828985,
+      "grad_norm": 3.509765625,
+      "learning_rate": 6.156104069379587e-05,
+      "loss": 1.4131,
+      "step": 5772
+    },
+    {
+      "epoch": 5.3293330256173554,
+      "grad_norm": 4.421875,
+      "learning_rate": 6.155436957971981e-05,
+      "loss": 1.4612,
+      "step": 5773
+    },
+    {
+      "epoch": 5.3302561735518115,
+      "grad_norm": 3.580078125,
+      "learning_rate": 6.154769846564376e-05,
+      "loss": 1.3057,
+      "step": 5774
+    },
+    {
+      "epoch": 5.331179321486268,
+      "grad_norm": 4.328125,
+      "learning_rate": 6.154102735156772e-05,
+      "loss": 1.3975,
+      "step": 5775
+    },
+    {
+      "epoch": 5.332102469420724,
+      "grad_norm": 4.1953125,
+      "learning_rate": 6.153435623749166e-05,
+      "loss": 1.3634,
+      "step": 5776
+    },
+    {
+      "epoch": 5.333025617355181,
+      "grad_norm": 3.943359375,
+      "learning_rate": 6.152768512341561e-05,
+      "loss": 1.3037,
+      "step": 5777
+    },
+    {
+      "epoch": 5.333948765289637,
+      "grad_norm": 4.28125,
+      "learning_rate": 6.152101400933956e-05,
+      "loss": 1.3951,
+      "step": 5778
+    },
+    {
+      "epoch": 5.334871913224094,
+      "grad_norm": 3.18359375,
+      "learning_rate": 6.151434289526351e-05,
+      "loss": 1.679,
+      "step": 5779
+    },
+    {
+      "epoch": 5.33579506115855,
+      "grad_norm": 3.470703125,
+      "learning_rate": 6.150767178118746e-05,
+      "loss": 1.4551,
+      "step": 5780
+    },
+    {
+      "epoch": 5.336718209093007,
+      "grad_norm": 3.70703125,
+      "learning_rate": 6.15010006671114e-05,
+      "loss": 1.5371,
+      "step": 5781
+    },
+    {
+      "epoch": 5.337641357027463,
+      "grad_norm": 4.3046875,
+      "learning_rate": 6.149432955303535e-05,
+      "loss": 1.2293,
+      "step": 5782
+    },
+    {
+      "epoch": 5.33856450496192,
+      "grad_norm": 4.23828125,
+      "learning_rate": 6.148765843895931e-05,
+      "loss": 1.438,
+      "step": 5783
+    },
+    {
+      "epoch": 5.339487652896377,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.148098732488326e-05,
+      "loss": 1.0004,
+      "step": 5784
+    },
+    {
+      "epoch": 5.340410800830833,
+      "grad_norm": 3.3203125,
+      "learning_rate": 6.14743162108072e-05,
+      "loss": 1.4592,
+      "step": 5785
+    },
+    {
+      "epoch": 5.34133394876529,
+      "grad_norm": 3.322265625,
+      "learning_rate": 6.146764509673115e-05,
+      "loss": 1.4006,
+      "step": 5786
+    },
+    {
+      "epoch": 5.342257096699746,
+      "grad_norm": 3.83203125,
+      "learning_rate": 6.14609739826551e-05,
+      "loss": 1.208,
+      "step": 5787
+    },
+    {
+      "epoch": 5.343180244634203,
+      "grad_norm": 3.6328125,
+      "learning_rate": 6.145430286857906e-05,
+      "loss": 1.2026,
+      "step": 5788
+    },
+    {
+      "epoch": 5.344103392568659,
+      "grad_norm": 3.064453125,
+      "learning_rate": 6.144763175450301e-05,
+      "loss": 1.2345,
+      "step": 5789
+    },
+    {
+      "epoch": 5.345026540503116,
+      "grad_norm": 3.625,
+      "learning_rate": 6.144096064042696e-05,
+      "loss": 1.2849,
+      "step": 5790
+    },
+    {
+      "epoch": 5.345949688437572,
+      "grad_norm": 3.57421875,
+      "learning_rate": 6.143428952635091e-05,
+      "loss": 1.2612,
+      "step": 5791
+    },
+    {
+      "epoch": 5.346872836372029,
+      "grad_norm": 3.623046875,
+      "learning_rate": 6.142761841227486e-05,
+      "loss": 1.003,
+      "step": 5792
+    },
+    {
+      "epoch": 5.347795984306485,
+      "grad_norm": 3.912109375,
+      "learning_rate": 6.14209472981988e-05,
+      "loss": 1.3541,
+      "step": 5793
+    },
+    {
+      "epoch": 5.348719132240942,
+      "grad_norm": 3.361328125,
+      "learning_rate": 6.141427618412275e-05,
+      "loss": 1.3708,
+      "step": 5794
+    },
+    {
+      "epoch": 5.349642280175398,
+      "grad_norm": 3.888671875,
+      "learning_rate": 6.140760507004671e-05,
+      "loss": 1.6245,
+      "step": 5795
+    },
+    {
+      "epoch": 5.350565428109855,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.140093395597065e-05,
+      "loss": 1.1605,
+      "step": 5796
+    },
+    {
+      "epoch": 5.351488576044311,
+      "grad_norm": 3.998046875,
+      "learning_rate": 6.13942628418946e-05,
+      "loss": 1.3052,
+      "step": 5797
+    },
+    {
+      "epoch": 5.352411723978768,
+      "grad_norm": 4.43359375,
+      "learning_rate": 6.138759172781855e-05,
+      "loss": 1.3784,
+      "step": 5798
+    },
+    {
+      "epoch": 5.3533348719132245,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.13809206137425e-05,
+      "loss": 1.4431,
+      "step": 5799
+    },
+    {
+      "epoch": 5.354258019847681,
+      "grad_norm": 4.95703125,
+      "learning_rate": 6.137424949966645e-05,
+      "loss": 1.3684,
+      "step": 5800
+    },
+    {
+      "epoch": 5.3551811677821375,
+      "grad_norm": 4.1171875,
+      "learning_rate": 6.13675783855904e-05,
+      "loss": 1.4768,
+      "step": 5801
+    },
+    {
+      "epoch": 5.3561043157165935,
+      "grad_norm": 3.244140625,
+      "learning_rate": 6.136090727151434e-05,
+      "loss": 1.5559,
+      "step": 5802
+    },
+    {
+      "epoch": 5.35702746365105,
+      "grad_norm": 3.228515625,
+      "learning_rate": 6.13542361574383e-05,
+      "loss": 1.5769,
+      "step": 5803
+    },
+    {
+      "epoch": 5.3579506115855065,
+      "grad_norm": 4.046875,
+      "learning_rate": 6.134756504336225e-05,
+      "loss": 1.438,
+      "step": 5804
+    },
+    {
+      "epoch": 5.358873759519963,
+      "grad_norm": 3.64453125,
+      "learning_rate": 6.134089392928619e-05,
+      "loss": 1.6506,
+      "step": 5805
+    },
+    {
+      "epoch": 5.359796907454419,
+      "grad_norm": 4.046875,
+      "learning_rate": 6.133422281521014e-05,
+      "loss": 1.4019,
+      "step": 5806
+    },
+    {
+      "epoch": 5.360720055388876,
+      "grad_norm": 3.72265625,
+      "learning_rate": 6.13275517011341e-05,
+      "loss": 1.3894,
+      "step": 5807
+    },
+    {
+      "epoch": 5.361643203323332,
+      "grad_norm": 3.943359375,
+      "learning_rate": 6.132088058705804e-05,
+      "loss": 1.3552,
+      "step": 5808
+    },
+    {
+      "epoch": 5.362566351257789,
+      "grad_norm": 3.375,
+      "learning_rate": 6.131420947298199e-05,
+      "loss": 1.3447,
+      "step": 5809
+    },
+    {
+      "epoch": 5.363489499192245,
+      "grad_norm": 3.787109375,
+      "learning_rate": 6.130753835890593e-05,
+      "loss": 1.1775,
+      "step": 5810
+    },
+    {
+      "epoch": 5.364412647126702,
+      "grad_norm": 3.568359375,
+      "learning_rate": 6.130086724482989e-05,
+      "loss": 1.488,
+      "step": 5811
+    },
+    {
+      "epoch": 5.365335795061158,
+      "grad_norm": 3.58984375,
+      "learning_rate": 6.129419613075384e-05,
+      "loss": 1.323,
+      "step": 5812
+    },
+    {
+      "epoch": 5.366258942995615,
+      "grad_norm": 3.6171875,
+      "learning_rate": 6.12875250166778e-05,
+      "loss": 1.1181,
+      "step": 5813
+    },
+    {
+      "epoch": 5.367182090930071,
+      "grad_norm": 3.7265625,
+      "learning_rate": 6.128085390260174e-05,
+      "loss": 1.5347,
+      "step": 5814
+    },
+    {
+      "epoch": 5.368105238864528,
+      "grad_norm": 3.970703125,
+      "learning_rate": 6.127418278852569e-05,
+      "loss": 1.3169,
+      "step": 5815
+    },
+    {
+      "epoch": 5.369028386798984,
+      "grad_norm": 3.666015625,
+      "learning_rate": 6.126751167444964e-05,
+      "loss": 1.3691,
+      "step": 5816
+    },
+    {
+      "epoch": 5.369951534733441,
+      "grad_norm": 4.33984375,
+      "learning_rate": 6.126084056037359e-05,
+      "loss": 1.5115,
+      "step": 5817
+    },
+    {
+      "epoch": 5.370874682667898,
+      "grad_norm": 4.05859375,
+      "learning_rate": 6.125416944629754e-05,
+      "loss": 1.2534,
+      "step": 5818
+    },
+    {
+      "epoch": 5.371797830602354,
+      "grad_norm": 3.904296875,
+      "learning_rate": 6.124749833222148e-05,
+      "loss": 1.5239,
+      "step": 5819
+    },
+    {
+      "epoch": 5.372720978536811,
+      "grad_norm": 3.87109375,
+      "learning_rate": 6.124082721814544e-05,
+      "loss": 1.5815,
+      "step": 5820
+    },
+    {
+      "epoch": 5.373644126471267,
+      "grad_norm": 3.763671875,
+      "learning_rate": 6.123415610406939e-05,
+      "loss": 1.4216,
+      "step": 5821
+    },
+    {
+      "epoch": 5.374567274405724,
+      "grad_norm": 4.01953125,
+      "learning_rate": 6.122748498999333e-05,
+      "loss": 1.4398,
+      "step": 5822
+    },
+    {
+      "epoch": 5.37549042234018,
+      "grad_norm": 3.791015625,
+      "learning_rate": 6.122081387591728e-05,
+      "loss": 1.1476,
+      "step": 5823
+    },
+    {
+      "epoch": 5.376413570274637,
+      "grad_norm": 4.4140625,
+      "learning_rate": 6.121414276184124e-05,
+      "loss": 1.3765,
+      "step": 5824
+    },
+    {
+      "epoch": 5.377336718209093,
+      "grad_norm": 3.822265625,
+      "learning_rate": 6.120747164776518e-05,
+      "loss": 1.5186,
+      "step": 5825
+    },
+    {
+      "epoch": 5.37825986614355,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.120080053368913e-05,
+      "loss": 1.3892,
+      "step": 5826
+    },
+    {
+      "epoch": 5.379183014078006,
+      "grad_norm": 3.458984375,
+      "learning_rate": 6.119412941961307e-05,
+      "loss": 1.2378,
+      "step": 5827
+    },
+    {
+      "epoch": 5.380106162012463,
+      "grad_norm": 3.7109375,
+      "learning_rate": 6.118745830553703e-05,
+      "loss": 1.2513,
+      "step": 5828
+    },
+    {
+      "epoch": 5.381029309946919,
+      "grad_norm": 3.849609375,
+      "learning_rate": 6.118078719146098e-05,
+      "loss": 1.5474,
+      "step": 5829
+    },
+    {
+      "epoch": 5.3819524578813756,
+      "grad_norm": 4.0859375,
+      "learning_rate": 6.117411607738492e-05,
+      "loss": 1.4053,
+      "step": 5830
+    },
+    {
+      "epoch": 5.382875605815832,
+      "grad_norm": 3.501953125,
+      "learning_rate": 6.116744496330887e-05,
+      "loss": 1.5984,
+      "step": 5831
+    },
+    {
+      "epoch": 5.3837987537502885,
+      "grad_norm": 2.787109375,
+      "learning_rate": 6.116077384923283e-05,
+      "loss": 1.6809,
+      "step": 5832
+    },
+    {
+      "epoch": 5.384721901684745,
+      "grad_norm": 3.927734375,
+      "learning_rate": 6.115410273515677e-05,
+      "loss": 1.4902,
+      "step": 5833
+    },
+    {
+      "epoch": 5.385645049619201,
+      "grad_norm": 3.783203125,
+      "learning_rate": 6.114743162108072e-05,
+      "loss": 1.322,
+      "step": 5834
+    },
+    {
+      "epoch": 5.386568197553658,
+      "grad_norm": 3.79296875,
+      "learning_rate": 6.114076050700466e-05,
+      "loss": 1.3506,
+      "step": 5835
+    },
+    {
+      "epoch": 5.387491345488114,
+      "grad_norm": 3.98046875,
+      "learning_rate": 6.113408939292862e-05,
+      "loss": 1.4456,
+      "step": 5836
+    },
+    {
+      "epoch": 5.388414493422571,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.112741827885257e-05,
+      "loss": 1.542,
+      "step": 5837
+    },
+    {
+      "epoch": 5.389337641357027,
+      "grad_norm": 3.283203125,
+      "learning_rate": 6.112074716477651e-05,
+      "loss": 1.5645,
+      "step": 5838
+    },
+    {
+      "epoch": 5.390260789291484,
+      "grad_norm": 3.6171875,
+      "learning_rate": 6.111407605070047e-05,
+      "loss": 1.4226,
+      "step": 5839
+    },
+    {
+      "epoch": 5.39118393722594,
+      "grad_norm": 4.1484375,
+      "learning_rate": 6.110740493662443e-05,
+      "loss": 1.3972,
+      "step": 5840
+    },
+    {
+      "epoch": 5.392107085160397,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.110073382254838e-05,
+      "loss": 1.429,
+      "step": 5841
+    },
+    {
+      "epoch": 5.393030233094853,
+      "grad_norm": 3.759765625,
+      "learning_rate": 6.109406270847232e-05,
+      "loss": 1.4861,
+      "step": 5842
+    },
+    {
+      "epoch": 5.39395338102931,
+      "grad_norm": 3.845703125,
+      "learning_rate": 6.108739159439627e-05,
+      "loss": 1.4622,
+      "step": 5843
+    },
+    {
+      "epoch": 5.394876528963766,
+      "grad_norm": 3.63671875,
+      "learning_rate": 6.108072048032023e-05,
+      "loss": 1.4468,
+      "step": 5844
+    },
+    {
+      "epoch": 5.395799676898223,
+      "grad_norm": 4.23828125,
+      "learning_rate": 6.107404936624417e-05,
+      "loss": 1.3958,
+      "step": 5845
+    },
+    {
+      "epoch": 5.396722824832679,
+      "grad_norm": 3.689453125,
+      "learning_rate": 6.106737825216812e-05,
+      "loss": 1.436,
+      "step": 5846
+    },
+    {
+      "epoch": 5.397645972767136,
+      "grad_norm": 3.9296875,
+      "learning_rate": 6.106070713809206e-05,
+      "loss": 1.6323,
+      "step": 5847
+    },
+    {
+      "epoch": 5.398569120701593,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.105403602401602e-05,
+      "loss": 1.3801,
+      "step": 5848
+    },
+    {
+      "epoch": 5.399492268636049,
+      "grad_norm": 4.609375,
+      "learning_rate": 6.104736490993997e-05,
+      "loss": 0.9199,
+      "step": 5849
+    },
+    {
+      "epoch": 5.400415416570506,
+      "grad_norm": 4.33984375,
+      "learning_rate": 6.104069379586391e-05,
+      "loss": 1.3835,
+      "step": 5850
+    },
+    {
+      "epoch": 5.401338564504962,
+      "grad_norm": 3.923828125,
+      "learning_rate": 6.1034022681787863e-05,
+      "loss": 1.5579,
+      "step": 5851
+    },
+    {
+      "epoch": 5.402261712439419,
+      "grad_norm": 3.091796875,
+      "learning_rate": 6.102735156771181e-05,
+      "loss": 1.4365,
+      "step": 5852
+    },
+    {
+      "epoch": 5.403184860373875,
+      "grad_norm": 4.1171875,
+      "learning_rate": 6.102068045363576e-05,
+      "loss": 1.3817,
+      "step": 5853
+    },
+    {
+      "epoch": 5.404108008308332,
+      "grad_norm": 4.5078125,
+      "learning_rate": 6.1014009339559707e-05,
+      "loss": 1.3928,
+      "step": 5854
+    },
+    {
+      "epoch": 5.405031156242788,
+      "grad_norm": 3.7578125,
+      "learning_rate": 6.100733822548366e-05,
+      "loss": 0.9588,
+      "step": 5855
+    },
+    {
+      "epoch": 5.405954304177245,
+      "grad_norm": 3.619140625,
+      "learning_rate": 6.1000667111407604e-05,
+      "loss": 1.0975,
+      "step": 5856
+    },
+    {
+      "epoch": 5.406877452111701,
+      "grad_norm": 3.90234375,
+      "learning_rate": 6.0993995997331556e-05,
+      "loss": 1.552,
+      "step": 5857
+    },
+    {
+      "epoch": 5.407800600046158,
+      "grad_norm": 3.1640625,
+      "learning_rate": 6.09873248832555e-05,
+      "loss": 1.3397,
+      "step": 5858
+    },
+    {
+      "epoch": 5.408723747980614,
+      "grad_norm": 3.544921875,
+      "learning_rate": 6.0980653769179454e-05,
+      "loss": 1.3884,
+      "step": 5859
+    },
+    {
+      "epoch": 5.4096468959150705,
+      "grad_norm": 3.302734375,
+      "learning_rate": 6.09739826551034e-05,
+      "loss": 1.2477,
+      "step": 5860
+    },
+    {
+      "epoch": 5.4105700438495266,
+      "grad_norm": 4.7421875,
+      "learning_rate": 6.096731154102735e-05,
+      "loss": 1.5254,
+      "step": 5861
+    },
+    {
+      "epoch": 5.4114931917839835,
+      "grad_norm": 3.294921875,
+      "learning_rate": 6.09606404269513e-05,
+      "loss": 1.1887,
+      "step": 5862
+    },
+    {
+      "epoch": 5.4124163397184395,
+      "grad_norm": 3.001953125,
+      "learning_rate": 6.095396931287525e-05,
+      "loss": 1.1503,
+      "step": 5863
+    },
+    {
+      "epoch": 5.413339487652896,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.094729819879921e-05,
+      "loss": 1.3232,
+      "step": 5864
+    },
+    {
+      "epoch": 5.414262635587352,
+      "grad_norm": 3.873046875,
+      "learning_rate": 6.0940627084723154e-05,
+      "loss": 1.3694,
+      "step": 5865
+    },
+    {
+      "epoch": 5.415185783521809,
+      "grad_norm": 3.4765625,
+      "learning_rate": 6.0933955970647106e-05,
+      "loss": 1.0951,
+      "step": 5866
+    },
+    {
+      "epoch": 5.416108931456266,
+      "grad_norm": 3.32421875,
+      "learning_rate": 6.092728485657105e-05,
+      "loss": 1.2856,
+      "step": 5867
+    },
+    {
+      "epoch": 5.417032079390722,
+      "grad_norm": 3.619140625,
+      "learning_rate": 6.0920613742495004e-05,
+      "loss": 1.3613,
+      "step": 5868
+    },
+    {
+      "epoch": 5.417955227325179,
+      "grad_norm": 3.607421875,
+      "learning_rate": 6.091394262841895e-05,
+      "loss": 1.4075,
+      "step": 5869
+    },
+    {
+      "epoch": 5.418878375259635,
+      "grad_norm": 3.712890625,
+      "learning_rate": 6.09072715143429e-05,
+      "loss": 1.521,
+      "step": 5870
+    },
+    {
+      "epoch": 5.419801523194092,
+      "grad_norm": 4.06640625,
+      "learning_rate": 6.090060040026685e-05,
+      "loss": 1.3813,
+      "step": 5871
+    },
+    {
+      "epoch": 5.420724671128548,
+      "grad_norm": 3.94140625,
+      "learning_rate": 6.08939292861908e-05,
+      "loss": 1.2863,
+      "step": 5872
+    },
+    {
+      "epoch": 5.421647819063005,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.0887258172114744e-05,
+      "loss": 1.4481,
+      "step": 5873
+    },
+    {
+      "epoch": 5.422570966997461,
+      "grad_norm": 3.6484375,
+      "learning_rate": 6.0880587058038696e-05,
+      "loss": 1.0432,
+      "step": 5874
+    },
+    {
+      "epoch": 5.423494114931918,
+      "grad_norm": 4.0234375,
+      "learning_rate": 6.087391594396264e-05,
+      "loss": 1.2932,
+      "step": 5875
+    },
+    {
+      "epoch": 5.424417262866374,
+      "grad_norm": 4.7109375,
+      "learning_rate": 6.0867244829886594e-05,
+      "loss": 1.7419,
+      "step": 5876
+    },
+    {
+      "epoch": 5.425340410800831,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.086057371581054e-05,
+      "loss": 1.1597,
+      "step": 5877
+    },
+    {
+      "epoch": 5.426263558735287,
+      "grad_norm": 4.3515625,
+      "learning_rate": 6.085390260173449e-05,
+      "loss": 0.9126,
+      "step": 5878
+    },
+    {
+      "epoch": 5.427186706669744,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.084723148765844e-05,
+      "loss": 1.3546,
+      "step": 5879
+    },
+    {
+      "epoch": 5.4281098546042,
+      "grad_norm": 3.42578125,
+      "learning_rate": 6.084056037358239e-05,
+      "loss": 1.2722,
+      "step": 5880
+    },
+    {
+      "epoch": 5.429033002538657,
+      "grad_norm": 3.462890625,
+      "learning_rate": 6.0833889259506335e-05,
+      "loss": 1.4517,
+      "step": 5881
+    },
+    {
+      "epoch": 5.429956150473114,
+      "grad_norm": 3.8671875,
+      "learning_rate": 6.082721814543029e-05,
+      "loss": 1.4001,
+      "step": 5882
+    },
+    {
+      "epoch": 5.43087929840757,
+      "grad_norm": 3.197265625,
+      "learning_rate": 6.082054703135423e-05,
+      "loss": 1.3507,
+      "step": 5883
+    },
+    {
+      "epoch": 5.431802446342027,
+      "grad_norm": 3.884765625,
+      "learning_rate": 6.0813875917278185e-05,
+      "loss": 1.5232,
+      "step": 5884
+    },
+    {
+      "epoch": 5.432725594276483,
+      "grad_norm": 3.7109375,
+      "learning_rate": 6.080720480320214e-05,
+      "loss": 1.5359,
+      "step": 5885
+    },
+    {
+      "epoch": 5.43364874221094,
+      "grad_norm": 3.66015625,
+      "learning_rate": 6.080053368912608e-05,
+      "loss": 1.108,
+      "step": 5886
+    },
+    {
+      "epoch": 5.434571890145396,
+      "grad_norm": 3.646484375,
+      "learning_rate": 6.0793862575050034e-05,
+      "loss": 1.4636,
+      "step": 5887
+    },
+    {
+      "epoch": 5.435495038079853,
+      "grad_norm": 3.904296875,
+      "learning_rate": 6.078719146097398e-05,
+      "loss": 1.2434,
+      "step": 5888
+    },
+    {
+      "epoch": 5.436418186014309,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.078052034689794e-05,
+      "loss": 1.4729,
+      "step": 5889
+    },
+    {
+      "epoch": 5.4373413339487655,
+      "grad_norm": 3.64453125,
+      "learning_rate": 6.077384923282189e-05,
+      "loss": 1.2478,
+      "step": 5890
+    },
+    {
+      "epoch": 5.4382644818832215,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.0767178118745837e-05,
+      "loss": 1.1958,
+      "step": 5891
+    },
+    {
+      "epoch": 5.4391876298176784,
+      "grad_norm": 3.17578125,
+      "learning_rate": 6.076050700466979e-05,
+      "loss": 1.0701,
+      "step": 5892
+    },
+    {
+      "epoch": 5.4401107777521345,
+      "grad_norm": 4.08984375,
+      "learning_rate": 6.0753835890593734e-05,
+      "loss": 1.4012,
+      "step": 5893
+    },
+    {
+      "epoch": 5.441033925686591,
+      "grad_norm": 4.3203125,
+      "learning_rate": 6.0747164776517686e-05,
+      "loss": 1.5984,
+      "step": 5894
+    },
+    {
+      "epoch": 5.441957073621047,
+      "grad_norm": 3.689453125,
+      "learning_rate": 6.074049366244163e-05,
+      "loss": 1.4421,
+      "step": 5895
+    },
+    {
+      "epoch": 5.442880221555504,
+      "grad_norm": 3.31640625,
+      "learning_rate": 6.0733822548365584e-05,
+      "loss": 1.4614,
+      "step": 5896
+    },
+    {
+      "epoch": 5.44380336948996,
+      "grad_norm": 3.56640625,
+      "learning_rate": 6.072715143428953e-05,
+      "loss": 1.5305,
+      "step": 5897
+    },
+    {
+      "epoch": 5.444726517424417,
+      "grad_norm": 3.064453125,
+      "learning_rate": 6.072048032021348e-05,
+      "loss": 1.6067,
+      "step": 5898
+    },
+    {
+      "epoch": 5.445649665358873,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.071380920613743e-05,
+      "loss": 1.4375,
+      "step": 5899
+    },
+    {
+      "epoch": 5.44657281329333,
+      "grad_norm": 4.484375,
+      "learning_rate": 6.070713809206138e-05,
+      "loss": 1.5261,
+      "step": 5900
+    },
+    {
+      "epoch": 5.447495961227787,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.0700466977985325e-05,
+      "loss": 1.4077,
+      "step": 5901
+    },
+    {
+      "epoch": 5.448419109162243,
+      "grad_norm": 3.740234375,
+      "learning_rate": 6.069379586390928e-05,
+      "loss": 1.2838,
+      "step": 5902
+    },
+    {
+      "epoch": 5.4493422570967,
+      "grad_norm": 4.15234375,
+      "learning_rate": 6.068712474983322e-05,
+      "loss": 1.2793,
+      "step": 5903
+    },
+    {
+      "epoch": 5.450265405031156,
+      "grad_norm": 3.310546875,
+      "learning_rate": 6.0680453635757175e-05,
+      "loss": 1.3933,
+      "step": 5904
+    },
+    {
+      "epoch": 5.451188552965613,
+      "grad_norm": 3.599609375,
+      "learning_rate": 6.067378252168112e-05,
+      "loss": 1.3354,
+      "step": 5905
+    },
+    {
+      "epoch": 5.452111700900069,
+      "grad_norm": 3.216796875,
+      "learning_rate": 6.066711140760507e-05,
+      "loss": 1.3186,
+      "step": 5906
+    },
+    {
+      "epoch": 5.453034848834526,
+      "grad_norm": 3.236328125,
+      "learning_rate": 6.066044029352902e-05,
+      "loss": 1.342,
+      "step": 5907
+    },
+    {
+      "epoch": 5.453957996768982,
+      "grad_norm": 4.29296875,
+      "learning_rate": 6.065376917945297e-05,
+      "loss": 1.2997,
+      "step": 5908
+    },
+    {
+      "epoch": 5.454881144703439,
+      "grad_norm": 4.06640625,
+      "learning_rate": 6.0647098065376915e-05,
+      "loss": 1.2571,
+      "step": 5909
+    },
+    {
+      "epoch": 5.455804292637895,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.064042695130087e-05,
+      "loss": 1.3257,
+      "step": 5910
+    },
+    {
+      "epoch": 5.456727440572352,
+      "grad_norm": 3.630859375,
+      "learning_rate": 6.063375583722481e-05,
+      "loss": 1.5005,
+      "step": 5911
+    },
+    {
+      "epoch": 5.457650588506808,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.0627084723148765e-05,
+      "loss": 1.7056,
+      "step": 5912
+    },
+    {
+      "epoch": 5.458573736441265,
+      "grad_norm": 3.80078125,
+      "learning_rate": 6.062041360907271e-05,
+      "loss": 1.5021,
+      "step": 5913
+    },
+    {
+      "epoch": 5.459496884375721,
+      "grad_norm": 3.083984375,
+      "learning_rate": 6.061374249499667e-05,
+      "loss": 0.8711,
+      "step": 5914
+    },
+    {
+      "epoch": 5.460420032310178,
+      "grad_norm": 3.916015625,
+      "learning_rate": 6.060707138092062e-05,
+      "loss": 1.5378,
+      "step": 5915
+    },
+    {
+      "epoch": 5.461343180244635,
+      "grad_norm": 4.28515625,
+      "learning_rate": 6.060040026684457e-05,
+      "loss": 1.5498,
+      "step": 5916
+    },
+    {
+      "epoch": 5.462266328179091,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.059372915276852e-05,
+      "loss": 1.4314,
+      "step": 5917
+    },
+    {
+      "epoch": 5.4631894761135475,
+      "grad_norm": 3.369140625,
+      "learning_rate": 6.0587058038692465e-05,
+      "loss": 1.3193,
+      "step": 5918
+    },
+    {
+      "epoch": 5.464112624048004,
+      "grad_norm": 3.646484375,
+      "learning_rate": 6.058038692461642e-05,
+      "loss": 1.2344,
+      "step": 5919
+    },
+    {
+      "epoch": 5.4650357719824605,
+      "grad_norm": 3.81640625,
+      "learning_rate": 6.057371581054036e-05,
+      "loss": 1.6267,
+      "step": 5920
+    },
+    {
+      "epoch": 5.4659589199169165,
+      "grad_norm": 3.5546875,
+      "learning_rate": 6.0567044696464315e-05,
+      "loss": 1.3091,
+      "step": 5921
+    },
+    {
+      "epoch": 5.466882067851373,
+      "grad_norm": 2.986328125,
+      "learning_rate": 6.056037358238826e-05,
+      "loss": 1.5669,
+      "step": 5922
+    },
+    {
+      "epoch": 5.4678052157858295,
+      "grad_norm": 4.12109375,
+      "learning_rate": 6.055370246831221e-05,
+      "loss": 1.3123,
+      "step": 5923
+    },
+    {
+      "epoch": 5.468728363720286,
+      "grad_norm": 3.037109375,
+      "learning_rate": 6.0547031354236165e-05,
+      "loss": 1.3591,
+      "step": 5924
+    },
+    {
+      "epoch": 5.469651511654742,
+      "grad_norm": 3.8671875,
+      "learning_rate": 6.054036024016011e-05,
+      "loss": 1.3687,
+      "step": 5925
+    },
+    {
+      "epoch": 5.470574659589199,
+      "grad_norm": 3.7265625,
+      "learning_rate": 6.053368912608406e-05,
+      "loss": 1.3855,
+      "step": 5926
+    },
+    {
+      "epoch": 5.471497807523655,
+      "grad_norm": 3.853515625,
+      "learning_rate": 6.052701801200801e-05,
+      "loss": 1.7095,
+      "step": 5927
+    },
+    {
+      "epoch": 5.472420955458112,
+      "grad_norm": 5.10546875,
+      "learning_rate": 6.052034689793196e-05,
+      "loss": 1.2994,
+      "step": 5928
+    },
+    {
+      "epoch": 5.473344103392568,
+      "grad_norm": 3.720703125,
+      "learning_rate": 6.0513675783855905e-05,
+      "loss": 1.114,
+      "step": 5929
+    },
+    {
+      "epoch": 5.474267251327025,
+      "grad_norm": 3.873046875,
+      "learning_rate": 6.050700466977986e-05,
+      "loss": 1.4663,
+      "step": 5930
+    },
+    {
+      "epoch": 5.475190399261482,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.05003335557038e-05,
+      "loss": 1.313,
+      "step": 5931
+    },
+    {
+      "epoch": 5.476113547195938,
+      "grad_norm": 4.12890625,
+      "learning_rate": 6.0493662441627755e-05,
+      "loss": 1.3923,
+      "step": 5932
+    },
+    {
+      "epoch": 5.477036695130395,
+      "grad_norm": 4.9296875,
+      "learning_rate": 6.04869913275517e-05,
+      "loss": 1.3628,
+      "step": 5933
+    },
+    {
+      "epoch": 5.477959843064851,
+      "grad_norm": 4.27734375,
+      "learning_rate": 6.048032021347565e-05,
+      "loss": 1.4556,
+      "step": 5934
+    },
+    {
+      "epoch": 5.478882990999308,
+      "grad_norm": 3.607421875,
+      "learning_rate": 6.04736490993996e-05,
+      "loss": 1.3328,
+      "step": 5935
+    },
+    {
+      "epoch": 5.479806138933764,
+      "grad_norm": 3.54296875,
+      "learning_rate": 6.046697798532355e-05,
+      "loss": 1.2527,
+      "step": 5936
+    },
+    {
+      "epoch": 5.480729286868221,
+      "grad_norm": 3.44140625,
+      "learning_rate": 6.0460306871247496e-05,
+      "loss": 1.3933,
+      "step": 5937
+    },
+    {
+      "epoch": 5.481652434802677,
+      "grad_norm": 3.896484375,
+      "learning_rate": 6.045363575717145e-05,
+      "loss": 1.5649,
+      "step": 5938
+    },
+    {
+      "epoch": 5.482575582737134,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.044696464309541e-05,
+      "loss": 1.6653,
+      "step": 5939
+    },
+    {
+      "epoch": 5.48349873067159,
+      "grad_norm": 3.677734375,
+      "learning_rate": 6.044029352901935e-05,
+      "loss": 1.2534,
+      "step": 5940
+    },
+    {
+      "epoch": 5.484421878606047,
+      "grad_norm": 4.07421875,
+      "learning_rate": 6.0433622414943305e-05,
+      "loss": 1.0222,
+      "step": 5941
+    },
+    {
+      "epoch": 5.485345026540503,
+      "grad_norm": 3.625,
+      "learning_rate": 6.042695130086725e-05,
+      "loss": 1.4021,
+      "step": 5942
+    },
+    {
+      "epoch": 5.48626817447496,
+      "grad_norm": 3.744140625,
+      "learning_rate": 6.04202801867912e-05,
+      "loss": 1.3689,
+      "step": 5943
+    },
+    {
+      "epoch": 5.487191322409416,
+      "grad_norm": 3.33984375,
+      "learning_rate": 6.041360907271515e-05,
+      "loss": 1.5012,
+      "step": 5944
+    },
+    {
+      "epoch": 5.488114470343873,
+      "grad_norm": 3.77734375,
+      "learning_rate": 6.04069379586391e-05,
+      "loss": 1.4614,
+      "step": 5945
+    },
+    {
+      "epoch": 5.489037618278329,
+      "grad_norm": 4.546875,
+      "learning_rate": 6.0400266844563045e-05,
+      "loss": 1.4243,
+      "step": 5946
+    },
+    {
+      "epoch": 5.489960766212786,
+      "grad_norm": 3.70703125,
+      "learning_rate": 6.0393595730487e-05,
+      "loss": 1.4109,
+      "step": 5947
+    },
+    {
+      "epoch": 5.490883914147242,
+      "grad_norm": 3.72265625,
+      "learning_rate": 6.038692461641094e-05,
+      "loss": 1.1722,
+      "step": 5948
+    },
+    {
+      "epoch": 5.4918070620816986,
+      "grad_norm": 3.322265625,
+      "learning_rate": 6.0380253502334895e-05,
+      "loss": 1.3123,
+      "step": 5949
+    },
+    {
+      "epoch": 5.4927302100161555,
+      "grad_norm": 3.5703125,
+      "learning_rate": 6.037358238825884e-05,
+      "loss": 1.5071,
+      "step": 5950
+    },
+    {
+      "epoch": 5.4936533579506115,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.036691127418279e-05,
+      "loss": 1.4326,
+      "step": 5951
+    },
+    {
+      "epoch": 5.494576505885068,
+      "grad_norm": 3.6015625,
+      "learning_rate": 6.036024016010674e-05,
+      "loss": 1.2939,
+      "step": 5952
+    },
+    {
+      "epoch": 5.495499653819524,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.035356904603069e-05,
+      "loss": 1.5547,
+      "step": 5953
+    },
+    {
+      "epoch": 5.496422801753981,
+      "grad_norm": 3.8984375,
+      "learning_rate": 6.0346897931954636e-05,
+      "loss": 1.2825,
+      "step": 5954
+    },
+    {
+      "epoch": 5.497345949688437,
+      "grad_norm": 3.814453125,
+      "learning_rate": 6.034022681787859e-05,
+      "loss": 1.5889,
+      "step": 5955
+    },
+    {
+      "epoch": 5.498269097622894,
+      "grad_norm": 3.974609375,
+      "learning_rate": 6.0333555703802533e-05,
+      "loss": 1.1237,
+      "step": 5956
+    },
+    {
+      "epoch": 5.49919224555735,
+      "grad_norm": 4.01171875,
+      "learning_rate": 6.0326884589726486e-05,
+      "loss": 1.4141,
+      "step": 5957
+    },
+    {
+      "epoch": 5.500115393491807,
+      "grad_norm": 3.986328125,
+      "learning_rate": 6.032021347565043e-05,
+      "loss": 1.4478,
+      "step": 5958
+    },
+    {
+      "epoch": 5.501038541426263,
+      "grad_norm": 3.630859375,
+      "learning_rate": 6.031354236157438e-05,
+      "loss": 1.4045,
+      "step": 5959
+    },
+    {
+      "epoch": 5.50196168936072,
+      "grad_norm": 3.59765625,
+      "learning_rate": 6.030687124749833e-05,
+      "loss": 1.0616,
+      "step": 5960
+    },
+    {
+      "epoch": 5.502884837295176,
+      "grad_norm": 4.05859375,
+      "learning_rate": 6.030020013342228e-05,
+      "loss": 1.5166,
+      "step": 5961
+    },
+    {
+      "epoch": 5.503807985229633,
+      "grad_norm": 3.763671875,
+      "learning_rate": 6.0293529019346226e-05,
+      "loss": 1.0962,
+      "step": 5962
+    },
+    {
+      "epoch": 5.504731133164089,
+      "grad_norm": 4.19140625,
+      "learning_rate": 6.028685790527018e-05,
+      "loss": 1.4485,
+      "step": 5963
+    },
+    {
+      "epoch": 5.505654281098546,
+      "grad_norm": 3.583984375,
+      "learning_rate": 6.028018679119414e-05,
+      "loss": 1.2627,
+      "step": 5964
+    },
+    {
+      "epoch": 5.506577429033003,
+      "grad_norm": 3.74609375,
+      "learning_rate": 6.027351567711809e-05,
+      "loss": 1.4448,
+      "step": 5965
+    },
+    {
+      "epoch": 5.507500576967459,
+      "grad_norm": 3.283203125,
+      "learning_rate": 6.0266844563042035e-05,
+      "loss": 1.4873,
+      "step": 5966
+    },
+    {
+      "epoch": 5.508423724901916,
+      "grad_norm": 3.416015625,
+      "learning_rate": 6.026017344896599e-05,
+      "loss": 1.4619,
+      "step": 5967
+    },
+    {
+      "epoch": 5.509346872836372,
+      "grad_norm": 2.955078125,
+      "learning_rate": 6.025350233488993e-05,
+      "loss": 1.6448,
+      "step": 5968
+    },
+    {
+      "epoch": 5.510270020770829,
+      "grad_norm": 3.341796875,
+      "learning_rate": 6.0246831220813885e-05,
+      "loss": 1.4263,
+      "step": 5969
+    },
+    {
+      "epoch": 5.511193168705285,
+      "grad_norm": 4.4453125,
+      "learning_rate": 6.024016010673783e-05,
+      "loss": 1.0507,
+      "step": 5970
+    },
+    {
+      "epoch": 5.512116316639742,
+      "grad_norm": 3.58984375,
+      "learning_rate": 6.023348899266178e-05,
+      "loss": 1.405,
+      "step": 5971
+    },
+    {
+      "epoch": 5.513039464574198,
+      "grad_norm": 3.880859375,
+      "learning_rate": 6.022681787858573e-05,
+      "loss": 1.5117,
+      "step": 5972
+    },
+    {
+      "epoch": 5.513962612508655,
+      "grad_norm": 4.484375,
+      "learning_rate": 6.022014676450968e-05,
+      "loss": 1.3367,
+      "step": 5973
+    },
+    {
+      "epoch": 5.514885760443111,
+      "grad_norm": 3.60546875,
+      "learning_rate": 6.0213475650433626e-05,
+      "loss": 0.8901,
+      "step": 5974
+    },
+    {
+      "epoch": 5.515808908377568,
+      "grad_norm": 4.7421875,
+      "learning_rate": 6.020680453635758e-05,
+      "loss": 1.6938,
+      "step": 5975
+    },
+    {
+      "epoch": 5.516732056312024,
+      "grad_norm": 4.16015625,
+      "learning_rate": 6.0200133422281523e-05,
+      "loss": 1.1868,
+      "step": 5976
+    },
+    {
+      "epoch": 5.517655204246481,
+      "grad_norm": 5.02734375,
+      "learning_rate": 6.0193462308205476e-05,
+      "loss": 1.291,
+      "step": 5977
+    },
+    {
+      "epoch": 5.518578352180937,
+      "grad_norm": 2.919921875,
+      "learning_rate": 6.018679119412942e-05,
+      "loss": 1.1462,
+      "step": 5978
+    },
+    {
+      "epoch": 5.5195015001153935,
+      "grad_norm": 5.41015625,
+      "learning_rate": 6.018012008005337e-05,
+      "loss": 1.4003,
+      "step": 5979
+    },
+    {
+      "epoch": 5.5204246480498504,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.017344896597732e-05,
+      "loss": 1.4968,
+      "step": 5980
+    },
+    {
+      "epoch": 5.5213477959843065,
+      "grad_norm": 3.66796875,
+      "learning_rate": 6.016677785190127e-05,
+      "loss": 1.3882,
+      "step": 5981
+    },
+    {
+      "epoch": 5.5222709439187625,
+      "grad_norm": 3.4140625,
+      "learning_rate": 6.0160106737825216e-05,
+      "loss": 1.4299,
+      "step": 5982
+    },
+    {
+      "epoch": 5.523194091853219,
+      "grad_norm": 4.1171875,
+      "learning_rate": 6.015343562374917e-05,
+      "loss": 1.407,
+      "step": 5983
+    },
+    {
+      "epoch": 5.524117239787676,
+      "grad_norm": 3.287109375,
+      "learning_rate": 6.0146764509673114e-05,
+      "loss": 1.3696,
+      "step": 5984
+    },
+    {
+      "epoch": 5.525040387722132,
+      "grad_norm": 4.08984375,
+      "learning_rate": 6.0140093395597066e-05,
+      "loss": 1.4724,
+      "step": 5985
+    },
+    {
+      "epoch": 5.525963535656589,
+      "grad_norm": 4.41796875,
+      "learning_rate": 6.013342228152101e-05,
+      "loss": 1.2505,
+      "step": 5986
+    },
+    {
+      "epoch": 5.526886683591045,
+      "grad_norm": 3.41015625,
+      "learning_rate": 6.0126751167444964e-05,
+      "loss": 1.5339,
+      "step": 5987
+    },
+    {
+      "epoch": 5.527809831525502,
+      "grad_norm": 4.078125,
+      "learning_rate": 6.012008005336891e-05,
+      "loss": 1.2532,
+      "step": 5988
+    },
+    {
+      "epoch": 5.528732979459958,
+      "grad_norm": 4.03515625,
+      "learning_rate": 6.011340893929287e-05,
+      "loss": 1.6077,
+      "step": 5989
+    },
+    {
+      "epoch": 5.529656127394415,
+      "grad_norm": 3.62109375,
+      "learning_rate": 6.010673782521682e-05,
+      "loss": 1.3058,
+      "step": 5990
+    },
+    {
+      "epoch": 5.530579275328871,
+      "grad_norm": 4.4140625,
+      "learning_rate": 6.0100066711140766e-05,
+      "loss": 1.5334,
+      "step": 5991
+    },
+    {
+      "epoch": 5.531502423263328,
+      "grad_norm": 3.697265625,
+      "learning_rate": 6.009339559706472e-05,
+      "loss": 1.6221,
+      "step": 5992
+    },
+    {
+      "epoch": 5.532425571197784,
+      "grad_norm": 4.87109375,
+      "learning_rate": 6.0086724482988663e-05,
+      "loss": 1.6829,
+      "step": 5993
+    },
+    {
+      "epoch": 5.533348719132241,
+      "grad_norm": 4.6484375,
+      "learning_rate": 6.0080053368912616e-05,
+      "loss": 0.8883,
+      "step": 5994
+    },
+    {
+      "epoch": 5.534271867066698,
+      "grad_norm": 3.142578125,
+      "learning_rate": 6.007338225483656e-05,
+      "loss": 1.0578,
+      "step": 5995
+    },
+    {
+      "epoch": 5.535195015001154,
+      "grad_norm": 3.83984375,
+      "learning_rate": 6.006671114076051e-05,
+      "loss": 1.4629,
+      "step": 5996
+    },
+    {
+      "epoch": 5.53611816293561,
+      "grad_norm": 3.51953125,
+      "learning_rate": 6.006004002668446e-05,
+      "loss": 1.4177,
+      "step": 5997
+    },
+    {
+      "epoch": 5.537041310870067,
+      "grad_norm": 4.07421875,
+      "learning_rate": 6.005336891260841e-05,
+      "loss": 1.4241,
+      "step": 5998
+    },
+    {
+      "epoch": 5.537964458804524,
+      "grad_norm": 3.861328125,
+      "learning_rate": 6.0046697798532356e-05,
+      "loss": 1.4246,
+      "step": 5999
+    },
+    {
+      "epoch": 5.53888760673898,
+      "grad_norm": 4.3828125,
+      "learning_rate": 6.004002668445631e-05,
+      "loss": 1.249,
+      "step": 6000
+    },
+    {
+      "epoch": 5.53888760673898,
+      "eval_loss": 0.8779296875,
+      "eval_runtime": 23.3579,
+      "eval_samples_per_second": 976.158,
+      "eval_steps_per_second": 61.05,
+      "step": 6000
+    },
+    {
+      "epoch": 5.539810754673437,
+      "grad_norm": 4.28125,
+      "learning_rate": 6.0033355570380254e-05,
+      "loss": 1.23,
+      "step": 6001
+    },
+    {
+      "epoch": 5.540733902607893,
+      "grad_norm": 3.755859375,
+      "learning_rate": 6.0026684456304206e-05,
+      "loss": 1.4207,
+      "step": 6002
+    },
+    {
+      "epoch": 5.54165705054235,
+      "grad_norm": 3.654296875,
+      "learning_rate": 6.002001334222815e-05,
+      "loss": 1.5137,
+      "step": 6003
+    },
+    {
+      "epoch": 5.542580198476806,
+      "grad_norm": 5.21484375,
+      "learning_rate": 6.0013342228152104e-05,
+      "loss": 1.2513,
+      "step": 6004
+    },
+    {
+      "epoch": 5.543503346411263,
+      "grad_norm": 3.470703125,
+      "learning_rate": 6.000667111407605e-05,
+      "loss": 1.4014,
+      "step": 6005
+    },
+    {
+      "epoch": 5.544426494345719,
+      "grad_norm": 3.140625,
+      "learning_rate": 6e-05,
+      "loss": 1.3884,
+      "step": 6006
+    },
+    {
+      "epoch": 5.545349642280176,
+      "grad_norm": 4.1484375,
+      "learning_rate": 5.999332888592395e-05,
+      "loss": 1.4412,
+      "step": 6007
+    },
+    {
+      "epoch": 5.546272790214632,
+      "grad_norm": 3.26953125,
+      "learning_rate": 5.99866577718479e-05,
+      "loss": 1.5669,
+      "step": 6008
+    },
+    {
+      "epoch": 5.5471959381490885,
+      "grad_norm": 3.54296875,
+      "learning_rate": 5.9979986657771845e-05,
+      "loss": 1.4343,
+      "step": 6009
+    },
+    {
+      "epoch": 5.5481190860835445,
+      "grad_norm": 3.568359375,
+      "learning_rate": 5.99733155436958e-05,
+      "loss": 1.4822,
+      "step": 6010
+    },
+    {
+      "epoch": 5.5490422340180015,
+      "grad_norm": 4.5234375,
+      "learning_rate": 5.996664442961974e-05,
+      "loss": 1.5674,
+      "step": 6011
+    },
+    {
+      "epoch": 5.5499653819524575,
+      "grad_norm": 4.49609375,
+      "learning_rate": 5.9959973315543694e-05,
+      "loss": 1.395,
+      "step": 6012
+    },
+    {
+      "epoch": 5.550888529886914,
+      "grad_norm": 3.921875,
+      "learning_rate": 5.995330220146764e-05,
+      "loss": 1.2556,
+      "step": 6013
+    },
+    {
+      "epoch": 5.551811677821371,
+      "grad_norm": 3.94921875,
+      "learning_rate": 5.9946631087391606e-05,
+      "loss": 1.2736,
+      "step": 6014
+    },
+    {
+      "epoch": 5.552734825755827,
+      "grad_norm": 3.748046875,
+      "learning_rate": 5.993995997331555e-05,
+      "loss": 1.374,
+      "step": 6015
+    },
+    {
+      "epoch": 5.553657973690284,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.99332888592395e-05,
+      "loss": 1.5579,
+      "step": 6016
+    },
+    {
+      "epoch": 5.55458112162474,
+      "grad_norm": 3.80078125,
+      "learning_rate": 5.992661774516345e-05,
+      "loss": 1.3259,
+      "step": 6017
+    },
+    {
+      "epoch": 5.555504269559197,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.99199466310874e-05,
+      "loss": 1.287,
+      "step": 6018
+    },
+    {
+      "epoch": 5.556427417493653,
+      "grad_norm": 3.783203125,
+      "learning_rate": 5.9913275517011346e-05,
+      "loss": 1.2178,
+      "step": 6019
+    },
+    {
+      "epoch": 5.55735056542811,
+      "grad_norm": 3.607421875,
+      "learning_rate": 5.99066044029353e-05,
+      "loss": 1.4624,
+      "step": 6020
+    },
+    {
+      "epoch": 5.558273713362566,
+      "grad_norm": 4.62109375,
+      "learning_rate": 5.9899933288859244e-05,
+      "loss": 1.3679,
+      "step": 6021
+    },
+    {
+      "epoch": 5.559196861297023,
+      "grad_norm": 3.552734375,
+      "learning_rate": 5.9893262174783196e-05,
+      "loss": 1.3667,
+      "step": 6022
+    },
+    {
+      "epoch": 5.560120009231479,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.988659106070714e-05,
+      "loss": 1.4246,
+      "step": 6023
+    },
+    {
+      "epoch": 5.561043157165936,
+      "grad_norm": 3.720703125,
+      "learning_rate": 5.9879919946631094e-05,
+      "loss": 1.2239,
+      "step": 6024
+    },
+    {
+      "epoch": 5.561966305100392,
+      "grad_norm": 4.0703125,
+      "learning_rate": 5.987324883255504e-05,
+      "loss": 1.0938,
+      "step": 6025
+    },
+    {
+      "epoch": 5.562889453034849,
+      "grad_norm": 3.1171875,
+      "learning_rate": 5.986657771847899e-05,
+      "loss": 1.296,
+      "step": 6026
+    },
+    {
+      "epoch": 5.563812600969305,
+      "grad_norm": 3.31640625,
+      "learning_rate": 5.985990660440294e-05,
+      "loss": 1.4409,
+      "step": 6027
+    },
+    {
+      "epoch": 5.564735748903762,
+      "grad_norm": 3.62890625,
+      "learning_rate": 5.985323549032689e-05,
+      "loss": 1.2273,
+      "step": 6028
+    },
+    {
+      "epoch": 5.565658896838219,
+      "grad_norm": 4.203125,
+      "learning_rate": 5.9846564376250835e-05,
+      "loss": 1.5134,
+      "step": 6029
+    },
+    {
+      "epoch": 5.566582044772675,
+      "grad_norm": 3.349609375,
+      "learning_rate": 5.983989326217479e-05,
+      "loss": 1.4282,
+      "step": 6030
+    },
+    {
+      "epoch": 5.567505192707131,
+      "grad_norm": 3.513671875,
+      "learning_rate": 5.983322214809873e-05,
+      "loss": 1.2615,
+      "step": 6031
+    },
+    {
+      "epoch": 5.568428340641588,
+      "grad_norm": 3.52734375,
+      "learning_rate": 5.9826551034022684e-05,
+      "loss": 1.4392,
+      "step": 6032
+    },
+    {
+      "epoch": 5.569351488576045,
+      "grad_norm": 3.62890625,
+      "learning_rate": 5.981987991994663e-05,
+      "loss": 1.3209,
+      "step": 6033
+    },
+    {
+      "epoch": 5.570274636510501,
+      "grad_norm": 3.486328125,
+      "learning_rate": 5.981320880587058e-05,
+      "loss": 1.3923,
+      "step": 6034
+    },
+    {
+      "epoch": 5.571197784444958,
+      "grad_norm": 4.01171875,
+      "learning_rate": 5.980653769179453e-05,
+      "loss": 1.4775,
+      "step": 6035
+    },
+    {
+      "epoch": 5.572120932379414,
+      "grad_norm": 3.87109375,
+      "learning_rate": 5.979986657771848e-05,
+      "loss": 1.3076,
+      "step": 6036
+    },
+    {
+      "epoch": 5.5730440803138706,
+      "grad_norm": 3.349609375,
+      "learning_rate": 5.9793195463642425e-05,
+      "loss": 0.8701,
+      "step": 6037
+    },
+    {
+      "epoch": 5.573967228248327,
+      "grad_norm": 4.35546875,
+      "learning_rate": 5.978652434956638e-05,
+      "loss": 1.3307,
+      "step": 6038
+    },
+    {
+      "epoch": 5.5748903761827835,
+      "grad_norm": 3.103515625,
+      "learning_rate": 5.9779853235490336e-05,
+      "loss": 0.919,
+      "step": 6039
+    },
+    {
+      "epoch": 5.5758135241172395,
+      "grad_norm": 3.36328125,
+      "learning_rate": 5.977318212141428e-05,
+      "loss": 1.3849,
+      "step": 6040
+    },
+    {
+      "epoch": 5.576736672051696,
+      "grad_norm": 3.603515625,
+      "learning_rate": 5.9766511007338234e-05,
+      "loss": 1.3076,
+      "step": 6041
+    },
+    {
+      "epoch": 5.5776598199861525,
+      "grad_norm": 4.63671875,
+      "learning_rate": 5.975983989326218e-05,
+      "loss": 1.3159,
+      "step": 6042
+    },
+    {
+      "epoch": 5.578582967920609,
+      "grad_norm": 3.505859375,
+      "learning_rate": 5.975316877918613e-05,
+      "loss": 1.2996,
+      "step": 6043
+    },
+    {
+      "epoch": 5.579506115855065,
+      "grad_norm": 3.580078125,
+      "learning_rate": 5.974649766511008e-05,
+      "loss": 1.449,
+      "step": 6044
+    },
+    {
+      "epoch": 5.580429263789522,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.973982655103403e-05,
+      "loss": 1.3035,
+      "step": 6045
+    },
+    {
+      "epoch": 5.581352411723978,
+      "grad_norm": 4.4375,
+      "learning_rate": 5.9733155436957975e-05,
+      "loss": 1.7415,
+      "step": 6046
+    },
+    {
+      "epoch": 5.582275559658435,
+      "grad_norm": 3.189453125,
+      "learning_rate": 5.972648432288193e-05,
+      "loss": 1.4597,
+      "step": 6047
+    },
+    {
+      "epoch": 5.583198707592892,
+      "grad_norm": 3.287109375,
+      "learning_rate": 5.971981320880587e-05,
+      "loss": 1.4543,
+      "step": 6048
+    },
+    {
+      "epoch": 5.584121855527348,
+      "grad_norm": 3.619140625,
+      "learning_rate": 5.9713142094729824e-05,
+      "loss": 1.4265,
+      "step": 6049
+    },
+    {
+      "epoch": 5.585045003461805,
+      "grad_norm": 4.27734375,
+      "learning_rate": 5.970647098065377e-05,
+      "loss": 1.6997,
+      "step": 6050
+    },
+    {
+      "epoch": 5.585968151396261,
+      "grad_norm": 3.48828125,
+      "learning_rate": 5.969979986657772e-05,
+      "loss": 1.4775,
+      "step": 6051
+    },
+    {
+      "epoch": 5.586891299330718,
+      "grad_norm": 3.30859375,
+      "learning_rate": 5.969312875250167e-05,
+      "loss": 1.4146,
+      "step": 6052
+    },
+    {
+      "epoch": 5.587814447265174,
+      "grad_norm": 3.94921875,
+      "learning_rate": 5.968645763842562e-05,
+      "loss": 1.4697,
+      "step": 6053
+    },
+    {
+      "epoch": 5.588737595199631,
+      "grad_norm": 4.0390625,
+      "learning_rate": 5.9679786524349565e-05,
+      "loss": 1.3899,
+      "step": 6054
+    },
+    {
+      "epoch": 5.589660743134087,
+      "grad_norm": 3.544921875,
+      "learning_rate": 5.967311541027352e-05,
+      "loss": 1.0502,
+      "step": 6055
+    },
+    {
+      "epoch": 5.590583891068544,
+      "grad_norm": 3.69140625,
+      "learning_rate": 5.966644429619746e-05,
+      "loss": 1.4065,
+      "step": 6056
+    },
+    {
+      "epoch": 5.591507039003,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.9659773182121415e-05,
+      "loss": 1.437,
+      "step": 6057
+    },
+    {
+      "epoch": 5.592430186937457,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.965310206804536e-05,
+      "loss": 1.0945,
+      "step": 6058
+    },
+    {
+      "epoch": 5.593353334871913,
+      "grad_norm": 4.6484375,
+      "learning_rate": 5.964643095396931e-05,
+      "loss": 1.4946,
+      "step": 6059
+    },
+    {
+      "epoch": 5.59427648280637,
+      "grad_norm": 4.83203125,
+      "learning_rate": 5.963975983989326e-05,
+      "loss": 1.3247,
+      "step": 6060
+    },
+    {
+      "epoch": 5.595199630740826,
+      "grad_norm": 3.763671875,
+      "learning_rate": 5.963308872581721e-05,
+      "loss": 1.401,
+      "step": 6061
+    },
+    {
+      "epoch": 5.596122778675283,
+      "grad_norm": 3.580078125,
+      "learning_rate": 5.9626417611741156e-05,
+      "loss": 1.2563,
+      "step": 6062
+    },
+    {
+      "epoch": 5.59704592660974,
+      "grad_norm": 3.97265625,
+      "learning_rate": 5.961974649766511e-05,
+      "loss": 1.4001,
+      "step": 6063
+    },
+    {
+      "epoch": 5.597969074544196,
+      "grad_norm": 4.00390625,
+      "learning_rate": 5.961307538358907e-05,
+      "loss": 1.2495,
+      "step": 6064
+    },
+    {
+      "epoch": 5.598892222478653,
+      "grad_norm": 4.19921875,
+      "learning_rate": 5.960640426951302e-05,
+      "loss": 1.4709,
+      "step": 6065
+    },
+    {
+      "epoch": 5.599815370413109,
+      "grad_norm": 3.908203125,
+      "learning_rate": 5.9599733155436965e-05,
+      "loss": 1.3865,
+      "step": 6066
+    },
+    {
+      "epoch": 5.6007385183475655,
+      "grad_norm": 3.517578125,
+      "learning_rate": 5.959306204136092e-05,
+      "loss": 1.5156,
+      "step": 6067
+    },
+    {
+      "epoch": 5.601661666282022,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.958639092728486e-05,
+      "loss": 1.0991,
+      "step": 6068
+    },
+    {
+      "epoch": 5.6025848142164785,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.9579719813208814e-05,
+      "loss": 1.8513,
+      "step": 6069
+    },
+    {
+      "epoch": 5.6035079621509345,
+      "grad_norm": 3.431640625,
+      "learning_rate": 5.957304869913276e-05,
+      "loss": 1.5691,
+      "step": 6070
+    },
+    {
+      "epoch": 5.604431110085391,
+      "grad_norm": 4.56640625,
+      "learning_rate": 5.956637758505671e-05,
+      "loss": 1.4153,
+      "step": 6071
+    },
+    {
+      "epoch": 5.605354258019847,
+      "grad_norm": 3.634765625,
+      "learning_rate": 5.955970647098066e-05,
+      "loss": 1.4243,
+      "step": 6072
+    },
+    {
+      "epoch": 5.606277405954304,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.955303535690461e-05,
+      "loss": 1.3546,
+      "step": 6073
+    },
+    {
+      "epoch": 5.60720055388876,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.9546364242828555e-05,
+      "loss": 1.3242,
+      "step": 6074
+    },
+    {
+      "epoch": 5.608123701823217,
+      "grad_norm": 3.267578125,
+      "learning_rate": 5.953969312875251e-05,
+      "loss": 1.4587,
+      "step": 6075
+    },
+    {
+      "epoch": 5.609046849757673,
+      "grad_norm": 3.8515625,
+      "learning_rate": 5.953302201467645e-05,
+      "loss": 1.2035,
+      "step": 6076
+    },
+    {
+      "epoch": 5.60996999769213,
+      "grad_norm": 4.43359375,
+      "learning_rate": 5.9526350900600405e-05,
+      "loss": 1.5071,
+      "step": 6077
+    },
+    {
+      "epoch": 5.610893145626587,
+      "grad_norm": 3.849609375,
+      "learning_rate": 5.951967978652435e-05,
+      "loss": 1.3357,
+      "step": 6078
+    },
+    {
+      "epoch": 5.611816293561043,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.95130086724483e-05,
+      "loss": 0.6704,
+      "step": 6079
+    },
+    {
+      "epoch": 5.612739441495499,
+      "grad_norm": 4.1171875,
+      "learning_rate": 5.950633755837225e-05,
+      "loss": 1.5808,
+      "step": 6080
+    },
+    {
+      "epoch": 5.613662589429956,
+      "grad_norm": 4.0546875,
+      "learning_rate": 5.94996664442962e-05,
+      "loss": 1.467,
+      "step": 6081
+    },
+    {
+      "epoch": 5.614585737364413,
+      "grad_norm": 3.2421875,
+      "learning_rate": 5.9492995330220146e-05,
+      "loss": 1.116,
+      "step": 6082
+    },
+    {
+      "epoch": 5.615508885298869,
+      "grad_norm": 3.521484375,
+      "learning_rate": 5.94863242161441e-05,
+      "loss": 1.5332,
+      "step": 6083
+    },
+    {
+      "epoch": 5.616432033233326,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.947965310206804e-05,
+      "loss": 1.5054,
+      "step": 6084
+    },
+    {
+      "epoch": 5.617355181167782,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.9472981987991995e-05,
+      "loss": 1.6367,
+      "step": 6085
+    },
+    {
+      "epoch": 5.618278329102239,
+      "grad_norm": 2.96484375,
+      "learning_rate": 5.946631087391594e-05,
+      "loss": 1.3876,
+      "step": 6086
+    },
+    {
+      "epoch": 5.619201477036695,
+      "grad_norm": 3.314453125,
+      "learning_rate": 5.945963975983989e-05,
+      "loss": 1.3307,
+      "step": 6087
+    },
+    {
+      "epoch": 5.620124624971152,
+      "grad_norm": 3.48828125,
+      "learning_rate": 5.945296864576384e-05,
+      "loss": 1.3157,
+      "step": 6088
+    },
+    {
+      "epoch": 5.621047772905608,
+      "grad_norm": 3.267578125,
+      "learning_rate": 5.94462975316878e-05,
+      "loss": 1.4536,
+      "step": 6089
+    },
+    {
+      "epoch": 5.621970920840065,
+      "grad_norm": 3.57421875,
+      "learning_rate": 5.943962641761175e-05,
+      "loss": 1.4802,
+      "step": 6090
+    },
+    {
+      "epoch": 5.622894068774521,
+      "grad_norm": 4.06640625,
+      "learning_rate": 5.9432955303535695e-05,
+      "loss": 1.5056,
+      "step": 6091
+    },
+    {
+      "epoch": 5.623817216708978,
+      "grad_norm": 4.83984375,
+      "learning_rate": 5.942628418945965e-05,
+      "loss": 1.4656,
+      "step": 6092
+    },
+    {
+      "epoch": 5.624740364643434,
+      "grad_norm": 3.7734375,
+      "learning_rate": 5.941961307538359e-05,
+      "loss": 1.4324,
+      "step": 6093
+    },
+    {
+      "epoch": 5.625663512577891,
+      "grad_norm": 3.126953125,
+      "learning_rate": 5.9412941961307545e-05,
+      "loss": 1.4749,
+      "step": 6094
+    },
+    {
+      "epoch": 5.626586660512347,
+      "grad_norm": 3.69921875,
+      "learning_rate": 5.940627084723149e-05,
+      "loss": 1.2469,
+      "step": 6095
+    },
+    {
+      "epoch": 5.627509808446804,
+      "grad_norm": 4.07421875,
+      "learning_rate": 5.939959973315544e-05,
+      "loss": 1.3862,
+      "step": 6096
+    },
+    {
+      "epoch": 5.6284329563812605,
+      "grad_norm": 3.017578125,
+      "learning_rate": 5.939292861907939e-05,
+      "loss": 1.0581,
+      "step": 6097
+    },
+    {
+      "epoch": 5.6293561043157165,
+      "grad_norm": 4.22265625,
+      "learning_rate": 5.938625750500334e-05,
+      "loss": 1.4355,
+      "step": 6098
+    },
+    {
+      "epoch": 5.6302792522501734,
+      "grad_norm": 3.47265625,
+      "learning_rate": 5.9379586390927286e-05,
+      "loss": 0.863,
+      "step": 6099
+    },
+    {
+      "epoch": 5.6312024001846295,
+      "grad_norm": 3.267578125,
+      "learning_rate": 5.937291527685124e-05,
+      "loss": 1.3047,
+      "step": 6100
+    },
+    {
+      "epoch": 5.632125548119086,
+      "grad_norm": 3.779296875,
+      "learning_rate": 5.936624416277518e-05,
+      "loss": 1.6946,
+      "step": 6101
+    },
+    {
+      "epoch": 5.633048696053542,
+      "grad_norm": 3.51171875,
+      "learning_rate": 5.9359573048699136e-05,
+      "loss": 1.2809,
+      "step": 6102
+    },
+    {
+      "epoch": 5.633971843987999,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.935290193462308e-05,
+      "loss": 1.5022,
+      "step": 6103
+    },
+    {
+      "epoch": 5.634894991922455,
+      "grad_norm": 3.921875,
+      "learning_rate": 5.934623082054703e-05,
+      "loss": 1.5811,
+      "step": 6104
+    },
+    {
+      "epoch": 5.635818139856912,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.933955970647098e-05,
+      "loss": 1.3806,
+      "step": 6105
+    },
+    {
+      "epoch": 5.636741287791368,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.933288859239493e-05,
+      "loss": 1.4983,
+      "step": 6106
+    },
+    {
+      "epoch": 5.637664435725825,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.9326217478318876e-05,
+      "loss": 1.2162,
+      "step": 6107
+    },
+    {
+      "epoch": 5.638587583660281,
+      "grad_norm": 3.41796875,
+      "learning_rate": 5.931954636424283e-05,
+      "loss": 1.2888,
+      "step": 6108
+    },
+    {
+      "epoch": 5.639510731594738,
+      "grad_norm": 3.19140625,
+      "learning_rate": 5.9312875250166774e-05,
+      "loss": 1.4365,
+      "step": 6109
+    },
+    {
+      "epoch": 5.640433879529194,
+      "grad_norm": 3.50390625,
+      "learning_rate": 5.9306204136090726e-05,
+      "loss": 1.4553,
+      "step": 6110
+    },
+    {
+      "epoch": 5.641357027463651,
+      "grad_norm": 3.3359375,
+      "learning_rate": 5.929953302201468e-05,
+      "loss": 1.498,
+      "step": 6111
+    },
+    {
+      "epoch": 5.642280175398108,
+      "grad_norm": 3.10546875,
+      "learning_rate": 5.9292861907938624e-05,
+      "loss": 1.4829,
+      "step": 6112
+    },
+    {
+      "epoch": 5.643203323332564,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.9286190793862576e-05,
+      "loss": 1.4565,
+      "step": 6113
+    },
+    {
+      "epoch": 5.64412647126702,
+      "grad_norm": 3.58984375,
+      "learning_rate": 5.9279519679786535e-05,
+      "loss": 0.9779,
+      "step": 6114
+    },
+    {
+      "epoch": 5.645049619201477,
+      "grad_norm": 4.21484375,
+      "learning_rate": 5.927284856571048e-05,
+      "loss": 1.5918,
+      "step": 6115
+    },
+    {
+      "epoch": 5.645972767135934,
+      "grad_norm": 4.24609375,
+      "learning_rate": 5.926617745163443e-05,
+      "loss": 1.4475,
+      "step": 6116
+    },
+    {
+      "epoch": 5.64689591507039,
+      "grad_norm": 3.48828125,
+      "learning_rate": 5.925950633755838e-05,
+      "loss": 1.5576,
+      "step": 6117
+    },
+    {
+      "epoch": 5.647819063004847,
+      "grad_norm": 3.591796875,
+      "learning_rate": 5.925283522348233e-05,
+      "loss": 1.4397,
+      "step": 6118
+    },
+    {
+      "epoch": 5.648742210939303,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.9246164109406276e-05,
+      "loss": 1.3094,
+      "step": 6119
+    },
+    {
+      "epoch": 5.64966535887376,
+      "grad_norm": 3.19921875,
+      "learning_rate": 5.923949299533023e-05,
+      "loss": 1.1337,
+      "step": 6120
+    },
+    {
+      "epoch": 5.650588506808216,
+      "grad_norm": 4.34765625,
+      "learning_rate": 5.923282188125417e-05,
+      "loss": 1.2513,
+      "step": 6121
+    },
+    {
+      "epoch": 5.651511654742673,
+      "grad_norm": 3.931640625,
+      "learning_rate": 5.9226150767178125e-05,
+      "loss": 1.4607,
+      "step": 6122
+    },
+    {
+      "epoch": 5.652434802677129,
+      "grad_norm": 3.103515625,
+      "learning_rate": 5.921947965310207e-05,
+      "loss": 0.5986,
+      "step": 6123
+    },
+    {
+      "epoch": 5.653357950611586,
+      "grad_norm": 3.7265625,
+      "learning_rate": 5.921280853902602e-05,
+      "loss": 1.5262,
+      "step": 6124
+    },
+    {
+      "epoch": 5.654281098546042,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.920613742494997e-05,
+      "loss": 1.4385,
+      "step": 6125
+    },
+    {
+      "epoch": 5.655204246480499,
+      "grad_norm": 3.68359375,
+      "learning_rate": 5.919946631087392e-05,
+      "loss": 1.3302,
+      "step": 6126
+    },
+    {
+      "epoch": 5.6561273944149555,
+      "grad_norm": 3.724609375,
+      "learning_rate": 5.9192795196797866e-05,
+      "loss": 1.1288,
+      "step": 6127
+    },
+    {
+      "epoch": 5.6570505423494115,
+      "grad_norm": 3.52734375,
+      "learning_rate": 5.918612408272182e-05,
+      "loss": 1.5239,
+      "step": 6128
+    },
+    {
+      "epoch": 5.6579736902838675,
+      "grad_norm": 3.08984375,
+      "learning_rate": 5.9179452968645764e-05,
+      "loss": 1.044,
+      "step": 6129
+    },
+    {
+      "epoch": 5.6588968382183245,
+      "grad_norm": 4.07421875,
+      "learning_rate": 5.9172781854569716e-05,
+      "loss": 1.3293,
+      "step": 6130
+    },
+    {
+      "epoch": 5.659819986152781,
+      "grad_norm": 4.59765625,
+      "learning_rate": 5.916611074049366e-05,
+      "loss": 1.4133,
+      "step": 6131
+    },
+    {
+      "epoch": 5.660743134087237,
+      "grad_norm": 3.802734375,
+      "learning_rate": 5.9159439626417614e-05,
+      "loss": 1.1392,
+      "step": 6132
+    },
+    {
+      "epoch": 5.661666282021694,
+      "grad_norm": 3.634765625,
+      "learning_rate": 5.915276851234156e-05,
+      "loss": 1.5774,
+      "step": 6133
+    },
+    {
+      "epoch": 5.66258942995615,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.914609739826551e-05,
+      "loss": 1.4619,
+      "step": 6134
+    },
+    {
+      "epoch": 5.663512577890607,
+      "grad_norm": 3.77734375,
+      "learning_rate": 5.913942628418946e-05,
+      "loss": 1.5591,
+      "step": 6135
+    },
+    {
+      "epoch": 5.664435725825063,
+      "grad_norm": 3.15234375,
+      "learning_rate": 5.913275517011341e-05,
+      "loss": 1.4277,
+      "step": 6136
+    },
+    {
+      "epoch": 5.66535887375952,
+      "grad_norm": 3.251953125,
+      "learning_rate": 5.9126084056037354e-05,
+      "loss": 1.1666,
+      "step": 6137
+    },
+    {
+      "epoch": 5.666282021693976,
+      "grad_norm": 3.068359375,
+      "learning_rate": 5.9119412941961307e-05,
+      "loss": 1.2751,
+      "step": 6138
+    },
+    {
+      "epoch": 5.667205169628433,
+      "grad_norm": 4.6640625,
+      "learning_rate": 5.9112741827885266e-05,
+      "loss": 1.4387,
+      "step": 6139
+    },
+    {
+      "epoch": 5.668128317562889,
+      "grad_norm": 3.66796875,
+      "learning_rate": 5.910607071380921e-05,
+      "loss": 1.2845,
+      "step": 6140
+    },
+    {
+      "epoch": 5.669051465497346,
+      "grad_norm": 3.2578125,
+      "learning_rate": 5.909939959973316e-05,
+      "loss": 1.3735,
+      "step": 6141
+    },
+    {
+      "epoch": 5.669974613431802,
+      "grad_norm": 3.69140625,
+      "learning_rate": 5.909272848565711e-05,
+      "loss": 1.4082,
+      "step": 6142
+    },
+    {
+      "epoch": 5.670897761366259,
+      "grad_norm": 3.509765625,
+      "learning_rate": 5.908605737158106e-05,
+      "loss": 1.4731,
+      "step": 6143
+    },
+    {
+      "epoch": 5.671820909300715,
+      "grad_norm": 3.365234375,
+      "learning_rate": 5.9079386257505006e-05,
+      "loss": 1.4817,
+      "step": 6144
+    },
+    {
+      "epoch": 5.672744057235172,
+      "grad_norm": 3.779296875,
+      "learning_rate": 5.907271514342896e-05,
+      "loss": 1.2839,
+      "step": 6145
+    },
+    {
+      "epoch": 5.673667205169629,
+      "grad_norm": 3.455078125,
+      "learning_rate": 5.9066044029352904e-05,
+      "loss": 1.3955,
+      "step": 6146
+    },
+    {
+      "epoch": 5.674590353104085,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.9059372915276856e-05,
+      "loss": 1.3154,
+      "step": 6147
+    },
+    {
+      "epoch": 5.675513501038542,
+      "grad_norm": 3.767578125,
+      "learning_rate": 5.90527018012008e-05,
+      "loss": 1.2645,
+      "step": 6148
+    },
+    {
+      "epoch": 5.676436648972998,
+      "grad_norm": 4.859375,
+      "learning_rate": 5.9046030687124754e-05,
+      "loss": 1.3774,
+      "step": 6149
+    },
+    {
+      "epoch": 5.677359796907455,
+      "grad_norm": 4.41015625,
+      "learning_rate": 5.9039359573048706e-05,
+      "loss": 1.4558,
+      "step": 6150
+    },
+    {
+      "epoch": 5.678282944841911,
+      "grad_norm": 3.98046875,
+      "learning_rate": 5.903268845897265e-05,
+      "loss": 1.1737,
+      "step": 6151
+    },
+    {
+      "epoch": 5.679206092776368,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.9026017344896604e-05,
+      "loss": 1.4707,
+      "step": 6152
+    },
+    {
+      "epoch": 5.680129240710824,
+      "grad_norm": 3.705078125,
+      "learning_rate": 5.901934623082055e-05,
+      "loss": 1.2849,
+      "step": 6153
+    },
+    {
+      "epoch": 5.681052388645281,
+      "grad_norm": 3.935546875,
+      "learning_rate": 5.90126751167445e-05,
+      "loss": 1.5276,
+      "step": 6154
+    },
+    {
+      "epoch": 5.681975536579737,
+      "grad_norm": 3.91796875,
+      "learning_rate": 5.900600400266845e-05,
+      "loss": 1.2329,
+      "step": 6155
+    },
+    {
+      "epoch": 5.6828986845141936,
+      "grad_norm": 3.33203125,
+      "learning_rate": 5.89993328885924e-05,
+      "loss": 1.4011,
+      "step": 6156
+    },
+    {
+      "epoch": 5.68382183244865,
+      "grad_norm": 4.04296875,
+      "learning_rate": 5.8992661774516344e-05,
+      "loss": 1.3579,
+      "step": 6157
+    },
+    {
+      "epoch": 5.6847449803831065,
+      "grad_norm": 3.748046875,
+      "learning_rate": 5.8985990660440296e-05,
+      "loss": 1.4685,
+      "step": 6158
+    },
+    {
+      "epoch": 5.6856681283175625,
+      "grad_norm": 4.05078125,
+      "learning_rate": 5.897931954636424e-05,
+      "loss": 1.4294,
+      "step": 6159
+    },
+    {
+      "epoch": 5.686591276252019,
+      "grad_norm": 3.791015625,
+      "learning_rate": 5.8972648432288194e-05,
+      "loss": 1.2361,
+      "step": 6160
+    },
+    {
+      "epoch": 5.687514424186476,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.896597731821214e-05,
+      "loss": 1.4078,
+      "step": 6161
+    },
+    {
+      "epoch": 5.688437572120932,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.895930620413609e-05,
+      "loss": 1.3999,
+      "step": 6162
+    },
+    {
+      "epoch": 5.689360720055388,
+      "grad_norm": 3.56640625,
+      "learning_rate": 5.895263509006004e-05,
+      "loss": 1.0733,
+      "step": 6163
+    },
+    {
+      "epoch": 5.690283867989845,
+      "grad_norm": 3.873046875,
+      "learning_rate": 5.8945963975983996e-05,
+      "loss": 1.2313,
+      "step": 6164
+    },
+    {
+      "epoch": 5.691207015924302,
+      "grad_norm": 4.41015625,
+      "learning_rate": 5.893929286190795e-05,
+      "loss": 1.6162,
+      "step": 6165
+    },
+    {
+      "epoch": 5.692130163858758,
+      "grad_norm": 3.572265625,
+      "learning_rate": 5.8932621747831894e-05,
+      "loss": 1.205,
+      "step": 6166
+    },
+    {
+      "epoch": 5.693053311793215,
+      "grad_norm": 4.1171875,
+      "learning_rate": 5.8925950633755846e-05,
+      "loss": 1.4424,
+      "step": 6167
+    },
+    {
+      "epoch": 5.693976459727671,
+      "grad_norm": 3.662109375,
+      "learning_rate": 5.891927951967979e-05,
+      "loss": 1.417,
+      "step": 6168
+    },
+    {
+      "epoch": 5.694899607662128,
+      "grad_norm": 3.76953125,
+      "learning_rate": 5.8912608405603744e-05,
+      "loss": 1.2,
+      "step": 6169
+    },
+    {
+      "epoch": 5.695822755596584,
+      "grad_norm": 3.884765625,
+      "learning_rate": 5.890593729152769e-05,
+      "loss": 1.4299,
+      "step": 6170
+    },
+    {
+      "epoch": 5.696745903531041,
+      "grad_norm": 3.17578125,
+      "learning_rate": 5.889926617745164e-05,
+      "loss": 1.4048,
+      "step": 6171
+    },
+    {
+      "epoch": 5.697669051465497,
+      "grad_norm": 4.01953125,
+      "learning_rate": 5.889259506337559e-05,
+      "loss": 1.3197,
+      "step": 6172
+    },
+    {
+      "epoch": 5.698592199399954,
+      "grad_norm": 5.09765625,
+      "learning_rate": 5.888592394929954e-05,
+      "loss": 1.2471,
+      "step": 6173
+    },
+    {
+      "epoch": 5.69951534733441,
+      "grad_norm": 3.326171875,
+      "learning_rate": 5.8879252835223484e-05,
+      "loss": 1.0881,
+      "step": 6174
+    },
+    {
+      "epoch": 5.700438495268867,
+      "grad_norm": 4.3359375,
+      "learning_rate": 5.8872581721147437e-05,
+      "loss": 1.6238,
+      "step": 6175
+    },
+    {
+      "epoch": 5.701361643203323,
+      "grad_norm": 3.37109375,
+      "learning_rate": 5.886591060707138e-05,
+      "loss": 1.4319,
+      "step": 6176
+    },
+    {
+      "epoch": 5.70228479113778,
+      "grad_norm": 4.125,
+      "learning_rate": 5.8859239492995334e-05,
+      "loss": 1.0023,
+      "step": 6177
+    },
+    {
+      "epoch": 5.703207939072236,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.885256837891928e-05,
+      "loss": 1.1271,
+      "step": 6178
+    },
+    {
+      "epoch": 5.704131087006693,
+      "grad_norm": 4.6875,
+      "learning_rate": 5.884589726484323e-05,
+      "loss": 1.5566,
+      "step": 6179
+    },
+    {
+      "epoch": 5.70505423494115,
+      "grad_norm": 3.958984375,
+      "learning_rate": 5.883922615076718e-05,
+      "loss": 1.2734,
+      "step": 6180
+    },
+    {
+      "epoch": 5.705977382875606,
+      "grad_norm": 3.458984375,
+      "learning_rate": 5.883255503669113e-05,
+      "loss": 1.2288,
+      "step": 6181
+    },
+    {
+      "epoch": 5.706900530810063,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.8825883922615075e-05,
+      "loss": 1.564,
+      "step": 6182
+    },
+    {
+      "epoch": 5.707823678744519,
+      "grad_norm": 3.19140625,
+      "learning_rate": 5.881921280853903e-05,
+      "loss": 0.8358,
+      "step": 6183
+    },
+    {
+      "epoch": 5.708746826678976,
+      "grad_norm": 4.2421875,
+      "learning_rate": 5.881254169446297e-05,
+      "loss": 1.3835,
+      "step": 6184
+    },
+    {
+      "epoch": 5.709669974613432,
+      "grad_norm": 3.251953125,
+      "learning_rate": 5.8805870580386925e-05,
+      "loss": 0.959,
+      "step": 6185
+    },
+    {
+      "epoch": 5.7105931225478885,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.879919946631087e-05,
+      "loss": 1.1154,
+      "step": 6186
+    },
+    {
+      "epoch": 5.711516270482345,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.879252835223482e-05,
+      "loss": 1.6702,
+      "step": 6187
+    },
+    {
+      "epoch": 5.7124394184168015,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.878585723815877e-05,
+      "loss": 1.301,
+      "step": 6188
+    },
+    {
+      "epoch": 5.7133625663512575,
+      "grad_norm": 3.763671875,
+      "learning_rate": 5.877918612408273e-05,
+      "loss": 1.0128,
+      "step": 6189
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 3.369140625,
+      "learning_rate": 5.877251501000668e-05,
+      "loss": 1.4753,
+      "step": 6190
+    },
+    {
+      "epoch": 5.71520886222017,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.876584389593063e-05,
+      "loss": 1.3386,
+      "step": 6191
+    },
+    {
+      "epoch": 5.716132010154627,
+      "grad_norm": 3.72265625,
+      "learning_rate": 5.875917278185458e-05,
+      "loss": 1.2844,
+      "step": 6192
+    },
+    {
+      "epoch": 5.717055158089083,
+      "grad_norm": 3.314453125,
+      "learning_rate": 5.875250166777853e-05,
+      "loss": 1.4929,
+      "step": 6193
+    },
+    {
+      "epoch": 5.71797830602354,
+      "grad_norm": 3.884765625,
+      "learning_rate": 5.8745830553702474e-05,
+      "loss": 1.4163,
+      "step": 6194
+    },
+    {
+      "epoch": 5.718901453957997,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.8739159439626427e-05,
+      "loss": 1.498,
+      "step": 6195
+    },
+    {
+      "epoch": 5.719824601892453,
+      "grad_norm": 3.158203125,
+      "learning_rate": 5.873248832555037e-05,
+      "loss": 1.6421,
+      "step": 6196
+    },
+    {
+      "epoch": 5.72074774982691,
+      "grad_norm": 4.7890625,
+      "learning_rate": 5.8725817211474324e-05,
+      "loss": 1.3804,
+      "step": 6197
+    },
+    {
+      "epoch": 5.721670897761366,
+      "grad_norm": 3.392578125,
+      "learning_rate": 5.871914609739827e-05,
+      "loss": 1.2261,
+      "step": 6198
+    },
+    {
+      "epoch": 5.722594045695823,
+      "grad_norm": 3.306640625,
+      "learning_rate": 5.871247498332222e-05,
+      "loss": 1.5615,
+      "step": 6199
+    },
+    {
+      "epoch": 5.723517193630279,
+      "grad_norm": 3.6953125,
+      "learning_rate": 5.870580386924617e-05,
+      "loss": 1.3337,
+      "step": 6200
+    },
+    {
+      "epoch": 5.724440341564736,
+      "grad_norm": 2.962890625,
+      "learning_rate": 5.869913275517012e-05,
+      "loss": 1.4067,
+      "step": 6201
+    },
+    {
+      "epoch": 5.725363489499192,
+      "grad_norm": 4.35546875,
+      "learning_rate": 5.8692461641094065e-05,
+      "loss": 1.8584,
+      "step": 6202
+    },
+    {
+      "epoch": 5.726286637433649,
+      "grad_norm": 3.57421875,
+      "learning_rate": 5.868579052701802e-05,
+      "loss": 1.2345,
+      "step": 6203
+    },
+    {
+      "epoch": 5.727209785368105,
+      "grad_norm": 3.7578125,
+      "learning_rate": 5.867911941294196e-05,
+      "loss": 1.1891,
+      "step": 6204
+    },
+    {
+      "epoch": 5.728132933302562,
+      "grad_norm": 4.79296875,
+      "learning_rate": 5.8672448298865915e-05,
+      "loss": 1.356,
+      "step": 6205
+    },
+    {
+      "epoch": 5.729056081237018,
+      "grad_norm": 4.796875,
+      "learning_rate": 5.866577718478986e-05,
+      "loss": 1.5371,
+      "step": 6206
+    },
+    {
+      "epoch": 5.729979229171475,
+      "grad_norm": 3.51171875,
+      "learning_rate": 5.865910607071381e-05,
+      "loss": 1.5564,
+      "step": 6207
+    },
+    {
+      "epoch": 5.730902377105931,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.865243495663776e-05,
+      "loss": 1.2124,
+      "step": 6208
+    },
+    {
+      "epoch": 5.731825525040388,
+      "grad_norm": 3.384765625,
+      "learning_rate": 5.864576384256171e-05,
+      "loss": 1.5178,
+      "step": 6209
+    },
+    {
+      "epoch": 5.732748672974845,
+      "grad_norm": 3.216796875,
+      "learning_rate": 5.8639092728485655e-05,
+      "loss": 1.2251,
+      "step": 6210
+    },
+    {
+      "epoch": 5.733671820909301,
+      "grad_norm": 4.3984375,
+      "learning_rate": 5.863242161440961e-05,
+      "loss": 1.3685,
+      "step": 6211
+    },
+    {
+      "epoch": 5.734594968843757,
+      "grad_norm": 4.62890625,
+      "learning_rate": 5.862575050033355e-05,
+      "loss": 1.4082,
+      "step": 6212
+    },
+    {
+      "epoch": 5.735518116778214,
+      "grad_norm": 3.552734375,
+      "learning_rate": 5.8619079386257505e-05,
+      "loss": 1.3945,
+      "step": 6213
+    },
+    {
+      "epoch": 5.736441264712671,
+      "grad_norm": 3.505859375,
+      "learning_rate": 5.861240827218145e-05,
+      "loss": 1.583,
+      "step": 6214
+    },
+    {
+      "epoch": 5.737364412647127,
+      "grad_norm": 3.572265625,
+      "learning_rate": 5.860573715810541e-05,
+      "loss": 1.3513,
+      "step": 6215
+    },
+    {
+      "epoch": 5.7382875605815835,
+      "grad_norm": 3.681640625,
+      "learning_rate": 5.859906604402936e-05,
+      "loss": 1.1238,
+      "step": 6216
+    },
+    {
+      "epoch": 5.7392107085160395,
+      "grad_norm": 3.9453125,
+      "learning_rate": 5.859239492995331e-05,
+      "loss": 1.5859,
+      "step": 6217
+    },
+    {
+      "epoch": 5.7401338564504965,
+      "grad_norm": 3.68359375,
+      "learning_rate": 5.858572381587726e-05,
+      "loss": 1.623,
+      "step": 6218
+    },
+    {
+      "epoch": 5.7410570043849525,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.8579052701801205e-05,
+      "loss": 1.5054,
+      "step": 6219
+    },
+    {
+      "epoch": 5.741980152319409,
+      "grad_norm": 3.375,
+      "learning_rate": 5.857238158772516e-05,
+      "loss": 1.3115,
+      "step": 6220
+    },
+    {
+      "epoch": 5.742903300253865,
+      "grad_norm": 3.607421875,
+      "learning_rate": 5.85657104736491e-05,
+      "loss": 1.4016,
+      "step": 6221
+    },
+    {
+      "epoch": 5.743826448188322,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.8559039359573055e-05,
+      "loss": 1.4458,
+      "step": 6222
+    },
+    {
+      "epoch": 5.744749596122778,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.8552368245497e-05,
+      "loss": 1.3647,
+      "step": 6223
+    },
+    {
+      "epoch": 5.745672744057235,
+      "grad_norm": 3.650390625,
+      "learning_rate": 5.854569713142095e-05,
+      "loss": 1.4136,
+      "step": 6224
+    },
+    {
+      "epoch": 5.746595891991691,
+      "grad_norm": 3.708984375,
+      "learning_rate": 5.85390260173449e-05,
+      "loss": 1.3994,
+      "step": 6225
+    },
+    {
+      "epoch": 5.747519039926148,
+      "grad_norm": 4.16015625,
+      "learning_rate": 5.853235490326885e-05,
+      "loss": 1.2524,
+      "step": 6226
+    },
+    {
+      "epoch": 5.748442187860604,
+      "grad_norm": 3.732421875,
+      "learning_rate": 5.8525683789192795e-05,
+      "loss": 1.3994,
+      "step": 6227
+    },
+    {
+      "epoch": 5.749365335795061,
+      "grad_norm": 3.8515625,
+      "learning_rate": 5.851901267511675e-05,
+      "loss": 1.1379,
+      "step": 6228
+    },
+    {
+      "epoch": 5.750288483729518,
+      "grad_norm": 3.3828125,
+      "learning_rate": 5.851234156104069e-05,
+      "loss": 1.3257,
+      "step": 6229
+    },
+    {
+      "epoch": 5.751211631663974,
+      "grad_norm": 3.75390625,
+      "learning_rate": 5.8505670446964645e-05,
+      "loss": 1.427,
+      "step": 6230
+    },
+    {
+      "epoch": 5.752134779598431,
+      "grad_norm": 3.572265625,
+      "learning_rate": 5.849899933288859e-05,
+      "loss": 1.4871,
+      "step": 6231
+    },
+    {
+      "epoch": 5.753057927532887,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.849232821881254e-05,
+      "loss": 1.623,
+      "step": 6232
+    },
+    {
+      "epoch": 5.753981075467344,
+      "grad_norm": 3.216796875,
+      "learning_rate": 5.848565710473649e-05,
+      "loss": 1.1406,
+      "step": 6233
+    },
+    {
+      "epoch": 5.7549042234018,
+      "grad_norm": 3.771484375,
+      "learning_rate": 5.847898599066044e-05,
+      "loss": 1.4792,
+      "step": 6234
+    },
+    {
+      "epoch": 5.755827371336257,
+      "grad_norm": 3.7109375,
+      "learning_rate": 5.8472314876584386e-05,
+      "loss": 1.3069,
+      "step": 6235
+    },
+    {
+      "epoch": 5.756750519270713,
+      "grad_norm": 3.119140625,
+      "learning_rate": 5.846564376250834e-05,
+      "loss": 1.584,
+      "step": 6236
+    },
+    {
+      "epoch": 5.75767366720517,
+      "grad_norm": 3.630859375,
+      "learning_rate": 5.8458972648432284e-05,
+      "loss": 1.6267,
+      "step": 6237
+    },
+    {
+      "epoch": 5.758596815139626,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.8452301534356236e-05,
+      "loss": 1.0154,
+      "step": 6238
+    },
+    {
+      "epoch": 5.759519963074083,
+      "grad_norm": 3.8046875,
+      "learning_rate": 5.844563042028018e-05,
+      "loss": 1.3594,
+      "step": 6239
+    },
+    {
+      "epoch": 5.760443111008539,
+      "grad_norm": 3.458984375,
+      "learning_rate": 5.843895930620415e-05,
+      "loss": 1.0629,
+      "step": 6240
+    },
+    {
+      "epoch": 5.761366258942996,
+      "grad_norm": 3.39453125,
+      "learning_rate": 5.843228819212809e-05,
+      "loss": 0.9824,
+      "step": 6241
+    },
+    {
+      "epoch": 5.762289406877452,
+      "grad_norm": 3.998046875,
+      "learning_rate": 5.8425617078052045e-05,
+      "loss": 1.2625,
+      "step": 6242
+    },
+    {
+      "epoch": 5.763212554811909,
+      "grad_norm": 3.66015625,
+      "learning_rate": 5.841894596397599e-05,
+      "loss": 1.0564,
+      "step": 6243
+    },
+    {
+      "epoch": 5.7641357027463656,
+      "grad_norm": 3.318359375,
+      "learning_rate": 5.841227484989994e-05,
+      "loss": 1.1921,
+      "step": 6244
+    },
+    {
+      "epoch": 5.765058850680822,
+      "grad_norm": 3.560546875,
+      "learning_rate": 5.840560373582389e-05,
+      "loss": 1.4524,
+      "step": 6245
+    },
+    {
+      "epoch": 5.765981998615278,
+      "grad_norm": 3.361328125,
+      "learning_rate": 5.839893262174784e-05,
+      "loss": 1.5686,
+      "step": 6246
+    },
+    {
+      "epoch": 5.7669051465497345,
+      "grad_norm": 3.896484375,
+      "learning_rate": 5.8392261507671785e-05,
+      "loss": 1.4253,
+      "step": 6247
+    },
+    {
+      "epoch": 5.767828294484191,
+      "grad_norm": 3.212890625,
+      "learning_rate": 5.838559039359574e-05,
+      "loss": 1.1886,
+      "step": 6248
+    },
+    {
+      "epoch": 5.7687514424186475,
+      "grad_norm": 3.96484375,
+      "learning_rate": 5.837891927951968e-05,
+      "loss": 1.7822,
+      "step": 6249
+    },
+    {
+      "epoch": 5.769674590353104,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.8372248165443635e-05,
+      "loss": 1.6584,
+      "step": 6250
+    },
+    {
+      "epoch": 5.77059773828756,
+      "grad_norm": 3.6171875,
+      "learning_rate": 5.836557705136758e-05,
+      "loss": 1.199,
+      "step": 6251
+    },
+    {
+      "epoch": 5.771520886222017,
+      "grad_norm": 4.19921875,
+      "learning_rate": 5.835890593729153e-05,
+      "loss": 1.2963,
+      "step": 6252
+    },
+    {
+      "epoch": 5.772444034156473,
+      "grad_norm": 3.388671875,
+      "learning_rate": 5.835223482321548e-05,
+      "loss": 1.4097,
+      "step": 6253
+    },
+    {
+      "epoch": 5.77336718209093,
+      "grad_norm": 3.5546875,
+      "learning_rate": 5.834556370913943e-05,
+      "loss": 1.4507,
+      "step": 6254
+    },
+    {
+      "epoch": 5.774290330025386,
+      "grad_norm": 3.705078125,
+      "learning_rate": 5.8338892595063376e-05,
+      "loss": 1.4275,
+      "step": 6255
+    },
+    {
+      "epoch": 5.775213477959843,
+      "grad_norm": 4.05078125,
+      "learning_rate": 5.833222148098733e-05,
+      "loss": 1.5188,
+      "step": 6256
+    },
+    {
+      "epoch": 5.776136625894299,
+      "grad_norm": 4.38671875,
+      "learning_rate": 5.8325550366911274e-05,
+      "loss": 1.3356,
+      "step": 6257
+    },
+    {
+      "epoch": 5.777059773828756,
+      "grad_norm": 3.935546875,
+      "learning_rate": 5.8318879252835226e-05,
+      "loss": 1.3838,
+      "step": 6258
+    },
+    {
+      "epoch": 5.777982921763213,
+      "grad_norm": 3.404296875,
+      "learning_rate": 5.831220813875917e-05,
+      "loss": 1.1558,
+      "step": 6259
+    },
+    {
+      "epoch": 5.778906069697669,
+      "grad_norm": 3.416015625,
+      "learning_rate": 5.8305537024683123e-05,
+      "loss": 1.3232,
+      "step": 6260
+    },
+    {
+      "epoch": 5.779829217632125,
+      "grad_norm": 3.90234375,
+      "learning_rate": 5.829886591060707e-05,
+      "loss": 1.4519,
+      "step": 6261
+    },
+    {
+      "epoch": 5.780752365566582,
+      "grad_norm": 3.564453125,
+      "learning_rate": 5.829219479653102e-05,
+      "loss": 1.3186,
+      "step": 6262
+    },
+    {
+      "epoch": 5.781675513501039,
+      "grad_norm": 4.79296875,
+      "learning_rate": 5.8285523682454967e-05,
+      "loss": 1.3835,
+      "step": 6263
+    },
+    {
+      "epoch": 5.782598661435495,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.827885256837892e-05,
+      "loss": 1.2034,
+      "step": 6264
+    },
+    {
+      "epoch": 5.783521809369952,
+      "grad_norm": 4.54296875,
+      "learning_rate": 5.827218145430288e-05,
+      "loss": 1.6111,
+      "step": 6265
+    },
+    {
+      "epoch": 5.784444957304408,
+      "grad_norm": 4.671875,
+      "learning_rate": 5.826551034022682e-05,
+      "loss": 1.3818,
+      "step": 6266
+    },
+    {
+      "epoch": 5.785368105238865,
+      "grad_norm": 2.939453125,
+      "learning_rate": 5.8258839226150775e-05,
+      "loss": 1.439,
+      "step": 6267
+    },
+    {
+      "epoch": 5.786291253173321,
+      "grad_norm": 3.57421875,
+      "learning_rate": 5.825216811207472e-05,
+      "loss": 1.5029,
+      "step": 6268
+    },
+    {
+      "epoch": 5.787214401107778,
+      "grad_norm": 3.34765625,
+      "learning_rate": 5.824549699799867e-05,
+      "loss": 1.429,
+      "step": 6269
+    },
+    {
+      "epoch": 5.788137549042234,
+      "grad_norm": 4.51953125,
+      "learning_rate": 5.823882588392262e-05,
+      "loss": 1.4669,
+      "step": 6270
+    },
+    {
+      "epoch": 5.789060696976691,
+      "grad_norm": 3.595703125,
+      "learning_rate": 5.823215476984657e-05,
+      "loss": 1.4062,
+      "step": 6271
+    },
+    {
+      "epoch": 5.789983844911147,
+      "grad_norm": 4.1015625,
+      "learning_rate": 5.8225483655770516e-05,
+      "loss": 1.3809,
+      "step": 6272
+    },
+    {
+      "epoch": 5.790906992845604,
+      "grad_norm": 3.423828125,
+      "learning_rate": 5.821881254169447e-05,
+      "loss": 1.3872,
+      "step": 6273
+    },
+    {
+      "epoch": 5.79183014078006,
+      "grad_norm": 4.01171875,
+      "learning_rate": 5.8212141427618414e-05,
+      "loss": 1.3411,
+      "step": 6274
+    },
+    {
+      "epoch": 5.792753288714517,
+      "grad_norm": 3.25390625,
+      "learning_rate": 5.8205470313542366e-05,
+      "loss": 1.3059,
+      "step": 6275
+    },
+    {
+      "epoch": 5.793676436648973,
+      "grad_norm": 3.375,
+      "learning_rate": 5.819879919946631e-05,
+      "loss": 1.2343,
+      "step": 6276
+    },
+    {
+      "epoch": 5.7945995845834295,
+      "grad_norm": 3.55859375,
+      "learning_rate": 5.8192128085390264e-05,
+      "loss": 1.1122,
+      "step": 6277
+    },
+    {
+      "epoch": 5.795522732517886,
+      "grad_norm": 3.310546875,
+      "learning_rate": 5.818545697131421e-05,
+      "loss": 1.3516,
+      "step": 6278
+    },
+    {
+      "epoch": 5.796445880452342,
+      "grad_norm": 3.27734375,
+      "learning_rate": 5.817878585723816e-05,
+      "loss": 1.5825,
+      "step": 6279
+    },
+    {
+      "epoch": 5.797369028386799,
+      "grad_norm": 3.615234375,
+      "learning_rate": 5.8172114743162107e-05,
+      "loss": 1.5959,
+      "step": 6280
+    },
+    {
+      "epoch": 5.798292176321255,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.816544362908606e-05,
+      "loss": 1.0569,
+      "step": 6281
+    },
+    {
+      "epoch": 5.799215324255712,
+      "grad_norm": 4.2265625,
+      "learning_rate": 5.8158772515010004e-05,
+      "loss": 1.6282,
+      "step": 6282
+    },
+    {
+      "epoch": 5.800138472190168,
+      "grad_norm": 3.580078125,
+      "learning_rate": 5.8152101400933956e-05,
+      "loss": 1.5244,
+      "step": 6283
+    },
+    {
+      "epoch": 5.801061620124625,
+      "grad_norm": 3.794921875,
+      "learning_rate": 5.81454302868579e-05,
+      "loss": 1.1729,
+      "step": 6284
+    },
+    {
+      "epoch": 5.801984768059081,
+      "grad_norm": 3.23046875,
+      "learning_rate": 5.8138759172781854e-05,
+      "loss": 1.0753,
+      "step": 6285
+    },
+    {
+      "epoch": 5.802907915993538,
+      "grad_norm": 3.302734375,
+      "learning_rate": 5.81320880587058e-05,
+      "loss": 1.4199,
+      "step": 6286
+    },
+    {
+      "epoch": 5.803831063927994,
+      "grad_norm": 3.466796875,
+      "learning_rate": 5.812541694462975e-05,
+      "loss": 1.5911,
+      "step": 6287
+    },
+    {
+      "epoch": 5.804754211862451,
+      "grad_norm": 3.177734375,
+      "learning_rate": 5.81187458305537e-05,
+      "loss": 1.1203,
+      "step": 6288
+    },
+    {
+      "epoch": 5.805677359796907,
+      "grad_norm": 3.400390625,
+      "learning_rate": 5.811207471647765e-05,
+      "loss": 1.2123,
+      "step": 6289
+    },
+    {
+      "epoch": 5.806600507731364,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.810540360240161e-05,
+      "loss": 1.2668,
+      "step": 6290
+    },
+    {
+      "epoch": 5.80752365566582,
+      "grad_norm": 3.87890625,
+      "learning_rate": 5.809873248832556e-05,
+      "loss": 1.4783,
+      "step": 6291
+    },
+    {
+      "epoch": 5.808446803600277,
+      "grad_norm": 3.177734375,
+      "learning_rate": 5.8092061374249506e-05,
+      "loss": 1.697,
+      "step": 6292
+    },
+    {
+      "epoch": 5.809369951534734,
+      "grad_norm": 3.666015625,
+      "learning_rate": 5.808539026017346e-05,
+      "loss": 1.3276,
+      "step": 6293
+    },
+    {
+      "epoch": 5.81029309946919,
+      "grad_norm": 3.826171875,
+      "learning_rate": 5.8078719146097404e-05,
+      "loss": 1.1379,
+      "step": 6294
+    },
+    {
+      "epoch": 5.811216247403646,
+      "grad_norm": 3.896484375,
+      "learning_rate": 5.8072048032021356e-05,
+      "loss": 1.4407,
+      "step": 6295
+    },
+    {
+      "epoch": 5.812139395338103,
+      "grad_norm": 3.537109375,
+      "learning_rate": 5.80653769179453e-05,
+      "loss": 1.0743,
+      "step": 6296
+    },
+    {
+      "epoch": 5.81306254327256,
+      "grad_norm": 3.44140625,
+      "learning_rate": 5.8058705803869253e-05,
+      "loss": 1.2802,
+      "step": 6297
+    },
+    {
+      "epoch": 5.813985691207016,
+      "grad_norm": 4.29296875,
+      "learning_rate": 5.80520346897932e-05,
+      "loss": 1.1885,
+      "step": 6298
+    },
+    {
+      "epoch": 5.814908839141473,
+      "grad_norm": 4.01171875,
+      "learning_rate": 5.804536357571715e-05,
+      "loss": 1.4016,
+      "step": 6299
+    },
+    {
+      "epoch": 5.815831987075929,
+      "grad_norm": 3.943359375,
+      "learning_rate": 5.8038692461641097e-05,
+      "loss": 1.2993,
+      "step": 6300
+    },
+    {
+      "epoch": 5.816755135010386,
+      "grad_norm": 3.763671875,
+      "learning_rate": 5.803202134756505e-05,
+      "loss": 1.3013,
+      "step": 6301
+    },
+    {
+      "epoch": 5.817678282944842,
+      "grad_norm": 3.623046875,
+      "learning_rate": 5.8025350233488994e-05,
+      "loss": 1.2483,
+      "step": 6302
+    },
+    {
+      "epoch": 5.818601430879299,
+      "grad_norm": 3.787109375,
+      "learning_rate": 5.8018679119412946e-05,
+      "loss": 1.5305,
+      "step": 6303
+    },
+    {
+      "epoch": 5.819524578813755,
+      "grad_norm": 3.47265625,
+      "learning_rate": 5.801200800533689e-05,
+      "loss": 0.9464,
+      "step": 6304
+    },
+    {
+      "epoch": 5.8204477267482115,
+      "grad_norm": 2.91796875,
+      "learning_rate": 5.8005336891260844e-05,
+      "loss": 0.9864,
+      "step": 6305
+    },
+    {
+      "epoch": 5.821370874682668,
+      "grad_norm": 3.7265625,
+      "learning_rate": 5.799866577718479e-05,
+      "loss": 1.2515,
+      "step": 6306
+    },
+    {
+      "epoch": 5.8222940226171245,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.799199466310874e-05,
+      "loss": 1.4048,
+      "step": 6307
+    },
+    {
+      "epoch": 5.823217170551581,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.798532354903269e-05,
+      "loss": 1.3284,
+      "step": 6308
+    },
+    {
+      "epoch": 5.824140318486037,
+      "grad_norm": 3.599609375,
+      "learning_rate": 5.797865243495664e-05,
+      "loss": 1.4441,
+      "step": 6309
+    },
+    {
+      "epoch": 5.825063466420493,
+      "grad_norm": 3.427734375,
+      "learning_rate": 5.7971981320880585e-05,
+      "loss": 1.4119,
+      "step": 6310
+    },
+    {
+      "epoch": 5.82598661435495,
+      "grad_norm": 3.728515625,
+      "learning_rate": 5.796531020680454e-05,
+      "loss": 1.3153,
+      "step": 6311
+    },
+    {
+      "epoch": 5.826909762289407,
+      "grad_norm": 3.55859375,
+      "learning_rate": 5.795863909272848e-05,
+      "loss": 1.4774,
+      "step": 6312
+    },
+    {
+      "epoch": 5.827832910223863,
+      "grad_norm": 3.58984375,
+      "learning_rate": 5.7951967978652435e-05,
+      "loss": 1.2834,
+      "step": 6313
+    },
+    {
+      "epoch": 5.82875605815832,
+      "grad_norm": 4.2109375,
+      "learning_rate": 5.794529686457638e-05,
+      "loss": 1.345,
+      "step": 6314
+    },
+    {
+      "epoch": 5.829679206092776,
+      "grad_norm": 3.73046875,
+      "learning_rate": 5.793862575050034e-05,
+      "loss": 1.5322,
+      "step": 6315
+    },
+    {
+      "epoch": 5.830602354027233,
+      "grad_norm": 4.09765625,
+      "learning_rate": 5.793195463642429e-05,
+      "loss": 1.2281,
+      "step": 6316
+    },
+    {
+      "epoch": 5.831525501961689,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.7925283522348237e-05,
+      "loss": 1.3718,
+      "step": 6317
+    },
+    {
+      "epoch": 5.832448649896146,
+      "grad_norm": 4.02734375,
+      "learning_rate": 5.791861240827219e-05,
+      "loss": 1.3296,
+      "step": 6318
+    },
+    {
+      "epoch": 5.833371797830602,
+      "grad_norm": 3.830078125,
+      "learning_rate": 5.7911941294196134e-05,
+      "loss": 1.4277,
+      "step": 6319
+    },
+    {
+      "epoch": 5.834294945765059,
+      "grad_norm": 3.845703125,
+      "learning_rate": 5.7905270180120086e-05,
+      "loss": 1.3359,
+      "step": 6320
+    },
+    {
+      "epoch": 5.835218093699515,
+      "grad_norm": 3.55078125,
+      "learning_rate": 5.789859906604403e-05,
+      "loss": 1.6331,
+      "step": 6321
+    },
+    {
+      "epoch": 5.836141241633972,
+      "grad_norm": 3.208984375,
+      "learning_rate": 5.7891927951967984e-05,
+      "loss": 0.9729,
+      "step": 6322
+    },
+    {
+      "epoch": 5.837064389568428,
+      "grad_norm": 3.4765625,
+      "learning_rate": 5.788525683789193e-05,
+      "loss": 1.3308,
+      "step": 6323
+    },
+    {
+      "epoch": 5.837987537502885,
+      "grad_norm": 4.16015625,
+      "learning_rate": 5.787858572381588e-05,
+      "loss": 1.4312,
+      "step": 6324
+    },
+    {
+      "epoch": 5.838910685437341,
+      "grad_norm": 3.9140625,
+      "learning_rate": 5.787191460973983e-05,
+      "loss": 1.4294,
+      "step": 6325
+    },
+    {
+      "epoch": 5.839833833371798,
+      "grad_norm": 3.681640625,
+      "learning_rate": 5.786524349566378e-05,
+      "loss": 1.3831,
+      "step": 6326
+    },
+    {
+      "epoch": 5.840756981306255,
+      "grad_norm": 4.05078125,
+      "learning_rate": 5.7858572381587725e-05,
+      "loss": 1.4297,
+      "step": 6327
+    },
+    {
+      "epoch": 5.841680129240711,
+      "grad_norm": 3.55078125,
+      "learning_rate": 5.785190126751168e-05,
+      "loss": 0.6685,
+      "step": 6328
+    },
+    {
+      "epoch": 5.842603277175168,
+      "grad_norm": 3.86328125,
+      "learning_rate": 5.784523015343562e-05,
+      "loss": 1.4377,
+      "step": 6329
+    },
+    {
+      "epoch": 5.843526425109624,
+      "grad_norm": 3.986328125,
+      "learning_rate": 5.7838559039359575e-05,
+      "loss": 1.4172,
+      "step": 6330
+    },
+    {
+      "epoch": 5.844449573044081,
+      "grad_norm": 3.322265625,
+      "learning_rate": 5.783188792528352e-05,
+      "loss": 1.5771,
+      "step": 6331
+    },
+    {
+      "epoch": 5.845372720978537,
+      "grad_norm": 3.423828125,
+      "learning_rate": 5.782521681120747e-05,
+      "loss": 1.2524,
+      "step": 6332
+    },
+    {
+      "epoch": 5.846295868912994,
+      "grad_norm": 2.931640625,
+      "learning_rate": 5.781854569713142e-05,
+      "loss": 1.5642,
+      "step": 6333
+    },
+    {
+      "epoch": 5.84721901684745,
+      "grad_norm": 3.5078125,
+      "learning_rate": 5.781187458305537e-05,
+      "loss": 1.2617,
+      "step": 6334
+    },
+    {
+      "epoch": 5.8481421647819065,
+      "grad_norm": 3.287109375,
+      "learning_rate": 5.7805203468979315e-05,
+      "loss": 1.3662,
+      "step": 6335
+    },
+    {
+      "epoch": 5.8490653127163625,
+      "grad_norm": 3.259765625,
+      "learning_rate": 5.779853235490327e-05,
+      "loss": 1.3706,
+      "step": 6336
+    },
+    {
+      "epoch": 5.8499884606508195,
+      "grad_norm": 4.18359375,
+      "learning_rate": 5.779186124082722e-05,
+      "loss": 1.4209,
+      "step": 6337
+    },
+    {
+      "epoch": 5.8509116085852755,
+      "grad_norm": 3.74609375,
+      "learning_rate": 5.7785190126751165e-05,
+      "loss": 1.3306,
+      "step": 6338
+    },
+    {
+      "epoch": 5.851834756519732,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.777851901267512e-05,
+      "loss": 1.3522,
+      "step": 6339
+    },
+    {
+      "epoch": 5.852757904454188,
+      "grad_norm": 3.3671875,
+      "learning_rate": 5.7771847898599076e-05,
+      "loss": 1.3999,
+      "step": 6340
+    },
+    {
+      "epoch": 5.853681052388645,
+      "grad_norm": 4.1875,
+      "learning_rate": 5.776517678452302e-05,
+      "loss": 1.2375,
+      "step": 6341
+    },
+    {
+      "epoch": 5.854604200323102,
+      "grad_norm": 3.486328125,
+      "learning_rate": 5.7758505670446974e-05,
+      "loss": 1.4856,
+      "step": 6342
+    },
+    {
+      "epoch": 5.855527348257558,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.775183455637092e-05,
+      "loss": 1.2905,
+      "step": 6343
+    },
+    {
+      "epoch": 5.856450496192014,
+      "grad_norm": 3.5234375,
+      "learning_rate": 5.774516344229487e-05,
+      "loss": 1.5437,
+      "step": 6344
+    },
+    {
+      "epoch": 5.857373644126471,
+      "grad_norm": 3.693359375,
+      "learning_rate": 5.773849232821882e-05,
+      "loss": 1.3131,
+      "step": 6345
+    },
+    {
+      "epoch": 5.858296792060928,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.773182121414277e-05,
+      "loss": 1.3232,
+      "step": 6346
+    },
+    {
+      "epoch": 5.859219939995384,
+      "grad_norm": 3.146484375,
+      "learning_rate": 5.7725150100066715e-05,
+      "loss": 1.2272,
+      "step": 6347
+    },
+    {
+      "epoch": 5.860143087929841,
+      "grad_norm": 3.62109375,
+      "learning_rate": 5.771847898599067e-05,
+      "loss": 1.302,
+      "step": 6348
+    },
+    {
+      "epoch": 5.861066235864297,
+      "grad_norm": 3.205078125,
+      "learning_rate": 5.771180787191461e-05,
+      "loss": 1.3567,
+      "step": 6349
+    },
+    {
+      "epoch": 5.861989383798754,
+      "grad_norm": 3.4609375,
+      "learning_rate": 5.7705136757838565e-05,
+      "loss": 1.4934,
+      "step": 6350
+    },
+    {
+      "epoch": 5.86291253173321,
+      "grad_norm": 3.6796875,
+      "learning_rate": 5.769846564376251e-05,
+      "loss": 1.2914,
+      "step": 6351
+    },
+    {
+      "epoch": 5.863835679667667,
+      "grad_norm": 3.724609375,
+      "learning_rate": 5.769179452968646e-05,
+      "loss": 1.3489,
+      "step": 6352
+    },
+    {
+      "epoch": 5.864758827602123,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.768512341561041e-05,
+      "loss": 1.376,
+      "step": 6353
+    },
+    {
+      "epoch": 5.86568197553658,
+      "grad_norm": 3.6171875,
+      "learning_rate": 5.767845230153436e-05,
+      "loss": 1.3723,
+      "step": 6354
+    },
+    {
+      "epoch": 5.866605123471036,
+      "grad_norm": 3.451171875,
+      "learning_rate": 5.7671781187458305e-05,
+      "loss": 1.2876,
+      "step": 6355
+    },
+    {
+      "epoch": 5.867528271405493,
+      "grad_norm": 3.52734375,
+      "learning_rate": 5.766511007338226e-05,
+      "loss": 1.3665,
+      "step": 6356
+    },
+    {
+      "epoch": 5.868451419339949,
+      "grad_norm": 3.154296875,
+      "learning_rate": 5.76584389593062e-05,
+      "loss": 1.2987,
+      "step": 6357
+    },
+    {
+      "epoch": 5.869374567274406,
+      "grad_norm": 3.826171875,
+      "learning_rate": 5.7651767845230155e-05,
+      "loss": 1.4956,
+      "step": 6358
+    },
+    {
+      "epoch": 5.870297715208862,
+      "grad_norm": 3.287109375,
+      "learning_rate": 5.76450967311541e-05,
+      "loss": 1.4644,
+      "step": 6359
+    },
+    {
+      "epoch": 5.871220863143319,
+      "grad_norm": 3.517578125,
+      "learning_rate": 5.763842561707805e-05,
+      "loss": 1.6211,
+      "step": 6360
+    },
+    {
+      "epoch": 5.872144011077776,
+      "grad_norm": 3.458984375,
+      "learning_rate": 5.7631754503002e-05,
+      "loss": 1.2588,
+      "step": 6361
+    },
+    {
+      "epoch": 5.873067159012232,
+      "grad_norm": 3.42578125,
+      "learning_rate": 5.762508338892595e-05,
+      "loss": 0.9788,
+      "step": 6362
+    },
+    {
+      "epoch": 5.8739903069466886,
+      "grad_norm": 4.1328125,
+      "learning_rate": 5.7618412274849896e-05,
+      "loss": 1.3589,
+      "step": 6363
+    },
+    {
+      "epoch": 5.874913454881145,
+      "grad_norm": 4.125,
+      "learning_rate": 5.761174116077385e-05,
+      "loss": 1.6099,
+      "step": 6364
+    },
+    {
+      "epoch": 5.8758366028156015,
+      "grad_norm": 3.732421875,
+      "learning_rate": 5.760507004669781e-05,
+      "loss": 1.5952,
+      "step": 6365
+    },
+    {
+      "epoch": 5.8767597507500575,
+      "grad_norm": 4.1953125,
+      "learning_rate": 5.759839893262175e-05,
+      "loss": 1.7529,
+      "step": 6366
+    },
+    {
+      "epoch": 5.877682898684514,
+      "grad_norm": 3.193359375,
+      "learning_rate": 5.7591727818545705e-05,
+      "loss": 1.5378,
+      "step": 6367
+    },
+    {
+      "epoch": 5.8786060466189705,
+      "grad_norm": 3.931640625,
+      "learning_rate": 5.758505670446965e-05,
+      "loss": 1.2942,
+      "step": 6368
+    },
+    {
+      "epoch": 5.879529194553427,
+      "grad_norm": 4.3671875,
+      "learning_rate": 5.75783855903936e-05,
+      "loss": 1.3708,
+      "step": 6369
+    },
+    {
+      "epoch": 5.880452342487883,
+      "grad_norm": 3.6953125,
+      "learning_rate": 5.757171447631755e-05,
+      "loss": 1.3035,
+      "step": 6370
+    },
+    {
+      "epoch": 5.88137549042234,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.75650433622415e-05,
+      "loss": 1.2815,
+      "step": 6371
+    },
+    {
+      "epoch": 5.882298638356796,
+      "grad_norm": 3.751953125,
+      "learning_rate": 5.7558372248165445e-05,
+      "loss": 1.4148,
+      "step": 6372
+    },
+    {
+      "epoch": 5.883221786291253,
+      "grad_norm": 3.255859375,
+      "learning_rate": 5.75517011340894e-05,
+      "loss": 1.0329,
+      "step": 6373
+    },
+    {
+      "epoch": 5.884144934225709,
+      "grad_norm": 3.25,
+      "learning_rate": 5.754503002001334e-05,
+      "loss": 1.7234,
+      "step": 6374
+    },
+    {
+      "epoch": 5.885068082160166,
+      "grad_norm": 3.8671875,
+      "learning_rate": 5.7538358905937295e-05,
+      "loss": 1.0155,
+      "step": 6375
+    },
+    {
+      "epoch": 5.885991230094623,
+      "grad_norm": 3.744140625,
+      "learning_rate": 5.753168779186125e-05,
+      "loss": 1.3413,
+      "step": 6376
+    },
+    {
+      "epoch": 5.886914378029079,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.752501667778519e-05,
+      "loss": 1.3076,
+      "step": 6377
+    },
+    {
+      "epoch": 5.887837525963536,
+      "grad_norm": 3.423828125,
+      "learning_rate": 5.7518345563709145e-05,
+      "loss": 1.2625,
+      "step": 6378
+    },
+    {
+      "epoch": 5.888760673897992,
+      "grad_norm": 3.166015625,
+      "learning_rate": 5.751167444963309e-05,
+      "loss": 1.381,
+      "step": 6379
+    },
+    {
+      "epoch": 5.889683821832449,
+      "grad_norm": 3.384765625,
+      "learning_rate": 5.750500333555704e-05,
+      "loss": 1.0643,
+      "step": 6380
+    },
+    {
+      "epoch": 5.890606969766905,
+      "grad_norm": 3.8359375,
+      "learning_rate": 5.749833222148099e-05,
+      "loss": 1.3899,
+      "step": 6381
+    },
+    {
+      "epoch": 5.891530117701362,
+      "grad_norm": 3.91796875,
+      "learning_rate": 5.749166110740494e-05,
+      "loss": 1.1931,
+      "step": 6382
+    },
+    {
+      "epoch": 5.892453265635818,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.7484989993328886e-05,
+      "loss": 1.6089,
+      "step": 6383
+    },
+    {
+      "epoch": 5.893376413570275,
+      "grad_norm": 3.69921875,
+      "learning_rate": 5.747831887925284e-05,
+      "loss": 1.3584,
+      "step": 6384
+    },
+    {
+      "epoch": 5.894299561504731,
+      "grad_norm": 3.697265625,
+      "learning_rate": 5.747164776517678e-05,
+      "loss": 1.323,
+      "step": 6385
+    },
+    {
+      "epoch": 5.895222709439188,
+      "grad_norm": 3.8046875,
+      "learning_rate": 5.7464976651100736e-05,
+      "loss": 1.2698,
+      "step": 6386
+    },
+    {
+      "epoch": 5.896145857373644,
+      "grad_norm": 2.880859375,
+      "learning_rate": 5.745830553702468e-05,
+      "loss": 1.277,
+      "step": 6387
+    },
+    {
+      "epoch": 5.897069005308101,
+      "grad_norm": 3.869140625,
+      "learning_rate": 5.745163442294863e-05,
+      "loss": 1.2642,
+      "step": 6388
+    },
+    {
+      "epoch": 5.897992153242557,
+      "grad_norm": 3.88671875,
+      "learning_rate": 5.744496330887258e-05,
+      "loss": 1.4675,
+      "step": 6389
+    },
+    {
+      "epoch": 5.898915301177014,
+      "grad_norm": 4.6015625,
+      "learning_rate": 5.743829219479654e-05,
+      "loss": 1.4456,
+      "step": 6390
+    },
+    {
+      "epoch": 5.899838449111471,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.743162108072049e-05,
+      "loss": 1.3579,
+      "step": 6391
+    },
+    {
+      "epoch": 5.900761597045927,
+      "grad_norm": 4.01171875,
+      "learning_rate": 5.7424949966644435e-05,
+      "loss": 1.5005,
+      "step": 6392
+    },
+    {
+      "epoch": 5.901684744980383,
+      "grad_norm": 4.22265625,
+      "learning_rate": 5.741827885256839e-05,
+      "loss": 1.0687,
+      "step": 6393
+    },
+    {
+      "epoch": 5.90260789291484,
+      "grad_norm": 3.75,
+      "learning_rate": 5.741160773849233e-05,
+      "loss": 1.6184,
+      "step": 6394
+    },
+    {
+      "epoch": 5.9035310408492965,
+      "grad_norm": 4.0703125,
+      "learning_rate": 5.7404936624416285e-05,
+      "loss": 1.3018,
+      "step": 6395
+    },
+    {
+      "epoch": 5.9044541887837525,
+      "grad_norm": 3.2421875,
+      "learning_rate": 5.739826551034023e-05,
+      "loss": 1.4033,
+      "step": 6396
+    },
+    {
+      "epoch": 5.905377336718209,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.739159439626418e-05,
+      "loss": 1.2996,
+      "step": 6397
+    },
+    {
+      "epoch": 5.906300484652665,
+      "grad_norm": 3.443359375,
+      "learning_rate": 5.738492328218813e-05,
+      "loss": 1.3245,
+      "step": 6398
+    },
+    {
+      "epoch": 5.907223632587122,
+      "grad_norm": 3.697265625,
+      "learning_rate": 5.737825216811208e-05,
+      "loss": 1.3831,
+      "step": 6399
+    },
+    {
+      "epoch": 5.908146780521578,
+      "grad_norm": 3.10546875,
+      "learning_rate": 5.7371581054036026e-05,
+      "loss": 1.1251,
+      "step": 6400
+    },
+    {
+      "epoch": 5.909069928456035,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.736490993995998e-05,
+      "loss": 1.3859,
+      "step": 6401
+    },
+    {
+      "epoch": 5.909993076390491,
+      "grad_norm": 3.541015625,
+      "learning_rate": 5.7358238825883923e-05,
+      "loss": 1.6841,
+      "step": 6402
+    },
+    {
+      "epoch": 5.910916224324948,
+      "grad_norm": 3.66015625,
+      "learning_rate": 5.7351567711807876e-05,
+      "loss": 1.3167,
+      "step": 6403
+    },
+    {
+      "epoch": 5.911839372259404,
+      "grad_norm": 3.21484375,
+      "learning_rate": 5.734489659773182e-05,
+      "loss": 1.4148,
+      "step": 6404
+    },
+    {
+      "epoch": 5.912762520193861,
+      "grad_norm": 3.341796875,
+      "learning_rate": 5.733822548365577e-05,
+      "loss": 1.5115,
+      "step": 6405
+    },
+    {
+      "epoch": 5.913685668128317,
+      "grad_norm": 3.87109375,
+      "learning_rate": 5.733155436957972e-05,
+      "loss": 1.262,
+      "step": 6406
+    },
+    {
+      "epoch": 5.914608816062774,
+      "grad_norm": 3.708984375,
+      "learning_rate": 5.732488325550367e-05,
+      "loss": 1.2744,
+      "step": 6407
+    },
+    {
+      "epoch": 5.91553196399723,
+      "grad_norm": 3.85546875,
+      "learning_rate": 5.7318212141427616e-05,
+      "loss": 1.324,
+      "step": 6408
+    },
+    {
+      "epoch": 5.916455111931687,
+      "grad_norm": 3.982421875,
+      "learning_rate": 5.731154102735157e-05,
+      "loss": 1.4036,
+      "step": 6409
+    },
+    {
+      "epoch": 5.917378259866144,
+      "grad_norm": 4.06640625,
+      "learning_rate": 5.7304869913275514e-05,
+      "loss": 1.3315,
+      "step": 6410
+    },
+    {
+      "epoch": 5.9183014078006,
+      "grad_norm": 3.634765625,
+      "learning_rate": 5.7298198799199466e-05,
+      "loss": 1.1603,
+      "step": 6411
+    },
+    {
+      "epoch": 5.919224555735057,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.729152768512341e-05,
+      "loss": 1.3672,
+      "step": 6412
+    },
+    {
+      "epoch": 5.920147703669513,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.7284856571047364e-05,
+      "loss": 1.4175,
+      "step": 6413
+    },
+    {
+      "epoch": 5.92107085160397,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.727818545697131e-05,
+      "loss": 1.1403,
+      "step": 6414
+    },
+    {
+      "epoch": 5.921993999538426,
+      "grad_norm": 3.326171875,
+      "learning_rate": 5.727151434289527e-05,
+      "loss": 1.616,
+      "step": 6415
+    },
+    {
+      "epoch": 5.922917147472883,
+      "grad_norm": 3.736328125,
+      "learning_rate": 5.726484322881922e-05,
+      "loss": 1.2317,
+      "step": 6416
+    },
+    {
+      "epoch": 5.923840295407339,
+      "grad_norm": 3.208984375,
+      "learning_rate": 5.725817211474317e-05,
+      "loss": 1.5032,
+      "step": 6417
+    },
+    {
+      "epoch": 5.924763443341796,
+      "grad_norm": 3.220703125,
+      "learning_rate": 5.725150100066712e-05,
+      "loss": 1.5132,
+      "step": 6418
+    },
+    {
+      "epoch": 5.925686591276252,
+      "grad_norm": 3.662109375,
+      "learning_rate": 5.724482988659107e-05,
+      "loss": 1.2578,
+      "step": 6419
+    },
+    {
+      "epoch": 5.926609739210709,
+      "grad_norm": 3.123046875,
+      "learning_rate": 5.7238158772515016e-05,
+      "loss": 1.5403,
+      "step": 6420
+    },
+    {
+      "epoch": 5.927532887145165,
+      "grad_norm": 3.55859375,
+      "learning_rate": 5.723148765843897e-05,
+      "loss": 1.4623,
+      "step": 6421
+    },
+    {
+      "epoch": 5.928456035079622,
+      "grad_norm": 3.845703125,
+      "learning_rate": 5.722481654436291e-05,
+      "loss": 1.4236,
+      "step": 6422
+    },
+    {
+      "epoch": 5.929379183014078,
+      "grad_norm": 3.8203125,
+      "learning_rate": 5.7218145430286866e-05,
+      "loss": 1.5566,
+      "step": 6423
+    },
+    {
+      "epoch": 5.9303023309485345,
+      "grad_norm": 3.666015625,
+      "learning_rate": 5.721147431621081e-05,
+      "loss": 1.3596,
+      "step": 6424
+    },
+    {
+      "epoch": 5.9312254788829915,
+      "grad_norm": 3.513671875,
+      "learning_rate": 5.720480320213476e-05,
+      "loss": 1.5532,
+      "step": 6425
+    },
+    {
+      "epoch": 5.9321486268174475,
+      "grad_norm": 3.044921875,
+      "learning_rate": 5.719813208805871e-05,
+      "loss": 1.2563,
+      "step": 6426
+    },
+    {
+      "epoch": 5.9330717747519035,
+      "grad_norm": 3.24609375,
+      "learning_rate": 5.719146097398266e-05,
+      "loss": 1.4033,
+      "step": 6427
+    },
+    {
+      "epoch": 5.93399492268636,
+      "grad_norm": 3.544921875,
+      "learning_rate": 5.7184789859906606e-05,
+      "loss": 1.4688,
+      "step": 6428
+    },
+    {
+      "epoch": 5.934918070620817,
+      "grad_norm": 3.23828125,
+      "learning_rate": 5.717811874583056e-05,
+      "loss": 1.3977,
+      "step": 6429
+    },
+    {
+      "epoch": 5.935841218555273,
+      "grad_norm": 4.453125,
+      "learning_rate": 5.7171447631754504e-05,
+      "loss": 1.1422,
+      "step": 6430
+    },
+    {
+      "epoch": 5.93676436648973,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.7164776517678456e-05,
+      "loss": 1.2386,
+      "step": 6431
+    },
+    {
+      "epoch": 5.937687514424186,
+      "grad_norm": 4.5390625,
+      "learning_rate": 5.71581054036024e-05,
+      "loss": 1.4329,
+      "step": 6432
+    },
+    {
+      "epoch": 5.938610662358643,
+      "grad_norm": 3.70703125,
+      "learning_rate": 5.7151434289526354e-05,
+      "loss": 1.4673,
+      "step": 6433
+    },
+    {
+      "epoch": 5.939533810293099,
+      "grad_norm": 3.1484375,
+      "learning_rate": 5.71447631754503e-05,
+      "loss": 1.5706,
+      "step": 6434
+    },
+    {
+      "epoch": 5.940456958227556,
+      "grad_norm": 3.392578125,
+      "learning_rate": 5.713809206137425e-05,
+      "loss": 1.2432,
+      "step": 6435
+    },
+    {
+      "epoch": 5.941380106162012,
+      "grad_norm": 4.015625,
+      "learning_rate": 5.71314209472982e-05,
+      "loss": 1.4132,
+      "step": 6436
+    },
+    {
+      "epoch": 5.942303254096469,
+      "grad_norm": 3.1640625,
+      "learning_rate": 5.712474983322215e-05,
+      "loss": 1.6086,
+      "step": 6437
+    },
+    {
+      "epoch": 5.943226402030925,
+      "grad_norm": 3.58984375,
+      "learning_rate": 5.7118078719146094e-05,
+      "loss": 1.3835,
+      "step": 6438
+    },
+    {
+      "epoch": 5.944149549965382,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.711140760507005e-05,
+      "loss": 1.4609,
+      "step": 6439
+    },
+    {
+      "epoch": 5.945072697899839,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.7104736490994006e-05,
+      "loss": 1.3774,
+      "step": 6440
+    },
+    {
+      "epoch": 5.945995845834295,
+      "grad_norm": 3.66015625,
+      "learning_rate": 5.709806537691795e-05,
+      "loss": 1.2893,
+      "step": 6441
+    },
+    {
+      "epoch": 5.946918993768751,
+      "grad_norm": 3.216796875,
+      "learning_rate": 5.70913942628419e-05,
+      "loss": 1.1987,
+      "step": 6442
+    },
+    {
+      "epoch": 5.947842141703208,
+      "grad_norm": 3.603515625,
+      "learning_rate": 5.708472314876585e-05,
+      "loss": 1.1254,
+      "step": 6443
+    },
+    {
+      "epoch": 5.948765289637665,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.70780520346898e-05,
+      "loss": 1.291,
+      "step": 6444
+    },
+    {
+      "epoch": 5.949688437572121,
+      "grad_norm": 3.8828125,
+      "learning_rate": 5.7071380920613746e-05,
+      "loss": 1.407,
+      "step": 6445
+    },
+    {
+      "epoch": 5.950611585506578,
+      "grad_norm": 3.181640625,
+      "learning_rate": 5.70647098065377e-05,
+      "loss": 1.4441,
+      "step": 6446
+    },
+    {
+      "epoch": 5.951534733441034,
+      "grad_norm": 4.1640625,
+      "learning_rate": 5.7058038692461644e-05,
+      "loss": 1.521,
+      "step": 6447
+    },
+    {
+      "epoch": 5.952457881375491,
+      "grad_norm": 3.8203125,
+      "learning_rate": 5.7051367578385596e-05,
+      "loss": 1.5659,
+      "step": 6448
+    },
+    {
+      "epoch": 5.953381029309947,
+      "grad_norm": 3.876953125,
+      "learning_rate": 5.704469646430954e-05,
+      "loss": 1.399,
+      "step": 6449
+    },
+    {
+      "epoch": 5.954304177244404,
+      "grad_norm": 3.45703125,
+      "learning_rate": 5.7038025350233494e-05,
+      "loss": 1.255,
+      "step": 6450
+    },
+    {
+      "epoch": 5.95522732517886,
+      "grad_norm": 3.89453125,
+      "learning_rate": 5.703135423615744e-05,
+      "loss": 1.4108,
+      "step": 6451
+    },
+    {
+      "epoch": 5.956150473113317,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.702468312208139e-05,
+      "loss": 1.4094,
+      "step": 6452
+    },
+    {
+      "epoch": 5.957073621047773,
+      "grad_norm": 3.4140625,
+      "learning_rate": 5.701801200800534e-05,
+      "loss": 0.8958,
+      "step": 6453
+    },
+    {
+      "epoch": 5.9579967689822295,
+      "grad_norm": 3.470703125,
+      "learning_rate": 5.701134089392929e-05,
+      "loss": 1.5522,
+      "step": 6454
+    },
+    {
+      "epoch": 5.9589199169166855,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.7004669779853235e-05,
+      "loss": 1.3333,
+      "step": 6455
+    },
+    {
+      "epoch": 5.9598430648511425,
+      "grad_norm": 3.4453125,
+      "learning_rate": 5.699799866577719e-05,
+      "loss": 1.5562,
+      "step": 6456
+    },
+    {
+      "epoch": 5.9607662127855985,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.699132755170113e-05,
+      "loss": 1.2629,
+      "step": 6457
+    },
+    {
+      "epoch": 5.961689360720055,
+      "grad_norm": 3.537109375,
+      "learning_rate": 5.6984656437625084e-05,
+      "loss": 1.4275,
+      "step": 6458
+    },
+    {
+      "epoch": 5.962612508654512,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.697798532354903e-05,
+      "loss": 1.4573,
+      "step": 6459
+    },
+    {
+      "epoch": 5.963535656588968,
+      "grad_norm": 3.3203125,
+      "learning_rate": 5.697131420947298e-05,
+      "loss": 1.3984,
+      "step": 6460
+    },
+    {
+      "epoch": 5.964458804523425,
+      "grad_norm": 4.078125,
+      "learning_rate": 5.696464309539693e-05,
+      "loss": 1.344,
+      "step": 6461
+    },
+    {
+      "epoch": 5.965381952457881,
+      "grad_norm": 3.46484375,
+      "learning_rate": 5.695797198132088e-05,
+      "loss": 1.3562,
+      "step": 6462
+    },
+    {
+      "epoch": 5.966305100392338,
+      "grad_norm": 3.70703125,
+      "learning_rate": 5.6951300867244825e-05,
+      "loss": 1.4802,
+      "step": 6463
+    },
+    {
+      "epoch": 5.967228248326794,
+      "grad_norm": 3.634765625,
+      "learning_rate": 5.694462975316878e-05,
+      "loss": 1.5486,
+      "step": 6464
+    },
+    {
+      "epoch": 5.968151396261251,
+      "grad_norm": 3.455078125,
+      "learning_rate": 5.6937958639092736e-05,
+      "loss": 1.3638,
+      "step": 6465
+    },
+    {
+      "epoch": 5.969074544195707,
+      "grad_norm": 4.23046875,
+      "learning_rate": 5.693128752501669e-05,
+      "loss": 1.113,
+      "step": 6466
+    },
+    {
+      "epoch": 5.969997692130164,
+      "grad_norm": 3.162109375,
+      "learning_rate": 5.6924616410940634e-05,
+      "loss": 1.3435,
+      "step": 6467
+    },
+    {
+      "epoch": 5.97092084006462,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.6917945296864586e-05,
+      "loss": 1.1125,
+      "step": 6468
+    },
+    {
+      "epoch": 5.971843987999077,
+      "grad_norm": 3.91796875,
+      "learning_rate": 5.691127418278853e-05,
+      "loss": 1.4165,
+      "step": 6469
+    },
+    {
+      "epoch": 5.972767135933533,
+      "grad_norm": 4.48046875,
+      "learning_rate": 5.6904603068712484e-05,
+      "loss": 1.4619,
+      "step": 6470
+    },
+    {
+      "epoch": 5.97369028386799,
+      "grad_norm": 3.615234375,
+      "learning_rate": 5.689793195463643e-05,
+      "loss": 1.1654,
+      "step": 6471
+    },
+    {
+      "epoch": 5.974613431802446,
+      "grad_norm": 3.873046875,
+      "learning_rate": 5.689126084056038e-05,
+      "loss": 1.4265,
+      "step": 6472
+    },
+    {
+      "epoch": 5.975536579736903,
+      "grad_norm": 3.2734375,
+      "learning_rate": 5.688458972648433e-05,
+      "loss": 1.4819,
+      "step": 6473
+    },
+    {
+      "epoch": 5.97645972767136,
+      "grad_norm": 3.212890625,
+      "learning_rate": 5.687791861240828e-05,
+      "loss": 1.0986,
+      "step": 6474
+    },
+    {
+      "epoch": 5.977382875605816,
+      "grad_norm": 3.271484375,
+      "learning_rate": 5.6871247498332224e-05,
+      "loss": 1.3191,
+      "step": 6475
+    },
+    {
+      "epoch": 5.978306023540272,
+      "grad_norm": 4.31640625,
+      "learning_rate": 5.686457638425618e-05,
+      "loss": 1.3066,
+      "step": 6476
+    },
+    {
+      "epoch": 5.979229171474729,
+      "grad_norm": 4.72265625,
+      "learning_rate": 5.685790527018012e-05,
+      "loss": 1.4717,
+      "step": 6477
+    },
+    {
+      "epoch": 5.980152319409186,
+      "grad_norm": 3.12890625,
+      "learning_rate": 5.6851234156104074e-05,
+      "loss": 1.3613,
+      "step": 6478
+    },
+    {
+      "epoch": 5.981075467343642,
+      "grad_norm": 4.12890625,
+      "learning_rate": 5.684456304202802e-05,
+      "loss": 1.4285,
+      "step": 6479
+    },
+    {
+      "epoch": 5.981998615278099,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.683789192795197e-05,
+      "loss": 1.6028,
+      "step": 6480
+    },
+    {
+      "epoch": 5.982921763212555,
+      "grad_norm": 4.09765625,
+      "learning_rate": 5.683122081387592e-05,
+      "loss": 1.2584,
+      "step": 6481
+    },
+    {
+      "epoch": 5.983844911147012,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.682454969979987e-05,
+      "loss": 1.4697,
+      "step": 6482
+    },
+    {
+      "epoch": 5.984768059081468,
+      "grad_norm": 3.57421875,
+      "learning_rate": 5.6817878585723815e-05,
+      "loss": 1.3794,
+      "step": 6483
+    },
+    {
+      "epoch": 5.9856912070159245,
+      "grad_norm": 3.42578125,
+      "learning_rate": 5.681120747164777e-05,
+      "loss": 1.4232,
+      "step": 6484
+    },
+    {
+      "epoch": 5.9866143549503805,
+      "grad_norm": 3.705078125,
+      "learning_rate": 5.680453635757171e-05,
+      "loss": 1.4939,
+      "step": 6485
+    },
+    {
+      "epoch": 5.987537502884837,
+      "grad_norm": 3.205078125,
+      "learning_rate": 5.6797865243495665e-05,
+      "loss": 1.2886,
+      "step": 6486
+    },
+    {
+      "epoch": 5.9884606508192935,
+      "grad_norm": 3.396484375,
+      "learning_rate": 5.679119412941961e-05,
+      "loss": 1.3137,
+      "step": 6487
+    },
+    {
+      "epoch": 5.98938379875375,
+      "grad_norm": 3.64453125,
+      "learning_rate": 5.678452301534356e-05,
+      "loss": 1.2562,
+      "step": 6488
+    },
+    {
+      "epoch": 5.990306946688206,
+      "grad_norm": 3.54296875,
+      "learning_rate": 5.677785190126751e-05,
+      "loss": 1.6001,
+      "step": 6489
+    },
+    {
+      "epoch": 5.991230094622663,
+      "grad_norm": 3.833984375,
+      "learning_rate": 5.677118078719147e-05,
+      "loss": 1.5928,
+      "step": 6490
+    },
+    {
+      "epoch": 5.992153242557119,
+      "grad_norm": 3.61328125,
+      "learning_rate": 5.676450967311542e-05,
+      "loss": 1.4978,
+      "step": 6491
+    },
+    {
+      "epoch": 5.993076390491576,
+      "grad_norm": 3.126953125,
+      "learning_rate": 5.6757838559039365e-05,
+      "loss": 1.3796,
+      "step": 6492
+    },
+    {
+      "epoch": 5.993999538426033,
+      "grad_norm": 3.84765625,
+      "learning_rate": 5.675116744496332e-05,
+      "loss": 1.5457,
+      "step": 6493
+    },
+    {
+      "epoch": 5.994922686360489,
+      "grad_norm": 3.37109375,
+      "learning_rate": 5.674449633088726e-05,
+      "loss": 1.402,
+      "step": 6494
+    },
+    {
+      "epoch": 5.995845834294946,
+      "grad_norm": 3.408203125,
+      "learning_rate": 5.6737825216811214e-05,
+      "loss": 1.1611,
+      "step": 6495
+    },
+    {
+      "epoch": 5.996768982229402,
+      "grad_norm": 3.599609375,
+      "learning_rate": 5.673115410273516e-05,
+      "loss": 1.772,
+      "step": 6496
+    },
+    {
+      "epoch": 5.997692130163859,
+      "grad_norm": 3.9296875,
+      "learning_rate": 5.672448298865911e-05,
+      "loss": 1.3123,
+      "step": 6497
+    },
+    {
+      "epoch": 5.998615278098315,
+      "grad_norm": 3.689453125,
+      "learning_rate": 5.671781187458306e-05,
+      "loss": 1.4548,
+      "step": 6498
+    },
+    {
+      "epoch": 5.999538426032772,
+      "grad_norm": 3.60546875,
+      "learning_rate": 5.671114076050701e-05,
+      "loss": 1.3698,
+      "step": 6499
+    },
+    {
+      "epoch": 6.000461573967228,
+      "grad_norm": 3.048828125,
+      "learning_rate": 5.6704469646430955e-05,
+      "loss": 0.9334,
+      "step": 6500
+    },
+    {
+      "epoch": 6.001384721901685,
+      "grad_norm": 3.69921875,
+      "learning_rate": 5.669779853235491e-05,
+      "loss": 1.3837,
+      "step": 6501
+    },
+    {
+      "epoch": 6.002307869836141,
+      "grad_norm": 4.18359375,
+      "learning_rate": 5.669112741827885e-05,
+      "loss": 1.2385,
+      "step": 6502
+    },
+    {
+      "epoch": 6.003231017770598,
+      "grad_norm": 3.5703125,
+      "learning_rate": 5.6684456304202805e-05,
+      "loss": 1.1698,
+      "step": 6503
+    },
+    {
+      "epoch": 6.004154165705054,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.667778519012675e-05,
+      "loss": 1.3308,
+      "step": 6504
+    },
+    {
+      "epoch": 6.005077313639511,
+      "grad_norm": 3.908203125,
+      "learning_rate": 5.66711140760507e-05,
+      "loss": 1.2617,
+      "step": 6505
+    },
+    {
+      "epoch": 6.006000461573967,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.666444296197465e-05,
+      "loss": 1.4382,
+      "step": 6506
+    },
+    {
+      "epoch": 6.006923609508424,
+      "grad_norm": 3.263671875,
+      "learning_rate": 5.66577718478986e-05,
+      "loss": 1.4885,
+      "step": 6507
+    },
+    {
+      "epoch": 6.007846757442881,
+      "grad_norm": 3.419921875,
+      "learning_rate": 5.6651100733822546e-05,
+      "loss": 1.302,
+      "step": 6508
+    },
+    {
+      "epoch": 6.008769905377337,
+      "grad_norm": 3.0859375,
+      "learning_rate": 5.66444296197465e-05,
+      "loss": 1.2512,
+      "step": 6509
+    },
+    {
+      "epoch": 6.009693053311794,
+      "grad_norm": 3.44140625,
+      "learning_rate": 5.663775850567044e-05,
+      "loss": 1.1892,
+      "step": 6510
+    },
+    {
+      "epoch": 6.01061620124625,
+      "grad_norm": 3.81640625,
+      "learning_rate": 5.6631087391594396e-05,
+      "loss": 1.2234,
+      "step": 6511
+    },
+    {
+      "epoch": 6.0115393491807065,
+      "grad_norm": 3.298828125,
+      "learning_rate": 5.662441627751834e-05,
+      "loss": 1.4414,
+      "step": 6512
+    },
+    {
+      "epoch": 6.012462497115163,
+      "grad_norm": 4.00390625,
+      "learning_rate": 5.661774516344229e-05,
+      "loss": 1.0004,
+      "step": 6513
+    },
+    {
+      "epoch": 6.0133856450496195,
+      "grad_norm": 3.548828125,
+      "learning_rate": 5.661107404936624e-05,
+      "loss": 1.6731,
+      "step": 6514
+    },
+    {
+      "epoch": 6.0143087929840755,
+      "grad_norm": 3.44921875,
+      "learning_rate": 5.6604402935290204e-05,
+      "loss": 1.2571,
+      "step": 6515
+    },
+    {
+      "epoch": 6.015231940918532,
+      "grad_norm": 3.544921875,
+      "learning_rate": 5.659773182121415e-05,
+      "loss": 1.2964,
+      "step": 6516
+    },
+    {
+      "epoch": 6.016155088852988,
+      "grad_norm": 3.177734375,
+      "learning_rate": 5.65910607071381e-05,
+      "loss": 1.6907,
+      "step": 6517
+    },
+    {
+      "epoch": 6.017078236787445,
+      "grad_norm": 3.693359375,
+      "learning_rate": 5.658438959306205e-05,
+      "loss": 1.345,
+      "step": 6518
+    },
+    {
+      "epoch": 6.018001384721901,
+      "grad_norm": 3.0390625,
+      "learning_rate": 5.6577718478986e-05,
+      "loss": 1.1409,
+      "step": 6519
+    },
+    {
+      "epoch": 6.018924532656358,
+      "grad_norm": 3.7890625,
+      "learning_rate": 5.6571047364909945e-05,
+      "loss": 1.522,
+      "step": 6520
+    },
+    {
+      "epoch": 6.019847680590814,
+      "grad_norm": 3.509765625,
+      "learning_rate": 5.65643762508339e-05,
+      "loss": 1.3784,
+      "step": 6521
+    },
+    {
+      "epoch": 6.020770828525271,
+      "grad_norm": 3.900390625,
+      "learning_rate": 5.655770513675784e-05,
+      "loss": 1.3804,
+      "step": 6522
+    },
+    {
+      "epoch": 6.021693976459727,
+      "grad_norm": 3.66796875,
+      "learning_rate": 5.6551034022681795e-05,
+      "loss": 1.6667,
+      "step": 6523
+    },
+    {
+      "epoch": 6.022617124394184,
+      "grad_norm": 4.484375,
+      "learning_rate": 5.654436290860574e-05,
+      "loss": 1.3579,
+      "step": 6524
+    },
+    {
+      "epoch": 6.023540272328641,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.653769179452969e-05,
+      "loss": 1.3259,
+      "step": 6525
+    },
+    {
+      "epoch": 6.024463420263097,
+      "grad_norm": 3.248046875,
+      "learning_rate": 5.653102068045364e-05,
+      "loss": 1.5596,
+      "step": 6526
+    },
+    {
+      "epoch": 6.025386568197554,
+      "grad_norm": 3.5703125,
+      "learning_rate": 5.652434956637759e-05,
+      "loss": 1.3937,
+      "step": 6527
+    },
+    {
+      "epoch": 6.02630971613201,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.6517678452301536e-05,
+      "loss": 1.4006,
+      "step": 6528
+    },
+    {
+      "epoch": 6.027232864066467,
+      "grad_norm": 4.3203125,
+      "learning_rate": 5.651100733822549e-05,
+      "loss": 1.4358,
+      "step": 6529
+    },
+    {
+      "epoch": 6.028156012000923,
+      "grad_norm": 3.658203125,
+      "learning_rate": 5.650433622414943e-05,
+      "loss": 1.4285,
+      "step": 6530
+    },
+    {
+      "epoch": 6.02907915993538,
+      "grad_norm": 3.611328125,
+      "learning_rate": 5.6497665110073385e-05,
+      "loss": 1.2913,
+      "step": 6531
+    },
+    {
+      "epoch": 6.030002307869836,
+      "grad_norm": 3.669921875,
+      "learning_rate": 5.649099399599733e-05,
+      "loss": 1.5435,
+      "step": 6532
+    },
+    {
+      "epoch": 6.030925455804293,
+      "grad_norm": 4.546875,
+      "learning_rate": 5.648432288192128e-05,
+      "loss": 1.4802,
+      "step": 6533
+    },
+    {
+      "epoch": 6.031848603738749,
+      "grad_norm": 2.982421875,
+      "learning_rate": 5.647765176784523e-05,
+      "loss": 1.1816,
+      "step": 6534
+    },
+    {
+      "epoch": 6.032771751673206,
+      "grad_norm": 4.0078125,
+      "learning_rate": 5.647098065376918e-05,
+      "loss": 1.4041,
+      "step": 6535
+    },
+    {
+      "epoch": 6.033694899607662,
+      "grad_norm": 3.27734375,
+      "learning_rate": 5.6464309539693126e-05,
+      "loss": 1.3955,
+      "step": 6536
+    },
+    {
+      "epoch": 6.034618047542119,
+      "grad_norm": 4.08984375,
+      "learning_rate": 5.645763842561708e-05,
+      "loss": 1.2653,
+      "step": 6537
+    },
+    {
+      "epoch": 6.035541195476575,
+      "grad_norm": 4.0078125,
+      "learning_rate": 5.6450967311541024e-05,
+      "loss": 1.3337,
+      "step": 6538
+    },
+    {
+      "epoch": 6.036464343411032,
+      "grad_norm": 3.96484375,
+      "learning_rate": 5.6444296197464976e-05,
+      "loss": 1.354,
+      "step": 6539
+    },
+    {
+      "epoch": 6.037387491345488,
+      "grad_norm": 3.736328125,
+      "learning_rate": 5.6437625083388935e-05,
+      "loss": 1.4319,
+      "step": 6540
+    },
+    {
+      "epoch": 6.038310639279945,
+      "grad_norm": 3.33203125,
+      "learning_rate": 5.643095396931288e-05,
+      "loss": 1.2415,
+      "step": 6541
+    },
+    {
+      "epoch": 6.0392337872144015,
+      "grad_norm": 4.22265625,
+      "learning_rate": 5.642428285523683e-05,
+      "loss": 1.376,
+      "step": 6542
+    },
+    {
+      "epoch": 6.0401569351488575,
+      "grad_norm": 4.56640625,
+      "learning_rate": 5.641761174116078e-05,
+      "loss": 1.4858,
+      "step": 6543
+    },
+    {
+      "epoch": 6.0410800830833145,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.641094062708473e-05,
+      "loss": 1.4487,
+      "step": 6544
+    },
+    {
+      "epoch": 6.0420032310177705,
+      "grad_norm": 4.5859375,
+      "learning_rate": 5.6404269513008676e-05,
+      "loss": 1.521,
+      "step": 6545
+    },
+    {
+      "epoch": 6.042926378952227,
+      "grad_norm": 3.701171875,
+      "learning_rate": 5.639759839893263e-05,
+      "loss": 1.218,
+      "step": 6546
+    },
+    {
+      "epoch": 6.043849526886683,
+      "grad_norm": 3.939453125,
+      "learning_rate": 5.639092728485657e-05,
+      "loss": 1.2557,
+      "step": 6547
+    },
+    {
+      "epoch": 6.04477267482114,
+      "grad_norm": 4.56640625,
+      "learning_rate": 5.6384256170780526e-05,
+      "loss": 1.2566,
+      "step": 6548
+    },
+    {
+      "epoch": 6.045695822755596,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.637758505670447e-05,
+      "loss": 1.4153,
+      "step": 6549
+    },
+    {
+      "epoch": 6.046618970690053,
+      "grad_norm": 4.1796875,
+      "learning_rate": 5.637091394262842e-05,
+      "loss": 1.243,
+      "step": 6550
+    },
+    {
+      "epoch": 6.047542118624509,
+      "grad_norm": 3.462890625,
+      "learning_rate": 5.636424282855237e-05,
+      "loss": 1.3096,
+      "step": 6551
+    },
+    {
+      "epoch": 6.048465266558966,
+      "grad_norm": 3.333984375,
+      "learning_rate": 5.635757171447632e-05,
+      "loss": 1.2147,
+      "step": 6552
+    },
+    {
+      "epoch": 6.049388414493422,
+      "grad_norm": 3.66015625,
+      "learning_rate": 5.6350900600400266e-05,
+      "loss": 1.1572,
+      "step": 6553
+    },
+    {
+      "epoch": 6.050311562427879,
+      "grad_norm": 3.6640625,
+      "learning_rate": 5.634422948632422e-05,
+      "loss": 1.3474,
+      "step": 6554
+    },
+    {
+      "epoch": 6.051234710362335,
+      "grad_norm": 3.740234375,
+      "learning_rate": 5.6337558372248164e-05,
+      "loss": 1.228,
+      "step": 6555
+    },
+    {
+      "epoch": 6.052157858296792,
+      "grad_norm": 3.509765625,
+      "learning_rate": 5.6330887258172116e-05,
+      "loss": 1.2548,
+      "step": 6556
+    },
+    {
+      "epoch": 6.053081006231248,
+      "grad_norm": 4.390625,
+      "learning_rate": 5.632421614409606e-05,
+      "loss": 1.4719,
+      "step": 6557
+    },
+    {
+      "epoch": 6.054004154165705,
+      "grad_norm": 3.478515625,
+      "learning_rate": 5.6317545030020014e-05,
+      "loss": 1.4717,
+      "step": 6558
+    },
+    {
+      "epoch": 6.054927302100162,
+      "grad_norm": 3.8046875,
+      "learning_rate": 5.631087391594396e-05,
+      "loss": 1.2478,
+      "step": 6559
+    },
+    {
+      "epoch": 6.055850450034618,
+      "grad_norm": 3.673828125,
+      "learning_rate": 5.630420280186791e-05,
+      "loss": 1.4001,
+      "step": 6560
+    },
+    {
+      "epoch": 6.056773597969075,
+      "grad_norm": 3.56640625,
+      "learning_rate": 5.629753168779186e-05,
+      "loss": 1.2053,
+      "step": 6561
+    },
+    {
+      "epoch": 6.057696745903531,
+      "grad_norm": 4.34765625,
+      "learning_rate": 5.629086057371581e-05,
+      "loss": 1.5693,
+      "step": 6562
+    },
+    {
+      "epoch": 6.058619893837988,
+      "grad_norm": 3.830078125,
+      "learning_rate": 5.628418945963976e-05,
+      "loss": 1.6392,
+      "step": 6563
+    },
+    {
+      "epoch": 6.059543041772444,
+      "grad_norm": 3.75,
+      "learning_rate": 5.627751834556371e-05,
+      "loss": 1.4817,
+      "step": 6564
+    },
+    {
+      "epoch": 6.060466189706901,
+      "grad_norm": 4.42578125,
+      "learning_rate": 5.6270847231487666e-05,
+      "loss": 1.215,
+      "step": 6565
+    },
+    {
+      "epoch": 6.061389337641357,
+      "grad_norm": 3.99609375,
+      "learning_rate": 5.626417611741162e-05,
+      "loss": 1.0902,
+      "step": 6566
+    },
+    {
+      "epoch": 6.062312485575814,
+      "grad_norm": 3.599609375,
+      "learning_rate": 5.625750500333556e-05,
+      "loss": 1.5076,
+      "step": 6567
+    },
+    {
+      "epoch": 6.06323563351027,
+      "grad_norm": 3.787109375,
+      "learning_rate": 5.6250833889259515e-05,
+      "loss": 1.3738,
+      "step": 6568
+    },
+    {
+      "epoch": 6.064158781444727,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.624416277518346e-05,
+      "loss": 1.3789,
+      "step": 6569
+    },
+    {
+      "epoch": 6.065081929379183,
+      "grad_norm": 3.478515625,
+      "learning_rate": 5.623749166110741e-05,
+      "loss": 1.0191,
+      "step": 6570
+    },
+    {
+      "epoch": 6.06600507731364,
+      "grad_norm": 3.70703125,
+      "learning_rate": 5.623082054703136e-05,
+      "loss": 1.3757,
+      "step": 6571
+    },
+    {
+      "epoch": 6.066928225248096,
+      "grad_norm": 3.455078125,
+      "learning_rate": 5.622414943295531e-05,
+      "loss": 1.0402,
+      "step": 6572
+    },
+    {
+      "epoch": 6.0678513731825525,
+      "grad_norm": 3.587890625,
+      "learning_rate": 5.6217478318879256e-05,
+      "loss": 1.3943,
+      "step": 6573
+    },
+    {
+      "epoch": 6.068774521117009,
+      "grad_norm": 4.125,
+      "learning_rate": 5.621080720480321e-05,
+      "loss": 1.4387,
+      "step": 6574
+    },
+    {
+      "epoch": 6.0696976690514655,
+      "grad_norm": 3.34765625,
+      "learning_rate": 5.6204136090727154e-05,
+      "loss": 1.2878,
+      "step": 6575
+    },
+    {
+      "epoch": 6.070620816985922,
+      "grad_norm": 3.02734375,
+      "learning_rate": 5.6197464976651106e-05,
+      "loss": 1.3713,
+      "step": 6576
+    },
+    {
+      "epoch": 6.071543964920378,
+      "grad_norm": 3.40234375,
+      "learning_rate": 5.619079386257505e-05,
+      "loss": 1.478,
+      "step": 6577
+    },
+    {
+      "epoch": 6.072467112854835,
+      "grad_norm": 3.576171875,
+      "learning_rate": 5.6184122748499004e-05,
+      "loss": 1.4944,
+      "step": 6578
+    },
+    {
+      "epoch": 6.073390260789291,
+      "grad_norm": 3.75390625,
+      "learning_rate": 5.617745163442295e-05,
+      "loss": 1.4446,
+      "step": 6579
+    },
+    {
+      "epoch": 6.074313408723748,
+      "grad_norm": 3.322265625,
+      "learning_rate": 5.61707805203469e-05,
+      "loss": 1.4409,
+      "step": 6580
+    },
+    {
+      "epoch": 6.075236556658204,
+      "grad_norm": 3.533203125,
+      "learning_rate": 5.616410940627085e-05,
+      "loss": 1.2139,
+      "step": 6581
+    },
+    {
+      "epoch": 6.076159704592661,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.61574382921948e-05,
+      "loss": 1.5632,
+      "step": 6582
+    },
+    {
+      "epoch": 6.077082852527117,
+      "grad_norm": 3.41015625,
+      "learning_rate": 5.6150767178118744e-05,
+      "loss": 1.2339,
+      "step": 6583
+    },
+    {
+      "epoch": 6.078006000461574,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.6144096064042697e-05,
+      "loss": 1.2891,
+      "step": 6584
+    },
+    {
+      "epoch": 6.07892914839603,
+      "grad_norm": 3.27734375,
+      "learning_rate": 5.613742494996664e-05,
+      "loss": 1.4846,
+      "step": 6585
+    },
+    {
+      "epoch": 6.079852296330487,
+      "grad_norm": 4.29296875,
+      "learning_rate": 5.6130753835890594e-05,
+      "loss": 1.4346,
+      "step": 6586
+    },
+    {
+      "epoch": 6.080775444264943,
+      "grad_norm": 3.294921875,
+      "learning_rate": 5.612408272181454e-05,
+      "loss": 1.2642,
+      "step": 6587
+    },
+    {
+      "epoch": 6.0816985921994,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.611741160773849e-05,
+      "loss": 1.4661,
+      "step": 6588
+    },
+    {
+      "epoch": 6.082621740133856,
+      "grad_norm": 3.439453125,
+      "learning_rate": 5.611074049366244e-05,
+      "loss": 1.3259,
+      "step": 6589
+    },
+    {
+      "epoch": 6.083544888068313,
+      "grad_norm": 3.21484375,
+      "learning_rate": 5.610406937958639e-05,
+      "loss": 1.3579,
+      "step": 6590
+    },
+    {
+      "epoch": 6.08446803600277,
+      "grad_norm": 3.556640625,
+      "learning_rate": 5.609739826551035e-05,
+      "loss": 1.7202,
+      "step": 6591
+    },
+    {
+      "epoch": 6.085391183937226,
+      "grad_norm": 2.830078125,
+      "learning_rate": 5.6090727151434294e-05,
+      "loss": 1.4438,
+      "step": 6592
+    },
+    {
+      "epoch": 6.086314331871683,
+      "grad_norm": 4.23828125,
+      "learning_rate": 5.6084056037358246e-05,
+      "loss": 1.4465,
+      "step": 6593
+    },
+    {
+      "epoch": 6.087237479806139,
+      "grad_norm": 3.94921875,
+      "learning_rate": 5.607738492328219e-05,
+      "loss": 1.668,
+      "step": 6594
+    },
+    {
+      "epoch": 6.088160627740596,
+      "grad_norm": 3.76953125,
+      "learning_rate": 5.6070713809206144e-05,
+      "loss": 1.2848,
+      "step": 6595
+    },
+    {
+      "epoch": 6.089083775675052,
+      "grad_norm": 3.685546875,
+      "learning_rate": 5.606404269513009e-05,
+      "loss": 1.4996,
+      "step": 6596
+    },
+    {
+      "epoch": 6.090006923609509,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.605737158105404e-05,
+      "loss": 1.6687,
+      "step": 6597
+    },
+    {
+      "epoch": 6.090930071543965,
+      "grad_norm": 3.55078125,
+      "learning_rate": 5.605070046697799e-05,
+      "loss": 1.5156,
+      "step": 6598
+    },
+    {
+      "epoch": 6.091853219478422,
+      "grad_norm": 3.88671875,
+      "learning_rate": 5.604402935290194e-05,
+      "loss": 1.0724,
+      "step": 6599
+    },
+    {
+      "epoch": 6.092776367412878,
+      "grad_norm": 3.7578125,
+      "learning_rate": 5.6037358238825884e-05,
+      "loss": 1.3264,
+      "step": 6600
+    },
+    {
+      "epoch": 6.093699515347335,
+      "grad_norm": 3.650390625,
+      "learning_rate": 5.603068712474984e-05,
+      "loss": 1.265,
+      "step": 6601
+    },
+    {
+      "epoch": 6.094622663281791,
+      "grad_norm": 3.791015625,
+      "learning_rate": 5.602401601067379e-05,
+      "loss": 1.3242,
+      "step": 6602
+    },
+    {
+      "epoch": 6.0955458112162475,
+      "grad_norm": 4.25,
+      "learning_rate": 5.6017344896597734e-05,
+      "loss": 1.605,
+      "step": 6603
+    },
+    {
+      "epoch": 6.0964689591507035,
+      "grad_norm": 3.294921875,
+      "learning_rate": 5.6010673782521686e-05,
+      "loss": 1.5261,
+      "step": 6604
+    },
+    {
+      "epoch": 6.09739210708516,
+      "grad_norm": 3.2265625,
+      "learning_rate": 5.600400266844563e-05,
+      "loss": 1.3562,
+      "step": 6605
+    },
+    {
+      "epoch": 6.0983152550196165,
+      "grad_norm": 3.23046875,
+      "learning_rate": 5.5997331554369584e-05,
+      "loss": 1.1987,
+      "step": 6606
+    },
+    {
+      "epoch": 6.099238402954073,
+      "grad_norm": 4.44140625,
+      "learning_rate": 5.599066044029353e-05,
+      "loss": 1.3687,
+      "step": 6607
+    },
+    {
+      "epoch": 6.10016155088853,
+      "grad_norm": 3.658203125,
+      "learning_rate": 5.598398932621748e-05,
+      "loss": 1.3809,
+      "step": 6608
+    },
+    {
+      "epoch": 6.101084698822986,
+      "grad_norm": 4.2890625,
+      "learning_rate": 5.597731821214143e-05,
+      "loss": 1.3494,
+      "step": 6609
+    },
+    {
+      "epoch": 6.102007846757443,
+      "grad_norm": 3.45703125,
+      "learning_rate": 5.597064709806538e-05,
+      "loss": 1.4607,
+      "step": 6610
+    },
+    {
+      "epoch": 6.102930994691899,
+      "grad_norm": 3.056640625,
+      "learning_rate": 5.5963975983989325e-05,
+      "loss": 1.385,
+      "step": 6611
+    },
+    {
+      "epoch": 6.103854142626356,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.595730486991328e-05,
+      "loss": 1.3442,
+      "step": 6612
+    },
+    {
+      "epoch": 6.104777290560812,
+      "grad_norm": 3.6328125,
+      "learning_rate": 5.595063375583722e-05,
+      "loss": 1.2544,
+      "step": 6613
+    },
+    {
+      "epoch": 6.105700438495269,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.5943962641761175e-05,
+      "loss": 1.1289,
+      "step": 6614
+    },
+    {
+      "epoch": 6.106623586429725,
+      "grad_norm": 3.888671875,
+      "learning_rate": 5.593729152768512e-05,
+      "loss": 1.3345,
+      "step": 6615
+    },
+    {
+      "epoch": 6.107546734364182,
+      "grad_norm": 2.849609375,
+      "learning_rate": 5.593062041360908e-05,
+      "loss": 1.108,
+      "step": 6616
+    },
+    {
+      "epoch": 6.108469882298638,
+      "grad_norm": 3.552734375,
+      "learning_rate": 5.592394929953303e-05,
+      "loss": 1.4028,
+      "step": 6617
+    },
+    {
+      "epoch": 6.109393030233095,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.591727818545698e-05,
+      "loss": 1.0032,
+      "step": 6618
+    },
+    {
+      "epoch": 6.110316178167551,
+      "grad_norm": 3.634765625,
+      "learning_rate": 5.591060707138093e-05,
+      "loss": 1.3386,
+      "step": 6619
+    },
+    {
+      "epoch": 6.111239326102008,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.5903935957304874e-05,
+      "loss": 1.2441,
+      "step": 6620
+    },
+    {
+      "epoch": 6.112162474036464,
+      "grad_norm": 4.12109375,
+      "learning_rate": 5.5897264843228827e-05,
+      "loss": 1.354,
+      "step": 6621
+    },
+    {
+      "epoch": 6.113085621970921,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.589059372915277e-05,
+      "loss": 1.0757,
+      "step": 6622
+    },
+    {
+      "epoch": 6.114008769905377,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.5883922615076724e-05,
+      "loss": 1.3069,
+      "step": 6623
+    },
+    {
+      "epoch": 6.114931917839834,
+      "grad_norm": 3.548828125,
+      "learning_rate": 5.587725150100067e-05,
+      "loss": 1.5178,
+      "step": 6624
+    },
+    {
+      "epoch": 6.115855065774291,
+      "grad_norm": 3.322265625,
+      "learning_rate": 5.587058038692462e-05,
+      "loss": 1.6416,
+      "step": 6625
+    },
+    {
+      "epoch": 6.116778213708747,
+      "grad_norm": 3.380859375,
+      "learning_rate": 5.586390927284857e-05,
+      "loss": 1.337,
+      "step": 6626
+    },
+    {
+      "epoch": 6.117701361643204,
+      "grad_norm": 3.541015625,
+      "learning_rate": 5.585723815877252e-05,
+      "loss": 1.4336,
+      "step": 6627
+    },
+    {
+      "epoch": 6.11862450957766,
+      "grad_norm": 3.365234375,
+      "learning_rate": 5.5850567044696465e-05,
+      "loss": 1.2124,
+      "step": 6628
+    },
+    {
+      "epoch": 6.119547657512117,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.584389593062042e-05,
+      "loss": 1.2483,
+      "step": 6629
+    },
+    {
+      "epoch": 6.120470805446573,
+      "grad_norm": 4.59375,
+      "learning_rate": 5.583722481654436e-05,
+      "loss": 1.3914,
+      "step": 6630
+    },
+    {
+      "epoch": 6.1213939533810295,
+      "grad_norm": 3.05078125,
+      "learning_rate": 5.5830553702468315e-05,
+      "loss": 1.5967,
+      "step": 6631
+    },
+    {
+      "epoch": 6.122317101315486,
+      "grad_norm": 3.322265625,
+      "learning_rate": 5.582388258839226e-05,
+      "loss": 1.3544,
+      "step": 6632
+    },
+    {
+      "epoch": 6.1232402492499425,
+      "grad_norm": 3.697265625,
+      "learning_rate": 5.581721147431621e-05,
+      "loss": 1.2063,
+      "step": 6633
+    },
+    {
+      "epoch": 6.1241633971843985,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.581054036024016e-05,
+      "loss": 1.4216,
+      "step": 6634
+    },
+    {
+      "epoch": 6.125086545118855,
+      "grad_norm": 3.486328125,
+      "learning_rate": 5.580386924616411e-05,
+      "loss": 1.3875,
+      "step": 6635
+    },
+    {
+      "epoch": 6.1260096930533114,
+      "grad_norm": 4.03515625,
+      "learning_rate": 5.5797198132088055e-05,
+      "loss": 1.3242,
+      "step": 6636
+    },
+    {
+      "epoch": 6.126932840987768,
+      "grad_norm": 3.814453125,
+      "learning_rate": 5.579052701801201e-05,
+      "loss": 1.4844,
+      "step": 6637
+    },
+    {
+      "epoch": 6.127855988922224,
+      "grad_norm": 4.1796875,
+      "learning_rate": 5.578385590393595e-05,
+      "loss": 1.4092,
+      "step": 6638
+    },
+    {
+      "epoch": 6.128779136856681,
+      "grad_norm": 3.43359375,
+      "learning_rate": 5.5777184789859905e-05,
+      "loss": 1.4016,
+      "step": 6639
+    },
+    {
+      "epoch": 6.129702284791138,
+      "grad_norm": 4.01171875,
+      "learning_rate": 5.577051367578385e-05,
+      "loss": 1.4637,
+      "step": 6640
+    },
+    {
+      "epoch": 6.130625432725594,
+      "grad_norm": 3.662109375,
+      "learning_rate": 5.576384256170781e-05,
+      "loss": 1.366,
+      "step": 6641
+    },
+    {
+      "epoch": 6.131548580660051,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.575717144763176e-05,
+      "loss": 1.3103,
+      "step": 6642
+    },
+    {
+      "epoch": 6.132471728594507,
+      "grad_norm": 3.076171875,
+      "learning_rate": 5.5750500333555714e-05,
+      "loss": 1.5112,
+      "step": 6643
+    },
+    {
+      "epoch": 6.133394876528964,
+      "grad_norm": 3.00390625,
+      "learning_rate": 5.574382921947966e-05,
+      "loss": 1.3319,
+      "step": 6644
+    },
+    {
+      "epoch": 6.13431802446342,
+      "grad_norm": 4.33984375,
+      "learning_rate": 5.573715810540361e-05,
+      "loss": 1.5393,
+      "step": 6645
+    },
+    {
+      "epoch": 6.135241172397877,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.573048699132756e-05,
+      "loss": 1.4065,
+      "step": 6646
+    },
+    {
+      "epoch": 6.136164320332333,
+      "grad_norm": 4.69140625,
+      "learning_rate": 5.572381587725151e-05,
+      "loss": 1.3481,
+      "step": 6647
+    },
+    {
+      "epoch": 6.13708746826679,
+      "grad_norm": 3.958984375,
+      "learning_rate": 5.5717144763175455e-05,
+      "loss": 1.054,
+      "step": 6648
+    },
+    {
+      "epoch": 6.138010616201246,
+      "grad_norm": 3.630859375,
+      "learning_rate": 5.571047364909941e-05,
+      "loss": 1.3738,
+      "step": 6649
+    },
+    {
+      "epoch": 6.138933764135703,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.570380253502335e-05,
+      "loss": 1.1,
+      "step": 6650
+    },
+    {
+      "epoch": 6.139856912070159,
+      "grad_norm": 3.326171875,
+      "learning_rate": 5.5697131420947305e-05,
+      "loss": 1.2739,
+      "step": 6651
+    },
+    {
+      "epoch": 6.140780060004616,
+      "grad_norm": 3.921875,
+      "learning_rate": 5.569046030687125e-05,
+      "loss": 1.4487,
+      "step": 6652
+    },
+    {
+      "epoch": 6.141703207939072,
+      "grad_norm": 3.943359375,
+      "learning_rate": 5.56837891927952e-05,
+      "loss": 1.1219,
+      "step": 6653
+    },
+    {
+      "epoch": 6.142626355873529,
+      "grad_norm": 3.724609375,
+      "learning_rate": 5.567711807871915e-05,
+      "loss": 1.0565,
+      "step": 6654
+    },
+    {
+      "epoch": 6.143549503807985,
+      "grad_norm": 3.939453125,
+      "learning_rate": 5.56704469646431e-05,
+      "loss": 1.3097,
+      "step": 6655
+    },
+    {
+      "epoch": 6.144472651742442,
+      "grad_norm": 4.50390625,
+      "learning_rate": 5.5663775850567045e-05,
+      "loss": 1.4834,
+      "step": 6656
+    },
+    {
+      "epoch": 6.145395799676899,
+      "grad_norm": 3.701171875,
+      "learning_rate": 5.5657104736491e-05,
+      "loss": 0.9825,
+      "step": 6657
+    },
+    {
+      "epoch": 6.146318947611355,
+      "grad_norm": 3.759765625,
+      "learning_rate": 5.565043362241494e-05,
+      "loss": 1.48,
+      "step": 6658
+    },
+    {
+      "epoch": 6.147242095545812,
+      "grad_norm": 3.9140625,
+      "learning_rate": 5.5643762508338895e-05,
+      "loss": 1.3118,
+      "step": 6659
+    },
+    {
+      "epoch": 6.148165243480268,
+      "grad_norm": 3.416015625,
+      "learning_rate": 5.563709139426284e-05,
+      "loss": 1.3445,
+      "step": 6660
+    },
+    {
+      "epoch": 6.1490883914147245,
+      "grad_norm": 3.486328125,
+      "learning_rate": 5.563042028018679e-05,
+      "loss": 1.2363,
+      "step": 6661
+    },
+    {
+      "epoch": 6.1500115393491805,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.562374916611074e-05,
+      "loss": 1.4821,
+      "step": 6662
+    },
+    {
+      "epoch": 6.1509346872836375,
+      "grad_norm": 3.9453125,
+      "learning_rate": 5.561707805203469e-05,
+      "loss": 1.2622,
+      "step": 6663
+    },
+    {
+      "epoch": 6.1518578352180935,
+      "grad_norm": 3.951171875,
+      "learning_rate": 5.5610406937958636e-05,
+      "loss": 1.2201,
+      "step": 6664
+    },
+    {
+      "epoch": 6.15278098315255,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.560373582388259e-05,
+      "loss": 1.1667,
+      "step": 6665
+    },
+    {
+      "epoch": 6.153704131087006,
+      "grad_norm": 3.12890625,
+      "learning_rate": 5.559706470980655e-05,
+      "loss": 1.7109,
+      "step": 6666
+    },
+    {
+      "epoch": 6.154627279021463,
+      "grad_norm": 3.21484375,
+      "learning_rate": 5.559039359573049e-05,
+      "loss": 1.5039,
+      "step": 6667
+    },
+    {
+      "epoch": 6.155550426955919,
+      "grad_norm": 3.419921875,
+      "learning_rate": 5.5583722481654445e-05,
+      "loss": 1.1874,
+      "step": 6668
+    },
+    {
+      "epoch": 6.156473574890376,
+      "grad_norm": 3.892578125,
+      "learning_rate": 5.557705136757839e-05,
+      "loss": 1.4644,
+      "step": 6669
+    },
+    {
+      "epoch": 6.157396722824832,
+      "grad_norm": 4.79296875,
+      "learning_rate": 5.557038025350234e-05,
+      "loss": 1.2426,
+      "step": 6670
+    },
+    {
+      "epoch": 6.158319870759289,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.556370913942629e-05,
+      "loss": 1.1155,
+      "step": 6671
+    },
+    {
+      "epoch": 6.159243018693745,
+      "grad_norm": 4.06640625,
+      "learning_rate": 5.555703802535024e-05,
+      "loss": 1.5967,
+      "step": 6672
+    },
+    {
+      "epoch": 6.160166166628202,
+      "grad_norm": 3.3203125,
+      "learning_rate": 5.5550366911274185e-05,
+      "loss": 1.3569,
+      "step": 6673
+    },
+    {
+      "epoch": 6.161089314562659,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.554369579719814e-05,
+      "loss": 1.3013,
+      "step": 6674
+    },
+    {
+      "epoch": 6.162012462497115,
+      "grad_norm": 3.66015625,
+      "learning_rate": 5.553702468312208e-05,
+      "loss": 1.0945,
+      "step": 6675
+    },
+    {
+      "epoch": 6.162935610431572,
+      "grad_norm": 3.568359375,
+      "learning_rate": 5.5530353569046035e-05,
+      "loss": 1.6179,
+      "step": 6676
+    },
+    {
+      "epoch": 6.163858758366028,
+      "grad_norm": 3.01953125,
+      "learning_rate": 5.552368245496998e-05,
+      "loss": 1.3655,
+      "step": 6677
+    },
+    {
+      "epoch": 6.164781906300485,
+      "grad_norm": 3.03515625,
+      "learning_rate": 5.551701134089393e-05,
+      "loss": 1.0352,
+      "step": 6678
+    },
+    {
+      "epoch": 6.165705054234941,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.551034022681788e-05,
+      "loss": 1.3018,
+      "step": 6679
+    },
+    {
+      "epoch": 6.166628202169398,
+      "grad_norm": 3.6640625,
+      "learning_rate": 5.550366911274183e-05,
+      "loss": 1.5222,
+      "step": 6680
+    },
+    {
+      "epoch": 6.167551350103854,
+      "grad_norm": 3.111328125,
+      "learning_rate": 5.5496997998665776e-05,
+      "loss": 1.1886,
+      "step": 6681
+    },
+    {
+      "epoch": 6.168474498038311,
+      "grad_norm": 3.33203125,
+      "learning_rate": 5.549032688458973e-05,
+      "loss": 1.3857,
+      "step": 6682
+    },
+    {
+      "epoch": 6.169397645972767,
+      "grad_norm": 3.568359375,
+      "learning_rate": 5.5483655770513674e-05,
+      "loss": 1.5188,
+      "step": 6683
+    },
+    {
+      "epoch": 6.170320793907224,
+      "grad_norm": 3.89453125,
+      "learning_rate": 5.5476984656437626e-05,
+      "loss": 1.5735,
+      "step": 6684
+    },
+    {
+      "epoch": 6.17124394184168,
+      "grad_norm": 4.7265625,
+      "learning_rate": 5.547031354236157e-05,
+      "loss": 1.3654,
+      "step": 6685
+    },
+    {
+      "epoch": 6.172167089776137,
+      "grad_norm": 4.25390625,
+      "learning_rate": 5.5463642428285523e-05,
+      "loss": 1.3264,
+      "step": 6686
+    },
+    {
+      "epoch": 6.173090237710593,
+      "grad_norm": 3.162109375,
+      "learning_rate": 5.545697131420947e-05,
+      "loss": 1.3345,
+      "step": 6687
+    },
+    {
+      "epoch": 6.17401338564505,
+      "grad_norm": 3.73046875,
+      "learning_rate": 5.545030020013342e-05,
+      "loss": 1.4346,
+      "step": 6688
+    },
+    {
+      "epoch": 6.174936533579507,
+      "grad_norm": 4.734375,
+      "learning_rate": 5.5443629086057367e-05,
+      "loss": 1.1561,
+      "step": 6689
+    },
+    {
+      "epoch": 6.175859681513963,
+      "grad_norm": 3.564453125,
+      "learning_rate": 5.543695797198132e-05,
+      "loss": 1.5703,
+      "step": 6690
+    },
+    {
+      "epoch": 6.1767828294484195,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.543028685790528e-05,
+      "loss": 1.2382,
+      "step": 6691
+    },
+    {
+      "epoch": 6.1777059773828755,
+      "grad_norm": 3.748046875,
+      "learning_rate": 5.542361574382923e-05,
+      "loss": 1.5205,
+      "step": 6692
+    },
+    {
+      "epoch": 6.178629125317332,
+      "grad_norm": 4.12109375,
+      "learning_rate": 5.5416944629753175e-05,
+      "loss": 1.4451,
+      "step": 6693
+    },
+    {
+      "epoch": 6.1795522732517885,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.541027351567713e-05,
+      "loss": 1.1207,
+      "step": 6694
+    },
+    {
+      "epoch": 6.180475421186245,
+      "grad_norm": 3.943359375,
+      "learning_rate": 5.540360240160107e-05,
+      "loss": 1.2515,
+      "step": 6695
+    },
+    {
+      "epoch": 6.181398569120701,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.5396931287525025e-05,
+      "loss": 1.1233,
+      "step": 6696
+    },
+    {
+      "epoch": 6.182321717055158,
+      "grad_norm": 3.70703125,
+      "learning_rate": 5.539026017344897e-05,
+      "loss": 1.1333,
+      "step": 6697
+    },
+    {
+      "epoch": 6.183244864989614,
+      "grad_norm": 3.88671875,
+      "learning_rate": 5.538358905937292e-05,
+      "loss": 1.5066,
+      "step": 6698
+    },
+    {
+      "epoch": 6.184168012924071,
+      "grad_norm": 3.701171875,
+      "learning_rate": 5.537691794529687e-05,
+      "loss": 1.5735,
+      "step": 6699
+    },
+    {
+      "epoch": 6.185091160858527,
+      "grad_norm": 3.59765625,
+      "learning_rate": 5.537024683122082e-05,
+      "loss": 1.447,
+      "step": 6700
+    },
+    {
+      "epoch": 6.186014308792984,
+      "grad_norm": 4.328125,
+      "learning_rate": 5.5363575717144766e-05,
+      "loss": 1.1437,
+      "step": 6701
+    },
+    {
+      "epoch": 6.18693745672744,
+      "grad_norm": 3.90234375,
+      "learning_rate": 5.535690460306872e-05,
+      "loss": 1.384,
+      "step": 6702
+    },
+    {
+      "epoch": 6.187860604661897,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.5350233488992664e-05,
+      "loss": 1.5122,
+      "step": 6703
+    },
+    {
+      "epoch": 6.188783752596353,
+      "grad_norm": 3.16796875,
+      "learning_rate": 5.5343562374916616e-05,
+      "loss": 1.4148,
+      "step": 6704
+    },
+    {
+      "epoch": 6.18970690053081,
+      "grad_norm": 3.630859375,
+      "learning_rate": 5.533689126084056e-05,
+      "loss": 1.5332,
+      "step": 6705
+    },
+    {
+      "epoch": 6.190630048465267,
+      "grad_norm": 3.50390625,
+      "learning_rate": 5.5330220146764513e-05,
+      "loss": 1.3043,
+      "step": 6706
+    },
+    {
+      "epoch": 6.191553196399723,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.532354903268846e-05,
+      "loss": 1.4709,
+      "step": 6707
+    },
+    {
+      "epoch": 6.19247634433418,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.531687791861241e-05,
+      "loss": 1.2488,
+      "step": 6708
+    },
+    {
+      "epoch": 6.193399492268636,
+      "grad_norm": 4.1953125,
+      "learning_rate": 5.5310206804536356e-05,
+      "loss": 1.4209,
+      "step": 6709
+    },
+    {
+      "epoch": 6.194322640203093,
+      "grad_norm": 3.556640625,
+      "learning_rate": 5.530353569046031e-05,
+      "loss": 1.0786,
+      "step": 6710
+    },
+    {
+      "epoch": 6.195245788137549,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.5296864576384254e-05,
+      "loss": 1.2765,
+      "step": 6711
+    },
+    {
+      "epoch": 6.196168936072006,
+      "grad_norm": 4.078125,
+      "learning_rate": 5.5290193462308206e-05,
+      "loss": 1.4131,
+      "step": 6712
+    },
+    {
+      "epoch": 6.197092084006462,
+      "grad_norm": 3.25,
+      "learning_rate": 5.528352234823215e-05,
+      "loss": 1.1175,
+      "step": 6713
+    },
+    {
+      "epoch": 6.198015231940919,
+      "grad_norm": 3.8515625,
+      "learning_rate": 5.5276851234156104e-05,
+      "loss": 1.595,
+      "step": 6714
+    },
+    {
+      "epoch": 6.198938379875375,
+      "grad_norm": 3.0703125,
+      "learning_rate": 5.527018012008005e-05,
+      "loss": 1.2346,
+      "step": 6715
+    },
+    {
+      "epoch": 6.199861527809832,
+      "grad_norm": 4.1484375,
+      "learning_rate": 5.526350900600401e-05,
+      "loss": 1.293,
+      "step": 6716
+    },
+    {
+      "epoch": 6.200784675744288,
+      "grad_norm": 3.26953125,
+      "learning_rate": 5.525683789192796e-05,
+      "loss": 1.1522,
+      "step": 6717
+    },
+    {
+      "epoch": 6.201707823678745,
+      "grad_norm": 3.74609375,
+      "learning_rate": 5.5250166777851906e-05,
+      "loss": 1.8279,
+      "step": 6718
+    },
+    {
+      "epoch": 6.202630971613201,
+      "grad_norm": 3.8046875,
+      "learning_rate": 5.524349566377586e-05,
+      "loss": 1.172,
+      "step": 6719
+    },
+    {
+      "epoch": 6.203554119547658,
+      "grad_norm": 3.8671875,
+      "learning_rate": 5.5236824549699804e-05,
+      "loss": 1.299,
+      "step": 6720
+    },
+    {
+      "epoch": 6.204477267482114,
+      "grad_norm": 3.95703125,
+      "learning_rate": 5.5230153435623756e-05,
+      "loss": 1.5735,
+      "step": 6721
+    },
+    {
+      "epoch": 6.2054004154165705,
+      "grad_norm": 3.72265625,
+      "learning_rate": 5.52234823215477e-05,
+      "loss": 1.3545,
+      "step": 6722
+    },
+    {
+      "epoch": 6.206323563351027,
+      "grad_norm": 4.48046875,
+      "learning_rate": 5.5216811207471653e-05,
+      "loss": 1.636,
+      "step": 6723
+    },
+    {
+      "epoch": 6.207246711285483,
+      "grad_norm": 3.595703125,
+      "learning_rate": 5.52101400933956e-05,
+      "loss": 1.1786,
+      "step": 6724
+    },
+    {
+      "epoch": 6.20816985921994,
+      "grad_norm": 3.658203125,
+      "learning_rate": 5.520346897931955e-05,
+      "loss": 1.334,
+      "step": 6725
+    },
+    {
+      "epoch": 6.209093007154396,
+      "grad_norm": 4.28515625,
+      "learning_rate": 5.5196797865243497e-05,
+      "loss": 1.2747,
+      "step": 6726
+    },
+    {
+      "epoch": 6.210016155088853,
+      "grad_norm": 3.427734375,
+      "learning_rate": 5.519012675116745e-05,
+      "loss": 1.2053,
+      "step": 6727
+    },
+    {
+      "epoch": 6.210939303023309,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.5183455637091394e-05,
+      "loss": 1.3282,
+      "step": 6728
+    },
+    {
+      "epoch": 6.211862450957766,
+      "grad_norm": 4.01953125,
+      "learning_rate": 5.5176784523015346e-05,
+      "loss": 1.4712,
+      "step": 6729
+    },
+    {
+      "epoch": 6.212785598892222,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.517011340893929e-05,
+      "loss": 1.6008,
+      "step": 6730
+    },
+    {
+      "epoch": 6.213708746826679,
+      "grad_norm": 3.40234375,
+      "learning_rate": 5.5163442294863244e-05,
+      "loss": 1.2744,
+      "step": 6731
+    },
+    {
+      "epoch": 6.214631894761135,
+      "grad_norm": 4.1796875,
+      "learning_rate": 5.515677118078719e-05,
+      "loss": 1.1556,
+      "step": 6732
+    },
+    {
+      "epoch": 6.215555042695592,
+      "grad_norm": 3.22265625,
+      "learning_rate": 5.515010006671114e-05,
+      "loss": 1.4727,
+      "step": 6733
+    },
+    {
+      "epoch": 6.216478190630048,
+      "grad_norm": 3.193359375,
+      "learning_rate": 5.514342895263509e-05,
+      "loss": 1.6616,
+      "step": 6734
+    },
+    {
+      "epoch": 6.217401338564505,
+      "grad_norm": 3.39453125,
+      "learning_rate": 5.513675783855904e-05,
+      "loss": 1.4088,
+      "step": 6735
+    },
+    {
+      "epoch": 6.218324486498961,
+      "grad_norm": 3.326171875,
+      "learning_rate": 5.5130086724482985e-05,
+      "loss": 1.459,
+      "step": 6736
+    },
+    {
+      "epoch": 6.219247634433418,
+      "grad_norm": 3.388671875,
+      "learning_rate": 5.512341561040694e-05,
+      "loss": 1.4973,
+      "step": 6737
+    },
+    {
+      "epoch": 6.220170782367875,
+      "grad_norm": 3.693359375,
+      "learning_rate": 5.511674449633088e-05,
+      "loss": 1.2722,
+      "step": 6738
+    },
+    {
+      "epoch": 6.221093930302331,
+      "grad_norm": 3.884765625,
+      "learning_rate": 5.5110073382254835e-05,
+      "loss": 1.4326,
+      "step": 6739
+    },
+    {
+      "epoch": 6.222017078236788,
+      "grad_norm": 3.154296875,
+      "learning_rate": 5.510340226817878e-05,
+      "loss": 1.0808,
+      "step": 6740
+    },
+    {
+      "epoch": 6.222940226171244,
+      "grad_norm": 3.474609375,
+      "learning_rate": 5.5096731154102746e-05,
+      "loss": 1.2025,
+      "step": 6741
+    },
+    {
+      "epoch": 6.223863374105701,
+      "grad_norm": 3.126953125,
+      "learning_rate": 5.509006004002669e-05,
+      "loss": 1.5977,
+      "step": 6742
+    },
+    {
+      "epoch": 6.224786522040157,
+      "grad_norm": 3.537109375,
+      "learning_rate": 5.5083388925950643e-05,
+      "loss": 1.6267,
+      "step": 6743
+    },
+    {
+      "epoch": 6.225709669974614,
+      "grad_norm": 3.53515625,
+      "learning_rate": 5.507671781187459e-05,
+      "loss": 0.9379,
+      "step": 6744
+    },
+    {
+      "epoch": 6.22663281790907,
+      "grad_norm": 3.427734375,
+      "learning_rate": 5.507004669779854e-05,
+      "loss": 1.4612,
+      "step": 6745
+    },
+    {
+      "epoch": 6.227555965843527,
+      "grad_norm": 4.2421875,
+      "learning_rate": 5.5063375583722486e-05,
+      "loss": 1.4412,
+      "step": 6746
+    },
+    {
+      "epoch": 6.228479113777983,
+      "grad_norm": 3.67578125,
+      "learning_rate": 5.505670446964644e-05,
+      "loss": 1.168,
+      "step": 6747
+    },
+    {
+      "epoch": 6.22940226171244,
+      "grad_norm": 3.63671875,
+      "learning_rate": 5.5050033355570384e-05,
+      "loss": 1.4438,
+      "step": 6748
+    },
+    {
+      "epoch": 6.230325409646896,
+      "grad_norm": 3.783203125,
+      "learning_rate": 5.5043362241494336e-05,
+      "loss": 1.3491,
+      "step": 6749
+    },
+    {
+      "epoch": 6.2312485575813525,
+      "grad_norm": 2.9609375,
+      "learning_rate": 5.503669112741828e-05,
+      "loss": 1.2648,
+      "step": 6750
+    },
+    {
+      "epoch": 6.232171705515809,
+      "grad_norm": 3.697265625,
+      "learning_rate": 5.5030020013342234e-05,
+      "loss": 1.411,
+      "step": 6751
+    },
+    {
+      "epoch": 6.2330948534502655,
+      "grad_norm": 3.64453125,
+      "learning_rate": 5.502334889926618e-05,
+      "loss": 1.1986,
+      "step": 6752
+    },
+    {
+      "epoch": 6.2340180013847215,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.501667778519013e-05,
+      "loss": 1.3296,
+      "step": 6753
+    },
+    {
+      "epoch": 6.234941149319178,
+      "grad_norm": 3.4921875,
+      "learning_rate": 5.501000667111408e-05,
+      "loss": 1.1936,
+      "step": 6754
+    },
+    {
+      "epoch": 6.2358642972536344,
+      "grad_norm": 3.12109375,
+      "learning_rate": 5.500333555703803e-05,
+      "loss": 1.6611,
+      "step": 6755
+    },
+    {
+      "epoch": 6.236787445188091,
+      "grad_norm": 3.5078125,
+      "learning_rate": 5.4996664442961975e-05,
+      "loss": 1.3032,
+      "step": 6756
+    },
+    {
+      "epoch": 6.237710593122548,
+      "grad_norm": 4.4375,
+      "learning_rate": 5.498999332888593e-05,
+      "loss": 1.4368,
+      "step": 6757
+    },
+    {
+      "epoch": 6.238633741057004,
+      "grad_norm": 3.076171875,
+      "learning_rate": 5.498332221480987e-05,
+      "loss": 1.4656,
+      "step": 6758
+    },
+    {
+      "epoch": 6.239556888991461,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.4976651100733825e-05,
+      "loss": 1.2346,
+      "step": 6759
+    },
+    {
+      "epoch": 6.240480036925917,
+      "grad_norm": 3.392578125,
+      "learning_rate": 5.496997998665777e-05,
+      "loss": 1.2589,
+      "step": 6760
+    },
+    {
+      "epoch": 6.241403184860374,
+      "grad_norm": 3.291015625,
+      "learning_rate": 5.496330887258172e-05,
+      "loss": 1.4221,
+      "step": 6761
+    },
+    {
+      "epoch": 6.24232633279483,
+      "grad_norm": 3.5703125,
+      "learning_rate": 5.495663775850567e-05,
+      "loss": 1.4587,
+      "step": 6762
+    },
+    {
+      "epoch": 6.243249480729287,
+      "grad_norm": 3.267578125,
+      "learning_rate": 5.494996664442962e-05,
+      "loss": 1.3491,
+      "step": 6763
+    },
+    {
+      "epoch": 6.244172628663743,
+      "grad_norm": 3.626953125,
+      "learning_rate": 5.4943295530353565e-05,
+      "loss": 1.4707,
+      "step": 6764
+    },
+    {
+      "epoch": 6.2450957765982,
+      "grad_norm": 3.7109375,
+      "learning_rate": 5.493662441627752e-05,
+      "loss": 1.4331,
+      "step": 6765
+    },
+    {
+      "epoch": 6.246018924532656,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.4929953302201476e-05,
+      "loss": 1.5103,
+      "step": 6766
+    },
+    {
+      "epoch": 6.246942072467113,
+      "grad_norm": 4.12109375,
+      "learning_rate": 5.492328218812542e-05,
+      "loss": 1.1946,
+      "step": 6767
+    },
+    {
+      "epoch": 6.247865220401569,
+      "grad_norm": 3.560546875,
+      "learning_rate": 5.4916611074049374e-05,
+      "loss": 1.6099,
+      "step": 6768
+    },
+    {
+      "epoch": 6.248788368336026,
+      "grad_norm": 3.4140625,
+      "learning_rate": 5.490993995997332e-05,
+      "loss": 1.1508,
+      "step": 6769
+    },
+    {
+      "epoch": 6.249711516270482,
+      "grad_norm": 4.078125,
+      "learning_rate": 5.490326884589727e-05,
+      "loss": 1.4224,
+      "step": 6770
+    },
+    {
+      "epoch": 6.250634664204939,
+      "grad_norm": 3.40234375,
+      "learning_rate": 5.489659773182122e-05,
+      "loss": 1.2913,
+      "step": 6771
+    },
+    {
+      "epoch": 6.251557812139396,
+      "grad_norm": 3.970703125,
+      "learning_rate": 5.488992661774517e-05,
+      "loss": 1.3207,
+      "step": 6772
+    },
+    {
+      "epoch": 6.252480960073852,
+      "grad_norm": 3.677734375,
+      "learning_rate": 5.4883255503669115e-05,
+      "loss": 1.3408,
+      "step": 6773
+    },
+    {
+      "epoch": 6.253404108008309,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.487658438959307e-05,
+      "loss": 1.236,
+      "step": 6774
+    },
+    {
+      "epoch": 6.254327255942765,
+      "grad_norm": 3.62109375,
+      "learning_rate": 5.486991327551701e-05,
+      "loss": 1.1371,
+      "step": 6775
+    },
+    {
+      "epoch": 6.255250403877222,
+      "grad_norm": 3.849609375,
+      "learning_rate": 5.4863242161440965e-05,
+      "loss": 1.3569,
+      "step": 6776
+    },
+    {
+      "epoch": 6.256173551811678,
+      "grad_norm": 4.38671875,
+      "learning_rate": 5.485657104736491e-05,
+      "loss": 1.4697,
+      "step": 6777
+    },
+    {
+      "epoch": 6.257096699746135,
+      "grad_norm": 3.373046875,
+      "learning_rate": 5.484989993328886e-05,
+      "loss": 1.3101,
+      "step": 6778
+    },
+    {
+      "epoch": 6.258019847680591,
+      "grad_norm": 3.26171875,
+      "learning_rate": 5.484322881921281e-05,
+      "loss": 1.2629,
+      "step": 6779
+    },
+    {
+      "epoch": 6.2589429956150475,
+      "grad_norm": 3.486328125,
+      "learning_rate": 5.483655770513676e-05,
+      "loss": 1.4568,
+      "step": 6780
+    },
+    {
+      "epoch": 6.2598661435495035,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.4829886591060705e-05,
+      "loss": 1.5444,
+      "step": 6781
+    },
+    {
+      "epoch": 6.2607892914839605,
+      "grad_norm": 3.3828125,
+      "learning_rate": 5.482321547698466e-05,
+      "loss": 1.1041,
+      "step": 6782
+    },
+    {
+      "epoch": 6.2617124394184165,
+      "grad_norm": 3.654296875,
+      "learning_rate": 5.48165443629086e-05,
+      "loss": 1.2924,
+      "step": 6783
+    },
+    {
+      "epoch": 6.262635587352873,
+      "grad_norm": 3.185546875,
+      "learning_rate": 5.4809873248832555e-05,
+      "loss": 1.3284,
+      "step": 6784
+    },
+    {
+      "epoch": 6.263558735287329,
+      "grad_norm": 3.37890625,
+      "learning_rate": 5.48032021347565e-05,
+      "loss": 1.4697,
+      "step": 6785
+    },
+    {
+      "epoch": 6.264481883221786,
+      "grad_norm": 3.318359375,
+      "learning_rate": 5.479653102068045e-05,
+      "loss": 1.2489,
+      "step": 6786
+    },
+    {
+      "epoch": 6.265405031156243,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.47898599066044e-05,
+      "loss": 1.3488,
+      "step": 6787
+    },
+    {
+      "epoch": 6.266328179090699,
+      "grad_norm": 3.580078125,
+      "learning_rate": 5.478318879252835e-05,
+      "loss": 1.2134,
+      "step": 6788
+    },
+    {
+      "epoch": 6.267251327025156,
+      "grad_norm": 3.1484375,
+      "learning_rate": 5.47765176784523e-05,
+      "loss": 1.2395,
+      "step": 6789
+    },
+    {
+      "epoch": 6.268174474959612,
+      "grad_norm": 3.05859375,
+      "learning_rate": 5.476984656437625e-05,
+      "loss": 1.266,
+      "step": 6790
+    },
+    {
+      "epoch": 6.269097622894069,
+      "grad_norm": 4.19140625,
+      "learning_rate": 5.476317545030021e-05,
+      "loss": 1.5376,
+      "step": 6791
+    },
+    {
+      "epoch": 6.270020770828525,
+      "grad_norm": 3.810546875,
+      "learning_rate": 5.475650433622416e-05,
+      "loss": 1.1108,
+      "step": 6792
+    },
+    {
+      "epoch": 6.270943918762982,
+      "grad_norm": 4.04296875,
+      "learning_rate": 5.4749833222148105e-05,
+      "loss": 1.0835,
+      "step": 6793
+    },
+    {
+      "epoch": 6.271867066697438,
+      "grad_norm": 4.11328125,
+      "learning_rate": 5.474316210807206e-05,
+      "loss": 1.1064,
+      "step": 6794
+    },
+    {
+      "epoch": 6.272790214631895,
+      "grad_norm": 3.8984375,
+      "learning_rate": 5.4736490993996e-05,
+      "loss": 1.394,
+      "step": 6795
+    },
+    {
+      "epoch": 6.273713362566351,
+      "grad_norm": 3.375,
+      "learning_rate": 5.4729819879919955e-05,
+      "loss": 1.4548,
+      "step": 6796
+    },
+    {
+      "epoch": 6.274636510500808,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.47231487658439e-05,
+      "loss": 1.0952,
+      "step": 6797
+    },
+    {
+      "epoch": 6.275559658435264,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.471647765176785e-05,
+      "loss": 1.3347,
+      "step": 6798
+    },
+    {
+      "epoch": 6.276482806369721,
+      "grad_norm": 3.56640625,
+      "learning_rate": 5.47098065376918e-05,
+      "loss": 1.469,
+      "step": 6799
+    },
+    {
+      "epoch": 6.277405954304177,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.470313542361575e-05,
+      "loss": 1.2263,
+      "step": 6800
+    },
+    {
+      "epoch": 6.278329102238634,
+      "grad_norm": 4.0546875,
+      "learning_rate": 5.4696464309539695e-05,
+      "loss": 1.3082,
+      "step": 6801
+    },
+    {
+      "epoch": 6.27925225017309,
+      "grad_norm": 3.326171875,
+      "learning_rate": 5.468979319546365e-05,
+      "loss": 1.3142,
+      "step": 6802
+    },
+    {
+      "epoch": 6.280175398107547,
+      "grad_norm": 3.48828125,
+      "learning_rate": 5.468312208138759e-05,
+      "loss": 1.479,
+      "step": 6803
+    },
+    {
+      "epoch": 6.281098546042003,
+      "grad_norm": 3.744140625,
+      "learning_rate": 5.4676450967311545e-05,
+      "loss": 1.301,
+      "step": 6804
+    },
+    {
+      "epoch": 6.28202169397646,
+      "grad_norm": 3.73828125,
+      "learning_rate": 5.466977985323549e-05,
+      "loss": 1.4575,
+      "step": 6805
+    },
+    {
+      "epoch": 6.282944841910917,
+      "grad_norm": 5.1953125,
+      "learning_rate": 5.466310873915944e-05,
+      "loss": 1.3501,
+      "step": 6806
+    },
+    {
+      "epoch": 6.283867989845373,
+      "grad_norm": 3.521484375,
+      "learning_rate": 5.465643762508339e-05,
+      "loss": 1.4417,
+      "step": 6807
+    },
+    {
+      "epoch": 6.28479113777983,
+      "grad_norm": 3.849609375,
+      "learning_rate": 5.464976651100734e-05,
+      "loss": 1.4795,
+      "step": 6808
+    },
+    {
+      "epoch": 6.285714285714286,
+      "grad_norm": 3.958984375,
+      "learning_rate": 5.4643095396931286e-05,
+      "loss": 1.2908,
+      "step": 6809
+    },
+    {
+      "epoch": 6.2866374336487425,
+      "grad_norm": 3.45703125,
+      "learning_rate": 5.463642428285524e-05,
+      "loss": 1.1825,
+      "step": 6810
+    },
+    {
+      "epoch": 6.2875605815831985,
+      "grad_norm": 3.84765625,
+      "learning_rate": 5.4629753168779183e-05,
+      "loss": 1.2786,
+      "step": 6811
+    },
+    {
+      "epoch": 6.288483729517655,
+      "grad_norm": 3.693359375,
+      "learning_rate": 5.4623082054703136e-05,
+      "loss": 1.5029,
+      "step": 6812
+    },
+    {
+      "epoch": 6.2894068774521115,
+      "grad_norm": 3.1953125,
+      "learning_rate": 5.461641094062708e-05,
+      "loss": 1.344,
+      "step": 6813
+    },
+    {
+      "epoch": 6.290330025386568,
+      "grad_norm": 4.07421875,
+      "learning_rate": 5.460973982655103e-05,
+      "loss": 1.4119,
+      "step": 6814
+    },
+    {
+      "epoch": 6.291253173321024,
+      "grad_norm": 3.537109375,
+      "learning_rate": 5.460306871247498e-05,
+      "loss": 1.1307,
+      "step": 6815
+    },
+    {
+      "epoch": 6.292176321255481,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.459639759839894e-05,
+      "loss": 1.4502,
+      "step": 6816
+    },
+    {
+      "epoch": 6.293099469189937,
+      "grad_norm": 4.02734375,
+      "learning_rate": 5.458972648432289e-05,
+      "loss": 1.4507,
+      "step": 6817
+    },
+    {
+      "epoch": 6.294022617124394,
+      "grad_norm": 3.185546875,
+      "learning_rate": 5.4583055370246835e-05,
+      "loss": 1.499,
+      "step": 6818
+    },
+    {
+      "epoch": 6.29494576505885,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.457638425617079e-05,
+      "loss": 1.0831,
+      "step": 6819
+    },
+    {
+      "epoch": 6.295868912993307,
+      "grad_norm": 3.330078125,
+      "learning_rate": 5.456971314209473e-05,
+      "loss": 1.2874,
+      "step": 6820
+    },
+    {
+      "epoch": 6.296792060927764,
+      "grad_norm": 3.732421875,
+      "learning_rate": 5.4563042028018685e-05,
+      "loss": 1.4668,
+      "step": 6821
+    },
+    {
+      "epoch": 6.29771520886222,
+      "grad_norm": 3.62890625,
+      "learning_rate": 5.455637091394263e-05,
+      "loss": 1.3191,
+      "step": 6822
+    },
+    {
+      "epoch": 6.298638356796677,
+      "grad_norm": 3.37109375,
+      "learning_rate": 5.454969979986658e-05,
+      "loss": 1.7666,
+      "step": 6823
+    },
+    {
+      "epoch": 6.299561504731133,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.454302868579053e-05,
+      "loss": 1.4382,
+      "step": 6824
+    },
+    {
+      "epoch": 6.30048465266559,
+      "grad_norm": 3.42578125,
+      "learning_rate": 5.453635757171448e-05,
+      "loss": 0.9918,
+      "step": 6825
+    },
+    {
+      "epoch": 6.301407800600046,
+      "grad_norm": 4.12890625,
+      "learning_rate": 5.4529686457638426e-05,
+      "loss": 1.5159,
+      "step": 6826
+    },
+    {
+      "epoch": 6.302330948534503,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.452301534356238e-05,
+      "loss": 1.3767,
+      "step": 6827
+    },
+    {
+      "epoch": 6.303254096468959,
+      "grad_norm": 3.6796875,
+      "learning_rate": 5.451634422948633e-05,
+      "loss": 1.3097,
+      "step": 6828
+    },
+    {
+      "epoch": 6.304177244403416,
+      "grad_norm": 3.091796875,
+      "learning_rate": 5.4509673115410276e-05,
+      "loss": 0.9822,
+      "step": 6829
+    },
+    {
+      "epoch": 6.305100392337872,
+      "grad_norm": 3.478515625,
+      "learning_rate": 5.450300200133423e-05,
+      "loss": 1.4448,
+      "step": 6830
+    },
+    {
+      "epoch": 6.306023540272329,
+      "grad_norm": 3.806640625,
+      "learning_rate": 5.449633088725817e-05,
+      "loss": 1.2845,
+      "step": 6831
+    },
+    {
+      "epoch": 6.306946688206785,
+      "grad_norm": 3.16015625,
+      "learning_rate": 5.4489659773182126e-05,
+      "loss": 1.26,
+      "step": 6832
+    },
+    {
+      "epoch": 6.307869836141242,
+      "grad_norm": 3.435546875,
+      "learning_rate": 5.448298865910607e-05,
+      "loss": 1.4015,
+      "step": 6833
+    },
+    {
+      "epoch": 6.308792984075698,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.447631754503002e-05,
+      "loss": 1.3748,
+      "step": 6834
+    },
+    {
+      "epoch": 6.309716132010155,
+      "grad_norm": 3.34765625,
+      "learning_rate": 5.446964643095397e-05,
+      "loss": 1.158,
+      "step": 6835
+    },
+    {
+      "epoch": 6.310639279944611,
+      "grad_norm": 3.5,
+      "learning_rate": 5.446297531687792e-05,
+      "loss": 1.3555,
+      "step": 6836
+    },
+    {
+      "epoch": 6.311562427879068,
+      "grad_norm": 3.64453125,
+      "learning_rate": 5.4456304202801866e-05,
+      "loss": 1.4065,
+      "step": 6837
+    },
+    {
+      "epoch": 6.312485575813524,
+      "grad_norm": 3.16796875,
+      "learning_rate": 5.444963308872582e-05,
+      "loss": 1.2036,
+      "step": 6838
+    },
+    {
+      "epoch": 6.313408723747981,
+      "grad_norm": 3.7734375,
+      "learning_rate": 5.4442961974649764e-05,
+      "loss": 1.1904,
+      "step": 6839
+    },
+    {
+      "epoch": 6.3143318716824375,
+      "grad_norm": 3.865234375,
+      "learning_rate": 5.4436290860573716e-05,
+      "loss": 1.3125,
+      "step": 6840
+    },
+    {
+      "epoch": 6.3152550196168935,
+      "grad_norm": 3.056640625,
+      "learning_rate": 5.4429619746497675e-05,
+      "loss": 1.3982,
+      "step": 6841
+    },
+    {
+      "epoch": 6.31617816755135,
+      "grad_norm": 3.51953125,
+      "learning_rate": 5.442294863242162e-05,
+      "loss": 1.384,
+      "step": 6842
+    },
+    {
+      "epoch": 6.3171013154858064,
+      "grad_norm": 2.904296875,
+      "learning_rate": 5.441627751834557e-05,
+      "loss": 1.3794,
+      "step": 6843
+    },
+    {
+      "epoch": 6.318024463420263,
+      "grad_norm": 3.732421875,
+      "learning_rate": 5.440960640426952e-05,
+      "loss": 1.4922,
+      "step": 6844
+    },
+    {
+      "epoch": 6.318947611354719,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.440293529019347e-05,
+      "loss": 1.3179,
+      "step": 6845
+    },
+    {
+      "epoch": 6.319870759289176,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.4396264176117416e-05,
+      "loss": 1.4414,
+      "step": 6846
+    },
+    {
+      "epoch": 6.320793907223632,
+      "grad_norm": 3.439453125,
+      "learning_rate": 5.438959306204137e-05,
+      "loss": 0.8601,
+      "step": 6847
+    },
+    {
+      "epoch": 6.321717055158089,
+      "grad_norm": 3.53515625,
+      "learning_rate": 5.4382921947965313e-05,
+      "loss": 1.3571,
+      "step": 6848
+    },
+    {
+      "epoch": 6.322640203092545,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.4376250833889266e-05,
+      "loss": 1.2408,
+      "step": 6849
+    },
+    {
+      "epoch": 6.323563351027002,
+      "grad_norm": 3.4765625,
+      "learning_rate": 5.436957971981321e-05,
+      "loss": 1.3308,
+      "step": 6850
+    },
+    {
+      "epoch": 6.324486498961458,
+      "grad_norm": 3.69140625,
+      "learning_rate": 5.436290860573716e-05,
+      "loss": 1.2952,
+      "step": 6851
+    },
+    {
+      "epoch": 6.325409646895915,
+      "grad_norm": 3.638671875,
+      "learning_rate": 5.435623749166111e-05,
+      "loss": 1.4143,
+      "step": 6852
+    },
+    {
+      "epoch": 6.326332794830371,
+      "grad_norm": 3.5703125,
+      "learning_rate": 5.434956637758506e-05,
+      "loss": 1.3818,
+      "step": 6853
+    },
+    {
+      "epoch": 6.327255942764828,
+      "grad_norm": 3.904296875,
+      "learning_rate": 5.4342895263509006e-05,
+      "loss": 1.3749,
+      "step": 6854
+    },
+    {
+      "epoch": 6.328179090699285,
+      "grad_norm": 4.26953125,
+      "learning_rate": 5.433622414943296e-05,
+      "loss": 1.1105,
+      "step": 6855
+    },
+    {
+      "epoch": 6.329102238633741,
+      "grad_norm": 3.396484375,
+      "learning_rate": 5.4329553035356904e-05,
+      "loss": 1.4294,
+      "step": 6856
+    },
+    {
+      "epoch": 6.330025386568198,
+      "grad_norm": 3.279296875,
+      "learning_rate": 5.4322881921280856e-05,
+      "loss": 1.5608,
+      "step": 6857
+    },
+    {
+      "epoch": 6.330948534502654,
+      "grad_norm": 3.173828125,
+      "learning_rate": 5.43162108072048e-05,
+      "loss": 1.614,
+      "step": 6858
+    },
+    {
+      "epoch": 6.331871682437111,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.4309539693128754e-05,
+      "loss": 1.5063,
+      "step": 6859
+    },
+    {
+      "epoch": 6.332794830371567,
+      "grad_norm": 4.0859375,
+      "learning_rate": 5.43028685790527e-05,
+      "loss": 1.3672,
+      "step": 6860
+    },
+    {
+      "epoch": 6.333717978306024,
+      "grad_norm": 3.896484375,
+      "learning_rate": 5.429619746497665e-05,
+      "loss": 1.4614,
+      "step": 6861
+    },
+    {
+      "epoch": 6.33464112624048,
+      "grad_norm": 3.251953125,
+      "learning_rate": 5.42895263509006e-05,
+      "loss": 1.4084,
+      "step": 6862
+    },
+    {
+      "epoch": 6.335564274174937,
+      "grad_norm": 3.76953125,
+      "learning_rate": 5.428285523682455e-05,
+      "loss": 1.3718,
+      "step": 6863
+    },
+    {
+      "epoch": 6.336487422109393,
+      "grad_norm": 3.873046875,
+      "learning_rate": 5.4276184122748495e-05,
+      "loss": 1.0338,
+      "step": 6864
+    },
+    {
+      "epoch": 6.33741057004385,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.426951300867245e-05,
+      "loss": 1.2366,
+      "step": 6865
+    },
+    {
+      "epoch": 6.338333717978306,
+      "grad_norm": 3.861328125,
+      "learning_rate": 5.4262841894596406e-05,
+      "loss": 0.9758,
+      "step": 6866
+    },
+    {
+      "epoch": 6.339256865912763,
+      "grad_norm": 4.10546875,
+      "learning_rate": 5.425617078052035e-05,
+      "loss": 1.4081,
+      "step": 6867
+    },
+    {
+      "epoch": 6.340180013847219,
+      "grad_norm": 3.56640625,
+      "learning_rate": 5.42494996664443e-05,
+      "loss": 1.4236,
+      "step": 6868
+    },
+    {
+      "epoch": 6.3411031617816755,
+      "grad_norm": 3.0,
+      "learning_rate": 5.4242828552368256e-05,
+      "loss": 1.0082,
+      "step": 6869
+    },
+    {
+      "epoch": 6.3420263097161325,
+      "grad_norm": 3.5859375,
+      "learning_rate": 5.42361574382922e-05,
+      "loss": 1.5042,
+      "step": 6870
+    },
+    {
+      "epoch": 6.3429494576505885,
+      "grad_norm": 3.8203125,
+      "learning_rate": 5.422948632421615e-05,
+      "loss": 1.2739,
+      "step": 6871
+    },
+    {
+      "epoch": 6.343872605585045,
+      "grad_norm": 3.32421875,
+      "learning_rate": 5.42228152101401e-05,
+      "loss": 1.3374,
+      "step": 6872
+    },
+    {
+      "epoch": 6.344795753519501,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.421614409606405e-05,
+      "loss": 1.0155,
+      "step": 6873
+    },
+    {
+      "epoch": 6.345718901453958,
+      "grad_norm": 4.3203125,
+      "learning_rate": 5.4209472981987996e-05,
+      "loss": 1.4067,
+      "step": 6874
+    },
+    {
+      "epoch": 6.346642049388414,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.420280186791195e-05,
+      "loss": 1.1032,
+      "step": 6875
+    },
+    {
+      "epoch": 6.347565197322871,
+      "grad_norm": 3.814453125,
+      "learning_rate": 5.4196130753835894e-05,
+      "loss": 1.4045,
+      "step": 6876
+    },
+    {
+      "epoch": 6.348488345257327,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.4189459639759846e-05,
+      "loss": 1.3721,
+      "step": 6877
+    },
+    {
+      "epoch": 6.349411493191784,
+      "grad_norm": 3.681640625,
+      "learning_rate": 5.418278852568379e-05,
+      "loss": 1.2839,
+      "step": 6878
+    },
+    {
+      "epoch": 6.35033464112624,
+      "grad_norm": 3.244140625,
+      "learning_rate": 5.4176117411607744e-05,
+      "loss": 1.0317,
+      "step": 6879
+    },
+    {
+      "epoch": 6.351257789060697,
+      "grad_norm": 3.724609375,
+      "learning_rate": 5.416944629753169e-05,
+      "loss": 1.2894,
+      "step": 6880
+    },
+    {
+      "epoch": 6.352180936995153,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.416277518345564e-05,
+      "loss": 1.4829,
+      "step": 6881
+    },
+    {
+      "epoch": 6.35310408492961,
+      "grad_norm": 3.228515625,
+      "learning_rate": 5.415610406937959e-05,
+      "loss": 1.2532,
+      "step": 6882
+    },
+    {
+      "epoch": 6.354027232864066,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.414943295530354e-05,
+      "loss": 1.5464,
+      "step": 6883
+    },
+    {
+      "epoch": 6.354950380798523,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.4142761841227484e-05,
+      "loss": 1.1166,
+      "step": 6884
+    },
+    {
+      "epoch": 6.355873528732979,
+      "grad_norm": 4.25390625,
+      "learning_rate": 5.413609072715144e-05,
+      "loss": 1.332,
+      "step": 6885
+    },
+    {
+      "epoch": 6.356796676667436,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.412941961307538e-05,
+      "loss": 1.2959,
+      "step": 6886
+    },
+    {
+      "epoch": 6.357719824601892,
+      "grad_norm": 3.1796875,
+      "learning_rate": 5.4122748498999334e-05,
+      "loss": 1.4358,
+      "step": 6887
+    },
+    {
+      "epoch": 6.358642972536349,
+      "grad_norm": 4.4140625,
+      "learning_rate": 5.411607738492328e-05,
+      "loss": 1.5376,
+      "step": 6888
+    },
+    {
+      "epoch": 6.359566120470806,
+      "grad_norm": 3.01171875,
+      "learning_rate": 5.410940627084723e-05,
+      "loss": 1.4114,
+      "step": 6889
+    },
+    {
+      "epoch": 6.360489268405262,
+      "grad_norm": 3.888671875,
+      "learning_rate": 5.410273515677118e-05,
+      "loss": 0.9583,
+      "step": 6890
+    },
+    {
+      "epoch": 6.361412416339719,
+      "grad_norm": 3.466796875,
+      "learning_rate": 5.4096064042695136e-05,
+      "loss": 1.2864,
+      "step": 6891
+    },
+    {
+      "epoch": 6.362335564274175,
+      "grad_norm": 3.755859375,
+      "learning_rate": 5.408939292861909e-05,
+      "loss": 1.5597,
+      "step": 6892
+    },
+    {
+      "epoch": 6.363258712208632,
+      "grad_norm": 3.24609375,
+      "learning_rate": 5.4082721814543034e-05,
+      "loss": 1.1542,
+      "step": 6893
+    },
+    {
+      "epoch": 6.364181860143088,
+      "grad_norm": 3.03515625,
+      "learning_rate": 5.4076050700466986e-05,
+      "loss": 1.3376,
+      "step": 6894
+    },
+    {
+      "epoch": 6.365105008077545,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.406937958639093e-05,
+      "loss": 1.3894,
+      "step": 6895
+    },
+    {
+      "epoch": 6.366028156012001,
+      "grad_norm": 3.474609375,
+      "learning_rate": 5.4062708472314884e-05,
+      "loss": 1.2485,
+      "step": 6896
+    },
+    {
+      "epoch": 6.366951303946458,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.405603735823883e-05,
+      "loss": 1.3499,
+      "step": 6897
+    },
+    {
+      "epoch": 6.367874451880914,
+      "grad_norm": 3.513671875,
+      "learning_rate": 5.404936624416278e-05,
+      "loss": 1.1428,
+      "step": 6898
+    },
+    {
+      "epoch": 6.3687975998153705,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.404269513008673e-05,
+      "loss": 1.3914,
+      "step": 6899
+    },
+    {
+      "epoch": 6.3697207477498266,
+      "grad_norm": 4.13671875,
+      "learning_rate": 5.403602401601068e-05,
+      "loss": 1.3171,
+      "step": 6900
+    },
+    {
+      "epoch": 6.3706438956842835,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.4029352901934625e-05,
+      "loss": 1.271,
+      "step": 6901
+    },
+    {
+      "epoch": 6.3715670436187395,
+      "grad_norm": 3.037109375,
+      "learning_rate": 5.402268178785858e-05,
+      "loss": 1.2803,
+      "step": 6902
+    },
+    {
+      "epoch": 6.372490191553196,
+      "grad_norm": 3.51953125,
+      "learning_rate": 5.401601067378252e-05,
+      "loss": 1.396,
+      "step": 6903
+    },
+    {
+      "epoch": 6.373413339487653,
+      "grad_norm": 4.0703125,
+      "learning_rate": 5.4009339559706474e-05,
+      "loss": 1.27,
+      "step": 6904
+    },
+    {
+      "epoch": 6.374336487422109,
+      "grad_norm": 4.4609375,
+      "learning_rate": 5.400266844563042e-05,
+      "loss": 1.2086,
+      "step": 6905
+    },
+    {
+      "epoch": 6.375259635356566,
+      "grad_norm": 3.439453125,
+      "learning_rate": 5.399599733155437e-05,
+      "loss": 1.3809,
+      "step": 6906
+    },
+    {
+      "epoch": 6.376182783291022,
+      "grad_norm": 3.3984375,
+      "learning_rate": 5.398932621747832e-05,
+      "loss": 1.3024,
+      "step": 6907
+    },
+    {
+      "epoch": 6.377105931225479,
+      "grad_norm": 3.287109375,
+      "learning_rate": 5.398265510340227e-05,
+      "loss": 1.2449,
+      "step": 6908
+    },
+    {
+      "epoch": 6.378029079159935,
+      "grad_norm": 3.333984375,
+      "learning_rate": 5.3975983989326215e-05,
+      "loss": 1.1808,
+      "step": 6909
+    },
+    {
+      "epoch": 6.378952227094392,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.396931287525017e-05,
+      "loss": 1.2864,
+      "step": 6910
+    },
+    {
+      "epoch": 6.379875375028848,
+      "grad_norm": 3.11328125,
+      "learning_rate": 5.396264176117411e-05,
+      "loss": 1.4768,
+      "step": 6911
+    },
+    {
+      "epoch": 6.380798522963305,
+      "grad_norm": 3.80078125,
+      "learning_rate": 5.3955970647098065e-05,
+      "loss": 1.697,
+      "step": 6912
+    },
+    {
+      "epoch": 6.381721670897761,
+      "grad_norm": 3.83203125,
+      "learning_rate": 5.394929953302201e-05,
+      "loss": 1.4478,
+      "step": 6913
+    },
+    {
+      "epoch": 6.382644818832218,
+      "grad_norm": 3.0390625,
+      "learning_rate": 5.394262841894596e-05,
+      "loss": 1.5786,
+      "step": 6914
+    },
+    {
+      "epoch": 6.383567966766674,
+      "grad_norm": 3.56640625,
+      "learning_rate": 5.393595730486991e-05,
+      "loss": 1.5034,
+      "step": 6915
+    },
+    {
+      "epoch": 6.384491114701131,
+      "grad_norm": 3.189453125,
+      "learning_rate": 5.3929286190793874e-05,
+      "loss": 1.0975,
+      "step": 6916
+    },
+    {
+      "epoch": 6.385414262635587,
+      "grad_norm": 3.841796875,
+      "learning_rate": 5.392261507671782e-05,
+      "loss": 1.1586,
+      "step": 6917
+    },
+    {
+      "epoch": 6.386337410570044,
+      "grad_norm": 3.13671875,
+      "learning_rate": 5.391594396264177e-05,
+      "loss": 1.4702,
+      "step": 6918
+    },
+    {
+      "epoch": 6.387260558504501,
+      "grad_norm": 4.02734375,
+      "learning_rate": 5.390927284856572e-05,
+      "loss": 1.561,
+      "step": 6919
+    },
+    {
+      "epoch": 6.388183706438957,
+      "grad_norm": 3.970703125,
+      "learning_rate": 5.390260173448967e-05,
+      "loss": 1.563,
+      "step": 6920
+    },
+    {
+      "epoch": 6.389106854373414,
+      "grad_norm": 3.87890625,
+      "learning_rate": 5.3895930620413614e-05,
+      "loss": 1.429,
+      "step": 6921
+    },
+    {
+      "epoch": 6.39003000230787,
+      "grad_norm": 3.443359375,
+      "learning_rate": 5.388925950633757e-05,
+      "loss": 1.491,
+      "step": 6922
+    },
+    {
+      "epoch": 6.390953150242327,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.388258839226151e-05,
+      "loss": 1.4062,
+      "step": 6923
+    },
+    {
+      "epoch": 6.391876298176783,
+      "grad_norm": 3.36328125,
+      "learning_rate": 5.3875917278185464e-05,
+      "loss": 1.244,
+      "step": 6924
+    },
+    {
+      "epoch": 6.39279944611124,
+      "grad_norm": 3.376953125,
+      "learning_rate": 5.386924616410941e-05,
+      "loss": 1.5671,
+      "step": 6925
+    },
+    {
+      "epoch": 6.393722594045696,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.386257505003336e-05,
+      "loss": 1.4587,
+      "step": 6926
+    },
+    {
+      "epoch": 6.394645741980153,
+      "grad_norm": 3.400390625,
+      "learning_rate": 5.385590393595731e-05,
+      "loss": 1.3737,
+      "step": 6927
+    },
+    {
+      "epoch": 6.395568889914609,
+      "grad_norm": 3.369140625,
+      "learning_rate": 5.384923282188126e-05,
+      "loss": 1.4043,
+      "step": 6928
+    },
+    {
+      "epoch": 6.3964920378490655,
+      "grad_norm": 3.134765625,
+      "learning_rate": 5.3842561707805205e-05,
+      "loss": 1.3723,
+      "step": 6929
+    },
+    {
+      "epoch": 6.3974151857835215,
+      "grad_norm": 3.5546875,
+      "learning_rate": 5.383589059372916e-05,
+      "loss": 1.3064,
+      "step": 6930
+    },
+    {
+      "epoch": 6.398338333717978,
+      "grad_norm": 3.619140625,
+      "learning_rate": 5.38292194796531e-05,
+      "loss": 1.0028,
+      "step": 6931
+    },
+    {
+      "epoch": 6.3992614816524345,
+      "grad_norm": 3.751953125,
+      "learning_rate": 5.3822548365577055e-05,
+      "loss": 1.324,
+      "step": 6932
+    },
+    {
+      "epoch": 6.400184629586891,
+      "grad_norm": 4.234375,
+      "learning_rate": 5.3815877251501e-05,
+      "loss": 1.2979,
+      "step": 6933
+    },
+    {
+      "epoch": 6.401107777521347,
+      "grad_norm": 4.16015625,
+      "learning_rate": 5.380920613742495e-05,
+      "loss": 1.3293,
+      "step": 6934
+    },
+    {
+      "epoch": 6.402030925455804,
+      "grad_norm": 3.8984375,
+      "learning_rate": 5.38025350233489e-05,
+      "loss": 1.4216,
+      "step": 6935
+    },
+    {
+      "epoch": 6.40295407339026,
+      "grad_norm": 3.580078125,
+      "learning_rate": 5.379586390927285e-05,
+      "loss": 1.5217,
+      "step": 6936
+    },
+    {
+      "epoch": 6.403877221324717,
+      "grad_norm": 3.865234375,
+      "learning_rate": 5.3789192795196796e-05,
+      "loss": 1.3259,
+      "step": 6937
+    },
+    {
+      "epoch": 6.404800369259174,
+      "grad_norm": 3.560546875,
+      "learning_rate": 5.378252168112075e-05,
+      "loss": 1.3113,
+      "step": 6938
+    },
+    {
+      "epoch": 6.40572351719363,
+      "grad_norm": 3.19921875,
+      "learning_rate": 5.377585056704469e-05,
+      "loss": 1.0858,
+      "step": 6939
+    },
+    {
+      "epoch": 6.406646665128087,
+      "grad_norm": 3.50390625,
+      "learning_rate": 5.3769179452968645e-05,
+      "loss": 1.3237,
+      "step": 6940
+    },
+    {
+      "epoch": 6.407569813062543,
+      "grad_norm": 3.251953125,
+      "learning_rate": 5.3762508338892604e-05,
+      "loss": 1.4854,
+      "step": 6941
+    },
+    {
+      "epoch": 6.408492960997,
+      "grad_norm": 3.404296875,
+      "learning_rate": 5.375583722481655e-05,
+      "loss": 1.4155,
+      "step": 6942
+    },
+    {
+      "epoch": 6.409416108931456,
+      "grad_norm": 3.54296875,
+      "learning_rate": 5.37491661107405e-05,
+      "loss": 1.1565,
+      "step": 6943
+    },
+    {
+      "epoch": 6.410339256865913,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.374249499666445e-05,
+      "loss": 1.196,
+      "step": 6944
+    },
+    {
+      "epoch": 6.411262404800369,
+      "grad_norm": 4.07421875,
+      "learning_rate": 5.37358238825884e-05,
+      "loss": 1.5803,
+      "step": 6945
+    },
+    {
+      "epoch": 6.412185552734826,
+      "grad_norm": 3.5859375,
+      "learning_rate": 5.3729152768512345e-05,
+      "loss": 1.4595,
+      "step": 6946
+    },
+    {
+      "epoch": 6.413108700669282,
+      "grad_norm": 3.50390625,
+      "learning_rate": 5.37224816544363e-05,
+      "loss": 1.4756,
+      "step": 6947
+    },
+    {
+      "epoch": 6.414031848603739,
+      "grad_norm": 3.908203125,
+      "learning_rate": 5.371581054036024e-05,
+      "loss": 1.3528,
+      "step": 6948
+    },
+    {
+      "epoch": 6.414954996538195,
+      "grad_norm": 3.548828125,
+      "learning_rate": 5.3709139426284195e-05,
+      "loss": 1.1576,
+      "step": 6949
+    },
+    {
+      "epoch": 6.415878144472652,
+      "grad_norm": 3.935546875,
+      "learning_rate": 5.370246831220814e-05,
+      "loss": 1.4451,
+      "step": 6950
+    },
+    {
+      "epoch": 6.416801292407108,
+      "grad_norm": 3.45703125,
+      "learning_rate": 5.369579719813209e-05,
+      "loss": 1.1334,
+      "step": 6951
+    },
+    {
+      "epoch": 6.417724440341565,
+      "grad_norm": 4.4921875,
+      "learning_rate": 5.368912608405604e-05,
+      "loss": 1.4399,
+      "step": 6952
+    },
+    {
+      "epoch": 6.418647588276022,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.368245496997999e-05,
+      "loss": 1.3796,
+      "step": 6953
+    },
+    {
+      "epoch": 6.419570736210478,
+      "grad_norm": 3.501953125,
+      "learning_rate": 5.3675783855903936e-05,
+      "loss": 1.4124,
+      "step": 6954
+    },
+    {
+      "epoch": 6.420493884144935,
+      "grad_norm": 3.779296875,
+      "learning_rate": 5.366911274182789e-05,
+      "loss": 1.1522,
+      "step": 6955
+    },
+    {
+      "epoch": 6.421417032079391,
+      "grad_norm": 2.763671875,
+      "learning_rate": 5.366244162775183e-05,
+      "loss": 1.0243,
+      "step": 6956
+    },
+    {
+      "epoch": 6.4223401800138475,
+      "grad_norm": 3.7890625,
+      "learning_rate": 5.3655770513675785e-05,
+      "loss": 1.3738,
+      "step": 6957
+    },
+    {
+      "epoch": 6.423263327948304,
+      "grad_norm": 3.18359375,
+      "learning_rate": 5.364909939959973e-05,
+      "loss": 1.29,
+      "step": 6958
+    },
+    {
+      "epoch": 6.4241864758827605,
+      "grad_norm": 3.60546875,
+      "learning_rate": 5.364242828552368e-05,
+      "loss": 1.5913,
+      "step": 6959
+    },
+    {
+      "epoch": 6.4251096238172165,
+      "grad_norm": 3.548828125,
+      "learning_rate": 5.363575717144763e-05,
+      "loss": 1.5205,
+      "step": 6960
+    },
+    {
+      "epoch": 6.426032771751673,
+      "grad_norm": 3.41015625,
+      "learning_rate": 5.362908605737158e-05,
+      "loss": 1.311,
+      "step": 6961
+    },
+    {
+      "epoch": 6.4269559196861294,
+      "grad_norm": 3.044921875,
+      "learning_rate": 5.3622414943295526e-05,
+      "loss": 1.4573,
+      "step": 6962
+    },
+    {
+      "epoch": 6.427879067620586,
+      "grad_norm": 4.609375,
+      "learning_rate": 5.361574382921948e-05,
+      "loss": 1.6174,
+      "step": 6963
+    },
+    {
+      "epoch": 6.428802215555042,
+      "grad_norm": 3.33203125,
+      "learning_rate": 5.3609072715143424e-05,
+      "loss": 1.3701,
+      "step": 6964
+    },
+    {
+      "epoch": 6.429725363489499,
+      "grad_norm": 4.37890625,
+      "learning_rate": 5.3602401601067376e-05,
+      "loss": 1.3131,
+      "step": 6965
+    },
+    {
+      "epoch": 6.430648511423955,
+      "grad_norm": 3.2265625,
+      "learning_rate": 5.359573048699132e-05,
+      "loss": 1.1587,
+      "step": 6966
+    },
+    {
+      "epoch": 6.431571659358412,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.358905937291529e-05,
+      "loss": 1.3586,
+      "step": 6967
+    },
+    {
+      "epoch": 6.432494807292868,
+      "grad_norm": 3.66015625,
+      "learning_rate": 5.358238825883923e-05,
+      "loss": 1.1651,
+      "step": 6968
+    },
+    {
+      "epoch": 6.433417955227325,
+      "grad_norm": 4.28515625,
+      "learning_rate": 5.3575717144763185e-05,
+      "loss": 1.5906,
+      "step": 6969
+    },
+    {
+      "epoch": 6.434341103161781,
+      "grad_norm": 3.111328125,
+      "learning_rate": 5.356904603068713e-05,
+      "loss": 1.4004,
+      "step": 6970
+    },
+    {
+      "epoch": 6.435264251096238,
+      "grad_norm": 3.470703125,
+      "learning_rate": 5.356237491661108e-05,
+      "loss": 1.4243,
+      "step": 6971
+    },
+    {
+      "epoch": 6.436187399030695,
+      "grad_norm": 4.29296875,
+      "learning_rate": 5.355570380253503e-05,
+      "loss": 1.3066,
+      "step": 6972
+    },
+    {
+      "epoch": 6.437110546965151,
+      "grad_norm": 3.4765625,
+      "learning_rate": 5.354903268845898e-05,
+      "loss": 1.6987,
+      "step": 6973
+    },
+    {
+      "epoch": 6.438033694899608,
+      "grad_norm": 3.93359375,
+      "learning_rate": 5.3542361574382926e-05,
+      "loss": 1.2737,
+      "step": 6974
+    },
+    {
+      "epoch": 6.438956842834064,
+      "grad_norm": 3.892578125,
+      "learning_rate": 5.353569046030688e-05,
+      "loss": 1.481,
+      "step": 6975
+    },
+    {
+      "epoch": 6.439879990768521,
+      "grad_norm": 3.58203125,
+      "learning_rate": 5.352901934623082e-05,
+      "loss": 1.2905,
+      "step": 6976
+    },
+    {
+      "epoch": 6.440803138702977,
+      "grad_norm": 3.40234375,
+      "learning_rate": 5.3522348232154775e-05,
+      "loss": 1.3418,
+      "step": 6977
+    },
+    {
+      "epoch": 6.441726286637434,
+      "grad_norm": 3.810546875,
+      "learning_rate": 5.351567711807872e-05,
+      "loss": 1.0471,
+      "step": 6978
+    },
+    {
+      "epoch": 6.44264943457189,
+      "grad_norm": 4.265625,
+      "learning_rate": 5.350900600400267e-05,
+      "loss": 1.2908,
+      "step": 6979
+    },
+    {
+      "epoch": 6.443572582506347,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.350233488992662e-05,
+      "loss": 1.2444,
+      "step": 6980
+    },
+    {
+      "epoch": 6.444495730440803,
+      "grad_norm": 3.708984375,
+      "learning_rate": 5.349566377585057e-05,
+      "loss": 1.3146,
+      "step": 6981
+    },
+    {
+      "epoch": 6.44541887837526,
+      "grad_norm": 3.03515625,
+      "learning_rate": 5.3488992661774516e-05,
+      "loss": 1.4043,
+      "step": 6982
+    },
+    {
+      "epoch": 6.446342026309716,
+      "grad_norm": 4.171875,
+      "learning_rate": 5.348232154769847e-05,
+      "loss": 1.427,
+      "step": 6983
+    },
+    {
+      "epoch": 6.447265174244173,
+      "grad_norm": 3.361328125,
+      "learning_rate": 5.3475650433622414e-05,
+      "loss": 1.502,
+      "step": 6984
+    },
+    {
+      "epoch": 6.448188322178629,
+      "grad_norm": 3.857421875,
+      "learning_rate": 5.3468979319546366e-05,
+      "loss": 1.5178,
+      "step": 6985
+    },
+    {
+      "epoch": 6.449111470113086,
+      "grad_norm": 3.494140625,
+      "learning_rate": 5.346230820547031e-05,
+      "loss": 1.511,
+      "step": 6986
+    },
+    {
+      "epoch": 6.4500346180475425,
+      "grad_norm": 3.625,
+      "learning_rate": 5.3455637091394264e-05,
+      "loss": 1.4092,
+      "step": 6987
+    },
+    {
+      "epoch": 6.4509577659819985,
+      "grad_norm": 4.6328125,
+      "learning_rate": 5.344896597731821e-05,
+      "loss": 1.4485,
+      "step": 6988
+    },
+    {
+      "epoch": 6.4518809139164555,
+      "grad_norm": 3.28515625,
+      "learning_rate": 5.344229486324216e-05,
+      "loss": 1.4148,
+      "step": 6989
+    },
+    {
+      "epoch": 6.4528040618509115,
+      "grad_norm": 3.431640625,
+      "learning_rate": 5.343562374916611e-05,
+      "loss": 1.511,
+      "step": 6990
+    },
+    {
+      "epoch": 6.453727209785368,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.342895263509006e-05,
+      "loss": 0.9821,
+      "step": 6991
+    },
+    {
+      "epoch": 6.454650357719824,
+      "grad_norm": 3.06640625,
+      "learning_rate": 5.342228152101402e-05,
+      "loss": 1.4888,
+      "step": 6992
+    },
+    {
+      "epoch": 6.455573505654281,
+      "grad_norm": 3.3828125,
+      "learning_rate": 5.341561040693796e-05,
+      "loss": 1.4026,
+      "step": 6993
+    },
+    {
+      "epoch": 6.456496653588737,
+      "grad_norm": 3.1171875,
+      "learning_rate": 5.3408939292861915e-05,
+      "loss": 1.2098,
+      "step": 6994
+    },
+    {
+      "epoch": 6.457419801523194,
+      "grad_norm": 3.541015625,
+      "learning_rate": 5.340226817878586e-05,
+      "loss": 1.5176,
+      "step": 6995
+    },
+    {
+      "epoch": 6.45834294945765,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.339559706470981e-05,
+      "loss": 0.8998,
+      "step": 6996
+    },
+    {
+      "epoch": 6.459266097392107,
+      "grad_norm": 3.396484375,
+      "learning_rate": 5.338892595063376e-05,
+      "loss": 1.5999,
+      "step": 6997
+    },
+    {
+      "epoch": 6.460189245326563,
+      "grad_norm": 3.75390625,
+      "learning_rate": 5.338225483655771e-05,
+      "loss": 1.2924,
+      "step": 6998
+    },
+    {
+      "epoch": 6.46111239326102,
+      "grad_norm": 3.30859375,
+      "learning_rate": 5.3375583722481656e-05,
+      "loss": 1.1412,
+      "step": 6999
+    },
+    {
+      "epoch": 6.462035541195476,
+      "grad_norm": 3.87109375,
+      "learning_rate": 5.336891260840561e-05,
+      "loss": 1.3413,
+      "step": 7000
+    },
+    {
+      "epoch": 6.462035541195476,
+      "eval_loss": 0.88134765625,
+      "eval_runtime": 23.3811,
+      "eval_samples_per_second": 975.187,
+      "eval_steps_per_second": 60.989,
+      "step": 7000
+    },
+    {
+      "epoch": 6.462958689129933,
+      "grad_norm": 3.271484375,
+      "learning_rate": 5.3362241494329554e-05,
+      "loss": 1.3523,
+      "step": 7001
+    },
+    {
+      "epoch": 6.46388183706439,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.3355570380253506e-05,
+      "loss": 0.9946,
+      "step": 7002
+    },
+    {
+      "epoch": 6.464804984998846,
+      "grad_norm": 3.30859375,
+      "learning_rate": 5.334889926617745e-05,
+      "loss": 1.406,
+      "step": 7003
+    },
+    {
+      "epoch": 6.465728132933303,
+      "grad_norm": 3.462890625,
+      "learning_rate": 5.3342228152101404e-05,
+      "loss": 1.6047,
+      "step": 7004
+    },
+    {
+      "epoch": 6.466651280867759,
+      "grad_norm": 3.05859375,
+      "learning_rate": 5.333555703802535e-05,
+      "loss": 1.342,
+      "step": 7005
+    },
+    {
+      "epoch": 6.467574428802216,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.33288859239493e-05,
+      "loss": 1.2163,
+      "step": 7006
+    },
+    {
+      "epoch": 6.468497576736672,
+      "grad_norm": 4.20703125,
+      "learning_rate": 5.332221480987325e-05,
+      "loss": 1.3816,
+      "step": 7007
+    },
+    {
+      "epoch": 6.469420724671129,
+      "grad_norm": 3.501953125,
+      "learning_rate": 5.33155436957972e-05,
+      "loss": 1.3052,
+      "step": 7008
+    },
+    {
+      "epoch": 6.470343872605585,
+      "grad_norm": 3.9140625,
+      "learning_rate": 5.3308872581721144e-05,
+      "loss": 1.3291,
+      "step": 7009
+    },
+    {
+      "epoch": 6.471267020540042,
+      "grad_norm": 3.87890625,
+      "learning_rate": 5.3302201467645097e-05,
+      "loss": 1.3528,
+      "step": 7010
+    },
+    {
+      "epoch": 6.472190168474498,
+      "grad_norm": 3.470703125,
+      "learning_rate": 5.329553035356904e-05,
+      "loss": 1.1499,
+      "step": 7011
+    },
+    {
+      "epoch": 6.473113316408955,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.3288859239492994e-05,
+      "loss": 1.1826,
+      "step": 7012
+    },
+    {
+      "epoch": 6.474036464343411,
+      "grad_norm": 3.9921875,
+      "learning_rate": 5.3282188125416946e-05,
+      "loss": 1.187,
+      "step": 7013
+    },
+    {
+      "epoch": 6.474959612277868,
+      "grad_norm": 3.189453125,
+      "learning_rate": 5.327551701134089e-05,
+      "loss": 1.3688,
+      "step": 7014
+    },
+    {
+      "epoch": 6.475882760212324,
+      "grad_norm": 3.494140625,
+      "learning_rate": 5.3268845897264844e-05,
+      "loss": 1.2064,
+      "step": 7015
+    },
+    {
+      "epoch": 6.476805908146781,
+      "grad_norm": 3.54296875,
+      "learning_rate": 5.326217478318879e-05,
+      "loss": 0.9094,
+      "step": 7016
+    },
+    {
+      "epoch": 6.477729056081237,
+      "grad_norm": 3.263671875,
+      "learning_rate": 5.325550366911275e-05,
+      "loss": 1.319,
+      "step": 7017
+    },
+    {
+      "epoch": 6.4786522040156935,
+      "grad_norm": 3.330078125,
+      "learning_rate": 5.32488325550367e-05,
+      "loss": 1.5046,
+      "step": 7018
+    },
+    {
+      "epoch": 6.4795753519501496,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.3242161440960646e-05,
+      "loss": 1.4121,
+      "step": 7019
+    },
+    {
+      "epoch": 6.4804984998846065,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.32354903268846e-05,
+      "loss": 1.3098,
+      "step": 7020
+    },
+    {
+      "epoch": 6.481421647819063,
+      "grad_norm": 3.396484375,
+      "learning_rate": 5.3228819212808544e-05,
+      "loss": 1.323,
+      "step": 7021
+    },
+    {
+      "epoch": 6.482344795753519,
+      "grad_norm": 4.25,
+      "learning_rate": 5.3222148098732496e-05,
+      "loss": 1.3789,
+      "step": 7022
+    },
+    {
+      "epoch": 6.483267943687976,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.321547698465644e-05,
+      "loss": 1.4562,
+      "step": 7023
+    },
+    {
+      "epoch": 6.484191091622432,
+      "grad_norm": 3.451171875,
+      "learning_rate": 5.3208805870580394e-05,
+      "loss": 1.6084,
+      "step": 7024
+    },
+    {
+      "epoch": 6.485114239556889,
+      "grad_norm": 3.3046875,
+      "learning_rate": 5.320213475650434e-05,
+      "loss": 1.3494,
+      "step": 7025
+    },
+    {
+      "epoch": 6.486037387491345,
+      "grad_norm": 3.302734375,
+      "learning_rate": 5.319546364242829e-05,
+      "loss": 1.4335,
+      "step": 7026
+    },
+    {
+      "epoch": 6.486960535425802,
+      "grad_norm": 3.46484375,
+      "learning_rate": 5.318879252835224e-05,
+      "loss": 1.4202,
+      "step": 7027
+    },
+    {
+      "epoch": 6.487883683360258,
+      "grad_norm": 3.802734375,
+      "learning_rate": 5.318212141427619e-05,
+      "loss": 1.4878,
+      "step": 7028
+    },
+    {
+      "epoch": 6.488806831294715,
+      "grad_norm": 3.69140625,
+      "learning_rate": 5.3175450300200134e-05,
+      "loss": 1.3352,
+      "step": 7029
+    },
+    {
+      "epoch": 6.489729979229171,
+      "grad_norm": 4.38671875,
+      "learning_rate": 5.3168779186124087e-05,
+      "loss": 1.5149,
+      "step": 7030
+    },
+    {
+      "epoch": 6.490653127163628,
+      "grad_norm": 3.185546875,
+      "learning_rate": 5.316210807204803e-05,
+      "loss": 1.0455,
+      "step": 7031
+    },
+    {
+      "epoch": 6.491576275098084,
+      "grad_norm": 3.634765625,
+      "learning_rate": 5.3155436957971984e-05,
+      "loss": 1.2599,
+      "step": 7032
+    },
+    {
+      "epoch": 6.492499423032541,
+      "grad_norm": 3.50390625,
+      "learning_rate": 5.314876584389593e-05,
+      "loss": 1.0415,
+      "step": 7033
+    },
+    {
+      "epoch": 6.493422570966997,
+      "grad_norm": 3.615234375,
+      "learning_rate": 5.314209472981988e-05,
+      "loss": 1.4862,
+      "step": 7034
+    },
+    {
+      "epoch": 6.494345718901454,
+      "grad_norm": 3.763671875,
+      "learning_rate": 5.313542361574383e-05,
+      "loss": 1.1523,
+      "step": 7035
+    },
+    {
+      "epoch": 6.495268866835911,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.312875250166778e-05,
+      "loss": 0.9863,
+      "step": 7036
+    },
+    {
+      "epoch": 6.496192014770367,
+      "grad_norm": 3.7734375,
+      "learning_rate": 5.3122081387591725e-05,
+      "loss": 1.5471,
+      "step": 7037
+    },
+    {
+      "epoch": 6.497115162704824,
+      "grad_norm": 4.01953125,
+      "learning_rate": 5.311541027351568e-05,
+      "loss": 1.0909,
+      "step": 7038
+    },
+    {
+      "epoch": 6.49803831063928,
+      "grad_norm": 3.46484375,
+      "learning_rate": 5.310873915943962e-05,
+      "loss": 1.1753,
+      "step": 7039
+    },
+    {
+      "epoch": 6.498961458573737,
+      "grad_norm": 3.830078125,
+      "learning_rate": 5.3102068045363575e-05,
+      "loss": 1.269,
+      "step": 7040
+    },
+    {
+      "epoch": 6.499884606508193,
+      "grad_norm": 3.646484375,
+      "learning_rate": 5.309539693128752e-05,
+      "loss": 1.5049,
+      "step": 7041
+    },
+    {
+      "epoch": 6.50080775444265,
+      "grad_norm": 3.634765625,
+      "learning_rate": 5.308872581721148e-05,
+      "loss": 1.1263,
+      "step": 7042
+    },
+    {
+      "epoch": 6.501730902377106,
+      "grad_norm": 3.326171875,
+      "learning_rate": 5.308205470313543e-05,
+      "loss": 1.4014,
+      "step": 7043
+    },
+    {
+      "epoch": 6.502654050311563,
+      "grad_norm": 3.853515625,
+      "learning_rate": 5.307538358905938e-05,
+      "loss": 1.3048,
+      "step": 7044
+    },
+    {
+      "epoch": 6.503577198246019,
+      "grad_norm": 4.0078125,
+      "learning_rate": 5.306871247498333e-05,
+      "loss": 1.3669,
+      "step": 7045
+    },
+    {
+      "epoch": 6.504500346180476,
+      "grad_norm": 3.5703125,
+      "learning_rate": 5.3062041360907274e-05,
+      "loss": 1.3545,
+      "step": 7046
+    },
+    {
+      "epoch": 6.505423494114932,
+      "grad_norm": 3.216796875,
+      "learning_rate": 5.3055370246831227e-05,
+      "loss": 1.2079,
+      "step": 7047
+    },
+    {
+      "epoch": 6.5063466420493885,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.304869913275517e-05,
+      "loss": 1.4517,
+      "step": 7048
+    },
+    {
+      "epoch": 6.5072697899838445,
+      "grad_norm": 4.01171875,
+      "learning_rate": 5.3042028018679124e-05,
+      "loss": 1.4885,
+      "step": 7049
+    },
+    {
+      "epoch": 6.5081929379183014,
+      "grad_norm": 3.392578125,
+      "learning_rate": 5.303535690460307e-05,
+      "loss": 1.3811,
+      "step": 7050
+    },
+    {
+      "epoch": 6.509116085852758,
+      "grad_norm": 3.529296875,
+      "learning_rate": 5.302868579052702e-05,
+      "loss": 1.3115,
+      "step": 7051
+    },
+    {
+      "epoch": 6.510039233787214,
+      "grad_norm": 3.630859375,
+      "learning_rate": 5.302201467645097e-05,
+      "loss": 0.9832,
+      "step": 7052
+    },
+    {
+      "epoch": 6.51096238172167,
+      "grad_norm": 4.1171875,
+      "learning_rate": 5.301534356237492e-05,
+      "loss": 1.2228,
+      "step": 7053
+    },
+    {
+      "epoch": 6.511885529656127,
+      "grad_norm": 3.423828125,
+      "learning_rate": 5.300867244829887e-05,
+      "loss": 1.2662,
+      "step": 7054
+    },
+    {
+      "epoch": 6.512808677590584,
+      "grad_norm": 4.17578125,
+      "learning_rate": 5.300200133422282e-05,
+      "loss": 1.2791,
+      "step": 7055
+    },
+    {
+      "epoch": 6.51373182552504,
+      "grad_norm": 3.15234375,
+      "learning_rate": 5.299533022014677e-05,
+      "loss": 1.249,
+      "step": 7056
+    },
+    {
+      "epoch": 6.514654973459497,
+      "grad_norm": 3.73046875,
+      "learning_rate": 5.2988659106070715e-05,
+      "loss": 1.3148,
+      "step": 7057
+    },
+    {
+      "epoch": 6.515578121393953,
+      "grad_norm": 3.2890625,
+      "learning_rate": 5.298198799199467e-05,
+      "loss": 1.2961,
+      "step": 7058
+    },
+    {
+      "epoch": 6.51650126932841,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.297531687791861e-05,
+      "loss": 1.3052,
+      "step": 7059
+    },
+    {
+      "epoch": 6.517424417262866,
+      "grad_norm": 3.61328125,
+      "learning_rate": 5.2968645763842565e-05,
+      "loss": 1.2832,
+      "step": 7060
+    },
+    {
+      "epoch": 6.518347565197323,
+      "grad_norm": 3.73046875,
+      "learning_rate": 5.296197464976651e-05,
+      "loss": 1.4822,
+      "step": 7061
+    },
+    {
+      "epoch": 6.519270713131779,
+      "grad_norm": 3.35546875,
+      "learning_rate": 5.295530353569046e-05,
+      "loss": 1.5027,
+      "step": 7062
+    },
+    {
+      "epoch": 6.520193861066236,
+      "grad_norm": 4.08984375,
+      "learning_rate": 5.294863242161441e-05,
+      "loss": 1.2521,
+      "step": 7063
+    },
+    {
+      "epoch": 6.521117009000692,
+      "grad_norm": 3.8984375,
+      "learning_rate": 5.294196130753836e-05,
+      "loss": 1.3365,
+      "step": 7064
+    },
+    {
+      "epoch": 6.522040156935149,
+      "grad_norm": 3.107421875,
+      "learning_rate": 5.2935290193462305e-05,
+      "loss": 1.5759,
+      "step": 7065
+    },
+    {
+      "epoch": 6.522963304869606,
+      "grad_norm": 3.77734375,
+      "learning_rate": 5.292861907938626e-05,
+      "loss": 1.3821,
+      "step": 7066
+    },
+    {
+      "epoch": 6.523886452804062,
+      "grad_norm": 3.533203125,
+      "learning_rate": 5.2921947965310217e-05,
+      "loss": 1.166,
+      "step": 7067
+    },
+    {
+      "epoch": 6.524809600738518,
+      "grad_norm": 3.197265625,
+      "learning_rate": 5.291527685123416e-05,
+      "loss": 1.035,
+      "step": 7068
+    },
+    {
+      "epoch": 6.525732748672975,
+      "grad_norm": 4.24609375,
+      "learning_rate": 5.2908605737158114e-05,
+      "loss": 1.282,
+      "step": 7069
+    },
+    {
+      "epoch": 6.526655896607432,
+      "grad_norm": 3.3359375,
+      "learning_rate": 5.290193462308206e-05,
+      "loss": 1.2548,
+      "step": 7070
+    },
+    {
+      "epoch": 6.527579044541888,
+      "grad_norm": 3.64453125,
+      "learning_rate": 5.289526350900601e-05,
+      "loss": 1.1401,
+      "step": 7071
+    },
+    {
+      "epoch": 6.528502192476345,
+      "grad_norm": 3.13671875,
+      "learning_rate": 5.288859239492996e-05,
+      "loss": 1.399,
+      "step": 7072
+    },
+    {
+      "epoch": 6.529425340410801,
+      "grad_norm": 3.26171875,
+      "learning_rate": 5.288192128085391e-05,
+      "loss": 1.3853,
+      "step": 7073
+    },
+    {
+      "epoch": 6.530348488345258,
+      "grad_norm": 3.82421875,
+      "learning_rate": 5.2875250166777855e-05,
+      "loss": 1.3698,
+      "step": 7074
+    },
+    {
+      "epoch": 6.531271636279714,
+      "grad_norm": 3.373046875,
+      "learning_rate": 5.286857905270181e-05,
+      "loss": 1.3738,
+      "step": 7075
+    },
+    {
+      "epoch": 6.5321947842141705,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.286190793862575e-05,
+      "loss": 1.4492,
+      "step": 7076
+    },
+    {
+      "epoch": 6.533117932148627,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.2855236824549705e-05,
+      "loss": 1.4724,
+      "step": 7077
+    },
+    {
+      "epoch": 6.5340410800830835,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.284856571047365e-05,
+      "loss": 1.3518,
+      "step": 7078
+    },
+    {
+      "epoch": 6.5349642280175395,
+      "grad_norm": 3.63671875,
+      "learning_rate": 5.28418945963976e-05,
+      "loss": 1.198,
+      "step": 7079
+    },
+    {
+      "epoch": 6.535887375951996,
+      "grad_norm": 3.486328125,
+      "learning_rate": 5.283522348232155e-05,
+      "loss": 1.4072,
+      "step": 7080
+    },
+    {
+      "epoch": 6.5368105238864525,
+      "grad_norm": 3.423828125,
+      "learning_rate": 5.28285523682455e-05,
+      "loss": 1.4387,
+      "step": 7081
+    },
+    {
+      "epoch": 6.537733671820909,
+      "grad_norm": 3.845703125,
+      "learning_rate": 5.2821881254169445e-05,
+      "loss": 1.301,
+      "step": 7082
+    },
+    {
+      "epoch": 6.538656819755365,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.28152101400934e-05,
+      "loss": 1.1927,
+      "step": 7083
+    },
+    {
+      "epoch": 6.539579967689822,
+      "grad_norm": 3.294921875,
+      "learning_rate": 5.280853902601734e-05,
+      "loss": 1.489,
+      "step": 7084
+    },
+    {
+      "epoch": 6.540503115624279,
+      "grad_norm": 4.24609375,
+      "learning_rate": 5.2801867911941295e-05,
+      "loss": 1.5127,
+      "step": 7085
+    },
+    {
+      "epoch": 6.541426263558735,
+      "grad_norm": 3.30078125,
+      "learning_rate": 5.279519679786524e-05,
+      "loss": 1.2361,
+      "step": 7086
+    },
+    {
+      "epoch": 6.542349411493192,
+      "grad_norm": 3.41796875,
+      "learning_rate": 5.278852568378919e-05,
+      "loss": 1.668,
+      "step": 7087
+    },
+    {
+      "epoch": 6.543272559427648,
+      "grad_norm": 2.93359375,
+      "learning_rate": 5.278185456971314e-05,
+      "loss": 1.2556,
+      "step": 7088
+    },
+    {
+      "epoch": 6.544195707362105,
+      "grad_norm": 3.412109375,
+      "learning_rate": 5.277518345563709e-05,
+      "loss": 1.458,
+      "step": 7089
+    },
+    {
+      "epoch": 6.545118855296561,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.2768512341561036e-05,
+      "loss": 1.3674,
+      "step": 7090
+    },
+    {
+      "epoch": 6.546042003231018,
+      "grad_norm": 4.46484375,
+      "learning_rate": 5.276184122748499e-05,
+      "loss": 1.4548,
+      "step": 7091
+    },
+    {
+      "epoch": 6.546965151165474,
+      "grad_norm": 4.50390625,
+      "learning_rate": 5.275517011340895e-05,
+      "loss": 1.1171,
+      "step": 7092
+    },
+    {
+      "epoch": 6.547888299099931,
+      "grad_norm": 3.6640625,
+      "learning_rate": 5.274849899933289e-05,
+      "loss": 1.1853,
+      "step": 7093
+    },
+    {
+      "epoch": 6.548811447034387,
+      "grad_norm": 3.41015625,
+      "learning_rate": 5.2741827885256845e-05,
+      "loss": 1.5588,
+      "step": 7094
+    },
+    {
+      "epoch": 6.549734594968844,
+      "grad_norm": 3.4453125,
+      "learning_rate": 5.27351567711808e-05,
+      "loss": 1.052,
+      "step": 7095
+    },
+    {
+      "epoch": 6.5506577429033,
+      "grad_norm": 3.16015625,
+      "learning_rate": 5.272848565710474e-05,
+      "loss": 1.3225,
+      "step": 7096
+    },
+    {
+      "epoch": 6.551580890837757,
+      "grad_norm": 3.76953125,
+      "learning_rate": 5.2721814543028695e-05,
+      "loss": 0.9952,
+      "step": 7097
+    },
+    {
+      "epoch": 6.552504038772213,
+      "grad_norm": 3.697265625,
+      "learning_rate": 5.271514342895264e-05,
+      "loss": 1.5347,
+      "step": 7098
+    },
+    {
+      "epoch": 6.55342718670667,
+      "grad_norm": 4.07421875,
+      "learning_rate": 5.270847231487659e-05,
+      "loss": 1.3838,
+      "step": 7099
+    },
+    {
+      "epoch": 6.554350334641127,
+      "grad_norm": 3.55078125,
+      "learning_rate": 5.270180120080054e-05,
+      "loss": 1.3647,
+      "step": 7100
+    },
+    {
+      "epoch": 6.555273482575583,
+      "grad_norm": 3.7890625,
+      "learning_rate": 5.269513008672449e-05,
+      "loss": 1.511,
+      "step": 7101
+    },
+    {
+      "epoch": 6.556196630510039,
+      "grad_norm": 4.54296875,
+      "learning_rate": 5.2688458972648435e-05,
+      "loss": 1.4036,
+      "step": 7102
+    },
+    {
+      "epoch": 6.557119778444496,
+      "grad_norm": 4.296875,
+      "learning_rate": 5.268178785857239e-05,
+      "loss": 1.5337,
+      "step": 7103
+    },
+    {
+      "epoch": 6.558042926378953,
+      "grad_norm": 2.931640625,
+      "learning_rate": 5.267511674449633e-05,
+      "loss": 1.5212,
+      "step": 7104
+    },
+    {
+      "epoch": 6.558966074313409,
+      "grad_norm": 3.5078125,
+      "learning_rate": 5.2668445630420285e-05,
+      "loss": 1.4518,
+      "step": 7105
+    },
+    {
+      "epoch": 6.5598892222478655,
+      "grad_norm": 3.94140625,
+      "learning_rate": 5.266177451634423e-05,
+      "loss": 1.4219,
+      "step": 7106
+    },
+    {
+      "epoch": 6.5608123701823216,
+      "grad_norm": 3.40234375,
+      "learning_rate": 5.265510340226818e-05,
+      "loss": 1.0319,
+      "step": 7107
+    },
+    {
+      "epoch": 6.5617355181167785,
+      "grad_norm": 3.64453125,
+      "learning_rate": 5.264843228819213e-05,
+      "loss": 1.2844,
+      "step": 7108
+    },
+    {
+      "epoch": 6.5626586660512345,
+      "grad_norm": 3.99609375,
+      "learning_rate": 5.264176117411608e-05,
+      "loss": 1.3414,
+      "step": 7109
+    },
+    {
+      "epoch": 6.563581813985691,
+      "grad_norm": 3.00390625,
+      "learning_rate": 5.2635090060040026e-05,
+      "loss": 1.2104,
+      "step": 7110
+    },
+    {
+      "epoch": 6.564504961920147,
+      "grad_norm": 3.83203125,
+      "learning_rate": 5.262841894596398e-05,
+      "loss": 1.1327,
+      "step": 7111
+    },
+    {
+      "epoch": 6.565428109854604,
+      "grad_norm": 3.47265625,
+      "learning_rate": 5.2621747831887924e-05,
+      "loss": 1.6023,
+      "step": 7112
+    },
+    {
+      "epoch": 6.56635125778906,
+      "grad_norm": 3.16796875,
+      "learning_rate": 5.2615076717811876e-05,
+      "loss": 1.4426,
+      "step": 7113
+    },
+    {
+      "epoch": 6.567274405723517,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.260840560373582e-05,
+      "loss": 1.343,
+      "step": 7114
+    },
+    {
+      "epoch": 6.568197553657973,
+      "grad_norm": 3.4296875,
+      "learning_rate": 5.260173448965977e-05,
+      "loss": 1.2776,
+      "step": 7115
+    },
+    {
+      "epoch": 6.56912070159243,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.259506337558372e-05,
+      "loss": 1.6289,
+      "step": 7116
+    },
+    {
+      "epoch": 6.570043849526886,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.258839226150768e-05,
+      "loss": 1.4822,
+      "step": 7117
+    },
+    {
+      "epoch": 6.570966997461343,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.258172114743163e-05,
+      "loss": 1.4058,
+      "step": 7118
+    },
+    {
+      "epoch": 6.5718901453958,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.2575050033355575e-05,
+      "loss": 1.0064,
+      "step": 7119
+    },
+    {
+      "epoch": 6.572813293330256,
+      "grad_norm": 4.77734375,
+      "learning_rate": 5.256837891927953e-05,
+      "loss": 1.4253,
+      "step": 7120
+    },
+    {
+      "epoch": 6.573736441264713,
+      "grad_norm": 3.3515625,
+      "learning_rate": 5.256170780520347e-05,
+      "loss": 1.3213,
+      "step": 7121
+    },
+    {
+      "epoch": 6.574659589199169,
+      "grad_norm": 3.951171875,
+      "learning_rate": 5.2555036691127425e-05,
+      "loss": 1.4048,
+      "step": 7122
+    },
+    {
+      "epoch": 6.575582737133626,
+      "grad_norm": 2.7578125,
+      "learning_rate": 5.254836557705137e-05,
+      "loss": 1.1682,
+      "step": 7123
+    },
+    {
+      "epoch": 6.576505885068082,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.254169446297532e-05,
+      "loss": 1.2125,
+      "step": 7124
+    },
+    {
+      "epoch": 6.577429033002539,
+      "grad_norm": 4.078125,
+      "learning_rate": 5.253502334889927e-05,
+      "loss": 1.2566,
+      "step": 7125
+    },
+    {
+      "epoch": 6.578352180936995,
+      "grad_norm": 2.947265625,
+      "learning_rate": 5.252835223482322e-05,
+      "loss": 1.3867,
+      "step": 7126
+    },
+    {
+      "epoch": 6.579275328871452,
+      "grad_norm": 3.548828125,
+      "learning_rate": 5.2521681120747166e-05,
+      "loss": 1.4299,
+      "step": 7127
+    },
+    {
+      "epoch": 6.580198476805908,
+      "grad_norm": 3.505859375,
+      "learning_rate": 5.251501000667112e-05,
+      "loss": 1.2358,
+      "step": 7128
+    },
+    {
+      "epoch": 6.581121624740365,
+      "grad_norm": 3.326171875,
+      "learning_rate": 5.2508338892595064e-05,
+      "loss": 1.3367,
+      "step": 7129
+    },
+    {
+      "epoch": 6.582044772674821,
+      "grad_norm": 3.55078125,
+      "learning_rate": 5.2501667778519016e-05,
+      "loss": 1.0878,
+      "step": 7130
+    },
+    {
+      "epoch": 6.582967920609278,
+      "grad_norm": 3.6171875,
+      "learning_rate": 5.249499666444296e-05,
+      "loss": 1.1844,
+      "step": 7131
+    },
+    {
+      "epoch": 6.583891068543734,
+      "grad_norm": 3.853515625,
+      "learning_rate": 5.2488325550366913e-05,
+      "loss": 1.1731,
+      "step": 7132
+    },
+    {
+      "epoch": 6.584814216478191,
+      "grad_norm": 3.037109375,
+      "learning_rate": 5.248165443629086e-05,
+      "loss": 1.614,
+      "step": 7133
+    },
+    {
+      "epoch": 6.585737364412648,
+      "grad_norm": 3.42578125,
+      "learning_rate": 5.247498332221481e-05,
+      "loss": 1.2771,
+      "step": 7134
+    },
+    {
+      "epoch": 6.586660512347104,
+      "grad_norm": 3.5234375,
+      "learning_rate": 5.2468312208138757e-05,
+      "loss": 1.4678,
+      "step": 7135
+    },
+    {
+      "epoch": 6.5875836602815605,
+      "grad_norm": 3.7890625,
+      "learning_rate": 5.246164109406271e-05,
+      "loss": 1.3142,
+      "step": 7136
+    },
+    {
+      "epoch": 6.5885068082160165,
+      "grad_norm": 3.09765625,
+      "learning_rate": 5.2454969979986654e-05,
+      "loss": 1.3306,
+      "step": 7137
+    },
+    {
+      "epoch": 6.5894299561504734,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.2448298865910606e-05,
+      "loss": 1.3584,
+      "step": 7138
+    },
+    {
+      "epoch": 6.5903531040849295,
+      "grad_norm": 3.462890625,
+      "learning_rate": 5.244162775183455e-05,
+      "loss": 0.9485,
+      "step": 7139
+    },
+    {
+      "epoch": 6.591276252019386,
+      "grad_norm": 3.57421875,
+      "learning_rate": 5.2434956637758504e-05,
+      "loss": 1.3296,
+      "step": 7140
+    },
+    {
+      "epoch": 6.592199399953842,
+      "grad_norm": 2.853515625,
+      "learning_rate": 5.242828552368245e-05,
+      "loss": 1.5244,
+      "step": 7141
+    },
+    {
+      "epoch": 6.593122547888299,
+      "grad_norm": 3.919921875,
+      "learning_rate": 5.2421614409606415e-05,
+      "loss": 1.3527,
+      "step": 7142
+    },
+    {
+      "epoch": 6.594045695822755,
+      "grad_norm": 3.130859375,
+      "learning_rate": 5.241494329553036e-05,
+      "loss": 1.3716,
+      "step": 7143
+    },
+    {
+      "epoch": 6.594968843757212,
+      "grad_norm": 3.947265625,
+      "learning_rate": 5.240827218145431e-05,
+      "loss": 1.322,
+      "step": 7144
+    },
+    {
+      "epoch": 6.595891991691668,
+      "grad_norm": 3.94921875,
+      "learning_rate": 5.240160106737826e-05,
+      "loss": 1.2175,
+      "step": 7145
+    },
+    {
+      "epoch": 6.596815139626125,
+      "grad_norm": 3.716796875,
+      "learning_rate": 5.239492995330221e-05,
+      "loss": 1.3049,
+      "step": 7146
+    },
+    {
+      "epoch": 6.597738287560581,
+      "grad_norm": 3.23046875,
+      "learning_rate": 5.2388258839226156e-05,
+      "loss": 1.2682,
+      "step": 7147
+    },
+    {
+      "epoch": 6.598661435495038,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.238158772515011e-05,
+      "loss": 1.5481,
+      "step": 7148
+    },
+    {
+      "epoch": 6.599584583429495,
+      "grad_norm": 3.166015625,
+      "learning_rate": 5.2374916611074054e-05,
+      "loss": 1.0714,
+      "step": 7149
+    },
+    {
+      "epoch": 6.600507731363951,
+      "grad_norm": 3.71484375,
+      "learning_rate": 5.2368245496998006e-05,
+      "loss": 1.6106,
+      "step": 7150
+    },
+    {
+      "epoch": 6.601430879298407,
+      "grad_norm": 3.736328125,
+      "learning_rate": 5.236157438292195e-05,
+      "loss": 1.292,
+      "step": 7151
+    },
+    {
+      "epoch": 6.602354027232864,
+      "grad_norm": 4.078125,
+      "learning_rate": 5.23549032688459e-05,
+      "loss": 1.3578,
+      "step": 7152
+    },
+    {
+      "epoch": 6.603277175167321,
+      "grad_norm": 2.759765625,
+      "learning_rate": 5.234823215476985e-05,
+      "loss": 1.4282,
+      "step": 7153
+    },
+    {
+      "epoch": 6.604200323101777,
+      "grad_norm": 2.974609375,
+      "learning_rate": 5.23415610406938e-05,
+      "loss": 1.5259,
+      "step": 7154
+    },
+    {
+      "epoch": 6.605123471036234,
+      "grad_norm": 3.68359375,
+      "learning_rate": 5.2334889926617746e-05,
+      "loss": 1.4827,
+      "step": 7155
+    },
+    {
+      "epoch": 6.60604661897069,
+      "grad_norm": 2.8671875,
+      "learning_rate": 5.23282188125417e-05,
+      "loss": 1.1416,
+      "step": 7156
+    },
+    {
+      "epoch": 6.606969766905147,
+      "grad_norm": 3.705078125,
+      "learning_rate": 5.2321547698465644e-05,
+      "loss": 1.2195,
+      "step": 7157
+    },
+    {
+      "epoch": 6.607892914839603,
+      "grad_norm": 4.08984375,
+      "learning_rate": 5.2314876584389596e-05,
+      "loss": 1.3599,
+      "step": 7158
+    },
+    {
+      "epoch": 6.60881606277406,
+      "grad_norm": 3.298828125,
+      "learning_rate": 5.230820547031354e-05,
+      "loss": 1.438,
+      "step": 7159
+    },
+    {
+      "epoch": 6.609739210708516,
+      "grad_norm": 3.962890625,
+      "learning_rate": 5.2301534356237494e-05,
+      "loss": 1.3628,
+      "step": 7160
+    },
+    {
+      "epoch": 6.610662358642973,
+      "grad_norm": 3.05859375,
+      "learning_rate": 5.229486324216144e-05,
+      "loss": 1.2391,
+      "step": 7161
+    },
+    {
+      "epoch": 6.611585506577429,
+      "grad_norm": 3.4140625,
+      "learning_rate": 5.228819212808539e-05,
+      "loss": 1.1454,
+      "step": 7162
+    },
+    {
+      "epoch": 6.612508654511886,
+      "grad_norm": 3.908203125,
+      "learning_rate": 5.228152101400934e-05,
+      "loss": 1.4565,
+      "step": 7163
+    },
+    {
+      "epoch": 6.613431802446342,
+      "grad_norm": 3.60546875,
+      "learning_rate": 5.227484989993329e-05,
+      "loss": 1.4966,
+      "step": 7164
+    },
+    {
+      "epoch": 6.614354950380799,
+      "grad_norm": 3.564453125,
+      "learning_rate": 5.2268178785857235e-05,
+      "loss": 1.5513,
+      "step": 7165
+    },
+    {
+      "epoch": 6.615278098315255,
+      "grad_norm": 3.537109375,
+      "learning_rate": 5.226150767178119e-05,
+      "loss": 1.4421,
+      "step": 7166
+    },
+    {
+      "epoch": 6.6162012462497115,
+      "grad_norm": 4.296875,
+      "learning_rate": 5.2254836557705146e-05,
+      "loss": 1.2699,
+      "step": 7167
+    },
+    {
+      "epoch": 6.617124394184168,
+      "grad_norm": 4.421875,
+      "learning_rate": 5.224816544362909e-05,
+      "loss": 1.1558,
+      "step": 7168
+    },
+    {
+      "epoch": 6.6180475421186244,
+      "grad_norm": 4.78125,
+      "learning_rate": 5.2241494329553043e-05,
+      "loss": 1.3594,
+      "step": 7169
+    },
+    {
+      "epoch": 6.618970690053081,
+      "grad_norm": 4.03515625,
+      "learning_rate": 5.223482321547699e-05,
+      "loss": 1.3162,
+      "step": 7170
+    },
+    {
+      "epoch": 6.619893837987537,
+      "grad_norm": 2.68359375,
+      "learning_rate": 5.222815210140094e-05,
+      "loss": 1.3396,
+      "step": 7171
+    },
+    {
+      "epoch": 6.620816985921994,
+      "grad_norm": 3.63671875,
+      "learning_rate": 5.2221480987324887e-05,
+      "loss": 1.3663,
+      "step": 7172
+    },
+    {
+      "epoch": 6.62174013385645,
+      "grad_norm": 3.189453125,
+      "learning_rate": 5.221480987324884e-05,
+      "loss": 1.3752,
+      "step": 7173
+    },
+    {
+      "epoch": 6.622663281790907,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.2208138759172784e-05,
+      "loss": 1.1407,
+      "step": 7174
+    },
+    {
+      "epoch": 6.623586429725363,
+      "grad_norm": 3.130859375,
+      "learning_rate": 5.2201467645096736e-05,
+      "loss": 1.3419,
+      "step": 7175
+    },
+    {
+      "epoch": 6.62450957765982,
+      "grad_norm": 3.388671875,
+      "learning_rate": 5.219479653102068e-05,
+      "loss": 1.1096,
+      "step": 7176
+    },
+    {
+      "epoch": 6.625432725594276,
+      "grad_norm": 3.3828125,
+      "learning_rate": 5.2188125416944634e-05,
+      "loss": 1.3529,
+      "step": 7177
+    },
+    {
+      "epoch": 6.626355873528733,
+      "grad_norm": 3.220703125,
+      "learning_rate": 5.218145430286858e-05,
+      "loss": 1.1702,
+      "step": 7178
+    },
+    {
+      "epoch": 6.627279021463189,
+      "grad_norm": 3.462890625,
+      "learning_rate": 5.217478318879253e-05,
+      "loss": 1.4446,
+      "step": 7179
+    },
+    {
+      "epoch": 6.628202169397646,
+      "grad_norm": 3.388671875,
+      "learning_rate": 5.216811207471648e-05,
+      "loss": 1.4722,
+      "step": 7180
+    },
+    {
+      "epoch": 6.629125317332102,
+      "grad_norm": 3.353515625,
+      "learning_rate": 5.216144096064043e-05,
+      "loss": 1.4214,
+      "step": 7181
+    },
+    {
+      "epoch": 6.630048465266559,
+      "grad_norm": 3.712890625,
+      "learning_rate": 5.2154769846564375e-05,
+      "loss": 1.3103,
+      "step": 7182
+    },
+    {
+      "epoch": 6.630971613201016,
+      "grad_norm": 3.345703125,
+      "learning_rate": 5.214809873248833e-05,
+      "loss": 1.2065,
+      "step": 7183
+    },
+    {
+      "epoch": 6.631894761135472,
+      "grad_norm": 3.17578125,
+      "learning_rate": 5.214142761841227e-05,
+      "loss": 1.2854,
+      "step": 7184
+    },
+    {
+      "epoch": 6.632817909069928,
+      "grad_norm": 3.427734375,
+      "learning_rate": 5.2134756504336225e-05,
+      "loss": 1.3159,
+      "step": 7185
+    },
+    {
+      "epoch": 6.633741057004385,
+      "grad_norm": 4.015625,
+      "learning_rate": 5.212808539026017e-05,
+      "loss": 1.707,
+      "step": 7186
+    },
+    {
+      "epoch": 6.634664204938842,
+      "grad_norm": 3.71484375,
+      "learning_rate": 5.212141427618412e-05,
+      "loss": 1.3634,
+      "step": 7187
+    },
+    {
+      "epoch": 6.635587352873298,
+      "grad_norm": 3.951171875,
+      "learning_rate": 5.211474316210807e-05,
+      "loss": 1.5525,
+      "step": 7188
+    },
+    {
+      "epoch": 6.636510500807755,
+      "grad_norm": 3.302734375,
+      "learning_rate": 5.210807204803202e-05,
+      "loss": 1.6759,
+      "step": 7189
+    },
+    {
+      "epoch": 6.637433648742211,
+      "grad_norm": 3.29296875,
+      "learning_rate": 5.2101400933955965e-05,
+      "loss": 1.4443,
+      "step": 7190
+    },
+    {
+      "epoch": 6.638356796676668,
+      "grad_norm": 3.94140625,
+      "learning_rate": 5.209472981987992e-05,
+      "loss": 1.3009,
+      "step": 7191
+    },
+    {
+      "epoch": 6.639279944611124,
+      "grad_norm": 3.607421875,
+      "learning_rate": 5.2088058705803876e-05,
+      "loss": 1.4077,
+      "step": 7192
+    },
+    {
+      "epoch": 6.640203092545581,
+      "grad_norm": 3.626953125,
+      "learning_rate": 5.208138759172783e-05,
+      "loss": 1.3899,
+      "step": 7193
+    },
+    {
+      "epoch": 6.641126240480037,
+      "grad_norm": 3.21484375,
+      "learning_rate": 5.2074716477651774e-05,
+      "loss": 1.6321,
+      "step": 7194
+    },
+    {
+      "epoch": 6.6420493884144935,
+      "grad_norm": 3.5234375,
+      "learning_rate": 5.2068045363575726e-05,
+      "loss": 1.0876,
+      "step": 7195
+    },
+    {
+      "epoch": 6.64297253634895,
+      "grad_norm": 3.525390625,
+      "learning_rate": 5.206137424949967e-05,
+      "loss": 1.098,
+      "step": 7196
+    },
+    {
+      "epoch": 6.6438956842834065,
+      "grad_norm": 3.197265625,
+      "learning_rate": 5.2054703135423624e-05,
+      "loss": 1.0084,
+      "step": 7197
+    },
+    {
+      "epoch": 6.644818832217863,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.204803202134757e-05,
+      "loss": 1.396,
+      "step": 7198
+    },
+    {
+      "epoch": 6.645741980152319,
+      "grad_norm": 4.34765625,
+      "learning_rate": 5.204136090727152e-05,
+      "loss": 1.4258,
+      "step": 7199
+    },
+    {
+      "epoch": 6.6466651280867755,
+      "grad_norm": 3.185546875,
+      "learning_rate": 5.203468979319547e-05,
+      "loss": 1.3965,
+      "step": 7200
+    },
+    {
+      "epoch": 6.647588276021232,
+      "grad_norm": 3.431640625,
+      "learning_rate": 5.202801867911942e-05,
+      "loss": 1.3723,
+      "step": 7201
+    },
+    {
+      "epoch": 6.648511423955689,
+      "grad_norm": 3.87109375,
+      "learning_rate": 5.2021347565043365e-05,
+      "loss": 1.1525,
+      "step": 7202
+    },
+    {
+      "epoch": 6.649434571890145,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.201467645096732e-05,
+      "loss": 1.3174,
+      "step": 7203
+    },
+    {
+      "epoch": 6.650357719824602,
+      "grad_norm": 3.61328125,
+      "learning_rate": 5.200800533689126e-05,
+      "loss": 1.3892,
+      "step": 7204
+    },
+    {
+      "epoch": 6.651280867759058,
+      "grad_norm": 3.841796875,
+      "learning_rate": 5.2001334222815214e-05,
+      "loss": 1.3363,
+      "step": 7205
+    },
+    {
+      "epoch": 6.652204015693515,
+      "grad_norm": 3.755859375,
+      "learning_rate": 5.199466310873916e-05,
+      "loss": 1.3591,
+      "step": 7206
+    },
+    {
+      "epoch": 6.653127163627971,
+      "grad_norm": 4.02734375,
+      "learning_rate": 5.198799199466311e-05,
+      "loss": 1.2371,
+      "step": 7207
+    },
+    {
+      "epoch": 6.654050311562428,
+      "grad_norm": 4.0390625,
+      "learning_rate": 5.198132088058706e-05,
+      "loss": 1.2573,
+      "step": 7208
+    },
+    {
+      "epoch": 6.654973459496884,
+      "grad_norm": 3.177734375,
+      "learning_rate": 5.197464976651101e-05,
+      "loss": 1.4695,
+      "step": 7209
+    },
+    {
+      "epoch": 6.655896607431341,
+      "grad_norm": 3.88671875,
+      "learning_rate": 5.1967978652434955e-05,
+      "loss": 1.2766,
+      "step": 7210
+    },
+    {
+      "epoch": 6.656819755365797,
+      "grad_norm": 3.255859375,
+      "learning_rate": 5.196130753835891e-05,
+      "loss": 1.1995,
+      "step": 7211
+    },
+    {
+      "epoch": 6.657742903300254,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.195463642428285e-05,
+      "loss": 1.3711,
+      "step": 7212
+    },
+    {
+      "epoch": 6.65866605123471,
+      "grad_norm": 3.4453125,
+      "learning_rate": 5.1947965310206805e-05,
+      "loss": 1.554,
+      "step": 7213
+    },
+    {
+      "epoch": 6.659589199169167,
+      "grad_norm": 3.67578125,
+      "learning_rate": 5.194129419613075e-05,
+      "loss": 1.3865,
+      "step": 7214
+    },
+    {
+      "epoch": 6.660512347103623,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.19346230820547e-05,
+      "loss": 1.4927,
+      "step": 7215
+    },
+    {
+      "epoch": 6.66143549503808,
+      "grad_norm": 3.65234375,
+      "learning_rate": 5.192795196797865e-05,
+      "loss": 1.3969,
+      "step": 7216
+    },
+    {
+      "epoch": 6.662358642972537,
+      "grad_norm": 3.38671875,
+      "learning_rate": 5.192128085390261e-05,
+      "loss": 1.3118,
+      "step": 7217
+    },
+    {
+      "epoch": 6.663281790906993,
+      "grad_norm": 4.04296875,
+      "learning_rate": 5.191460973982656e-05,
+      "loss": 0.7992,
+      "step": 7218
+    },
+    {
+      "epoch": 6.66420493884145,
+      "grad_norm": 4.08203125,
+      "learning_rate": 5.1907938625750505e-05,
+      "loss": 1.4016,
+      "step": 7219
+    },
+    {
+      "epoch": 6.665128086775906,
+      "grad_norm": 3.63671875,
+      "learning_rate": 5.190126751167446e-05,
+      "loss": 1.3625,
+      "step": 7220
+    },
+    {
+      "epoch": 6.666051234710363,
+      "grad_norm": 3.177734375,
+      "learning_rate": 5.18945963975984e-05,
+      "loss": 1.3656,
+      "step": 7221
+    },
+    {
+      "epoch": 6.666974382644819,
+      "grad_norm": 3.236328125,
+      "learning_rate": 5.1887925283522355e-05,
+      "loss": 1.3926,
+      "step": 7222
+    },
+    {
+      "epoch": 6.667897530579276,
+      "grad_norm": 3.076171875,
+      "learning_rate": 5.18812541694463e-05,
+      "loss": 1.6128,
+      "step": 7223
+    },
+    {
+      "epoch": 6.668820678513732,
+      "grad_norm": 3.642578125,
+      "learning_rate": 5.187458305537025e-05,
+      "loss": 1.4238,
+      "step": 7224
+    },
+    {
+      "epoch": 6.6697438264481885,
+      "grad_norm": 3.369140625,
+      "learning_rate": 5.18679119412942e-05,
+      "loss": 1.4746,
+      "step": 7225
+    },
+    {
+      "epoch": 6.6706669743826446,
+      "grad_norm": 3.373046875,
+      "learning_rate": 5.186124082721815e-05,
+      "loss": 1.2024,
+      "step": 7226
+    },
+    {
+      "epoch": 6.6715901223171015,
+      "grad_norm": 3.017578125,
+      "learning_rate": 5.1854569713142095e-05,
+      "loss": 1.3115,
+      "step": 7227
+    },
+    {
+      "epoch": 6.6725132702515575,
+      "grad_norm": 3.697265625,
+      "learning_rate": 5.184789859906605e-05,
+      "loss": 1.3293,
+      "step": 7228
+    },
+    {
+      "epoch": 6.673436418186014,
+      "grad_norm": 3.29296875,
+      "learning_rate": 5.184122748498999e-05,
+      "loss": 1.2994,
+      "step": 7229
+    },
+    {
+      "epoch": 6.67435956612047,
+      "grad_norm": 3.353515625,
+      "learning_rate": 5.1834556370913945e-05,
+      "loss": 1.5415,
+      "step": 7230
+    },
+    {
+      "epoch": 6.675282714054927,
+      "grad_norm": 3.7109375,
+      "learning_rate": 5.182788525683789e-05,
+      "loss": 1.4666,
+      "step": 7231
+    },
+    {
+      "epoch": 6.676205861989384,
+      "grad_norm": 3.880859375,
+      "learning_rate": 5.182121414276184e-05,
+      "loss": 1.4368,
+      "step": 7232
+    },
+    {
+      "epoch": 6.67712900992384,
+      "grad_norm": 3.375,
+      "learning_rate": 5.181454302868579e-05,
+      "loss": 1.2689,
+      "step": 7233
+    },
+    {
+      "epoch": 6.678052157858296,
+      "grad_norm": 3.951171875,
+      "learning_rate": 5.180787191460974e-05,
+      "loss": 1.3967,
+      "step": 7234
+    },
+    {
+      "epoch": 6.678975305792753,
+      "grad_norm": 4.07421875,
+      "learning_rate": 5.1801200800533686e-05,
+      "loss": 1.478,
+      "step": 7235
+    },
+    {
+      "epoch": 6.67989845372721,
+      "grad_norm": 3.43359375,
+      "learning_rate": 5.179452968645764e-05,
+      "loss": 1.5396,
+      "step": 7236
+    },
+    {
+      "epoch": 6.680821601661666,
+      "grad_norm": 4.0,
+      "learning_rate": 5.1787858572381583e-05,
+      "loss": 1.3242,
+      "step": 7237
+    },
+    {
+      "epoch": 6.681744749596123,
+      "grad_norm": 3.693359375,
+      "learning_rate": 5.1781187458305536e-05,
+      "loss": 1.222,
+      "step": 7238
+    },
+    {
+      "epoch": 6.682667897530579,
+      "grad_norm": 3.865234375,
+      "learning_rate": 5.177451634422949e-05,
+      "loss": 1.4861,
+      "step": 7239
+    },
+    {
+      "epoch": 6.683591045465036,
+      "grad_norm": 2.97265625,
+      "learning_rate": 5.176784523015343e-05,
+      "loss": 1.3577,
+      "step": 7240
+    },
+    {
+      "epoch": 6.684514193399492,
+      "grad_norm": 3.990234375,
+      "learning_rate": 5.1761174116077386e-05,
+      "loss": 1.4663,
+      "step": 7241
+    },
+    {
+      "epoch": 6.685437341333949,
+      "grad_norm": 3.98828125,
+      "learning_rate": 5.1754503002001344e-05,
+      "loss": 1.2974,
+      "step": 7242
+    },
+    {
+      "epoch": 6.686360489268405,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.174783188792529e-05,
+      "loss": 1.3767,
+      "step": 7243
+    },
+    {
+      "epoch": 6.687283637202862,
+      "grad_norm": 3.48828125,
+      "learning_rate": 5.174116077384924e-05,
+      "loss": 1.2333,
+      "step": 7244
+    },
+    {
+      "epoch": 6.688206785137318,
+      "grad_norm": 3.72265625,
+      "learning_rate": 5.173448965977319e-05,
+      "loss": 1.2445,
+      "step": 7245
+    },
+    {
+      "epoch": 6.689129933071775,
+      "grad_norm": 3.71484375,
+      "learning_rate": 5.172781854569714e-05,
+      "loss": 1.5366,
+      "step": 7246
+    },
+    {
+      "epoch": 6.690053081006231,
+      "grad_norm": 4.05859375,
+      "learning_rate": 5.1721147431621085e-05,
+      "loss": 1.51,
+      "step": 7247
+    },
+    {
+      "epoch": 6.690976228940688,
+      "grad_norm": 2.865234375,
+      "learning_rate": 5.171447631754504e-05,
+      "loss": 1.1987,
+      "step": 7248
+    },
+    {
+      "epoch": 6.691899376875144,
+      "grad_norm": 3.1953125,
+      "learning_rate": 5.170780520346898e-05,
+      "loss": 1.1997,
+      "step": 7249
+    },
+    {
+      "epoch": 6.692822524809601,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.1701134089392935e-05,
+      "loss": 1.2622,
+      "step": 7250
+    },
+    {
+      "epoch": 6.693745672744058,
+      "grad_norm": 3.2421875,
+      "learning_rate": 5.169446297531688e-05,
+      "loss": 0.8822,
+      "step": 7251
+    },
+    {
+      "epoch": 6.694668820678514,
+      "grad_norm": 3.521484375,
+      "learning_rate": 5.168779186124083e-05,
+      "loss": 1.1559,
+      "step": 7252
+    },
+    {
+      "epoch": 6.695591968612971,
+      "grad_norm": 3.85546875,
+      "learning_rate": 5.168112074716478e-05,
+      "loss": 1.2141,
+      "step": 7253
+    },
+    {
+      "epoch": 6.696515116547427,
+      "grad_norm": 3.236328125,
+      "learning_rate": 5.167444963308873e-05,
+      "loss": 1.5344,
+      "step": 7254
+    },
+    {
+      "epoch": 6.6974382644818835,
+      "grad_norm": 4.16015625,
+      "learning_rate": 5.1667778519012676e-05,
+      "loss": 1.3704,
+      "step": 7255
+    },
+    {
+      "epoch": 6.6983614124163395,
+      "grad_norm": 3.419921875,
+      "learning_rate": 5.166110740493663e-05,
+      "loss": 1.1702,
+      "step": 7256
+    },
+    {
+      "epoch": 6.6992845603507964,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.1654436290860573e-05,
+      "loss": 1.2345,
+      "step": 7257
+    },
+    {
+      "epoch": 6.7002077082852525,
+      "grad_norm": 4.0390625,
+      "learning_rate": 5.1647765176784526e-05,
+      "loss": 1.2927,
+      "step": 7258
+    },
+    {
+      "epoch": 6.701130856219709,
+      "grad_norm": 3.6640625,
+      "learning_rate": 5.164109406270847e-05,
+      "loss": 1.5513,
+      "step": 7259
+    },
+    {
+      "epoch": 6.702054004154165,
+      "grad_norm": 3.49609375,
+      "learning_rate": 5.163442294863242e-05,
+      "loss": 1.2349,
+      "step": 7260
+    },
+    {
+      "epoch": 6.702977152088622,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.162775183455637e-05,
+      "loss": 1.3372,
+      "step": 7261
+    },
+    {
+      "epoch": 6.703900300023078,
+      "grad_norm": 3.490234375,
+      "learning_rate": 5.162108072048032e-05,
+      "loss": 1.2621,
+      "step": 7262
+    },
+    {
+      "epoch": 6.704823447957535,
+      "grad_norm": 3.583984375,
+      "learning_rate": 5.1614409606404266e-05,
+      "loss": 1.4224,
+      "step": 7263
+    },
+    {
+      "epoch": 6.705746595891991,
+      "grad_norm": 3.365234375,
+      "learning_rate": 5.160773849232822e-05,
+      "loss": 1.1014,
+      "step": 7264
+    },
+    {
+      "epoch": 6.706669743826448,
+      "grad_norm": 3.53515625,
+      "learning_rate": 5.1601067378252164e-05,
+      "loss": 1.3015,
+      "step": 7265
+    },
+    {
+      "epoch": 6.707592891760905,
+      "grad_norm": 3.7890625,
+      "learning_rate": 5.1594396264176116e-05,
+      "loss": 1.5024,
+      "step": 7266
+    },
+    {
+      "epoch": 6.708516039695361,
+      "grad_norm": 3.693359375,
+      "learning_rate": 5.1587725150100075e-05,
+      "loss": 1.3805,
+      "step": 7267
+    },
+    {
+      "epoch": 6.709439187629818,
+      "grad_norm": 3.201171875,
+      "learning_rate": 5.158105403602402e-05,
+      "loss": 1.1681,
+      "step": 7268
+    },
+    {
+      "epoch": 6.710362335564274,
+      "grad_norm": 3.869140625,
+      "learning_rate": 5.157438292194797e-05,
+      "loss": 1.4077,
+      "step": 7269
+    },
+    {
+      "epoch": 6.711285483498731,
+      "grad_norm": 3.46484375,
+      "learning_rate": 5.156771180787192e-05,
+      "loss": 1.4727,
+      "step": 7270
+    },
+    {
+      "epoch": 6.712208631433187,
+      "grad_norm": 3.4765625,
+      "learning_rate": 5.156104069379587e-05,
+      "loss": 1.2812,
+      "step": 7271
+    },
+    {
+      "epoch": 6.713131779367644,
+      "grad_norm": 4.1328125,
+      "learning_rate": 5.1554369579719816e-05,
+      "loss": 1.7766,
+      "step": 7272
+    },
+    {
+      "epoch": 6.7140549273021,
+      "grad_norm": 3.568359375,
+      "learning_rate": 5.154769846564377e-05,
+      "loss": 0.8437,
+      "step": 7273
+    },
+    {
+      "epoch": 6.714978075236557,
+      "grad_norm": 3.255859375,
+      "learning_rate": 5.1541027351567713e-05,
+      "loss": 1.3303,
+      "step": 7274
+    },
+    {
+      "epoch": 6.715901223171013,
+      "grad_norm": 3.857421875,
+      "learning_rate": 5.1534356237491666e-05,
+      "loss": 1.426,
+      "step": 7275
+    },
+    {
+      "epoch": 6.71682437110547,
+      "grad_norm": 3.33203125,
+      "learning_rate": 5.152768512341561e-05,
+      "loss": 1.3425,
+      "step": 7276
+    },
+    {
+      "epoch": 6.717747519039926,
+      "grad_norm": 3.34765625,
+      "learning_rate": 5.152101400933956e-05,
+      "loss": 1.2281,
+      "step": 7277
+    },
+    {
+      "epoch": 6.718670666974383,
+      "grad_norm": 3.513671875,
+      "learning_rate": 5.151434289526351e-05,
+      "loss": 1.5156,
+      "step": 7278
+    },
+    {
+      "epoch": 6.719593814908839,
+      "grad_norm": 3.82421875,
+      "learning_rate": 5.150767178118746e-05,
+      "loss": 1.2917,
+      "step": 7279
+    },
+    {
+      "epoch": 6.720516962843296,
+      "grad_norm": 3.626953125,
+      "learning_rate": 5.150100066711141e-05,
+      "loss": 1.4563,
+      "step": 7280
+    },
+    {
+      "epoch": 6.721440110777753,
+      "grad_norm": 3.341796875,
+      "learning_rate": 5.149432955303536e-05,
+      "loss": 1.0701,
+      "step": 7281
+    },
+    {
+      "epoch": 6.722363258712209,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.148765843895931e-05,
+      "loss": 1.355,
+      "step": 7282
+    },
+    {
+      "epoch": 6.723286406646665,
+      "grad_norm": 3.25390625,
+      "learning_rate": 5.1480987324883256e-05,
+      "loss": 1.3081,
+      "step": 7283
+    },
+    {
+      "epoch": 6.724209554581122,
+      "grad_norm": 3.173828125,
+      "learning_rate": 5.147431621080721e-05,
+      "loss": 1.5386,
+      "step": 7284
+    },
+    {
+      "epoch": 6.7251327025155785,
+      "grad_norm": 3.138671875,
+      "learning_rate": 5.1467645096731154e-05,
+      "loss": 1.572,
+      "step": 7285
+    },
+    {
+      "epoch": 6.7260558504500345,
+      "grad_norm": 3.498046875,
+      "learning_rate": 5.1460973982655106e-05,
+      "loss": 1.0743,
+      "step": 7286
+    },
+    {
+      "epoch": 6.726978998384491,
+      "grad_norm": 3.263671875,
+      "learning_rate": 5.145430286857905e-05,
+      "loss": 1.2493,
+      "step": 7287
+    },
+    {
+      "epoch": 6.7279021463189475,
+      "grad_norm": 3.498046875,
+      "learning_rate": 5.1447631754503004e-05,
+      "loss": 1.5146,
+      "step": 7288
+    },
+    {
+      "epoch": 6.728825294253404,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.144096064042695e-05,
+      "loss": 1.2698,
+      "step": 7289
+    },
+    {
+      "epoch": 6.72974844218786,
+      "grad_norm": 2.99609375,
+      "learning_rate": 5.14342895263509e-05,
+      "loss": 1.3372,
+      "step": 7290
+    },
+    {
+      "epoch": 6.730671590122317,
+      "grad_norm": 3.0546875,
+      "learning_rate": 5.142761841227485e-05,
+      "loss": 1.2903,
+      "step": 7291
+    },
+    {
+      "epoch": 6.731594738056773,
+      "grad_norm": 3.458984375,
+      "learning_rate": 5.1420947298198806e-05,
+      "loss": 1.4963,
+      "step": 7292
+    },
+    {
+      "epoch": 6.73251788599123,
+      "grad_norm": 3.28515625,
+      "learning_rate": 5.141427618412276e-05,
+      "loss": 1.2847,
+      "step": 7293
+    },
+    {
+      "epoch": 6.733441033925686,
+      "grad_norm": 3.34765625,
+      "learning_rate": 5.1407605070046703e-05,
+      "loss": 1.5779,
+      "step": 7294
+    },
+    {
+      "epoch": 6.734364181860143,
+      "grad_norm": 4.25,
+      "learning_rate": 5.1400933955970656e-05,
+      "loss": 0.7534,
+      "step": 7295
+    },
+    {
+      "epoch": 6.735287329794599,
+      "grad_norm": 4.8671875,
+      "learning_rate": 5.13942628418946e-05,
+      "loss": 1.5178,
+      "step": 7296
+    },
+    {
+      "epoch": 6.736210477729056,
+      "grad_norm": 3.365234375,
+      "learning_rate": 5.138759172781855e-05,
+      "loss": 1.689,
+      "step": 7297
+    },
+    {
+      "epoch": 6.737133625663512,
+      "grad_norm": 3.87109375,
+      "learning_rate": 5.13809206137425e-05,
+      "loss": 1.3484,
+      "step": 7298
+    },
+    {
+      "epoch": 6.738056773597969,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.137424949966645e-05,
+      "loss": 1.3738,
+      "step": 7299
+    },
+    {
+      "epoch": 6.738979921532426,
+      "grad_norm": 3.740234375,
+      "learning_rate": 5.1367578385590396e-05,
+      "loss": 1.4661,
+      "step": 7300
+    },
+    {
+      "epoch": 6.739903069466882,
+      "grad_norm": 3.884765625,
+      "learning_rate": 5.136090727151435e-05,
+      "loss": 1.2817,
+      "step": 7301
+    },
+    {
+      "epoch": 6.740826217401339,
+      "grad_norm": 3.09765625,
+      "learning_rate": 5.1354236157438294e-05,
+      "loss": 1.4829,
+      "step": 7302
+    },
+    {
+      "epoch": 6.741749365335795,
+      "grad_norm": 3.7578125,
+      "learning_rate": 5.1347565043362246e-05,
+      "loss": 1.4814,
+      "step": 7303
+    },
+    {
+      "epoch": 6.742672513270252,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.134089392928619e-05,
+      "loss": 1.3555,
+      "step": 7304
+    },
+    {
+      "epoch": 6.743595661204708,
+      "grad_norm": 3.580078125,
+      "learning_rate": 5.1334222815210144e-05,
+      "loss": 1.6355,
+      "step": 7305
+    },
+    {
+      "epoch": 6.744518809139165,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.132755170113409e-05,
+      "loss": 1.2069,
+      "step": 7306
+    },
+    {
+      "epoch": 6.745441957073621,
+      "grad_norm": 3.33984375,
+      "learning_rate": 5.132088058705804e-05,
+      "loss": 1.5756,
+      "step": 7307
+    },
+    {
+      "epoch": 6.746365105008078,
+      "grad_norm": 2.884765625,
+      "learning_rate": 5.131420947298199e-05,
+      "loss": 1.1991,
+      "step": 7308
+    },
+    {
+      "epoch": 6.747288252942534,
+      "grad_norm": 3.865234375,
+      "learning_rate": 5.130753835890594e-05,
+      "loss": 1.4219,
+      "step": 7309
+    },
+    {
+      "epoch": 6.748211400876991,
+      "grad_norm": 3.576171875,
+      "learning_rate": 5.1300867244829884e-05,
+      "loss": 1.4248,
+      "step": 7310
+    },
+    {
+      "epoch": 6.749134548811447,
+      "grad_norm": 4.23046875,
+      "learning_rate": 5.129419613075384e-05,
+      "loss": 1.4656,
+      "step": 7311
+    },
+    {
+      "epoch": 6.750057696745904,
+      "grad_norm": 3.2890625,
+      "learning_rate": 5.128752501667778e-05,
+      "loss": 1.3623,
+      "step": 7312
+    },
+    {
+      "epoch": 6.75098084468036,
+      "grad_norm": 3.52734375,
+      "learning_rate": 5.1280853902601734e-05,
+      "loss": 1.4363,
+      "step": 7313
+    },
+    {
+      "epoch": 6.7519039926148166,
+      "grad_norm": 4.1328125,
+      "learning_rate": 5.127418278852568e-05,
+      "loss": 1.6736,
+      "step": 7314
+    },
+    {
+      "epoch": 6.7528271405492735,
+      "grad_norm": 2.98828125,
+      "learning_rate": 5.126751167444963e-05,
+      "loss": 1.3938,
+      "step": 7315
+    },
+    {
+      "epoch": 6.7537502884837295,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.126084056037358e-05,
+      "loss": 1.4568,
+      "step": 7316
+    },
+    {
+      "epoch": 6.7546734364181855,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.1254169446297536e-05,
+      "loss": 1.4285,
+      "step": 7317
+    },
+    {
+      "epoch": 6.755596584352642,
+      "grad_norm": 3.54296875,
+      "learning_rate": 5.124749833222149e-05,
+      "loss": 1.4064,
+      "step": 7318
+    },
+    {
+      "epoch": 6.756519732287099,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.1240827218145434e-05,
+      "loss": 1.0732,
+      "step": 7319
+    },
+    {
+      "epoch": 6.757442880221555,
+      "grad_norm": 3.9453125,
+      "learning_rate": 5.1234156104069386e-05,
+      "loss": 1.3359,
+      "step": 7320
+    },
+    {
+      "epoch": 6.758366028156012,
+      "grad_norm": 3.63671875,
+      "learning_rate": 5.122748498999334e-05,
+      "loss": 1.4226,
+      "step": 7321
+    },
+    {
+      "epoch": 6.759289176090468,
+      "grad_norm": 4.19140625,
+      "learning_rate": 5.1220813875917284e-05,
+      "loss": 1.1025,
+      "step": 7322
+    },
+    {
+      "epoch": 6.760212324024925,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.1214142761841236e-05,
+      "loss": 1.2441,
+      "step": 7323
+    },
+    {
+      "epoch": 6.761135471959381,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.120747164776518e-05,
+      "loss": 1.1299,
+      "step": 7324
+    },
+    {
+      "epoch": 6.762058619893838,
+      "grad_norm": 3.099609375,
+      "learning_rate": 5.1200800533689134e-05,
+      "loss": 1.2798,
+      "step": 7325
+    },
+    {
+      "epoch": 6.762981767828294,
+      "grad_norm": 3.150390625,
+      "learning_rate": 5.119412941961308e-05,
+      "loss": 1.2822,
+      "step": 7326
+    },
+    {
+      "epoch": 6.763904915762751,
+      "grad_norm": 3.607421875,
+      "learning_rate": 5.118745830553703e-05,
+      "loss": 1.5986,
+      "step": 7327
+    },
+    {
+      "epoch": 6.764828063697207,
+      "grad_norm": 3.0703125,
+      "learning_rate": 5.118078719146098e-05,
+      "loss": 1.6294,
+      "step": 7328
+    },
+    {
+      "epoch": 6.765751211631664,
+      "grad_norm": 3.86328125,
+      "learning_rate": 5.117411607738493e-05,
+      "loss": 1.7,
+      "step": 7329
+    },
+    {
+      "epoch": 6.766674359566121,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.1167444963308874e-05,
+      "loss": 1.1494,
+      "step": 7330
+    },
+    {
+      "epoch": 6.767597507500577,
+      "grad_norm": 3.314453125,
+      "learning_rate": 5.116077384923283e-05,
+      "loss": 1.3525,
+      "step": 7331
+    },
+    {
+      "epoch": 6.768520655435033,
+      "grad_norm": 3.603515625,
+      "learning_rate": 5.115410273515677e-05,
+      "loss": 1.1042,
+      "step": 7332
+    },
+    {
+      "epoch": 6.76944380336949,
+      "grad_norm": 3.41796875,
+      "learning_rate": 5.1147431621080724e-05,
+      "loss": 1.46,
+      "step": 7333
+    },
+    {
+      "epoch": 6.770366951303947,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.114076050700467e-05,
+      "loss": 1.4282,
+      "step": 7334
+    },
+    {
+      "epoch": 6.771290099238403,
+      "grad_norm": 3.240234375,
+      "learning_rate": 5.113408939292862e-05,
+      "loss": 1.5146,
+      "step": 7335
+    },
+    {
+      "epoch": 6.77221324717286,
+      "grad_norm": 3.97265625,
+      "learning_rate": 5.112741827885257e-05,
+      "loss": 1.5806,
+      "step": 7336
+    },
+    {
+      "epoch": 6.773136395107316,
+      "grad_norm": 3.302734375,
+      "learning_rate": 5.112074716477652e-05,
+      "loss": 0.958,
+      "step": 7337
+    },
+    {
+      "epoch": 6.774059543041773,
+      "grad_norm": 3.7578125,
+      "learning_rate": 5.1114076050700465e-05,
+      "loss": 1.3824,
+      "step": 7338
+    },
+    {
+      "epoch": 6.774982690976229,
+      "grad_norm": 3.45703125,
+      "learning_rate": 5.110740493662442e-05,
+      "loss": 1.2532,
+      "step": 7339
+    },
+    {
+      "epoch": 6.775905838910686,
+      "grad_norm": 3.4296875,
+      "learning_rate": 5.110073382254836e-05,
+      "loss": 1.3113,
+      "step": 7340
+    },
+    {
+      "epoch": 6.776828986845142,
+      "grad_norm": 3.94921875,
+      "learning_rate": 5.1094062708472315e-05,
+      "loss": 1.2927,
+      "step": 7341
+    },
+    {
+      "epoch": 6.777752134779599,
+      "grad_norm": 4.3515625,
+      "learning_rate": 5.108739159439626e-05,
+      "loss": 1.3596,
+      "step": 7342
+    },
+    {
+      "epoch": 6.778675282714055,
+      "grad_norm": 3.310546875,
+      "learning_rate": 5.108072048032022e-05,
+      "loss": 1.2042,
+      "step": 7343
+    },
+    {
+      "epoch": 6.7795984306485115,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.107404936624417e-05,
+      "loss": 1.5522,
+      "step": 7344
+    },
+    {
+      "epoch": 6.780521578582968,
+      "grad_norm": 3.861328125,
+      "learning_rate": 5.106737825216812e-05,
+      "loss": 1.5039,
+      "step": 7345
+    },
+    {
+      "epoch": 6.7814447265174245,
+      "grad_norm": 3.689453125,
+      "learning_rate": 5.106070713809207e-05,
+      "loss": 1.4249,
+      "step": 7346
+    },
+    {
+      "epoch": 6.7823678744518805,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.1054036024016015e-05,
+      "loss": 1.2578,
+      "step": 7347
+    },
+    {
+      "epoch": 6.783291022386337,
+      "grad_norm": 3.17578125,
+      "learning_rate": 5.104736490993997e-05,
+      "loss": 1.3503,
+      "step": 7348
+    },
+    {
+      "epoch": 6.784214170320794,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.104069379586391e-05,
+      "loss": 1.3879,
+      "step": 7349
+    },
+    {
+      "epoch": 6.78513731825525,
+      "grad_norm": 3.75390625,
+      "learning_rate": 5.1034022681787864e-05,
+      "loss": 1.1265,
+      "step": 7350
+    },
+    {
+      "epoch": 6.786060466189707,
+      "grad_norm": 2.83203125,
+      "learning_rate": 5.102735156771181e-05,
+      "loss": 0.9399,
+      "step": 7351
+    },
+    {
+      "epoch": 6.786983614124163,
+      "grad_norm": 3.169921875,
+      "learning_rate": 5.102068045363576e-05,
+      "loss": 1.2705,
+      "step": 7352
+    },
+    {
+      "epoch": 6.78790676205862,
+      "grad_norm": 3.1171875,
+      "learning_rate": 5.101400933955971e-05,
+      "loss": 1.3962,
+      "step": 7353
+    },
+    {
+      "epoch": 6.788829909993076,
+      "grad_norm": 4.37890625,
+      "learning_rate": 5.100733822548366e-05,
+      "loss": 1.3452,
+      "step": 7354
+    },
+    {
+      "epoch": 6.789753057927533,
+      "grad_norm": 3.04296875,
+      "learning_rate": 5.1000667111407605e-05,
+      "loss": 1.1477,
+      "step": 7355
+    },
+    {
+      "epoch": 6.790676205861989,
+      "grad_norm": 4.0390625,
+      "learning_rate": 5.099399599733156e-05,
+      "loss": 1.2201,
+      "step": 7356
+    },
+    {
+      "epoch": 6.791599353796446,
+      "grad_norm": 3.537109375,
+      "learning_rate": 5.09873248832555e-05,
+      "loss": 1.1033,
+      "step": 7357
+    },
+    {
+      "epoch": 6.792522501730902,
+      "grad_norm": 4.66015625,
+      "learning_rate": 5.0980653769179455e-05,
+      "loss": 1.3376,
+      "step": 7358
+    },
+    {
+      "epoch": 6.793445649665359,
+      "grad_norm": 3.748046875,
+      "learning_rate": 5.09739826551034e-05,
+      "loss": 1.0811,
+      "step": 7359
+    },
+    {
+      "epoch": 6.794368797599815,
+      "grad_norm": 3.166015625,
+      "learning_rate": 5.096731154102735e-05,
+      "loss": 1.2317,
+      "step": 7360
+    },
+    {
+      "epoch": 6.795291945534272,
+      "grad_norm": 3.201171875,
+      "learning_rate": 5.09606404269513e-05,
+      "loss": 1.3918,
+      "step": 7361
+    },
+    {
+      "epoch": 6.796215093468728,
+      "grad_norm": 3.47265625,
+      "learning_rate": 5.095396931287525e-05,
+      "loss": 1.3545,
+      "step": 7362
+    },
+    {
+      "epoch": 6.797138241403185,
+      "grad_norm": 3.48828125,
+      "learning_rate": 5.0947298198799196e-05,
+      "loss": 1.762,
+      "step": 7363
+    },
+    {
+      "epoch": 6.798061389337642,
+      "grad_norm": 3.76171875,
+      "learning_rate": 5.094062708472315e-05,
+      "loss": 1.3552,
+      "step": 7364
+    },
+    {
+      "epoch": 6.798984537272098,
+      "grad_norm": 3.279296875,
+      "learning_rate": 5.093395597064709e-05,
+      "loss": 1.6545,
+      "step": 7365
+    },
+    {
+      "epoch": 6.799907685206554,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.0927284856571045e-05,
+      "loss": 1.3866,
+      "step": 7366
+    },
+    {
+      "epoch": 6.800830833141011,
+      "grad_norm": 3.44140625,
+      "learning_rate": 5.092061374249499e-05,
+      "loss": 1.6077,
+      "step": 7367
+    },
+    {
+      "epoch": 6.801753981075468,
+      "grad_norm": 3.46484375,
+      "learning_rate": 5.091394262841896e-05,
+      "loss": 1.2383,
+      "step": 7368
+    },
+    {
+      "epoch": 6.802677129009924,
+      "grad_norm": 3.74609375,
+      "learning_rate": 5.09072715143429e-05,
+      "loss": 1.1665,
+      "step": 7369
+    },
+    {
+      "epoch": 6.803600276944381,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.0900600400266854e-05,
+      "loss": 1.5212,
+      "step": 7370
+    },
+    {
+      "epoch": 6.804523424878837,
+      "grad_norm": 4.27734375,
+      "learning_rate": 5.08939292861908e-05,
+      "loss": 1.0491,
+      "step": 7371
+    },
+    {
+      "epoch": 6.805446572813294,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.088725817211475e-05,
+      "loss": 1.1401,
+      "step": 7372
+    },
+    {
+      "epoch": 6.80636972074775,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.08805870580387e-05,
+      "loss": 1.3694,
+      "step": 7373
+    },
+    {
+      "epoch": 6.8072928686822065,
+      "grad_norm": 4.23828125,
+      "learning_rate": 5.087391594396265e-05,
+      "loss": 1.3535,
+      "step": 7374
+    },
+    {
+      "epoch": 6.8082160166166625,
+      "grad_norm": 3.44921875,
+      "learning_rate": 5.0867244829886595e-05,
+      "loss": 1.6023,
+      "step": 7375
+    },
+    {
+      "epoch": 6.8091391645511194,
+      "grad_norm": 4.19921875,
+      "learning_rate": 5.086057371581055e-05,
+      "loss": 1.5837,
+      "step": 7376
+    },
+    {
+      "epoch": 6.8100623124855755,
+      "grad_norm": 3.3671875,
+      "learning_rate": 5.085390260173449e-05,
+      "loss": 1.436,
+      "step": 7377
+    },
+    {
+      "epoch": 6.810985460420032,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.0847231487658445e-05,
+      "loss": 1.4165,
+      "step": 7378
+    },
+    {
+      "epoch": 6.811908608354489,
+      "grad_norm": 3.8828125,
+      "learning_rate": 5.084056037358239e-05,
+      "loss": 1.5581,
+      "step": 7379
+    },
+    {
+      "epoch": 6.812831756288945,
+      "grad_norm": 3.974609375,
+      "learning_rate": 5.083388925950634e-05,
+      "loss": 1.5383,
+      "step": 7380
+    },
+    {
+      "epoch": 6.813754904223401,
+      "grad_norm": 3.318359375,
+      "learning_rate": 5.082721814543029e-05,
+      "loss": 1.3004,
+      "step": 7381
+    },
+    {
+      "epoch": 6.814678052157858,
+      "grad_norm": 4.55859375,
+      "learning_rate": 5.082054703135424e-05,
+      "loss": 1.4861,
+      "step": 7382
+    },
+    {
+      "epoch": 6.815601200092315,
+      "grad_norm": 3.669921875,
+      "learning_rate": 5.0813875917278186e-05,
+      "loss": 1.48,
+      "step": 7383
+    },
+    {
+      "epoch": 6.816524348026771,
+      "grad_norm": 3.34765625,
+      "learning_rate": 5.080720480320214e-05,
+      "loss": 1.2526,
+      "step": 7384
+    },
+    {
+      "epoch": 6.817447495961228,
+      "grad_norm": 3.861328125,
+      "learning_rate": 5.080053368912608e-05,
+      "loss": 1.5061,
+      "step": 7385
+    },
+    {
+      "epoch": 6.818370643895684,
+      "grad_norm": 3.46484375,
+      "learning_rate": 5.0793862575050035e-05,
+      "loss": 1.1921,
+      "step": 7386
+    },
+    {
+      "epoch": 6.819293791830141,
+      "grad_norm": 3.6796875,
+      "learning_rate": 5.078719146097398e-05,
+      "loss": 1.5569,
+      "step": 7387
+    },
+    {
+      "epoch": 6.820216939764597,
+      "grad_norm": 3.142578125,
+      "learning_rate": 5.078052034689793e-05,
+      "loss": 1.3083,
+      "step": 7388
+    },
+    {
+      "epoch": 6.821140087699054,
+      "grad_norm": 3.69921875,
+      "learning_rate": 5.077384923282188e-05,
+      "loss": 1.142,
+      "step": 7389
+    },
+    {
+      "epoch": 6.82206323563351,
+      "grad_norm": 3.59765625,
+      "learning_rate": 5.076717811874583e-05,
+      "loss": 1.2791,
+      "step": 7390
+    },
+    {
+      "epoch": 6.822986383567967,
+      "grad_norm": 3.6015625,
+      "learning_rate": 5.0760507004669776e-05,
+      "loss": 1.5142,
+      "step": 7391
+    },
+    {
+      "epoch": 6.823909531502423,
+      "grad_norm": 2.904296875,
+      "learning_rate": 5.075383589059373e-05,
+      "loss": 1.1596,
+      "step": 7392
+    },
+    {
+      "epoch": 6.82483267943688,
+      "grad_norm": 4.046875,
+      "learning_rate": 5.074716477651769e-05,
+      "loss": 1.3933,
+      "step": 7393
+    },
+    {
+      "epoch": 6.825755827371336,
+      "grad_norm": 4.7109375,
+      "learning_rate": 5.074049366244163e-05,
+      "loss": 1.4951,
+      "step": 7394
+    },
+    {
+      "epoch": 6.826678975305793,
+      "grad_norm": 3.3046875,
+      "learning_rate": 5.0733822548365585e-05,
+      "loss": 0.9647,
+      "step": 7395
+    },
+    {
+      "epoch": 6.827602123240249,
+      "grad_norm": 3.447265625,
+      "learning_rate": 5.072715143428953e-05,
+      "loss": 1.6182,
+      "step": 7396
+    },
+    {
+      "epoch": 6.828525271174706,
+      "grad_norm": 4.4375,
+      "learning_rate": 5.072048032021348e-05,
+      "loss": 1.3601,
+      "step": 7397
+    },
+    {
+      "epoch": 6.829448419109163,
+      "grad_norm": 3.388671875,
+      "learning_rate": 5.071380920613743e-05,
+      "loss": 1.2256,
+      "step": 7398
+    },
+    {
+      "epoch": 6.830371567043619,
+      "grad_norm": 4.1640625,
+      "learning_rate": 5.070713809206138e-05,
+      "loss": 1.4962,
+      "step": 7399
+    },
+    {
+      "epoch": 6.831294714978076,
+      "grad_norm": 3.7734375,
+      "learning_rate": 5.0700466977985326e-05,
+      "loss": 1.2822,
+      "step": 7400
+    },
+    {
+      "epoch": 6.832217862912532,
+      "grad_norm": 3.58984375,
+      "learning_rate": 5.069379586390928e-05,
+      "loss": 1.4048,
+      "step": 7401
+    },
+    {
+      "epoch": 6.8331410108469886,
+      "grad_norm": 3.916015625,
+      "learning_rate": 5.068712474983322e-05,
+      "loss": 1.4824,
+      "step": 7402
+    },
+    {
+      "epoch": 6.834064158781445,
+      "grad_norm": 3.26171875,
+      "learning_rate": 5.0680453635757175e-05,
+      "loss": 1.3633,
+      "step": 7403
+    },
+    {
+      "epoch": 6.8349873067159015,
+      "grad_norm": 3.58203125,
+      "learning_rate": 5.067378252168112e-05,
+      "loss": 1.3479,
+      "step": 7404
+    },
+    {
+      "epoch": 6.8359104546503575,
+      "grad_norm": 4.52734375,
+      "learning_rate": 5.066711140760507e-05,
+      "loss": 1.3094,
+      "step": 7405
+    },
+    {
+      "epoch": 6.836833602584814,
+      "grad_norm": 3.646484375,
+      "learning_rate": 5.066044029352902e-05,
+      "loss": 0.9625,
+      "step": 7406
+    },
+    {
+      "epoch": 6.8377567505192705,
+      "grad_norm": 3.05859375,
+      "learning_rate": 5.065376917945297e-05,
+      "loss": 1.3843,
+      "step": 7407
+    },
+    {
+      "epoch": 6.838679898453727,
+      "grad_norm": 4.203125,
+      "learning_rate": 5.0647098065376916e-05,
+      "loss": 1.3335,
+      "step": 7408
+    },
+    {
+      "epoch": 6.839603046388183,
+      "grad_norm": 3.935546875,
+      "learning_rate": 5.064042695130087e-05,
+      "loss": 1.2032,
+      "step": 7409
+    },
+    {
+      "epoch": 6.84052619432264,
+      "grad_norm": 4.16796875,
+      "learning_rate": 5.0633755837224814e-05,
+      "loss": 1.4468,
+      "step": 7410
+    },
+    {
+      "epoch": 6.841449342257096,
+      "grad_norm": 3.435546875,
+      "learning_rate": 5.0627084723148766e-05,
+      "loss": 1.1589,
+      "step": 7411
+    },
+    {
+      "epoch": 6.842372490191553,
+      "grad_norm": 3.365234375,
+      "learning_rate": 5.062041360907271e-05,
+      "loss": 1.2356,
+      "step": 7412
+    },
+    {
+      "epoch": 6.84329563812601,
+      "grad_norm": 3.60546875,
+      "learning_rate": 5.0613742494996664e-05,
+      "loss": 1.4636,
+      "step": 7413
+    },
+    {
+      "epoch": 6.844218786060466,
+      "grad_norm": 3.20703125,
+      "learning_rate": 5.060707138092061e-05,
+      "loss": 1.3707,
+      "step": 7414
+    },
+    {
+      "epoch": 6.845141933994922,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.060040026684456e-05,
+      "loss": 1.5464,
+      "step": 7415
+    },
+    {
+      "epoch": 6.846065081929379,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.059372915276851e-05,
+      "loss": 1.4675,
+      "step": 7416
+    },
+    {
+      "epoch": 6.846988229863836,
+      "grad_norm": 3.623046875,
+      "learning_rate": 5.058705803869246e-05,
+      "loss": 1.4419,
+      "step": 7417
+    },
+    {
+      "epoch": 6.847911377798292,
+      "grad_norm": 3.599609375,
+      "learning_rate": 5.058038692461642e-05,
+      "loss": 1.5408,
+      "step": 7418
+    },
+    {
+      "epoch": 6.848834525732749,
+      "grad_norm": 3.79296875,
+      "learning_rate": 5.057371581054037e-05,
+      "loss": 1.1428,
+      "step": 7419
+    },
+    {
+      "epoch": 6.849757673667205,
+      "grad_norm": 3.1015625,
+      "learning_rate": 5.0567044696464316e-05,
+      "loss": 1.5852,
+      "step": 7420
+    },
+    {
+      "epoch": 6.850680821601662,
+      "grad_norm": 3.48828125,
+      "learning_rate": 5.056037358238827e-05,
+      "loss": 1.3723,
+      "step": 7421
+    },
+    {
+      "epoch": 6.851603969536118,
+      "grad_norm": 4.76953125,
+      "learning_rate": 5.055370246831221e-05,
+      "loss": 1.5264,
+      "step": 7422
+    },
+    {
+      "epoch": 6.852527117470575,
+      "grad_norm": 3.134765625,
+      "learning_rate": 5.0547031354236165e-05,
+      "loss": 1.5964,
+      "step": 7423
+    },
+    {
+      "epoch": 6.853450265405031,
+      "grad_norm": 3.205078125,
+      "learning_rate": 5.054036024016011e-05,
+      "loss": 1.2625,
+      "step": 7424
+    },
+    {
+      "epoch": 6.854373413339488,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.053368912608406e-05,
+      "loss": 1.3162,
+      "step": 7425
+    },
+    {
+      "epoch": 6.855296561273944,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.052701801200801e-05,
+      "loss": 1.3911,
+      "step": 7426
+    },
+    {
+      "epoch": 6.856219709208401,
+      "grad_norm": 4.1171875,
+      "learning_rate": 5.052034689793196e-05,
+      "loss": 1.1833,
+      "step": 7427
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 3.451171875,
+      "learning_rate": 5.0513675783855906e-05,
+      "loss": 1.4509,
+      "step": 7428
+    },
+    {
+      "epoch": 6.858066005077314,
+      "grad_norm": 3.123046875,
+      "learning_rate": 5.050700466977986e-05,
+      "loss": 1.5115,
+      "step": 7429
+    },
+    {
+      "epoch": 6.85898915301177,
+      "grad_norm": 3.435546875,
+      "learning_rate": 5.0500333555703804e-05,
+      "loss": 1.5933,
+      "step": 7430
+    },
+    {
+      "epoch": 6.859912300946227,
+      "grad_norm": 3.40234375,
+      "learning_rate": 5.0493662441627756e-05,
+      "loss": 0.8779,
+      "step": 7431
+    },
+    {
+      "epoch": 6.8608354488806835,
+      "grad_norm": 3.330078125,
+      "learning_rate": 5.04869913275517e-05,
+      "loss": 1.5088,
+      "step": 7432
+    },
+    {
+      "epoch": 6.8617585968151396,
+      "grad_norm": 3.00390625,
+      "learning_rate": 5.0480320213475654e-05,
+      "loss": 1.4331,
+      "step": 7433
+    },
+    {
+      "epoch": 6.8626817447495965,
+      "grad_norm": 3.64453125,
+      "learning_rate": 5.04736490993996e-05,
+      "loss": 1.4233,
+      "step": 7434
+    },
+    {
+      "epoch": 6.8636048926840525,
+      "grad_norm": 3.158203125,
+      "learning_rate": 5.046697798532355e-05,
+      "loss": 1.3384,
+      "step": 7435
+    },
+    {
+      "epoch": 6.864528040618509,
+      "grad_norm": 3.384765625,
+      "learning_rate": 5.04603068712475e-05,
+      "loss": 1.3058,
+      "step": 7436
+    },
+    {
+      "epoch": 6.865451188552965,
+      "grad_norm": 3.287109375,
+      "learning_rate": 5.045363575717145e-05,
+      "loss": 1.2819,
+      "step": 7437
+    },
+    {
+      "epoch": 6.866374336487422,
+      "grad_norm": 3.685546875,
+      "learning_rate": 5.0446964643095394e-05,
+      "loss": 1.1512,
+      "step": 7438
+    },
+    {
+      "epoch": 6.867297484421878,
+      "grad_norm": 3.544921875,
+      "learning_rate": 5.0440293529019346e-05,
+      "loss": 1.3044,
+      "step": 7439
+    },
+    {
+      "epoch": 6.868220632356335,
+      "grad_norm": 3.6953125,
+      "learning_rate": 5.043362241494329e-05,
+      "loss": 1.375,
+      "step": 7440
+    },
+    {
+      "epoch": 6.869143780290791,
+      "grad_norm": 3.361328125,
+      "learning_rate": 5.0426951300867244e-05,
+      "loss": 1.4172,
+      "step": 7441
+    },
+    {
+      "epoch": 6.870066928225248,
+      "grad_norm": 3.76171875,
+      "learning_rate": 5.042028018679119e-05,
+      "loss": 1.1353,
+      "step": 7442
+    },
+    {
+      "epoch": 6.870990076159704,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.041360907271515e-05,
+      "loss": 1.4458,
+      "step": 7443
+    },
+    {
+      "epoch": 6.871913224094161,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.04069379586391e-05,
+      "loss": 1.1887,
+      "step": 7444
+    },
+    {
+      "epoch": 6.872836372028617,
+      "grad_norm": 3.51171875,
+      "learning_rate": 5.0400266844563046e-05,
+      "loss": 1.4258,
+      "step": 7445
+    },
+    {
+      "epoch": 6.873759519963074,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.0393595730487e-05,
+      "loss": 1.4041,
+      "step": 7446
+    },
+    {
+      "epoch": 6.874682667897531,
+      "grad_norm": 3.677734375,
+      "learning_rate": 5.0386924616410944e-05,
+      "loss": 1.3357,
+      "step": 7447
+    },
+    {
+      "epoch": 6.875605815831987,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.0380253502334896e-05,
+      "loss": 1.1896,
+      "step": 7448
+    },
+    {
+      "epoch": 6.876528963766444,
+      "grad_norm": 3.201171875,
+      "learning_rate": 5.037358238825884e-05,
+      "loss": 1.2748,
+      "step": 7449
+    },
+    {
+      "epoch": 6.8774521117009,
+      "grad_norm": 3.181640625,
+      "learning_rate": 5.0366911274182794e-05,
+      "loss": 1.5308,
+      "step": 7450
+    },
+    {
+      "epoch": 6.878375259635357,
+      "grad_norm": 3.400390625,
+      "learning_rate": 5.036024016010674e-05,
+      "loss": 1.3811,
+      "step": 7451
+    },
+    {
+      "epoch": 6.879298407569813,
+      "grad_norm": 3.623046875,
+      "learning_rate": 5.035356904603069e-05,
+      "loss": 1.532,
+      "step": 7452
+    },
+    {
+      "epoch": 6.88022155550427,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.034689793195464e-05,
+      "loss": 1.219,
+      "step": 7453
+    },
+    {
+      "epoch": 6.881144703438726,
+      "grad_norm": 3.02734375,
+      "learning_rate": 5.034022681787859e-05,
+      "loss": 1.3999,
+      "step": 7454
+    },
+    {
+      "epoch": 6.882067851373183,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.0333555703802534e-05,
+      "loss": 1.4243,
+      "step": 7455
+    },
+    {
+      "epoch": 6.882990999307639,
+      "grad_norm": 3.75,
+      "learning_rate": 5.0326884589726487e-05,
+      "loss": 1.1208,
+      "step": 7456
+    },
+    {
+      "epoch": 6.883914147242096,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.032021347565043e-05,
+      "loss": 1.3062,
+      "step": 7457
+    },
+    {
+      "epoch": 6.884837295176552,
+      "grad_norm": 3.8359375,
+      "learning_rate": 5.0313542361574384e-05,
+      "loss": 1.375,
+      "step": 7458
+    },
+    {
+      "epoch": 6.885760443111009,
+      "grad_norm": 3.5,
+      "learning_rate": 5.030687124749833e-05,
+      "loss": 1.1168,
+      "step": 7459
+    },
+    {
+      "epoch": 6.886683591045465,
+      "grad_norm": 4.46875,
+      "learning_rate": 5.030020013342228e-05,
+      "loss": 1.5613,
+      "step": 7460
+    },
+    {
+      "epoch": 6.887606738979922,
+      "grad_norm": 3.751953125,
+      "learning_rate": 5.029352901934623e-05,
+      "loss": 1.4475,
+      "step": 7461
+    },
+    {
+      "epoch": 6.8885298869143785,
+      "grad_norm": 3.3203125,
+      "learning_rate": 5.028685790527018e-05,
+      "loss": 1.6245,
+      "step": 7462
+    },
+    {
+      "epoch": 6.8894530348488345,
+      "grad_norm": 3.931640625,
+      "learning_rate": 5.0280186791194125e-05,
+      "loss": 1.2216,
+      "step": 7463
+    },
+    {
+      "epoch": 6.890376182783291,
+      "grad_norm": 3.470703125,
+      "learning_rate": 5.027351567711808e-05,
+      "loss": 1.5686,
+      "step": 7464
+    },
+    {
+      "epoch": 6.8912993307177475,
+      "grad_norm": 3.25390625,
+      "learning_rate": 5.026684456304203e-05,
+      "loss": 1.46,
+      "step": 7465
+    },
+    {
+      "epoch": 6.892222478652204,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.0260173448965975e-05,
+      "loss": 1.4238,
+      "step": 7466
+    },
+    {
+      "epoch": 6.89314562658666,
+      "grad_norm": 3.95703125,
+      "learning_rate": 5.025350233488993e-05,
+      "loss": 1.6333,
+      "step": 7467
+    },
+    {
+      "epoch": 6.894068774521117,
+      "grad_norm": 3.3984375,
+      "learning_rate": 5.0246831220813886e-05,
+      "loss": 1.332,
+      "step": 7468
+    },
+    {
+      "epoch": 6.894991922455573,
+      "grad_norm": 3.60546875,
+      "learning_rate": 5.024016010673783e-05,
+      "loss": 1.2451,
+      "step": 7469
+    },
+    {
+      "epoch": 6.89591507039003,
+      "grad_norm": 3.724609375,
+      "learning_rate": 5.0233488992661784e-05,
+      "loss": 1.5125,
+      "step": 7470
+    },
+    {
+      "epoch": 6.896838218324486,
+      "grad_norm": 4.07421875,
+      "learning_rate": 5.022681787858573e-05,
+      "loss": 1.4951,
+      "step": 7471
+    },
+    {
+      "epoch": 6.897761366258943,
+      "grad_norm": 3.025390625,
+      "learning_rate": 5.022014676450968e-05,
+      "loss": 1.4021,
+      "step": 7472
+    },
+    {
+      "epoch": 6.898684514193399,
+      "grad_norm": 3.79296875,
+      "learning_rate": 5.021347565043363e-05,
+      "loss": 1.318,
+      "step": 7473
+    },
+    {
+      "epoch": 6.899607662127856,
+      "grad_norm": 3.658203125,
+      "learning_rate": 5.020680453635758e-05,
+      "loss": 1.5046,
+      "step": 7474
+    },
+    {
+      "epoch": 6.900530810062312,
+      "grad_norm": 3.607421875,
+      "learning_rate": 5.0200133422281524e-05,
+      "loss": 1.4509,
+      "step": 7475
+    },
+    {
+      "epoch": 6.901453957996769,
+      "grad_norm": 3.341796875,
+      "learning_rate": 5.0193462308205476e-05,
+      "loss": 0.9574,
+      "step": 7476
+    },
+    {
+      "epoch": 6.902377105931225,
+      "grad_norm": 3.419921875,
+      "learning_rate": 5.018679119412942e-05,
+      "loss": 1.3186,
+      "step": 7477
+    },
+    {
+      "epoch": 6.903300253865682,
+      "grad_norm": 3.25,
+      "learning_rate": 5.0180120080053374e-05,
+      "loss": 0.779,
+      "step": 7478
+    },
+    {
+      "epoch": 6.904223401800138,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.017344896597732e-05,
+      "loss": 1.0942,
+      "step": 7479
+    },
+    {
+      "epoch": 6.905146549734595,
+      "grad_norm": 3.42578125,
+      "learning_rate": 5.016677785190127e-05,
+      "loss": 1.3987,
+      "step": 7480
+    },
+    {
+      "epoch": 6.906069697669052,
+      "grad_norm": 3.5859375,
+      "learning_rate": 5.016010673782522e-05,
+      "loss": 1.5884,
+      "step": 7481
+    },
+    {
+      "epoch": 6.906992845603508,
+      "grad_norm": 3.05859375,
+      "learning_rate": 5.015343562374917e-05,
+      "loss": 1.4189,
+      "step": 7482
+    },
+    {
+      "epoch": 6.907915993537965,
+      "grad_norm": 3.96484375,
+      "learning_rate": 5.0146764509673115e-05,
+      "loss": 1.2585,
+      "step": 7483
+    },
+    {
+      "epoch": 6.908839141472421,
+      "grad_norm": 3.22265625,
+      "learning_rate": 5.014009339559707e-05,
+      "loss": 1.3331,
+      "step": 7484
+    },
+    {
+      "epoch": 6.909762289406878,
+      "grad_norm": 3.974609375,
+      "learning_rate": 5.013342228152101e-05,
+      "loss": 1.3203,
+      "step": 7485
+    },
+    {
+      "epoch": 6.910685437341334,
+      "grad_norm": 2.958984375,
+      "learning_rate": 5.0126751167444965e-05,
+      "loss": 1.3301,
+      "step": 7486
+    },
+    {
+      "epoch": 6.911608585275791,
+      "grad_norm": 3.966796875,
+      "learning_rate": 5.012008005336891e-05,
+      "loss": 1.3271,
+      "step": 7487
+    },
+    {
+      "epoch": 6.912531733210247,
+      "grad_norm": 3.1953125,
+      "learning_rate": 5.011340893929286e-05,
+      "loss": 1.4253,
+      "step": 7488
+    },
+    {
+      "epoch": 6.913454881144704,
+      "grad_norm": 3.7265625,
+      "learning_rate": 5.010673782521681e-05,
+      "loss": 1.2012,
+      "step": 7489
+    },
+    {
+      "epoch": 6.91437802907916,
+      "grad_norm": 3.0,
+      "learning_rate": 5.010006671114076e-05,
+      "loss": 0.9941,
+      "step": 7490
+    },
+    {
+      "epoch": 6.915301177013617,
+      "grad_norm": 4.10546875,
+      "learning_rate": 5.0093395597064705e-05,
+      "loss": 1.1433,
+      "step": 7491
+    },
+    {
+      "epoch": 6.916224324948073,
+      "grad_norm": 3.474609375,
+      "learning_rate": 5.008672448298866e-05,
+      "loss": 1.2698,
+      "step": 7492
+    },
+    {
+      "epoch": 6.9171474728825295,
+      "grad_norm": 3.16796875,
+      "learning_rate": 5.0080053368912617e-05,
+      "loss": 1.0828,
+      "step": 7493
+    },
+    {
+      "epoch": 6.9180706208169855,
+      "grad_norm": 3.62890625,
+      "learning_rate": 5.007338225483656e-05,
+      "loss": 1.4236,
+      "step": 7494
+    },
+    {
+      "epoch": 6.9189937687514425,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.0066711140760514e-05,
+      "loss": 1.2205,
+      "step": 7495
+    },
+    {
+      "epoch": 6.919916916685899,
+      "grad_norm": 4.296875,
+      "learning_rate": 5.006004002668446e-05,
+      "loss": 1.272,
+      "step": 7496
+    },
+    {
+      "epoch": 6.920840064620355,
+      "grad_norm": 3.130859375,
+      "learning_rate": 5.005336891260841e-05,
+      "loss": 1.2517,
+      "step": 7497
+    },
+    {
+      "epoch": 6.921763212554811,
+      "grad_norm": 3.71484375,
+      "learning_rate": 5.004669779853236e-05,
+      "loss": 1.5137,
+      "step": 7498
+    },
+    {
+      "epoch": 6.922686360489268,
+      "grad_norm": 3.126953125,
+      "learning_rate": 5.004002668445631e-05,
+      "loss": 1.3413,
+      "step": 7499
+    },
+    {
+      "epoch": 6.923609508423725,
+      "grad_norm": 3.998046875,
+      "learning_rate": 5.0033355570380255e-05,
+      "loss": 1.7144,
+      "step": 7500
+    },
+    {
+      "epoch": 6.924532656358181,
+      "grad_norm": 3.26171875,
+      "learning_rate": 5.002668445630421e-05,
+      "loss": 1.4072,
+      "step": 7501
+    },
+    {
+      "epoch": 6.925455804292638,
+      "grad_norm": 4.35546875,
+      "learning_rate": 5.002001334222815e-05,
+      "loss": 1.8015,
+      "step": 7502
+    },
+    {
+      "epoch": 6.926378952227094,
+      "grad_norm": 3.5703125,
+      "learning_rate": 5.0013342228152105e-05,
+      "loss": 1.4243,
+      "step": 7503
+    },
+    {
+      "epoch": 6.927302100161551,
+      "grad_norm": 3.384765625,
+      "learning_rate": 5.000667111407605e-05,
+      "loss": 1.3499,
+      "step": 7504
+    },
+    {
+      "epoch": 6.928225248096007,
+      "grad_norm": 3.294921875,
+      "learning_rate": 5e-05,
+      "loss": 1.6389,
+      "step": 7505
+    },
+    {
+      "epoch": 6.929148396030464,
+      "grad_norm": 3.138671875,
+      "learning_rate": 4.9993328885923955e-05,
+      "loss": 1.6074,
+      "step": 7506
+    },
+    {
+      "epoch": 6.93007154396492,
+      "grad_norm": 3.30078125,
+      "learning_rate": 4.99866577718479e-05,
+      "loss": 1.4893,
+      "step": 7507
+    },
+    {
+      "epoch": 6.930994691899377,
+      "grad_norm": 2.78515625,
+      "learning_rate": 4.997998665777185e-05,
+      "loss": 1.1433,
+      "step": 7508
+    },
+    {
+      "epoch": 6.931917839833833,
+      "grad_norm": 3.39453125,
+      "learning_rate": 4.99733155436958e-05,
+      "loss": 1.3721,
+      "step": 7509
+    },
+    {
+      "epoch": 6.93284098776829,
+      "grad_norm": 3.490234375,
+      "learning_rate": 4.996664442961975e-05,
+      "loss": 1.3389,
+      "step": 7510
+    },
+    {
+      "epoch": 6.933764135702747,
+      "grad_norm": 4.43359375,
+      "learning_rate": 4.9959973315543695e-05,
+      "loss": 1.3568,
+      "step": 7511
+    },
+    {
+      "epoch": 6.934687283637203,
+      "grad_norm": 3.240234375,
+      "learning_rate": 4.995330220146765e-05,
+      "loss": 1.3673,
+      "step": 7512
+    },
+    {
+      "epoch": 6.935610431571659,
+      "grad_norm": 3.560546875,
+      "learning_rate": 4.99466310873916e-05,
+      "loss": 1.697,
+      "step": 7513
+    },
+    {
+      "epoch": 6.936533579506116,
+      "grad_norm": 3.66015625,
+      "learning_rate": 4.9939959973315545e-05,
+      "loss": 1.4216,
+      "step": 7514
+    },
+    {
+      "epoch": 6.937456727440573,
+      "grad_norm": 3.529296875,
+      "learning_rate": 4.99332888592395e-05,
+      "loss": 1.1658,
+      "step": 7515
+    },
+    {
+      "epoch": 6.938379875375029,
+      "grad_norm": 3.228515625,
+      "learning_rate": 4.992661774516344e-05,
+      "loss": 1.3336,
+      "step": 7516
+    },
+    {
+      "epoch": 6.939303023309486,
+      "grad_norm": 4.015625,
+      "learning_rate": 4.9919946631087395e-05,
+      "loss": 1.269,
+      "step": 7517
+    },
+    {
+      "epoch": 6.940226171243942,
+      "grad_norm": 3.4453125,
+      "learning_rate": 4.991327551701134e-05,
+      "loss": 1.2031,
+      "step": 7518
+    },
+    {
+      "epoch": 6.941149319178399,
+      "grad_norm": 3.68359375,
+      "learning_rate": 4.990660440293529e-05,
+      "loss": 1.2985,
+      "step": 7519
+    },
+    {
+      "epoch": 6.942072467112855,
+      "grad_norm": 3.08984375,
+      "learning_rate": 4.989993328885924e-05,
+      "loss": 1.4192,
+      "step": 7520
+    },
+    {
+      "epoch": 6.9429956150473116,
+      "grad_norm": 3.400390625,
+      "learning_rate": 4.989326217478319e-05,
+      "loss": 1.3835,
+      "step": 7521
+    },
+    {
+      "epoch": 6.943918762981768,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9886591060707136e-05,
+      "loss": 1.3286,
+      "step": 7522
+    },
+    {
+      "epoch": 6.9448419109162245,
+      "grad_norm": 3.404296875,
+      "learning_rate": 4.987991994663109e-05,
+      "loss": 1.4968,
+      "step": 7523
+    },
+    {
+      "epoch": 6.9457650588506805,
+      "grad_norm": 3.37109375,
+      "learning_rate": 4.987324883255504e-05,
+      "loss": 1.3967,
+      "step": 7524
+    },
+    {
+      "epoch": 6.946688206785137,
+      "grad_norm": 3.955078125,
+      "learning_rate": 4.986657771847899e-05,
+      "loss": 1.2827,
+      "step": 7525
+    },
+    {
+      "epoch": 6.9476113547195935,
+      "grad_norm": 3.451171875,
+      "learning_rate": 4.985990660440294e-05,
+      "loss": 1.2678,
+      "step": 7526
+    },
+    {
+      "epoch": 6.94853450265405,
+      "grad_norm": 3.361328125,
+      "learning_rate": 4.985323549032689e-05,
+      "loss": 1.55,
+      "step": 7527
+    },
+    {
+      "epoch": 6.949457650588506,
+      "grad_norm": 3.1171875,
+      "learning_rate": 4.9846564376250835e-05,
+      "loss": 1.5457,
+      "step": 7528
+    },
+    {
+      "epoch": 6.950380798522963,
+      "grad_norm": 2.9921875,
+      "learning_rate": 4.983989326217479e-05,
+      "loss": 1.3628,
+      "step": 7529
+    },
+    {
+      "epoch": 6.95130394645742,
+      "grad_norm": 3.380859375,
+      "learning_rate": 4.983322214809873e-05,
+      "loss": 1.2281,
+      "step": 7530
+    },
+    {
+      "epoch": 6.952227094391876,
+      "grad_norm": 4.02734375,
+      "learning_rate": 4.9826551034022685e-05,
+      "loss": 1.6245,
+      "step": 7531
+    },
+    {
+      "epoch": 6.953150242326333,
+      "grad_norm": 3.72265625,
+      "learning_rate": 4.981987991994663e-05,
+      "loss": 1.3577,
+      "step": 7532
+    },
+    {
+      "epoch": 6.954073390260789,
+      "grad_norm": 3.162109375,
+      "learning_rate": 4.981320880587058e-05,
+      "loss": 0.9041,
+      "step": 7533
+    },
+    {
+      "epoch": 6.954996538195246,
+      "grad_norm": 4.05078125,
+      "learning_rate": 4.980653769179453e-05,
+      "loss": 1.3723,
+      "step": 7534
+    },
+    {
+      "epoch": 6.955919686129702,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.979986657771848e-05,
+      "loss": 1.2988,
+      "step": 7535
+    },
+    {
+      "epoch": 6.956842834064159,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9793195463642426e-05,
+      "loss": 1.2799,
+      "step": 7536
+    },
+    {
+      "epoch": 6.957765981998615,
+      "grad_norm": 3.302734375,
+      "learning_rate": 4.9786524349566385e-05,
+      "loss": 1.4128,
+      "step": 7537
+    },
+    {
+      "epoch": 6.958689129933072,
+      "grad_norm": 3.943359375,
+      "learning_rate": 4.977985323549033e-05,
+      "loss": 1.2676,
+      "step": 7538
+    },
+    {
+      "epoch": 6.959612277867528,
+      "grad_norm": 4.04296875,
+      "learning_rate": 4.977318212141428e-05,
+      "loss": 1.1971,
+      "step": 7539
+    },
+    {
+      "epoch": 6.960535425801985,
+      "grad_norm": 4.29296875,
+      "learning_rate": 4.976651100733823e-05,
+      "loss": 1.2108,
+      "step": 7540
+    },
+    {
+      "epoch": 6.961458573736441,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.975983989326218e-05,
+      "loss": 1.5608,
+      "step": 7541
+    },
+    {
+      "epoch": 6.962381721670898,
+      "grad_norm": 3.513671875,
+      "learning_rate": 4.9753168779186126e-05,
+      "loss": 1.3384,
+      "step": 7542
+    },
+    {
+      "epoch": 6.963304869605354,
+      "grad_norm": 3.75,
+      "learning_rate": 4.974649766511008e-05,
+      "loss": 1.3882,
+      "step": 7543
+    },
+    {
+      "epoch": 6.964228017539811,
+      "grad_norm": 3.22265625,
+      "learning_rate": 4.973982655103402e-05,
+      "loss": 1.4933,
+      "step": 7544
+    },
+    {
+      "epoch": 6.965151165474268,
+      "grad_norm": 3.529296875,
+      "learning_rate": 4.9733155436957975e-05,
+      "loss": 1.5349,
+      "step": 7545
+    },
+    {
+      "epoch": 6.966074313408724,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.972648432288192e-05,
+      "loss": 1.1868,
+      "step": 7546
+    },
+    {
+      "epoch": 6.96699746134318,
+      "grad_norm": 3.357421875,
+      "learning_rate": 4.971981320880587e-05,
+      "loss": 1.3325,
+      "step": 7547
+    },
+    {
+      "epoch": 6.967920609277637,
+      "grad_norm": 3.87890625,
+      "learning_rate": 4.971314209472982e-05,
+      "loss": 1.4556,
+      "step": 7548
+    },
+    {
+      "epoch": 6.968843757212094,
+      "grad_norm": 3.130859375,
+      "learning_rate": 4.970647098065378e-05,
+      "loss": 1.4575,
+      "step": 7549
+    },
+    {
+      "epoch": 6.96976690514655,
+      "grad_norm": 3.82421875,
+      "learning_rate": 4.969979986657772e-05,
+      "loss": 1.3435,
+      "step": 7550
+    },
+    {
+      "epoch": 6.9706900530810065,
+      "grad_norm": 3.404296875,
+      "learning_rate": 4.9693128752501675e-05,
+      "loss": 1.3016,
+      "step": 7551
+    },
+    {
+      "epoch": 6.971613201015463,
+      "grad_norm": 3.306640625,
+      "learning_rate": 4.968645763842562e-05,
+      "loss": 1.2209,
+      "step": 7552
+    },
+    {
+      "epoch": 6.9725363489499195,
+      "grad_norm": 4.19921875,
+      "learning_rate": 4.967978652434957e-05,
+      "loss": 1.3875,
+      "step": 7553
+    },
+    {
+      "epoch": 6.9734594968843755,
+      "grad_norm": 3.623046875,
+      "learning_rate": 4.967311541027352e-05,
+      "loss": 1.3177,
+      "step": 7554
+    },
+    {
+      "epoch": 6.974382644818832,
+      "grad_norm": 4.1015625,
+      "learning_rate": 4.966644429619747e-05,
+      "loss": 1.3818,
+      "step": 7555
+    },
+    {
+      "epoch": 6.975305792753288,
+      "grad_norm": 2.830078125,
+      "learning_rate": 4.9659773182121416e-05,
+      "loss": 1.2023,
+      "step": 7556
+    },
+    {
+      "epoch": 6.976228940687745,
+      "grad_norm": 3.931640625,
+      "learning_rate": 4.965310206804537e-05,
+      "loss": 1.6108,
+      "step": 7557
+    },
+    {
+      "epoch": 6.977152088622201,
+      "grad_norm": 3.0859375,
+      "learning_rate": 4.9646430953969314e-05,
+      "loss": 1.4753,
+      "step": 7558
+    },
+    {
+      "epoch": 6.978075236556658,
+      "grad_norm": 3.615234375,
+      "learning_rate": 4.9639759839893266e-05,
+      "loss": 1.354,
+      "step": 7559
+    },
+    {
+      "epoch": 6.978998384491114,
+      "grad_norm": 3.298828125,
+      "learning_rate": 4.963308872581721e-05,
+      "loss": 1.5159,
+      "step": 7560
+    },
+    {
+      "epoch": 6.979921532425571,
+      "grad_norm": 3.294921875,
+      "learning_rate": 4.962641761174116e-05,
+      "loss": 1.5898,
+      "step": 7561
+    },
+    {
+      "epoch": 6.980844680360027,
+      "grad_norm": 3.177734375,
+      "learning_rate": 4.9619746497665116e-05,
+      "loss": 1.4583,
+      "step": 7562
+    },
+    {
+      "epoch": 6.981767828294484,
+      "grad_norm": 3.478515625,
+      "learning_rate": 4.961307538358906e-05,
+      "loss": 1.3325,
+      "step": 7563
+    },
+    {
+      "epoch": 6.982690976228941,
+      "grad_norm": 3.63671875,
+      "learning_rate": 4.960640426951301e-05,
+      "loss": 1.5181,
+      "step": 7564
+    },
+    {
+      "epoch": 6.983614124163397,
+      "grad_norm": 3.99609375,
+      "learning_rate": 4.959973315543696e-05,
+      "loss": 1.5947,
+      "step": 7565
+    },
+    {
+      "epoch": 6.984537272097854,
+      "grad_norm": 3.57421875,
+      "learning_rate": 4.959306204136091e-05,
+      "loss": 1.3936,
+      "step": 7566
+    },
+    {
+      "epoch": 6.98546042003231,
+      "grad_norm": 3.416015625,
+      "learning_rate": 4.9586390927284856e-05,
+      "loss": 1.4802,
+      "step": 7567
+    },
+    {
+      "epoch": 6.986383567966767,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.957971981320881e-05,
+      "loss": 1.3076,
+      "step": 7568
+    },
+    {
+      "epoch": 6.987306715901223,
+      "grad_norm": 3.478515625,
+      "learning_rate": 4.9573048699132754e-05,
+      "loss": 1.2946,
+      "step": 7569
+    },
+    {
+      "epoch": 6.98822986383568,
+      "grad_norm": 3.193359375,
+      "learning_rate": 4.9566377585056706e-05,
+      "loss": 1.584,
+      "step": 7570
+    },
+    {
+      "epoch": 6.989153011770136,
+      "grad_norm": 4.1015625,
+      "learning_rate": 4.955970647098065e-05,
+      "loss": 1.2433,
+      "step": 7571
+    },
+    {
+      "epoch": 6.990076159704593,
+      "grad_norm": 3.8046875,
+      "learning_rate": 4.9553035356904604e-05,
+      "loss": 1.2969,
+      "step": 7572
+    },
+    {
+      "epoch": 6.990999307639049,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.954636424282855e-05,
+      "loss": 1.376,
+      "step": 7573
+    },
+    {
+      "epoch": 6.991922455573506,
+      "grad_norm": 3.072265625,
+      "learning_rate": 4.953969312875251e-05,
+      "loss": 1.2209,
+      "step": 7574
+    },
+    {
+      "epoch": 6.992845603507962,
+      "grad_norm": 3.146484375,
+      "learning_rate": 4.9533022014676454e-05,
+      "loss": 0.9789,
+      "step": 7575
+    },
+    {
+      "epoch": 6.993768751442419,
+      "grad_norm": 3.435546875,
+      "learning_rate": 4.9526350900600406e-05,
+      "loss": 1.137,
+      "step": 7576
+    },
+    {
+      "epoch": 6.994691899376875,
+      "grad_norm": 3.904296875,
+      "learning_rate": 4.951967978652435e-05,
+      "loss": 1.6416,
+      "step": 7577
+    },
+    {
+      "epoch": 6.995615047311332,
+      "grad_norm": 4.33203125,
+      "learning_rate": 4.9513008672448303e-05,
+      "loss": 1.6042,
+      "step": 7578
+    },
+    {
+      "epoch": 6.996538195245789,
+      "grad_norm": 3.771484375,
+      "learning_rate": 4.950633755837225e-05,
+      "loss": 1.5908,
+      "step": 7579
+    },
+    {
+      "epoch": 6.997461343180245,
+      "grad_norm": 3.28515625,
+      "learning_rate": 4.94996664442962e-05,
+      "loss": 1.4631,
+      "step": 7580
+    },
+    {
+      "epoch": 6.9983844911147015,
+      "grad_norm": 3.357421875,
+      "learning_rate": 4.9492995330220146e-05,
+      "loss": 1.3572,
+      "step": 7581
+    },
+    {
+      "epoch": 6.9993076390491575,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.94863242161441e-05,
+      "loss": 1.4084,
+      "step": 7582
+    },
+    {
+      "epoch": 7.0002307869836145,
+      "grad_norm": 3.423828125,
+      "learning_rate": 4.9479653102068044e-05,
+      "loss": 1.0191,
+      "step": 7583
+    },
+    {
+      "epoch": 7.0011539349180705,
+      "grad_norm": 3.603515625,
+      "learning_rate": 4.9472981987991996e-05,
+      "loss": 1.3113,
+      "step": 7584
+    },
+    {
+      "epoch": 7.002077082852527,
+      "grad_norm": 3.587890625,
+      "learning_rate": 4.946631087391594e-05,
+      "loss": 1.2993,
+      "step": 7585
+    },
+    {
+      "epoch": 7.003000230786983,
+      "grad_norm": 3.78515625,
+      "learning_rate": 4.9459639759839894e-05,
+      "loss": 1.356,
+      "step": 7586
+    },
+    {
+      "epoch": 7.00392337872144,
+      "grad_norm": 3.970703125,
+      "learning_rate": 4.9452968645763846e-05,
+      "loss": 1.3601,
+      "step": 7587
+    },
+    {
+      "epoch": 7.004846526655896,
+      "grad_norm": 3.556640625,
+      "learning_rate": 4.94462975316878e-05,
+      "loss": 1.3499,
+      "step": 7588
+    },
+    {
+      "epoch": 7.005769674590353,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.9439626417611744e-05,
+      "loss": 1.1628,
+      "step": 7589
+    },
+    {
+      "epoch": 7.006692822524809,
+      "grad_norm": 3.291015625,
+      "learning_rate": 4.9432955303535696e-05,
+      "loss": 0.8774,
+      "step": 7590
+    },
+    {
+      "epoch": 7.007615970459266,
+      "grad_norm": 3.333984375,
+      "learning_rate": 4.942628418945964e-05,
+      "loss": 1.4336,
+      "step": 7591
+    },
+    {
+      "epoch": 7.008539118393722,
+      "grad_norm": 3.658203125,
+      "learning_rate": 4.9419613075383594e-05,
+      "loss": 1.0062,
+      "step": 7592
+    },
+    {
+      "epoch": 7.009462266328179,
+      "grad_norm": 3.408203125,
+      "learning_rate": 4.941294196130754e-05,
+      "loss": 1.4866,
+      "step": 7593
+    },
+    {
+      "epoch": 7.010385414262635,
+      "grad_norm": 3.763671875,
+      "learning_rate": 4.940627084723149e-05,
+      "loss": 1.3324,
+      "step": 7594
+    },
+    {
+      "epoch": 7.011308562197092,
+      "grad_norm": 3.310546875,
+      "learning_rate": 4.939959973315544e-05,
+      "loss": 1.4248,
+      "step": 7595
+    },
+    {
+      "epoch": 7.012231710131549,
+      "grad_norm": 3.529296875,
+      "learning_rate": 4.939292861907939e-05,
+      "loss": 1.1633,
+      "step": 7596
+    },
+    {
+      "epoch": 7.013154858066005,
+      "grad_norm": 3.5859375,
+      "learning_rate": 4.9386257505003334e-05,
+      "loss": 1.3038,
+      "step": 7597
+    },
+    {
+      "epoch": 7.014078006000462,
+      "grad_norm": 3.716796875,
+      "learning_rate": 4.9379586390927287e-05,
+      "loss": 1.2119,
+      "step": 7598
+    },
+    {
+      "epoch": 7.015001153934918,
+      "grad_norm": 3.322265625,
+      "learning_rate": 4.937291527685123e-05,
+      "loss": 1.0555,
+      "step": 7599
+    },
+    {
+      "epoch": 7.015924301869375,
+      "grad_norm": 3.681640625,
+      "learning_rate": 4.936624416277519e-05,
+      "loss": 1.3647,
+      "step": 7600
+    },
+    {
+      "epoch": 7.016847449803831,
+      "grad_norm": 3.69140625,
+      "learning_rate": 4.9359573048699136e-05,
+      "loss": 1.4109,
+      "step": 7601
+    },
+    {
+      "epoch": 7.017770597738288,
+      "grad_norm": 3.7109375,
+      "learning_rate": 4.935290193462309e-05,
+      "loss": 1.5034,
+      "step": 7602
+    },
+    {
+      "epoch": 7.018693745672744,
+      "grad_norm": 3.119140625,
+      "learning_rate": 4.9346230820547034e-05,
+      "loss": 1.2417,
+      "step": 7603
+    },
+    {
+      "epoch": 7.019616893607201,
+      "grad_norm": 4.640625,
+      "learning_rate": 4.9339559706470986e-05,
+      "loss": 1.2051,
+      "step": 7604
+    },
+    {
+      "epoch": 7.020540041541657,
+      "grad_norm": 2.833984375,
+      "learning_rate": 4.933288859239493e-05,
+      "loss": 1.2822,
+      "step": 7605
+    },
+    {
+      "epoch": 7.021463189476114,
+      "grad_norm": 3.951171875,
+      "learning_rate": 4.9326217478318884e-05,
+      "loss": 1.3789,
+      "step": 7606
+    },
+    {
+      "epoch": 7.02238633741057,
+      "grad_norm": 2.990234375,
+      "learning_rate": 4.931954636424283e-05,
+      "loss": 1.0805,
+      "step": 7607
+    },
+    {
+      "epoch": 7.023309485345027,
+      "grad_norm": 3.654296875,
+      "learning_rate": 4.931287525016678e-05,
+      "loss": 1.2946,
+      "step": 7608
+    },
+    {
+      "epoch": 7.024232633279483,
+      "grad_norm": 3.115234375,
+      "learning_rate": 4.930620413609073e-05,
+      "loss": 1.3093,
+      "step": 7609
+    },
+    {
+      "epoch": 7.02515578121394,
+      "grad_norm": 3.8671875,
+      "learning_rate": 4.929953302201468e-05,
+      "loss": 1.0112,
+      "step": 7610
+    },
+    {
+      "epoch": 7.026078929148396,
+      "grad_norm": 3.42578125,
+      "learning_rate": 4.9292861907938625e-05,
+      "loss": 1.6016,
+      "step": 7611
+    },
+    {
+      "epoch": 7.0270020770828525,
+      "grad_norm": 4.5,
+      "learning_rate": 4.928619079386258e-05,
+      "loss": 1.3818,
+      "step": 7612
+    },
+    {
+      "epoch": 7.027925225017309,
+      "grad_norm": 4.43359375,
+      "learning_rate": 4.927951967978653e-05,
+      "loss": 1.0424,
+      "step": 7613
+    },
+    {
+      "epoch": 7.0288483729517655,
+      "grad_norm": 4.0234375,
+      "learning_rate": 4.9272848565710474e-05,
+      "loss": 1.3239,
+      "step": 7614
+    },
+    {
+      "epoch": 7.029771520886222,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.926617745163443e-05,
+      "loss": 1.5303,
+      "step": 7615
+    },
+    {
+      "epoch": 7.030694668820678,
+      "grad_norm": 3.56640625,
+      "learning_rate": 4.925950633755837e-05,
+      "loss": 1.2825,
+      "step": 7616
+    },
+    {
+      "epoch": 7.031617816755135,
+      "grad_norm": 2.83984375,
+      "learning_rate": 4.9252835223482324e-05,
+      "loss": 1.354,
+      "step": 7617
+    },
+    {
+      "epoch": 7.032540964689591,
+      "grad_norm": 3.2734375,
+      "learning_rate": 4.9246164109406277e-05,
+      "loss": 1.2117,
+      "step": 7618
+    },
+    {
+      "epoch": 7.033464112624048,
+      "grad_norm": 3.017578125,
+      "learning_rate": 4.923949299533022e-05,
+      "loss": 1.3147,
+      "step": 7619
+    },
+    {
+      "epoch": 7.034387260558504,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.9232821881254174e-05,
+      "loss": 1.3287,
+      "step": 7620
+    },
+    {
+      "epoch": 7.035310408492961,
+      "grad_norm": 3.94921875,
+      "learning_rate": 4.922615076717812e-05,
+      "loss": 1.3843,
+      "step": 7621
+    },
+    {
+      "epoch": 7.036233556427417,
+      "grad_norm": 2.962890625,
+      "learning_rate": 4.921947965310207e-05,
+      "loss": 1.5461,
+      "step": 7622
+    },
+    {
+      "epoch": 7.037156704361874,
+      "grad_norm": 3.587890625,
+      "learning_rate": 4.921280853902602e-05,
+      "loss": 1.5515,
+      "step": 7623
+    },
+    {
+      "epoch": 7.03807985229633,
+      "grad_norm": 3.435546875,
+      "learning_rate": 4.920613742494997e-05,
+      "loss": 1.177,
+      "step": 7624
+    },
+    {
+      "epoch": 7.039003000230787,
+      "grad_norm": 3.69921875,
+      "learning_rate": 4.919946631087392e-05,
+      "loss": 1.3237,
+      "step": 7625
+    },
+    {
+      "epoch": 7.039926148165243,
+      "grad_norm": 3.375,
+      "learning_rate": 4.919279519679787e-05,
+      "loss": 1.2749,
+      "step": 7626
+    },
+    {
+      "epoch": 7.0408492960997,
+      "grad_norm": 3.4140625,
+      "learning_rate": 4.918612408272182e-05,
+      "loss": 1.3645,
+      "step": 7627
+    },
+    {
+      "epoch": 7.041772444034157,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9179452968645765e-05,
+      "loss": 1.2094,
+      "step": 7628
+    },
+    {
+      "epoch": 7.042695591968613,
+      "grad_norm": 3.375,
+      "learning_rate": 4.917278185456972e-05,
+      "loss": 1.2878,
+      "step": 7629
+    },
+    {
+      "epoch": 7.04361873990307,
+      "grad_norm": 3.216796875,
+      "learning_rate": 4.916611074049366e-05,
+      "loss": 1.3328,
+      "step": 7630
+    },
+    {
+      "epoch": 7.044541887837526,
+      "grad_norm": 3.060546875,
+      "learning_rate": 4.9159439626417615e-05,
+      "loss": 1.3328,
+      "step": 7631
+    },
+    {
+      "epoch": 7.045465035771983,
+      "grad_norm": 3.755859375,
+      "learning_rate": 4.915276851234156e-05,
+      "loss": 1.4382,
+      "step": 7632
+    },
+    {
+      "epoch": 7.046388183706439,
+      "grad_norm": 3.314453125,
+      "learning_rate": 4.914609739826551e-05,
+      "loss": 1.2264,
+      "step": 7633
+    },
+    {
+      "epoch": 7.047311331640896,
+      "grad_norm": 3.4140625,
+      "learning_rate": 4.913942628418946e-05,
+      "loss": 1.2667,
+      "step": 7634
+    },
+    {
+      "epoch": 7.048234479575352,
+      "grad_norm": 3.193359375,
+      "learning_rate": 4.913275517011341e-05,
+      "loss": 1.2688,
+      "step": 7635
+    },
+    {
+      "epoch": 7.049157627509809,
+      "grad_norm": 2.92578125,
+      "learning_rate": 4.9126084056037355e-05,
+      "loss": 1.4763,
+      "step": 7636
+    },
+    {
+      "epoch": 7.050080775444265,
+      "grad_norm": 3.6015625,
+      "learning_rate": 4.9119412941961314e-05,
+      "loss": 1.4712,
+      "step": 7637
+    },
+    {
+      "epoch": 7.051003923378722,
+      "grad_norm": 3.615234375,
+      "learning_rate": 4.911274182788526e-05,
+      "loss": 1.4133,
+      "step": 7638
+    },
+    {
+      "epoch": 7.051927071313178,
+      "grad_norm": 4.6640625,
+      "learning_rate": 4.910607071380921e-05,
+      "loss": 1.4111,
+      "step": 7639
+    },
+    {
+      "epoch": 7.0528502192476346,
+      "grad_norm": 3.576171875,
+      "learning_rate": 4.909939959973316e-05,
+      "loss": 1.3276,
+      "step": 7640
+    },
+    {
+      "epoch": 7.053773367182091,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.909272848565711e-05,
+      "loss": 1.1204,
+      "step": 7641
+    },
+    {
+      "epoch": 7.0546965151165475,
+      "grad_norm": 2.96484375,
+      "learning_rate": 4.9086057371581055e-05,
+      "loss": 1.0435,
+      "step": 7642
+    },
+    {
+      "epoch": 7.0556196630510035,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.907938625750501e-05,
+      "loss": 1.6575,
+      "step": 7643
+    },
+    {
+      "epoch": 7.05654281098546,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.907271514342895e-05,
+      "loss": 1.4279,
+      "step": 7644
+    },
+    {
+      "epoch": 7.057465958919917,
+      "grad_norm": 3.51953125,
+      "learning_rate": 4.9066044029352905e-05,
+      "loss": 1.3121,
+      "step": 7645
+    },
+    {
+      "epoch": 7.058389106854373,
+      "grad_norm": 3.537109375,
+      "learning_rate": 4.905937291527685e-05,
+      "loss": 1.303,
+      "step": 7646
+    },
+    {
+      "epoch": 7.05931225478883,
+      "grad_norm": 3.931640625,
+      "learning_rate": 4.90527018012008e-05,
+      "loss": 1.4338,
+      "step": 7647
+    },
+    {
+      "epoch": 7.060235402723286,
+      "grad_norm": 4.18359375,
+      "learning_rate": 4.904603068712475e-05,
+      "loss": 1.3896,
+      "step": 7648
+    },
+    {
+      "epoch": 7.061158550657743,
+      "grad_norm": 3.537109375,
+      "learning_rate": 4.90393595730487e-05,
+      "loss": 1.333,
+      "step": 7649
+    },
+    {
+      "epoch": 7.062081698592199,
+      "grad_norm": 3.619140625,
+      "learning_rate": 4.903268845897265e-05,
+      "loss": 1.4058,
+      "step": 7650
+    },
+    {
+      "epoch": 7.063004846526656,
+      "grad_norm": 3.529296875,
+      "learning_rate": 4.9026017344896604e-05,
+      "loss": 0.9402,
+      "step": 7651
+    },
+    {
+      "epoch": 7.063927994461112,
+      "grad_norm": 3.509765625,
+      "learning_rate": 4.901934623082055e-05,
+      "loss": 1.2026,
+      "step": 7652
+    },
+    {
+      "epoch": 7.064851142395569,
+      "grad_norm": 4.73046875,
+      "learning_rate": 4.90126751167445e-05,
+      "loss": 1.4871,
+      "step": 7653
+    },
+    {
+      "epoch": 7.065774290330025,
+      "grad_norm": 3.25390625,
+      "learning_rate": 4.900600400266845e-05,
+      "loss": 1.2646,
+      "step": 7654
+    },
+    {
+      "epoch": 7.066697438264482,
+      "grad_norm": 3.4453125,
+      "learning_rate": 4.89993328885924e-05,
+      "loss": 1.416,
+      "step": 7655
+    },
+    {
+      "epoch": 7.067620586198938,
+      "grad_norm": 3.70703125,
+      "learning_rate": 4.8992661774516345e-05,
+      "loss": 0.9898,
+      "step": 7656
+    },
+    {
+      "epoch": 7.068543734133395,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.89859906604403e-05,
+      "loss": 1.3547,
+      "step": 7657
+    },
+    {
+      "epoch": 7.069466882067851,
+      "grad_norm": 4.48828125,
+      "learning_rate": 4.897931954636424e-05,
+      "loss": 1.479,
+      "step": 7658
+    },
+    {
+      "epoch": 7.070390030002308,
+      "grad_norm": 3.07421875,
+      "learning_rate": 4.8972648432288195e-05,
+      "loss": 1.4102,
+      "step": 7659
+    },
+    {
+      "epoch": 7.071313177936764,
+      "grad_norm": 3.490234375,
+      "learning_rate": 4.896597731821214e-05,
+      "loss": 1.3547,
+      "step": 7660
+    },
+    {
+      "epoch": 7.072236325871221,
+      "grad_norm": 3.47265625,
+      "learning_rate": 4.895930620413609e-05,
+      "loss": 0.9969,
+      "step": 7661
+    },
+    {
+      "epoch": 7.073159473805678,
+      "grad_norm": 3.43359375,
+      "learning_rate": 4.8952635090060045e-05,
+      "loss": 1.3005,
+      "step": 7662
+    },
+    {
+      "epoch": 7.074082621740134,
+      "grad_norm": 3.591796875,
+      "learning_rate": 4.8945963975984e-05,
+      "loss": 1.3955,
+      "step": 7663
+    },
+    {
+      "epoch": 7.075005769674591,
+      "grad_norm": 3.099609375,
+      "learning_rate": 4.893929286190794e-05,
+      "loss": 0.9302,
+      "step": 7664
+    },
+    {
+      "epoch": 7.075928917609047,
+      "grad_norm": 3.5859375,
+      "learning_rate": 4.8932621747831895e-05,
+      "loss": 1.5508,
+      "step": 7665
+    },
+    {
+      "epoch": 7.076852065543504,
+      "grad_norm": 3.697265625,
+      "learning_rate": 4.892595063375584e-05,
+      "loss": 1.452,
+      "step": 7666
+    },
+    {
+      "epoch": 7.07777521347796,
+      "grad_norm": 3.119140625,
+      "learning_rate": 4.891927951967979e-05,
+      "loss": 1.3528,
+      "step": 7667
+    },
+    {
+      "epoch": 7.078698361412417,
+      "grad_norm": 3.99609375,
+      "learning_rate": 4.891260840560374e-05,
+      "loss": 0.985,
+      "step": 7668
+    },
+    {
+      "epoch": 7.079621509346873,
+      "grad_norm": 3.88671875,
+      "learning_rate": 4.890593729152769e-05,
+      "loss": 1.2755,
+      "step": 7669
+    },
+    {
+      "epoch": 7.0805446572813295,
+      "grad_norm": 3.404296875,
+      "learning_rate": 4.8899266177451635e-05,
+      "loss": 1.5281,
+      "step": 7670
+    },
+    {
+      "epoch": 7.081467805215786,
+      "grad_norm": 3.43359375,
+      "learning_rate": 4.889259506337559e-05,
+      "loss": 1.4226,
+      "step": 7671
+    },
+    {
+      "epoch": 7.0823909531502425,
+      "grad_norm": 3.2421875,
+      "learning_rate": 4.888592394929953e-05,
+      "loss": 1.4661,
+      "step": 7672
+    },
+    {
+      "epoch": 7.0833141010846985,
+      "grad_norm": 3.416015625,
+      "learning_rate": 4.8879252835223485e-05,
+      "loss": 1.376,
+      "step": 7673
+    },
+    {
+      "epoch": 7.084237249019155,
+      "grad_norm": 3.681640625,
+      "learning_rate": 4.887258172114743e-05,
+      "loss": 1.3381,
+      "step": 7674
+    },
+    {
+      "epoch": 7.085160396953611,
+      "grad_norm": 4.62890625,
+      "learning_rate": 4.886591060707138e-05,
+      "loss": 1.5044,
+      "step": 7675
+    },
+    {
+      "epoch": 7.086083544888068,
+      "grad_norm": 3.810546875,
+      "learning_rate": 4.8859239492995335e-05,
+      "loss": 1.236,
+      "step": 7676
+    },
+    {
+      "epoch": 7.087006692822524,
+      "grad_norm": 4.26171875,
+      "learning_rate": 4.885256837891928e-05,
+      "loss": 1.2844,
+      "step": 7677
+    },
+    {
+      "epoch": 7.087929840756981,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.884589726484323e-05,
+      "loss": 1.3361,
+      "step": 7678
+    },
+    {
+      "epoch": 7.088852988691438,
+      "grad_norm": 4.06640625,
+      "learning_rate": 4.883922615076718e-05,
+      "loss": 1.4446,
+      "step": 7679
+    },
+    {
+      "epoch": 7.089776136625894,
+      "grad_norm": 3.091796875,
+      "learning_rate": 4.883255503669113e-05,
+      "loss": 1.2483,
+      "step": 7680
+    },
+    {
+      "epoch": 7.090699284560351,
+      "grad_norm": 3.669921875,
+      "learning_rate": 4.8825883922615076e-05,
+      "loss": 1.3142,
+      "step": 7681
+    },
+    {
+      "epoch": 7.091622432494807,
+      "grad_norm": 3.78515625,
+      "learning_rate": 4.881921280853903e-05,
+      "loss": 1.4092,
+      "step": 7682
+    },
+    {
+      "epoch": 7.092545580429264,
+      "grad_norm": 3.216796875,
+      "learning_rate": 4.8812541694462973e-05,
+      "loss": 1.5261,
+      "step": 7683
+    },
+    {
+      "epoch": 7.09346872836372,
+      "grad_norm": 3.89453125,
+      "learning_rate": 4.8805870580386926e-05,
+      "loss": 1.2081,
+      "step": 7684
+    },
+    {
+      "epoch": 7.094391876298177,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.879919946631087e-05,
+      "loss": 1.2549,
+      "step": 7685
+    },
+    {
+      "epoch": 7.095315024232633,
+      "grad_norm": 3.51953125,
+      "learning_rate": 4.879252835223482e-05,
+      "loss": 1.0868,
+      "step": 7686
+    },
+    {
+      "epoch": 7.09623817216709,
+      "grad_norm": 3.48828125,
+      "learning_rate": 4.8785857238158775e-05,
+      "loss": 0.9575,
+      "step": 7687
+    },
+    {
+      "epoch": 7.097161320101546,
+      "grad_norm": 3.07421875,
+      "learning_rate": 4.877918612408273e-05,
+      "loss": 1.207,
+      "step": 7688
+    },
+    {
+      "epoch": 7.098084468036003,
+      "grad_norm": 4.265625,
+      "learning_rate": 4.877251501000667e-05,
+      "loss": 1.3799,
+      "step": 7689
+    },
+    {
+      "epoch": 7.099007615970459,
+      "grad_norm": 3.611328125,
+      "learning_rate": 4.8765843895930625e-05,
+      "loss": 1.1044,
+      "step": 7690
+    },
+    {
+      "epoch": 7.099930763904916,
+      "grad_norm": 3.240234375,
+      "learning_rate": 4.875917278185457e-05,
+      "loss": 1.2567,
+      "step": 7691
+    },
+    {
+      "epoch": 7.100853911839372,
+      "grad_norm": 3.732421875,
+      "learning_rate": 4.875250166777852e-05,
+      "loss": 1.2446,
+      "step": 7692
+    },
+    {
+      "epoch": 7.101777059773829,
+      "grad_norm": 3.5234375,
+      "learning_rate": 4.874583055370247e-05,
+      "loss": 1.5432,
+      "step": 7693
+    },
+    {
+      "epoch": 7.102700207708285,
+      "grad_norm": 4.546875,
+      "learning_rate": 4.873915943962642e-05,
+      "loss": 1.3071,
+      "step": 7694
+    },
+    {
+      "epoch": 7.103623355642742,
+      "grad_norm": 2.91015625,
+      "learning_rate": 4.8732488325550366e-05,
+      "loss": 1.3584,
+      "step": 7695
+    },
+    {
+      "epoch": 7.104546503577199,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.872581721147432e-05,
+      "loss": 1.2001,
+      "step": 7696
+    },
+    {
+      "epoch": 7.105469651511655,
+      "grad_norm": 3.482421875,
+      "learning_rate": 4.8719146097398264e-05,
+      "loss": 1.2268,
+      "step": 7697
+    },
+    {
+      "epoch": 7.106392799446112,
+      "grad_norm": 3.275390625,
+      "learning_rate": 4.8712474983322216e-05,
+      "loss": 1.5022,
+      "step": 7698
+    },
+    {
+      "epoch": 7.107315947380568,
+      "grad_norm": 3.375,
+      "learning_rate": 4.870580386924616e-05,
+      "loss": 1.4678,
+      "step": 7699
+    },
+    {
+      "epoch": 7.1082390953150245,
+      "grad_norm": 4.49609375,
+      "learning_rate": 4.869913275517012e-05,
+      "loss": 1.6184,
+      "step": 7700
+    },
+    {
+      "epoch": 7.1091622432494805,
+      "grad_norm": 3.724609375,
+      "learning_rate": 4.8692461641094066e-05,
+      "loss": 1.0704,
+      "step": 7701
+    },
+    {
+      "epoch": 7.1100853911839375,
+      "grad_norm": 3.51171875,
+      "learning_rate": 4.868579052701802e-05,
+      "loss": 1.5312,
+      "step": 7702
+    },
+    {
+      "epoch": 7.1110085391183935,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.867911941294196e-05,
+      "loss": 1.5049,
+      "step": 7703
+    },
+    {
+      "epoch": 7.11193168705285,
+      "grad_norm": 3.86328125,
+      "learning_rate": 4.8672448298865916e-05,
+      "loss": 1.0784,
+      "step": 7704
+    },
+    {
+      "epoch": 7.112854834987306,
+      "grad_norm": 3.517578125,
+      "learning_rate": 4.866577718478986e-05,
+      "loss": 1.5935,
+      "step": 7705
+    },
+    {
+      "epoch": 7.113777982921763,
+      "grad_norm": 3.103515625,
+      "learning_rate": 4.865910607071381e-05,
+      "loss": 1.5942,
+      "step": 7706
+    },
+    {
+      "epoch": 7.114701130856219,
+      "grad_norm": 3.978515625,
+      "learning_rate": 4.865243495663776e-05,
+      "loss": 1.3267,
+      "step": 7707
+    },
+    {
+      "epoch": 7.115624278790676,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.864576384256171e-05,
+      "loss": 1.5469,
+      "step": 7708
+    },
+    {
+      "epoch": 7.116547426725132,
+      "grad_norm": 3.6015625,
+      "learning_rate": 4.8639092728485656e-05,
+      "loss": 1.5786,
+      "step": 7709
+    },
+    {
+      "epoch": 7.117470574659589,
+      "grad_norm": 3.142578125,
+      "learning_rate": 4.863242161440961e-05,
+      "loss": 1.3723,
+      "step": 7710
+    },
+    {
+      "epoch": 7.118393722594046,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.8625750500333554e-05,
+      "loss": 1.4038,
+      "step": 7711
+    },
+    {
+      "epoch": 7.119316870528502,
+      "grad_norm": 4.0390625,
+      "learning_rate": 4.861907938625751e-05,
+      "loss": 1.606,
+      "step": 7712
+    },
+    {
+      "epoch": 7.120240018462959,
+      "grad_norm": 3.37109375,
+      "learning_rate": 4.861240827218146e-05,
+      "loss": 1.3475,
+      "step": 7713
+    },
+    {
+      "epoch": 7.121163166397415,
+      "grad_norm": 3.470703125,
+      "learning_rate": 4.860573715810541e-05,
+      "loss": 1.4875,
+      "step": 7714
+    },
+    {
+      "epoch": 7.122086314331872,
+      "grad_norm": 3.41015625,
+      "learning_rate": 4.8599066044029356e-05,
+      "loss": 1.255,
+      "step": 7715
+    },
+    {
+      "epoch": 7.123009462266328,
+      "grad_norm": 3.201171875,
+      "learning_rate": 4.859239492995331e-05,
+      "loss": 0.9826,
+      "step": 7716
+    },
+    {
+      "epoch": 7.123932610200785,
+      "grad_norm": 3.84765625,
+      "learning_rate": 4.8585723815877254e-05,
+      "loss": 1.5208,
+      "step": 7717
+    },
+    {
+      "epoch": 7.124855758135241,
+      "grad_norm": 3.318359375,
+      "learning_rate": 4.8579052701801206e-05,
+      "loss": 1.26,
+      "step": 7718
+    },
+    {
+      "epoch": 7.125778906069698,
+      "grad_norm": 3.177734375,
+      "learning_rate": 4.857238158772515e-05,
+      "loss": 1.3273,
+      "step": 7719
+    },
+    {
+      "epoch": 7.126702054004154,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.8565710473649103e-05,
+      "loss": 1.0487,
+      "step": 7720
+    },
+    {
+      "epoch": 7.127625201938611,
+      "grad_norm": 3.2265625,
+      "learning_rate": 4.855903935957305e-05,
+      "loss": 1.3987,
+      "step": 7721
+    },
+    {
+      "epoch": 7.128548349873067,
+      "grad_norm": 3.97265625,
+      "learning_rate": 4.8552368245497e-05,
+      "loss": 1.4331,
+      "step": 7722
+    },
+    {
+      "epoch": 7.129471497807524,
+      "grad_norm": 3.564453125,
+      "learning_rate": 4.8545697131420947e-05,
+      "loss": 1.3396,
+      "step": 7723
+    },
+    {
+      "epoch": 7.13039464574198,
+      "grad_norm": 3.3203125,
+      "learning_rate": 4.85390260173449e-05,
+      "loss": 1.2542,
+      "step": 7724
+    },
+    {
+      "epoch": 7.131317793676437,
+      "grad_norm": 3.2734375,
+      "learning_rate": 4.853235490326885e-05,
+      "loss": 1.1393,
+      "step": 7725
+    },
+    {
+      "epoch": 7.132240941610893,
+      "grad_norm": 3.509765625,
+      "learning_rate": 4.8525683789192796e-05,
+      "loss": 1.5762,
+      "step": 7726
+    },
+    {
+      "epoch": 7.13316408954535,
+      "grad_norm": 3.626953125,
+      "learning_rate": 4.851901267511675e-05,
+      "loss": 1.2327,
+      "step": 7727
+    },
+    {
+      "epoch": 7.1340872374798066,
+      "grad_norm": 3.181640625,
+      "learning_rate": 4.8512341561040694e-05,
+      "loss": 1.4111,
+      "step": 7728
+    },
+    {
+      "epoch": 7.135010385414263,
+      "grad_norm": 3.513671875,
+      "learning_rate": 4.8505670446964646e-05,
+      "loss": 1.4346,
+      "step": 7729
+    },
+    {
+      "epoch": 7.1359335333487195,
+      "grad_norm": 3.751953125,
+      "learning_rate": 4.849899933288859e-05,
+      "loss": 1.0321,
+      "step": 7730
+    },
+    {
+      "epoch": 7.1368566812831755,
+      "grad_norm": 4.0234375,
+      "learning_rate": 4.8492328218812544e-05,
+      "loss": 1.1987,
+      "step": 7731
+    },
+    {
+      "epoch": 7.137779829217632,
+      "grad_norm": 3.61328125,
+      "learning_rate": 4.8485657104736496e-05,
+      "loss": 1.2444,
+      "step": 7732
+    },
+    {
+      "epoch": 7.1387029771520885,
+      "grad_norm": 3.7109375,
+      "learning_rate": 4.847898599066044e-05,
+      "loss": 1.7161,
+      "step": 7733
+    },
+    {
+      "epoch": 7.139626125086545,
+      "grad_norm": 3.041015625,
+      "learning_rate": 4.8472314876584394e-05,
+      "loss": 1.333,
+      "step": 7734
+    },
+    {
+      "epoch": 7.140549273021001,
+      "grad_norm": 3.48828125,
+      "learning_rate": 4.846564376250834e-05,
+      "loss": 1.364,
+      "step": 7735
+    },
+    {
+      "epoch": 7.141472420955458,
+      "grad_norm": 3.298828125,
+      "learning_rate": 4.845897264843229e-05,
+      "loss": 1.1986,
+      "step": 7736
+    },
+    {
+      "epoch": 7.142395568889914,
+      "grad_norm": 4.23828125,
+      "learning_rate": 4.8452301534356244e-05,
+      "loss": 1.324,
+      "step": 7737
+    },
+    {
+      "epoch": 7.143318716824371,
+      "grad_norm": 2.8984375,
+      "learning_rate": 4.844563042028019e-05,
+      "loss": 1.3906,
+      "step": 7738
+    },
+    {
+      "epoch": 7.144241864758827,
+      "grad_norm": 3.33203125,
+      "learning_rate": 4.843895930620414e-05,
+      "loss": 1.3625,
+      "step": 7739
+    },
+    {
+      "epoch": 7.145165012693284,
+      "grad_norm": 3.58984375,
+      "learning_rate": 4.8432288192128087e-05,
+      "loss": 1.2313,
+      "step": 7740
+    },
+    {
+      "epoch": 7.14608816062774,
+      "grad_norm": 3.603515625,
+      "learning_rate": 4.842561707805204e-05,
+      "loss": 1.5457,
+      "step": 7741
+    },
+    {
+      "epoch": 7.147011308562197,
+      "grad_norm": 3.619140625,
+      "learning_rate": 4.8418945963975984e-05,
+      "loss": 1.2312,
+      "step": 7742
+    },
+    {
+      "epoch": 7.147934456496653,
+      "grad_norm": 3.056640625,
+      "learning_rate": 4.8412274849899936e-05,
+      "loss": 1.4133,
+      "step": 7743
+    },
+    {
+      "epoch": 7.14885760443111,
+      "grad_norm": 3.943359375,
+      "learning_rate": 4.840560373582388e-05,
+      "loss": 1.2611,
+      "step": 7744
+    },
+    {
+      "epoch": 7.149780752365567,
+      "grad_norm": 3.134765625,
+      "learning_rate": 4.8398932621747834e-05,
+      "loss": 1.3391,
+      "step": 7745
+    },
+    {
+      "epoch": 7.150703900300023,
+      "grad_norm": 3.5234375,
+      "learning_rate": 4.839226150767178e-05,
+      "loss": 1.3313,
+      "step": 7746
+    },
+    {
+      "epoch": 7.15162704823448,
+      "grad_norm": 3.375,
+      "learning_rate": 4.838559039359573e-05,
+      "loss": 1.2671,
+      "step": 7747
+    },
+    {
+      "epoch": 7.152550196168936,
+      "grad_norm": 3.322265625,
+      "learning_rate": 4.837891927951968e-05,
+      "loss": 1.2321,
+      "step": 7748
+    },
+    {
+      "epoch": 7.153473344103393,
+      "grad_norm": 3.9140625,
+      "learning_rate": 4.837224816544363e-05,
+      "loss": 1.1555,
+      "step": 7749
+    },
+    {
+      "epoch": 7.154396492037849,
+      "grad_norm": 3.595703125,
+      "learning_rate": 4.836557705136758e-05,
+      "loss": 1.3792,
+      "step": 7750
+    },
+    {
+      "epoch": 7.155319639972306,
+      "grad_norm": 3.595703125,
+      "learning_rate": 4.8358905937291534e-05,
+      "loss": 1.2051,
+      "step": 7751
+    },
+    {
+      "epoch": 7.156242787906762,
+      "grad_norm": 3.208984375,
+      "learning_rate": 4.835223482321548e-05,
+      "loss": 1.3116,
+      "step": 7752
+    },
+    {
+      "epoch": 7.157165935841219,
+      "grad_norm": 3.431640625,
+      "learning_rate": 4.834556370913943e-05,
+      "loss": 1.3506,
+      "step": 7753
+    },
+    {
+      "epoch": 7.158089083775675,
+      "grad_norm": 3.548828125,
+      "learning_rate": 4.833889259506338e-05,
+      "loss": 1.0767,
+      "step": 7754
+    },
+    {
+      "epoch": 7.159012231710132,
+      "grad_norm": 3.232421875,
+      "learning_rate": 4.833222148098733e-05,
+      "loss": 1.3503,
+      "step": 7755
+    },
+    {
+      "epoch": 7.159935379644588,
+      "grad_norm": 3.404296875,
+      "learning_rate": 4.8325550366911274e-05,
+      "loss": 1.5098,
+      "step": 7756
+    },
+    {
+      "epoch": 7.160858527579045,
+      "grad_norm": 3.123046875,
+      "learning_rate": 4.831887925283523e-05,
+      "loss": 1.1556,
+      "step": 7757
+    },
+    {
+      "epoch": 7.161781675513501,
+      "grad_norm": 3.43359375,
+      "learning_rate": 4.831220813875917e-05,
+      "loss": 1.2703,
+      "step": 7758
+    },
+    {
+      "epoch": 7.162704823447958,
+      "grad_norm": 3.43359375,
+      "learning_rate": 4.8305537024683124e-05,
+      "loss": 1.1154,
+      "step": 7759
+    },
+    {
+      "epoch": 7.1636279713824145,
+      "grad_norm": 3.8515625,
+      "learning_rate": 4.829886591060707e-05,
+      "loss": 1.3989,
+      "step": 7760
+    },
+    {
+      "epoch": 7.1645511193168705,
+      "grad_norm": 3.71484375,
+      "learning_rate": 4.829219479653102e-05,
+      "loss": 1.4099,
+      "step": 7761
+    },
+    {
+      "epoch": 7.165474267251327,
+      "grad_norm": 3.74609375,
+      "learning_rate": 4.8285523682454974e-05,
+      "loss": 1.446,
+      "step": 7762
+    },
+    {
+      "epoch": 7.166397415185783,
+      "grad_norm": 3.0546875,
+      "learning_rate": 4.8278852568378926e-05,
+      "loss": 1.499,
+      "step": 7763
+    },
+    {
+      "epoch": 7.16732056312024,
+      "grad_norm": 3.23828125,
+      "learning_rate": 4.827218145430287e-05,
+      "loss": 1.4358,
+      "step": 7764
+    },
+    {
+      "epoch": 7.168243711054696,
+      "grad_norm": 3.240234375,
+      "learning_rate": 4.8265510340226824e-05,
+      "loss": 1.5283,
+      "step": 7765
+    },
+    {
+      "epoch": 7.169166858989153,
+      "grad_norm": 3.560546875,
+      "learning_rate": 4.825883922615077e-05,
+      "loss": 1.0503,
+      "step": 7766
+    },
+    {
+      "epoch": 7.170090006923609,
+      "grad_norm": 3.6953125,
+      "learning_rate": 4.825216811207472e-05,
+      "loss": 1.2288,
+      "step": 7767
+    },
+    {
+      "epoch": 7.171013154858066,
+      "grad_norm": 2.939453125,
+      "learning_rate": 4.824549699799867e-05,
+      "loss": 1.0872,
+      "step": 7768
+    },
+    {
+      "epoch": 7.171936302792522,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.823882588392262e-05,
+      "loss": 1.0491,
+      "step": 7769
+    },
+    {
+      "epoch": 7.172859450726979,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.8232154769846565e-05,
+      "loss": 1.269,
+      "step": 7770
+    },
+    {
+      "epoch": 7.173782598661435,
+      "grad_norm": 3.12890625,
+      "learning_rate": 4.822548365577052e-05,
+      "loss": 1.1392,
+      "step": 7771
+    },
+    {
+      "epoch": 7.174705746595892,
+      "grad_norm": 3.705078125,
+      "learning_rate": 4.821881254169446e-05,
+      "loss": 1.325,
+      "step": 7772
+    },
+    {
+      "epoch": 7.175628894530348,
+      "grad_norm": 3.87890625,
+      "learning_rate": 4.8212141427618415e-05,
+      "loss": 1.0663,
+      "step": 7773
+    },
+    {
+      "epoch": 7.176552042464805,
+      "grad_norm": 3.3984375,
+      "learning_rate": 4.820547031354236e-05,
+      "loss": 1.4739,
+      "step": 7774
+    },
+    {
+      "epoch": 7.177475190399261,
+      "grad_norm": 3.849609375,
+      "learning_rate": 4.819879919946632e-05,
+      "loss": 1.4834,
+      "step": 7775
+    },
+    {
+      "epoch": 7.178398338333718,
+      "grad_norm": 2.947265625,
+      "learning_rate": 4.8192128085390264e-05,
+      "loss": 1.3032,
+      "step": 7776
+    },
+    {
+      "epoch": 7.179321486268175,
+      "grad_norm": 3.240234375,
+      "learning_rate": 4.8185456971314217e-05,
+      "loss": 1.5969,
+      "step": 7777
+    },
+    {
+      "epoch": 7.180244634202631,
+      "grad_norm": 3.388671875,
+      "learning_rate": 4.817878585723816e-05,
+      "loss": 1.4626,
+      "step": 7778
+    },
+    {
+      "epoch": 7.181167782137088,
+      "grad_norm": 3.82421875,
+      "learning_rate": 4.8172114743162114e-05,
+      "loss": 1.1459,
+      "step": 7779
+    },
+    {
+      "epoch": 7.182090930071544,
+      "grad_norm": 3.125,
+      "learning_rate": 4.816544362908606e-05,
+      "loss": 1.4001,
+      "step": 7780
+    },
+    {
+      "epoch": 7.183014078006001,
+      "grad_norm": 3.490234375,
+      "learning_rate": 4.815877251501001e-05,
+      "loss": 1.3804,
+      "step": 7781
+    },
+    {
+      "epoch": 7.183937225940457,
+      "grad_norm": 3.134765625,
+      "learning_rate": 4.815210140093396e-05,
+      "loss": 1.1023,
+      "step": 7782
+    },
+    {
+      "epoch": 7.184860373874914,
+      "grad_norm": 3.03515625,
+      "learning_rate": 4.814543028685791e-05,
+      "loss": 0.9559,
+      "step": 7783
+    },
+    {
+      "epoch": 7.18578352180937,
+      "grad_norm": 3.673828125,
+      "learning_rate": 4.8138759172781855e-05,
+      "loss": 1.5237,
+      "step": 7784
+    },
+    {
+      "epoch": 7.186706669743827,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.813208805870581e-05,
+      "loss": 1.4668,
+      "step": 7785
+    },
+    {
+      "epoch": 7.187629817678283,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.812541694462975e-05,
+      "loss": 1.314,
+      "step": 7786
+    },
+    {
+      "epoch": 7.18855296561274,
+      "grad_norm": 3.580078125,
+      "learning_rate": 4.8118745830553705e-05,
+      "loss": 1.0553,
+      "step": 7787
+    },
+    {
+      "epoch": 7.189476113547196,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.811207471647766e-05,
+      "loss": 0.9425,
+      "step": 7788
+    },
+    {
+      "epoch": 7.1903992614816525,
+      "grad_norm": 3.75,
+      "learning_rate": 4.81054036024016e-05,
+      "loss": 1.0907,
+      "step": 7789
+    },
+    {
+      "epoch": 7.191322409416109,
+      "grad_norm": 3.48828125,
+      "learning_rate": 4.8098732488325555e-05,
+      "loss": 1.3163,
+      "step": 7790
+    },
+    {
+      "epoch": 7.1922455573505655,
+      "grad_norm": 3.42578125,
+      "learning_rate": 4.80920613742495e-05,
+      "loss": 1.4189,
+      "step": 7791
+    },
+    {
+      "epoch": 7.1931687052850215,
+      "grad_norm": 3.19921875,
+      "learning_rate": 4.808539026017345e-05,
+      "loss": 1.4138,
+      "step": 7792
+    },
+    {
+      "epoch": 7.194091853219478,
+      "grad_norm": 3.576171875,
+      "learning_rate": 4.80787191460974e-05,
+      "loss": 1.3315,
+      "step": 7793
+    },
+    {
+      "epoch": 7.195015001153935,
+      "grad_norm": 2.8984375,
+      "learning_rate": 4.807204803202135e-05,
+      "loss": 1.4033,
+      "step": 7794
+    },
+    {
+      "epoch": 7.195938149088391,
+      "grad_norm": 3.490234375,
+      "learning_rate": 4.8065376917945295e-05,
+      "loss": 1.2019,
+      "step": 7795
+    },
+    {
+      "epoch": 7.196861297022848,
+      "grad_norm": 3.568359375,
+      "learning_rate": 4.805870580386925e-05,
+      "loss": 1.5469,
+      "step": 7796
+    },
+    {
+      "epoch": 7.197784444957304,
+      "grad_norm": 3.66015625,
+      "learning_rate": 4.805203468979319e-05,
+      "loss": 1.4631,
+      "step": 7797
+    },
+    {
+      "epoch": 7.198707592891761,
+      "grad_norm": 3.263671875,
+      "learning_rate": 4.8045363575717145e-05,
+      "loss": 1.4856,
+      "step": 7798
+    },
+    {
+      "epoch": 7.199630740826217,
+      "grad_norm": 3.302734375,
+      "learning_rate": 4.803869246164109e-05,
+      "loss": 1.304,
+      "step": 7799
+    },
+    {
+      "epoch": 7.200553888760674,
+      "grad_norm": 4.00390625,
+      "learning_rate": 4.803202134756505e-05,
+      "loss": 1.3875,
+      "step": 7800
+    },
+    {
+      "epoch": 7.20147703669513,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.8025350233488995e-05,
+      "loss": 1.5544,
+      "step": 7801
+    },
+    {
+      "epoch": 7.202400184629587,
+      "grad_norm": 3.564453125,
+      "learning_rate": 4.801867911941295e-05,
+      "loss": 1.5562,
+      "step": 7802
+    },
+    {
+      "epoch": 7.203323332564043,
+      "grad_norm": 3.271484375,
+      "learning_rate": 4.801200800533689e-05,
+      "loss": 1.4595,
+      "step": 7803
+    },
+    {
+      "epoch": 7.2042464804985,
+      "grad_norm": 3.509765625,
+      "learning_rate": 4.8005336891260845e-05,
+      "loss": 1.3535,
+      "step": 7804
+    },
+    {
+      "epoch": 7.205169628432956,
+      "grad_norm": 3.287109375,
+      "learning_rate": 4.799866577718479e-05,
+      "loss": 1.3821,
+      "step": 7805
+    },
+    {
+      "epoch": 7.206092776367413,
+      "grad_norm": 3.52734375,
+      "learning_rate": 4.799199466310874e-05,
+      "loss": 1.3083,
+      "step": 7806
+    },
+    {
+      "epoch": 7.207015924301869,
+      "grad_norm": 3.853515625,
+      "learning_rate": 4.798532354903269e-05,
+      "loss": 1.5449,
+      "step": 7807
+    },
+    {
+      "epoch": 7.207939072236326,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.797865243495664e-05,
+      "loss": 1.5234,
+      "step": 7808
+    },
+    {
+      "epoch": 7.208862220170783,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.7971981320880586e-05,
+      "loss": 1.3896,
+      "step": 7809
+    },
+    {
+      "epoch": 7.209785368105239,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.796531020680454e-05,
+      "loss": 1.2671,
+      "step": 7810
+    },
+    {
+      "epoch": 7.210708516039696,
+      "grad_norm": 3.576171875,
+      "learning_rate": 4.795863909272848e-05,
+      "loss": 1.2534,
+      "step": 7811
+    },
+    {
+      "epoch": 7.211631663974152,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.7951967978652435e-05,
+      "loss": 1.2709,
+      "step": 7812
+    },
+    {
+      "epoch": 7.212554811908609,
+      "grad_norm": 3.095703125,
+      "learning_rate": 4.794529686457639e-05,
+      "loss": 1.2114,
+      "step": 7813
+    },
+    {
+      "epoch": 7.213477959843065,
+      "grad_norm": 3.73046875,
+      "learning_rate": 4.793862575050034e-05,
+      "loss": 1.4399,
+      "step": 7814
+    },
+    {
+      "epoch": 7.214401107777522,
+      "grad_norm": 3.892578125,
+      "learning_rate": 4.7931954636424285e-05,
+      "loss": 1.6985,
+      "step": 7815
+    },
+    {
+      "epoch": 7.215324255711978,
+      "grad_norm": 3.36328125,
+      "learning_rate": 4.792528352234824e-05,
+      "loss": 1.3076,
+      "step": 7816
+    },
+    {
+      "epoch": 7.216247403646435,
+      "grad_norm": 3.205078125,
+      "learning_rate": 4.791861240827218e-05,
+      "loss": 1.3217,
+      "step": 7817
+    },
+    {
+      "epoch": 7.217170551580891,
+      "grad_norm": 3.2109375,
+      "learning_rate": 4.7911941294196135e-05,
+      "loss": 1.1302,
+      "step": 7818
+    },
+    {
+      "epoch": 7.2180936995153475,
+      "grad_norm": 3.38671875,
+      "learning_rate": 4.790527018012008e-05,
+      "loss": 1.2911,
+      "step": 7819
+    },
+    {
+      "epoch": 7.2190168474498035,
+      "grad_norm": 3.09765625,
+      "learning_rate": 4.789859906604403e-05,
+      "loss": 1.2715,
+      "step": 7820
+    },
+    {
+      "epoch": 7.2199399953842605,
+      "grad_norm": 3.6796875,
+      "learning_rate": 4.789192795196798e-05,
+      "loss": 1.4448,
+      "step": 7821
+    },
+    {
+      "epoch": 7.2208631433187165,
+      "grad_norm": 3.521484375,
+      "learning_rate": 4.788525683789193e-05,
+      "loss": 1.3044,
+      "step": 7822
+    },
+    {
+      "epoch": 7.221786291253173,
+      "grad_norm": 3.380859375,
+      "learning_rate": 4.7878585723815876e-05,
+      "loss": 1.4099,
+      "step": 7823
+    },
+    {
+      "epoch": 7.222709439187629,
+      "grad_norm": 3.740234375,
+      "learning_rate": 4.787191460973983e-05,
+      "loss": 1.0389,
+      "step": 7824
+    },
+    {
+      "epoch": 7.223632587122086,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.786524349566378e-05,
+      "loss": 1.2883,
+      "step": 7825
+    },
+    {
+      "epoch": 7.224555735056542,
+      "grad_norm": 3.91015625,
+      "learning_rate": 4.785857238158773e-05,
+      "loss": 1.1914,
+      "step": 7826
+    },
+    {
+      "epoch": 7.225478882990999,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.785190126751168e-05,
+      "loss": 1.4773,
+      "step": 7827
+    },
+    {
+      "epoch": 7.226402030925456,
+      "grad_norm": 3.759765625,
+      "learning_rate": 4.784523015343563e-05,
+      "loss": 1.345,
+      "step": 7828
+    },
+    {
+      "epoch": 7.227325178859912,
+      "grad_norm": 3.56640625,
+      "learning_rate": 4.7838559039359576e-05,
+      "loss": 1.49,
+      "step": 7829
+    },
+    {
+      "epoch": 7.228248326794369,
+      "grad_norm": 3.74609375,
+      "learning_rate": 4.783188792528353e-05,
+      "loss": 1.0691,
+      "step": 7830
+    },
+    {
+      "epoch": 7.229171474728825,
+      "grad_norm": 3.541015625,
+      "learning_rate": 4.782521681120747e-05,
+      "loss": 1.2379,
+      "step": 7831
+    },
+    {
+      "epoch": 7.230094622663282,
+      "grad_norm": 3.72265625,
+      "learning_rate": 4.7818545697131425e-05,
+      "loss": 1.3407,
+      "step": 7832
+    },
+    {
+      "epoch": 7.231017770597738,
+      "grad_norm": 3.646484375,
+      "learning_rate": 4.781187458305537e-05,
+      "loss": 1.1613,
+      "step": 7833
+    },
+    {
+      "epoch": 7.231940918532195,
+      "grad_norm": 4.234375,
+      "learning_rate": 4.780520346897932e-05,
+      "loss": 1.2737,
+      "step": 7834
+    },
+    {
+      "epoch": 7.232864066466651,
+      "grad_norm": 3.1640625,
+      "learning_rate": 4.779853235490327e-05,
+      "loss": 1.2573,
+      "step": 7835
+    },
+    {
+      "epoch": 7.233787214401108,
+      "grad_norm": 3.5234375,
+      "learning_rate": 4.779186124082722e-05,
+      "loss": 1.2791,
+      "step": 7836
+    },
+    {
+      "epoch": 7.234710362335564,
+      "grad_norm": 3.09765625,
+      "learning_rate": 4.7785190126751166e-05,
+      "loss": 1.2747,
+      "step": 7837
+    },
+    {
+      "epoch": 7.235633510270021,
+      "grad_norm": 3.73046875,
+      "learning_rate": 4.777851901267512e-05,
+      "loss": 1.2797,
+      "step": 7838
+    },
+    {
+      "epoch": 7.236556658204477,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.777184789859907e-05,
+      "loss": 1.0626,
+      "step": 7839
+    },
+    {
+      "epoch": 7.237479806138934,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.7765176784523016e-05,
+      "loss": 1.4805,
+      "step": 7840
+    },
+    {
+      "epoch": 7.23840295407339,
+      "grad_norm": 3.654296875,
+      "learning_rate": 4.775850567044697e-05,
+      "loss": 1.6357,
+      "step": 7841
+    },
+    {
+      "epoch": 7.239326102007847,
+      "grad_norm": 2.845703125,
+      "learning_rate": 4.7751834556370914e-05,
+      "loss": 1.2537,
+      "step": 7842
+    },
+    {
+      "epoch": 7.240249249942304,
+      "grad_norm": 3.33984375,
+      "learning_rate": 4.7745163442294866e-05,
+      "loss": 1.491,
+      "step": 7843
+    },
+    {
+      "epoch": 7.24117239787676,
+      "grad_norm": 3.40234375,
+      "learning_rate": 4.773849232821882e-05,
+      "loss": 1.0663,
+      "step": 7844
+    },
+    {
+      "epoch": 7.242095545811217,
+      "grad_norm": 3.443359375,
+      "learning_rate": 4.773182121414276e-05,
+      "loss": 1.5498,
+      "step": 7845
+    },
+    {
+      "epoch": 7.243018693745673,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.7725150100066716e-05,
+      "loss": 1.5815,
+      "step": 7846
+    },
+    {
+      "epoch": 7.24394184168013,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.771847898599066e-05,
+      "loss": 1.5117,
+      "step": 7847
+    },
+    {
+      "epoch": 7.244864989614586,
+      "grad_norm": 3.3203125,
+      "learning_rate": 4.771180787191461e-05,
+      "loss": 1.0869,
+      "step": 7848
+    },
+    {
+      "epoch": 7.2457881375490425,
+      "grad_norm": 3.6640625,
+      "learning_rate": 4.770513675783856e-05,
+      "loss": 1.203,
+      "step": 7849
+    },
+    {
+      "epoch": 7.2467112854834985,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.769846564376251e-05,
+      "loss": 1.4022,
+      "step": 7850
+    },
+    {
+      "epoch": 7.247634433417955,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.769179452968646e-05,
+      "loss": 1.3711,
+      "step": 7851
+    },
+    {
+      "epoch": 7.2485575813524115,
+      "grad_norm": 3.302734375,
+      "learning_rate": 4.768512341561041e-05,
+      "loss": 1.2649,
+      "step": 7852
+    },
+    {
+      "epoch": 7.249480729286868,
+      "grad_norm": 3.353515625,
+      "learning_rate": 4.767845230153436e-05,
+      "loss": 1.0251,
+      "step": 7853
+    },
+    {
+      "epoch": 7.250403877221324,
+      "grad_norm": 3.662109375,
+      "learning_rate": 4.7671781187458306e-05,
+      "loss": 1.3817,
+      "step": 7854
+    },
+    {
+      "epoch": 7.251327025155781,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.766511007338226e-05,
+      "loss": 1.4019,
+      "step": 7855
+    },
+    {
+      "epoch": 7.252250173090237,
+      "grad_norm": 3.24609375,
+      "learning_rate": 4.7658438959306204e-05,
+      "loss": 1.2896,
+      "step": 7856
+    },
+    {
+      "epoch": 7.253173321024694,
+      "grad_norm": 4.0078125,
+      "learning_rate": 4.7651767845230156e-05,
+      "loss": 1.1278,
+      "step": 7857
+    },
+    {
+      "epoch": 7.254096468959151,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.76450967311541e-05,
+      "loss": 1.3652,
+      "step": 7858
+    },
+    {
+      "epoch": 7.255019616893607,
+      "grad_norm": 4.15234375,
+      "learning_rate": 4.7638425617078054e-05,
+      "loss": 1.2651,
+      "step": 7859
+    },
+    {
+      "epoch": 7.255942764828064,
+      "grad_norm": 3.517578125,
+      "learning_rate": 4.7631754503002e-05,
+      "loss": 1.098,
+      "step": 7860
+    },
+    {
+      "epoch": 7.25686591276252,
+      "grad_norm": 3.638671875,
+      "learning_rate": 4.762508338892595e-05,
+      "loss": 1.5139,
+      "step": 7861
+    },
+    {
+      "epoch": 7.257789060696977,
+      "grad_norm": 2.98828125,
+      "learning_rate": 4.76184122748499e-05,
+      "loss": 1.3223,
+      "step": 7862
+    },
+    {
+      "epoch": 7.258712208631433,
+      "grad_norm": 3.283203125,
+      "learning_rate": 4.7611741160773856e-05,
+      "loss": 1.4551,
+      "step": 7863
+    },
+    {
+      "epoch": 7.25963535656589,
+      "grad_norm": 4.1015625,
+      "learning_rate": 4.76050700466978e-05,
+      "loss": 1.4768,
+      "step": 7864
+    },
+    {
+      "epoch": 7.260558504500346,
+      "grad_norm": 4.05078125,
+      "learning_rate": 4.759839893262175e-05,
+      "loss": 1.6082,
+      "step": 7865
+    },
+    {
+      "epoch": 7.261481652434803,
+      "grad_norm": 3.802734375,
+      "learning_rate": 4.75917278185457e-05,
+      "loss": 0.9702,
+      "step": 7866
+    },
+    {
+      "epoch": 7.262404800369259,
+      "grad_norm": 3.591796875,
+      "learning_rate": 4.758505670446965e-05,
+      "loss": 1.1841,
+      "step": 7867
+    },
+    {
+      "epoch": 7.263327948303716,
+      "grad_norm": 3.03515625,
+      "learning_rate": 4.7578385590393596e-05,
+      "loss": 1.5133,
+      "step": 7868
+    },
+    {
+      "epoch": 7.264251096238172,
+      "grad_norm": 3.73828125,
+      "learning_rate": 4.757171447631755e-05,
+      "loss": 1.344,
+      "step": 7869
+    },
+    {
+      "epoch": 7.265174244172629,
+      "grad_norm": 3.2890625,
+      "learning_rate": 4.7565043362241494e-05,
+      "loss": 1.4321,
+      "step": 7870
+    },
+    {
+      "epoch": 7.266097392107085,
+      "grad_norm": 3.474609375,
+      "learning_rate": 4.7558372248165446e-05,
+      "loss": 1.407,
+      "step": 7871
+    },
+    {
+      "epoch": 7.267020540041542,
+      "grad_norm": 3.119140625,
+      "learning_rate": 4.755170113408939e-05,
+      "loss": 1.5769,
+      "step": 7872
+    },
+    {
+      "epoch": 7.267943687975998,
+      "grad_norm": 4.1953125,
+      "learning_rate": 4.7545030020013344e-05,
+      "loss": 1.334,
+      "step": 7873
+    },
+    {
+      "epoch": 7.268866835910455,
+      "grad_norm": 3.61328125,
+      "learning_rate": 4.753835890593729e-05,
+      "loss": 1.396,
+      "step": 7874
+    },
+    {
+      "epoch": 7.269789983844911,
+      "grad_norm": 3.666015625,
+      "learning_rate": 4.753168779186125e-05,
+      "loss": 1.3064,
+      "step": 7875
+    },
+    {
+      "epoch": 7.270713131779368,
+      "grad_norm": 3.302734375,
+      "learning_rate": 4.7525016677785194e-05,
+      "loss": 1.1093,
+      "step": 7876
+    },
+    {
+      "epoch": 7.2716362797138245,
+      "grad_norm": 3.083984375,
+      "learning_rate": 4.7518345563709146e-05,
+      "loss": 1.2228,
+      "step": 7877
+    },
+    {
+      "epoch": 7.272559427648281,
+      "grad_norm": 3.572265625,
+      "learning_rate": 4.751167444963309e-05,
+      "loss": 1.5386,
+      "step": 7878
+    },
+    {
+      "epoch": 7.2734825755827375,
+      "grad_norm": 3.9765625,
+      "learning_rate": 4.7505003335557044e-05,
+      "loss": 1.3931,
+      "step": 7879
+    },
+    {
+      "epoch": 7.2744057235171935,
+      "grad_norm": 3.013671875,
+      "learning_rate": 4.749833222148099e-05,
+      "loss": 1.3853,
+      "step": 7880
+    },
+    {
+      "epoch": 7.27532887145165,
+      "grad_norm": 4.02734375,
+      "learning_rate": 4.749166110740494e-05,
+      "loss": 1.5117,
+      "step": 7881
+    },
+    {
+      "epoch": 7.276252019386106,
+      "grad_norm": 3.419921875,
+      "learning_rate": 4.7484989993328887e-05,
+      "loss": 1.3044,
+      "step": 7882
+    },
+    {
+      "epoch": 7.277175167320563,
+      "grad_norm": 4.06640625,
+      "learning_rate": 4.747831887925284e-05,
+      "loss": 1.5144,
+      "step": 7883
+    },
+    {
+      "epoch": 7.278098315255019,
+      "grad_norm": 3.125,
+      "learning_rate": 4.7471647765176784e-05,
+      "loss": 1.3518,
+      "step": 7884
+    },
+    {
+      "epoch": 7.279021463189476,
+      "grad_norm": 4.01953125,
+      "learning_rate": 4.7464976651100736e-05,
+      "loss": 1.3054,
+      "step": 7885
+    },
+    {
+      "epoch": 7.279944611123932,
+      "grad_norm": 3.603515625,
+      "learning_rate": 4.745830553702468e-05,
+      "loss": 1.3738,
+      "step": 7886
+    },
+    {
+      "epoch": 7.280867759058389,
+      "grad_norm": 3.58203125,
+      "learning_rate": 4.7451634422948634e-05,
+      "loss": 1.2354,
+      "step": 7887
+    },
+    {
+      "epoch": 7.281790906992845,
+      "grad_norm": 3.521484375,
+      "learning_rate": 4.7444963308872586e-05,
+      "loss": 1.4009,
+      "step": 7888
+    },
+    {
+      "epoch": 7.282714054927302,
+      "grad_norm": 3.869140625,
+      "learning_rate": 4.743829219479654e-05,
+      "loss": 1.1392,
+      "step": 7889
+    },
+    {
+      "epoch": 7.283637202861758,
+      "grad_norm": 3.564453125,
+      "learning_rate": 4.7431621080720484e-05,
+      "loss": 1.4109,
+      "step": 7890
+    },
+    {
+      "epoch": 7.284560350796215,
+      "grad_norm": 3.73046875,
+      "learning_rate": 4.7424949966644436e-05,
+      "loss": 1.3722,
+      "step": 7891
+    },
+    {
+      "epoch": 7.285483498730672,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.741827885256838e-05,
+      "loss": 1.3879,
+      "step": 7892
+    },
+    {
+      "epoch": 7.286406646665128,
+      "grad_norm": 3.302734375,
+      "learning_rate": 4.7411607738492334e-05,
+      "loss": 1.6631,
+      "step": 7893
+    },
+    {
+      "epoch": 7.287329794599585,
+      "grad_norm": 3.861328125,
+      "learning_rate": 4.740493662441628e-05,
+      "loss": 1.408,
+      "step": 7894
+    },
+    {
+      "epoch": 7.288252942534041,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.739826551034023e-05,
+      "loss": 1.1669,
+      "step": 7895
+    },
+    {
+      "epoch": 7.289176090468498,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.739159439626418e-05,
+      "loss": 1.5347,
+      "step": 7896
+    },
+    {
+      "epoch": 7.290099238402954,
+      "grad_norm": 3.3828125,
+      "learning_rate": 4.738492328218813e-05,
+      "loss": 1.4119,
+      "step": 7897
+    },
+    {
+      "epoch": 7.291022386337411,
+      "grad_norm": 3.908203125,
+      "learning_rate": 4.7378252168112074e-05,
+      "loss": 1.2473,
+      "step": 7898
+    },
+    {
+      "epoch": 7.291945534271867,
+      "grad_norm": 3.13671875,
+      "learning_rate": 4.737158105403603e-05,
+      "loss": 1.3035,
+      "step": 7899
+    },
+    {
+      "epoch": 7.292868682206324,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.736490993995998e-05,
+      "loss": 1.1603,
+      "step": 7900
+    },
+    {
+      "epoch": 7.29379183014078,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.7358238825883924e-05,
+      "loss": 1.4028,
+      "step": 7901
+    },
+    {
+      "epoch": 7.294714978075237,
+      "grad_norm": 3.220703125,
+      "learning_rate": 4.7351567711807877e-05,
+      "loss": 1.3865,
+      "step": 7902
+    },
+    {
+      "epoch": 7.295638126009693,
+      "grad_norm": 3.158203125,
+      "learning_rate": 4.734489659773182e-05,
+      "loss": 1.3494,
+      "step": 7903
+    },
+    {
+      "epoch": 7.29656127394415,
+      "grad_norm": 3.8671875,
+      "learning_rate": 4.7338225483655774e-05,
+      "loss": 1.2937,
+      "step": 7904
+    },
+    {
+      "epoch": 7.297484421878606,
+      "grad_norm": 3.3671875,
+      "learning_rate": 4.733155436957972e-05,
+      "loss": 0.9621,
+      "step": 7905
+    },
+    {
+      "epoch": 7.298407569813063,
+      "grad_norm": 3.681640625,
+      "learning_rate": 4.732488325550367e-05,
+      "loss": 1.222,
+      "step": 7906
+    },
+    {
+      "epoch": 7.299330717747519,
+      "grad_norm": 4.0390625,
+      "learning_rate": 4.731821214142762e-05,
+      "loss": 1.261,
+      "step": 7907
+    },
+    {
+      "epoch": 7.3002538656819755,
+      "grad_norm": 3.333984375,
+      "learning_rate": 4.731154102735157e-05,
+      "loss": 1.1233,
+      "step": 7908
+    },
+    {
+      "epoch": 7.301177013616432,
+      "grad_norm": 3.39453125,
+      "learning_rate": 4.7304869913275515e-05,
+      "loss": 1.5396,
+      "step": 7909
+    },
+    {
+      "epoch": 7.3021001615508885,
+      "grad_norm": 3.443359375,
+      "learning_rate": 4.729819879919947e-05,
+      "loss": 1.3865,
+      "step": 7910
+    },
+    {
+      "epoch": 7.303023309485345,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.729152768512341e-05,
+      "loss": 1.3923,
+      "step": 7911
+    },
+    {
+      "epoch": 7.303946457419801,
+      "grad_norm": 3.60546875,
+      "learning_rate": 4.7284856571047365e-05,
+      "loss": 1.4558,
+      "step": 7912
+    },
+    {
+      "epoch": 7.304869605354258,
+      "grad_norm": 3.13671875,
+      "learning_rate": 4.727818545697132e-05,
+      "loss": 1.604,
+      "step": 7913
+    },
+    {
+      "epoch": 7.305792753288714,
+      "grad_norm": 3.31640625,
+      "learning_rate": 4.727151434289527e-05,
+      "loss": 1.2659,
+      "step": 7914
+    },
+    {
+      "epoch": 7.306715901223171,
+      "grad_norm": 3.263671875,
+      "learning_rate": 4.7264843228819215e-05,
+      "loss": 1.0531,
+      "step": 7915
+    },
+    {
+      "epoch": 7.307639049157627,
+      "grad_norm": 3.099609375,
+      "learning_rate": 4.725817211474317e-05,
+      "loss": 1.2532,
+      "step": 7916
+    },
+    {
+      "epoch": 7.308562197092084,
+      "grad_norm": 3.72265625,
+      "learning_rate": 4.725150100066711e-05,
+      "loss": 1.3435,
+      "step": 7917
+    },
+    {
+      "epoch": 7.30948534502654,
+      "grad_norm": 3.435546875,
+      "learning_rate": 4.7244829886591064e-05,
+      "loss": 1.1597,
+      "step": 7918
+    },
+    {
+      "epoch": 7.310408492960997,
+      "grad_norm": 3.138671875,
+      "learning_rate": 4.723815877251501e-05,
+      "loss": 1.3448,
+      "step": 7919
+    },
+    {
+      "epoch": 7.311331640895453,
+      "grad_norm": 2.876953125,
+      "learning_rate": 4.723148765843896e-05,
+      "loss": 1.0861,
+      "step": 7920
+    },
+    {
+      "epoch": 7.31225478882991,
+      "grad_norm": 3.220703125,
+      "learning_rate": 4.722481654436291e-05,
+      "loss": 1.2949,
+      "step": 7921
+    },
+    {
+      "epoch": 7.313177936764366,
+      "grad_norm": 3.76171875,
+      "learning_rate": 4.721814543028686e-05,
+      "loss": 1.2324,
+      "step": 7922
+    },
+    {
+      "epoch": 7.314101084698823,
+      "grad_norm": 3.873046875,
+      "learning_rate": 4.7211474316210805e-05,
+      "loss": 1.3853,
+      "step": 7923
+    },
+    {
+      "epoch": 7.315024232633279,
+      "grad_norm": 3.634765625,
+      "learning_rate": 4.720480320213476e-05,
+      "loss": 1.3328,
+      "step": 7924
+    },
+    {
+      "epoch": 7.315947380567736,
+      "grad_norm": 3.576171875,
+      "learning_rate": 4.719813208805871e-05,
+      "loss": 1.407,
+      "step": 7925
+    },
+    {
+      "epoch": 7.316870528502193,
+      "grad_norm": 3.54296875,
+      "learning_rate": 4.719146097398266e-05,
+      "loss": 1.3223,
+      "step": 7926
+    },
+    {
+      "epoch": 7.317793676436649,
+      "grad_norm": 3.533203125,
+      "learning_rate": 4.718478985990661e-05,
+      "loss": 1.364,
+      "step": 7927
+    },
+    {
+      "epoch": 7.318716824371106,
+      "grad_norm": 3.94140625,
+      "learning_rate": 4.717811874583056e-05,
+      "loss": 1.1013,
+      "step": 7928
+    },
+    {
+      "epoch": 7.319639972305562,
+      "grad_norm": 4.23828125,
+      "learning_rate": 4.7171447631754505e-05,
+      "loss": 1.2047,
+      "step": 7929
+    },
+    {
+      "epoch": 7.320563120240019,
+      "grad_norm": 3.63671875,
+      "learning_rate": 4.716477651767846e-05,
+      "loss": 1.4558,
+      "step": 7930
+    },
+    {
+      "epoch": 7.321486268174475,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.71581054036024e-05,
+      "loss": 1.3508,
+      "step": 7931
+    },
+    {
+      "epoch": 7.322409416108932,
+      "grad_norm": 3.27734375,
+      "learning_rate": 4.7151434289526355e-05,
+      "loss": 1.3129,
+      "step": 7932
+    },
+    {
+      "epoch": 7.323332564043388,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.71447631754503e-05,
+      "loss": 1.0449,
+      "step": 7933
+    },
+    {
+      "epoch": 7.324255711977845,
+      "grad_norm": 3.884765625,
+      "learning_rate": 4.713809206137425e-05,
+      "loss": 1.3254,
+      "step": 7934
+    },
+    {
+      "epoch": 7.325178859912301,
+      "grad_norm": 3.298828125,
+      "learning_rate": 4.71314209472982e-05,
+      "loss": 0.9703,
+      "step": 7935
+    },
+    {
+      "epoch": 7.326102007846758,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.712474983322215e-05,
+      "loss": 0.9668,
+      "step": 7936
+    },
+    {
+      "epoch": 7.327025155781214,
+      "grad_norm": 3.1015625,
+      "learning_rate": 4.7118078719146095e-05,
+      "loss": 1.1742,
+      "step": 7937
+    },
+    {
+      "epoch": 7.3279483037156705,
+      "grad_norm": 3.443359375,
+      "learning_rate": 4.7111407605070054e-05,
+      "loss": 1.1512,
+      "step": 7938
+    },
+    {
+      "epoch": 7.3288714516501265,
+      "grad_norm": 3.064453125,
+      "learning_rate": 4.7104736490994e-05,
+      "loss": 1.5464,
+      "step": 7939
+    },
+    {
+      "epoch": 7.3297945995845835,
+      "grad_norm": 3.85546875,
+      "learning_rate": 4.709806537691795e-05,
+      "loss": 1.1206,
+      "step": 7940
+    },
+    {
+      "epoch": 7.33071774751904,
+      "grad_norm": 3.2265625,
+      "learning_rate": 4.70913942628419e-05,
+      "loss": 1.1753,
+      "step": 7941
+    },
+    {
+      "epoch": 7.331640895453496,
+      "grad_norm": 3.294921875,
+      "learning_rate": 4.708472314876585e-05,
+      "loss": 1.3472,
+      "step": 7942
+    },
+    {
+      "epoch": 7.332564043387953,
+      "grad_norm": 3.505859375,
+      "learning_rate": 4.7078052034689795e-05,
+      "loss": 1.6306,
+      "step": 7943
+    },
+    {
+      "epoch": 7.333487191322409,
+      "grad_norm": 2.97265625,
+      "learning_rate": 4.707138092061375e-05,
+      "loss": 1.049,
+      "step": 7944
+    },
+    {
+      "epoch": 7.334410339256866,
+      "grad_norm": 3.248046875,
+      "learning_rate": 4.706470980653769e-05,
+      "loss": 1.2856,
+      "step": 7945
+    },
+    {
+      "epoch": 7.335333487191322,
+      "grad_norm": 3.208984375,
+      "learning_rate": 4.7058038692461645e-05,
+      "loss": 1.4072,
+      "step": 7946
+    },
+    {
+      "epoch": 7.336256635125779,
+      "grad_norm": 3.791015625,
+      "learning_rate": 4.705136757838559e-05,
+      "loss": 1.5212,
+      "step": 7947
+    },
+    {
+      "epoch": 7.337179783060235,
+      "grad_norm": 3.32421875,
+      "learning_rate": 4.704469646430954e-05,
+      "loss": 1.4082,
+      "step": 7948
+    },
+    {
+      "epoch": 7.338102930994692,
+      "grad_norm": 3.962890625,
+      "learning_rate": 4.703802535023349e-05,
+      "loss": 1.2954,
+      "step": 7949
+    },
+    {
+      "epoch": 7.339026078929148,
+      "grad_norm": 3.658203125,
+      "learning_rate": 4.703135423615744e-05,
+      "loss": 1.4614,
+      "step": 7950
+    },
+    {
+      "epoch": 7.339949226863605,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.702468312208139e-05,
+      "loss": 1.2654,
+      "step": 7951
+    },
+    {
+      "epoch": 7.340872374798061,
+      "grad_norm": 3.75,
+      "learning_rate": 4.701801200800534e-05,
+      "loss": 1.1521,
+      "step": 7952
+    },
+    {
+      "epoch": 7.341795522732518,
+      "grad_norm": 3.599609375,
+      "learning_rate": 4.701134089392929e-05,
+      "loss": 1.3889,
+      "step": 7953
+    },
+    {
+      "epoch": 7.342718670666974,
+      "grad_norm": 3.435546875,
+      "learning_rate": 4.7004669779853235e-05,
+      "loss": 1.3523,
+      "step": 7954
+    },
+    {
+      "epoch": 7.343641818601431,
+      "grad_norm": 3.6015625,
+      "learning_rate": 4.699799866577719e-05,
+      "loss": 1.1603,
+      "step": 7955
+    },
+    {
+      "epoch": 7.344564966535887,
+      "grad_norm": 3.072265625,
+      "learning_rate": 4.699132755170113e-05,
+      "loss": 1.6321,
+      "step": 7956
+    },
+    {
+      "epoch": 7.345488114470344,
+      "grad_norm": 4.08984375,
+      "learning_rate": 4.6984656437625085e-05,
+      "loss": 1.5161,
+      "step": 7957
+    },
+    {
+      "epoch": 7.3464112624048,
+      "grad_norm": 3.978515625,
+      "learning_rate": 4.697798532354904e-05,
+      "loss": 1.4626,
+      "step": 7958
+    },
+    {
+      "epoch": 7.347334410339257,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.697131420947298e-05,
+      "loss": 1.5522,
+      "step": 7959
+    },
+    {
+      "epoch": 7.348257558273714,
+      "grad_norm": 3.58203125,
+      "learning_rate": 4.6964643095396935e-05,
+      "loss": 1.4897,
+      "step": 7960
+    },
+    {
+      "epoch": 7.34918070620817,
+      "grad_norm": 3.6484375,
+      "learning_rate": 4.695797198132088e-05,
+      "loss": 1.311,
+      "step": 7961
+    },
+    {
+      "epoch": 7.350103854142627,
+      "grad_norm": 4.39453125,
+      "learning_rate": 4.695130086724483e-05,
+      "loss": 1.2843,
+      "step": 7962
+    },
+    {
+      "epoch": 7.351027002077083,
+      "grad_norm": 3.48828125,
+      "learning_rate": 4.6944629753168785e-05,
+      "loss": 1.1947,
+      "step": 7963
+    },
+    {
+      "epoch": 7.35195015001154,
+      "grad_norm": 3.115234375,
+      "learning_rate": 4.693795863909273e-05,
+      "loss": 1.2312,
+      "step": 7964
+    },
+    {
+      "epoch": 7.352873297945996,
+      "grad_norm": 3.638671875,
+      "learning_rate": 4.693128752501668e-05,
+      "loss": 1.3177,
+      "step": 7965
+    },
+    {
+      "epoch": 7.353796445880453,
+      "grad_norm": 2.75390625,
+      "learning_rate": 4.692461641094063e-05,
+      "loss": 0.9474,
+      "step": 7966
+    },
+    {
+      "epoch": 7.354719593814909,
+      "grad_norm": 3.591796875,
+      "learning_rate": 4.691794529686458e-05,
+      "loss": 1.2974,
+      "step": 7967
+    },
+    {
+      "epoch": 7.3556427417493655,
+      "grad_norm": 3.22265625,
+      "learning_rate": 4.6911274182788526e-05,
+      "loss": 1.563,
+      "step": 7968
+    },
+    {
+      "epoch": 7.3565658896838215,
+      "grad_norm": 3.6328125,
+      "learning_rate": 4.690460306871248e-05,
+      "loss": 1.365,
+      "step": 7969
+    },
+    {
+      "epoch": 7.357489037618278,
+      "grad_norm": 3.3203125,
+      "learning_rate": 4.689793195463642e-05,
+      "loss": 1.553,
+      "step": 7970
+    },
+    {
+      "epoch": 7.3584121855527345,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.6891260840560376e-05,
+      "loss": 1.3049,
+      "step": 7971
+    },
+    {
+      "epoch": 7.359335333487191,
+      "grad_norm": 3.005859375,
+      "learning_rate": 4.688458972648432e-05,
+      "loss": 1.2445,
+      "step": 7972
+    },
+    {
+      "epoch": 7.360258481421647,
+      "grad_norm": 3.173828125,
+      "learning_rate": 4.687791861240827e-05,
+      "loss": 1.4915,
+      "step": 7973
+    },
+    {
+      "epoch": 7.361181629356104,
+      "grad_norm": 3.861328125,
+      "learning_rate": 4.687124749833222e-05,
+      "loss": 1.3774,
+      "step": 7974
+    },
+    {
+      "epoch": 7.362104777290561,
+      "grad_norm": 3.251953125,
+      "learning_rate": 4.686457638425617e-05,
+      "loss": 1.3813,
+      "step": 7975
+    },
+    {
+      "epoch": 7.363027925225017,
+      "grad_norm": 3.013671875,
+      "learning_rate": 4.685790527018012e-05,
+      "loss": 1.3992,
+      "step": 7976
+    },
+    {
+      "epoch": 7.363951073159474,
+      "grad_norm": 3.806640625,
+      "learning_rate": 4.6851234156104075e-05,
+      "loss": 0.8969,
+      "step": 7977
+    },
+    {
+      "epoch": 7.36487422109393,
+      "grad_norm": 3.455078125,
+      "learning_rate": 4.684456304202802e-05,
+      "loss": 1.3188,
+      "step": 7978
+    },
+    {
+      "epoch": 7.365797369028387,
+      "grad_norm": 3.544921875,
+      "learning_rate": 4.683789192795197e-05,
+      "loss": 1.6116,
+      "step": 7979
+    },
+    {
+      "epoch": 7.366720516962843,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.683122081387592e-05,
+      "loss": 1.2244,
+      "step": 7980
+    },
+    {
+      "epoch": 7.3676436648973,
+      "grad_norm": 3.3359375,
+      "learning_rate": 4.682454969979987e-05,
+      "loss": 1.1219,
+      "step": 7981
+    },
+    {
+      "epoch": 7.368566812831756,
+      "grad_norm": 4.2265625,
+      "learning_rate": 4.6817878585723816e-05,
+      "loss": 1.3196,
+      "step": 7982
+    },
+    {
+      "epoch": 7.369489960766213,
+      "grad_norm": 3.1484375,
+      "learning_rate": 4.681120747164777e-05,
+      "loss": 1.4231,
+      "step": 7983
+    },
+    {
+      "epoch": 7.370413108700669,
+      "grad_norm": 3.681640625,
+      "learning_rate": 4.6804536357571714e-05,
+      "loss": 1.2942,
+      "step": 7984
+    },
+    {
+      "epoch": 7.371336256635126,
+      "grad_norm": 4.10546875,
+      "learning_rate": 4.6797865243495666e-05,
+      "loss": 1.6016,
+      "step": 7985
+    },
+    {
+      "epoch": 7.372259404569582,
+      "grad_norm": 3.294921875,
+      "learning_rate": 4.679119412941961e-05,
+      "loss": 1.4988,
+      "step": 7986
+    },
+    {
+      "epoch": 7.373182552504039,
+      "grad_norm": 3.978515625,
+      "learning_rate": 4.6784523015343563e-05,
+      "loss": 1.5125,
+      "step": 7987
+    },
+    {
+      "epoch": 7.374105700438495,
+      "grad_norm": 4.03515625,
+      "learning_rate": 4.6777851901267516e-05,
+      "loss": 1.0538,
+      "step": 7988
+    },
+    {
+      "epoch": 7.375028848372952,
+      "grad_norm": 3.216796875,
+      "learning_rate": 4.677118078719147e-05,
+      "loss": 1.2454,
+      "step": 7989
+    },
+    {
+      "epoch": 7.375951996307409,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.676450967311541e-05,
+      "loss": 1.1202,
+      "step": 7990
+    },
+    {
+      "epoch": 7.376875144241865,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.6757838559039365e-05,
+      "loss": 1.3958,
+      "step": 7991
+    },
+    {
+      "epoch": 7.377798292176322,
+      "grad_norm": 3.31640625,
+      "learning_rate": 4.675116744496331e-05,
+      "loss": 1.5513,
+      "step": 7992
+    },
+    {
+      "epoch": 7.378721440110778,
+      "grad_norm": 3.826171875,
+      "learning_rate": 4.674449633088726e-05,
+      "loss": 1.4011,
+      "step": 7993
+    },
+    {
+      "epoch": 7.379644588045235,
+      "grad_norm": 3.314453125,
+      "learning_rate": 4.673782521681121e-05,
+      "loss": 1.3047,
+      "step": 7994
+    },
+    {
+      "epoch": 7.380567735979691,
+      "grad_norm": 2.8203125,
+      "learning_rate": 4.673115410273516e-05,
+      "loss": 1.438,
+      "step": 7995
+    },
+    {
+      "epoch": 7.3814908839141475,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.6724482988659106e-05,
+      "loss": 1.3269,
+      "step": 7996
+    },
+    {
+      "epoch": 7.382414031848604,
+      "grad_norm": 3.181640625,
+      "learning_rate": 4.671781187458306e-05,
+      "loss": 1.5981,
+      "step": 7997
+    },
+    {
+      "epoch": 7.3833371797830605,
+      "grad_norm": 3.18359375,
+      "learning_rate": 4.6711140760507004e-05,
+      "loss": 0.9033,
+      "step": 7998
+    },
+    {
+      "epoch": 7.3842603277175165,
+      "grad_norm": 4.48046875,
+      "learning_rate": 4.6704469646430956e-05,
+      "loss": 1.3599,
+      "step": 7999
+    },
+    {
+      "epoch": 7.385183475651973,
+      "grad_norm": 4.08203125,
+      "learning_rate": 4.66977985323549e-05,
+      "loss": 1.3546,
+      "step": 8000
+    },
+    {
+      "epoch": 7.385183475651973,
+      "eval_loss": 0.87890625,
+      "eval_runtime": 23.3429,
+      "eval_samples_per_second": 976.783,
+      "eval_steps_per_second": 61.089,
+      "step": 8000
+    },
+    {
+      "epoch": 7.386106623586429,
+      "grad_norm": 3.947265625,
+      "learning_rate": 4.669112741827886e-05,
+      "loss": 1.6038,
+      "step": 8001
+    },
+    {
+      "epoch": 7.387029771520886,
+      "grad_norm": 3.6796875,
+      "learning_rate": 4.6684456304202806e-05,
+      "loss": 1.4983,
+      "step": 8002
+    },
+    {
+      "epoch": 7.387952919455342,
+      "grad_norm": 3.705078125,
+      "learning_rate": 4.667778519012676e-05,
+      "loss": 1.321,
+      "step": 8003
+    },
+    {
+      "epoch": 7.388876067389799,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.6671114076050703e-05,
+      "loss": 1.3933,
+      "step": 8004
+    },
+    {
+      "epoch": 7.389799215324255,
+      "grad_norm": 3.58203125,
+      "learning_rate": 4.6664442961974656e-05,
+      "loss": 1.3152,
+      "step": 8005
+    },
+    {
+      "epoch": 7.390722363258712,
+      "grad_norm": 3.103515625,
+      "learning_rate": 4.66577718478986e-05,
+      "loss": 1.3872,
+      "step": 8006
+    },
+    {
+      "epoch": 7.391645511193168,
+      "grad_norm": 4.03515625,
+      "learning_rate": 4.665110073382255e-05,
+      "loss": 1.2938,
+      "step": 8007
+    },
+    {
+      "epoch": 7.392568659127625,
+      "grad_norm": 3.5,
+      "learning_rate": 4.66444296197465e-05,
+      "loss": 1.3412,
+      "step": 8008
+    },
+    {
+      "epoch": 7.393491807062082,
+      "grad_norm": 3.892578125,
+      "learning_rate": 4.663775850567045e-05,
+      "loss": 0.7166,
+      "step": 8009
+    },
+    {
+      "epoch": 7.394414954996538,
+      "grad_norm": 3.373046875,
+      "learning_rate": 4.6631087391594396e-05,
+      "loss": 1.1545,
+      "step": 8010
+    },
+    {
+      "epoch": 7.395338102930995,
+      "grad_norm": 3.173828125,
+      "learning_rate": 4.662441627751835e-05,
+      "loss": 0.9023,
+      "step": 8011
+    },
+    {
+      "epoch": 7.396261250865451,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.6617745163442294e-05,
+      "loss": 1.3215,
+      "step": 8012
+    },
+    {
+      "epoch": 7.397184398799908,
+      "grad_norm": 3.9609375,
+      "learning_rate": 4.6611074049366246e-05,
+      "loss": 1.2546,
+      "step": 8013
+    },
+    {
+      "epoch": 7.398107546734364,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.66044029352902e-05,
+      "loss": 1.4343,
+      "step": 8014
+    },
+    {
+      "epoch": 7.399030694668821,
+      "grad_norm": 3.03515625,
+      "learning_rate": 4.6597731821214144e-05,
+      "loss": 1.2384,
+      "step": 8015
+    },
+    {
+      "epoch": 7.399953842603277,
+      "grad_norm": 3.806640625,
+      "learning_rate": 4.6591060707138096e-05,
+      "loss": 1.3755,
+      "step": 8016
+    },
+    {
+      "epoch": 7.400876990537734,
+      "grad_norm": 3.76953125,
+      "learning_rate": 4.658438959306204e-05,
+      "loss": 1.2102,
+      "step": 8017
+    },
+    {
+      "epoch": 7.40180013847219,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.6577718478985994e-05,
+      "loss": 1.2986,
+      "step": 8018
+    },
+    {
+      "epoch": 7.402723286406647,
+      "grad_norm": 3.87109375,
+      "learning_rate": 4.657104736490994e-05,
+      "loss": 1.2634,
+      "step": 8019
+    },
+    {
+      "epoch": 7.403646434341103,
+      "grad_norm": 2.91796875,
+      "learning_rate": 4.656437625083389e-05,
+      "loss": 1.291,
+      "step": 8020
+    },
+    {
+      "epoch": 7.40456958227556,
+      "grad_norm": 3.3671875,
+      "learning_rate": 4.655770513675784e-05,
+      "loss": 1.218,
+      "step": 8021
+    },
+    {
+      "epoch": 7.405492730210016,
+      "grad_norm": 3.400390625,
+      "learning_rate": 4.655103402268179e-05,
+      "loss": 1.2517,
+      "step": 8022
+    },
+    {
+      "epoch": 7.406415878144473,
+      "grad_norm": 3.431640625,
+      "learning_rate": 4.6544362908605734e-05,
+      "loss": 1.1848,
+      "step": 8023
+    },
+    {
+      "epoch": 7.40733902607893,
+      "grad_norm": 3.462890625,
+      "learning_rate": 4.653769179452969e-05,
+      "loss": 1.5327,
+      "step": 8024
+    },
+    {
+      "epoch": 7.408262174013386,
+      "grad_norm": 3.666015625,
+      "learning_rate": 4.653102068045363e-05,
+      "loss": 1.2649,
+      "step": 8025
+    },
+    {
+      "epoch": 7.4091853219478425,
+      "grad_norm": 3.189453125,
+      "learning_rate": 4.652434956637759e-05,
+      "loss": 1.0297,
+      "step": 8026
+    },
+    {
+      "epoch": 7.4101084698822985,
+      "grad_norm": 3.11328125,
+      "learning_rate": 4.6517678452301536e-05,
+      "loss": 1.2773,
+      "step": 8027
+    },
+    {
+      "epoch": 7.4110316178167555,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.651100733822549e-05,
+      "loss": 1.4441,
+      "step": 8028
+    },
+    {
+      "epoch": 7.4119547657512115,
+      "grad_norm": 3.8515625,
+      "learning_rate": 4.6504336224149434e-05,
+      "loss": 1.434,
+      "step": 8029
+    },
+    {
+      "epoch": 7.412877913685668,
+      "grad_norm": 3.3984375,
+      "learning_rate": 4.6497665110073386e-05,
+      "loss": 1.4451,
+      "step": 8030
+    },
+    {
+      "epoch": 7.413801061620124,
+      "grad_norm": 4.0703125,
+      "learning_rate": 4.649099399599733e-05,
+      "loss": 1.248,
+      "step": 8031
+    },
+    {
+      "epoch": 7.414724209554581,
+      "grad_norm": 3.1796875,
+      "learning_rate": 4.6484322881921284e-05,
+      "loss": 1.2601,
+      "step": 8032
+    },
+    {
+      "epoch": 7.415647357489037,
+      "grad_norm": 3.787109375,
+      "learning_rate": 4.647765176784523e-05,
+      "loss": 1.4705,
+      "step": 8033
+    },
+    {
+      "epoch": 7.416570505423494,
+      "grad_norm": 3.64453125,
+      "learning_rate": 4.647098065376918e-05,
+      "loss": 1.2179,
+      "step": 8034
+    },
+    {
+      "epoch": 7.41749365335795,
+      "grad_norm": 3.10546875,
+      "learning_rate": 4.646430953969313e-05,
+      "loss": 1.0914,
+      "step": 8035
+    },
+    {
+      "epoch": 7.418416801292407,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.645763842561708e-05,
+      "loss": 1.3386,
+      "step": 8036
+    },
+    {
+      "epoch": 7.419339949226863,
+      "grad_norm": 3.837890625,
+      "learning_rate": 4.6450967311541025e-05,
+      "loss": 1.2954,
+      "step": 8037
+    },
+    {
+      "epoch": 7.42026309716132,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.6444296197464984e-05,
+      "loss": 1.3295,
+      "step": 8038
+    },
+    {
+      "epoch": 7.421186245095777,
+      "grad_norm": 3.580078125,
+      "learning_rate": 4.643762508338893e-05,
+      "loss": 1.3945,
+      "step": 8039
+    },
+    {
+      "epoch": 7.422109393030233,
+      "grad_norm": 3.291015625,
+      "learning_rate": 4.643095396931288e-05,
+      "loss": 1.3936,
+      "step": 8040
+    },
+    {
+      "epoch": 7.42303254096469,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.642428285523683e-05,
+      "loss": 1.2672,
+      "step": 8041
+    },
+    {
+      "epoch": 7.423955688899146,
+      "grad_norm": 3.3046875,
+      "learning_rate": 4.641761174116078e-05,
+      "loss": 1.24,
+      "step": 8042
+    },
+    {
+      "epoch": 7.424878836833603,
+      "grad_norm": 2.96484375,
+      "learning_rate": 4.6410940627084724e-05,
+      "loss": 0.976,
+      "step": 8043
+    },
+    {
+      "epoch": 7.425801984768059,
+      "grad_norm": 3.732421875,
+      "learning_rate": 4.6404269513008677e-05,
+      "loss": 1.5088,
+      "step": 8044
+    },
+    {
+      "epoch": 7.426725132702516,
+      "grad_norm": 3.1953125,
+      "learning_rate": 4.639759839893262e-05,
+      "loss": 1.4777,
+      "step": 8045
+    },
+    {
+      "epoch": 7.427648280636972,
+      "grad_norm": 4.0078125,
+      "learning_rate": 4.6390927284856574e-05,
+      "loss": 1.3643,
+      "step": 8046
+    },
+    {
+      "epoch": 7.428571428571429,
+      "grad_norm": 3.900390625,
+      "learning_rate": 4.638425617078052e-05,
+      "loss": 1.3743,
+      "step": 8047
+    },
+    {
+      "epoch": 7.429494576505885,
+      "grad_norm": 4.20703125,
+      "learning_rate": 4.637758505670447e-05,
+      "loss": 1.2788,
+      "step": 8048
+    },
+    {
+      "epoch": 7.430417724440342,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.637091394262842e-05,
+      "loss": 1.4448,
+      "step": 8049
+    },
+    {
+      "epoch": 7.431340872374798,
+      "grad_norm": 3.6796875,
+      "learning_rate": 4.636424282855237e-05,
+      "loss": 1.5208,
+      "step": 8050
+    },
+    {
+      "epoch": 7.432264020309255,
+      "grad_norm": 3.990234375,
+      "learning_rate": 4.635757171447632e-05,
+      "loss": 1.006,
+      "step": 8051
+    },
+    {
+      "epoch": 7.433187168243711,
+      "grad_norm": 3.62109375,
+      "learning_rate": 4.6350900600400274e-05,
+      "loss": 1.0548,
+      "step": 8052
+    },
+    {
+      "epoch": 7.434110316178168,
+      "grad_norm": 3.1640625,
+      "learning_rate": 4.634422948632422e-05,
+      "loss": 1.0959,
+      "step": 8053
+    },
+    {
+      "epoch": 7.435033464112624,
+      "grad_norm": 2.994140625,
+      "learning_rate": 4.633755837224817e-05,
+      "loss": 1.248,
+      "step": 8054
+    },
+    {
+      "epoch": 7.435956612047081,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.633088725817212e-05,
+      "loss": 1.2937,
+      "step": 8055
+    },
+    {
+      "epoch": 7.436879759981537,
+      "grad_norm": 3.705078125,
+      "learning_rate": 4.632421614409607e-05,
+      "loss": 1.2701,
+      "step": 8056
+    },
+    {
+      "epoch": 7.4378029079159935,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.6317545030020015e-05,
+      "loss": 1.522,
+      "step": 8057
+    },
+    {
+      "epoch": 7.43872605585045,
+      "grad_norm": 3.990234375,
+      "learning_rate": 4.631087391594397e-05,
+      "loss": 1.1815,
+      "step": 8058
+    },
+    {
+      "epoch": 7.4396492037849065,
+      "grad_norm": 2.96484375,
+      "learning_rate": 4.630420280186791e-05,
+      "loss": 1.1989,
+      "step": 8059
+    },
+    {
+      "epoch": 7.440572351719363,
+      "grad_norm": 3.439453125,
+      "learning_rate": 4.6297531687791864e-05,
+      "loss": 1.4683,
+      "step": 8060
+    },
+    {
+      "epoch": 7.441495499653819,
+      "grad_norm": 3.46484375,
+      "learning_rate": 4.629086057371581e-05,
+      "loss": 1.1423,
+      "step": 8061
+    },
+    {
+      "epoch": 7.442418647588276,
+      "grad_norm": 3.35546875,
+      "learning_rate": 4.628418945963976e-05,
+      "loss": 1.2405,
+      "step": 8062
+    },
+    {
+      "epoch": 7.443341795522732,
+      "grad_norm": 3.08203125,
+      "learning_rate": 4.6277518345563714e-05,
+      "loss": 1.0422,
+      "step": 8063
+    },
+    {
+      "epoch": 7.444264943457189,
+      "grad_norm": 3.322265625,
+      "learning_rate": 4.627084723148766e-05,
+      "loss": 1.4856,
+      "step": 8064
+    },
+    {
+      "epoch": 7.445188091391645,
+      "grad_norm": 4.13671875,
+      "learning_rate": 4.626417611741161e-05,
+      "loss": 1.1487,
+      "step": 8065
+    },
+    {
+      "epoch": 7.446111239326102,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.625750500333556e-05,
+      "loss": 1.1002,
+      "step": 8066
+    },
+    {
+      "epoch": 7.447034387260558,
+      "grad_norm": 3.591796875,
+      "learning_rate": 4.625083388925951e-05,
+      "loss": 1.7256,
+      "step": 8067
+    },
+    {
+      "epoch": 7.447957535195015,
+      "grad_norm": 3.587890625,
+      "learning_rate": 4.6244162775183455e-05,
+      "loss": 1.2146,
+      "step": 8068
+    },
+    {
+      "epoch": 7.448880683129471,
+      "grad_norm": 3.181640625,
+      "learning_rate": 4.623749166110741e-05,
+      "loss": 1.3816,
+      "step": 8069
+    },
+    {
+      "epoch": 7.449803831063928,
+      "grad_norm": 3.3671875,
+      "learning_rate": 4.623082054703136e-05,
+      "loss": 1.0813,
+      "step": 8070
+    },
+    {
+      "epoch": 7.450726978998384,
+      "grad_norm": 3.46484375,
+      "learning_rate": 4.6224149432955305e-05,
+      "loss": 1.1875,
+      "step": 8071
+    },
+    {
+      "epoch": 7.451650126932841,
+      "grad_norm": 2.916015625,
+      "learning_rate": 4.621747831887926e-05,
+      "loss": 1.5752,
+      "step": 8072
+    },
+    {
+      "epoch": 7.452573274867298,
+      "grad_norm": 4.359375,
+      "learning_rate": 4.62108072048032e-05,
+      "loss": 1.2808,
+      "step": 8073
+    },
+    {
+      "epoch": 7.453496422801754,
+      "grad_norm": 3.287109375,
+      "learning_rate": 4.6204136090727155e-05,
+      "loss": 1.3623,
+      "step": 8074
+    },
+    {
+      "epoch": 7.454419570736211,
+      "grad_norm": 3.232421875,
+      "learning_rate": 4.61974649766511e-05,
+      "loss": 1.3831,
+      "step": 8075
+    },
+    {
+      "epoch": 7.455342718670667,
+      "grad_norm": 3.123046875,
+      "learning_rate": 4.619079386257505e-05,
+      "loss": 1.426,
+      "step": 8076
+    },
+    {
+      "epoch": 7.456265866605124,
+      "grad_norm": 3.86328125,
+      "learning_rate": 4.6184122748499005e-05,
+      "loss": 1.3679,
+      "step": 8077
+    },
+    {
+      "epoch": 7.45718901453958,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.617745163442295e-05,
+      "loss": 1.1832,
+      "step": 8078
+    },
+    {
+      "epoch": 7.458112162474037,
+      "grad_norm": 3.759765625,
+      "learning_rate": 4.61707805203469e-05,
+      "loss": 1.2729,
+      "step": 8079
+    },
+    {
+      "epoch": 7.459035310408493,
+      "grad_norm": 3.416015625,
+      "learning_rate": 4.616410940627085e-05,
+      "loss": 1.4836,
+      "step": 8080
+    },
+    {
+      "epoch": 7.45995845834295,
+      "grad_norm": 3.013671875,
+      "learning_rate": 4.61574382921948e-05,
+      "loss": 1.386,
+      "step": 8081
+    },
+    {
+      "epoch": 7.460881606277406,
+      "grad_norm": 3.80859375,
+      "learning_rate": 4.6150767178118745e-05,
+      "loss": 1.3105,
+      "step": 8082
+    },
+    {
+      "epoch": 7.461804754211863,
+      "grad_norm": 2.7734375,
+      "learning_rate": 4.61440960640427e-05,
+      "loss": 1.4856,
+      "step": 8083
+    },
+    {
+      "epoch": 7.462727902146319,
+      "grad_norm": 3.552734375,
+      "learning_rate": 4.613742494996664e-05,
+      "loss": 1.181,
+      "step": 8084
+    },
+    {
+      "epoch": 7.463651050080776,
+      "grad_norm": 3.36328125,
+      "learning_rate": 4.6130753835890595e-05,
+      "loss": 1.421,
+      "step": 8085
+    },
+    {
+      "epoch": 7.464574198015232,
+      "grad_norm": 3.89453125,
+      "learning_rate": 4.612408272181454e-05,
+      "loss": 1.3621,
+      "step": 8086
+    },
+    {
+      "epoch": 7.4654973459496885,
+      "grad_norm": 3.64453125,
+      "learning_rate": 4.611741160773849e-05,
+      "loss": 1.019,
+      "step": 8087
+    },
+    {
+      "epoch": 7.4664204938841445,
+      "grad_norm": 3.43359375,
+      "learning_rate": 4.6110740493662445e-05,
+      "loss": 1.3226,
+      "step": 8088
+    },
+    {
+      "epoch": 7.467343641818601,
+      "grad_norm": 3.126953125,
+      "learning_rate": 4.61040693795864e-05,
+      "loss": 1.4954,
+      "step": 8089
+    },
+    {
+      "epoch": 7.4682667897530575,
+      "grad_norm": 3.150390625,
+      "learning_rate": 4.609739826551034e-05,
+      "loss": 1.5107,
+      "step": 8090
+    },
+    {
+      "epoch": 7.469189937687514,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.6090727151434295e-05,
+      "loss": 1.3213,
+      "step": 8091
+    },
+    {
+      "epoch": 7.470113085621971,
+      "grad_norm": 3.232421875,
+      "learning_rate": 4.608405603735824e-05,
+      "loss": 1.459,
+      "step": 8092
+    },
+    {
+      "epoch": 7.471036233556427,
+      "grad_norm": 3.3828125,
+      "learning_rate": 4.607738492328219e-05,
+      "loss": 1.4071,
+      "step": 8093
+    },
+    {
+      "epoch": 7.471959381490884,
+      "grad_norm": 3.474609375,
+      "learning_rate": 4.607071380920614e-05,
+      "loss": 1.2649,
+      "step": 8094
+    },
+    {
+      "epoch": 7.47288252942534,
+      "grad_norm": 3.060546875,
+      "learning_rate": 4.606404269513009e-05,
+      "loss": 1.259,
+      "step": 8095
+    },
+    {
+      "epoch": 7.473805677359797,
+      "grad_norm": 3.66796875,
+      "learning_rate": 4.6057371581054035e-05,
+      "loss": 1.3926,
+      "step": 8096
+    },
+    {
+      "epoch": 7.474728825294253,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.605070046697799e-05,
+      "loss": 1.2737,
+      "step": 8097
+    },
+    {
+      "epoch": 7.47565197322871,
+      "grad_norm": 3.73828125,
+      "learning_rate": 4.604402935290193e-05,
+      "loss": 1.314,
+      "step": 8098
+    },
+    {
+      "epoch": 7.476575121163166,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.6037358238825885e-05,
+      "loss": 1.3638,
+      "step": 8099
+    },
+    {
+      "epoch": 7.477498269097623,
+      "grad_norm": 4.0390625,
+      "learning_rate": 4.603068712474983e-05,
+      "loss": 1.4504,
+      "step": 8100
+    },
+    {
+      "epoch": 7.478421417032079,
+      "grad_norm": 3.658203125,
+      "learning_rate": 4.602401601067379e-05,
+      "loss": 1.4021,
+      "step": 8101
+    },
+    {
+      "epoch": 7.479344564966536,
+      "grad_norm": 3.37890625,
+      "learning_rate": 4.6017344896597735e-05,
+      "loss": 1.3208,
+      "step": 8102
+    },
+    {
+      "epoch": 7.480267712900992,
+      "grad_norm": 3.10546875,
+      "learning_rate": 4.601067378252169e-05,
+      "loss": 1.3992,
+      "step": 8103
+    },
+    {
+      "epoch": 7.481190860835449,
+      "grad_norm": 3.666015625,
+      "learning_rate": 4.600400266844563e-05,
+      "loss": 1.2736,
+      "step": 8104
+    },
+    {
+      "epoch": 7.482114008769905,
+      "grad_norm": 3.220703125,
+      "learning_rate": 4.5997331554369585e-05,
+      "loss": 1.5039,
+      "step": 8105
+    },
+    {
+      "epoch": 7.483037156704362,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.599066044029353e-05,
+      "loss": 1.504,
+      "step": 8106
+    },
+    {
+      "epoch": 7.483960304638819,
+      "grad_norm": 2.607421875,
+      "learning_rate": 4.598398932621748e-05,
+      "loss": 1.126,
+      "step": 8107
+    },
+    {
+      "epoch": 7.484883452573275,
+      "grad_norm": 4.0078125,
+      "learning_rate": 4.597731821214143e-05,
+      "loss": 1.1709,
+      "step": 8108
+    },
+    {
+      "epoch": 7.485806600507732,
+      "grad_norm": 3.931640625,
+      "learning_rate": 4.597064709806538e-05,
+      "loss": 1.3875,
+      "step": 8109
+    },
+    {
+      "epoch": 7.486729748442188,
+      "grad_norm": 4.59765625,
+      "learning_rate": 4.5963975983989326e-05,
+      "loss": 0.9321,
+      "step": 8110
+    },
+    {
+      "epoch": 7.487652896376645,
+      "grad_norm": 3.314453125,
+      "learning_rate": 4.595730486991328e-05,
+      "loss": 1.229,
+      "step": 8111
+    },
+    {
+      "epoch": 7.488576044311101,
+      "grad_norm": 3.4921875,
+      "learning_rate": 4.595063375583722e-05,
+      "loss": 1.1777,
+      "step": 8112
+    },
+    {
+      "epoch": 7.489499192245558,
+      "grad_norm": 3.576171875,
+      "learning_rate": 4.594396264176118e-05,
+      "loss": 1.2832,
+      "step": 8113
+    },
+    {
+      "epoch": 7.490422340180014,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.593729152768513e-05,
+      "loss": 1.3115,
+      "step": 8114
+    },
+    {
+      "epoch": 7.4913454881144705,
+      "grad_norm": 4.23828125,
+      "learning_rate": 4.593062041360908e-05,
+      "loss": 1.3403,
+      "step": 8115
+    },
+    {
+      "epoch": 7.492268636048927,
+      "grad_norm": 3.275390625,
+      "learning_rate": 4.5923949299533025e-05,
+      "loss": 1.2019,
+      "step": 8116
+    },
+    {
+      "epoch": 7.4931917839833835,
+      "grad_norm": 3.62109375,
+      "learning_rate": 4.591727818545698e-05,
+      "loss": 1.218,
+      "step": 8117
+    },
+    {
+      "epoch": 7.4941149319178395,
+      "grad_norm": 3.76171875,
+      "learning_rate": 4.591060707138092e-05,
+      "loss": 1.6367,
+      "step": 8118
+    },
+    {
+      "epoch": 7.495038079852296,
+      "grad_norm": 4.02734375,
+      "learning_rate": 4.5903935957304875e-05,
+      "loss": 1.574,
+      "step": 8119
+    },
+    {
+      "epoch": 7.4959612277867524,
+      "grad_norm": 3.40234375,
+      "learning_rate": 4.589726484322882e-05,
+      "loss": 1.2839,
+      "step": 8120
+    },
+    {
+      "epoch": 7.496884375721209,
+      "grad_norm": 3.958984375,
+      "learning_rate": 4.589059372915277e-05,
+      "loss": 1.2288,
+      "step": 8121
+    },
+    {
+      "epoch": 7.497807523655666,
+      "grad_norm": 3.08203125,
+      "learning_rate": 4.588392261507672e-05,
+      "loss": 1.1923,
+      "step": 8122
+    },
+    {
+      "epoch": 7.498730671590122,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.587725150100067e-05,
+      "loss": 1.3042,
+      "step": 8123
+    },
+    {
+      "epoch": 7.499653819524579,
+      "grad_norm": 3.39453125,
+      "learning_rate": 4.5870580386924616e-05,
+      "loss": 1.3828,
+      "step": 8124
+    },
+    {
+      "epoch": 7.500576967459035,
+      "grad_norm": 3.630859375,
+      "learning_rate": 4.586390927284857e-05,
+      "loss": 1.3074,
+      "step": 8125
+    },
+    {
+      "epoch": 7.501500115393492,
+      "grad_norm": 3.275390625,
+      "learning_rate": 4.585723815877252e-05,
+      "loss": 1.4473,
+      "step": 8126
+    },
+    {
+      "epoch": 7.502423263327948,
+      "grad_norm": 3.443359375,
+      "learning_rate": 4.5850567044696466e-05,
+      "loss": 1.2227,
+      "step": 8127
+    },
+    {
+      "epoch": 7.503346411262405,
+      "grad_norm": 2.822265625,
+      "learning_rate": 4.584389593062042e-05,
+      "loss": 1.1121,
+      "step": 8128
+    },
+    {
+      "epoch": 7.504269559196861,
+      "grad_norm": 3.5390625,
+      "learning_rate": 4.5837224816544363e-05,
+      "loss": 1.2888,
+      "step": 8129
+    },
+    {
+      "epoch": 7.505192707131318,
+      "grad_norm": 3.119140625,
+      "learning_rate": 4.5830553702468316e-05,
+      "loss": 1.4651,
+      "step": 8130
+    },
+    {
+      "epoch": 7.506115855065774,
+      "grad_norm": 3.365234375,
+      "learning_rate": 4.582388258839226e-05,
+      "loss": 1.3007,
+      "step": 8131
+    },
+    {
+      "epoch": 7.507039003000231,
+      "grad_norm": 3.5,
+      "learning_rate": 4.581721147431621e-05,
+      "loss": 1.322,
+      "step": 8132
+    },
+    {
+      "epoch": 7.507962150934687,
+      "grad_norm": 3.685546875,
+      "learning_rate": 4.581054036024016e-05,
+      "loss": 1.467,
+      "step": 8133
+    },
+    {
+      "epoch": 7.508885298869144,
+      "grad_norm": 3.189453125,
+      "learning_rate": 4.580386924616411e-05,
+      "loss": 0.8982,
+      "step": 8134
+    },
+    {
+      "epoch": 7.5098084468036,
+      "grad_norm": 3.677734375,
+      "learning_rate": 4.5797198132088056e-05,
+      "loss": 1.46,
+      "step": 8135
+    },
+    {
+      "epoch": 7.510731594738057,
+      "grad_norm": 3.5859375,
+      "learning_rate": 4.579052701801201e-05,
+      "loss": 1.1051,
+      "step": 8136
+    },
+    {
+      "epoch": 7.511654742672514,
+      "grad_norm": 3.853515625,
+      "learning_rate": 4.5783855903935954e-05,
+      "loss": 1.4041,
+      "step": 8137
+    },
+    {
+      "epoch": 7.51257789060697,
+      "grad_norm": 3.94921875,
+      "learning_rate": 4.577718478985991e-05,
+      "loss": 1.2905,
+      "step": 8138
+    },
+    {
+      "epoch": 7.513501038541426,
+      "grad_norm": 3.69140625,
+      "learning_rate": 4.577051367578386e-05,
+      "loss": 1.5261,
+      "step": 8139
+    },
+    {
+      "epoch": 7.514424186475883,
+      "grad_norm": 3.662109375,
+      "learning_rate": 4.576384256170781e-05,
+      "loss": 1.3931,
+      "step": 8140
+    },
+    {
+      "epoch": 7.51534733441034,
+      "grad_norm": 2.64453125,
+      "learning_rate": 4.5757171447631756e-05,
+      "loss": 1.1419,
+      "step": 8141
+    },
+    {
+      "epoch": 7.516270482344796,
+      "grad_norm": 3.009765625,
+      "learning_rate": 4.575050033355571e-05,
+      "loss": 1.0113,
+      "step": 8142
+    },
+    {
+      "epoch": 7.517193630279253,
+      "grad_norm": 3.41015625,
+      "learning_rate": 4.5743829219479654e-05,
+      "loss": 1.2545,
+      "step": 8143
+    },
+    {
+      "epoch": 7.518116778213709,
+      "grad_norm": 2.998046875,
+      "learning_rate": 4.5737158105403606e-05,
+      "loss": 1.1824,
+      "step": 8144
+    },
+    {
+      "epoch": 7.5190399261481655,
+      "grad_norm": 2.9453125,
+      "learning_rate": 4.573048699132755e-05,
+      "loss": 1.4807,
+      "step": 8145
+    },
+    {
+      "epoch": 7.5199630740826215,
+      "grad_norm": 3.642578125,
+      "learning_rate": 4.5723815877251503e-05,
+      "loss": 1.2114,
+      "step": 8146
+    },
+    {
+      "epoch": 7.5208862220170785,
+      "grad_norm": 3.20703125,
+      "learning_rate": 4.571714476317545e-05,
+      "loss": 1.498,
+      "step": 8147
+    },
+    {
+      "epoch": 7.5218093699515345,
+      "grad_norm": 3.5703125,
+      "learning_rate": 4.57104736490994e-05,
+      "loss": 1.2543,
+      "step": 8148
+    },
+    {
+      "epoch": 7.522732517885991,
+      "grad_norm": 3.46484375,
+      "learning_rate": 4.5703802535023347e-05,
+      "loss": 1.2303,
+      "step": 8149
+    },
+    {
+      "epoch": 7.523655665820447,
+      "grad_norm": 3.48046875,
+      "learning_rate": 4.56971314209473e-05,
+      "loss": 0.9994,
+      "step": 8150
+    },
+    {
+      "epoch": 7.524578813754904,
+      "grad_norm": 3.3203125,
+      "learning_rate": 4.569046030687125e-05,
+      "loss": 1.3197,
+      "step": 8151
+    },
+    {
+      "epoch": 7.52550196168936,
+      "grad_norm": 3.91015625,
+      "learning_rate": 4.56837891927952e-05,
+      "loss": 1.2676,
+      "step": 8152
+    },
+    {
+      "epoch": 7.526425109623817,
+      "grad_norm": 3.23828125,
+      "learning_rate": 4.567711807871915e-05,
+      "loss": 1.1759,
+      "step": 8153
+    },
+    {
+      "epoch": 7.527348257558273,
+      "grad_norm": 3.30859375,
+      "learning_rate": 4.56704469646431e-05,
+      "loss": 1.2393,
+      "step": 8154
+    },
+    {
+      "epoch": 7.52827140549273,
+      "grad_norm": 3.396484375,
+      "learning_rate": 4.5663775850567046e-05,
+      "loss": 1.449,
+      "step": 8155
+    },
+    {
+      "epoch": 7.529194553427187,
+      "grad_norm": 4.0390625,
+      "learning_rate": 4.5657104736491e-05,
+      "loss": 1.4646,
+      "step": 8156
+    },
+    {
+      "epoch": 7.530117701361643,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5650433622414944e-05,
+      "loss": 1.0804,
+      "step": 8157
+    },
+    {
+      "epoch": 7.5310408492961,
+      "grad_norm": 4.0859375,
+      "learning_rate": 4.5643762508338896e-05,
+      "loss": 1.2966,
+      "step": 8158
+    },
+    {
+      "epoch": 7.531963997230556,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.563709139426284e-05,
+      "loss": 1.3508,
+      "step": 8159
+    },
+    {
+      "epoch": 7.532887145165013,
+      "grad_norm": 4.17578125,
+      "learning_rate": 4.5630420280186794e-05,
+      "loss": 1.4448,
+      "step": 8160
+    },
+    {
+      "epoch": 7.533810293099469,
+      "grad_norm": 3.39453125,
+      "learning_rate": 4.562374916611074e-05,
+      "loss": 1.3975,
+      "step": 8161
+    },
+    {
+      "epoch": 7.534733441033926,
+      "grad_norm": 3.822265625,
+      "learning_rate": 4.561707805203469e-05,
+      "loss": 1.5068,
+      "step": 8162
+    },
+    {
+      "epoch": 7.535656588968382,
+      "grad_norm": 3.0078125,
+      "learning_rate": 4.561040693795864e-05,
+      "loss": 1.3193,
+      "step": 8163
+    },
+    {
+      "epoch": 7.536579736902839,
+      "grad_norm": 3.478515625,
+      "learning_rate": 4.5603735823882596e-05,
+      "loss": 1.3499,
+      "step": 8164
+    },
+    {
+      "epoch": 7.537502884837295,
+      "grad_norm": 3.107421875,
+      "learning_rate": 4.559706470980654e-05,
+      "loss": 1.4814,
+      "step": 8165
+    },
+    {
+      "epoch": 7.538426032771752,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.5590393595730493e-05,
+      "loss": 1.2307,
+      "step": 8166
+    },
+    {
+      "epoch": 7.539349180706208,
+      "grad_norm": 3.09765625,
+      "learning_rate": 4.558372248165444e-05,
+      "loss": 1.1666,
+      "step": 8167
+    },
+    {
+      "epoch": 7.540272328640665,
+      "grad_norm": 3.46484375,
+      "learning_rate": 4.557705136757839e-05,
+      "loss": 1.2576,
+      "step": 8168
+    },
+    {
+      "epoch": 7.541195476575121,
+      "grad_norm": 3.451171875,
+      "learning_rate": 4.5570380253502336e-05,
+      "loss": 1.3389,
+      "step": 8169
+    },
+    {
+      "epoch": 7.542118624509578,
+      "grad_norm": 3.408203125,
+      "learning_rate": 4.556370913942629e-05,
+      "loss": 1.4177,
+      "step": 8170
+    },
+    {
+      "epoch": 7.543041772444035,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.5557038025350234e-05,
+      "loss": 1.3296,
+      "step": 8171
+    },
+    {
+      "epoch": 7.543964920378491,
+      "grad_norm": 4.046875,
+      "learning_rate": 4.5550366911274186e-05,
+      "loss": 1.3209,
+      "step": 8172
+    },
+    {
+      "epoch": 7.544888068312947,
+      "grad_norm": 4.2734375,
+      "learning_rate": 4.554369579719813e-05,
+      "loss": 1.3101,
+      "step": 8173
+    },
+    {
+      "epoch": 7.545811216247404,
+      "grad_norm": 3.951171875,
+      "learning_rate": 4.5537024683122084e-05,
+      "loss": 1.3701,
+      "step": 8174
+    },
+    {
+      "epoch": 7.5467343641818605,
+      "grad_norm": 4.6328125,
+      "learning_rate": 4.553035356904603e-05,
+      "loss": 1.459,
+      "step": 8175
+    },
+    {
+      "epoch": 7.5476575121163165,
+      "grad_norm": 2.91796875,
+      "learning_rate": 4.552368245496998e-05,
+      "loss": 1.4255,
+      "step": 8176
+    },
+    {
+      "epoch": 7.548580660050773,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.5517011340893934e-05,
+      "loss": 1.3975,
+      "step": 8177
+    },
+    {
+      "epoch": 7.5495038079852295,
+      "grad_norm": 3.54296875,
+      "learning_rate": 4.551034022681788e-05,
+      "loss": 1.1282,
+      "step": 8178
+    },
+    {
+      "epoch": 7.550426955919686,
+      "grad_norm": 2.978515625,
+      "learning_rate": 4.550366911274183e-05,
+      "loss": 1.415,
+      "step": 8179
+    },
+    {
+      "epoch": 7.551350103854142,
+      "grad_norm": 3.22265625,
+      "learning_rate": 4.549699799866578e-05,
+      "loss": 1.2893,
+      "step": 8180
+    },
+    {
+      "epoch": 7.552273251788599,
+      "grad_norm": 2.939453125,
+      "learning_rate": 4.549032688458973e-05,
+      "loss": 1.3342,
+      "step": 8181
+    },
+    {
+      "epoch": 7.553196399723055,
+      "grad_norm": 4.0546875,
+      "learning_rate": 4.5483655770513675e-05,
+      "loss": 1.3416,
+      "step": 8182
+    },
+    {
+      "epoch": 7.554119547657512,
+      "grad_norm": 3.48046875,
+      "learning_rate": 4.547698465643763e-05,
+      "loss": 1.3713,
+      "step": 8183
+    },
+    {
+      "epoch": 7.555042695591968,
+      "grad_norm": 4.0078125,
+      "learning_rate": 4.547031354236158e-05,
+      "loss": 0.824,
+      "step": 8184
+    },
+    {
+      "epoch": 7.555965843526425,
+      "grad_norm": 3.5078125,
+      "learning_rate": 4.5463642428285524e-05,
+      "loss": 1.6353,
+      "step": 8185
+    },
+    {
+      "epoch": 7.556888991460881,
+      "grad_norm": 3.36328125,
+      "learning_rate": 4.5456971314209477e-05,
+      "loss": 1.348,
+      "step": 8186
+    },
+    {
+      "epoch": 7.557812139395338,
+      "grad_norm": 3.810546875,
+      "learning_rate": 4.545030020013342e-05,
+      "loss": 1.2468,
+      "step": 8187
+    },
+    {
+      "epoch": 7.558735287329794,
+      "grad_norm": 3.61328125,
+      "learning_rate": 4.5443629086057374e-05,
+      "loss": 1.647,
+      "step": 8188
+    },
+    {
+      "epoch": 7.559658435264251,
+      "grad_norm": 3.314453125,
+      "learning_rate": 4.5436957971981326e-05,
+      "loss": 1.5129,
+      "step": 8189
+    },
+    {
+      "epoch": 7.560581583198708,
+      "grad_norm": 3.833984375,
+      "learning_rate": 4.543028685790527e-05,
+      "loss": 1.2424,
+      "step": 8190
+    },
+    {
+      "epoch": 7.561504731133164,
+      "grad_norm": 2.759765625,
+      "learning_rate": 4.5423615743829224e-05,
+      "loss": 0.9568,
+      "step": 8191
+    },
+    {
+      "epoch": 7.562427879067621,
+      "grad_norm": 3.427734375,
+      "learning_rate": 4.541694462975317e-05,
+      "loss": 1.2898,
+      "step": 8192
+    },
+    {
+      "epoch": 7.563351027002077,
+      "grad_norm": 3.7890625,
+      "learning_rate": 4.541027351567712e-05,
+      "loss": 1.3646,
+      "step": 8193
+    },
+    {
+      "epoch": 7.564274174936534,
+      "grad_norm": 3.216796875,
+      "learning_rate": 4.540360240160107e-05,
+      "loss": 1.3704,
+      "step": 8194
+    },
+    {
+      "epoch": 7.56519732287099,
+      "grad_norm": 3.73828125,
+      "learning_rate": 4.539693128752502e-05,
+      "loss": 1.3455,
+      "step": 8195
+    },
+    {
+      "epoch": 7.566120470805447,
+      "grad_norm": 3.45703125,
+      "learning_rate": 4.5390260173448965e-05,
+      "loss": 1.0508,
+      "step": 8196
+    },
+    {
+      "epoch": 7.567043618739903,
+      "grad_norm": 3.6796875,
+      "learning_rate": 4.538358905937292e-05,
+      "loss": 1.4187,
+      "step": 8197
+    },
+    {
+      "epoch": 7.56796676667436,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.537691794529686e-05,
+      "loss": 1.3844,
+      "step": 8198
+    },
+    {
+      "epoch": 7.568889914608816,
+      "grad_norm": 3.3359375,
+      "learning_rate": 4.5370246831220815e-05,
+      "loss": 1.1714,
+      "step": 8199
+    },
+    {
+      "epoch": 7.569813062543273,
+      "grad_norm": 3.423828125,
+      "learning_rate": 4.536357571714476e-05,
+      "loss": 1.0525,
+      "step": 8200
+    },
+    {
+      "epoch": 7.570736210477729,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.535690460306872e-05,
+      "loss": 1.3981,
+      "step": 8201
+    },
+    {
+      "epoch": 7.571659358412186,
+      "grad_norm": 3.509765625,
+      "learning_rate": 4.5350233488992664e-05,
+      "loss": 1.2954,
+      "step": 8202
+    },
+    {
+      "epoch": 7.572582506346642,
+      "grad_norm": 3.705078125,
+      "learning_rate": 4.534356237491662e-05,
+      "loss": 1.3521,
+      "step": 8203
+    },
+    {
+      "epoch": 7.573505654281099,
+      "grad_norm": 3.099609375,
+      "learning_rate": 4.533689126084056e-05,
+      "loss": 1.4326,
+      "step": 8204
+    },
+    {
+      "epoch": 7.5744288022155555,
+      "grad_norm": 3.365234375,
+      "learning_rate": 4.5330220146764514e-05,
+      "loss": 1.5439,
+      "step": 8205
+    },
+    {
+      "epoch": 7.5753519501500115,
+      "grad_norm": 3.052734375,
+      "learning_rate": 4.532354903268846e-05,
+      "loss": 1.5281,
+      "step": 8206
+    },
+    {
+      "epoch": 7.576275098084468,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.531687791861241e-05,
+      "loss": 1.5437,
+      "step": 8207
+    },
+    {
+      "epoch": 7.577198246018924,
+      "grad_norm": 3.30859375,
+      "learning_rate": 4.531020680453636e-05,
+      "loss": 1.4109,
+      "step": 8208
+    },
+    {
+      "epoch": 7.578121393953381,
+      "grad_norm": 3.291015625,
+      "learning_rate": 4.530353569046031e-05,
+      "loss": 1.4459,
+      "step": 8209
+    },
+    {
+      "epoch": 7.579044541887837,
+      "grad_norm": 3.794921875,
+      "learning_rate": 4.5296864576384255e-05,
+      "loss": 1.2852,
+      "step": 8210
+    },
+    {
+      "epoch": 7.579967689822294,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.529019346230821e-05,
+      "loss": 1.3713,
+      "step": 8211
+    },
+    {
+      "epoch": 7.58089083775675,
+      "grad_norm": 3.388671875,
+      "learning_rate": 4.528352234823215e-05,
+      "loss": 1.5369,
+      "step": 8212
+    },
+    {
+      "epoch": 7.581813985691207,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.5276851234156105e-05,
+      "loss": 1.3416,
+      "step": 8213
+    },
+    {
+      "epoch": 7.582737133625663,
+      "grad_norm": 3.33984375,
+      "learning_rate": 4.527018012008006e-05,
+      "loss": 1.343,
+      "step": 8214
+    },
+    {
+      "epoch": 7.58366028156012,
+      "grad_norm": 3.224609375,
+      "learning_rate": 4.526350900600401e-05,
+      "loss": 1.3975,
+      "step": 8215
+    },
+    {
+      "epoch": 7.584583429494576,
+      "grad_norm": 3.888671875,
+      "learning_rate": 4.5256837891927955e-05,
+      "loss": 1.4606,
+      "step": 8216
+    },
+    {
+      "epoch": 7.585506577429033,
+      "grad_norm": 3.37109375,
+      "learning_rate": 4.525016677785191e-05,
+      "loss": 1.4924,
+      "step": 8217
+    },
+    {
+      "epoch": 7.586429725363489,
+      "grad_norm": 3.400390625,
+      "learning_rate": 4.524349566377585e-05,
+      "loss": 1.4167,
+      "step": 8218
+    },
+    {
+      "epoch": 7.587352873297946,
+      "grad_norm": 3.728515625,
+      "learning_rate": 4.5236824549699805e-05,
+      "loss": 1.276,
+      "step": 8219
+    },
+    {
+      "epoch": 7.588276021232403,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.523015343562375e-05,
+      "loss": 1.2477,
+      "step": 8220
+    },
+    {
+      "epoch": 7.589199169166859,
+      "grad_norm": 3.279296875,
+      "learning_rate": 4.52234823215477e-05,
+      "loss": 1.5635,
+      "step": 8221
+    },
+    {
+      "epoch": 7.590122317101315,
+      "grad_norm": 3.201171875,
+      "learning_rate": 4.521681120747165e-05,
+      "loss": 1.467,
+      "step": 8222
+    },
+    {
+      "epoch": 7.591045465035772,
+      "grad_norm": 3.654296875,
+      "learning_rate": 4.52101400933956e-05,
+      "loss": 1.4275,
+      "step": 8223
+    },
+    {
+      "epoch": 7.591968612970229,
+      "grad_norm": 3.931640625,
+      "learning_rate": 4.5203468979319545e-05,
+      "loss": 1.4895,
+      "step": 8224
+    },
+    {
+      "epoch": 7.592891760904685,
+      "grad_norm": 3.287109375,
+      "learning_rate": 4.51967978652435e-05,
+      "loss": 1.1716,
+      "step": 8225
+    },
+    {
+      "epoch": 7.593814908839142,
+      "grad_norm": 3.759765625,
+      "learning_rate": 4.519012675116745e-05,
+      "loss": 1.1903,
+      "step": 8226
+    },
+    {
+      "epoch": 7.594738056773598,
+      "grad_norm": 3.365234375,
+      "learning_rate": 4.51834556370914e-05,
+      "loss": 1.5034,
+      "step": 8227
+    },
+    {
+      "epoch": 7.595661204708055,
+      "grad_norm": 3.54296875,
+      "learning_rate": 4.517678452301535e-05,
+      "loss": 1.3142,
+      "step": 8228
+    },
+    {
+      "epoch": 7.596584352642511,
+      "grad_norm": 4.17578125,
+      "learning_rate": 4.51701134089393e-05,
+      "loss": 1.3047,
+      "step": 8229
+    },
+    {
+      "epoch": 7.597507500576968,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.5163442294863245e-05,
+      "loss": 1.2744,
+      "step": 8230
+    },
+    {
+      "epoch": 7.598430648511424,
+      "grad_norm": 3.2421875,
+      "learning_rate": 4.51567711807872e-05,
+      "loss": 1.1711,
+      "step": 8231
+    },
+    {
+      "epoch": 7.599353796445881,
+      "grad_norm": 3.68359375,
+      "learning_rate": 4.515010006671114e-05,
+      "loss": 1.4692,
+      "step": 8232
+    },
+    {
+      "epoch": 7.600276944380337,
+      "grad_norm": 3.90234375,
+      "learning_rate": 4.5143428952635095e-05,
+      "loss": 1.2734,
+      "step": 8233
+    },
+    {
+      "epoch": 7.6012000923147935,
+      "grad_norm": 3.1640625,
+      "learning_rate": 4.513675783855904e-05,
+      "loss": 1.2319,
+      "step": 8234
+    },
+    {
+      "epoch": 7.60212324024925,
+      "grad_norm": 3.556640625,
+      "learning_rate": 4.513008672448299e-05,
+      "loss": 1.4426,
+      "step": 8235
+    },
+    {
+      "epoch": 7.6030463881837065,
+      "grad_norm": 3.298828125,
+      "learning_rate": 4.512341561040694e-05,
+      "loss": 1.2778,
+      "step": 8236
+    },
+    {
+      "epoch": 7.6039695361181625,
+      "grad_norm": 4.1171875,
+      "learning_rate": 4.511674449633089e-05,
+      "loss": 1.2798,
+      "step": 8237
+    },
+    {
+      "epoch": 7.604892684052619,
+      "grad_norm": 3.07421875,
+      "learning_rate": 4.5110073382254835e-05,
+      "loss": 1.0986,
+      "step": 8238
+    },
+    {
+      "epoch": 7.605815831987076,
+      "grad_norm": 3.736328125,
+      "learning_rate": 4.510340226817879e-05,
+      "loss": 1.1392,
+      "step": 8239
+    },
+    {
+      "epoch": 7.606738979921532,
+      "grad_norm": 3.853515625,
+      "learning_rate": 4.509673115410274e-05,
+      "loss": 1.2241,
+      "step": 8240
+    },
+    {
+      "epoch": 7.607662127855989,
+      "grad_norm": 3.283203125,
+      "learning_rate": 4.5090060040026685e-05,
+      "loss": 1.1355,
+      "step": 8241
+    },
+    {
+      "epoch": 7.608585275790445,
+      "grad_norm": 3.662109375,
+      "learning_rate": 4.508338892595064e-05,
+      "loss": 1.2897,
+      "step": 8242
+    },
+    {
+      "epoch": 7.609508423724902,
+      "grad_norm": 3.962890625,
+      "learning_rate": 4.507671781187458e-05,
+      "loss": 1.2642,
+      "step": 8243
+    },
+    {
+      "epoch": 7.610431571659358,
+      "grad_norm": 3.541015625,
+      "learning_rate": 4.5070046697798535e-05,
+      "loss": 1.1454,
+      "step": 8244
+    },
+    {
+      "epoch": 7.611354719593815,
+      "grad_norm": 3.33984375,
+      "learning_rate": 4.506337558372248e-05,
+      "loss": 1.182,
+      "step": 8245
+    },
+    {
+      "epoch": 7.612277867528271,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.505670446964643e-05,
+      "loss": 1.436,
+      "step": 8246
+    },
+    {
+      "epoch": 7.613201015462728,
+      "grad_norm": 3.701171875,
+      "learning_rate": 4.505003335557038e-05,
+      "loss": 1.1201,
+      "step": 8247
+    },
+    {
+      "epoch": 7.614124163397184,
+      "grad_norm": 3.4609375,
+      "learning_rate": 4.504336224149433e-05,
+      "loss": 1.5776,
+      "step": 8248
+    },
+    {
+      "epoch": 7.615047311331641,
+      "grad_norm": 3.228515625,
+      "learning_rate": 4.5036691127418276e-05,
+      "loss": 1.249,
+      "step": 8249
+    },
+    {
+      "epoch": 7.615970459266097,
+      "grad_norm": 3.54296875,
+      "learning_rate": 4.503002001334223e-05,
+      "loss": 1.2006,
+      "step": 8250
+    },
+    {
+      "epoch": 7.616893607200554,
+      "grad_norm": 3.740234375,
+      "learning_rate": 4.502334889926618e-05,
+      "loss": 1.4084,
+      "step": 8251
+    },
+    {
+      "epoch": 7.61781675513501,
+      "grad_norm": 3.15234375,
+      "learning_rate": 4.501667778519013e-05,
+      "loss": 1.3997,
+      "step": 8252
+    },
+    {
+      "epoch": 7.618739903069467,
+      "grad_norm": 3.751953125,
+      "learning_rate": 4.501000667111408e-05,
+      "loss": 1.2874,
+      "step": 8253
+    },
+    {
+      "epoch": 7.619663051003924,
+      "grad_norm": 4.18359375,
+      "learning_rate": 4.500333555703803e-05,
+      "loss": 1.2263,
+      "step": 8254
+    },
+    {
+      "epoch": 7.62058619893838,
+      "grad_norm": 2.951171875,
+      "learning_rate": 4.4996664442961976e-05,
+      "loss": 1.0895,
+      "step": 8255
+    },
+    {
+      "epoch": 7.621509346872836,
+      "grad_norm": 3.37109375,
+      "learning_rate": 4.498999332888593e-05,
+      "loss": 1.0753,
+      "step": 8256
+    },
+    {
+      "epoch": 7.622432494807293,
+      "grad_norm": 3.751953125,
+      "learning_rate": 4.498332221480987e-05,
+      "loss": 1.1292,
+      "step": 8257
+    },
+    {
+      "epoch": 7.62335564274175,
+      "grad_norm": 3.21484375,
+      "learning_rate": 4.4976651100733825e-05,
+      "loss": 1.4353,
+      "step": 8258
+    },
+    {
+      "epoch": 7.624278790676206,
+      "grad_norm": 3.462890625,
+      "learning_rate": 4.496997998665777e-05,
+      "loss": 1.5979,
+      "step": 8259
+    },
+    {
+      "epoch": 7.625201938610663,
+      "grad_norm": 3.357421875,
+      "learning_rate": 4.496330887258172e-05,
+      "loss": 1.458,
+      "step": 8260
+    },
+    {
+      "epoch": 7.626125086545119,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.495663775850567e-05,
+      "loss": 1.2676,
+      "step": 8261
+    },
+    {
+      "epoch": 7.627048234479576,
+      "grad_norm": 4.33984375,
+      "learning_rate": 4.494996664442962e-05,
+      "loss": 1.4182,
+      "step": 8262
+    },
+    {
+      "epoch": 7.627971382414032,
+      "grad_norm": 3.14453125,
+      "learning_rate": 4.4943295530353566e-05,
+      "loss": 1.5161,
+      "step": 8263
+    },
+    {
+      "epoch": 7.6288945303484885,
+      "grad_norm": 3.33984375,
+      "learning_rate": 4.4936624416277525e-05,
+      "loss": 1.3049,
+      "step": 8264
+    },
+    {
+      "epoch": 7.6298176782829445,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.492995330220147e-05,
+      "loss": 1.1678,
+      "step": 8265
+    },
+    {
+      "epoch": 7.6307408262174015,
+      "grad_norm": 3.732421875,
+      "learning_rate": 4.492328218812542e-05,
+      "loss": 1.3745,
+      "step": 8266
+    },
+    {
+      "epoch": 7.6316639741518575,
+      "grad_norm": 3.314453125,
+      "learning_rate": 4.491661107404937e-05,
+      "loss": 1.4648,
+      "step": 8267
+    },
+    {
+      "epoch": 7.632587122086314,
+      "grad_norm": 3.716796875,
+      "learning_rate": 4.490993995997332e-05,
+      "loss": 0.9432,
+      "step": 8268
+    },
+    {
+      "epoch": 7.633510270020771,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.4903268845897266e-05,
+      "loss": 0.8483,
+      "step": 8269
+    },
+    {
+      "epoch": 7.634433417955227,
+      "grad_norm": 3.099609375,
+      "learning_rate": 4.489659773182122e-05,
+      "loss": 1.4688,
+      "step": 8270
+    },
+    {
+      "epoch": 7.635356565889683,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.4889926617745163e-05,
+      "loss": 1.5916,
+      "step": 8271
+    },
+    {
+      "epoch": 7.63627971382414,
+      "grad_norm": 3.3046875,
+      "learning_rate": 4.4883255503669116e-05,
+      "loss": 1.3105,
+      "step": 8272
+    },
+    {
+      "epoch": 7.637202861758597,
+      "grad_norm": 3.90234375,
+      "learning_rate": 4.487658438959306e-05,
+      "loss": 1.2577,
+      "step": 8273
+    },
+    {
+      "epoch": 7.638126009693053,
+      "grad_norm": 3.619140625,
+      "learning_rate": 4.486991327551701e-05,
+      "loss": 1.5132,
+      "step": 8274
+    },
+    {
+      "epoch": 7.63904915762751,
+      "grad_norm": 4.203125,
+      "learning_rate": 4.486324216144096e-05,
+      "loss": 1.3401,
+      "step": 8275
+    },
+    {
+      "epoch": 7.639972305561966,
+      "grad_norm": 3.232421875,
+      "learning_rate": 4.485657104736492e-05,
+      "loss": 0.9807,
+      "step": 8276
+    },
+    {
+      "epoch": 7.640895453496423,
+      "grad_norm": 3.8984375,
+      "learning_rate": 4.484989993328886e-05,
+      "loss": 1.3936,
+      "step": 8277
+    },
+    {
+      "epoch": 7.641818601430879,
+      "grad_norm": 3.416015625,
+      "learning_rate": 4.4843228819212815e-05,
+      "loss": 1.2877,
+      "step": 8278
+    },
+    {
+      "epoch": 7.642741749365336,
+      "grad_norm": 3.88671875,
+      "learning_rate": 4.483655770513676e-05,
+      "loss": 1.4734,
+      "step": 8279
+    },
+    {
+      "epoch": 7.643664897299792,
+      "grad_norm": 3.24609375,
+      "learning_rate": 4.482988659106071e-05,
+      "loss": 1.5662,
+      "step": 8280
+    },
+    {
+      "epoch": 7.644588045234249,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.482321547698466e-05,
+      "loss": 0.961,
+      "step": 8281
+    },
+    {
+      "epoch": 7.645511193168705,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.481654436290861e-05,
+      "loss": 1.2434,
+      "step": 8282
+    },
+    {
+      "epoch": 7.646434341103162,
+      "grad_norm": 3.228515625,
+      "learning_rate": 4.4809873248832556e-05,
+      "loss": 1.3748,
+      "step": 8283
+    },
+    {
+      "epoch": 7.647357489037618,
+      "grad_norm": 3.80859375,
+      "learning_rate": 4.480320213475651e-05,
+      "loss": 1.3734,
+      "step": 8284
+    },
+    {
+      "epoch": 7.648280636972075,
+      "grad_norm": 2.681640625,
+      "learning_rate": 4.4796531020680454e-05,
+      "loss": 1.3652,
+      "step": 8285
+    },
+    {
+      "epoch": 7.649203784906531,
+      "grad_norm": 4.00390625,
+      "learning_rate": 4.4789859906604406e-05,
+      "loss": 1.303,
+      "step": 8286
+    },
+    {
+      "epoch": 7.650126932840988,
+      "grad_norm": 3.185546875,
+      "learning_rate": 4.478318879252835e-05,
+      "loss": 1.3999,
+      "step": 8287
+    },
+    {
+      "epoch": 7.651050080775445,
+      "grad_norm": 3.275390625,
+      "learning_rate": 4.4776517678452304e-05,
+      "loss": 0.9505,
+      "step": 8288
+    },
+    {
+      "epoch": 7.651973228709901,
+      "grad_norm": 3.443359375,
+      "learning_rate": 4.4769846564376256e-05,
+      "loss": 1.2042,
+      "step": 8289
+    },
+    {
+      "epoch": 7.652896376644358,
+      "grad_norm": 3.455078125,
+      "learning_rate": 4.47631754503002e-05,
+      "loss": 1.1101,
+      "step": 8290
+    },
+    {
+      "epoch": 7.653819524578814,
+      "grad_norm": 3.693359375,
+      "learning_rate": 4.475650433622415e-05,
+      "loss": 1.3503,
+      "step": 8291
+    },
+    {
+      "epoch": 7.654742672513271,
+      "grad_norm": 3.87890625,
+      "learning_rate": 4.47498332221481e-05,
+      "loss": 1.4368,
+      "step": 8292
+    },
+    {
+      "epoch": 7.655665820447727,
+      "grad_norm": 3.06640625,
+      "learning_rate": 4.474316210807205e-05,
+      "loss": 1.1714,
+      "step": 8293
+    },
+    {
+      "epoch": 7.6565889683821835,
+      "grad_norm": 3.51171875,
+      "learning_rate": 4.4736490993995996e-05,
+      "loss": 1.5049,
+      "step": 8294
+    },
+    {
+      "epoch": 7.6575121163166395,
+      "grad_norm": 3.94921875,
+      "learning_rate": 4.472981987991995e-05,
+      "loss": 1.1687,
+      "step": 8295
+    },
+    {
+      "epoch": 7.658435264251096,
+      "grad_norm": 3.201171875,
+      "learning_rate": 4.47231487658439e-05,
+      "loss": 1.1279,
+      "step": 8296
+    },
+    {
+      "epoch": 7.6593584121855525,
+      "grad_norm": 3.48828125,
+      "learning_rate": 4.4716477651767846e-05,
+      "loss": 1.2064,
+      "step": 8297
+    },
+    {
+      "epoch": 7.660281560120009,
+      "grad_norm": 3.107421875,
+      "learning_rate": 4.47098065376918e-05,
+      "loss": 1.3845,
+      "step": 8298
+    },
+    {
+      "epoch": 7.661204708054465,
+      "grad_norm": 3.6796875,
+      "learning_rate": 4.4703135423615744e-05,
+      "loss": 1.0261,
+      "step": 8299
+    },
+    {
+      "epoch": 7.662127855988922,
+      "grad_norm": 3.5703125,
+      "learning_rate": 4.4696464309539696e-05,
+      "loss": 1.4048,
+      "step": 8300
+    },
+    {
+      "epoch": 7.663051003923378,
+      "grad_norm": 4.0,
+      "learning_rate": 4.468979319546365e-05,
+      "loss": 1.552,
+      "step": 8301
+    },
+    {
+      "epoch": 7.663974151857835,
+      "grad_norm": 3.158203125,
+      "learning_rate": 4.4683122081387594e-05,
+      "loss": 1.4858,
+      "step": 8302
+    },
+    {
+      "epoch": 7.664897299792292,
+      "grad_norm": 2.908203125,
+      "learning_rate": 4.4676450967311546e-05,
+      "loss": 1.5527,
+      "step": 8303
+    },
+    {
+      "epoch": 7.665820447726748,
+      "grad_norm": 3.875,
+      "learning_rate": 4.466977985323549e-05,
+      "loss": 1.4519,
+      "step": 8304
+    },
+    {
+      "epoch": 7.666743595661204,
+      "grad_norm": 3.494140625,
+      "learning_rate": 4.4663108739159444e-05,
+      "loss": 1.457,
+      "step": 8305
+    },
+    {
+      "epoch": 7.667666743595661,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.465643762508339e-05,
+      "loss": 1.4248,
+      "step": 8306
+    },
+    {
+      "epoch": 7.668589891530118,
+      "grad_norm": 3.369140625,
+      "learning_rate": 4.464976651100734e-05,
+      "loss": 1.4592,
+      "step": 8307
+    },
+    {
+      "epoch": 7.669513039464574,
+      "grad_norm": 3.189453125,
+      "learning_rate": 4.464309539693129e-05,
+      "loss": 1.3198,
+      "step": 8308
+    },
+    {
+      "epoch": 7.670436187399031,
+      "grad_norm": 2.75,
+      "learning_rate": 4.463642428285524e-05,
+      "loss": 1.1098,
+      "step": 8309
+    },
+    {
+      "epoch": 7.671359335333487,
+      "grad_norm": 3.93359375,
+      "learning_rate": 4.4629753168779184e-05,
+      "loss": 1.3894,
+      "step": 8310
+    },
+    {
+      "epoch": 7.672282483267944,
+      "grad_norm": 3.99609375,
+      "learning_rate": 4.4623082054703137e-05,
+      "loss": 1.3333,
+      "step": 8311
+    },
+    {
+      "epoch": 7.6732056312024,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.461641094062708e-05,
+      "loss": 1.3669,
+      "step": 8312
+    },
+    {
+      "epoch": 7.674128779136857,
+      "grad_norm": 4.01953125,
+      "learning_rate": 4.4609739826551034e-05,
+      "loss": 1.5625,
+      "step": 8313
+    },
+    {
+      "epoch": 7.675051927071313,
+      "grad_norm": 3.02734375,
+      "learning_rate": 4.4603068712474986e-05,
+      "loss": 1.5784,
+      "step": 8314
+    },
+    {
+      "epoch": 7.67597507500577,
+      "grad_norm": 3.689453125,
+      "learning_rate": 4.459639759839894e-05,
+      "loss": 1.412,
+      "step": 8315
+    },
+    {
+      "epoch": 7.676898222940226,
+      "grad_norm": 3.59765625,
+      "learning_rate": 4.4589726484322884e-05,
+      "loss": 1.3508,
+      "step": 8316
+    },
+    {
+      "epoch": 7.677821370874683,
+      "grad_norm": 3.1953125,
+      "learning_rate": 4.4583055370246836e-05,
+      "loss": 1.239,
+      "step": 8317
+    },
+    {
+      "epoch": 7.67874451880914,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.457638425617078e-05,
+      "loss": 1.3428,
+      "step": 8318
+    },
+    {
+      "epoch": 7.679667666743596,
+      "grad_norm": 4.11328125,
+      "learning_rate": 4.4569713142094734e-05,
+      "loss": 1.2189,
+      "step": 8319
+    },
+    {
+      "epoch": 7.680590814678052,
+      "grad_norm": 2.978515625,
+      "learning_rate": 4.456304202801868e-05,
+      "loss": 1.1729,
+      "step": 8320
+    },
+    {
+      "epoch": 7.681513962612509,
+      "grad_norm": 3.80078125,
+      "learning_rate": 4.455637091394263e-05,
+      "loss": 1.3292,
+      "step": 8321
+    },
+    {
+      "epoch": 7.6824371105469655,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.454969979986658e-05,
+      "loss": 0.9389,
+      "step": 8322
+    },
+    {
+      "epoch": 7.683360258481422,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.454302868579053e-05,
+      "loss": 1.377,
+      "step": 8323
+    },
+    {
+      "epoch": 7.6842834064158785,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.4536357571714475e-05,
+      "loss": 1.6008,
+      "step": 8324
+    },
+    {
+      "epoch": 7.6852065543503345,
+      "grad_norm": 3.0859375,
+      "learning_rate": 4.452968645763843e-05,
+      "loss": 1.6919,
+      "step": 8325
+    },
+    {
+      "epoch": 7.686129702284791,
+      "grad_norm": 4.0703125,
+      "learning_rate": 4.452301534356238e-05,
+      "loss": 1.4717,
+      "step": 8326
+    },
+    {
+      "epoch": 7.6870528502192474,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.451634422948633e-05,
+      "loss": 1.4902,
+      "step": 8327
+    },
+    {
+      "epoch": 7.687975998153704,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.4509673115410277e-05,
+      "loss": 1.3379,
+      "step": 8328
+    },
+    {
+      "epoch": 7.68889914608816,
+      "grad_norm": 3.533203125,
+      "learning_rate": 4.450300200133423e-05,
+      "loss": 1.3806,
+      "step": 8329
+    },
+    {
+      "epoch": 7.689822294022617,
+      "grad_norm": 3.3984375,
+      "learning_rate": 4.4496330887258174e-05,
+      "loss": 1.2069,
+      "step": 8330
+    },
+    {
+      "epoch": 7.690745441957073,
+      "grad_norm": 3.720703125,
+      "learning_rate": 4.4489659773182126e-05,
+      "loss": 1.2268,
+      "step": 8331
+    },
+    {
+      "epoch": 7.69166858989153,
+      "grad_norm": 3.716796875,
+      "learning_rate": 4.448298865910607e-05,
+      "loss": 1.5383,
+      "step": 8332
+    },
+    {
+      "epoch": 7.692591737825986,
+      "grad_norm": 3.75390625,
+      "learning_rate": 4.4476317545030024e-05,
+      "loss": 1.1788,
+      "step": 8333
+    },
+    {
+      "epoch": 7.693514885760443,
+      "grad_norm": 3.21484375,
+      "learning_rate": 4.446964643095397e-05,
+      "loss": 1.3597,
+      "step": 8334
+    },
+    {
+      "epoch": 7.694438033694899,
+      "grad_norm": 3.6953125,
+      "learning_rate": 4.446297531687792e-05,
+      "loss": 1.2211,
+      "step": 8335
+    },
+    {
+      "epoch": 7.695361181629356,
+      "grad_norm": 3.46484375,
+      "learning_rate": 4.445630420280187e-05,
+      "loss": 1.5691,
+      "step": 8336
+    },
+    {
+      "epoch": 7.696284329563813,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.444963308872582e-05,
+      "loss": 1.2844,
+      "step": 8337
+    },
+    {
+      "epoch": 7.697207477498269,
+      "grad_norm": 3.599609375,
+      "learning_rate": 4.4442961974649765e-05,
+      "loss": 1.3489,
+      "step": 8338
+    },
+    {
+      "epoch": 7.698130625432726,
+      "grad_norm": 2.94140625,
+      "learning_rate": 4.4436290860573724e-05,
+      "loss": 1.3203,
+      "step": 8339
+    },
+    {
+      "epoch": 7.699053773367182,
+      "grad_norm": 3.53515625,
+      "learning_rate": 4.442961974649767e-05,
+      "loss": 1.1329,
+      "step": 8340
+    },
+    {
+      "epoch": 7.699976921301639,
+      "grad_norm": 3.9765625,
+      "learning_rate": 4.442294863242162e-05,
+      "loss": 1.3359,
+      "step": 8341
+    },
+    {
+      "epoch": 7.700900069236095,
+      "grad_norm": 3.8515625,
+      "learning_rate": 4.441627751834557e-05,
+      "loss": 1.0773,
+      "step": 8342
+    },
+    {
+      "epoch": 7.701823217170552,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.440960640426952e-05,
+      "loss": 1.4078,
+      "step": 8343
+    },
+    {
+      "epoch": 7.702746365105008,
+      "grad_norm": 3.798828125,
+      "learning_rate": 4.4402935290193464e-05,
+      "loss": 0.9725,
+      "step": 8344
+    },
+    {
+      "epoch": 7.703669513039465,
+      "grad_norm": 3.275390625,
+      "learning_rate": 4.439626417611742e-05,
+      "loss": 1.2537,
+      "step": 8345
+    },
+    {
+      "epoch": 7.704592660973921,
+      "grad_norm": 4.00390625,
+      "learning_rate": 4.438959306204136e-05,
+      "loss": 1.2476,
+      "step": 8346
+    },
+    {
+      "epoch": 7.705515808908378,
+      "grad_norm": 4.296875,
+      "learning_rate": 4.4382921947965314e-05,
+      "loss": 1.4939,
+      "step": 8347
+    },
+    {
+      "epoch": 7.706438956842834,
+      "grad_norm": 3.65234375,
+      "learning_rate": 4.437625083388926e-05,
+      "loss": 0.9813,
+      "step": 8348
+    },
+    {
+      "epoch": 7.707362104777291,
+      "grad_norm": 3.5390625,
+      "learning_rate": 4.436957971981321e-05,
+      "loss": 1.2993,
+      "step": 8349
+    },
+    {
+      "epoch": 7.708285252711747,
+      "grad_norm": 3.0078125,
+      "learning_rate": 4.436290860573716e-05,
+      "loss": 1.1926,
+      "step": 8350
+    },
+    {
+      "epoch": 7.709208400646204,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.435623749166111e-05,
+      "loss": 1.0612,
+      "step": 8351
+    },
+    {
+      "epoch": 7.7101315485806605,
+      "grad_norm": 3.416015625,
+      "learning_rate": 4.434956637758506e-05,
+      "loss": 1.3188,
+      "step": 8352
+    },
+    {
+      "epoch": 7.7110546965151165,
+      "grad_norm": 3.927734375,
+      "learning_rate": 4.434289526350901e-05,
+      "loss": 1.3779,
+      "step": 8353
+    },
+    {
+      "epoch": 7.711977844449573,
+      "grad_norm": 3.998046875,
+      "learning_rate": 4.433622414943296e-05,
+      "loss": 1.4944,
+      "step": 8354
+    },
+    {
+      "epoch": 7.7129009923840295,
+      "grad_norm": 3.001953125,
+      "learning_rate": 4.4329553035356905e-05,
+      "loss": 0.9866,
+      "step": 8355
+    },
+    {
+      "epoch": 7.713824140318486,
+      "grad_norm": 3.455078125,
+      "learning_rate": 4.432288192128086e-05,
+      "loss": 1.4399,
+      "step": 8356
+    },
+    {
+      "epoch": 7.714747288252942,
+      "grad_norm": 3.76171875,
+      "learning_rate": 4.43162108072048e-05,
+      "loss": 1.257,
+      "step": 8357
+    },
+    {
+      "epoch": 7.715670436187399,
+      "grad_norm": 3.244140625,
+      "learning_rate": 4.4309539693128755e-05,
+      "loss": 1.4701,
+      "step": 8358
+    },
+    {
+      "epoch": 7.716593584121855,
+      "grad_norm": 3.6328125,
+      "learning_rate": 4.43028685790527e-05,
+      "loss": 1.3799,
+      "step": 8359
+    },
+    {
+      "epoch": 7.717516732056312,
+      "grad_norm": 3.57421875,
+      "learning_rate": 4.429619746497665e-05,
+      "loss": 1.1134,
+      "step": 8360
+    },
+    {
+      "epoch": 7.718439879990768,
+      "grad_norm": 3.78515625,
+      "learning_rate": 4.42895263509006e-05,
+      "loss": 1.2822,
+      "step": 8361
+    },
+    {
+      "epoch": 7.719363027925225,
+      "grad_norm": 2.8515625,
+      "learning_rate": 4.428285523682455e-05,
+      "loss": 1.297,
+      "step": 8362
+    },
+    {
+      "epoch": 7.720286175859681,
+      "grad_norm": 4.0078125,
+      "learning_rate": 4.4276184122748495e-05,
+      "loss": 1.4142,
+      "step": 8363
+    },
+    {
+      "epoch": 7.721209323794138,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.4269513008672454e-05,
+      "loss": 1.3959,
+      "step": 8364
+    },
+    {
+      "epoch": 7.722132471728594,
+      "grad_norm": 3.90234375,
+      "learning_rate": 4.42628418945964e-05,
+      "loss": 1.3423,
+      "step": 8365
+    },
+    {
+      "epoch": 7.723055619663051,
+      "grad_norm": 3.396484375,
+      "learning_rate": 4.425617078052035e-05,
+      "loss": 1.3323,
+      "step": 8366
+    },
+    {
+      "epoch": 7.723978767597507,
+      "grad_norm": 3.177734375,
+      "learning_rate": 4.42494996664443e-05,
+      "loss": 1.2539,
+      "step": 8367
+    },
+    {
+      "epoch": 7.724901915531964,
+      "grad_norm": 3.658203125,
+      "learning_rate": 4.424282855236825e-05,
+      "loss": 1.5474,
+      "step": 8368
+    },
+    {
+      "epoch": 7.72582506346642,
+      "grad_norm": 3.37890625,
+      "learning_rate": 4.4236157438292195e-05,
+      "loss": 1.2866,
+      "step": 8369
+    },
+    {
+      "epoch": 7.726748211400877,
+      "grad_norm": 3.05859375,
+      "learning_rate": 4.422948632421615e-05,
+      "loss": 1.4941,
+      "step": 8370
+    },
+    {
+      "epoch": 7.727671359335334,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.422281521014009e-05,
+      "loss": 1.6001,
+      "step": 8371
+    },
+    {
+      "epoch": 7.72859450726979,
+      "grad_norm": 2.947265625,
+      "learning_rate": 4.4216144096064045e-05,
+      "loss": 1.4353,
+      "step": 8372
+    },
+    {
+      "epoch": 7.729517655204247,
+      "grad_norm": 2.96484375,
+      "learning_rate": 4.420947298198799e-05,
+      "loss": 1.1917,
+      "step": 8373
+    },
+    {
+      "epoch": 7.730440803138703,
+      "grad_norm": 3.369140625,
+      "learning_rate": 4.420280186791194e-05,
+      "loss": 1.1171,
+      "step": 8374
+    },
+    {
+      "epoch": 7.73136395107316,
+      "grad_norm": 4.0078125,
+      "learning_rate": 4.419613075383589e-05,
+      "loss": 1.4458,
+      "step": 8375
+    },
+    {
+      "epoch": 7.732287099007616,
+      "grad_norm": 3.318359375,
+      "learning_rate": 4.418945963975984e-05,
+      "loss": 1.3193,
+      "step": 8376
+    },
+    {
+      "epoch": 7.733210246942073,
+      "grad_norm": 3.8984375,
+      "learning_rate": 4.418278852568379e-05,
+      "loss": 1.2986,
+      "step": 8377
+    },
+    {
+      "epoch": 7.734133394876529,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.4176117411607745e-05,
+      "loss": 1.4102,
+      "step": 8378
+    },
+    {
+      "epoch": 7.735056542810986,
+      "grad_norm": 3.99609375,
+      "learning_rate": 4.416944629753169e-05,
+      "loss": 1.1422,
+      "step": 8379
+    },
+    {
+      "epoch": 7.735979690745442,
+      "grad_norm": 3.35546875,
+      "learning_rate": 4.416277518345564e-05,
+      "loss": 1.116,
+      "step": 8380
+    },
+    {
+      "epoch": 7.736902838679899,
+      "grad_norm": 3.10546875,
+      "learning_rate": 4.415610406937959e-05,
+      "loss": 1.2661,
+      "step": 8381
+    },
+    {
+      "epoch": 7.737825986614355,
+      "grad_norm": 3.357421875,
+      "learning_rate": 4.414943295530354e-05,
+      "loss": 1.254,
+      "step": 8382
+    },
+    {
+      "epoch": 7.7387491345488115,
+      "grad_norm": 3.212890625,
+      "learning_rate": 4.4142761841227485e-05,
+      "loss": 1.4561,
+      "step": 8383
+    },
+    {
+      "epoch": 7.7396722824832676,
+      "grad_norm": 3.673828125,
+      "learning_rate": 4.413609072715144e-05,
+      "loss": 1.4978,
+      "step": 8384
+    },
+    {
+      "epoch": 7.7405954304177245,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.412941961307538e-05,
+      "loss": 1.3486,
+      "step": 8385
+    },
+    {
+      "epoch": 7.741518578352181,
+      "grad_norm": 3.505859375,
+      "learning_rate": 4.4122748498999335e-05,
+      "loss": 1.5247,
+      "step": 8386
+    },
+    {
+      "epoch": 7.742441726286637,
+      "grad_norm": 3.353515625,
+      "learning_rate": 4.411607738492328e-05,
+      "loss": 1.4277,
+      "step": 8387
+    },
+    {
+      "epoch": 7.743364874221094,
+      "grad_norm": 3.0859375,
+      "learning_rate": 4.410940627084723e-05,
+      "loss": 1.1168,
+      "step": 8388
+    },
+    {
+      "epoch": 7.74428802215555,
+      "grad_norm": 2.92578125,
+      "learning_rate": 4.4102735156771185e-05,
+      "loss": 1.5847,
+      "step": 8389
+    },
+    {
+      "epoch": 7.745211170090007,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.409606404269514e-05,
+      "loss": 1.3298,
+      "step": 8390
+    },
+    {
+      "epoch": 7.746134318024463,
+      "grad_norm": 3.2109375,
+      "learning_rate": 4.408939292861908e-05,
+      "loss": 1.4658,
+      "step": 8391
+    },
+    {
+      "epoch": 7.74705746595892,
+      "grad_norm": 3.751953125,
+      "learning_rate": 4.4082721814543035e-05,
+      "loss": 1.3821,
+      "step": 8392
+    },
+    {
+      "epoch": 7.747980613893376,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.407605070046698e-05,
+      "loss": 1.3582,
+      "step": 8393
+    },
+    {
+      "epoch": 7.748903761827833,
+      "grad_norm": 3.16796875,
+      "learning_rate": 4.406937958639093e-05,
+      "loss": 1.0482,
+      "step": 8394
+    },
+    {
+      "epoch": 7.749826909762289,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.406270847231488e-05,
+      "loss": 1.3862,
+      "step": 8395
+    },
+    {
+      "epoch": 7.750750057696746,
+      "grad_norm": 3.01953125,
+      "learning_rate": 4.405603735823883e-05,
+      "loss": 1.2876,
+      "step": 8396
+    },
+    {
+      "epoch": 7.751673205631202,
+      "grad_norm": 3.173828125,
+      "learning_rate": 4.4049366244162776e-05,
+      "loss": 1.1794,
+      "step": 8397
+    },
+    {
+      "epoch": 7.752596353565659,
+      "grad_norm": 3.67578125,
+      "learning_rate": 4.404269513008673e-05,
+      "loss": 1.267,
+      "step": 8398
+    },
+    {
+      "epoch": 7.753519501500115,
+      "grad_norm": 3.306640625,
+      "learning_rate": 4.403602401601067e-05,
+      "loss": 1.2388,
+      "step": 8399
+    },
+    {
+      "epoch": 7.754442649434572,
+      "grad_norm": 3.7265625,
+      "learning_rate": 4.4029352901934625e-05,
+      "loss": 1.2429,
+      "step": 8400
+    },
+    {
+      "epoch": 7.755365797369029,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.402268178785857e-05,
+      "loss": 1.1538,
+      "step": 8401
+    },
+    {
+      "epoch": 7.756288945303485,
+      "grad_norm": 2.99609375,
+      "learning_rate": 4.401601067378252e-05,
+      "loss": 1.4329,
+      "step": 8402
+    },
+    {
+      "epoch": 7.757212093237941,
+      "grad_norm": 3.54296875,
+      "learning_rate": 4.4009339559706475e-05,
+      "loss": 1.4434,
+      "step": 8403
+    },
+    {
+      "epoch": 7.758135241172398,
+      "grad_norm": 3.494140625,
+      "learning_rate": 4.400266844563042e-05,
+      "loss": 1.4065,
+      "step": 8404
+    },
+    {
+      "epoch": 7.759058389106855,
+      "grad_norm": 3.341796875,
+      "learning_rate": 4.399599733155437e-05,
+      "loss": 1.2079,
+      "step": 8405
+    },
+    {
+      "epoch": 7.759981537041311,
+      "grad_norm": 2.998046875,
+      "learning_rate": 4.398932621747832e-05,
+      "loss": 1.4973,
+      "step": 8406
+    },
+    {
+      "epoch": 7.760904684975768,
+      "grad_norm": 3.052734375,
+      "learning_rate": 4.398265510340227e-05,
+      "loss": 1.5833,
+      "step": 8407
+    },
+    {
+      "epoch": 7.761827832910224,
+      "grad_norm": 3.54296875,
+      "learning_rate": 4.3975983989326216e-05,
+      "loss": 1.2788,
+      "step": 8408
+    },
+    {
+      "epoch": 7.762750980844681,
+      "grad_norm": 3.357421875,
+      "learning_rate": 4.396931287525017e-05,
+      "loss": 1.3322,
+      "step": 8409
+    },
+    {
+      "epoch": 7.763674128779137,
+      "grad_norm": 3.228515625,
+      "learning_rate": 4.396264176117412e-05,
+      "loss": 1.4539,
+      "step": 8410
+    },
+    {
+      "epoch": 7.764597276713594,
+      "grad_norm": 3.365234375,
+      "learning_rate": 4.3955970647098066e-05,
+      "loss": 1.6023,
+      "step": 8411
+    },
+    {
+      "epoch": 7.76552042464805,
+      "grad_norm": 3.189453125,
+      "learning_rate": 4.394929953302202e-05,
+      "loss": 1.3636,
+      "step": 8412
+    },
+    {
+      "epoch": 7.7664435725825065,
+      "grad_norm": 4.0,
+      "learning_rate": 4.3942628418945963e-05,
+      "loss": 1.3208,
+      "step": 8413
+    },
+    {
+      "epoch": 7.7673667205169625,
+      "grad_norm": 3.361328125,
+      "learning_rate": 4.3935957304869916e-05,
+      "loss": 1.174,
+      "step": 8414
+    },
+    {
+      "epoch": 7.7682898684514194,
+      "grad_norm": 3.134765625,
+      "learning_rate": 4.392928619079387e-05,
+      "loss": 1.5559,
+      "step": 8415
+    },
+    {
+      "epoch": 7.7692130163858755,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.392261507671781e-05,
+      "loss": 1.6096,
+      "step": 8416
+    },
+    {
+      "epoch": 7.770136164320332,
+      "grad_norm": 3.48828125,
+      "learning_rate": 4.3915943962641765e-05,
+      "loss": 1.3901,
+      "step": 8417
+    },
+    {
+      "epoch": 7.771059312254788,
+      "grad_norm": 3.455078125,
+      "learning_rate": 4.390927284856571e-05,
+      "loss": 1.3174,
+      "step": 8418
+    },
+    {
+      "epoch": 7.771982460189245,
+      "grad_norm": 3.1796875,
+      "learning_rate": 4.390260173448966e-05,
+      "loss": 1.5115,
+      "step": 8419
+    },
+    {
+      "epoch": 7.772905608123702,
+      "grad_norm": 3.7265625,
+      "learning_rate": 4.389593062041361e-05,
+      "loss": 1.3047,
+      "step": 8420
+    },
+    {
+      "epoch": 7.773828756058158,
+      "grad_norm": 3.408203125,
+      "learning_rate": 4.388925950633756e-05,
+      "loss": 1.325,
+      "step": 8421
+    },
+    {
+      "epoch": 7.774751903992615,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.3882588392261506e-05,
+      "loss": 1.2947,
+      "step": 8422
+    },
+    {
+      "epoch": 7.775675051927071,
+      "grad_norm": 3.42578125,
+      "learning_rate": 4.387591727818546e-05,
+      "loss": 1.2849,
+      "step": 8423
+    },
+    {
+      "epoch": 7.776598199861528,
+      "grad_norm": 3.994140625,
+      "learning_rate": 4.3869246164109404e-05,
+      "loss": 1.4114,
+      "step": 8424
+    },
+    {
+      "epoch": 7.777521347795984,
+      "grad_norm": 3.2109375,
+      "learning_rate": 4.3862575050033356e-05,
+      "loss": 1.2711,
+      "step": 8425
+    },
+    {
+      "epoch": 7.778444495730441,
+      "grad_norm": 3.615234375,
+      "learning_rate": 4.38559039359573e-05,
+      "loss": 1.355,
+      "step": 8426
+    },
+    {
+      "epoch": 7.779367643664897,
+      "grad_norm": 2.935546875,
+      "learning_rate": 4.384923282188126e-05,
+      "loss": 1.3793,
+      "step": 8427
+    },
+    {
+      "epoch": 7.780290791599354,
+      "grad_norm": 3.451171875,
+      "learning_rate": 4.3842561707805206e-05,
+      "loss": 1.4324,
+      "step": 8428
+    },
+    {
+      "epoch": 7.78121393953381,
+      "grad_norm": 2.998046875,
+      "learning_rate": 4.383589059372916e-05,
+      "loss": 1.551,
+      "step": 8429
+    },
+    {
+      "epoch": 7.782137087468267,
+      "grad_norm": 4.08984375,
+      "learning_rate": 4.3829219479653104e-05,
+      "loss": 1.4885,
+      "step": 8430
+    },
+    {
+      "epoch": 7.783060235402723,
+      "grad_norm": 2.9453125,
+      "learning_rate": 4.3822548365577056e-05,
+      "loss": 1.6221,
+      "step": 8431
+    },
+    {
+      "epoch": 7.78398338333718,
+      "grad_norm": 4.19921875,
+      "learning_rate": 4.3815877251501e-05,
+      "loss": 1.3865,
+      "step": 8432
+    },
+    {
+      "epoch": 7.784906531271636,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.380920613742495e-05,
+      "loss": 1.2715,
+      "step": 8433
+    },
+    {
+      "epoch": 7.785829679206093,
+      "grad_norm": 3.994140625,
+      "learning_rate": 4.38025350233489e-05,
+      "loss": 1.3011,
+      "step": 8434
+    },
+    {
+      "epoch": 7.78675282714055,
+      "grad_norm": 4.18359375,
+      "learning_rate": 4.379586390927285e-05,
+      "loss": 1.5437,
+      "step": 8435
+    },
+    {
+      "epoch": 7.787675975075006,
+      "grad_norm": 2.7734375,
+      "learning_rate": 4.3789192795196796e-05,
+      "loss": 1.3546,
+      "step": 8436
+    },
+    {
+      "epoch": 7.788599123009462,
+      "grad_norm": 3.642578125,
+      "learning_rate": 4.378252168112075e-05,
+      "loss": 1.3779,
+      "step": 8437
+    },
+    {
+      "epoch": 7.789522270943919,
+      "grad_norm": 3.037109375,
+      "learning_rate": 4.3775850567044694e-05,
+      "loss": 1.3135,
+      "step": 8438
+    },
+    {
+      "epoch": 7.790445418878376,
+      "grad_norm": 3.123046875,
+      "learning_rate": 4.376917945296865e-05,
+      "loss": 1.3821,
+      "step": 8439
+    },
+    {
+      "epoch": 7.791368566812832,
+      "grad_norm": 3.73828125,
+      "learning_rate": 4.37625083388926e-05,
+      "loss": 1.3557,
+      "step": 8440
+    },
+    {
+      "epoch": 7.7922917147472885,
+      "grad_norm": 4.05078125,
+      "learning_rate": 4.375583722481655e-05,
+      "loss": 0.9692,
+      "step": 8441
+    },
+    {
+      "epoch": 7.793214862681745,
+      "grad_norm": 2.97265625,
+      "learning_rate": 4.3749166110740496e-05,
+      "loss": 1.363,
+      "step": 8442
+    },
+    {
+      "epoch": 7.7941380106162015,
+      "grad_norm": 3.275390625,
+      "learning_rate": 4.374249499666445e-05,
+      "loss": 0.9971,
+      "step": 8443
+    },
+    {
+      "epoch": 7.7950611585506575,
+      "grad_norm": 3.00390625,
+      "learning_rate": 4.3735823882588394e-05,
+      "loss": 1.3164,
+      "step": 8444
+    },
+    {
+      "epoch": 7.795984306485114,
+      "grad_norm": 3.75,
+      "learning_rate": 4.3729152768512346e-05,
+      "loss": 1.366,
+      "step": 8445
+    },
+    {
+      "epoch": 7.7969074544195704,
+      "grad_norm": 3.337890625,
+      "learning_rate": 4.372248165443629e-05,
+      "loss": 1.4246,
+      "step": 8446
+    },
+    {
+      "epoch": 7.797830602354027,
+      "grad_norm": 3.529296875,
+      "learning_rate": 4.3715810540360244e-05,
+      "loss": 1.4827,
+      "step": 8447
+    },
+    {
+      "epoch": 7.798753750288483,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.370913942628419e-05,
+      "loss": 1.5715,
+      "step": 8448
+    },
+    {
+      "epoch": 7.79967689822294,
+      "grad_norm": 3.40234375,
+      "learning_rate": 4.370246831220814e-05,
+      "loss": 1.2888,
+      "step": 8449
+    },
+    {
+      "epoch": 7.800600046157397,
+      "grad_norm": 3.353515625,
+      "learning_rate": 4.369579719813209e-05,
+      "loss": 1.2415,
+      "step": 8450
+    },
+    {
+      "epoch": 7.801523194091853,
+      "grad_norm": 3.189453125,
+      "learning_rate": 4.368912608405604e-05,
+      "loss": 1.5215,
+      "step": 8451
+    },
+    {
+      "epoch": 7.802446342026309,
+      "grad_norm": 3.470703125,
+      "learning_rate": 4.368245496997999e-05,
+      "loss": 1.4297,
+      "step": 8452
+    },
+    {
+      "epoch": 7.803369489960766,
+      "grad_norm": 3.92578125,
+      "learning_rate": 4.367578385590394e-05,
+      "loss": 1.4025,
+      "step": 8453
+    },
+    {
+      "epoch": 7.804292637895223,
+      "grad_norm": 2.9921875,
+      "learning_rate": 4.366911274182789e-05,
+      "loss": 1.1276,
+      "step": 8454
+    },
+    {
+      "epoch": 7.805215785829679,
+      "grad_norm": 3.642578125,
+      "learning_rate": 4.366244162775184e-05,
+      "loss": 1.2759,
+      "step": 8455
+    },
+    {
+      "epoch": 7.806138933764136,
+      "grad_norm": 3.93359375,
+      "learning_rate": 4.3655770513675786e-05,
+      "loss": 1.1719,
+      "step": 8456
+    },
+    {
+      "epoch": 7.807062081698592,
+      "grad_norm": 3.884765625,
+      "learning_rate": 4.364909939959974e-05,
+      "loss": 1.4348,
+      "step": 8457
+    },
+    {
+      "epoch": 7.807985229633049,
+      "grad_norm": 3.599609375,
+      "learning_rate": 4.3642428285523684e-05,
+      "loss": 1.1465,
+      "step": 8458
+    },
+    {
+      "epoch": 7.808908377567505,
+      "grad_norm": 3.33203125,
+      "learning_rate": 4.3635757171447636e-05,
+      "loss": 1.2788,
+      "step": 8459
+    },
+    {
+      "epoch": 7.809831525501962,
+      "grad_norm": 3.29296875,
+      "learning_rate": 4.362908605737158e-05,
+      "loss": 1.136,
+      "step": 8460
+    },
+    {
+      "epoch": 7.810754673436418,
+      "grad_norm": 3.982421875,
+      "learning_rate": 4.3622414943295534e-05,
+      "loss": 1.1792,
+      "step": 8461
+    },
+    {
+      "epoch": 7.811677821370875,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.361574382921948e-05,
+      "loss": 1.4043,
+      "step": 8462
+    },
+    {
+      "epoch": 7.812600969305331,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.360907271514343e-05,
+      "loss": 1.2898,
+      "step": 8463
+    },
+    {
+      "epoch": 7.813524117239788,
+      "grad_norm": 3.255859375,
+      "learning_rate": 4.3602401601067384e-05,
+      "loss": 0.9893,
+      "step": 8464
+    },
+    {
+      "epoch": 7.814447265174244,
+      "grad_norm": 3.3828125,
+      "learning_rate": 4.359573048699133e-05,
+      "loss": 1.2471,
+      "step": 8465
+    },
+    {
+      "epoch": 7.815370413108701,
+      "grad_norm": 3.158203125,
+      "learning_rate": 4.358905937291528e-05,
+      "loss": 1.3716,
+      "step": 8466
+    },
+    {
+      "epoch": 7.816293561043157,
+      "grad_norm": 3.88671875,
+      "learning_rate": 4.358238825883923e-05,
+      "loss": 1.4524,
+      "step": 8467
+    },
+    {
+      "epoch": 7.817216708977614,
+      "grad_norm": 3.162109375,
+      "learning_rate": 4.357571714476318e-05,
+      "loss": 1.1853,
+      "step": 8468
+    },
+    {
+      "epoch": 7.818139856912071,
+      "grad_norm": 3.19921875,
+      "learning_rate": 4.3569046030687124e-05,
+      "loss": 1.3572,
+      "step": 8469
+    },
+    {
+      "epoch": 7.819063004846527,
+      "grad_norm": 3.619140625,
+      "learning_rate": 4.3562374916611077e-05,
+      "loss": 0.6434,
+      "step": 8470
+    },
+    {
+      "epoch": 7.8199861527809835,
+      "grad_norm": 3.845703125,
+      "learning_rate": 4.355570380253502e-05,
+      "loss": 1.6414,
+      "step": 8471
+    },
+    {
+      "epoch": 7.8209093007154395,
+      "grad_norm": 2.728515625,
+      "learning_rate": 4.3549032688458974e-05,
+      "loss": 0.789,
+      "step": 8472
+    },
+    {
+      "epoch": 7.8218324486498965,
+      "grad_norm": 3.400390625,
+      "learning_rate": 4.354236157438292e-05,
+      "loss": 1.592,
+      "step": 8473
+    },
+    {
+      "epoch": 7.8227555965843525,
+      "grad_norm": 3.529296875,
+      "learning_rate": 4.353569046030687e-05,
+      "loss": 1.4194,
+      "step": 8474
+    },
+    {
+      "epoch": 7.823678744518809,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.352901934623082e-05,
+      "loss": 1.4578,
+      "step": 8475
+    },
+    {
+      "epoch": 7.824601892453265,
+      "grad_norm": 4.49609375,
+      "learning_rate": 4.352234823215477e-05,
+      "loss": 1.209,
+      "step": 8476
+    },
+    {
+      "epoch": 7.825525040387722,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.351567711807872e-05,
+      "loss": 1.3635,
+      "step": 8477
+    },
+    {
+      "epoch": 7.826448188322178,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.3509006004002674e-05,
+      "loss": 1.4868,
+      "step": 8478
+    },
+    {
+      "epoch": 7.827371336256635,
+      "grad_norm": 3.205078125,
+      "learning_rate": 4.350233488992662e-05,
+      "loss": 1.2806,
+      "step": 8479
+    },
+    {
+      "epoch": 7.828294484191091,
+      "grad_norm": 3.130859375,
+      "learning_rate": 4.349566377585057e-05,
+      "loss": 1.2522,
+      "step": 8480
+    },
+    {
+      "epoch": 7.829217632125548,
+      "grad_norm": 3.125,
+      "learning_rate": 4.348899266177452e-05,
+      "loss": 1.2139,
+      "step": 8481
+    },
+    {
+      "epoch": 7.830140780060004,
+      "grad_norm": 3.39453125,
+      "learning_rate": 4.348232154769847e-05,
+      "loss": 1.2847,
+      "step": 8482
+    },
+    {
+      "epoch": 7.831063927994461,
+      "grad_norm": 3.5234375,
+      "learning_rate": 4.3475650433622415e-05,
+      "loss": 1.3806,
+      "step": 8483
+    },
+    {
+      "epoch": 7.831987075928918,
+      "grad_norm": 3.62890625,
+      "learning_rate": 4.346897931954637e-05,
+      "loss": 1.3007,
+      "step": 8484
+    },
+    {
+      "epoch": 7.832910223863374,
+      "grad_norm": 2.982421875,
+      "learning_rate": 4.346230820547031e-05,
+      "loss": 1.4663,
+      "step": 8485
+    },
+    {
+      "epoch": 7.83383337179783,
+      "grad_norm": 3.63671875,
+      "learning_rate": 4.3455637091394264e-05,
+      "loss": 1.1742,
+      "step": 8486
+    },
+    {
+      "epoch": 7.834756519732287,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.344896597731821e-05,
+      "loss": 1.5974,
+      "step": 8487
+    },
+    {
+      "epoch": 7.835679667666744,
+      "grad_norm": 3.505859375,
+      "learning_rate": 4.344229486324216e-05,
+      "loss": 1.2224,
+      "step": 8488
+    },
+    {
+      "epoch": 7.8366028156012,
+      "grad_norm": 3.28515625,
+      "learning_rate": 4.3435623749166114e-05,
+      "loss": 1.2344,
+      "step": 8489
+    },
+    {
+      "epoch": 7.837525963535657,
+      "grad_norm": 3.283203125,
+      "learning_rate": 4.3428952635090067e-05,
+      "loss": 1.2637,
+      "step": 8490
+    },
+    {
+      "epoch": 7.838449111470113,
+      "grad_norm": 3.46484375,
+      "learning_rate": 4.342228152101401e-05,
+      "loss": 1.2047,
+      "step": 8491
+    },
+    {
+      "epoch": 7.83937225940457,
+      "grad_norm": 3.24609375,
+      "learning_rate": 4.3415610406937964e-05,
+      "loss": 1.1505,
+      "step": 8492
+    },
+    {
+      "epoch": 7.840295407339026,
+      "grad_norm": 3.619140625,
+      "learning_rate": 4.340893929286191e-05,
+      "loss": 1.3018,
+      "step": 8493
+    },
+    {
+      "epoch": 7.841218555273483,
+      "grad_norm": 3.30859375,
+      "learning_rate": 4.340226817878586e-05,
+      "loss": 1.438,
+      "step": 8494
+    },
+    {
+      "epoch": 7.842141703207939,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.339559706470981e-05,
+      "loss": 1.4504,
+      "step": 8495
+    },
+    {
+      "epoch": 7.843064851142396,
+      "grad_norm": 3.849609375,
+      "learning_rate": 4.338892595063376e-05,
+      "loss": 1.3616,
+      "step": 8496
+    },
+    {
+      "epoch": 7.843987999076852,
+      "grad_norm": 3.9296875,
+      "learning_rate": 4.3382254836557705e-05,
+      "loss": 1.3462,
+      "step": 8497
+    },
+    {
+      "epoch": 7.844911147011309,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.337558372248166e-05,
+      "loss": 1.5608,
+      "step": 8498
+    },
+    {
+      "epoch": 7.845834294945765,
+      "grad_norm": 3.88671875,
+      "learning_rate": 4.33689126084056e-05,
+      "loss": 1.6616,
+      "step": 8499
+    },
+    {
+      "epoch": 7.846757442880222,
+      "grad_norm": 3.21484375,
+      "learning_rate": 4.3362241494329555e-05,
+      "loss": 0.9882,
+      "step": 8500
+    },
+    {
+      "epoch": 7.847680590814678,
+      "grad_norm": 3.005859375,
+      "learning_rate": 4.33555703802535e-05,
+      "loss": 1.3828,
+      "step": 8501
+    },
+    {
+      "epoch": 7.8486037387491345,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.334889926617746e-05,
+      "loss": 1.4189,
+      "step": 8502
+    },
+    {
+      "epoch": 7.849526886683591,
+      "grad_norm": 3.4921875,
+      "learning_rate": 4.3342228152101405e-05,
+      "loss": 1.4612,
+      "step": 8503
+    },
+    {
+      "epoch": 7.8504500346180475,
+      "grad_norm": 4.015625,
+      "learning_rate": 4.333555703802536e-05,
+      "loss": 1.0648,
+      "step": 8504
+    },
+    {
+      "epoch": 7.851373182552504,
+      "grad_norm": 3.2265625,
+      "learning_rate": 4.33288859239493e-05,
+      "loss": 1.0757,
+      "step": 8505
+    },
+    {
+      "epoch": 7.85229633048696,
+      "grad_norm": 3.42578125,
+      "learning_rate": 4.3322214809873254e-05,
+      "loss": 1.4087,
+      "step": 8506
+    },
+    {
+      "epoch": 7.853219478421417,
+      "grad_norm": 3.33203125,
+      "learning_rate": 4.33155436957972e-05,
+      "loss": 1.4305,
+      "step": 8507
+    },
+    {
+      "epoch": 7.854142626355873,
+      "grad_norm": 3.15234375,
+      "learning_rate": 4.330887258172115e-05,
+      "loss": 0.8625,
+      "step": 8508
+    },
+    {
+      "epoch": 7.85506577429033,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.33022014676451e-05,
+      "loss": 1.4666,
+      "step": 8509
+    },
+    {
+      "epoch": 7.855988922224786,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.329553035356905e-05,
+      "loss": 1.244,
+      "step": 8510
+    },
+    {
+      "epoch": 7.856912070159243,
+      "grad_norm": 3.673828125,
+      "learning_rate": 4.3288859239492995e-05,
+      "loss": 1.2769,
+      "step": 8511
+    },
+    {
+      "epoch": 7.857835218093699,
+      "grad_norm": 3.0703125,
+      "learning_rate": 4.328218812541695e-05,
+      "loss": 1.3875,
+      "step": 8512
+    },
+    {
+      "epoch": 7.858758366028156,
+      "grad_norm": 3.3203125,
+      "learning_rate": 4.327551701134089e-05,
+      "loss": 1.235,
+      "step": 8513
+    },
+    {
+      "epoch": 7.859681513962612,
+      "grad_norm": 3.888671875,
+      "learning_rate": 4.3268845897264845e-05,
+      "loss": 1.4197,
+      "step": 8514
+    },
+    {
+      "epoch": 7.860604661897069,
+      "grad_norm": 3.24609375,
+      "learning_rate": 4.32621747831888e-05,
+      "loss": 1.3027,
+      "step": 8515
+    },
+    {
+      "epoch": 7.861527809831525,
+      "grad_norm": 3.341796875,
+      "learning_rate": 4.325550366911274e-05,
+      "loss": 1.5977,
+      "step": 8516
+    },
+    {
+      "epoch": 7.862450957765982,
+      "grad_norm": 3.755859375,
+      "learning_rate": 4.3248832555036695e-05,
+      "loss": 1.3704,
+      "step": 8517
+    },
+    {
+      "epoch": 7.863374105700439,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.324216144096064e-05,
+      "loss": 1.1377,
+      "step": 8518
+    },
+    {
+      "epoch": 7.864297253634895,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.323549032688459e-05,
+      "loss": 1.2129,
+      "step": 8519
+    },
+    {
+      "epoch": 7.865220401569352,
+      "grad_norm": 3.29296875,
+      "learning_rate": 4.322881921280854e-05,
+      "loss": 1.4136,
+      "step": 8520
+    },
+    {
+      "epoch": 7.866143549503808,
+      "grad_norm": 3.00390625,
+      "learning_rate": 4.322214809873249e-05,
+      "loss": 1.1428,
+      "step": 8521
+    },
+    {
+      "epoch": 7.867066697438265,
+      "grad_norm": 3.25,
+      "learning_rate": 4.321547698465644e-05,
+      "loss": 1.1935,
+      "step": 8522
+    },
+    {
+      "epoch": 7.867989845372721,
+      "grad_norm": 3.5,
+      "learning_rate": 4.320880587058039e-05,
+      "loss": 1.1641,
+      "step": 8523
+    },
+    {
+      "epoch": 7.868912993307178,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.320213475650434e-05,
+      "loss": 1.3198,
+      "step": 8524
+    },
+    {
+      "epoch": 7.869836141241634,
+      "grad_norm": 3.814453125,
+      "learning_rate": 4.3195463642428285e-05,
+      "loss": 1.3787,
+      "step": 8525
+    },
+    {
+      "epoch": 7.870759289176091,
+      "grad_norm": 3.603515625,
+      "learning_rate": 4.318879252835224e-05,
+      "loss": 1.7542,
+      "step": 8526
+    },
+    {
+      "epoch": 7.871682437110547,
+      "grad_norm": 3.591796875,
+      "learning_rate": 4.318212141427619e-05,
+      "loss": 1.3735,
+      "step": 8527
+    },
+    {
+      "epoch": 7.872605585045004,
+      "grad_norm": 4.05859375,
+      "learning_rate": 4.3175450300200135e-05,
+      "loss": 1.5004,
+      "step": 8528
+    },
+    {
+      "epoch": 7.87352873297946,
+      "grad_norm": 3.2421875,
+      "learning_rate": 4.316877918612409e-05,
+      "loss": 1.28,
+      "step": 8529
+    },
+    {
+      "epoch": 7.874451880913917,
+      "grad_norm": 3.3515625,
+      "learning_rate": 4.316210807204803e-05,
+      "loss": 1.0334,
+      "step": 8530
+    },
+    {
+      "epoch": 7.875375028848373,
+      "grad_norm": 3.15234375,
+      "learning_rate": 4.3155436957971985e-05,
+      "loss": 1.2383,
+      "step": 8531
+    },
+    {
+      "epoch": 7.8762981767828295,
+      "grad_norm": 3.15234375,
+      "learning_rate": 4.314876584389593e-05,
+      "loss": 1.1909,
+      "step": 8532
+    },
+    {
+      "epoch": 7.877221324717286,
+      "grad_norm": 3.6328125,
+      "learning_rate": 4.314209472981988e-05,
+      "loss": 1.3684,
+      "step": 8533
+    },
+    {
+      "epoch": 7.8781444726517424,
+      "grad_norm": 3.1796875,
+      "learning_rate": 4.313542361574383e-05,
+      "loss": 1.4023,
+      "step": 8534
+    },
+    {
+      "epoch": 7.8790676205861985,
+      "grad_norm": 3.826171875,
+      "learning_rate": 4.312875250166778e-05,
+      "loss": 1.2183,
+      "step": 8535
+    },
+    {
+      "epoch": 7.879990768520655,
+      "grad_norm": 3.69140625,
+      "learning_rate": 4.3122081387591726e-05,
+      "loss": 1.0259,
+      "step": 8536
+    },
+    {
+      "epoch": 7.880913916455112,
+      "grad_norm": 4.0078125,
+      "learning_rate": 4.311541027351568e-05,
+      "loss": 1.5166,
+      "step": 8537
+    },
+    {
+      "epoch": 7.881837064389568,
+      "grad_norm": 3.12109375,
+      "learning_rate": 4.310873915943962e-05,
+      "loss": 1.171,
+      "step": 8538
+    },
+    {
+      "epoch": 7.882760212324025,
+      "grad_norm": 3.2109375,
+      "learning_rate": 4.3102068045363576e-05,
+      "loss": 1.3014,
+      "step": 8539
+    },
+    {
+      "epoch": 7.883683360258481,
+      "grad_norm": 3.041015625,
+      "learning_rate": 4.309539693128753e-05,
+      "loss": 1.0382,
+      "step": 8540
+    },
+    {
+      "epoch": 7.884606508192938,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.308872581721148e-05,
+      "loss": 1.408,
+      "step": 8541
+    },
+    {
+      "epoch": 7.885529656127394,
+      "grad_norm": 2.84765625,
+      "learning_rate": 4.3082054703135425e-05,
+      "loss": 1.3594,
+      "step": 8542
+    },
+    {
+      "epoch": 7.886452804061851,
+      "grad_norm": 3.466796875,
+      "learning_rate": 4.307538358905938e-05,
+      "loss": 1.4802,
+      "step": 8543
+    },
+    {
+      "epoch": 7.887375951996307,
+      "grad_norm": 3.392578125,
+      "learning_rate": 4.306871247498332e-05,
+      "loss": 1.3088,
+      "step": 8544
+    },
+    {
+      "epoch": 7.888299099930764,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.3062041360907275e-05,
+      "loss": 1.5662,
+      "step": 8545
+    },
+    {
+      "epoch": 7.88922224786522,
+      "grad_norm": 3.802734375,
+      "learning_rate": 4.305537024683122e-05,
+      "loss": 1.3376,
+      "step": 8546
+    },
+    {
+      "epoch": 7.890145395799677,
+      "grad_norm": 3.556640625,
+      "learning_rate": 4.304869913275517e-05,
+      "loss": 1.0573,
+      "step": 8547
+    },
+    {
+      "epoch": 7.891068543734133,
+      "grad_norm": 3.30078125,
+      "learning_rate": 4.304202801867912e-05,
+      "loss": 1.4873,
+      "step": 8548
+    },
+    {
+      "epoch": 7.89199169166859,
+      "grad_norm": 3.51953125,
+      "learning_rate": 4.303535690460307e-05,
+      "loss": 1.3564,
+      "step": 8549
+    },
+    {
+      "epoch": 7.892914839603046,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.3028685790527016e-05,
+      "loss": 1.5259,
+      "step": 8550
+    },
+    {
+      "epoch": 7.893837987537503,
+      "grad_norm": 3.830078125,
+      "learning_rate": 4.302201467645097e-05,
+      "loss": 1.4709,
+      "step": 8551
+    },
+    {
+      "epoch": 7.89476113547196,
+      "grad_norm": 3.064453125,
+      "learning_rate": 4.301534356237492e-05,
+      "loss": 1.3596,
+      "step": 8552
+    },
+    {
+      "epoch": 7.895684283406416,
+      "grad_norm": 3.244140625,
+      "learning_rate": 4.300867244829887e-05,
+      "loss": 1.4775,
+      "step": 8553
+    },
+    {
+      "epoch": 7.896607431340873,
+      "grad_norm": 4.10546875,
+      "learning_rate": 4.300200133422282e-05,
+      "loss": 1.3253,
+      "step": 8554
+    },
+    {
+      "epoch": 7.897530579275329,
+      "grad_norm": 3.7890625,
+      "learning_rate": 4.299533022014677e-05,
+      "loss": 1.4189,
+      "step": 8555
+    },
+    {
+      "epoch": 7.898453727209786,
+      "grad_norm": 3.07421875,
+      "learning_rate": 4.2988659106070716e-05,
+      "loss": 1.4631,
+      "step": 8556
+    },
+    {
+      "epoch": 7.899376875144242,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.298198799199467e-05,
+      "loss": 1.6133,
+      "step": 8557
+    },
+    {
+      "epoch": 7.900300023078699,
+      "grad_norm": 2.857421875,
+      "learning_rate": 4.297531687791861e-05,
+      "loss": 1.1658,
+      "step": 8558
+    },
+    {
+      "epoch": 7.901223171013155,
+      "grad_norm": 3.466796875,
+      "learning_rate": 4.2968645763842566e-05,
+      "loss": 1.4103,
+      "step": 8559
+    },
+    {
+      "epoch": 7.9021463189476115,
+      "grad_norm": 4.29296875,
+      "learning_rate": 4.296197464976651e-05,
+      "loss": 1.4963,
+      "step": 8560
+    },
+    {
+      "epoch": 7.903069466882068,
+      "grad_norm": 3.3515625,
+      "learning_rate": 4.295530353569046e-05,
+      "loss": 1.2472,
+      "step": 8561
+    },
+    {
+      "epoch": 7.9039926148165245,
+      "grad_norm": 3.0,
+      "learning_rate": 4.294863242161441e-05,
+      "loss": 1.019,
+      "step": 8562
+    },
+    {
+      "epoch": 7.9049157627509805,
+      "grad_norm": 3.595703125,
+      "learning_rate": 4.294196130753836e-05,
+      "loss": 1.2516,
+      "step": 8563
+    },
+    {
+      "epoch": 7.905838910685437,
+      "grad_norm": 3.470703125,
+      "learning_rate": 4.2935290193462306e-05,
+      "loss": 1.3964,
+      "step": 8564
+    },
+    {
+      "epoch": 7.9067620586198935,
+      "grad_norm": 3.767578125,
+      "learning_rate": 4.2928619079386265e-05,
+      "loss": 1.3107,
+      "step": 8565
+    },
+    {
+      "epoch": 7.90768520655435,
+      "grad_norm": 3.20703125,
+      "learning_rate": 4.292194796531021e-05,
+      "loss": 1.345,
+      "step": 8566
+    },
+    {
+      "epoch": 7.908608354488807,
+      "grad_norm": 3.990234375,
+      "learning_rate": 4.291527685123416e-05,
+      "loss": 1.5935,
+      "step": 8567
+    },
+    {
+      "epoch": 7.909531502423263,
+      "grad_norm": 3.474609375,
+      "learning_rate": 4.290860573715811e-05,
+      "loss": 1.1651,
+      "step": 8568
+    },
+    {
+      "epoch": 7.910454650357719,
+      "grad_norm": 3.76953125,
+      "learning_rate": 4.290193462308206e-05,
+      "loss": 1.2769,
+      "step": 8569
+    },
+    {
+      "epoch": 7.911377798292176,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.2895263509006006e-05,
+      "loss": 1.1997,
+      "step": 8570
+    },
+    {
+      "epoch": 7.912300946226633,
+      "grad_norm": 3.294921875,
+      "learning_rate": 4.288859239492996e-05,
+      "loss": 1.1274,
+      "step": 8571
+    },
+    {
+      "epoch": 7.913224094161089,
+      "grad_norm": 3.333984375,
+      "learning_rate": 4.2881921280853904e-05,
+      "loss": 1.3831,
+      "step": 8572
+    },
+    {
+      "epoch": 7.914147242095546,
+      "grad_norm": 3.3359375,
+      "learning_rate": 4.2875250166777856e-05,
+      "loss": 1.6868,
+      "step": 8573
+    },
+    {
+      "epoch": 7.915070390030002,
+      "grad_norm": 3.259765625,
+      "learning_rate": 4.28685790527018e-05,
+      "loss": 1.0232,
+      "step": 8574
+    },
+    {
+      "epoch": 7.915993537964459,
+      "grad_norm": 3.501953125,
+      "learning_rate": 4.286190793862575e-05,
+      "loss": 1.509,
+      "step": 8575
+    },
+    {
+      "epoch": 7.916916685898915,
+      "grad_norm": 3.15234375,
+      "learning_rate": 4.28552368245497e-05,
+      "loss": 1.2916,
+      "step": 8576
+    },
+    {
+      "epoch": 7.917839833833372,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.284856571047365e-05,
+      "loss": 1.3259,
+      "step": 8577
+    },
+    {
+      "epoch": 7.918762981767828,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.28418945963976e-05,
+      "loss": 1.3811,
+      "step": 8578
+    },
+    {
+      "epoch": 7.919686129702285,
+      "grad_norm": 3.2421875,
+      "learning_rate": 4.283522348232155e-05,
+      "loss": 1.25,
+      "step": 8579
+    },
+    {
+      "epoch": 7.920609277636741,
+      "grad_norm": 4.18359375,
+      "learning_rate": 4.28285523682455e-05,
+      "loss": 1.2893,
+      "step": 8580
+    },
+    {
+      "epoch": 7.921532425571198,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.2821881254169446e-05,
+      "loss": 1.3412,
+      "step": 8581
+    },
+    {
+      "epoch": 7.922455573505655,
+      "grad_norm": 3.310546875,
+      "learning_rate": 4.28152101400934e-05,
+      "loss": 1.3982,
+      "step": 8582
+    },
+    {
+      "epoch": 7.923378721440111,
+      "grad_norm": 3.939453125,
+      "learning_rate": 4.2808539026017344e-05,
+      "loss": 1.0435,
+      "step": 8583
+    },
+    {
+      "epoch": 7.924301869374567,
+      "grad_norm": 2.927734375,
+      "learning_rate": 4.2801867911941296e-05,
+      "loss": 1.0376,
+      "step": 8584
+    },
+    {
+      "epoch": 7.925225017309024,
+      "grad_norm": 3.767578125,
+      "learning_rate": 4.279519679786524e-05,
+      "loss": 1.2976,
+      "step": 8585
+    },
+    {
+      "epoch": 7.926148165243481,
+      "grad_norm": 3.408203125,
+      "learning_rate": 4.2788525683789194e-05,
+      "loss": 1.6104,
+      "step": 8586
+    },
+    {
+      "epoch": 7.927071313177937,
+      "grad_norm": 3.634765625,
+      "learning_rate": 4.278185456971314e-05,
+      "loss": 1.2876,
+      "step": 8587
+    },
+    {
+      "epoch": 7.927994461112394,
+      "grad_norm": 3.0859375,
+      "learning_rate": 4.277518345563709e-05,
+      "loss": 1.6431,
+      "step": 8588
+    },
+    {
+      "epoch": 7.92891760904685,
+      "grad_norm": 3.119140625,
+      "learning_rate": 4.276851234156104e-05,
+      "loss": 1.2634,
+      "step": 8589
+    },
+    {
+      "epoch": 7.9298407569813065,
+      "grad_norm": 3.310546875,
+      "learning_rate": 4.2761841227484996e-05,
+      "loss": 1.1246,
+      "step": 8590
+    },
+    {
+      "epoch": 7.9307639049157626,
+      "grad_norm": 3.8515625,
+      "learning_rate": 4.275517011340894e-05,
+      "loss": 1.1833,
+      "step": 8591
+    },
+    {
+      "epoch": 7.9316870528502195,
+      "grad_norm": 3.716796875,
+      "learning_rate": 4.2748498999332893e-05,
+      "loss": 1.5278,
+      "step": 8592
+    },
+    {
+      "epoch": 7.9326102007846755,
+      "grad_norm": 3.98828125,
+      "learning_rate": 4.274182788525684e-05,
+      "loss": 1.2507,
+      "step": 8593
+    },
+    {
+      "epoch": 7.933533348719132,
+      "grad_norm": 3.18359375,
+      "learning_rate": 4.273515677118079e-05,
+      "loss": 1.1259,
+      "step": 8594
+    },
+    {
+      "epoch": 7.934456496653588,
+      "grad_norm": 3.14453125,
+      "learning_rate": 4.2728485657104737e-05,
+      "loss": 1.4281,
+      "step": 8595
+    },
+    {
+      "epoch": 7.935379644588045,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.272181454302869e-05,
+      "loss": 1.0082,
+      "step": 8596
+    },
+    {
+      "epoch": 7.936302792522501,
+      "grad_norm": 3.642578125,
+      "learning_rate": 4.2715143428952634e-05,
+      "loss": 1.4106,
+      "step": 8597
+    },
+    {
+      "epoch": 7.937225940456958,
+      "grad_norm": 3.501953125,
+      "learning_rate": 4.2708472314876586e-05,
+      "loss": 1.4226,
+      "step": 8598
+    },
+    {
+      "epoch": 7.938149088391414,
+      "grad_norm": 3.462890625,
+      "learning_rate": 4.270180120080053e-05,
+      "loss": 1.3333,
+      "step": 8599
+    },
+    {
+      "epoch": 7.939072236325871,
+      "grad_norm": 3.4765625,
+      "learning_rate": 4.2695130086724484e-05,
+      "loss": 1.1648,
+      "step": 8600
+    },
+    {
+      "epoch": 7.939995384260328,
+      "grad_norm": 3.591796875,
+      "learning_rate": 4.268845897264843e-05,
+      "loss": 1.488,
+      "step": 8601
+    },
+    {
+      "epoch": 7.940918532194784,
+      "grad_norm": 3.51171875,
+      "learning_rate": 4.268178785857239e-05,
+      "loss": 1.4814,
+      "step": 8602
+    },
+    {
+      "epoch": 7.941841680129241,
+      "grad_norm": 3.673828125,
+      "learning_rate": 4.2675116744496334e-05,
+      "loss": 1.1005,
+      "step": 8603
+    },
+    {
+      "epoch": 7.942764828063697,
+      "grad_norm": 3.23046875,
+      "learning_rate": 4.2668445630420286e-05,
+      "loss": 1.2924,
+      "step": 8604
+    },
+    {
+      "epoch": 7.943687975998154,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.266177451634423e-05,
+      "loss": 1.2766,
+      "step": 8605
+    },
+    {
+      "epoch": 7.94461112393261,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.2655103402268184e-05,
+      "loss": 1.2952,
+      "step": 8606
+    },
+    {
+      "epoch": 7.945534271867067,
+      "grad_norm": 3.232421875,
+      "learning_rate": 4.264843228819213e-05,
+      "loss": 1.1166,
+      "step": 8607
+    },
+    {
+      "epoch": 7.946457419801523,
+      "grad_norm": 3.115234375,
+      "learning_rate": 4.264176117411608e-05,
+      "loss": 1.4006,
+      "step": 8608
+    },
+    {
+      "epoch": 7.94738056773598,
+      "grad_norm": 3.583984375,
+      "learning_rate": 4.263509006004003e-05,
+      "loss": 1.4308,
+      "step": 8609
+    },
+    {
+      "epoch": 7.948303715670436,
+      "grad_norm": 3.2890625,
+      "learning_rate": 4.262841894596398e-05,
+      "loss": 1.5281,
+      "step": 8610
+    },
+    {
+      "epoch": 7.949226863604893,
+      "grad_norm": 2.9921875,
+      "learning_rate": 4.2621747831887924e-05,
+      "loss": 1.5596,
+      "step": 8611
+    },
+    {
+      "epoch": 7.950150011539349,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.261507671781188e-05,
+      "loss": 1.5444,
+      "step": 8612
+    },
+    {
+      "epoch": 7.951073159473806,
+      "grad_norm": 3.287109375,
+      "learning_rate": 4.260840560373582e-05,
+      "loss": 1.3281,
+      "step": 8613
+    },
+    {
+      "epoch": 7.951996307408262,
+      "grad_norm": 2.92578125,
+      "learning_rate": 4.2601734489659774e-05,
+      "loss": 1.1112,
+      "step": 8614
+    },
+    {
+      "epoch": 7.952919455342719,
+      "grad_norm": 3.12109375,
+      "learning_rate": 4.2595063375583726e-05,
+      "loss": 1.4722,
+      "step": 8615
+    },
+    {
+      "epoch": 7.953842603277176,
+      "grad_norm": 3.861328125,
+      "learning_rate": 4.258839226150768e-05,
+      "loss": 1.2039,
+      "step": 8616
+    },
+    {
+      "epoch": 7.954765751211632,
+      "grad_norm": 3.248046875,
+      "learning_rate": 4.2581721147431624e-05,
+      "loss": 1.4097,
+      "step": 8617
+    },
+    {
+      "epoch": 7.955688899146088,
+      "grad_norm": 3.623046875,
+      "learning_rate": 4.2575050033355576e-05,
+      "loss": 1.5195,
+      "step": 8618
+    },
+    {
+      "epoch": 7.956612047080545,
+      "grad_norm": 3.107421875,
+      "learning_rate": 4.256837891927952e-05,
+      "loss": 1.4641,
+      "step": 8619
+    },
+    {
+      "epoch": 7.9575351950150015,
+      "grad_norm": 3.58203125,
+      "learning_rate": 4.2561707805203474e-05,
+      "loss": 1.6016,
+      "step": 8620
+    },
+    {
+      "epoch": 7.9584583429494575,
+      "grad_norm": 2.990234375,
+      "learning_rate": 4.255503669112742e-05,
+      "loss": 1.4609,
+      "step": 8621
+    },
+    {
+      "epoch": 7.9593814908839144,
+      "grad_norm": 3.80078125,
+      "learning_rate": 4.254836557705137e-05,
+      "loss": 1.2714,
+      "step": 8622
+    },
+    {
+      "epoch": 7.9603046388183705,
+      "grad_norm": 3.431640625,
+      "learning_rate": 4.254169446297532e-05,
+      "loss": 1.2234,
+      "step": 8623
+    },
+    {
+      "epoch": 7.961227786752827,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.253502334889927e-05,
+      "loss": 1.4241,
+      "step": 8624
+    },
+    {
+      "epoch": 7.962150934687283,
+      "grad_norm": 3.47265625,
+      "learning_rate": 4.2528352234823215e-05,
+      "loss": 1.2643,
+      "step": 8625
+    },
+    {
+      "epoch": 7.96307408262174,
+      "grad_norm": 3.04296875,
+      "learning_rate": 4.252168112074717e-05,
+      "loss": 1.4281,
+      "step": 8626
+    },
+    {
+      "epoch": 7.963997230556196,
+      "grad_norm": 3.1796875,
+      "learning_rate": 4.251501000667112e-05,
+      "loss": 1.3804,
+      "step": 8627
+    },
+    {
+      "epoch": 7.964920378490653,
+      "grad_norm": 3.33203125,
+      "learning_rate": 4.2508338892595064e-05,
+      "loss": 1.4368,
+      "step": 8628
+    },
+    {
+      "epoch": 7.965843526425109,
+      "grad_norm": 3.509765625,
+      "learning_rate": 4.250166777851902e-05,
+      "loss": 1.3142,
+      "step": 8629
+    },
+    {
+      "epoch": 7.966766674359566,
+      "grad_norm": 2.48828125,
+      "learning_rate": 4.249499666444296e-05,
+      "loss": 1.0308,
+      "step": 8630
+    },
+    {
+      "epoch": 7.967689822294023,
+      "grad_norm": 3.751953125,
+      "learning_rate": 4.2488325550366914e-05,
+      "loss": 1.2847,
+      "step": 8631
+    },
+    {
+      "epoch": 7.968612970228479,
+      "grad_norm": 3.81640625,
+      "learning_rate": 4.248165443629086e-05,
+      "loss": 1.3386,
+      "step": 8632
+    },
+    {
+      "epoch": 7.969536118162935,
+      "grad_norm": 3.52734375,
+      "learning_rate": 4.247498332221481e-05,
+      "loss": 1.5771,
+      "step": 8633
+    },
+    {
+      "epoch": 7.970459266097392,
+      "grad_norm": 4.0390625,
+      "learning_rate": 4.246831220813876e-05,
+      "loss": 1.3198,
+      "step": 8634
+    },
+    {
+      "epoch": 7.971382414031849,
+      "grad_norm": 3.439453125,
+      "learning_rate": 4.246164109406271e-05,
+      "loss": 1.2524,
+      "step": 8635
+    },
+    {
+      "epoch": 7.972305561966305,
+      "grad_norm": 4.1171875,
+      "learning_rate": 4.245496997998666e-05,
+      "loss": 1.481,
+      "step": 8636
+    },
+    {
+      "epoch": 7.973228709900762,
+      "grad_norm": 3.162109375,
+      "learning_rate": 4.244829886591061e-05,
+      "loss": 1.3362,
+      "step": 8637
+    },
+    {
+      "epoch": 7.974151857835218,
+      "grad_norm": 3.568359375,
+      "learning_rate": 4.244162775183456e-05,
+      "loss": 1.4023,
+      "step": 8638
+    },
+    {
+      "epoch": 7.975075005769675,
+      "grad_norm": 3.37890625,
+      "learning_rate": 4.2434956637758505e-05,
+      "loss": 1.3801,
+      "step": 8639
+    },
+    {
+      "epoch": 7.975998153704131,
+      "grad_norm": 3.15234375,
+      "learning_rate": 4.242828552368246e-05,
+      "loss": 1.3728,
+      "step": 8640
+    },
+    {
+      "epoch": 7.976921301638588,
+      "grad_norm": 3.685546875,
+      "learning_rate": 4.242161440960641e-05,
+      "loss": 1.2227,
+      "step": 8641
+    },
+    {
+      "epoch": 7.977844449573044,
+      "grad_norm": 3.44921875,
+      "learning_rate": 4.2414943295530355e-05,
+      "loss": 1.3425,
+      "step": 8642
+    },
+    {
+      "epoch": 7.978767597507501,
+      "grad_norm": 3.275390625,
+      "learning_rate": 4.240827218145431e-05,
+      "loss": 1.2993,
+      "step": 8643
+    },
+    {
+      "epoch": 7.979690745441957,
+      "grad_norm": 3.31640625,
+      "learning_rate": 4.240160106737825e-05,
+      "loss": 1.2336,
+      "step": 8644
+    },
+    {
+      "epoch": 7.980613893376414,
+      "grad_norm": 3.103515625,
+      "learning_rate": 4.2394929953302205e-05,
+      "loss": 1.2599,
+      "step": 8645
+    },
+    {
+      "epoch": 7.98153704131087,
+      "grad_norm": 3.755859375,
+      "learning_rate": 4.238825883922615e-05,
+      "loss": 1.3262,
+      "step": 8646
+    },
+    {
+      "epoch": 7.982460189245327,
+      "grad_norm": 3.4296875,
+      "learning_rate": 4.23815877251501e-05,
+      "loss": 1.2017,
+      "step": 8647
+    },
+    {
+      "epoch": 7.983383337179783,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.237491661107405e-05,
+      "loss": 1.236,
+      "step": 8648
+    },
+    {
+      "epoch": 7.98430648511424,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.2368245496998e-05,
+      "loss": 1.3867,
+      "step": 8649
+    },
+    {
+      "epoch": 7.9852296330486965,
+      "grad_norm": 3.89453125,
+      "learning_rate": 4.2361574382921945e-05,
+      "loss": 1.5403,
+      "step": 8650
+    },
+    {
+      "epoch": 7.9861527809831525,
+      "grad_norm": 3.20703125,
+      "learning_rate": 4.23549032688459e-05,
+      "loss": 1.4341,
+      "step": 8651
+    },
+    {
+      "epoch": 7.987075928917609,
+      "grad_norm": 3.25390625,
+      "learning_rate": 4.234823215476985e-05,
+      "loss": 1.2075,
+      "step": 8652
+    },
+    {
+      "epoch": 7.9879990768520654,
+      "grad_norm": 3.521484375,
+      "learning_rate": 4.23415610406938e-05,
+      "loss": 1.4802,
+      "step": 8653
+    },
+    {
+      "epoch": 7.988922224786522,
+      "grad_norm": 3.61328125,
+      "learning_rate": 4.233488992661775e-05,
+      "loss": 1.1848,
+      "step": 8654
+    },
+    {
+      "epoch": 7.989845372720978,
+      "grad_norm": 3.091796875,
+      "learning_rate": 4.23282188125417e-05,
+      "loss": 1.2722,
+      "step": 8655
+    },
+    {
+      "epoch": 7.990768520655435,
+      "grad_norm": 3.134765625,
+      "learning_rate": 4.2321547698465645e-05,
+      "loss": 1.4314,
+      "step": 8656
+    },
+    {
+      "epoch": 7.991691668589891,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.23148765843896e-05,
+      "loss": 1.4368,
+      "step": 8657
+    },
+    {
+      "epoch": 7.992614816524348,
+      "grad_norm": 2.84765625,
+      "learning_rate": 4.230820547031354e-05,
+      "loss": 1.6831,
+      "step": 8658
+    },
+    {
+      "epoch": 7.993537964458804,
+      "grad_norm": 3.39453125,
+      "learning_rate": 4.2301534356237495e-05,
+      "loss": 1.1815,
+      "step": 8659
+    },
+    {
+      "epoch": 7.994461112393261,
+      "grad_norm": 3.20703125,
+      "learning_rate": 4.229486324216144e-05,
+      "loss": 1.2322,
+      "step": 8660
+    },
+    {
+      "epoch": 7.995384260327717,
+      "grad_norm": 3.396484375,
+      "learning_rate": 4.228819212808539e-05,
+      "loss": 1.3162,
+      "step": 8661
+    },
+    {
+      "epoch": 7.996307408262174,
+      "grad_norm": 3.02734375,
+      "learning_rate": 4.228152101400934e-05,
+      "loss": 1.1552,
+      "step": 8662
+    },
+    {
+      "epoch": 7.99723055619663,
+      "grad_norm": 3.1015625,
+      "learning_rate": 4.227484989993329e-05,
+      "loss": 1.3289,
+      "step": 8663
+    },
+    {
+      "epoch": 7.998153704131087,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.2268178785857236e-05,
+      "loss": 1.3306,
+      "step": 8664
+    },
+    {
+      "epoch": 7.999076852065544,
+      "grad_norm": 2.86328125,
+      "learning_rate": 4.2261507671781194e-05,
+      "loss": 1.5149,
+      "step": 8665
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.225483655770514e-05,
+      "loss": 1.0592,
+      "step": 8666
+    },
+    {
+      "epoch": 8.000923147934456,
+      "grad_norm": 3.701171875,
+      "learning_rate": 4.224816544362909e-05,
+      "loss": 1.2983,
+      "step": 8667
+    },
+    {
+      "epoch": 8.001846295868914,
+      "grad_norm": 3.07421875,
+      "learning_rate": 4.224149432955304e-05,
+      "loss": 1.48,
+      "step": 8668
+    },
+    {
+      "epoch": 8.00276944380337,
+      "grad_norm": 3.517578125,
+      "learning_rate": 4.223482321547699e-05,
+      "loss": 1.395,
+      "step": 8669
+    },
+    {
+      "epoch": 8.003692591737826,
+      "grad_norm": 3.30078125,
+      "learning_rate": 4.2228152101400935e-05,
+      "loss": 1.4714,
+      "step": 8670
+    },
+    {
+      "epoch": 8.004615739672282,
+      "grad_norm": 3.125,
+      "learning_rate": 4.222148098732489e-05,
+      "loss": 1.1021,
+      "step": 8671
+    },
+    {
+      "epoch": 8.00553888760674,
+      "grad_norm": 3.208984375,
+      "learning_rate": 4.221480987324883e-05,
+      "loss": 1.2107,
+      "step": 8672
+    },
+    {
+      "epoch": 8.006462035541196,
+      "grad_norm": 2.951171875,
+      "learning_rate": 4.2208138759172785e-05,
+      "loss": 1.4209,
+      "step": 8673
+    },
+    {
+      "epoch": 8.007385183475652,
+      "grad_norm": 3.62890625,
+      "learning_rate": 4.220146764509673e-05,
+      "loss": 1.2312,
+      "step": 8674
+    },
+    {
+      "epoch": 8.008308331410108,
+      "grad_norm": 3.505859375,
+      "learning_rate": 4.219479653102068e-05,
+      "loss": 1.2645,
+      "step": 8675
+    },
+    {
+      "epoch": 8.009231479344566,
+      "grad_norm": 3.271484375,
+      "learning_rate": 4.218812541694463e-05,
+      "loss": 1.2856,
+      "step": 8676
+    },
+    {
+      "epoch": 8.010154627279022,
+      "grad_norm": 4.00390625,
+      "learning_rate": 4.218145430286859e-05,
+      "loss": 1.5139,
+      "step": 8677
+    },
+    {
+      "epoch": 8.011077775213478,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.217478318879253e-05,
+      "loss": 1.3347,
+      "step": 8678
+    },
+    {
+      "epoch": 8.012000923147934,
+      "grad_norm": 3.125,
+      "learning_rate": 4.2168112074716485e-05,
+      "loss": 1.1055,
+      "step": 8679
+    },
+    {
+      "epoch": 8.012924071082391,
+      "grad_norm": 3.08984375,
+      "learning_rate": 4.216144096064043e-05,
+      "loss": 1.5901,
+      "step": 8680
+    },
+    {
+      "epoch": 8.013847219016847,
+      "grad_norm": 3.642578125,
+      "learning_rate": 4.215476984656438e-05,
+      "loss": 1.265,
+      "step": 8681
+    },
+    {
+      "epoch": 8.014770366951304,
+      "grad_norm": 3.162109375,
+      "learning_rate": 4.214809873248833e-05,
+      "loss": 1.3843,
+      "step": 8682
+    },
+    {
+      "epoch": 8.015693514885761,
+      "grad_norm": 2.974609375,
+      "learning_rate": 4.214142761841228e-05,
+      "loss": 1.333,
+      "step": 8683
+    },
+    {
+      "epoch": 8.016616662820217,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.2134756504336225e-05,
+      "loss": 1.4016,
+      "step": 8684
+    },
+    {
+      "epoch": 8.017539810754673,
+      "grad_norm": 3.537109375,
+      "learning_rate": 4.212808539026018e-05,
+      "loss": 1.259,
+      "step": 8685
+    },
+    {
+      "epoch": 8.01846295868913,
+      "grad_norm": 3.298828125,
+      "learning_rate": 4.212141427618412e-05,
+      "loss": 1.3201,
+      "step": 8686
+    },
+    {
+      "epoch": 8.019386106623587,
+      "grad_norm": 3.142578125,
+      "learning_rate": 4.2114743162108075e-05,
+      "loss": 1.5496,
+      "step": 8687
+    },
+    {
+      "epoch": 8.020309254558043,
+      "grad_norm": 3.279296875,
+      "learning_rate": 4.210807204803202e-05,
+      "loss": 1.4248,
+      "step": 8688
+    },
+    {
+      "epoch": 8.0212324024925,
+      "grad_norm": 2.81640625,
+      "learning_rate": 4.210140093395597e-05,
+      "loss": 1.1637,
+      "step": 8689
+    },
+    {
+      "epoch": 8.022155550426955,
+      "grad_norm": 3.708984375,
+      "learning_rate": 4.2094729819879925e-05,
+      "loss": 1.0988,
+      "step": 8690
+    },
+    {
+      "epoch": 8.023078698361413,
+      "grad_norm": 3.96484375,
+      "learning_rate": 4.208805870580387e-05,
+      "loss": 1.2668,
+      "step": 8691
+    },
+    {
+      "epoch": 8.024001846295869,
+      "grad_norm": 3.814453125,
+      "learning_rate": 4.208138759172782e-05,
+      "loss": 1.343,
+      "step": 8692
+    },
+    {
+      "epoch": 8.024924994230325,
+      "grad_norm": 3.427734375,
+      "learning_rate": 4.207471647765177e-05,
+      "loss": 1.2432,
+      "step": 8693
+    },
+    {
+      "epoch": 8.025848142164781,
+      "grad_norm": 4.29296875,
+      "learning_rate": 4.206804536357572e-05,
+      "loss": 1.4036,
+      "step": 8694
+    },
+    {
+      "epoch": 8.026771290099239,
+      "grad_norm": 3.255859375,
+      "learning_rate": 4.2061374249499666e-05,
+      "loss": 1.4155,
+      "step": 8695
+    },
+    {
+      "epoch": 8.027694438033695,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.205470313542362e-05,
+      "loss": 0.6936,
+      "step": 8696
+    },
+    {
+      "epoch": 8.028617585968151,
+      "grad_norm": 3.375,
+      "learning_rate": 4.2048032021347563e-05,
+      "loss": 1.5579,
+      "step": 8697
+    },
+    {
+      "epoch": 8.029540733902607,
+      "grad_norm": 3.791015625,
+      "learning_rate": 4.2041360907271516e-05,
+      "loss": 1.7068,
+      "step": 8698
+    },
+    {
+      "epoch": 8.030463881837065,
+      "grad_norm": 3.197265625,
+      "learning_rate": 4.203468979319546e-05,
+      "loss": 1.1108,
+      "step": 8699
+    },
+    {
+      "epoch": 8.03138702977152,
+      "grad_norm": 3.384765625,
+      "learning_rate": 4.202801867911941e-05,
+      "loss": 1.4385,
+      "step": 8700
+    },
+    {
+      "epoch": 8.032310177705977,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.202134756504336e-05,
+      "loss": 1.1538,
+      "step": 8701
+    },
+    {
+      "epoch": 8.033233325640435,
+      "grad_norm": 3.4140625,
+      "learning_rate": 4.201467645096732e-05,
+      "loss": 1.4688,
+      "step": 8702
+    },
+    {
+      "epoch": 8.03415647357489,
+      "grad_norm": 3.939453125,
+      "learning_rate": 4.200800533689126e-05,
+      "loss": 1.3887,
+      "step": 8703
+    },
+    {
+      "epoch": 8.035079621509347,
+      "grad_norm": 3.322265625,
+      "learning_rate": 4.2001334222815215e-05,
+      "loss": 1.069,
+      "step": 8704
+    },
+    {
+      "epoch": 8.036002769443803,
+      "grad_norm": 3.3515625,
+      "learning_rate": 4.199466310873916e-05,
+      "loss": 1.4011,
+      "step": 8705
+    },
+    {
+      "epoch": 8.03692591737826,
+      "grad_norm": 3.646484375,
+      "learning_rate": 4.198799199466311e-05,
+      "loss": 1.1705,
+      "step": 8706
+    },
+    {
+      "epoch": 8.037849065312717,
+      "grad_norm": 4.24609375,
+      "learning_rate": 4.198132088058706e-05,
+      "loss": 1.1797,
+      "step": 8707
+    },
+    {
+      "epoch": 8.038772213247173,
+      "grad_norm": 3.4140625,
+      "learning_rate": 4.197464976651101e-05,
+      "loss": 1.1547,
+      "step": 8708
+    },
+    {
+      "epoch": 8.039695361181629,
+      "grad_norm": 3.072265625,
+      "learning_rate": 4.1967978652434956e-05,
+      "loss": 1.0522,
+      "step": 8709
+    },
+    {
+      "epoch": 8.040618509116086,
+      "grad_norm": 3.84765625,
+      "learning_rate": 4.196130753835891e-05,
+      "loss": 1.0392,
+      "step": 8710
+    },
+    {
+      "epoch": 8.041541657050542,
+      "grad_norm": 3.60546875,
+      "learning_rate": 4.1954636424282854e-05,
+      "loss": 1.3167,
+      "step": 8711
+    },
+    {
+      "epoch": 8.042464804984998,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.1947965310206806e-05,
+      "loss": 1.3389,
+      "step": 8712
+    },
+    {
+      "epoch": 8.043387952919455,
+      "grad_norm": 3.55859375,
+      "learning_rate": 4.194129419613075e-05,
+      "loss": 1.446,
+      "step": 8713
+    },
+    {
+      "epoch": 8.044311100853912,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.1934623082054704e-05,
+      "loss": 1.4177,
+      "step": 8714
+    },
+    {
+      "epoch": 8.045234248788368,
+      "grad_norm": 3.017578125,
+      "learning_rate": 4.1927951967978656e-05,
+      "loss": 0.9457,
+      "step": 8715
+    },
+    {
+      "epoch": 8.046157396722824,
+      "grad_norm": 3.5234375,
+      "learning_rate": 4.192128085390261e-05,
+      "loss": 1.5417,
+      "step": 8716
+    },
+    {
+      "epoch": 8.047080544657282,
+      "grad_norm": 3.685546875,
+      "learning_rate": 4.1914609739826553e-05,
+      "loss": 1.3997,
+      "step": 8717
+    },
+    {
+      "epoch": 8.048003692591738,
+      "grad_norm": 2.91796875,
+      "learning_rate": 4.1907938625750506e-05,
+      "loss": 1.3503,
+      "step": 8718
+    },
+    {
+      "epoch": 8.048926840526194,
+      "grad_norm": 3.1796875,
+      "learning_rate": 4.190126751167445e-05,
+      "loss": 1.4297,
+      "step": 8719
+    },
+    {
+      "epoch": 8.04984998846065,
+      "grad_norm": 3.33984375,
+      "learning_rate": 4.18945963975984e-05,
+      "loss": 1.25,
+      "step": 8720
+    },
+    {
+      "epoch": 8.050773136395108,
+      "grad_norm": 3.494140625,
+      "learning_rate": 4.188792528352235e-05,
+      "loss": 1.3618,
+      "step": 8721
+    },
+    {
+      "epoch": 8.051696284329564,
+      "grad_norm": 3.798828125,
+      "learning_rate": 4.18812541694463e-05,
+      "loss": 1.3602,
+      "step": 8722
+    },
+    {
+      "epoch": 8.05261943226402,
+      "grad_norm": 3.09765625,
+      "learning_rate": 4.1874583055370246e-05,
+      "loss": 1.3022,
+      "step": 8723
+    },
+    {
+      "epoch": 8.053542580198476,
+      "grad_norm": 3.193359375,
+      "learning_rate": 4.18679119412942e-05,
+      "loss": 1.1598,
+      "step": 8724
+    },
+    {
+      "epoch": 8.054465728132934,
+      "grad_norm": 3.18359375,
+      "learning_rate": 4.1861240827218144e-05,
+      "loss": 1.5779,
+      "step": 8725
+    },
+    {
+      "epoch": 8.05538887606739,
+      "grad_norm": 3.658203125,
+      "learning_rate": 4.1854569713142096e-05,
+      "loss": 1.4189,
+      "step": 8726
+    },
+    {
+      "epoch": 8.056312024001846,
+      "grad_norm": 3.037109375,
+      "learning_rate": 4.184789859906604e-05,
+      "loss": 1.041,
+      "step": 8727
+    },
+    {
+      "epoch": 8.057235171936302,
+      "grad_norm": 3.572265625,
+      "learning_rate": 4.184122748499e-05,
+      "loss": 1.5256,
+      "step": 8728
+    },
+    {
+      "epoch": 8.05815831987076,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.1834556370913946e-05,
+      "loss": 1.4531,
+      "step": 8729
+    },
+    {
+      "epoch": 8.059081467805216,
+      "grad_norm": 4.18359375,
+      "learning_rate": 4.18278852568379e-05,
+      "loss": 1.48,
+      "step": 8730
+    },
+    {
+      "epoch": 8.060004615739672,
+      "grad_norm": 3.21484375,
+      "learning_rate": 4.1821214142761844e-05,
+      "loss": 1.2097,
+      "step": 8731
+    },
+    {
+      "epoch": 8.06092776367413,
+      "grad_norm": 3.10546875,
+      "learning_rate": 4.1814543028685796e-05,
+      "loss": 1.2366,
+      "step": 8732
+    },
+    {
+      "epoch": 8.061850911608586,
+      "grad_norm": 3.1484375,
+      "learning_rate": 4.180787191460974e-05,
+      "loss": 1.4163,
+      "step": 8733
+    },
+    {
+      "epoch": 8.062774059543042,
+      "grad_norm": 4.02734375,
+      "learning_rate": 4.1801200800533693e-05,
+      "loss": 1.533,
+      "step": 8734
+    },
+    {
+      "epoch": 8.063697207477498,
+      "grad_norm": 3.23828125,
+      "learning_rate": 4.179452968645764e-05,
+      "loss": 1.4532,
+      "step": 8735
+    },
+    {
+      "epoch": 8.064620355411956,
+      "grad_norm": 3.201171875,
+      "learning_rate": 4.178785857238159e-05,
+      "loss": 1.3816,
+      "step": 8736
+    },
+    {
+      "epoch": 8.065543503346412,
+      "grad_norm": 3.744140625,
+      "learning_rate": 4.1781187458305537e-05,
+      "loss": 1.5574,
+      "step": 8737
+    },
+    {
+      "epoch": 8.066466651280868,
+      "grad_norm": 3.21484375,
+      "learning_rate": 4.177451634422949e-05,
+      "loss": 1.4558,
+      "step": 8738
+    },
+    {
+      "epoch": 8.067389799215324,
+      "grad_norm": 3.9921875,
+      "learning_rate": 4.1767845230153434e-05,
+      "loss": 1.293,
+      "step": 8739
+    },
+    {
+      "epoch": 8.068312947149781,
+      "grad_norm": 3.083984375,
+      "learning_rate": 4.1761174116077386e-05,
+      "loss": 1.199,
+      "step": 8740
+    },
+    {
+      "epoch": 8.069236095084237,
+      "grad_norm": 3.365234375,
+      "learning_rate": 4.175450300200134e-05,
+      "loss": 1.4243,
+      "step": 8741
+    },
+    {
+      "epoch": 8.070159243018693,
+      "grad_norm": 3.5078125,
+      "learning_rate": 4.1747831887925284e-05,
+      "loss": 1.4102,
+      "step": 8742
+    },
+    {
+      "epoch": 8.07108239095315,
+      "grad_norm": 4.37890625,
+      "learning_rate": 4.1741160773849236e-05,
+      "loss": 1.2307,
+      "step": 8743
+    },
+    {
+      "epoch": 8.072005538887607,
+      "grad_norm": 3.44140625,
+      "learning_rate": 4.173448965977318e-05,
+      "loss": 1.5015,
+      "step": 8744
+    },
+    {
+      "epoch": 8.072928686822063,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.1727818545697134e-05,
+      "loss": 1.3088,
+      "step": 8745
+    },
+    {
+      "epoch": 8.07385183475652,
+      "grad_norm": 3.423828125,
+      "learning_rate": 4.172114743162108e-05,
+      "loss": 1.1714,
+      "step": 8746
+    },
+    {
+      "epoch": 8.074774982690975,
+      "grad_norm": 3.201171875,
+      "learning_rate": 4.171447631754503e-05,
+      "loss": 1.2145,
+      "step": 8747
+    },
+    {
+      "epoch": 8.075698130625433,
+      "grad_norm": 3.224609375,
+      "learning_rate": 4.1707805203468984e-05,
+      "loss": 1.4199,
+      "step": 8748
+    },
+    {
+      "epoch": 8.07662127855989,
+      "grad_norm": 3.76171875,
+      "learning_rate": 4.170113408939293e-05,
+      "loss": 1.3699,
+      "step": 8749
+    },
+    {
+      "epoch": 8.077544426494345,
+      "grad_norm": 3.16796875,
+      "learning_rate": 4.169446297531688e-05,
+      "loss": 1.5046,
+      "step": 8750
+    },
+    {
+      "epoch": 8.078467574428803,
+      "grad_norm": 3.22265625,
+      "learning_rate": 4.168779186124083e-05,
+      "loss": 1.478,
+      "step": 8751
+    },
+    {
+      "epoch": 8.079390722363259,
+      "grad_norm": 3.076171875,
+      "learning_rate": 4.168112074716478e-05,
+      "loss": 1.4194,
+      "step": 8752
+    },
+    {
+      "epoch": 8.080313870297715,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.167444963308873e-05,
+      "loss": 1.3774,
+      "step": 8753
+    },
+    {
+      "epoch": 8.081237018232171,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.166777851901268e-05,
+      "loss": 1.4226,
+      "step": 8754
+    },
+    {
+      "epoch": 8.082160166166629,
+      "grad_norm": 3.712890625,
+      "learning_rate": 4.166110740493663e-05,
+      "loss": 0.8111,
+      "step": 8755
+    },
+    {
+      "epoch": 8.083083314101085,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.1654436290860574e-05,
+      "loss": 1.4229,
+      "step": 8756
+    },
+    {
+      "epoch": 8.084006462035541,
+      "grad_norm": 3.75390625,
+      "learning_rate": 4.1647765176784526e-05,
+      "loss": 1.3589,
+      "step": 8757
+    },
+    {
+      "epoch": 8.084929609969997,
+      "grad_norm": 3.669921875,
+      "learning_rate": 4.164109406270847e-05,
+      "loss": 1.2842,
+      "step": 8758
+    },
+    {
+      "epoch": 8.085852757904455,
+      "grad_norm": 3.185546875,
+      "learning_rate": 4.1634422948632424e-05,
+      "loss": 1.1898,
+      "step": 8759
+    },
+    {
+      "epoch": 8.08677590583891,
+      "grad_norm": 3.26953125,
+      "learning_rate": 4.162775183455637e-05,
+      "loss": 1.2409,
+      "step": 8760
+    },
+    {
+      "epoch": 8.087699053773367,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.162108072048032e-05,
+      "loss": 1.4539,
+      "step": 8761
+    },
+    {
+      "epoch": 8.088622201707823,
+      "grad_norm": 3.029296875,
+      "learning_rate": 4.161440960640427e-05,
+      "loss": 1.3291,
+      "step": 8762
+    },
+    {
+      "epoch": 8.08954534964228,
+      "grad_norm": 3.267578125,
+      "learning_rate": 4.160773849232822e-05,
+      "loss": 1.4009,
+      "step": 8763
+    },
+    {
+      "epoch": 8.090468497576737,
+      "grad_norm": 3.181640625,
+      "learning_rate": 4.1601067378252165e-05,
+      "loss": 1.3533,
+      "step": 8764
+    },
+    {
+      "epoch": 8.091391645511193,
+      "grad_norm": 3.53515625,
+      "learning_rate": 4.1594396264176124e-05,
+      "loss": 1.1499,
+      "step": 8765
+    },
+    {
+      "epoch": 8.09231479344565,
+      "grad_norm": 3.5,
+      "learning_rate": 4.158772515010007e-05,
+      "loss": 1.3524,
+      "step": 8766
+    },
+    {
+      "epoch": 8.093237941380107,
+      "grad_norm": 3.16796875,
+      "learning_rate": 4.158105403602402e-05,
+      "loss": 1.5869,
+      "step": 8767
+    },
+    {
+      "epoch": 8.094161089314563,
+      "grad_norm": 3.646484375,
+      "learning_rate": 4.157438292194797e-05,
+      "loss": 1.2043,
+      "step": 8768
+    },
+    {
+      "epoch": 8.095084237249019,
+      "grad_norm": 3.1171875,
+      "learning_rate": 4.156771180787192e-05,
+      "loss": 1.1818,
+      "step": 8769
+    },
+    {
+      "epoch": 8.096007385183476,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.1561040693795865e-05,
+      "loss": 1.2057,
+      "step": 8770
+    },
+    {
+      "epoch": 8.096930533117932,
+      "grad_norm": 3.583984375,
+      "learning_rate": 4.155436957971982e-05,
+      "loss": 1.4028,
+      "step": 8771
+    },
+    {
+      "epoch": 8.097853681052388,
+      "grad_norm": 3.99609375,
+      "learning_rate": 4.154769846564376e-05,
+      "loss": 1.3933,
+      "step": 8772
+    },
+    {
+      "epoch": 8.098776828986844,
+      "grad_norm": 3.708984375,
+      "learning_rate": 4.1541027351567714e-05,
+      "loss": 1.4055,
+      "step": 8773
+    },
+    {
+      "epoch": 8.099699976921302,
+      "grad_norm": 3.44921875,
+      "learning_rate": 4.153435623749166e-05,
+      "loss": 1.5261,
+      "step": 8774
+    },
+    {
+      "epoch": 8.100623124855758,
+      "grad_norm": 3.69921875,
+      "learning_rate": 4.152768512341561e-05,
+      "loss": 1.4375,
+      "step": 8775
+    },
+    {
+      "epoch": 8.101546272790214,
+      "grad_norm": 3.439453125,
+      "learning_rate": 4.152101400933956e-05,
+      "loss": 1.3098,
+      "step": 8776
+    },
+    {
+      "epoch": 8.10246942072467,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.151434289526351e-05,
+      "loss": 1.2925,
+      "step": 8777
+    },
+    {
+      "epoch": 8.103392568659128,
+      "grad_norm": 4.16796875,
+      "learning_rate": 4.150767178118746e-05,
+      "loss": 1.2395,
+      "step": 8778
+    },
+    {
+      "epoch": 8.104315716593584,
+      "grad_norm": 3.4453125,
+      "learning_rate": 4.1501000667111414e-05,
+      "loss": 1.2306,
+      "step": 8779
+    },
+    {
+      "epoch": 8.10523886452804,
+      "grad_norm": 4.2578125,
+      "learning_rate": 4.149432955303536e-05,
+      "loss": 1.5845,
+      "step": 8780
+    },
+    {
+      "epoch": 8.106162012462496,
+      "grad_norm": 3.529296875,
+      "learning_rate": 4.148765843895931e-05,
+      "loss": 1.2593,
+      "step": 8781
+    },
+    {
+      "epoch": 8.107085160396954,
+      "grad_norm": 3.302734375,
+      "learning_rate": 4.148098732488326e-05,
+      "loss": 1.4641,
+      "step": 8782
+    },
+    {
+      "epoch": 8.10800830833141,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.147431621080721e-05,
+      "loss": 1.4114,
+      "step": 8783
+    },
+    {
+      "epoch": 8.108931456265866,
+      "grad_norm": 3.7421875,
+      "learning_rate": 4.1467645096731155e-05,
+      "loss": 1.4238,
+      "step": 8784
+    },
+    {
+      "epoch": 8.109854604200324,
+      "grad_norm": 3.1796875,
+      "learning_rate": 4.146097398265511e-05,
+      "loss": 1.5291,
+      "step": 8785
+    },
+    {
+      "epoch": 8.11077775213478,
+      "grad_norm": 3.69921875,
+      "learning_rate": 4.145430286857905e-05,
+      "loss": 1.1707,
+      "step": 8786
+    },
+    {
+      "epoch": 8.111700900069236,
+      "grad_norm": 3.40234375,
+      "learning_rate": 4.1447631754503005e-05,
+      "loss": 1.1161,
+      "step": 8787
+    },
+    {
+      "epoch": 8.112624048003692,
+      "grad_norm": 3.369140625,
+      "learning_rate": 4.144096064042695e-05,
+      "loss": 1.4629,
+      "step": 8788
+    },
+    {
+      "epoch": 8.11354719593815,
+      "grad_norm": 3.412109375,
+      "learning_rate": 4.14342895263509e-05,
+      "loss": 1.4077,
+      "step": 8789
+    },
+    {
+      "epoch": 8.114470343872606,
+      "grad_norm": 3.7109375,
+      "learning_rate": 4.1427618412274854e-05,
+      "loss": 1.2673,
+      "step": 8790
+    },
+    {
+      "epoch": 8.115393491807062,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.142094729819881e-05,
+      "loss": 1.3417,
+      "step": 8791
+    },
+    {
+      "epoch": 8.116316639741518,
+      "grad_norm": 3.13671875,
+      "learning_rate": 4.141427618412275e-05,
+      "loss": 1.4253,
+      "step": 8792
+    },
+    {
+      "epoch": 8.117239787675976,
+      "grad_norm": 3.052734375,
+      "learning_rate": 4.1407605070046704e-05,
+      "loss": 1.502,
+      "step": 8793
+    },
+    {
+      "epoch": 8.118162935610432,
+      "grad_norm": 3.166015625,
+      "learning_rate": 4.140093395597065e-05,
+      "loss": 1.3972,
+      "step": 8794
+    },
+    {
+      "epoch": 8.119086083544888,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.13942628418946e-05,
+      "loss": 1.2778,
+      "step": 8795
+    },
+    {
+      "epoch": 8.120009231479344,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.138759172781855e-05,
+      "loss": 1.229,
+      "step": 8796
+    },
+    {
+      "epoch": 8.120932379413802,
+      "grad_norm": 3.466796875,
+      "learning_rate": 4.13809206137425e-05,
+      "loss": 1.0753,
+      "step": 8797
+    },
+    {
+      "epoch": 8.121855527348258,
+      "grad_norm": 2.9921875,
+      "learning_rate": 4.1374249499666445e-05,
+      "loss": 1.0578,
+      "step": 8798
+    },
+    {
+      "epoch": 8.122778675282714,
+      "grad_norm": 3.21484375,
+      "learning_rate": 4.13675783855904e-05,
+      "loss": 1.416,
+      "step": 8799
+    },
+    {
+      "epoch": 8.123701823217171,
+      "grad_norm": 3.341796875,
+      "learning_rate": 4.136090727151434e-05,
+      "loss": 1.4031,
+      "step": 8800
+    },
+    {
+      "epoch": 8.124624971151627,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.1354236157438295e-05,
+      "loss": 1.3295,
+      "step": 8801
+    },
+    {
+      "epoch": 8.125548119086083,
+      "grad_norm": 3.861328125,
+      "learning_rate": 4.134756504336224e-05,
+      "loss": 1.2622,
+      "step": 8802
+    },
+    {
+      "epoch": 8.12647126702054,
+      "grad_norm": 3.5546875,
+      "learning_rate": 4.134089392928619e-05,
+      "loss": 1.335,
+      "step": 8803
+    },
+    {
+      "epoch": 8.127394414954997,
+      "grad_norm": 3.044921875,
+      "learning_rate": 4.1334222815210145e-05,
+      "loss": 1.3389,
+      "step": 8804
+    },
+    {
+      "epoch": 8.128317562889453,
+      "grad_norm": 3.115234375,
+      "learning_rate": 4.132755170113409e-05,
+      "loss": 1.3018,
+      "step": 8805
+    },
+    {
+      "epoch": 8.12924071082391,
+      "grad_norm": 3.29296875,
+      "learning_rate": 4.132088058705804e-05,
+      "loss": 1.2192,
+      "step": 8806
+    },
+    {
+      "epoch": 8.130163858758365,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.131420947298199e-05,
+      "loss": 1.2139,
+      "step": 8807
+    },
+    {
+      "epoch": 8.131087006692823,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.130753835890594e-05,
+      "loss": 1.4255,
+      "step": 8808
+    },
+    {
+      "epoch": 8.13201015462728,
+      "grad_norm": 3.392578125,
+      "learning_rate": 4.1300867244829885e-05,
+      "loss": 1.4166,
+      "step": 8809
+    },
+    {
+      "epoch": 8.132933302561735,
+      "grad_norm": 3.181640625,
+      "learning_rate": 4.129419613075384e-05,
+      "loss": 1.3043,
+      "step": 8810
+    },
+    {
+      "epoch": 8.133856450496191,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.128752501667778e-05,
+      "loss": 1.4197,
+      "step": 8811
+    },
+    {
+      "epoch": 8.134779598430649,
+      "grad_norm": 3.30859375,
+      "learning_rate": 4.1280853902601735e-05,
+      "loss": 1.3442,
+      "step": 8812
+    },
+    {
+      "epoch": 8.135702746365105,
+      "grad_norm": 2.818359375,
+      "learning_rate": 4.127418278852568e-05,
+      "loss": 1.426,
+      "step": 8813
+    },
+    {
+      "epoch": 8.136625894299561,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.126751167444963e-05,
+      "loss": 1.3752,
+      "step": 8814
+    },
+    {
+      "epoch": 8.137549042234019,
+      "grad_norm": 3.845703125,
+      "learning_rate": 4.1260840560373585e-05,
+      "loss": 1.4109,
+      "step": 8815
+    },
+    {
+      "epoch": 8.138472190168475,
+      "grad_norm": 3.166015625,
+      "learning_rate": 4.125416944629754e-05,
+      "loss": 1.1804,
+      "step": 8816
+    },
+    {
+      "epoch": 8.139395338102931,
+      "grad_norm": 3.205078125,
+      "learning_rate": 4.124749833222148e-05,
+      "loss": 1.4077,
+      "step": 8817
+    },
+    {
+      "epoch": 8.140318486037387,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.1240827218145435e-05,
+      "loss": 1.2261,
+      "step": 8818
+    },
+    {
+      "epoch": 8.141241633971845,
+      "grad_norm": 3.0703125,
+      "learning_rate": 4.123415610406938e-05,
+      "loss": 1.3396,
+      "step": 8819
+    },
+    {
+      "epoch": 8.1421647819063,
+      "grad_norm": 3.4921875,
+      "learning_rate": 4.122748498999333e-05,
+      "loss": 1.3157,
+      "step": 8820
+    },
+    {
+      "epoch": 8.143087929840757,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.122081387591728e-05,
+      "loss": 1.4585,
+      "step": 8821
+    },
+    {
+      "epoch": 8.144011077775213,
+      "grad_norm": 3.75390625,
+      "learning_rate": 4.121414276184123e-05,
+      "loss": 1.4016,
+      "step": 8822
+    },
+    {
+      "epoch": 8.14493422570967,
+      "grad_norm": 3.298828125,
+      "learning_rate": 4.1207471647765176e-05,
+      "loss": 1.405,
+      "step": 8823
+    },
+    {
+      "epoch": 8.145857373644127,
+      "grad_norm": 3.1640625,
+      "learning_rate": 4.120080053368913e-05,
+      "loss": 1.3975,
+      "step": 8824
+    },
+    {
+      "epoch": 8.146780521578583,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.119412941961307e-05,
+      "loss": 1.5496,
+      "step": 8825
+    },
+    {
+      "epoch": 8.147703669513039,
+      "grad_norm": 3.19921875,
+      "learning_rate": 4.1187458305537025e-05,
+      "loss": 1.4106,
+      "step": 8826
+    },
+    {
+      "epoch": 8.148626817447497,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.118078719146097e-05,
+      "loss": 0.9629,
+      "step": 8827
+    },
+    {
+      "epoch": 8.149549965381953,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.117411607738493e-05,
+      "loss": 1.3955,
+      "step": 8828
+    },
+    {
+      "epoch": 8.150473113316409,
+      "grad_norm": 3.9296875,
+      "learning_rate": 4.1167444963308875e-05,
+      "loss": 1.3442,
+      "step": 8829
+    },
+    {
+      "epoch": 8.151396261250865,
+      "grad_norm": 2.962890625,
+      "learning_rate": 4.116077384923283e-05,
+      "loss": 1.4004,
+      "step": 8830
+    },
+    {
+      "epoch": 8.152319409185322,
+      "grad_norm": 3.544921875,
+      "learning_rate": 4.115410273515677e-05,
+      "loss": 1.1411,
+      "step": 8831
+    },
+    {
+      "epoch": 8.153242557119778,
+      "grad_norm": 3.923828125,
+      "learning_rate": 4.1147431621080725e-05,
+      "loss": 1.2064,
+      "step": 8832
+    },
+    {
+      "epoch": 8.154165705054234,
+      "grad_norm": 3.443359375,
+      "learning_rate": 4.114076050700467e-05,
+      "loss": 1.1412,
+      "step": 8833
+    },
+    {
+      "epoch": 8.155088852988692,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.113408939292862e-05,
+      "loss": 1.4263,
+      "step": 8834
+    },
+    {
+      "epoch": 8.156012000923148,
+      "grad_norm": 3.689453125,
+      "learning_rate": 4.112741827885257e-05,
+      "loss": 1.3367,
+      "step": 8835
+    },
+    {
+      "epoch": 8.156935148857604,
+      "grad_norm": 3.24609375,
+      "learning_rate": 4.112074716477652e-05,
+      "loss": 1.3408,
+      "step": 8836
+    },
+    {
+      "epoch": 8.15785829679206,
+      "grad_norm": 3.552734375,
+      "learning_rate": 4.1114076050700466e-05,
+      "loss": 1.4678,
+      "step": 8837
+    },
+    {
+      "epoch": 8.158781444726518,
+      "grad_norm": 3.6953125,
+      "learning_rate": 4.110740493662442e-05,
+      "loss": 1.4221,
+      "step": 8838
+    },
+    {
+      "epoch": 8.159704592660974,
+      "grad_norm": 3.126953125,
+      "learning_rate": 4.1100733822548363e-05,
+      "loss": 1.3289,
+      "step": 8839
+    },
+    {
+      "epoch": 8.16062774059543,
+      "grad_norm": 3.412109375,
+      "learning_rate": 4.109406270847232e-05,
+      "loss": 1.366,
+      "step": 8840
+    },
+    {
+      "epoch": 8.161550888529886,
+      "grad_norm": 3.611328125,
+      "learning_rate": 4.108739159439627e-05,
+      "loss": 1.0857,
+      "step": 8841
+    },
+    {
+      "epoch": 8.162474036464344,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.108072048032022e-05,
+      "loss": 1.4529,
+      "step": 8842
+    },
+    {
+      "epoch": 8.1633971843988,
+      "grad_norm": 3.41015625,
+      "learning_rate": 4.1074049366244166e-05,
+      "loss": 1.4067,
+      "step": 8843
+    },
+    {
+      "epoch": 8.164320332333256,
+      "grad_norm": 3.373046875,
+      "learning_rate": 4.106737825216812e-05,
+      "loss": 1.2094,
+      "step": 8844
+    },
+    {
+      "epoch": 8.165243480267712,
+      "grad_norm": 3.4609375,
+      "learning_rate": 4.106070713809206e-05,
+      "loss": 1.3945,
+      "step": 8845
+    },
+    {
+      "epoch": 8.16616662820217,
+      "grad_norm": 3.294921875,
+      "learning_rate": 4.1054036024016015e-05,
+      "loss": 1.373,
+      "step": 8846
+    },
+    {
+      "epoch": 8.167089776136626,
+      "grad_norm": 4.2890625,
+      "learning_rate": 4.104736490993996e-05,
+      "loss": 1.2211,
+      "step": 8847
+    },
+    {
+      "epoch": 8.168012924071082,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.104069379586391e-05,
+      "loss": 1.4771,
+      "step": 8848
+    },
+    {
+      "epoch": 8.16893607200554,
+      "grad_norm": 3.0,
+      "learning_rate": 4.103402268178786e-05,
+      "loss": 1.2644,
+      "step": 8849
+    },
+    {
+      "epoch": 8.169859219939996,
+      "grad_norm": 3.126953125,
+      "learning_rate": 4.102735156771181e-05,
+      "loss": 1.3135,
+      "step": 8850
+    },
+    {
+      "epoch": 8.170782367874452,
+      "grad_norm": 3.64453125,
+      "learning_rate": 4.1020680453635756e-05,
+      "loss": 1.2778,
+      "step": 8851
+    },
+    {
+      "epoch": 8.171705515808908,
+      "grad_norm": 3.38671875,
+      "learning_rate": 4.101400933955971e-05,
+      "loss": 1.2834,
+      "step": 8852
+    },
+    {
+      "epoch": 8.172628663743366,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.100733822548366e-05,
+      "loss": 1.3657,
+      "step": 8853
+    },
+    {
+      "epoch": 8.173551811677822,
+      "grad_norm": 3.861328125,
+      "learning_rate": 4.1000667111407606e-05,
+      "loss": 0.9442,
+      "step": 8854
+    },
+    {
+      "epoch": 8.174474959612278,
+      "grad_norm": 3.38671875,
+      "learning_rate": 4.099399599733156e-05,
+      "loss": 1.5334,
+      "step": 8855
+    },
+    {
+      "epoch": 8.175398107546734,
+      "grad_norm": 3.58984375,
+      "learning_rate": 4.0987324883255504e-05,
+      "loss": 1.5984,
+      "step": 8856
+    },
+    {
+      "epoch": 8.176321255481191,
+      "grad_norm": 3.865234375,
+      "learning_rate": 4.0980653769179456e-05,
+      "loss": 1.3835,
+      "step": 8857
+    },
+    {
+      "epoch": 8.177244403415648,
+      "grad_norm": 3.44140625,
+      "learning_rate": 4.09739826551034e-05,
+      "loss": 1.4343,
+      "step": 8858
+    },
+    {
+      "epoch": 8.178167551350104,
+      "grad_norm": 3.2421875,
+      "learning_rate": 4.0967311541027353e-05,
+      "loss": 1.3489,
+      "step": 8859
+    },
+    {
+      "epoch": 8.17909069928456,
+      "grad_norm": 3.70703125,
+      "learning_rate": 4.09606404269513e-05,
+      "loss": 1.3145,
+      "step": 8860
+    },
+    {
+      "epoch": 8.180013847219017,
+      "grad_norm": 3.134765625,
+      "learning_rate": 4.095396931287525e-05,
+      "loss": 1.3375,
+      "step": 8861
+    },
+    {
+      "epoch": 8.180936995153473,
+      "grad_norm": 3.47265625,
+      "learning_rate": 4.09472981987992e-05,
+      "loss": 1.1229,
+      "step": 8862
+    },
+    {
+      "epoch": 8.18186014308793,
+      "grad_norm": 3.212890625,
+      "learning_rate": 4.094062708472315e-05,
+      "loss": 1.2782,
+      "step": 8863
+    },
+    {
+      "epoch": 8.182783291022385,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.09339559706471e-05,
+      "loss": 1.3396,
+      "step": 8864
+    },
+    {
+      "epoch": 8.183706438956843,
+      "grad_norm": 3.240234375,
+      "learning_rate": 4.092728485657105e-05,
+      "loss": 1.3035,
+      "step": 8865
+    },
+    {
+      "epoch": 8.1846295868913,
+      "grad_norm": 3.794921875,
+      "learning_rate": 4.0920613742495e-05,
+      "loss": 1.1721,
+      "step": 8866
+    },
+    {
+      "epoch": 8.185552734825755,
+      "grad_norm": 2.998046875,
+      "learning_rate": 4.091394262841895e-05,
+      "loss": 1.2795,
+      "step": 8867
+    },
+    {
+      "epoch": 8.186475882760213,
+      "grad_norm": 3.775390625,
+      "learning_rate": 4.0907271514342896e-05,
+      "loss": 1.3296,
+      "step": 8868
+    },
+    {
+      "epoch": 8.18739903069467,
+      "grad_norm": 3.439453125,
+      "learning_rate": 4.090060040026685e-05,
+      "loss": 1.2827,
+      "step": 8869
+    },
+    {
+      "epoch": 8.188322178629125,
+      "grad_norm": 3.267578125,
+      "learning_rate": 4.0893929286190794e-05,
+      "loss": 1.4095,
+      "step": 8870
+    },
+    {
+      "epoch": 8.189245326563581,
+      "grad_norm": 3.33984375,
+      "learning_rate": 4.0887258172114746e-05,
+      "loss": 1.2924,
+      "step": 8871
+    },
+    {
+      "epoch": 8.190168474498039,
+      "grad_norm": 3.75,
+      "learning_rate": 4.088058705803869e-05,
+      "loss": 1.3195,
+      "step": 8872
+    },
+    {
+      "epoch": 8.191091622432495,
+      "grad_norm": 3.3046875,
+      "learning_rate": 4.0873915943962644e-05,
+      "loss": 1.4067,
+      "step": 8873
+    },
+    {
+      "epoch": 8.192014770366951,
+      "grad_norm": 3.19921875,
+      "learning_rate": 4.086724482988659e-05,
+      "loss": 1.594,
+      "step": 8874
+    },
+    {
+      "epoch": 8.192937918301407,
+      "grad_norm": 4.16015625,
+      "learning_rate": 4.086057371581054e-05,
+      "loss": 1.1887,
+      "step": 8875
+    },
+    {
+      "epoch": 8.193861066235865,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.085390260173449e-05,
+      "loss": 1.415,
+      "step": 8876
+    },
+    {
+      "epoch": 8.19478421417032,
+      "grad_norm": 3.7890625,
+      "learning_rate": 4.084723148765844e-05,
+      "loss": 1.4124,
+      "step": 8877
+    },
+    {
+      "epoch": 8.195707362104777,
+      "grad_norm": 3.556640625,
+      "learning_rate": 4.084056037358239e-05,
+      "loss": 1.1218,
+      "step": 8878
+    },
+    {
+      "epoch": 8.196630510039233,
+      "grad_norm": 3.751953125,
+      "learning_rate": 4.083388925950634e-05,
+      "loss": 1.3916,
+      "step": 8879
+    },
+    {
+      "epoch": 8.19755365797369,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.082721814543029e-05,
+      "loss": 1.3704,
+      "step": 8880
+    },
+    {
+      "epoch": 8.198476805908147,
+      "grad_norm": 3.380859375,
+      "learning_rate": 4.082054703135424e-05,
+      "loss": 1.6228,
+      "step": 8881
+    },
+    {
+      "epoch": 8.199399953842603,
+      "grad_norm": 3.1328125,
+      "learning_rate": 4.0813875917278186e-05,
+      "loss": 1.321,
+      "step": 8882
+    },
+    {
+      "epoch": 8.20032310177706,
+      "grad_norm": 3.169921875,
+      "learning_rate": 4.080720480320214e-05,
+      "loss": 1.3005,
+      "step": 8883
+    },
+    {
+      "epoch": 8.201246249711517,
+      "grad_norm": 3.494140625,
+      "learning_rate": 4.0800533689126084e-05,
+      "loss": 1.1873,
+      "step": 8884
+    },
+    {
+      "epoch": 8.202169397645973,
+      "grad_norm": 3.1015625,
+      "learning_rate": 4.0793862575050036e-05,
+      "loss": 1.2692,
+      "step": 8885
+    },
+    {
+      "epoch": 8.203092545580429,
+      "grad_norm": 3.037109375,
+      "learning_rate": 4.078719146097398e-05,
+      "loss": 0.932,
+      "step": 8886
+    },
+    {
+      "epoch": 8.204015693514886,
+      "grad_norm": 3.8671875,
+      "learning_rate": 4.0780520346897934e-05,
+      "loss": 1.4019,
+      "step": 8887
+    },
+    {
+      "epoch": 8.204938841449342,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.077384923282188e-05,
+      "loss": 1.4724,
+      "step": 8888
+    },
+    {
+      "epoch": 8.205861989383799,
+      "grad_norm": 3.125,
+      "learning_rate": 4.076717811874583e-05,
+      "loss": 1.2295,
+      "step": 8889
+    },
+    {
+      "epoch": 8.206785137318255,
+      "grad_norm": 2.873046875,
+      "learning_rate": 4.0760507004669784e-05,
+      "loss": 1.0173,
+      "step": 8890
+    },
+    {
+      "epoch": 8.207708285252712,
+      "grad_norm": 3.4609375,
+      "learning_rate": 4.0753835890593736e-05,
+      "loss": 1.3713,
+      "step": 8891
+    },
+    {
+      "epoch": 8.208631433187168,
+      "grad_norm": 3.51171875,
+      "learning_rate": 4.074716477651768e-05,
+      "loss": 1.24,
+      "step": 8892
+    },
+    {
+      "epoch": 8.209554581121624,
+      "grad_norm": 3.306640625,
+      "learning_rate": 4.0740493662441634e-05,
+      "loss": 1.2797,
+      "step": 8893
+    },
+    {
+      "epoch": 8.21047772905608,
+      "grad_norm": 3.3515625,
+      "learning_rate": 4.073382254836558e-05,
+      "loss": 1.4685,
+      "step": 8894
+    },
+    {
+      "epoch": 8.211400876990538,
+      "grad_norm": 3.2109375,
+      "learning_rate": 4.072715143428953e-05,
+      "loss": 1.238,
+      "step": 8895
+    },
+    {
+      "epoch": 8.212324024924994,
+      "grad_norm": 3.916015625,
+      "learning_rate": 4.072048032021348e-05,
+      "loss": 1.5,
+      "step": 8896
+    },
+    {
+      "epoch": 8.21324717285945,
+      "grad_norm": 3.41796875,
+      "learning_rate": 4.071380920613743e-05,
+      "loss": 1.1477,
+      "step": 8897
+    },
+    {
+      "epoch": 8.214170320793908,
+      "grad_norm": 3.244140625,
+      "learning_rate": 4.0707138092061374e-05,
+      "loss": 1.1691,
+      "step": 8898
+    },
+    {
+      "epoch": 8.215093468728364,
+      "grad_norm": 3.05859375,
+      "learning_rate": 4.0700466977985326e-05,
+      "loss": 1.1445,
+      "step": 8899
+    },
+    {
+      "epoch": 8.21601661666282,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.069379586390927e-05,
+      "loss": 1.1194,
+      "step": 8900
+    },
+    {
+      "epoch": 8.216939764597276,
+      "grad_norm": 3.03515625,
+      "learning_rate": 4.0687124749833224e-05,
+      "loss": 1.3857,
+      "step": 8901
+    },
+    {
+      "epoch": 8.217862912531734,
+      "grad_norm": 3.13671875,
+      "learning_rate": 4.068045363575717e-05,
+      "loss": 1.1508,
+      "step": 8902
+    },
+    {
+      "epoch": 8.21878606046619,
+      "grad_norm": 3.90234375,
+      "learning_rate": 4.067378252168113e-05,
+      "loss": 1.1752,
+      "step": 8903
+    },
+    {
+      "epoch": 8.219709208400646,
+      "grad_norm": 4.3359375,
+      "learning_rate": 4.0667111407605074e-05,
+      "loss": 1.3699,
+      "step": 8904
+    },
+    {
+      "epoch": 8.220632356335102,
+      "grad_norm": 3.05078125,
+      "learning_rate": 4.0660440293529026e-05,
+      "loss": 1.5142,
+      "step": 8905
+    },
+    {
+      "epoch": 8.22155550426956,
+      "grad_norm": 3.27734375,
+      "learning_rate": 4.065376917945297e-05,
+      "loss": 1.3655,
+      "step": 8906
+    },
+    {
+      "epoch": 8.222478652204016,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.0647098065376924e-05,
+      "loss": 1.2344,
+      "step": 8907
+    },
+    {
+      "epoch": 8.223401800138472,
+      "grad_norm": 2.96484375,
+      "learning_rate": 4.064042695130087e-05,
+      "loss": 1.1958,
+      "step": 8908
+    },
+    {
+      "epoch": 8.224324948072928,
+      "grad_norm": 3.669921875,
+      "learning_rate": 4.063375583722482e-05,
+      "loss": 1.2773,
+      "step": 8909
+    },
+    {
+      "epoch": 8.225248096007386,
+      "grad_norm": 3.806640625,
+      "learning_rate": 4.062708472314877e-05,
+      "loss": 1.5022,
+      "step": 8910
+    },
+    {
+      "epoch": 8.226171243941842,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.062041360907272e-05,
+      "loss": 1.5562,
+      "step": 8911
+    },
+    {
+      "epoch": 8.227094391876298,
+      "grad_norm": 3.130859375,
+      "learning_rate": 4.0613742494996665e-05,
+      "loss": 1.3906,
+      "step": 8912
+    },
+    {
+      "epoch": 8.228017539810754,
+      "grad_norm": 3.779296875,
+      "learning_rate": 4.060707138092062e-05,
+      "loss": 1.1034,
+      "step": 8913
+    },
+    {
+      "epoch": 8.228940687745212,
+      "grad_norm": 3.681640625,
+      "learning_rate": 4.060040026684456e-05,
+      "loss": 1.1509,
+      "step": 8914
+    },
+    {
+      "epoch": 8.229863835679668,
+      "grad_norm": 3.56640625,
+      "learning_rate": 4.0593729152768514e-05,
+      "loss": 1.2485,
+      "step": 8915
+    },
+    {
+      "epoch": 8.230786983614124,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.0587058038692467e-05,
+      "loss": 1.2159,
+      "step": 8916
+    },
+    {
+      "epoch": 8.231710131548581,
+      "grad_norm": 3.1953125,
+      "learning_rate": 4.058038692461641e-05,
+      "loss": 1.5657,
+      "step": 8917
+    },
+    {
+      "epoch": 8.232633279483037,
+      "grad_norm": 3.32421875,
+      "learning_rate": 4.0573715810540364e-05,
+      "loss": 1.2301,
+      "step": 8918
+    },
+    {
+      "epoch": 8.233556427417493,
+      "grad_norm": 3.611328125,
+      "learning_rate": 4.056704469646431e-05,
+      "loss": 1.4878,
+      "step": 8919
+    },
+    {
+      "epoch": 8.23447957535195,
+      "grad_norm": 3.2734375,
+      "learning_rate": 4.056037358238826e-05,
+      "loss": 1.3611,
+      "step": 8920
+    },
+    {
+      "epoch": 8.235402723286407,
+      "grad_norm": 3.646484375,
+      "learning_rate": 4.055370246831221e-05,
+      "loss": 1.1573,
+      "step": 8921
+    },
+    {
+      "epoch": 8.236325871220863,
+      "grad_norm": 3.560546875,
+      "learning_rate": 4.054703135423616e-05,
+      "loss": 1.4678,
+      "step": 8922
+    },
+    {
+      "epoch": 8.23724901915532,
+      "grad_norm": 3.724609375,
+      "learning_rate": 4.0540360240160105e-05,
+      "loss": 1.4048,
+      "step": 8923
+    },
+    {
+      "epoch": 8.238172167089775,
+      "grad_norm": 3.033203125,
+      "learning_rate": 4.053368912608406e-05,
+      "loss": 1.3064,
+      "step": 8924
+    },
+    {
+      "epoch": 8.239095315024233,
+      "grad_norm": 3.076171875,
+      "learning_rate": 4.0527018012008e-05,
+      "loss": 0.9751,
+      "step": 8925
+    },
+    {
+      "epoch": 8.24001846295869,
+      "grad_norm": 4.046875,
+      "learning_rate": 4.0520346897931955e-05,
+      "loss": 1.4346,
+      "step": 8926
+    },
+    {
+      "epoch": 8.240941610893145,
+      "grad_norm": 3.123046875,
+      "learning_rate": 4.05136757838559e-05,
+      "loss": 1.3586,
+      "step": 8927
+    },
+    {
+      "epoch": 8.241864758827601,
+      "grad_norm": 3.478515625,
+      "learning_rate": 4.050700466977986e-05,
+      "loss": 1.4041,
+      "step": 8928
+    },
+    {
+      "epoch": 8.242787906762059,
+      "grad_norm": 3.892578125,
+      "learning_rate": 4.0500333555703805e-05,
+      "loss": 1.1309,
+      "step": 8929
+    },
+    {
+      "epoch": 8.243711054696515,
+      "grad_norm": 3.486328125,
+      "learning_rate": 4.049366244162776e-05,
+      "loss": 1.1774,
+      "step": 8930
+    },
+    {
+      "epoch": 8.244634202630971,
+      "grad_norm": 3.626953125,
+      "learning_rate": 4.04869913275517e-05,
+      "loss": 1.3665,
+      "step": 8931
+    },
+    {
+      "epoch": 8.245557350565429,
+      "grad_norm": 3.55078125,
+      "learning_rate": 4.0480320213475654e-05,
+      "loss": 1.1676,
+      "step": 8932
+    },
+    {
+      "epoch": 8.246480498499885,
+      "grad_norm": 3.19921875,
+      "learning_rate": 4.04736490993996e-05,
+      "loss": 1.2441,
+      "step": 8933
+    },
+    {
+      "epoch": 8.247403646434341,
+      "grad_norm": 3.041015625,
+      "learning_rate": 4.046697798532355e-05,
+      "loss": 1.3718,
+      "step": 8934
+    },
+    {
+      "epoch": 8.248326794368797,
+      "grad_norm": 3.23828125,
+      "learning_rate": 4.04603068712475e-05,
+      "loss": 1.0609,
+      "step": 8935
+    },
+    {
+      "epoch": 8.249249942303255,
+      "grad_norm": 3.5390625,
+      "learning_rate": 4.045363575717145e-05,
+      "loss": 1.1539,
+      "step": 8936
+    },
+    {
+      "epoch": 8.25017309023771,
+      "grad_norm": 3.2734375,
+      "learning_rate": 4.0446964643095395e-05,
+      "loss": 1.3633,
+      "step": 8937
+    },
+    {
+      "epoch": 8.251096238172167,
+      "grad_norm": 3.802734375,
+      "learning_rate": 4.044029352901935e-05,
+      "loss": 1.2302,
+      "step": 8938
+    },
+    {
+      "epoch": 8.252019386106623,
+      "grad_norm": 2.384765625,
+      "learning_rate": 4.043362241494329e-05,
+      "loss": 1.0687,
+      "step": 8939
+    },
+    {
+      "epoch": 8.25294253404108,
+      "grad_norm": 3.322265625,
+      "learning_rate": 4.0426951300867245e-05,
+      "loss": 1.499,
+      "step": 8940
+    },
+    {
+      "epoch": 8.253865681975537,
+      "grad_norm": 3.3828125,
+      "learning_rate": 4.04202801867912e-05,
+      "loss": 1.4102,
+      "step": 8941
+    },
+    {
+      "epoch": 8.254788829909993,
+      "grad_norm": 3.1015625,
+      "learning_rate": 4.041360907271515e-05,
+      "loss": 1.354,
+      "step": 8942
+    },
+    {
+      "epoch": 8.255711977844449,
+      "grad_norm": 2.998046875,
+      "learning_rate": 4.0406937958639095e-05,
+      "loss": 1.6616,
+      "step": 8943
+    },
+    {
+      "epoch": 8.256635125778907,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.040026684456305e-05,
+      "loss": 1.519,
+      "step": 8944
+    },
+    {
+      "epoch": 8.257558273713363,
+      "grad_norm": 3.283203125,
+      "learning_rate": 4.039359573048699e-05,
+      "loss": 1.3672,
+      "step": 8945
+    },
+    {
+      "epoch": 8.258481421647819,
+      "grad_norm": 3.451171875,
+      "learning_rate": 4.0386924616410945e-05,
+      "loss": 1.1671,
+      "step": 8946
+    },
+    {
+      "epoch": 8.259404569582276,
+      "grad_norm": 3.05859375,
+      "learning_rate": 4.038025350233489e-05,
+      "loss": 1.3547,
+      "step": 8947
+    },
+    {
+      "epoch": 8.260327717516732,
+      "grad_norm": 3.583984375,
+      "learning_rate": 4.037358238825884e-05,
+      "loss": 1.2604,
+      "step": 8948
+    },
+    {
+      "epoch": 8.261250865451188,
+      "grad_norm": 3.427734375,
+      "learning_rate": 4.036691127418279e-05,
+      "loss": 1.1727,
+      "step": 8949
+    },
+    {
+      "epoch": 8.262174013385644,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.036024016010674e-05,
+      "loss": 1.1461,
+      "step": 8950
+    },
+    {
+      "epoch": 8.263097161320102,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.0353569046030685e-05,
+      "loss": 1.5303,
+      "step": 8951
+    },
+    {
+      "epoch": 8.264020309254558,
+      "grad_norm": 3.03515625,
+      "learning_rate": 4.034689793195464e-05,
+      "loss": 1.2949,
+      "step": 8952
+    },
+    {
+      "epoch": 8.264943457189014,
+      "grad_norm": 3.9296875,
+      "learning_rate": 4.034022681787859e-05,
+      "loss": 1.2803,
+      "step": 8953
+    },
+    {
+      "epoch": 8.26586660512347,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.033355570380254e-05,
+      "loss": 1.1899,
+      "step": 8954
+    },
+    {
+      "epoch": 8.266789753057928,
+      "grad_norm": 3.6171875,
+      "learning_rate": 4.032688458972649e-05,
+      "loss": 1.2363,
+      "step": 8955
+    },
+    {
+      "epoch": 8.267712900992384,
+      "grad_norm": 3.41015625,
+      "learning_rate": 4.032021347565044e-05,
+      "loss": 1.3684,
+      "step": 8956
+    },
+    {
+      "epoch": 8.26863604892684,
+      "grad_norm": 3.072265625,
+      "learning_rate": 4.0313542361574385e-05,
+      "loss": 1.3619,
+      "step": 8957
+    },
+    {
+      "epoch": 8.269559196861296,
+      "grad_norm": 3.36328125,
+      "learning_rate": 4.030687124749834e-05,
+      "loss": 1.2333,
+      "step": 8958
+    },
+    {
+      "epoch": 8.270482344795754,
+      "grad_norm": 3.33203125,
+      "learning_rate": 4.030020013342228e-05,
+      "loss": 1.3142,
+      "step": 8959
+    },
+    {
+      "epoch": 8.27140549273021,
+      "grad_norm": 3.603515625,
+      "learning_rate": 4.0293529019346235e-05,
+      "loss": 1.2654,
+      "step": 8960
+    },
+    {
+      "epoch": 8.272328640664666,
+      "grad_norm": 3.767578125,
+      "learning_rate": 4.028685790527018e-05,
+      "loss": 1.2208,
+      "step": 8961
+    },
+    {
+      "epoch": 8.273251788599122,
+      "grad_norm": 3.462890625,
+      "learning_rate": 4.028018679119413e-05,
+      "loss": 1.23,
+      "step": 8962
+    },
+    {
+      "epoch": 8.27417493653358,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.027351567711808e-05,
+      "loss": 1.3027,
+      "step": 8963
+    },
+    {
+      "epoch": 8.275098084468036,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.026684456304203e-05,
+      "loss": 1.3389,
+      "step": 8964
+    },
+    {
+      "epoch": 8.276021232402492,
+      "grad_norm": 3.685546875,
+      "learning_rate": 4.0260173448965976e-05,
+      "loss": 1.1565,
+      "step": 8965
+    },
+    {
+      "epoch": 8.27694438033695,
+      "grad_norm": 3.3984375,
+      "learning_rate": 4.025350233488993e-05,
+      "loss": 1.3351,
+      "step": 8966
+    },
+    {
+      "epoch": 8.277867528271406,
+      "grad_norm": 3.228515625,
+      "learning_rate": 4.024683122081388e-05,
+      "loss": 1.2034,
+      "step": 8967
+    },
+    {
+      "epoch": 8.278790676205862,
+      "grad_norm": 3.232421875,
+      "learning_rate": 4.0240160106737825e-05,
+      "loss": 1.2457,
+      "step": 8968
+    },
+    {
+      "epoch": 8.279713824140318,
+      "grad_norm": 3.53515625,
+      "learning_rate": 4.023348899266178e-05,
+      "loss": 1.2761,
+      "step": 8969
+    },
+    {
+      "epoch": 8.280636972074776,
+      "grad_norm": 3.345703125,
+      "learning_rate": 4.022681787858572e-05,
+      "loss": 1.3396,
+      "step": 8970
+    },
+    {
+      "epoch": 8.281560120009232,
+      "grad_norm": 3.60546875,
+      "learning_rate": 4.0220146764509675e-05,
+      "loss": 1.2747,
+      "step": 8971
+    },
+    {
+      "epoch": 8.282483267943688,
+      "grad_norm": 3.943359375,
+      "learning_rate": 4.021347565043362e-05,
+      "loss": 1.384,
+      "step": 8972
+    },
+    {
+      "epoch": 8.283406415878144,
+      "grad_norm": 3.705078125,
+      "learning_rate": 4.020680453635757e-05,
+      "loss": 1.3486,
+      "step": 8973
+    },
+    {
+      "epoch": 8.284329563812602,
+      "grad_norm": 3.52734375,
+      "learning_rate": 4.0200133422281525e-05,
+      "loss": 1.4568,
+      "step": 8974
+    },
+    {
+      "epoch": 8.285252711747058,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.019346230820547e-05,
+      "loss": 1.5154,
+      "step": 8975
+    },
+    {
+      "epoch": 8.286175859681514,
+      "grad_norm": 3.10546875,
+      "learning_rate": 4.018679119412942e-05,
+      "loss": 1.1974,
+      "step": 8976
+    },
+    {
+      "epoch": 8.28709900761597,
+      "grad_norm": 3.005859375,
+      "learning_rate": 4.018012008005337e-05,
+      "loss": 1.4165,
+      "step": 8977
+    },
+    {
+      "epoch": 8.288022155550427,
+      "grad_norm": 3.42578125,
+      "learning_rate": 4.017344896597732e-05,
+      "loss": 1.3113,
+      "step": 8978
+    },
+    {
+      "epoch": 8.288945303484883,
+      "grad_norm": 3.7734375,
+      "learning_rate": 4.016677785190127e-05,
+      "loss": 1.2192,
+      "step": 8979
+    },
+    {
+      "epoch": 8.28986845141934,
+      "grad_norm": 3.24609375,
+      "learning_rate": 4.016010673782522e-05,
+      "loss": 1.4775,
+      "step": 8980
+    },
+    {
+      "epoch": 8.290791599353797,
+      "grad_norm": 3.4921875,
+      "learning_rate": 4.015343562374917e-05,
+      "loss": 1.3589,
+      "step": 8981
+    },
+    {
+      "epoch": 8.291714747288253,
+      "grad_norm": 3.634765625,
+      "learning_rate": 4.0146764509673116e-05,
+      "loss": 1.5027,
+      "step": 8982
+    },
+    {
+      "epoch": 8.29263789522271,
+      "grad_norm": 3.517578125,
+      "learning_rate": 4.014009339559707e-05,
+      "loss": 1.4841,
+      "step": 8983
+    },
+    {
+      "epoch": 8.293561043157165,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.013342228152101e-05,
+      "loss": 1.2449,
+      "step": 8984
+    },
+    {
+      "epoch": 8.294484191091623,
+      "grad_norm": 3.716796875,
+      "learning_rate": 4.0126751167444966e-05,
+      "loss": 1.3623,
+      "step": 8985
+    },
+    {
+      "epoch": 8.29540733902608,
+      "grad_norm": 3.822265625,
+      "learning_rate": 4.012008005336891e-05,
+      "loss": 1.3047,
+      "step": 8986
+    },
+    {
+      "epoch": 8.296330486960535,
+      "grad_norm": 3.1171875,
+      "learning_rate": 4.011340893929286e-05,
+      "loss": 1.408,
+      "step": 8987
+    },
+    {
+      "epoch": 8.297253634894991,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.010673782521681e-05,
+      "loss": 1.2189,
+      "step": 8988
+    },
+    {
+      "epoch": 8.298176782829449,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.010006671114076e-05,
+      "loss": 1.2537,
+      "step": 8989
+    },
+    {
+      "epoch": 8.299099930763905,
+      "grad_norm": 3.58984375,
+      "learning_rate": 4.0093395597064706e-05,
+      "loss": 1.1875,
+      "step": 8990
+    },
+    {
+      "epoch": 8.300023078698361,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.0086724482988665e-05,
+      "loss": 1.2585,
+      "step": 8991
+    },
+    {
+      "epoch": 8.300946226632817,
+      "grad_norm": 3.064453125,
+      "learning_rate": 4.008005336891261e-05,
+      "loss": 1.3989,
+      "step": 8992
+    },
+    {
+      "epoch": 8.301869374567275,
+      "grad_norm": 2.853515625,
+      "learning_rate": 4.007338225483656e-05,
+      "loss": 1.1,
+      "step": 8993
+    },
+    {
+      "epoch": 8.302792522501731,
+      "grad_norm": 3.625,
+      "learning_rate": 4.006671114076051e-05,
+      "loss": 1.0814,
+      "step": 8994
+    },
+    {
+      "epoch": 8.303715670436187,
+      "grad_norm": 3.435546875,
+      "learning_rate": 4.006004002668446e-05,
+      "loss": 1.3112,
+      "step": 8995
+    },
+    {
+      "epoch": 8.304638818370645,
+      "grad_norm": 3.044921875,
+      "learning_rate": 4.0053368912608406e-05,
+      "loss": 1.377,
+      "step": 8996
+    },
+    {
+      "epoch": 8.3055619663051,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.004669779853236e-05,
+      "loss": 1.1033,
+      "step": 8997
+    },
+    {
+      "epoch": 8.306485114239557,
+      "grad_norm": 3.755859375,
+      "learning_rate": 4.0040026684456304e-05,
+      "loss": 1.3997,
+      "step": 8998
+    },
+    {
+      "epoch": 8.307408262174013,
+      "grad_norm": 2.994140625,
+      "learning_rate": 4.0033355570380256e-05,
+      "loss": 1.344,
+      "step": 8999
+    },
+    {
+      "epoch": 8.30833141010847,
+      "grad_norm": 3.4765625,
+      "learning_rate": 4.00266844563042e-05,
+      "loss": 1.3845,
+      "step": 9000
+    },
+    {
+      "epoch": 8.30833141010847,
+      "eval_loss": 0.876953125,
+      "eval_runtime": 23.3612,
+      "eval_samples_per_second": 976.022,
+      "eval_steps_per_second": 61.041,
+      "step": 9000
+    },
+    {
+      "epoch": 8.309254558042927,
+      "grad_norm": 3.2578125,
+      "learning_rate": 4.0020013342228153e-05,
+      "loss": 1.5564,
+      "step": 9001
+    },
+    {
+      "epoch": 8.310177705977383,
+      "grad_norm": 3.306640625,
+      "learning_rate": 4.00133422281521e-05,
+      "loss": 1.1855,
+      "step": 9002
+    },
+    {
+      "epoch": 8.311100853911839,
+      "grad_norm": 3.76171875,
+      "learning_rate": 4.000667111407606e-05,
+      "loss": 1.5393,
+      "step": 9003
+    },
+    {
+      "epoch": 8.312024001846297,
+      "grad_norm": 3.53125,
+      "learning_rate": 4e-05,
+      "loss": 1.2102,
+      "step": 9004
+    },
+    {
+      "epoch": 8.312947149780753,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.9993328885923955e-05,
+      "loss": 1.4893,
+      "step": 9005
+    },
+    {
+      "epoch": 8.313870297715209,
+      "grad_norm": 3.19921875,
+      "learning_rate": 3.99866577718479e-05,
+      "loss": 1.2437,
+      "step": 9006
+    },
+    {
+      "epoch": 8.314793445649665,
+      "grad_norm": 3.25,
+      "learning_rate": 3.997998665777185e-05,
+      "loss": 1.177,
+      "step": 9007
+    },
+    {
+      "epoch": 8.315716593584122,
+      "grad_norm": 3.373046875,
+      "learning_rate": 3.99733155436958e-05,
+      "loss": 1.1193,
+      "step": 9008
+    },
+    {
+      "epoch": 8.316639741518578,
+      "grad_norm": 4.06640625,
+      "learning_rate": 3.996664442961975e-05,
+      "loss": 1.1733,
+      "step": 9009
+    },
+    {
+      "epoch": 8.317562889453034,
+      "grad_norm": 3.5234375,
+      "learning_rate": 3.9959973315543696e-05,
+      "loss": 1.4382,
+      "step": 9010
+    },
+    {
+      "epoch": 8.31848603738749,
+      "grad_norm": 3.498046875,
+      "learning_rate": 3.995330220146765e-05,
+      "loss": 1.2434,
+      "step": 9011
+    },
+    {
+      "epoch": 8.319409185321948,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.9946631087391594e-05,
+      "loss": 1.3839,
+      "step": 9012
+    },
+    {
+      "epoch": 8.320332333256404,
+      "grad_norm": 3.568359375,
+      "learning_rate": 3.9939959973315546e-05,
+      "loss": 1.3105,
+      "step": 9013
+    },
+    {
+      "epoch": 8.32125548119086,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.993328885923949e-05,
+      "loss": 1.245,
+      "step": 9014
+    },
+    {
+      "epoch": 8.322178629125318,
+      "grad_norm": 3.958984375,
+      "learning_rate": 3.9926617745163444e-05,
+      "loss": 1.0032,
+      "step": 9015
+    },
+    {
+      "epoch": 8.323101777059774,
+      "grad_norm": 3.029296875,
+      "learning_rate": 3.9919946631087396e-05,
+      "loss": 1.4097,
+      "step": 9016
+    },
+    {
+      "epoch": 8.32402492499423,
+      "grad_norm": 2.990234375,
+      "learning_rate": 3.991327551701135e-05,
+      "loss": 1.1003,
+      "step": 9017
+    },
+    {
+      "epoch": 8.324948072928686,
+      "grad_norm": 3.060546875,
+      "learning_rate": 3.9906604402935294e-05,
+      "loss": 1.2717,
+      "step": 9018
+    },
+    {
+      "epoch": 8.325871220863144,
+      "grad_norm": 3.376953125,
+      "learning_rate": 3.9899933288859246e-05,
+      "loss": 1.3533,
+      "step": 9019
+    },
+    {
+      "epoch": 8.3267943687976,
+      "grad_norm": 3.48046875,
+      "learning_rate": 3.989326217478319e-05,
+      "loss": 1.3499,
+      "step": 9020
+    },
+    {
+      "epoch": 8.327717516732056,
+      "grad_norm": 3.06640625,
+      "learning_rate": 3.988659106070714e-05,
+      "loss": 1.3203,
+      "step": 9021
+    },
+    {
+      "epoch": 8.328640664666512,
+      "grad_norm": 3.337890625,
+      "learning_rate": 3.987991994663109e-05,
+      "loss": 1.2582,
+      "step": 9022
+    },
+    {
+      "epoch": 8.32956381260097,
+      "grad_norm": 4.27734375,
+      "learning_rate": 3.987324883255504e-05,
+      "loss": 1.2097,
+      "step": 9023
+    },
+    {
+      "epoch": 8.330486960535426,
+      "grad_norm": 3.39453125,
+      "learning_rate": 3.9866577718478986e-05,
+      "loss": 1.3142,
+      "step": 9024
+    },
+    {
+      "epoch": 8.331410108469882,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.985990660440294e-05,
+      "loss": 1.3987,
+      "step": 9025
+    },
+    {
+      "epoch": 8.332333256404338,
+      "grad_norm": 2.798828125,
+      "learning_rate": 3.9853235490326884e-05,
+      "loss": 1.1303,
+      "step": 9026
+    },
+    {
+      "epoch": 8.333256404338796,
+      "grad_norm": 3.466796875,
+      "learning_rate": 3.9846564376250836e-05,
+      "loss": 1.3518,
+      "step": 9027
+    },
+    {
+      "epoch": 8.334179552273252,
+      "grad_norm": 2.791015625,
+      "learning_rate": 3.983989326217479e-05,
+      "loss": 1.2668,
+      "step": 9028
+    },
+    {
+      "epoch": 8.335102700207708,
+      "grad_norm": 3.375,
+      "learning_rate": 3.9833222148098734e-05,
+      "loss": 1.4985,
+      "step": 9029
+    },
+    {
+      "epoch": 8.336025848142166,
+      "grad_norm": 4.1796875,
+      "learning_rate": 3.9826551034022686e-05,
+      "loss": 0.9865,
+      "step": 9030
+    },
+    {
+      "epoch": 8.336948996076622,
+      "grad_norm": 3.244140625,
+      "learning_rate": 3.981987991994663e-05,
+      "loss": 1.193,
+      "step": 9031
+    },
+    {
+      "epoch": 8.337872144011078,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.9813208805870584e-05,
+      "loss": 0.9051,
+      "step": 9032
+    },
+    {
+      "epoch": 8.338795291945534,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.980653769179453e-05,
+      "loss": 1.3105,
+      "step": 9033
+    },
+    {
+      "epoch": 8.339718439879992,
+      "grad_norm": 3.263671875,
+      "learning_rate": 3.979986657771848e-05,
+      "loss": 1.3748,
+      "step": 9034
+    },
+    {
+      "epoch": 8.340641587814448,
+      "grad_norm": 2.998046875,
+      "learning_rate": 3.979319546364243e-05,
+      "loss": 1.2866,
+      "step": 9035
+    },
+    {
+      "epoch": 8.341564735748904,
+      "grad_norm": 3.64453125,
+      "learning_rate": 3.978652434956638e-05,
+      "loss": 1.5391,
+      "step": 9036
+    },
+    {
+      "epoch": 8.34248788368336,
+      "grad_norm": 2.869140625,
+      "learning_rate": 3.9779853235490324e-05,
+      "loss": 1.4187,
+      "step": 9037
+    },
+    {
+      "epoch": 8.343411031617817,
+      "grad_norm": 3.220703125,
+      "learning_rate": 3.977318212141428e-05,
+      "loss": 1.4141,
+      "step": 9038
+    },
+    {
+      "epoch": 8.344334179552273,
+      "grad_norm": 2.82421875,
+      "learning_rate": 3.976651100733822e-05,
+      "loss": 1.4358,
+      "step": 9039
+    },
+    {
+      "epoch": 8.34525732748673,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.9759839893262174e-05,
+      "loss": 1.2896,
+      "step": 9040
+    },
+    {
+      "epoch": 8.346180475421185,
+      "grad_norm": 2.9453125,
+      "learning_rate": 3.9753168779186127e-05,
+      "loss": 1.4688,
+      "step": 9041
+    },
+    {
+      "epoch": 8.347103623355643,
+      "grad_norm": 3.517578125,
+      "learning_rate": 3.974649766511008e-05,
+      "loss": 1.3435,
+      "step": 9042
+    },
+    {
+      "epoch": 8.3480267712901,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.9739826551034024e-05,
+      "loss": 1.0772,
+      "step": 9043
+    },
+    {
+      "epoch": 8.348949919224555,
+      "grad_norm": 3.458984375,
+      "learning_rate": 3.9733155436957976e-05,
+      "loss": 1.3245,
+      "step": 9044
+    },
+    {
+      "epoch": 8.349873067159013,
+      "grad_norm": 3.349609375,
+      "learning_rate": 3.972648432288192e-05,
+      "loss": 1.0114,
+      "step": 9045
+    },
+    {
+      "epoch": 8.35079621509347,
+      "grad_norm": 3.208984375,
+      "learning_rate": 3.9719813208805874e-05,
+      "loss": 1.1206,
+      "step": 9046
+    },
+    {
+      "epoch": 8.351719363027925,
+      "grad_norm": 3.494140625,
+      "learning_rate": 3.971314209472982e-05,
+      "loss": 1.291,
+      "step": 9047
+    },
+    {
+      "epoch": 8.352642510962381,
+      "grad_norm": 2.884765625,
+      "learning_rate": 3.970647098065377e-05,
+      "loss": 1.3604,
+      "step": 9048
+    },
+    {
+      "epoch": 8.353565658896839,
+      "grad_norm": 3.1171875,
+      "learning_rate": 3.969979986657772e-05,
+      "loss": 1.5068,
+      "step": 9049
+    },
+    {
+      "epoch": 8.354488806831295,
+      "grad_norm": 3.228515625,
+      "learning_rate": 3.969312875250167e-05,
+      "loss": 1.4026,
+      "step": 9050
+    },
+    {
+      "epoch": 8.355411954765751,
+      "grad_norm": 3.11328125,
+      "learning_rate": 3.9686457638425615e-05,
+      "loss": 0.7751,
+      "step": 9051
+    },
+    {
+      "epoch": 8.356335102700207,
+      "grad_norm": 3.21484375,
+      "learning_rate": 3.967978652434957e-05,
+      "loss": 1.2029,
+      "step": 9052
+    },
+    {
+      "epoch": 8.357258250634665,
+      "grad_norm": 3.05859375,
+      "learning_rate": 3.967311541027352e-05,
+      "loss": 1.365,
+      "step": 9053
+    },
+    {
+      "epoch": 8.358181398569121,
+      "grad_norm": 3.451171875,
+      "learning_rate": 3.966644429619747e-05,
+      "loss": 1.3073,
+      "step": 9054
+    },
+    {
+      "epoch": 8.359104546503577,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.965977318212142e-05,
+      "loss": 1.3359,
+      "step": 9055
+    },
+    {
+      "epoch": 8.360027694438033,
+      "grad_norm": 3.85546875,
+      "learning_rate": 3.965310206804537e-05,
+      "loss": 1.2661,
+      "step": 9056
+    },
+    {
+      "epoch": 8.36095084237249,
+      "grad_norm": 3.46484375,
+      "learning_rate": 3.9646430953969314e-05,
+      "loss": 1.3228,
+      "step": 9057
+    },
+    {
+      "epoch": 8.361873990306947,
+      "grad_norm": 3.80078125,
+      "learning_rate": 3.9639759839893267e-05,
+      "loss": 1.3102,
+      "step": 9058
+    },
+    {
+      "epoch": 8.362797138241403,
+      "grad_norm": 3.228515625,
+      "learning_rate": 3.963308872581721e-05,
+      "loss": 1.3831,
+      "step": 9059
+    },
+    {
+      "epoch": 8.363720286175859,
+      "grad_norm": 3.04296875,
+      "learning_rate": 3.9626417611741164e-05,
+      "loss": 1.3894,
+      "step": 9060
+    },
+    {
+      "epoch": 8.364643434110317,
+      "grad_norm": 3.3046875,
+      "learning_rate": 3.961974649766511e-05,
+      "loss": 1.3462,
+      "step": 9061
+    },
+    {
+      "epoch": 8.365566582044773,
+      "grad_norm": 3.064453125,
+      "learning_rate": 3.961307538358906e-05,
+      "loss": 1.3982,
+      "step": 9062
+    },
+    {
+      "epoch": 8.366489729979229,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.960640426951301e-05,
+      "loss": 1.2319,
+      "step": 9063
+    },
+    {
+      "epoch": 8.367412877913686,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.959973315543696e-05,
+      "loss": 1.0468,
+      "step": 9064
+    },
+    {
+      "epoch": 8.368336025848143,
+      "grad_norm": 3.572265625,
+      "learning_rate": 3.9593062041360905e-05,
+      "loss": 1.2946,
+      "step": 9065
+    },
+    {
+      "epoch": 8.369259173782599,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.9586390927284864e-05,
+      "loss": 1.0515,
+      "step": 9066
+    },
+    {
+      "epoch": 8.370182321717055,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.957971981320881e-05,
+      "loss": 1.3855,
+      "step": 9067
+    },
+    {
+      "epoch": 8.371105469651512,
+      "grad_norm": 3.94140625,
+      "learning_rate": 3.957304869913276e-05,
+      "loss": 0.9866,
+      "step": 9068
+    },
+    {
+      "epoch": 8.372028617585968,
+      "grad_norm": 2.783203125,
+      "learning_rate": 3.956637758505671e-05,
+      "loss": 1.3752,
+      "step": 9069
+    },
+    {
+      "epoch": 8.372951765520424,
+      "grad_norm": 3.7578125,
+      "learning_rate": 3.955970647098066e-05,
+      "loss": 1.3108,
+      "step": 9070
+    },
+    {
+      "epoch": 8.37387491345488,
+      "grad_norm": 3.5390625,
+      "learning_rate": 3.9553035356904605e-05,
+      "loss": 1.3031,
+      "step": 9071
+    },
+    {
+      "epoch": 8.374798061389338,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.954636424282856e-05,
+      "loss": 1.3147,
+      "step": 9072
+    },
+    {
+      "epoch": 8.375721209323794,
+      "grad_norm": 3.330078125,
+      "learning_rate": 3.95396931287525e-05,
+      "loss": 1.3157,
+      "step": 9073
+    },
+    {
+      "epoch": 8.37664435725825,
+      "grad_norm": 3.0859375,
+      "learning_rate": 3.9533022014676454e-05,
+      "loss": 1.3594,
+      "step": 9074
+    },
+    {
+      "epoch": 8.377567505192706,
+      "grad_norm": 3.107421875,
+      "learning_rate": 3.95263509006004e-05,
+      "loss": 1.1691,
+      "step": 9075
+    },
+    {
+      "epoch": 8.378490653127164,
+      "grad_norm": 2.958984375,
+      "learning_rate": 3.951967978652435e-05,
+      "loss": 1.1595,
+      "step": 9076
+    },
+    {
+      "epoch": 8.37941380106162,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.95130086724483e-05,
+      "loss": 1.2708,
+      "step": 9077
+    },
+    {
+      "epoch": 8.380336948996076,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.950633755837225e-05,
+      "loss": 0.9403,
+      "step": 9078
+    },
+    {
+      "epoch": 8.381260096930534,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.94996664442962e-05,
+      "loss": 1.2444,
+      "step": 9079
+    },
+    {
+      "epoch": 8.38218324486499,
+      "grad_norm": 3.271484375,
+      "learning_rate": 3.949299533022015e-05,
+      "loss": 1.2353,
+      "step": 9080
+    },
+    {
+      "epoch": 8.383106392799446,
+      "grad_norm": 3.310546875,
+      "learning_rate": 3.94863242161441e-05,
+      "loss": 1.3027,
+      "step": 9081
+    },
+    {
+      "epoch": 8.384029540733902,
+      "grad_norm": 3.353515625,
+      "learning_rate": 3.9479653102068045e-05,
+      "loss": 1.3379,
+      "step": 9082
+    },
+    {
+      "epoch": 8.38495268866836,
+      "grad_norm": 3.400390625,
+      "learning_rate": 3.9472981987992e-05,
+      "loss": 1.3251,
+      "step": 9083
+    },
+    {
+      "epoch": 8.385875836602816,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.946631087391594e-05,
+      "loss": 1.3818,
+      "step": 9084
+    },
+    {
+      "epoch": 8.386798984537272,
+      "grad_norm": 3.16015625,
+      "learning_rate": 3.9459639759839895e-05,
+      "loss": 0.9318,
+      "step": 9085
+    },
+    {
+      "epoch": 8.387722132471728,
+      "grad_norm": 3.638671875,
+      "learning_rate": 3.945296864576384e-05,
+      "loss": 1.3708,
+      "step": 9086
+    },
+    {
+      "epoch": 8.388645280406186,
+      "grad_norm": 2.994140625,
+      "learning_rate": 3.944629753168779e-05,
+      "loss": 1.3364,
+      "step": 9087
+    },
+    {
+      "epoch": 8.389568428340642,
+      "grad_norm": 3.5703125,
+      "learning_rate": 3.9439626417611745e-05,
+      "loss": 1.1826,
+      "step": 9088
+    },
+    {
+      "epoch": 8.390491576275098,
+      "grad_norm": 3.205078125,
+      "learning_rate": 3.943295530353569e-05,
+      "loss": 1.7507,
+      "step": 9089
+    },
+    {
+      "epoch": 8.391414724209554,
+      "grad_norm": 3.7109375,
+      "learning_rate": 3.942628418945964e-05,
+      "loss": 1.4443,
+      "step": 9090
+    },
+    {
+      "epoch": 8.392337872144012,
+      "grad_norm": 4.0234375,
+      "learning_rate": 3.9419613075383595e-05,
+      "loss": 1.4243,
+      "step": 9091
+    },
+    {
+      "epoch": 8.393261020078468,
+      "grad_norm": 3.095703125,
+      "learning_rate": 3.941294196130754e-05,
+      "loss": 1.4661,
+      "step": 9092
+    },
+    {
+      "epoch": 8.394184168012924,
+      "grad_norm": 3.583984375,
+      "learning_rate": 3.940627084723149e-05,
+      "loss": 1.4443,
+      "step": 9093
+    },
+    {
+      "epoch": 8.395107315947381,
+      "grad_norm": 3.451171875,
+      "learning_rate": 3.939959973315544e-05,
+      "loss": 1.2129,
+      "step": 9094
+    },
+    {
+      "epoch": 8.396030463881837,
+      "grad_norm": 3.55078125,
+      "learning_rate": 3.939292861907939e-05,
+      "loss": 1.5637,
+      "step": 9095
+    },
+    {
+      "epoch": 8.396953611816294,
+      "grad_norm": 3.4609375,
+      "learning_rate": 3.9386257505003335e-05,
+      "loss": 1.4736,
+      "step": 9096
+    },
+    {
+      "epoch": 8.39787675975075,
+      "grad_norm": 3.154296875,
+      "learning_rate": 3.937958639092729e-05,
+      "loss": 1.2512,
+      "step": 9097
+    },
+    {
+      "epoch": 8.398799907685207,
+      "grad_norm": 3.314453125,
+      "learning_rate": 3.937291527685123e-05,
+      "loss": 1.2485,
+      "step": 9098
+    },
+    {
+      "epoch": 8.399723055619663,
+      "grad_norm": 4.0234375,
+      "learning_rate": 3.9366244162775185e-05,
+      "loss": 1.4376,
+      "step": 9099
+    },
+    {
+      "epoch": 8.40064620355412,
+      "grad_norm": 3.259765625,
+      "learning_rate": 3.935957304869913e-05,
+      "loss": 1.4407,
+      "step": 9100
+    },
+    {
+      "epoch": 8.401569351488575,
+      "grad_norm": 3.39453125,
+      "learning_rate": 3.935290193462308e-05,
+      "loss": 1.0023,
+      "step": 9101
+    },
+    {
+      "epoch": 8.402492499423033,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.934623082054703e-05,
+      "loss": 1.2592,
+      "step": 9102
+    },
+    {
+      "epoch": 8.40341564735749,
+      "grad_norm": 3.896484375,
+      "learning_rate": 3.933955970647098e-05,
+      "loss": 1.4834,
+      "step": 9103
+    },
+    {
+      "epoch": 8.404338795291945,
+      "grad_norm": 4.04296875,
+      "learning_rate": 3.933288859239493e-05,
+      "loss": 1.1541,
+      "step": 9104
+    },
+    {
+      "epoch": 8.405261943226401,
+      "grad_norm": 3.875,
+      "learning_rate": 3.9326217478318885e-05,
+      "loss": 1.2399,
+      "step": 9105
+    },
+    {
+      "epoch": 8.406185091160859,
+      "grad_norm": 3.67578125,
+      "learning_rate": 3.931954636424283e-05,
+      "loss": 1.1388,
+      "step": 9106
+    },
+    {
+      "epoch": 8.407108239095315,
+      "grad_norm": 3.759765625,
+      "learning_rate": 3.931287525016678e-05,
+      "loss": 1.3213,
+      "step": 9107
+    },
+    {
+      "epoch": 8.408031387029771,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.930620413609073e-05,
+      "loss": 1.5144,
+      "step": 9108
+    },
+    {
+      "epoch": 8.408954534964227,
+      "grad_norm": 3.8515625,
+      "learning_rate": 3.929953302201468e-05,
+      "loss": 1.2532,
+      "step": 9109
+    },
+    {
+      "epoch": 8.409877682898685,
+      "grad_norm": 4.1015625,
+      "learning_rate": 3.9292861907938625e-05,
+      "loss": 1.1134,
+      "step": 9110
+    },
+    {
+      "epoch": 8.410800830833141,
+      "grad_norm": 3.37890625,
+      "learning_rate": 3.928619079386258e-05,
+      "loss": 1.4216,
+      "step": 9111
+    },
+    {
+      "epoch": 8.411723978767597,
+      "grad_norm": 3.2578125,
+      "learning_rate": 3.927951967978652e-05,
+      "loss": 1.4441,
+      "step": 9112
+    },
+    {
+      "epoch": 8.412647126702055,
+      "grad_norm": 3.583984375,
+      "learning_rate": 3.9272848565710475e-05,
+      "loss": 1.1652,
+      "step": 9113
+    },
+    {
+      "epoch": 8.41357027463651,
+      "grad_norm": 3.939453125,
+      "learning_rate": 3.926617745163442e-05,
+      "loss": 1.4365,
+      "step": 9114
+    },
+    {
+      "epoch": 8.414493422570967,
+      "grad_norm": 3.099609375,
+      "learning_rate": 3.925950633755837e-05,
+      "loss": 1.1127,
+      "step": 9115
+    },
+    {
+      "epoch": 8.415416570505423,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.9252835223482325e-05,
+      "loss": 1.0809,
+      "step": 9116
+    },
+    {
+      "epoch": 8.41633971843988,
+      "grad_norm": 3.283203125,
+      "learning_rate": 3.924616410940628e-05,
+      "loss": 1.4507,
+      "step": 9117
+    },
+    {
+      "epoch": 8.417262866374337,
+      "grad_norm": 3.283203125,
+      "learning_rate": 3.923949299533022e-05,
+      "loss": 0.9915,
+      "step": 9118
+    },
+    {
+      "epoch": 8.418186014308793,
+      "grad_norm": 3.294921875,
+      "learning_rate": 3.9232821881254175e-05,
+      "loss": 1.5525,
+      "step": 9119
+    },
+    {
+      "epoch": 8.419109162243249,
+      "grad_norm": 3.6640625,
+      "learning_rate": 3.922615076717812e-05,
+      "loss": 1.3149,
+      "step": 9120
+    },
+    {
+      "epoch": 8.420032310177707,
+      "grad_norm": 2.81640625,
+      "learning_rate": 3.921947965310207e-05,
+      "loss": 1.3452,
+      "step": 9121
+    },
+    {
+      "epoch": 8.420955458112163,
+      "grad_norm": 3.017578125,
+      "learning_rate": 3.921280853902602e-05,
+      "loss": 1.2657,
+      "step": 9122
+    },
+    {
+      "epoch": 8.421878606046619,
+      "grad_norm": 3.232421875,
+      "learning_rate": 3.920613742494997e-05,
+      "loss": 1.3574,
+      "step": 9123
+    },
+    {
+      "epoch": 8.422801753981075,
+      "grad_norm": 3.337890625,
+      "learning_rate": 3.9199466310873916e-05,
+      "loss": 1.3308,
+      "step": 9124
+    },
+    {
+      "epoch": 8.423724901915532,
+      "grad_norm": 3.037109375,
+      "learning_rate": 3.919279519679787e-05,
+      "loss": 1.6143,
+      "step": 9125
+    },
+    {
+      "epoch": 8.424648049849988,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.918612408272181e-05,
+      "loss": 1.291,
+      "step": 9126
+    },
+    {
+      "epoch": 8.425571197784445,
+      "grad_norm": 3.228515625,
+      "learning_rate": 3.9179452968645766e-05,
+      "loss": 1.2985,
+      "step": 9127
+    },
+    {
+      "epoch": 8.4264943457189,
+      "grad_norm": 3.2890625,
+      "learning_rate": 3.917278185456971e-05,
+      "loss": 1.3694,
+      "step": 9128
+    },
+    {
+      "epoch": 8.427417493653358,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.916611074049367e-05,
+      "loss": 1.3625,
+      "step": 9129
+    },
+    {
+      "epoch": 8.428340641587814,
+      "grad_norm": 3.244140625,
+      "learning_rate": 3.9159439626417615e-05,
+      "loss": 1.5969,
+      "step": 9130
+    },
+    {
+      "epoch": 8.42926378952227,
+      "grad_norm": 3.44921875,
+      "learning_rate": 3.915276851234157e-05,
+      "loss": 1.2096,
+      "step": 9131
+    },
+    {
+      "epoch": 8.430186937456728,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.914609739826551e-05,
+      "loss": 1.3044,
+      "step": 9132
+    },
+    {
+      "epoch": 8.431110085391184,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.9139426284189465e-05,
+      "loss": 1.4084,
+      "step": 9133
+    },
+    {
+      "epoch": 8.43203323332564,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.913275517011341e-05,
+      "loss": 1.1862,
+      "step": 9134
+    },
+    {
+      "epoch": 8.432956381260096,
+      "grad_norm": 3.619140625,
+      "learning_rate": 3.912608405603736e-05,
+      "loss": 1.4128,
+      "step": 9135
+    },
+    {
+      "epoch": 8.433879529194554,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.911941294196131e-05,
+      "loss": 0.9605,
+      "step": 9136
+    },
+    {
+      "epoch": 8.43480267712901,
+      "grad_norm": 3.97265625,
+      "learning_rate": 3.911274182788526e-05,
+      "loss": 1.5117,
+      "step": 9137
+    },
+    {
+      "epoch": 8.435725825063466,
+      "grad_norm": 3.447265625,
+      "learning_rate": 3.9106070713809206e-05,
+      "loss": 1.4009,
+      "step": 9138
+    },
+    {
+      "epoch": 8.436648972997922,
+      "grad_norm": 3.330078125,
+      "learning_rate": 3.909939959973316e-05,
+      "loss": 1.2872,
+      "step": 9139
+    },
+    {
+      "epoch": 8.43757212093238,
+      "grad_norm": 3.43359375,
+      "learning_rate": 3.9092728485657104e-05,
+      "loss": 1.4851,
+      "step": 9140
+    },
+    {
+      "epoch": 8.438495268866836,
+      "grad_norm": 3.634765625,
+      "learning_rate": 3.9086057371581056e-05,
+      "loss": 1.0912,
+      "step": 9141
+    },
+    {
+      "epoch": 8.439418416801292,
+      "grad_norm": 3.84765625,
+      "learning_rate": 3.907938625750501e-05,
+      "loss": 1.4172,
+      "step": 9142
+    },
+    {
+      "epoch": 8.44034156473575,
+      "grad_norm": 3.189453125,
+      "learning_rate": 3.9072715143428953e-05,
+      "loss": 1.2152,
+      "step": 9143
+    },
+    {
+      "epoch": 8.441264712670206,
+      "grad_norm": 3.435546875,
+      "learning_rate": 3.9066044029352906e-05,
+      "loss": 1.5217,
+      "step": 9144
+    },
+    {
+      "epoch": 8.442187860604662,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.905937291527685e-05,
+      "loss": 1.2991,
+      "step": 9145
+    },
+    {
+      "epoch": 8.443111008539118,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.90527018012008e-05,
+      "loss": 1.3776,
+      "step": 9146
+    },
+    {
+      "epoch": 8.444034156473576,
+      "grad_norm": 3.08203125,
+      "learning_rate": 3.904603068712475e-05,
+      "loss": 1.2096,
+      "step": 9147
+    },
+    {
+      "epoch": 8.444957304408032,
+      "grad_norm": 3.392578125,
+      "learning_rate": 3.90393595730487e-05,
+      "loss": 1.3794,
+      "step": 9148
+    },
+    {
+      "epoch": 8.445880452342488,
+      "grad_norm": 3.6640625,
+      "learning_rate": 3.9032688458972646e-05,
+      "loss": 1.269,
+      "step": 9149
+    },
+    {
+      "epoch": 8.446803600276944,
+      "grad_norm": 3.212890625,
+      "learning_rate": 3.90260173448966e-05,
+      "loss": 1.2168,
+      "step": 9150
+    },
+    {
+      "epoch": 8.447726748211402,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.9019346230820544e-05,
+      "loss": 1.3765,
+      "step": 9151
+    },
+    {
+      "epoch": 8.448649896145858,
+      "grad_norm": 3.47265625,
+      "learning_rate": 3.9012675116744496e-05,
+      "loss": 1.2778,
+      "step": 9152
+    },
+    {
+      "epoch": 8.449573044080314,
+      "grad_norm": 3.04296875,
+      "learning_rate": 3.900600400266844e-05,
+      "loss": 1.3647,
+      "step": 9153
+    },
+    {
+      "epoch": 8.45049619201477,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.89993328885924e-05,
+      "loss": 1.1289,
+      "step": 9154
+    },
+    {
+      "epoch": 8.451419339949227,
+      "grad_norm": 3.443359375,
+      "learning_rate": 3.8992661774516346e-05,
+      "loss": 1.4512,
+      "step": 9155
+    },
+    {
+      "epoch": 8.452342487883683,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.89859906604403e-05,
+      "loss": 1.3516,
+      "step": 9156
+    },
+    {
+      "epoch": 8.45326563581814,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.8979319546364244e-05,
+      "loss": 1.4646,
+      "step": 9157
+    },
+    {
+      "epoch": 8.454188783752596,
+      "grad_norm": 2.93359375,
+      "learning_rate": 3.8972648432288196e-05,
+      "loss": 1.5312,
+      "step": 9158
+    },
+    {
+      "epoch": 8.455111931687053,
+      "grad_norm": 3.580078125,
+      "learning_rate": 3.896597731821214e-05,
+      "loss": 1.3457,
+      "step": 9159
+    },
+    {
+      "epoch": 8.45603507962151,
+      "grad_norm": 3.642578125,
+      "learning_rate": 3.8959306204136094e-05,
+      "loss": 1.4331,
+      "step": 9160
+    },
+    {
+      "epoch": 8.456958227555965,
+      "grad_norm": 2.9921875,
+      "learning_rate": 3.895263509006004e-05,
+      "loss": 1.4165,
+      "step": 9161
+    },
+    {
+      "epoch": 8.457881375490423,
+      "grad_norm": 3.814453125,
+      "learning_rate": 3.894596397598399e-05,
+      "loss": 0.9016,
+      "step": 9162
+    },
+    {
+      "epoch": 8.45880452342488,
+      "grad_norm": 3.013671875,
+      "learning_rate": 3.8939292861907937e-05,
+      "loss": 1.469,
+      "step": 9163
+    },
+    {
+      "epoch": 8.459727671359335,
+      "grad_norm": 3.66015625,
+      "learning_rate": 3.893262174783189e-05,
+      "loss": 1.3149,
+      "step": 9164
+    },
+    {
+      "epoch": 8.460650819293791,
+      "grad_norm": 4.14453125,
+      "learning_rate": 3.8925950633755834e-05,
+      "loss": 1.2118,
+      "step": 9165
+    },
+    {
+      "epoch": 8.461573967228249,
+      "grad_norm": 3.580078125,
+      "learning_rate": 3.891927951967979e-05,
+      "loss": 1.0162,
+      "step": 9166
+    },
+    {
+      "epoch": 8.462497115162705,
+      "grad_norm": 3.00390625,
+      "learning_rate": 3.891260840560374e-05,
+      "loss": 1.4434,
+      "step": 9167
+    },
+    {
+      "epoch": 8.463420263097161,
+      "grad_norm": 3.42578125,
+      "learning_rate": 3.890593729152769e-05,
+      "loss": 1.4702,
+      "step": 9168
+    },
+    {
+      "epoch": 8.464343411031617,
+      "grad_norm": 3.115234375,
+      "learning_rate": 3.8899266177451636e-05,
+      "loss": 1.2942,
+      "step": 9169
+    },
+    {
+      "epoch": 8.465266558966075,
+      "grad_norm": 3.19921875,
+      "learning_rate": 3.889259506337559e-05,
+      "loss": 1.323,
+      "step": 9170
+    },
+    {
+      "epoch": 8.466189706900531,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.8885923949299534e-05,
+      "loss": 1.3403,
+      "step": 9171
+    },
+    {
+      "epoch": 8.467112854834987,
+      "grad_norm": 3.5703125,
+      "learning_rate": 3.8879252835223486e-05,
+      "loss": 1.0384,
+      "step": 9172
+    },
+    {
+      "epoch": 8.468036002769443,
+      "grad_norm": 3.189453125,
+      "learning_rate": 3.887258172114743e-05,
+      "loss": 1.5608,
+      "step": 9173
+    },
+    {
+      "epoch": 8.4689591507039,
+      "grad_norm": 3.34765625,
+      "learning_rate": 3.8865910607071384e-05,
+      "loss": 1.1917,
+      "step": 9174
+    },
+    {
+      "epoch": 8.469882298638357,
+      "grad_norm": 3.775390625,
+      "learning_rate": 3.885923949299533e-05,
+      "loss": 1.4873,
+      "step": 9175
+    },
+    {
+      "epoch": 8.470805446572813,
+      "grad_norm": 3.52734375,
+      "learning_rate": 3.885256837891928e-05,
+      "loss": 1.1941,
+      "step": 9176
+    },
+    {
+      "epoch": 8.471728594507269,
+      "grad_norm": 3.005859375,
+      "learning_rate": 3.884589726484323e-05,
+      "loss": 1.366,
+      "step": 9177
+    },
+    {
+      "epoch": 8.472651742441727,
+      "grad_norm": 3.15234375,
+      "learning_rate": 3.883922615076718e-05,
+      "loss": 1.0477,
+      "step": 9178
+    },
+    {
+      "epoch": 8.473574890376183,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.883255503669113e-05,
+      "loss": 1.3,
+      "step": 9179
+    },
+    {
+      "epoch": 8.474498038310639,
+      "grad_norm": 3.52734375,
+      "learning_rate": 3.8825883922615083e-05,
+      "loss": 1.2103,
+      "step": 9180
+    },
+    {
+      "epoch": 8.475421186245097,
+      "grad_norm": 3.13671875,
+      "learning_rate": 3.881921280853903e-05,
+      "loss": 1.4224,
+      "step": 9181
+    },
+    {
+      "epoch": 8.476344334179553,
+      "grad_norm": 3.283203125,
+      "learning_rate": 3.881254169446298e-05,
+      "loss": 1.3733,
+      "step": 9182
+    },
+    {
+      "epoch": 8.477267482114009,
+      "grad_norm": 3.0546875,
+      "learning_rate": 3.8805870580386927e-05,
+      "loss": 1.3147,
+      "step": 9183
+    },
+    {
+      "epoch": 8.478190630048465,
+      "grad_norm": 2.837890625,
+      "learning_rate": 3.879919946631088e-05,
+      "loss": 1.1392,
+      "step": 9184
+    },
+    {
+      "epoch": 8.479113777982922,
+      "grad_norm": 3.2578125,
+      "learning_rate": 3.8792528352234824e-05,
+      "loss": 1.303,
+      "step": 9185
+    },
+    {
+      "epoch": 8.480036925917378,
+      "grad_norm": 3.80859375,
+      "learning_rate": 3.8785857238158776e-05,
+      "loss": 1.3394,
+      "step": 9186
+    },
+    {
+      "epoch": 8.480960073851834,
+      "grad_norm": 3.52734375,
+      "learning_rate": 3.877918612408272e-05,
+      "loss": 1.2963,
+      "step": 9187
+    },
+    {
+      "epoch": 8.48188322178629,
+      "grad_norm": 3.35546875,
+      "learning_rate": 3.8772515010006674e-05,
+      "loss": 1.0581,
+      "step": 9188
+    },
+    {
+      "epoch": 8.482806369720748,
+      "grad_norm": 3.236328125,
+      "learning_rate": 3.876584389593062e-05,
+      "loss": 0.7809,
+      "step": 9189
+    },
+    {
+      "epoch": 8.483729517655204,
+      "grad_norm": 4.05859375,
+      "learning_rate": 3.875917278185457e-05,
+      "loss": 1.6128,
+      "step": 9190
+    },
+    {
+      "epoch": 8.48465266558966,
+      "grad_norm": 3.494140625,
+      "learning_rate": 3.8752501667778524e-05,
+      "loss": 1.4497,
+      "step": 9191
+    },
+    {
+      "epoch": 8.485575813524116,
+      "grad_norm": 3.12890625,
+      "learning_rate": 3.874583055370247e-05,
+      "loss": 1.2421,
+      "step": 9192
+    },
+    {
+      "epoch": 8.486498961458574,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.873915943962642e-05,
+      "loss": 1.6572,
+      "step": 9193
+    },
+    {
+      "epoch": 8.48742210939303,
+      "grad_norm": 3.447265625,
+      "learning_rate": 3.873248832555037e-05,
+      "loss": 1.4023,
+      "step": 9194
+    },
+    {
+      "epoch": 8.488345257327486,
+      "grad_norm": 3.189453125,
+      "learning_rate": 3.872581721147432e-05,
+      "loss": 1.4153,
+      "step": 9195
+    },
+    {
+      "epoch": 8.489268405261944,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.8719146097398265e-05,
+      "loss": 1.1163,
+      "step": 9196
+    },
+    {
+      "epoch": 8.4901915531964,
+      "grad_norm": 3.2265625,
+      "learning_rate": 3.871247498332222e-05,
+      "loss": 1.2793,
+      "step": 9197
+    },
+    {
+      "epoch": 8.491114701130856,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.870580386924616e-05,
+      "loss": 1.3059,
+      "step": 9198
+    },
+    {
+      "epoch": 8.492037849065312,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.8699132755170114e-05,
+      "loss": 1.1687,
+      "step": 9199
+    },
+    {
+      "epoch": 8.49296099699977,
+      "grad_norm": 2.978515625,
+      "learning_rate": 3.8692461641094067e-05,
+      "loss": 1.3441,
+      "step": 9200
+    },
+    {
+      "epoch": 8.493884144934226,
+      "grad_norm": 3.544921875,
+      "learning_rate": 3.868579052701801e-05,
+      "loss": 1.5835,
+      "step": 9201
+    },
+    {
+      "epoch": 8.494807292868682,
+      "grad_norm": 3.751953125,
+      "learning_rate": 3.8679119412941964e-05,
+      "loss": 1.4885,
+      "step": 9202
+    },
+    {
+      "epoch": 8.495730440803138,
+      "grad_norm": 3.052734375,
+      "learning_rate": 3.867244829886591e-05,
+      "loss": 1.3297,
+      "step": 9203
+    },
+    {
+      "epoch": 8.496653588737596,
+      "grad_norm": 3.208984375,
+      "learning_rate": 3.866577718478986e-05,
+      "loss": 1.436,
+      "step": 9204
+    },
+    {
+      "epoch": 8.497576736672052,
+      "grad_norm": 3.875,
+      "learning_rate": 3.8659106070713814e-05,
+      "loss": 1.078,
+      "step": 9205
+    },
+    {
+      "epoch": 8.498499884606508,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.865243495663776e-05,
+      "loss": 1.2864,
+      "step": 9206
+    },
+    {
+      "epoch": 8.499423032540964,
+      "grad_norm": 3.708984375,
+      "learning_rate": 3.864576384256171e-05,
+      "loss": 1.3767,
+      "step": 9207
+    },
+    {
+      "epoch": 8.500346180475422,
+      "grad_norm": 3.27734375,
+      "learning_rate": 3.863909272848566e-05,
+      "loss": 1.4338,
+      "step": 9208
+    },
+    {
+      "epoch": 8.501269328409878,
+      "grad_norm": 3.51953125,
+      "learning_rate": 3.863242161440961e-05,
+      "loss": 1.5334,
+      "step": 9209
+    },
+    {
+      "epoch": 8.502192476344334,
+      "grad_norm": 3.955078125,
+      "learning_rate": 3.8625750500333555e-05,
+      "loss": 1.4814,
+      "step": 9210
+    },
+    {
+      "epoch": 8.503115624278792,
+      "grad_norm": 3.658203125,
+      "learning_rate": 3.861907938625751e-05,
+      "loss": 1.416,
+      "step": 9211
+    },
+    {
+      "epoch": 8.504038772213248,
+      "grad_norm": 3.23828125,
+      "learning_rate": 3.861240827218145e-05,
+      "loss": 1.4214,
+      "step": 9212
+    },
+    {
+      "epoch": 8.504961920147704,
+      "grad_norm": 3.115234375,
+      "learning_rate": 3.8605737158105405e-05,
+      "loss": 1.1526,
+      "step": 9213
+    },
+    {
+      "epoch": 8.50588506808216,
+      "grad_norm": 3.099609375,
+      "learning_rate": 3.859906604402935e-05,
+      "loss": 1.2888,
+      "step": 9214
+    },
+    {
+      "epoch": 8.506808216016617,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.85923949299533e-05,
+      "loss": 0.985,
+      "step": 9215
+    },
+    {
+      "epoch": 8.507731363951073,
+      "grad_norm": 2.939453125,
+      "learning_rate": 3.8585723815877254e-05,
+      "loss": 1.3408,
+      "step": 9216
+    },
+    {
+      "epoch": 8.50865451188553,
+      "grad_norm": 3.435546875,
+      "learning_rate": 3.857905270180121e-05,
+      "loss": 1.16,
+      "step": 9217
+    },
+    {
+      "epoch": 8.509577659819985,
+      "grad_norm": 3.19140625,
+      "learning_rate": 3.857238158772515e-05,
+      "loss": 1.3984,
+      "step": 9218
+    },
+    {
+      "epoch": 8.510500807754443,
+      "grad_norm": 3.384765625,
+      "learning_rate": 3.8565710473649104e-05,
+      "loss": 1.0582,
+      "step": 9219
+    },
+    {
+      "epoch": 8.5114239556889,
+      "grad_norm": 3.689453125,
+      "learning_rate": 3.855903935957305e-05,
+      "loss": 1.3821,
+      "step": 9220
+    },
+    {
+      "epoch": 8.512347103623355,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.8552368245497e-05,
+      "loss": 1.0586,
+      "step": 9221
+    },
+    {
+      "epoch": 8.513270251557811,
+      "grad_norm": 3.751953125,
+      "learning_rate": 3.854569713142095e-05,
+      "loss": 1.509,
+      "step": 9222
+    },
+    {
+      "epoch": 8.51419339949227,
+      "grad_norm": 3.53515625,
+      "learning_rate": 3.85390260173449e-05,
+      "loss": 1.4648,
+      "step": 9223
+    },
+    {
+      "epoch": 8.515116547426725,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.8532354903268845e-05,
+      "loss": 1.3074,
+      "step": 9224
+    },
+    {
+      "epoch": 8.516039695361181,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.85256837891928e-05,
+      "loss": 1.3325,
+      "step": 9225
+    },
+    {
+      "epoch": 8.516962843295637,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.851901267511674e-05,
+      "loss": 1.3448,
+      "step": 9226
+    },
+    {
+      "epoch": 8.517885991230095,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.8512341561040695e-05,
+      "loss": 1.332,
+      "step": 9227
+    },
+    {
+      "epoch": 8.518809139164551,
+      "grad_norm": 3.466796875,
+      "learning_rate": 3.850567044696464e-05,
+      "loss": 1.4265,
+      "step": 9228
+    },
+    {
+      "epoch": 8.519732287099007,
+      "grad_norm": 3.189453125,
+      "learning_rate": 3.84989993328886e-05,
+      "loss": 1.1936,
+      "step": 9229
+    },
+    {
+      "epoch": 8.520655435033465,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.8492328218812545e-05,
+      "loss": 1.2571,
+      "step": 9230
+    },
+    {
+      "epoch": 8.521578582967921,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.84856571047365e-05,
+      "loss": 1.3083,
+      "step": 9231
+    },
+    {
+      "epoch": 8.522501730902377,
+      "grad_norm": 3.6171875,
+      "learning_rate": 3.847898599066044e-05,
+      "loss": 1.2717,
+      "step": 9232
+    },
+    {
+      "epoch": 8.523424878836833,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.8472314876584395e-05,
+      "loss": 1.2694,
+      "step": 9233
+    },
+    {
+      "epoch": 8.52434802677129,
+      "grad_norm": 3.748046875,
+      "learning_rate": 3.846564376250834e-05,
+      "loss": 1.0554,
+      "step": 9234
+    },
+    {
+      "epoch": 8.525271174705747,
+      "grad_norm": 3.0546875,
+      "learning_rate": 3.845897264843229e-05,
+      "loss": 1.4858,
+      "step": 9235
+    },
+    {
+      "epoch": 8.526194322640203,
+      "grad_norm": 4.0078125,
+      "learning_rate": 3.845230153435624e-05,
+      "loss": 1.1456,
+      "step": 9236
+    },
+    {
+      "epoch": 8.527117470574659,
+      "grad_norm": 3.65234375,
+      "learning_rate": 3.844563042028019e-05,
+      "loss": 1.2915,
+      "step": 9237
+    },
+    {
+      "epoch": 8.528040618509117,
+      "grad_norm": 4.23828125,
+      "learning_rate": 3.8438959306204135e-05,
+      "loss": 1.3757,
+      "step": 9238
+    },
+    {
+      "epoch": 8.528963766443573,
+      "grad_norm": 3.087890625,
+      "learning_rate": 3.843228819212809e-05,
+      "loss": 1.6807,
+      "step": 9239
+    },
+    {
+      "epoch": 8.529886914378029,
+      "grad_norm": 2.986328125,
+      "learning_rate": 3.842561707805203e-05,
+      "loss": 1.3381,
+      "step": 9240
+    },
+    {
+      "epoch": 8.530810062312487,
+      "grad_norm": 3.630859375,
+      "learning_rate": 3.841894596397599e-05,
+      "loss": 1.3375,
+      "step": 9241
+    },
+    {
+      "epoch": 8.531733210246943,
+      "grad_norm": 3.52734375,
+      "learning_rate": 3.841227484989994e-05,
+      "loss": 1.604,
+      "step": 9242
+    },
+    {
+      "epoch": 8.532656358181399,
+      "grad_norm": 3.462890625,
+      "learning_rate": 3.840560373582389e-05,
+      "loss": 1.3132,
+      "step": 9243
+    },
+    {
+      "epoch": 8.533579506115855,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.8398932621747835e-05,
+      "loss": 1.1055,
+      "step": 9244
+    },
+    {
+      "epoch": 8.534502654050312,
+      "grad_norm": 3.509765625,
+      "learning_rate": 3.839226150767179e-05,
+      "loss": 1.2175,
+      "step": 9245
+    },
+    {
+      "epoch": 8.535425801984768,
+      "grad_norm": 3.32421875,
+      "learning_rate": 3.838559039359573e-05,
+      "loss": 1.157,
+      "step": 9246
+    },
+    {
+      "epoch": 8.536348949919224,
+      "grad_norm": 3.673828125,
+      "learning_rate": 3.8378919279519685e-05,
+      "loss": 1.2961,
+      "step": 9247
+    },
+    {
+      "epoch": 8.53727209785368,
+      "grad_norm": 3.427734375,
+      "learning_rate": 3.837224816544363e-05,
+      "loss": 1.3184,
+      "step": 9248
+    },
+    {
+      "epoch": 8.538195245788138,
+      "grad_norm": 3.177734375,
+      "learning_rate": 3.836557705136758e-05,
+      "loss": 1.2954,
+      "step": 9249
+    },
+    {
+      "epoch": 8.539118393722594,
+      "grad_norm": 2.98046875,
+      "learning_rate": 3.835890593729153e-05,
+      "loss": 1.119,
+      "step": 9250
+    },
+    {
+      "epoch": 8.54004154165705,
+      "grad_norm": 2.896484375,
+      "learning_rate": 3.835223482321548e-05,
+      "loss": 1.3992,
+      "step": 9251
+    },
+    {
+      "epoch": 8.540964689591506,
+      "grad_norm": 2.9609375,
+      "learning_rate": 3.8345563709139426e-05,
+      "loss": 1.3992,
+      "step": 9252
+    },
+    {
+      "epoch": 8.541887837525964,
+      "grad_norm": 3.755859375,
+      "learning_rate": 3.833889259506338e-05,
+      "loss": 1.262,
+      "step": 9253
+    },
+    {
+      "epoch": 8.54281098546042,
+      "grad_norm": 3.021484375,
+      "learning_rate": 3.833222148098733e-05,
+      "loss": 1.3445,
+      "step": 9254
+    },
+    {
+      "epoch": 8.543734133394876,
+      "grad_norm": 4.6015625,
+      "learning_rate": 3.8325550366911275e-05,
+      "loss": 1.1956,
+      "step": 9255
+    },
+    {
+      "epoch": 8.544657281329332,
+      "grad_norm": 3.642578125,
+      "learning_rate": 3.831887925283523e-05,
+      "loss": 1.3911,
+      "step": 9256
+    },
+    {
+      "epoch": 8.54558042926379,
+      "grad_norm": 3.60546875,
+      "learning_rate": 3.831220813875917e-05,
+      "loss": 1.3553,
+      "step": 9257
+    },
+    {
+      "epoch": 8.546503577198246,
+      "grad_norm": 3.25390625,
+      "learning_rate": 3.8305537024683125e-05,
+      "loss": 1.2516,
+      "step": 9258
+    },
+    {
+      "epoch": 8.547426725132702,
+      "grad_norm": 3.341796875,
+      "learning_rate": 3.829886591060707e-05,
+      "loss": 1.3376,
+      "step": 9259
+    },
+    {
+      "epoch": 8.54834987306716,
+      "grad_norm": 3.55078125,
+      "learning_rate": 3.829219479653102e-05,
+      "loss": 1.5654,
+      "step": 9260
+    },
+    {
+      "epoch": 8.549273021001616,
+      "grad_norm": 2.978515625,
+      "learning_rate": 3.828552368245497e-05,
+      "loss": 1.3007,
+      "step": 9261
+    },
+    {
+      "epoch": 8.550196168936072,
+      "grad_norm": 3.2109375,
+      "learning_rate": 3.827885256837892e-05,
+      "loss": 1.199,
+      "step": 9262
+    },
+    {
+      "epoch": 8.551119316870528,
+      "grad_norm": 3.74609375,
+      "learning_rate": 3.8272181454302866e-05,
+      "loss": 1.3499,
+      "step": 9263
+    },
+    {
+      "epoch": 8.552042464804986,
+      "grad_norm": 3.458984375,
+      "learning_rate": 3.826551034022682e-05,
+      "loss": 1.5317,
+      "step": 9264
+    },
+    {
+      "epoch": 8.552965612739442,
+      "grad_norm": 3.767578125,
+      "learning_rate": 3.8258839226150764e-05,
+      "loss": 1.4927,
+      "step": 9265
+    },
+    {
+      "epoch": 8.553888760673898,
+      "grad_norm": 3.580078125,
+      "learning_rate": 3.825216811207472e-05,
+      "loss": 1.1855,
+      "step": 9266
+    },
+    {
+      "epoch": 8.554811908608354,
+      "grad_norm": 3.158203125,
+      "learning_rate": 3.824549699799867e-05,
+      "loss": 1.2712,
+      "step": 9267
+    },
+    {
+      "epoch": 8.555735056542812,
+      "grad_norm": 3.15234375,
+      "learning_rate": 3.823882588392262e-05,
+      "loss": 1.4077,
+      "step": 9268
+    },
+    {
+      "epoch": 8.556658204477268,
+      "grad_norm": 3.3671875,
+      "learning_rate": 3.8232154769846566e-05,
+      "loss": 1.5583,
+      "step": 9269
+    },
+    {
+      "epoch": 8.557581352411724,
+      "grad_norm": 3.884765625,
+      "learning_rate": 3.822548365577052e-05,
+      "loss": 1.4268,
+      "step": 9270
+    },
+    {
+      "epoch": 8.55850450034618,
+      "grad_norm": 3.748046875,
+      "learning_rate": 3.821881254169446e-05,
+      "loss": 1.3257,
+      "step": 9271
+    },
+    {
+      "epoch": 8.559427648280638,
+      "grad_norm": 3.396484375,
+      "learning_rate": 3.8212141427618415e-05,
+      "loss": 1.1707,
+      "step": 9272
+    },
+    {
+      "epoch": 8.560350796215094,
+      "grad_norm": 3.62109375,
+      "learning_rate": 3.820547031354236e-05,
+      "loss": 1.3555,
+      "step": 9273
+    },
+    {
+      "epoch": 8.56127394414955,
+      "grad_norm": 3.552734375,
+      "learning_rate": 3.819879919946631e-05,
+      "loss": 1.2645,
+      "step": 9274
+    },
+    {
+      "epoch": 8.562197092084006,
+      "grad_norm": 3.416015625,
+      "learning_rate": 3.819212808539026e-05,
+      "loss": 1.0637,
+      "step": 9275
+    },
+    {
+      "epoch": 8.563120240018463,
+      "grad_norm": 3.42578125,
+      "learning_rate": 3.818545697131421e-05,
+      "loss": 1.4832,
+      "step": 9276
+    },
+    {
+      "epoch": 8.56404338795292,
+      "grad_norm": 3.84765625,
+      "learning_rate": 3.8178785857238156e-05,
+      "loss": 1.3381,
+      "step": 9277
+    },
+    {
+      "epoch": 8.564966535887375,
+      "grad_norm": 3.689453125,
+      "learning_rate": 3.817211474316211e-05,
+      "loss": 1.2162,
+      "step": 9278
+    },
+    {
+      "epoch": 8.565889683821833,
+      "grad_norm": 3.47265625,
+      "learning_rate": 3.816544362908606e-05,
+      "loss": 1.1448,
+      "step": 9279
+    },
+    {
+      "epoch": 8.56681283175629,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.815877251501001e-05,
+      "loss": 1.4067,
+      "step": 9280
+    },
+    {
+      "epoch": 8.567735979690745,
+      "grad_norm": 3.7265625,
+      "learning_rate": 3.815210140093396e-05,
+      "loss": 1.2736,
+      "step": 9281
+    },
+    {
+      "epoch": 8.568659127625201,
+      "grad_norm": 3.337890625,
+      "learning_rate": 3.814543028685791e-05,
+      "loss": 1.3992,
+      "step": 9282
+    },
+    {
+      "epoch": 8.56958227555966,
+      "grad_norm": 3.572265625,
+      "learning_rate": 3.8138759172781856e-05,
+      "loss": 1.439,
+      "step": 9283
+    },
+    {
+      "epoch": 8.570505423494115,
+      "grad_norm": 3.232421875,
+      "learning_rate": 3.813208805870581e-05,
+      "loss": 1.1143,
+      "step": 9284
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 3.220703125,
+      "learning_rate": 3.8125416944629753e-05,
+      "loss": 1.5132,
+      "step": 9285
+    },
+    {
+      "epoch": 8.572351719363027,
+      "grad_norm": 3.2109375,
+      "learning_rate": 3.8118745830553706e-05,
+      "loss": 1.2068,
+      "step": 9286
+    },
+    {
+      "epoch": 8.573274867297485,
+      "grad_norm": 3.365234375,
+      "learning_rate": 3.811207471647765e-05,
+      "loss": 1.4216,
+      "step": 9287
+    },
+    {
+      "epoch": 8.574198015231941,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.81054036024016e-05,
+      "loss": 1.0383,
+      "step": 9288
+    },
+    {
+      "epoch": 8.575121163166397,
+      "grad_norm": 3.986328125,
+      "learning_rate": 3.809873248832555e-05,
+      "loss": 1.4905,
+      "step": 9289
+    },
+    {
+      "epoch": 8.576044311100853,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.80920613742495e-05,
+      "loss": 1.2358,
+      "step": 9290
+    },
+    {
+      "epoch": 8.57696745903531,
+      "grad_norm": 2.978515625,
+      "learning_rate": 3.8085390260173446e-05,
+      "loss": 1.3077,
+      "step": 9291
+    },
+    {
+      "epoch": 8.577890606969767,
+      "grad_norm": 4.2578125,
+      "learning_rate": 3.8078719146097405e-05,
+      "loss": 1.3694,
+      "step": 9292
+    },
+    {
+      "epoch": 8.578813754904223,
+      "grad_norm": 2.962890625,
+      "learning_rate": 3.807204803202135e-05,
+      "loss": 1.323,
+      "step": 9293
+    },
+    {
+      "epoch": 8.579736902838679,
+      "grad_norm": 3.69140625,
+      "learning_rate": 3.80653769179453e-05,
+      "loss": 1.3607,
+      "step": 9294
+    },
+    {
+      "epoch": 8.580660050773137,
+      "grad_norm": 3.40234375,
+      "learning_rate": 3.805870580386925e-05,
+      "loss": 1.283,
+      "step": 9295
+    },
+    {
+      "epoch": 8.581583198707593,
+      "grad_norm": 3.220703125,
+      "learning_rate": 3.80520346897932e-05,
+      "loss": 1.3354,
+      "step": 9296
+    },
+    {
+      "epoch": 8.582506346642049,
+      "grad_norm": 2.857421875,
+      "learning_rate": 3.8045363575717146e-05,
+      "loss": 1.635,
+      "step": 9297
+    },
+    {
+      "epoch": 8.583429494576507,
+      "grad_norm": 4.046875,
+      "learning_rate": 3.80386924616411e-05,
+      "loss": 1.3159,
+      "step": 9298
+    },
+    {
+      "epoch": 8.584352642510963,
+      "grad_norm": 3.076171875,
+      "learning_rate": 3.8032021347565044e-05,
+      "loss": 1.5889,
+      "step": 9299
+    },
+    {
+      "epoch": 8.585275790445419,
+      "grad_norm": 3.865234375,
+      "learning_rate": 3.8025350233488996e-05,
+      "loss": 1.1779,
+      "step": 9300
+    },
+    {
+      "epoch": 8.586198938379875,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.801867911941294e-05,
+      "loss": 1.4429,
+      "step": 9301
+    },
+    {
+      "epoch": 8.587122086314332,
+      "grad_norm": 3.3046875,
+      "learning_rate": 3.8012008005336894e-05,
+      "loss": 1.4668,
+      "step": 9302
+    },
+    {
+      "epoch": 8.588045234248789,
+      "grad_norm": 3.810546875,
+      "learning_rate": 3.800533689126084e-05,
+      "loss": 1.5264,
+      "step": 9303
+    },
+    {
+      "epoch": 8.588968382183245,
+      "grad_norm": 3.767578125,
+      "learning_rate": 3.799866577718479e-05,
+      "loss": 1.21,
+      "step": 9304
+    },
+    {
+      "epoch": 8.5898915301177,
+      "grad_norm": 3.83984375,
+      "learning_rate": 3.799199466310874e-05,
+      "loss": 1.4297,
+      "step": 9305
+    },
+    {
+      "epoch": 8.590814678052158,
+      "grad_norm": 3.375,
+      "learning_rate": 3.798532354903269e-05,
+      "loss": 1.2609,
+      "step": 9306
+    },
+    {
+      "epoch": 8.591737825986614,
+      "grad_norm": 4.17578125,
+      "learning_rate": 3.797865243495664e-05,
+      "loss": 1.1277,
+      "step": 9307
+    },
+    {
+      "epoch": 8.59266097392107,
+      "grad_norm": 3.193359375,
+      "learning_rate": 3.7971981320880586e-05,
+      "loss": 1.5037,
+      "step": 9308
+    },
+    {
+      "epoch": 8.593584121855528,
+      "grad_norm": 3.55078125,
+      "learning_rate": 3.796531020680454e-05,
+      "loss": 1.4597,
+      "step": 9309
+    },
+    {
+      "epoch": 8.594507269789984,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.7958639092728484e-05,
+      "loss": 1.2994,
+      "step": 9310
+    },
+    {
+      "epoch": 8.59543041772444,
+      "grad_norm": 3.205078125,
+      "learning_rate": 3.7951967978652436e-05,
+      "loss": 1.2766,
+      "step": 9311
+    },
+    {
+      "epoch": 8.596353565658896,
+      "grad_norm": 3.1796875,
+      "learning_rate": 3.794529686457638e-05,
+      "loss": 0.9358,
+      "step": 9312
+    },
+    {
+      "epoch": 8.597276713593354,
+      "grad_norm": 3.1484375,
+      "learning_rate": 3.7938625750500334e-05,
+      "loss": 1.3313,
+      "step": 9313
+    },
+    {
+      "epoch": 8.59819986152781,
+      "grad_norm": 3.4453125,
+      "learning_rate": 3.7931954636424286e-05,
+      "loss": 1.2419,
+      "step": 9314
+    },
+    {
+      "epoch": 8.599123009462266,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.792528352234823e-05,
+      "loss": 1.3201,
+      "step": 9315
+    },
+    {
+      "epoch": 8.600046157396722,
+      "grad_norm": 3.189453125,
+      "learning_rate": 3.7918612408272184e-05,
+      "loss": 1.4072,
+      "step": 9316
+    },
+    {
+      "epoch": 8.60096930533118,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.7911941294196136e-05,
+      "loss": 1.5986,
+      "step": 9317
+    },
+    {
+      "epoch": 8.601892453265636,
+      "grad_norm": 3.298828125,
+      "learning_rate": 3.790527018012008e-05,
+      "loss": 1.353,
+      "step": 9318
+    },
+    {
+      "epoch": 8.602815601200092,
+      "grad_norm": 3.330078125,
+      "learning_rate": 3.7898599066044034e-05,
+      "loss": 1.0813,
+      "step": 9319
+    },
+    {
+      "epoch": 8.603738749134548,
+      "grad_norm": 3.62109375,
+      "learning_rate": 3.789192795196798e-05,
+      "loss": 1.3798,
+      "step": 9320
+    },
+    {
+      "epoch": 8.604661897069006,
+      "grad_norm": 3.39453125,
+      "learning_rate": 3.788525683789193e-05,
+      "loss": 1.3569,
+      "step": 9321
+    },
+    {
+      "epoch": 8.605585045003462,
+      "grad_norm": 4.19921875,
+      "learning_rate": 3.787858572381588e-05,
+      "loss": 1.3552,
+      "step": 9322
+    },
+    {
+      "epoch": 8.606508192937918,
+      "grad_norm": 3.607421875,
+      "learning_rate": 3.787191460973983e-05,
+      "loss": 1.3923,
+      "step": 9323
+    },
+    {
+      "epoch": 8.607431340872374,
+      "grad_norm": 3.11328125,
+      "learning_rate": 3.7865243495663774e-05,
+      "loss": 1.3713,
+      "step": 9324
+    },
+    {
+      "epoch": 8.608354488806832,
+      "grad_norm": 3.630859375,
+      "learning_rate": 3.7858572381587727e-05,
+      "loss": 1.2109,
+      "step": 9325
+    },
+    {
+      "epoch": 8.609277636741288,
+      "grad_norm": 3.193359375,
+      "learning_rate": 3.785190126751167e-05,
+      "loss": 1.3875,
+      "step": 9326
+    },
+    {
+      "epoch": 8.610200784675744,
+      "grad_norm": 3.712890625,
+      "learning_rate": 3.7845230153435624e-05,
+      "loss": 1.2499,
+      "step": 9327
+    },
+    {
+      "epoch": 8.611123932610202,
+      "grad_norm": 3.02734375,
+      "learning_rate": 3.783855903935957e-05,
+      "loss": 1.3086,
+      "step": 9328
+    },
+    {
+      "epoch": 8.612047080544658,
+      "grad_norm": 3.224609375,
+      "learning_rate": 3.783188792528353e-05,
+      "loss": 1.5061,
+      "step": 9329
+    },
+    {
+      "epoch": 8.612970228479114,
+      "grad_norm": 3.591796875,
+      "learning_rate": 3.7825216811207474e-05,
+      "loss": 1.0386,
+      "step": 9330
+    },
+    {
+      "epoch": 8.61389337641357,
+      "grad_norm": 3.552734375,
+      "learning_rate": 3.7818545697131426e-05,
+      "loss": 1.3809,
+      "step": 9331
+    },
+    {
+      "epoch": 8.614816524348027,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.781187458305537e-05,
+      "loss": 0.918,
+      "step": 9332
+    },
+    {
+      "epoch": 8.615739672282483,
+      "grad_norm": 3.7578125,
+      "learning_rate": 3.7805203468979324e-05,
+      "loss": 1.5364,
+      "step": 9333
+    },
+    {
+      "epoch": 8.61666282021694,
+      "grad_norm": 3.224609375,
+      "learning_rate": 3.779853235490327e-05,
+      "loss": 1.2769,
+      "step": 9334
+    },
+    {
+      "epoch": 8.617585968151396,
+      "grad_norm": 3.66015625,
+      "learning_rate": 3.779186124082722e-05,
+      "loss": 1.2969,
+      "step": 9335
+    },
+    {
+      "epoch": 8.618509116085853,
+      "grad_norm": 3.193359375,
+      "learning_rate": 3.778519012675117e-05,
+      "loss": 1.4961,
+      "step": 9336
+    },
+    {
+      "epoch": 8.61943226402031,
+      "grad_norm": 3.962890625,
+      "learning_rate": 3.777851901267512e-05,
+      "loss": 1.1136,
+      "step": 9337
+    },
+    {
+      "epoch": 8.620355411954765,
+      "grad_norm": 3.048828125,
+      "learning_rate": 3.7771847898599065e-05,
+      "loss": 1.1562,
+      "step": 9338
+    },
+    {
+      "epoch": 8.621278559889221,
+      "grad_norm": 3.25,
+      "learning_rate": 3.776517678452302e-05,
+      "loss": 1.1862,
+      "step": 9339
+    },
+    {
+      "epoch": 8.62220170782368,
+      "grad_norm": 3.357421875,
+      "learning_rate": 3.775850567044696e-05,
+      "loss": 1.5642,
+      "step": 9340
+    },
+    {
+      "epoch": 8.623124855758135,
+      "grad_norm": 3.611328125,
+      "learning_rate": 3.7751834556370914e-05,
+      "loss": 1.5549,
+      "step": 9341
+    },
+    {
+      "epoch": 8.624048003692591,
+      "grad_norm": 3.064453125,
+      "learning_rate": 3.774516344229487e-05,
+      "loss": 1.2798,
+      "step": 9342
+    },
+    {
+      "epoch": 8.624971151627047,
+      "grad_norm": 3.6953125,
+      "learning_rate": 3.773849232821882e-05,
+      "loss": 1.1514,
+      "step": 9343
+    },
+    {
+      "epoch": 8.625894299561505,
+      "grad_norm": 3.625,
+      "learning_rate": 3.7731821214142764e-05,
+      "loss": 1.2388,
+      "step": 9344
+    },
+    {
+      "epoch": 8.626817447495961,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.7725150100066716e-05,
+      "loss": 1.1975,
+      "step": 9345
+    },
+    {
+      "epoch": 8.627740595430417,
+      "grad_norm": 2.96484375,
+      "learning_rate": 3.771847898599066e-05,
+      "loss": 1.0221,
+      "step": 9346
+    },
+    {
+      "epoch": 8.628663743364875,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.7711807871914614e-05,
+      "loss": 1.2443,
+      "step": 9347
+    },
+    {
+      "epoch": 8.629586891299331,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.770513675783856e-05,
+      "loss": 1.1213,
+      "step": 9348
+    },
+    {
+      "epoch": 8.630510039233787,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.769846564376251e-05,
+      "loss": 1.3352,
+      "step": 9349
+    },
+    {
+      "epoch": 8.631433187168243,
+      "grad_norm": 3.98046875,
+      "learning_rate": 3.769179452968646e-05,
+      "loss": 1.0547,
+      "step": 9350
+    },
+    {
+      "epoch": 8.6323563351027,
+      "grad_norm": 3.748046875,
+      "learning_rate": 3.768512341561041e-05,
+      "loss": 1.1216,
+      "step": 9351
+    },
+    {
+      "epoch": 8.633279483037157,
+      "grad_norm": 3.5078125,
+      "learning_rate": 3.7678452301534355e-05,
+      "loss": 1.208,
+      "step": 9352
+    },
+    {
+      "epoch": 8.634202630971613,
+      "grad_norm": 3.216796875,
+      "learning_rate": 3.767178118745831e-05,
+      "loss": 1.3813,
+      "step": 9353
+    },
+    {
+      "epoch": 8.635125778906069,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.766511007338226e-05,
+      "loss": 1.5796,
+      "step": 9354
+    },
+    {
+      "epoch": 8.636048926840527,
+      "grad_norm": 3.33203125,
+      "learning_rate": 3.765843895930621e-05,
+      "loss": 1.1249,
+      "step": 9355
+    },
+    {
+      "epoch": 8.636972074774983,
+      "grad_norm": 3.259765625,
+      "learning_rate": 3.765176784523016e-05,
+      "loss": 1.2324,
+      "step": 9356
+    },
+    {
+      "epoch": 8.637895222709439,
+      "grad_norm": 3.73046875,
+      "learning_rate": 3.764509673115411e-05,
+      "loss": 1.5974,
+      "step": 9357
+    },
+    {
+      "epoch": 8.638818370643897,
+      "grad_norm": 3.2109375,
+      "learning_rate": 3.7638425617078054e-05,
+      "loss": 1.1969,
+      "step": 9358
+    },
+    {
+      "epoch": 8.639741518578353,
+      "grad_norm": 3.857421875,
+      "learning_rate": 3.763175450300201e-05,
+      "loss": 1.2532,
+      "step": 9359
+    },
+    {
+      "epoch": 8.640664666512809,
+      "grad_norm": 4.24609375,
+      "learning_rate": 3.762508338892595e-05,
+      "loss": 1.4211,
+      "step": 9360
+    },
+    {
+      "epoch": 8.641587814447265,
+      "grad_norm": 3.65234375,
+      "learning_rate": 3.7618412274849904e-05,
+      "loss": 1.29,
+      "step": 9361
+    },
+    {
+      "epoch": 8.642510962381722,
+      "grad_norm": 3.458984375,
+      "learning_rate": 3.761174116077385e-05,
+      "loss": 1.197,
+      "step": 9362
+    },
+    {
+      "epoch": 8.643434110316178,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.76050700466978e-05,
+      "loss": 1.3538,
+      "step": 9363
+    },
+    {
+      "epoch": 8.644357258250635,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.759839893262175e-05,
+      "loss": 1.262,
+      "step": 9364
+    },
+    {
+      "epoch": 8.64528040618509,
+      "grad_norm": 2.94140625,
+      "learning_rate": 3.75917278185457e-05,
+      "loss": 1.1772,
+      "step": 9365
+    },
+    {
+      "epoch": 8.646203554119548,
+      "grad_norm": 3.595703125,
+      "learning_rate": 3.7585056704469645e-05,
+      "loss": 1.2324,
+      "step": 9366
+    },
+    {
+      "epoch": 8.647126702054004,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.75783855903936e-05,
+      "loss": 1.6143,
+      "step": 9367
+    },
+    {
+      "epoch": 8.64804984998846,
+      "grad_norm": 3.6484375,
+      "learning_rate": 3.757171447631755e-05,
+      "loss": 0.9663,
+      "step": 9368
+    },
+    {
+      "epoch": 8.648972997922916,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.7565043362241495e-05,
+      "loss": 1.3824,
+      "step": 9369
+    },
+    {
+      "epoch": 8.649896145857374,
+      "grad_norm": 3.41796875,
+      "learning_rate": 3.755837224816545e-05,
+      "loss": 1.2838,
+      "step": 9370
+    },
+    {
+      "epoch": 8.65081929379183,
+      "grad_norm": 3.560546875,
+      "learning_rate": 3.755170113408939e-05,
+      "loss": 1.3591,
+      "step": 9371
+    },
+    {
+      "epoch": 8.651742441726286,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.7545030020013345e-05,
+      "loss": 1.0935,
+      "step": 9372
+    },
+    {
+      "epoch": 8.652665589660742,
+      "grad_norm": 3.45703125,
+      "learning_rate": 3.753835890593729e-05,
+      "loss": 1.4409,
+      "step": 9373
+    },
+    {
+      "epoch": 8.6535887375952,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.753168779186124e-05,
+      "loss": 1.4353,
+      "step": 9374
+    },
+    {
+      "epoch": 8.654511885529656,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.752501667778519e-05,
+      "loss": 0.8143,
+      "step": 9375
+    },
+    {
+      "epoch": 8.655435033464112,
+      "grad_norm": 3.787109375,
+      "learning_rate": 3.751834556370914e-05,
+      "loss": 1.4583,
+      "step": 9376
+    },
+    {
+      "epoch": 8.65635818139857,
+      "grad_norm": 3.30078125,
+      "learning_rate": 3.7511674449633085e-05,
+      "loss": 1.1759,
+      "step": 9377
+    },
+    {
+      "epoch": 8.657281329333026,
+      "grad_norm": 3.6171875,
+      "learning_rate": 3.750500333555704e-05,
+      "loss": 1.4365,
+      "step": 9378
+    },
+    {
+      "epoch": 8.658204477267482,
+      "grad_norm": 3.412109375,
+      "learning_rate": 3.749833222148099e-05,
+      "loss": 1.3389,
+      "step": 9379
+    },
+    {
+      "epoch": 8.659127625201938,
+      "grad_norm": 3.0390625,
+      "learning_rate": 3.749166110740494e-05,
+      "loss": 1.4146,
+      "step": 9380
+    },
+    {
+      "epoch": 8.660050773136396,
+      "grad_norm": 4.01171875,
+      "learning_rate": 3.748498999332889e-05,
+      "loss": 1.1082,
+      "step": 9381
+    },
+    {
+      "epoch": 8.660973921070852,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.747831887925284e-05,
+      "loss": 1.4785,
+      "step": 9382
+    },
+    {
+      "epoch": 8.661897069005308,
+      "grad_norm": 3.3359375,
+      "learning_rate": 3.7471647765176785e-05,
+      "loss": 1.2388,
+      "step": 9383
+    },
+    {
+      "epoch": 8.662820216939764,
+      "grad_norm": 3.232421875,
+      "learning_rate": 3.746497665110074e-05,
+      "loss": 1.262,
+      "step": 9384
+    },
+    {
+      "epoch": 8.663743364874222,
+      "grad_norm": 3.033203125,
+      "learning_rate": 3.745830553702468e-05,
+      "loss": 1.2231,
+      "step": 9385
+    },
+    {
+      "epoch": 8.664666512808678,
+      "grad_norm": 3.35546875,
+      "learning_rate": 3.7451634422948635e-05,
+      "loss": 1.427,
+      "step": 9386
+    },
+    {
+      "epoch": 8.665589660743134,
+      "grad_norm": 3.220703125,
+      "learning_rate": 3.744496330887258e-05,
+      "loss": 1.396,
+      "step": 9387
+    },
+    {
+      "epoch": 8.66651280867759,
+      "grad_norm": 3.6328125,
+      "learning_rate": 3.743829219479653e-05,
+      "loss": 1.3403,
+      "step": 9388
+    },
+    {
+      "epoch": 8.667435956612048,
+      "grad_norm": 3.568359375,
+      "learning_rate": 3.743162108072048e-05,
+      "loss": 1.3752,
+      "step": 9389
+    },
+    {
+      "epoch": 8.668359104546504,
+      "grad_norm": 3.0546875,
+      "learning_rate": 3.742494996664443e-05,
+      "loss": 1.184,
+      "step": 9390
+    },
+    {
+      "epoch": 8.66928225248096,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.7418278852568376e-05,
+      "loss": 1.4607,
+      "step": 9391
+    },
+    {
+      "epoch": 8.670205400415416,
+      "grad_norm": 3.0859375,
+      "learning_rate": 3.7411607738492335e-05,
+      "loss": 1.0922,
+      "step": 9392
+    },
+    {
+      "epoch": 8.671128548349873,
+      "grad_norm": 3.08203125,
+      "learning_rate": 3.740493662441628e-05,
+      "loss": 1.4006,
+      "step": 9393
+    },
+    {
+      "epoch": 8.67205169628433,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.739826551034023e-05,
+      "loss": 1.1423,
+      "step": 9394
+    },
+    {
+      "epoch": 8.672974844218786,
+      "grad_norm": 3.076171875,
+      "learning_rate": 3.739159439626418e-05,
+      "loss": 1.4592,
+      "step": 9395
+    },
+    {
+      "epoch": 8.673897992153243,
+      "grad_norm": 3.400390625,
+      "learning_rate": 3.738492328218813e-05,
+      "loss": 1.1308,
+      "step": 9396
+    },
+    {
+      "epoch": 8.6748211400877,
+      "grad_norm": 3.869140625,
+      "learning_rate": 3.7378252168112075e-05,
+      "loss": 1.4363,
+      "step": 9397
+    },
+    {
+      "epoch": 8.675744288022155,
+      "grad_norm": 2.654296875,
+      "learning_rate": 3.737158105403603e-05,
+      "loss": 1.3077,
+      "step": 9398
+    },
+    {
+      "epoch": 8.676667435956611,
+      "grad_norm": 3.4921875,
+      "learning_rate": 3.736490993995997e-05,
+      "loss": 1.3838,
+      "step": 9399
+    },
+    {
+      "epoch": 8.67759058389107,
+      "grad_norm": 3.150390625,
+      "learning_rate": 3.7358238825883925e-05,
+      "loss": 1.5051,
+      "step": 9400
+    },
+    {
+      "epoch": 8.678513731825525,
+      "grad_norm": 3.80859375,
+      "learning_rate": 3.735156771180787e-05,
+      "loss": 1.4592,
+      "step": 9401
+    },
+    {
+      "epoch": 8.679436879759981,
+      "grad_norm": 3.30078125,
+      "learning_rate": 3.734489659773182e-05,
+      "loss": 1.0416,
+      "step": 9402
+    },
+    {
+      "epoch": 8.680360027694437,
+      "grad_norm": 3.2734375,
+      "learning_rate": 3.733822548365577e-05,
+      "loss": 1.4109,
+      "step": 9403
+    },
+    {
+      "epoch": 8.681283175628895,
+      "grad_norm": 3.01171875,
+      "learning_rate": 3.733155436957973e-05,
+      "loss": 1.4749,
+      "step": 9404
+    },
+    {
+      "epoch": 8.682206323563351,
+      "grad_norm": 3.369140625,
+      "learning_rate": 3.732488325550367e-05,
+      "loss": 1.1676,
+      "step": 9405
+    },
+    {
+      "epoch": 8.683129471497807,
+      "grad_norm": 3.556640625,
+      "learning_rate": 3.7318212141427625e-05,
+      "loss": 1.2959,
+      "step": 9406
+    },
+    {
+      "epoch": 8.684052619432265,
+      "grad_norm": 3.248046875,
+      "learning_rate": 3.731154102735157e-05,
+      "loss": 1.3435,
+      "step": 9407
+    },
+    {
+      "epoch": 8.684975767366721,
+      "grad_norm": 3.294921875,
+      "learning_rate": 3.730486991327552e-05,
+      "loss": 1.2197,
+      "step": 9408
+    },
+    {
+      "epoch": 8.685898915301177,
+      "grad_norm": 3.072265625,
+      "learning_rate": 3.729819879919947e-05,
+      "loss": 1.1047,
+      "step": 9409
+    },
+    {
+      "epoch": 8.686822063235633,
+      "grad_norm": 3.931640625,
+      "learning_rate": 3.729152768512342e-05,
+      "loss": 1.2283,
+      "step": 9410
+    },
+    {
+      "epoch": 8.68774521117009,
+      "grad_norm": 3.869140625,
+      "learning_rate": 3.7284856571047366e-05,
+      "loss": 1.293,
+      "step": 9411
+    },
+    {
+      "epoch": 8.688668359104547,
+      "grad_norm": 3.087890625,
+      "learning_rate": 3.727818545697132e-05,
+      "loss": 1.4592,
+      "step": 9412
+    },
+    {
+      "epoch": 8.689591507039003,
+      "grad_norm": 3.3359375,
+      "learning_rate": 3.727151434289526e-05,
+      "loss": 0.96,
+      "step": 9413
+    },
+    {
+      "epoch": 8.690514654973459,
+      "grad_norm": 3.533203125,
+      "learning_rate": 3.7264843228819215e-05,
+      "loss": 1.4404,
+      "step": 9414
+    },
+    {
+      "epoch": 8.691437802907917,
+      "grad_norm": 3.521484375,
+      "learning_rate": 3.725817211474316e-05,
+      "loss": 1.1995,
+      "step": 9415
+    },
+    {
+      "epoch": 8.692360950842373,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.725150100066711e-05,
+      "loss": 1.0782,
+      "step": 9416
+    },
+    {
+      "epoch": 8.693284098776829,
+      "grad_norm": 3.27734375,
+      "learning_rate": 3.7244829886591065e-05,
+      "loss": 1.3247,
+      "step": 9417
+    },
+    {
+      "epoch": 8.694207246711285,
+      "grad_norm": 3.525390625,
+      "learning_rate": 3.723815877251501e-05,
+      "loss": 1.1734,
+      "step": 9418
+    },
+    {
+      "epoch": 8.695130394645743,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.723148765843896e-05,
+      "loss": 1.1354,
+      "step": 9419
+    },
+    {
+      "epoch": 8.696053542580199,
+      "grad_norm": 3.12890625,
+      "learning_rate": 3.722481654436291e-05,
+      "loss": 1.6655,
+      "step": 9420
+    },
+    {
+      "epoch": 8.696976690514655,
+      "grad_norm": 3.25390625,
+      "learning_rate": 3.721814543028686e-05,
+      "loss": 1.2928,
+      "step": 9421
+    },
+    {
+      "epoch": 8.69789983844911,
+      "grad_norm": 3.37109375,
+      "learning_rate": 3.7211474316210806e-05,
+      "loss": 1.3159,
+      "step": 9422
+    },
+    {
+      "epoch": 8.698822986383568,
+      "grad_norm": 3.8671875,
+      "learning_rate": 3.720480320213476e-05,
+      "loss": 1.3214,
+      "step": 9423
+    },
+    {
+      "epoch": 8.699746134318024,
+      "grad_norm": 3.5,
+      "learning_rate": 3.7198132088058704e-05,
+      "loss": 1.1757,
+      "step": 9424
+    },
+    {
+      "epoch": 8.70066928225248,
+      "grad_norm": 3.89453125,
+      "learning_rate": 3.7191460973982656e-05,
+      "loss": 1.4597,
+      "step": 9425
+    },
+    {
+      "epoch": 8.701592430186938,
+      "grad_norm": 3.65234375,
+      "learning_rate": 3.718478985990661e-05,
+      "loss": 1.2844,
+      "step": 9426
+    },
+    {
+      "epoch": 8.702515578121394,
+      "grad_norm": 3.6015625,
+      "learning_rate": 3.7178118745830553e-05,
+      "loss": 1.3176,
+      "step": 9427
+    },
+    {
+      "epoch": 8.70343872605585,
+      "grad_norm": 3.318359375,
+      "learning_rate": 3.7171447631754506e-05,
+      "loss": 1.3228,
+      "step": 9428
+    },
+    {
+      "epoch": 8.704361873990306,
+      "grad_norm": 3.228515625,
+      "learning_rate": 3.716477651767846e-05,
+      "loss": 1.2521,
+      "step": 9429
+    },
+    {
+      "epoch": 8.705285021924764,
+      "grad_norm": 3.736328125,
+      "learning_rate": 3.71581054036024e-05,
+      "loss": 0.8902,
+      "step": 9430
+    },
+    {
+      "epoch": 8.70620816985922,
+      "grad_norm": 3.39453125,
+      "learning_rate": 3.7151434289526356e-05,
+      "loss": 1.3452,
+      "step": 9431
+    },
+    {
+      "epoch": 8.707131317793676,
+      "grad_norm": 3.724609375,
+      "learning_rate": 3.71447631754503e-05,
+      "loss": 1.4019,
+      "step": 9432
+    },
+    {
+      "epoch": 8.708054465728132,
+      "grad_norm": 3.357421875,
+      "learning_rate": 3.713809206137425e-05,
+      "loss": 1.3127,
+      "step": 9433
+    },
+    {
+      "epoch": 8.70897761366259,
+      "grad_norm": 2.951171875,
+      "learning_rate": 3.71314209472982e-05,
+      "loss": 1.0598,
+      "step": 9434
+    },
+    {
+      "epoch": 8.709900761597046,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.712474983322215e-05,
+      "loss": 1.1044,
+      "step": 9435
+    },
+    {
+      "epoch": 8.710823909531502,
+      "grad_norm": 3.763671875,
+      "learning_rate": 3.7118078719146096e-05,
+      "loss": 1.6064,
+      "step": 9436
+    },
+    {
+      "epoch": 8.711747057465958,
+      "grad_norm": 3.345703125,
+      "learning_rate": 3.711140760507005e-05,
+      "loss": 1.3921,
+      "step": 9437
+    },
+    {
+      "epoch": 8.712670205400416,
+      "grad_norm": 3.44921875,
+      "learning_rate": 3.7104736490993994e-05,
+      "loss": 1.3271,
+      "step": 9438
+    },
+    {
+      "epoch": 8.713593353334872,
+      "grad_norm": 2.92578125,
+      "learning_rate": 3.7098065376917946e-05,
+      "loss": 1.1193,
+      "step": 9439
+    },
+    {
+      "epoch": 8.714516501269328,
+      "grad_norm": 4.16796875,
+      "learning_rate": 3.709139426284189e-05,
+      "loss": 1.3682,
+      "step": 9440
+    },
+    {
+      "epoch": 8.715439649203784,
+      "grad_norm": 3.16796875,
+      "learning_rate": 3.7084723148765844e-05,
+      "loss": 1.1061,
+      "step": 9441
+    },
+    {
+      "epoch": 8.716362797138242,
+      "grad_norm": 3.19921875,
+      "learning_rate": 3.7078052034689796e-05,
+      "loss": 1.2935,
+      "step": 9442
+    },
+    {
+      "epoch": 8.717285945072698,
+      "grad_norm": 3.3359375,
+      "learning_rate": 3.707138092061375e-05,
+      "loss": 1.4175,
+      "step": 9443
+    },
+    {
+      "epoch": 8.718209093007154,
+      "grad_norm": 3.357421875,
+      "learning_rate": 3.7064709806537694e-05,
+      "loss": 1.3167,
+      "step": 9444
+    },
+    {
+      "epoch": 8.719132240941612,
+      "grad_norm": 3.25,
+      "learning_rate": 3.7058038692461646e-05,
+      "loss": 1.2297,
+      "step": 9445
+    },
+    {
+      "epoch": 8.720055388876068,
+      "grad_norm": 3.166015625,
+      "learning_rate": 3.705136757838559e-05,
+      "loss": 1.4084,
+      "step": 9446
+    },
+    {
+      "epoch": 8.720978536810524,
+      "grad_norm": 4.0078125,
+      "learning_rate": 3.7044696464309543e-05,
+      "loss": 1.3354,
+      "step": 9447
+    },
+    {
+      "epoch": 8.72190168474498,
+      "grad_norm": 3.4609375,
+      "learning_rate": 3.703802535023349e-05,
+      "loss": 1.1576,
+      "step": 9448
+    },
+    {
+      "epoch": 8.722824832679438,
+      "grad_norm": 3.623046875,
+      "learning_rate": 3.703135423615744e-05,
+      "loss": 1.2683,
+      "step": 9449
+    },
+    {
+      "epoch": 8.723747980613894,
+      "grad_norm": 3.70703125,
+      "learning_rate": 3.7024683122081386e-05,
+      "loss": 1.3123,
+      "step": 9450
+    },
+    {
+      "epoch": 8.72467112854835,
+      "grad_norm": 3.626953125,
+      "learning_rate": 3.701801200800534e-05,
+      "loss": 1.2441,
+      "step": 9451
+    },
+    {
+      "epoch": 8.725594276482806,
+      "grad_norm": 3.19140625,
+      "learning_rate": 3.7011340893929284e-05,
+      "loss": 1.5439,
+      "step": 9452
+    },
+    {
+      "epoch": 8.726517424417263,
+      "grad_norm": 3.517578125,
+      "learning_rate": 3.7004669779853236e-05,
+      "loss": 1.0814,
+      "step": 9453
+    },
+    {
+      "epoch": 8.72744057235172,
+      "grad_norm": 3.447265625,
+      "learning_rate": 3.699799866577719e-05,
+      "loss": 1.291,
+      "step": 9454
+    },
+    {
+      "epoch": 8.728363720286175,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.699132755170114e-05,
+      "loss": 1.3184,
+      "step": 9455
+    },
+    {
+      "epoch": 8.729286868220633,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.6984656437625086e-05,
+      "loss": 1.3735,
+      "step": 9456
+    },
+    {
+      "epoch": 8.73021001615509,
+      "grad_norm": 3.248046875,
+      "learning_rate": 3.697798532354904e-05,
+      "loss": 1.3232,
+      "step": 9457
+    },
+    {
+      "epoch": 8.731133164089545,
+      "grad_norm": 4.1171875,
+      "learning_rate": 3.6971314209472984e-05,
+      "loss": 1.2336,
+      "step": 9458
+    },
+    {
+      "epoch": 8.732056312024001,
+      "grad_norm": 3.208984375,
+      "learning_rate": 3.6964643095396936e-05,
+      "loss": 1.1686,
+      "step": 9459
+    },
+    {
+      "epoch": 8.73297945995846,
+      "grad_norm": 2.92578125,
+      "learning_rate": 3.695797198132088e-05,
+      "loss": 1.5073,
+      "step": 9460
+    },
+    {
+      "epoch": 8.733902607892915,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.6951300867244834e-05,
+      "loss": 1.2737,
+      "step": 9461
+    },
+    {
+      "epoch": 8.734825755827371,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.694462975316878e-05,
+      "loss": 1.3835,
+      "step": 9462
+    },
+    {
+      "epoch": 8.735748903761827,
+      "grad_norm": 3.673828125,
+      "learning_rate": 3.693795863909273e-05,
+      "loss": 1.448,
+      "step": 9463
+    },
+    {
+      "epoch": 8.736672051696285,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.693128752501668e-05,
+      "loss": 1.623,
+      "step": 9464
+    },
+    {
+      "epoch": 8.737595199630741,
+      "grad_norm": 3.548828125,
+      "learning_rate": 3.692461641094063e-05,
+      "loss": 1.3252,
+      "step": 9465
+    },
+    {
+      "epoch": 8.738518347565197,
+      "grad_norm": 2.74609375,
+      "learning_rate": 3.6917945296864574e-05,
+      "loss": 1.4937,
+      "step": 9466
+    },
+    {
+      "epoch": 8.739441495499653,
+      "grad_norm": 3.716796875,
+      "learning_rate": 3.691127418278853e-05,
+      "loss": 1.4529,
+      "step": 9467
+    },
+    {
+      "epoch": 8.740364643434111,
+      "grad_norm": 3.5390625,
+      "learning_rate": 3.690460306871248e-05,
+      "loss": 1.2234,
+      "step": 9468
+    },
+    {
+      "epoch": 8.741287791368567,
+      "grad_norm": 3.431640625,
+      "learning_rate": 3.689793195463643e-05,
+      "loss": 1.4453,
+      "step": 9469
+    },
+    {
+      "epoch": 8.742210939303023,
+      "grad_norm": 3.630859375,
+      "learning_rate": 3.6891260840560376e-05,
+      "loss": 1.5269,
+      "step": 9470
+    },
+    {
+      "epoch": 8.743134087237479,
+      "grad_norm": 3.634765625,
+      "learning_rate": 3.688458972648433e-05,
+      "loss": 1.3687,
+      "step": 9471
+    },
+    {
+      "epoch": 8.744057235171937,
+      "grad_norm": 3.5234375,
+      "learning_rate": 3.6877918612408274e-05,
+      "loss": 1.3362,
+      "step": 9472
+    },
+    {
+      "epoch": 8.744980383106393,
+      "grad_norm": 3.001953125,
+      "learning_rate": 3.6871247498332226e-05,
+      "loss": 1.2477,
+      "step": 9473
+    },
+    {
+      "epoch": 8.745903531040849,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.686457638425617e-05,
+      "loss": 1.3896,
+      "step": 9474
+    },
+    {
+      "epoch": 8.746826678975307,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.6857905270180124e-05,
+      "loss": 1.3853,
+      "step": 9475
+    },
+    {
+      "epoch": 8.747749826909763,
+      "grad_norm": 2.7734375,
+      "learning_rate": 3.685123415610407e-05,
+      "loss": 1.2281,
+      "step": 9476
+    },
+    {
+      "epoch": 8.748672974844219,
+      "grad_norm": 3.39453125,
+      "learning_rate": 3.684456304202802e-05,
+      "loss": 1.2473,
+      "step": 9477
+    },
+    {
+      "epoch": 8.749596122778675,
+      "grad_norm": 3.396484375,
+      "learning_rate": 3.683789192795197e-05,
+      "loss": 1.4958,
+      "step": 9478
+    },
+    {
+      "epoch": 8.750519270713133,
+      "grad_norm": 3.626953125,
+      "learning_rate": 3.683122081387592e-05,
+      "loss": 1.3682,
+      "step": 9479
+    },
+    {
+      "epoch": 8.751442418647589,
+      "grad_norm": 3.560546875,
+      "learning_rate": 3.682454969979987e-05,
+      "loss": 1.6287,
+      "step": 9480
+    },
+    {
+      "epoch": 8.752365566582045,
+      "grad_norm": 3.244140625,
+      "learning_rate": 3.681787858572382e-05,
+      "loss": 1.2192,
+      "step": 9481
+    },
+    {
+      "epoch": 8.7532887145165,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.681120747164777e-05,
+      "loss": 1.2723,
+      "step": 9482
+    },
+    {
+      "epoch": 8.754211862450958,
+      "grad_norm": 3.673828125,
+      "learning_rate": 3.6804536357571714e-05,
+      "loss": 1.2502,
+      "step": 9483
+    },
+    {
+      "epoch": 8.755135010385414,
+      "grad_norm": 3.076171875,
+      "learning_rate": 3.679786524349567e-05,
+      "loss": 1.3289,
+      "step": 9484
+    },
+    {
+      "epoch": 8.75605815831987,
+      "grad_norm": 3.27734375,
+      "learning_rate": 3.679119412941961e-05,
+      "loss": 1.3494,
+      "step": 9485
+    },
+    {
+      "epoch": 8.756981306254326,
+      "grad_norm": 3.638671875,
+      "learning_rate": 3.6784523015343564e-05,
+      "loss": 1.337,
+      "step": 9486
+    },
+    {
+      "epoch": 8.757904454188784,
+      "grad_norm": 3.3828125,
+      "learning_rate": 3.677785190126751e-05,
+      "loss": 1.2091,
+      "step": 9487
+    },
+    {
+      "epoch": 8.75882760212324,
+      "grad_norm": 3.3359375,
+      "learning_rate": 3.677118078719146e-05,
+      "loss": 1.3101,
+      "step": 9488
+    },
+    {
+      "epoch": 8.759750750057696,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.676450967311541e-05,
+      "loss": 1.1736,
+      "step": 9489
+    },
+    {
+      "epoch": 8.760673897992152,
+      "grad_norm": 3.814453125,
+      "learning_rate": 3.675783855903936e-05,
+      "loss": 1.2798,
+      "step": 9490
+    },
+    {
+      "epoch": 8.76159704592661,
+      "grad_norm": 2.98828125,
+      "learning_rate": 3.6751167444963305e-05,
+      "loss": 1.1973,
+      "step": 9491
+    },
+    {
+      "epoch": 8.762520193861066,
+      "grad_norm": 3.201171875,
+      "learning_rate": 3.6744496330887264e-05,
+      "loss": 1.3254,
+      "step": 9492
+    },
+    {
+      "epoch": 8.763443341795522,
+      "grad_norm": 3.6015625,
+      "learning_rate": 3.673782521681121e-05,
+      "loss": 1.1703,
+      "step": 9493
+    },
+    {
+      "epoch": 8.76436648972998,
+      "grad_norm": 3.2734375,
+      "learning_rate": 3.673115410273516e-05,
+      "loss": 1.312,
+      "step": 9494
+    },
+    {
+      "epoch": 8.765289637664436,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.672448298865911e-05,
+      "loss": 1.0847,
+      "step": 9495
+    },
+    {
+      "epoch": 8.766212785598892,
+      "grad_norm": 3.197265625,
+      "learning_rate": 3.671781187458306e-05,
+      "loss": 1.3591,
+      "step": 9496
+    },
+    {
+      "epoch": 8.767135933533348,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.6711140760507005e-05,
+      "loss": 1.4121,
+      "step": 9497
+    },
+    {
+      "epoch": 8.768059081467806,
+      "grad_norm": 4.06640625,
+      "learning_rate": 3.670446964643096e-05,
+      "loss": 1.131,
+      "step": 9498
+    },
+    {
+      "epoch": 8.768982229402262,
+      "grad_norm": 3.54296875,
+      "learning_rate": 3.66977985323549e-05,
+      "loss": 1.1664,
+      "step": 9499
+    },
+    {
+      "epoch": 8.769905377336718,
+      "grad_norm": 3.033203125,
+      "learning_rate": 3.6691127418278855e-05,
+      "loss": 1.1523,
+      "step": 9500
+    },
+    {
+      "epoch": 8.770828525271174,
+      "grad_norm": 3.986328125,
+      "learning_rate": 3.66844563042028e-05,
+      "loss": 1.2013,
+      "step": 9501
+    },
+    {
+      "epoch": 8.771751673205632,
+      "grad_norm": 3.6171875,
+      "learning_rate": 3.667778519012675e-05,
+      "loss": 1.3303,
+      "step": 9502
+    },
+    {
+      "epoch": 8.772674821140088,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.66711140760507e-05,
+      "loss": 1.261,
+      "step": 9503
+    },
+    {
+      "epoch": 8.773597969074544,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.666444296197465e-05,
+      "loss": 1.3582,
+      "step": 9504
+    },
+    {
+      "epoch": 8.774521117009002,
+      "grad_norm": 3.28515625,
+      "learning_rate": 3.66577718478986e-05,
+      "loss": 0.8389,
+      "step": 9505
+    },
+    {
+      "epoch": 8.775444264943458,
+      "grad_norm": 3.166015625,
+      "learning_rate": 3.6651100733822554e-05,
+      "loss": 1.2537,
+      "step": 9506
+    },
+    {
+      "epoch": 8.776367412877914,
+      "grad_norm": 3.244140625,
+      "learning_rate": 3.66444296197465e-05,
+      "loss": 1.3578,
+      "step": 9507
+    },
+    {
+      "epoch": 8.77729056081237,
+      "grad_norm": 2.7421875,
+      "learning_rate": 3.663775850567045e-05,
+      "loss": 1.1526,
+      "step": 9508
+    },
+    {
+      "epoch": 8.778213708746827,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.66310873915944e-05,
+      "loss": 1.2878,
+      "step": 9509
+    },
+    {
+      "epoch": 8.779136856681284,
+      "grad_norm": 3.509765625,
+      "learning_rate": 3.662441627751835e-05,
+      "loss": 1.3279,
+      "step": 9510
+    },
+    {
+      "epoch": 8.78006000461574,
+      "grad_norm": 3.55078125,
+      "learning_rate": 3.6617745163442295e-05,
+      "loss": 1.3992,
+      "step": 9511
+    },
+    {
+      "epoch": 8.780983152550196,
+      "grad_norm": 3.26171875,
+      "learning_rate": 3.661107404936625e-05,
+      "loss": 1.2014,
+      "step": 9512
+    },
+    {
+      "epoch": 8.781906300484653,
+      "grad_norm": 3.033203125,
+      "learning_rate": 3.660440293529019e-05,
+      "loss": 1.3777,
+      "step": 9513
+    },
+    {
+      "epoch": 8.78282944841911,
+      "grad_norm": 3.62109375,
+      "learning_rate": 3.6597731821214145e-05,
+      "loss": 1.4333,
+      "step": 9514
+    },
+    {
+      "epoch": 8.783752596353565,
+      "grad_norm": 3.0703125,
+      "learning_rate": 3.659106070713809e-05,
+      "loss": 1.4941,
+      "step": 9515
+    },
+    {
+      "epoch": 8.784675744288021,
+      "grad_norm": 4.41796875,
+      "learning_rate": 3.658438959306204e-05,
+      "loss": 1.0068,
+      "step": 9516
+    },
+    {
+      "epoch": 8.78559889222248,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.6577718478985995e-05,
+      "loss": 1.4189,
+      "step": 9517
+    },
+    {
+      "epoch": 8.786522040156935,
+      "grad_norm": 3.4609375,
+      "learning_rate": 3.657104736490995e-05,
+      "loss": 1.2634,
+      "step": 9518
+    },
+    {
+      "epoch": 8.787445188091391,
+      "grad_norm": 2.966796875,
+      "learning_rate": 3.656437625083389e-05,
+      "loss": 1.1591,
+      "step": 9519
+    },
+    {
+      "epoch": 8.788368336025847,
+      "grad_norm": 3.5703125,
+      "learning_rate": 3.6557705136757844e-05,
+      "loss": 1.3574,
+      "step": 9520
+    },
+    {
+      "epoch": 8.789291483960305,
+      "grad_norm": 3.755859375,
+      "learning_rate": 3.655103402268179e-05,
+      "loss": 1.0121,
+      "step": 9521
+    },
+    {
+      "epoch": 8.790214631894761,
+      "grad_norm": 3.044921875,
+      "learning_rate": 3.654436290860574e-05,
+      "loss": 1.2686,
+      "step": 9522
+    },
+    {
+      "epoch": 8.791137779829217,
+      "grad_norm": 2.7734375,
+      "learning_rate": 3.653769179452969e-05,
+      "loss": 1.3501,
+      "step": 9523
+    },
+    {
+      "epoch": 8.792060927763675,
+      "grad_norm": 4.04296875,
+      "learning_rate": 3.653102068045364e-05,
+      "loss": 1.3848,
+      "step": 9524
+    },
+    {
+      "epoch": 8.792984075698131,
+      "grad_norm": 3.369140625,
+      "learning_rate": 3.6524349566377585e-05,
+      "loss": 1.4199,
+      "step": 9525
+    },
+    {
+      "epoch": 8.793907223632587,
+      "grad_norm": 3.638671875,
+      "learning_rate": 3.651767845230154e-05,
+      "loss": 1.3213,
+      "step": 9526
+    },
+    {
+      "epoch": 8.794830371567043,
+      "grad_norm": 3.138671875,
+      "learning_rate": 3.651100733822548e-05,
+      "loss": 1.2407,
+      "step": 9527
+    },
+    {
+      "epoch": 8.7957535195015,
+      "grad_norm": 3.3828125,
+      "learning_rate": 3.6504336224149435e-05,
+      "loss": 1.2693,
+      "step": 9528
+    },
+    {
+      "epoch": 8.796676667435957,
+      "grad_norm": 3.33203125,
+      "learning_rate": 3.649766511007338e-05,
+      "loss": 1.2778,
+      "step": 9529
+    },
+    {
+      "epoch": 8.797599815370413,
+      "grad_norm": 3.439453125,
+      "learning_rate": 3.649099399599733e-05,
+      "loss": 1.2976,
+      "step": 9530
+    },
+    {
+      "epoch": 8.798522963304869,
+      "grad_norm": 2.990234375,
+      "learning_rate": 3.6484322881921285e-05,
+      "loss": 1.3843,
+      "step": 9531
+    },
+    {
+      "epoch": 8.799446111239327,
+      "grad_norm": 2.83984375,
+      "learning_rate": 3.647765176784523e-05,
+      "loss": 1.197,
+      "step": 9532
+    },
+    {
+      "epoch": 8.800369259173783,
+      "grad_norm": 4.23828125,
+      "learning_rate": 3.647098065376918e-05,
+      "loss": 0.913,
+      "step": 9533
+    },
+    {
+      "epoch": 8.801292407108239,
+      "grad_norm": 3.173828125,
+      "learning_rate": 3.646430953969313e-05,
+      "loss": 1.417,
+      "step": 9534
+    },
+    {
+      "epoch": 8.802215555042695,
+      "grad_norm": 3.013671875,
+      "learning_rate": 3.645763842561708e-05,
+      "loss": 1.2581,
+      "step": 9535
+    },
+    {
+      "epoch": 8.803138702977153,
+      "grad_norm": 3.64453125,
+      "learning_rate": 3.6450967311541026e-05,
+      "loss": 1.3252,
+      "step": 9536
+    },
+    {
+      "epoch": 8.804061850911609,
+      "grad_norm": 3.376953125,
+      "learning_rate": 3.644429619746498e-05,
+      "loss": 1.0704,
+      "step": 9537
+    },
+    {
+      "epoch": 8.804984998846065,
+      "grad_norm": 3.521484375,
+      "learning_rate": 3.643762508338892e-05,
+      "loss": 1.426,
+      "step": 9538
+    },
+    {
+      "epoch": 8.80590814678052,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.6430953969312875e-05,
+      "loss": 1.3562,
+      "step": 9539
+    },
+    {
+      "epoch": 8.806831294714978,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.642428285523683e-05,
+      "loss": 1.2457,
+      "step": 9540
+    },
+    {
+      "epoch": 8.807754442649435,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.641761174116077e-05,
+      "loss": 1.2365,
+      "step": 9541
+    },
+    {
+      "epoch": 8.80867759058389,
+      "grad_norm": 3.658203125,
+      "learning_rate": 3.6410940627084725e-05,
+      "loss": 1.2673,
+      "step": 9542
+    },
+    {
+      "epoch": 8.809600738518348,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.640426951300868e-05,
+      "loss": 1.0746,
+      "step": 9543
+    },
+    {
+      "epoch": 8.810523886452804,
+      "grad_norm": 3.388671875,
+      "learning_rate": 3.639759839893262e-05,
+      "loss": 1.0988,
+      "step": 9544
+    },
+    {
+      "epoch": 8.81144703438726,
+      "grad_norm": 3.6328125,
+      "learning_rate": 3.6390927284856575e-05,
+      "loss": 1.2405,
+      "step": 9545
+    },
+    {
+      "epoch": 8.812370182321716,
+      "grad_norm": 3.310546875,
+      "learning_rate": 3.638425617078052e-05,
+      "loss": 0.9821,
+      "step": 9546
+    },
+    {
+      "epoch": 8.813293330256174,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.637758505670447e-05,
+      "loss": 1.3826,
+      "step": 9547
+    },
+    {
+      "epoch": 8.81421647819063,
+      "grad_norm": 3.111328125,
+      "learning_rate": 3.637091394262842e-05,
+      "loss": 1.2781,
+      "step": 9548
+    },
+    {
+      "epoch": 8.815139626125086,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.636424282855237e-05,
+      "loss": 1.3325,
+      "step": 9549
+    },
+    {
+      "epoch": 8.816062774059542,
+      "grad_norm": 3.62890625,
+      "learning_rate": 3.6357571714476316e-05,
+      "loss": 1.2546,
+      "step": 9550
+    },
+    {
+      "epoch": 8.816985921994,
+      "grad_norm": 3.787109375,
+      "learning_rate": 3.635090060040027e-05,
+      "loss": 1.5029,
+      "step": 9551
+    },
+    {
+      "epoch": 8.817909069928456,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.6344229486324213e-05,
+      "loss": 1.3328,
+      "step": 9552
+    },
+    {
+      "epoch": 8.818832217862912,
+      "grad_norm": 2.70703125,
+      "learning_rate": 3.6337558372248166e-05,
+      "loss": 1.0584,
+      "step": 9553
+    },
+    {
+      "epoch": 8.81975536579737,
+      "grad_norm": 3.232421875,
+      "learning_rate": 3.633088725817211e-05,
+      "loss": 1.3621,
+      "step": 9554
+    },
+    {
+      "epoch": 8.820678513731826,
+      "grad_norm": 3.498046875,
+      "learning_rate": 3.632421614409607e-05,
+      "loss": 1.3965,
+      "step": 9555
+    },
+    {
+      "epoch": 8.821601661666282,
+      "grad_norm": 4.078125,
+      "learning_rate": 3.6317545030020015e-05,
+      "loss": 1.3228,
+      "step": 9556
+    },
+    {
+      "epoch": 8.822524809600738,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.631087391594397e-05,
+      "loss": 1.2798,
+      "step": 9557
+    },
+    {
+      "epoch": 8.823447957535196,
+      "grad_norm": 2.98828125,
+      "learning_rate": 3.630420280186791e-05,
+      "loss": 1.3572,
+      "step": 9558
+    },
+    {
+      "epoch": 8.824371105469652,
+      "grad_norm": 3.490234375,
+      "learning_rate": 3.6297531687791865e-05,
+      "loss": 1.3517,
+      "step": 9559
+    },
+    {
+      "epoch": 8.825294253404108,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.629086057371581e-05,
+      "loss": 1.3232,
+      "step": 9560
+    },
+    {
+      "epoch": 8.826217401338564,
+      "grad_norm": 3.00390625,
+      "learning_rate": 3.628418945963976e-05,
+      "loss": 1.1486,
+      "step": 9561
+    },
+    {
+      "epoch": 8.827140549273022,
+      "grad_norm": 3.341796875,
+      "learning_rate": 3.627751834556371e-05,
+      "loss": 1.3977,
+      "step": 9562
+    },
+    {
+      "epoch": 8.828063697207478,
+      "grad_norm": 3.23828125,
+      "learning_rate": 3.627084723148766e-05,
+      "loss": 1.3091,
+      "step": 9563
+    },
+    {
+      "epoch": 8.828986845141934,
+      "grad_norm": 3.55859375,
+      "learning_rate": 3.6264176117411606e-05,
+      "loss": 1.1241,
+      "step": 9564
+    },
+    {
+      "epoch": 8.82990999307639,
+      "grad_norm": 3.017578125,
+      "learning_rate": 3.625750500333556e-05,
+      "loss": 1.3474,
+      "step": 9565
+    },
+    {
+      "epoch": 8.830833141010848,
+      "grad_norm": 3.12890625,
+      "learning_rate": 3.6250833889259504e-05,
+      "loss": 1.1358,
+      "step": 9566
+    },
+    {
+      "epoch": 8.831756288945304,
+      "grad_norm": 3.431640625,
+      "learning_rate": 3.624416277518346e-05,
+      "loss": 1.4832,
+      "step": 9567
+    },
+    {
+      "epoch": 8.83267943687976,
+      "grad_norm": 3.830078125,
+      "learning_rate": 3.623749166110741e-05,
+      "loss": 1.2871,
+      "step": 9568
+    },
+    {
+      "epoch": 8.833602584814216,
+      "grad_norm": 3.56640625,
+      "learning_rate": 3.623082054703136e-05,
+      "loss": 1.173,
+      "step": 9569
+    },
+    {
+      "epoch": 8.834525732748673,
+      "grad_norm": 3.5546875,
+      "learning_rate": 3.6224149432955306e-05,
+      "loss": 1.5276,
+      "step": 9570
+    },
+    {
+      "epoch": 8.83544888068313,
+      "grad_norm": 3.48828125,
+      "learning_rate": 3.621747831887926e-05,
+      "loss": 1.2173,
+      "step": 9571
+    },
+    {
+      "epoch": 8.836372028617586,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.62108072048032e-05,
+      "loss": 1.2903,
+      "step": 9572
+    },
+    {
+      "epoch": 8.837295176552043,
+      "grad_norm": 3.056640625,
+      "learning_rate": 3.6204136090727156e-05,
+      "loss": 1.2754,
+      "step": 9573
+    },
+    {
+      "epoch": 8.8382183244865,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.61974649766511e-05,
+      "loss": 1.2411,
+      "step": 9574
+    },
+    {
+      "epoch": 8.839141472420955,
+      "grad_norm": 3.02734375,
+      "learning_rate": 3.619079386257505e-05,
+      "loss": 1.1838,
+      "step": 9575
+    },
+    {
+      "epoch": 8.840064620355411,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.6184122748499e-05,
+      "loss": 1.5085,
+      "step": 9576
+    },
+    {
+      "epoch": 8.84098776828987,
+      "grad_norm": 3.287109375,
+      "learning_rate": 3.617745163442295e-05,
+      "loss": 1.4812,
+      "step": 9577
+    },
+    {
+      "epoch": 8.841910916224325,
+      "grad_norm": 3.482421875,
+      "learning_rate": 3.6170780520346896e-05,
+      "loss": 1.4976,
+      "step": 9578
+    },
+    {
+      "epoch": 8.842834064158781,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.616410940627085e-05,
+      "loss": 1.3972,
+      "step": 9579
+    },
+    {
+      "epoch": 8.843757212093237,
+      "grad_norm": 3.017578125,
+      "learning_rate": 3.61574382921948e-05,
+      "loss": 1.6487,
+      "step": 9580
+    },
+    {
+      "epoch": 8.844680360027695,
+      "grad_norm": 3.322265625,
+      "learning_rate": 3.615076717811875e-05,
+      "loss": 1.4885,
+      "step": 9581
+    },
+    {
+      "epoch": 8.845603507962151,
+      "grad_norm": 3.205078125,
+      "learning_rate": 3.61440960640427e-05,
+      "loss": 1.3813,
+      "step": 9582
+    },
+    {
+      "epoch": 8.846526655896607,
+      "grad_norm": 3.2890625,
+      "learning_rate": 3.613742494996665e-05,
+      "loss": 1.4097,
+      "step": 9583
+    },
+    {
+      "epoch": 8.847449803831063,
+      "grad_norm": 3.056640625,
+      "learning_rate": 3.6130753835890596e-05,
+      "loss": 1.334,
+      "step": 9584
+    },
+    {
+      "epoch": 8.848372951765521,
+      "grad_norm": 3.2578125,
+      "learning_rate": 3.612408272181455e-05,
+      "loss": 1.2445,
+      "step": 9585
+    },
+    {
+      "epoch": 8.849296099699977,
+      "grad_norm": 3.158203125,
+      "learning_rate": 3.6117411607738494e-05,
+      "loss": 1.3677,
+      "step": 9586
+    },
+    {
+      "epoch": 8.850219247634433,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.6110740493662446e-05,
+      "loss": 1.1665,
+      "step": 9587
+    },
+    {
+      "epoch": 8.851142395568889,
+      "grad_norm": 3.611328125,
+      "learning_rate": 3.610406937958639e-05,
+      "loss": 1.1394,
+      "step": 9588
+    },
+    {
+      "epoch": 8.852065543503347,
+      "grad_norm": 3.375,
+      "learning_rate": 3.6097398265510343e-05,
+      "loss": 1.5068,
+      "step": 9589
+    },
+    {
+      "epoch": 8.852988691437803,
+      "grad_norm": 3.9296875,
+      "learning_rate": 3.609072715143429e-05,
+      "loss": 1.1858,
+      "step": 9590
+    },
+    {
+      "epoch": 8.853911839372259,
+      "grad_norm": 3.341796875,
+      "learning_rate": 3.608405603735824e-05,
+      "loss": 1.1968,
+      "step": 9591
+    },
+    {
+      "epoch": 8.854834987306717,
+      "grad_norm": 3.07421875,
+      "learning_rate": 3.607738492328219e-05,
+      "loss": 1.084,
+      "step": 9592
+    },
+    {
+      "epoch": 8.855758135241173,
+      "grad_norm": 3.1640625,
+      "learning_rate": 3.607071380920614e-05,
+      "loss": 1.3027,
+      "step": 9593
+    },
+    {
+      "epoch": 8.856681283175629,
+      "grad_norm": 3.529296875,
+      "learning_rate": 3.606404269513009e-05,
+      "loss": 1.2927,
+      "step": 9594
+    },
+    {
+      "epoch": 8.857604431110085,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.6057371581054036e-05,
+      "loss": 1.0313,
+      "step": 9595
+    },
+    {
+      "epoch": 8.858527579044543,
+      "grad_norm": 3.3203125,
+      "learning_rate": 3.605070046697799e-05,
+      "loss": 1.3113,
+      "step": 9596
+    },
+    {
+      "epoch": 8.859450726978999,
+      "grad_norm": 3.333984375,
+      "learning_rate": 3.6044029352901934e-05,
+      "loss": 1.2383,
+      "step": 9597
+    },
+    {
+      "epoch": 8.860373874913455,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.6037358238825886e-05,
+      "loss": 1.1,
+      "step": 9598
+    },
+    {
+      "epoch": 8.86129702284791,
+      "grad_norm": 3.462890625,
+      "learning_rate": 3.603068712474983e-05,
+      "loss": 1.3115,
+      "step": 9599
+    },
+    {
+      "epoch": 8.862220170782368,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.6024016010673784e-05,
+      "loss": 1.3442,
+      "step": 9600
+    },
+    {
+      "epoch": 8.863143318716824,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.601734489659773e-05,
+      "loss": 1.0155,
+      "step": 9601
+    },
+    {
+      "epoch": 8.86406646665128,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.601067378252168e-05,
+      "loss": 1.3121,
+      "step": 9602
+    },
+    {
+      "epoch": 8.864989614585737,
+      "grad_norm": 3.37890625,
+      "learning_rate": 3.600400266844563e-05,
+      "loss": 1.1608,
+      "step": 9603
+    },
+    {
+      "epoch": 8.865912762520194,
+      "grad_norm": 3.337890625,
+      "learning_rate": 3.599733155436958e-05,
+      "loss": 1.1566,
+      "step": 9604
+    },
+    {
+      "epoch": 8.86683591045465,
+      "grad_norm": 3.525390625,
+      "learning_rate": 3.599066044029353e-05,
+      "loss": 1.2336,
+      "step": 9605
+    },
+    {
+      "epoch": 8.867759058389106,
+      "grad_norm": 3.6015625,
+      "learning_rate": 3.5983989326217483e-05,
+      "loss": 1.2236,
+      "step": 9606
+    },
+    {
+      "epoch": 8.868682206323562,
+      "grad_norm": 3.248046875,
+      "learning_rate": 3.597731821214143e-05,
+      "loss": 1.1709,
+      "step": 9607
+    },
+    {
+      "epoch": 8.86960535425802,
+      "grad_norm": 2.826171875,
+      "learning_rate": 3.597064709806538e-05,
+      "loss": 1.3116,
+      "step": 9608
+    },
+    {
+      "epoch": 8.870528502192476,
+      "grad_norm": 3.814453125,
+      "learning_rate": 3.5963975983989327e-05,
+      "loss": 1.1863,
+      "step": 9609
+    },
+    {
+      "epoch": 8.871451650126932,
+      "grad_norm": 3.572265625,
+      "learning_rate": 3.595730486991328e-05,
+      "loss": 1.2869,
+      "step": 9610
+    },
+    {
+      "epoch": 8.87237479806139,
+      "grad_norm": 3.498046875,
+      "learning_rate": 3.5950633755837224e-05,
+      "loss": 1.2207,
+      "step": 9611
+    },
+    {
+      "epoch": 8.873297945995846,
+      "grad_norm": 3.71484375,
+      "learning_rate": 3.5943962641761176e-05,
+      "loss": 1.5317,
+      "step": 9612
+    },
+    {
+      "epoch": 8.874221093930302,
+      "grad_norm": 3.365234375,
+      "learning_rate": 3.593729152768512e-05,
+      "loss": 1.219,
+      "step": 9613
+    },
+    {
+      "epoch": 8.875144241864758,
+      "grad_norm": 4.328125,
+      "learning_rate": 3.5930620413609074e-05,
+      "loss": 1.406,
+      "step": 9614
+    },
+    {
+      "epoch": 8.876067389799216,
+      "grad_norm": 3.232421875,
+      "learning_rate": 3.592394929953302e-05,
+      "loss": 1.2015,
+      "step": 9615
+    },
+    {
+      "epoch": 8.876990537733672,
+      "grad_norm": 3.564453125,
+      "learning_rate": 3.591727818545697e-05,
+      "loss": 1.226,
+      "step": 9616
+    },
+    {
+      "epoch": 8.877913685668128,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.5910607071380924e-05,
+      "loss": 1.4644,
+      "step": 9617
+    },
+    {
+      "epoch": 8.878836833602584,
+      "grad_norm": 2.591796875,
+      "learning_rate": 3.5903935957304876e-05,
+      "loss": 1.1613,
+      "step": 9618
+    },
+    {
+      "epoch": 8.879759981537042,
+      "grad_norm": 3.451171875,
+      "learning_rate": 3.589726484322882e-05,
+      "loss": 1.3845,
+      "step": 9619
+    },
+    {
+      "epoch": 8.880683129471498,
+      "grad_norm": 3.060546875,
+      "learning_rate": 3.5890593729152774e-05,
+      "loss": 1.4163,
+      "step": 9620
+    },
+    {
+      "epoch": 8.881606277405954,
+      "grad_norm": 2.876953125,
+      "learning_rate": 3.588392261507672e-05,
+      "loss": 1.1907,
+      "step": 9621
+    },
+    {
+      "epoch": 8.882529425340412,
+      "grad_norm": 3.0078125,
+      "learning_rate": 3.587725150100067e-05,
+      "loss": 1.1875,
+      "step": 9622
+    },
+    {
+      "epoch": 8.883452573274868,
+      "grad_norm": 3.02734375,
+      "learning_rate": 3.587058038692462e-05,
+      "loss": 1.3672,
+      "step": 9623
+    },
+    {
+      "epoch": 8.884375721209324,
+      "grad_norm": 3.232421875,
+      "learning_rate": 3.586390927284857e-05,
+      "loss": 0.9009,
+      "step": 9624
+    },
+    {
+      "epoch": 8.88529886914378,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.5857238158772514e-05,
+      "loss": 1.2195,
+      "step": 9625
+    },
+    {
+      "epoch": 8.886222017078238,
+      "grad_norm": 3.529296875,
+      "learning_rate": 3.585056704469647e-05,
+      "loss": 1.1445,
+      "step": 9626
+    },
+    {
+      "epoch": 8.887145165012694,
+      "grad_norm": 3.5234375,
+      "learning_rate": 3.584389593062041e-05,
+      "loss": 1.3557,
+      "step": 9627
+    },
+    {
+      "epoch": 8.88806831294715,
+      "grad_norm": 3.23828125,
+      "learning_rate": 3.5837224816544364e-05,
+      "loss": 1.3247,
+      "step": 9628
+    },
+    {
+      "epoch": 8.888991460881606,
+      "grad_norm": 3.173828125,
+      "learning_rate": 3.583055370246831e-05,
+      "loss": 1.2509,
+      "step": 9629
+    },
+    {
+      "epoch": 8.889914608816063,
+      "grad_norm": 3.20703125,
+      "learning_rate": 3.582388258839227e-05,
+      "loss": 1.4563,
+      "step": 9630
+    },
+    {
+      "epoch": 8.89083775675052,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.5817211474316214e-05,
+      "loss": 1.5288,
+      "step": 9631
+    },
+    {
+      "epoch": 8.891760904684975,
+      "grad_norm": 3.529296875,
+      "learning_rate": 3.5810540360240166e-05,
+      "loss": 1.5637,
+      "step": 9632
+    },
+    {
+      "epoch": 8.892684052619432,
+      "grad_norm": 3.103515625,
+      "learning_rate": 3.580386924616411e-05,
+      "loss": 1.4412,
+      "step": 9633
+    },
+    {
+      "epoch": 8.89360720055389,
+      "grad_norm": 3.33984375,
+      "learning_rate": 3.5797198132088064e-05,
+      "loss": 1.0063,
+      "step": 9634
+    },
+    {
+      "epoch": 8.894530348488345,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.579052701801201e-05,
+      "loss": 1.3428,
+      "step": 9635
+    },
+    {
+      "epoch": 8.895453496422801,
+      "grad_norm": 3.091796875,
+      "learning_rate": 3.578385590393596e-05,
+      "loss": 1.3015,
+      "step": 9636
+    },
+    {
+      "epoch": 8.896376644357257,
+      "grad_norm": 3.662109375,
+      "learning_rate": 3.577718478985991e-05,
+      "loss": 1.3901,
+      "step": 9637
+    },
+    {
+      "epoch": 8.897299792291715,
+      "grad_norm": 2.6953125,
+      "learning_rate": 3.577051367578386e-05,
+      "loss": 1.4197,
+      "step": 9638
+    },
+    {
+      "epoch": 8.898222940226171,
+      "grad_norm": 3.419921875,
+      "learning_rate": 3.5763842561707805e-05,
+      "loss": 1.386,
+      "step": 9639
+    },
+    {
+      "epoch": 8.899146088160627,
+      "grad_norm": 3.927734375,
+      "learning_rate": 3.575717144763176e-05,
+      "loss": 1.0955,
+      "step": 9640
+    },
+    {
+      "epoch": 8.900069236095085,
+      "grad_norm": 3.083984375,
+      "learning_rate": 3.57505003335557e-05,
+      "loss": 1.4783,
+      "step": 9641
+    },
+    {
+      "epoch": 8.900992384029541,
+      "grad_norm": 4.11328125,
+      "learning_rate": 3.5743829219479655e-05,
+      "loss": 1.2461,
+      "step": 9642
+    },
+    {
+      "epoch": 8.901915531963997,
+      "grad_norm": 3.4765625,
+      "learning_rate": 3.573715810540361e-05,
+      "loss": 1.0947,
+      "step": 9643
+    },
+    {
+      "epoch": 8.902838679898453,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.573048699132755e-05,
+      "loss": 1.2472,
+      "step": 9644
+    },
+    {
+      "epoch": 8.903761827832911,
+      "grad_norm": 3.134765625,
+      "learning_rate": 3.5723815877251504e-05,
+      "loss": 1.3416,
+      "step": 9645
+    },
+    {
+      "epoch": 8.904684975767367,
+      "grad_norm": 4.16015625,
+      "learning_rate": 3.571714476317545e-05,
+      "loss": 1.634,
+      "step": 9646
+    },
+    {
+      "epoch": 8.905608123701823,
+      "grad_norm": 3.037109375,
+      "learning_rate": 3.57104736490994e-05,
+      "loss": 1.3196,
+      "step": 9647
+    },
+    {
+      "epoch": 8.906531271636279,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.570380253502335e-05,
+      "loss": 1.6052,
+      "step": 9648
+    },
+    {
+      "epoch": 8.907454419570737,
+      "grad_norm": 3.353515625,
+      "learning_rate": 3.56971314209473e-05,
+      "loss": 1.478,
+      "step": 9649
+    },
+    {
+      "epoch": 8.908377567505193,
+      "grad_norm": 3.01953125,
+      "learning_rate": 3.5690460306871245e-05,
+      "loss": 1.4102,
+      "step": 9650
+    },
+    {
+      "epoch": 8.909300715439649,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.56837891927952e-05,
+      "loss": 1.3875,
+      "step": 9651
+    },
+    {
+      "epoch": 8.910223863374105,
+      "grad_norm": 3.50390625,
+      "learning_rate": 3.567711807871915e-05,
+      "loss": 1.2731,
+      "step": 9652
+    },
+    {
+      "epoch": 8.911147011308563,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.5670446964643095e-05,
+      "loss": 1.2723,
+      "step": 9653
+    },
+    {
+      "epoch": 8.912070159243019,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.566377585056705e-05,
+      "loss": 1.3374,
+      "step": 9654
+    },
+    {
+      "epoch": 8.912993307177475,
+      "grad_norm": 3.505859375,
+      "learning_rate": 3.5657104736491e-05,
+      "loss": 1.2617,
+      "step": 9655
+    },
+    {
+      "epoch": 8.91391645511193,
+      "grad_norm": 2.876953125,
+      "learning_rate": 3.5650433622414945e-05,
+      "loss": 1.0443,
+      "step": 9656
+    },
+    {
+      "epoch": 8.914839603046389,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.56437625083389e-05,
+      "loss": 1.21,
+      "step": 9657
+    },
+    {
+      "epoch": 8.915762750980845,
+      "grad_norm": 3.65234375,
+      "learning_rate": 3.563709139426284e-05,
+      "loss": 1.212,
+      "step": 9658
+    },
+    {
+      "epoch": 8.9166858989153,
+      "grad_norm": 3.193359375,
+      "learning_rate": 3.5630420280186795e-05,
+      "loss": 1.3658,
+      "step": 9659
+    },
+    {
+      "epoch": 8.917609046849758,
+      "grad_norm": 3.486328125,
+      "learning_rate": 3.562374916611074e-05,
+      "loss": 1.4106,
+      "step": 9660
+    },
+    {
+      "epoch": 8.918532194784214,
+      "grad_norm": 3.482421875,
+      "learning_rate": 3.561707805203469e-05,
+      "loss": 1.2476,
+      "step": 9661
+    },
+    {
+      "epoch": 8.91945534271867,
+      "grad_norm": 3.07421875,
+      "learning_rate": 3.561040693795864e-05,
+      "loss": 1.3972,
+      "step": 9662
+    },
+    {
+      "epoch": 8.920378490653126,
+      "grad_norm": 3.244140625,
+      "learning_rate": 3.560373582388259e-05,
+      "loss": 1.2606,
+      "step": 9663
+    },
+    {
+      "epoch": 8.921301638587584,
+      "grad_norm": 3.201171875,
+      "learning_rate": 3.5597064709806535e-05,
+      "loss": 1.3292,
+      "step": 9664
+    },
+    {
+      "epoch": 8.92222478652204,
+      "grad_norm": 4.12890625,
+      "learning_rate": 3.559039359573049e-05,
+      "loss": 1.3633,
+      "step": 9665
+    },
+    {
+      "epoch": 8.923147934456496,
+      "grad_norm": 2.98046875,
+      "learning_rate": 3.558372248165443e-05,
+      "loss": 1.4192,
+      "step": 9666
+    },
+    {
+      "epoch": 8.924071082390952,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.5577051367578385e-05,
+      "loss": 1.3293,
+      "step": 9667
+    },
+    {
+      "epoch": 8.92499423032541,
+      "grad_norm": 3.0703125,
+      "learning_rate": 3.557038025350234e-05,
+      "loss": 1.6226,
+      "step": 9668
+    },
+    {
+      "epoch": 8.925917378259866,
+      "grad_norm": 3.291015625,
+      "learning_rate": 3.556370913942629e-05,
+      "loss": 1.6292,
+      "step": 9669
+    },
+    {
+      "epoch": 8.926840526194322,
+      "grad_norm": 3.08984375,
+      "learning_rate": 3.5557038025350235e-05,
+      "loss": 1.6333,
+      "step": 9670
+    },
+    {
+      "epoch": 8.92776367412878,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.555036691127419e-05,
+      "loss": 1.2419,
+      "step": 9671
+    },
+    {
+      "epoch": 8.928686822063236,
+      "grad_norm": 3.158203125,
+      "learning_rate": 3.554369579719813e-05,
+      "loss": 1.3208,
+      "step": 9672
+    },
+    {
+      "epoch": 8.929609969997692,
+      "grad_norm": 3.14453125,
+      "learning_rate": 3.5537024683122085e-05,
+      "loss": 1.3,
+      "step": 9673
+    },
+    {
+      "epoch": 8.930533117932148,
+      "grad_norm": 3.1796875,
+      "learning_rate": 3.553035356904603e-05,
+      "loss": 1.4119,
+      "step": 9674
+    },
+    {
+      "epoch": 8.931456265866606,
+      "grad_norm": 3.525390625,
+      "learning_rate": 3.552368245496998e-05,
+      "loss": 1.634,
+      "step": 9675
+    },
+    {
+      "epoch": 8.932379413801062,
+      "grad_norm": 3.697265625,
+      "learning_rate": 3.551701134089393e-05,
+      "loss": 1.6277,
+      "step": 9676
+    },
+    {
+      "epoch": 8.933302561735518,
+      "grad_norm": 3.044921875,
+      "learning_rate": 3.551034022681788e-05,
+      "loss": 1.006,
+      "step": 9677
+    },
+    {
+      "epoch": 8.934225709669974,
+      "grad_norm": 2.94140625,
+      "learning_rate": 3.5503669112741826e-05,
+      "loss": 1.4524,
+      "step": 9678
+    },
+    {
+      "epoch": 8.935148857604432,
+      "grad_norm": 3.650390625,
+      "learning_rate": 3.549699799866578e-05,
+      "loss": 1.1855,
+      "step": 9679
+    },
+    {
+      "epoch": 8.936072005538888,
+      "grad_norm": 2.95703125,
+      "learning_rate": 3.549032688458973e-05,
+      "loss": 1.3284,
+      "step": 9680
+    },
+    {
+      "epoch": 8.936995153473344,
+      "grad_norm": 3.259765625,
+      "learning_rate": 3.548365577051368e-05,
+      "loss": 0.7054,
+      "step": 9681
+    },
+    {
+      "epoch": 8.9379183014078,
+      "grad_norm": 3.490234375,
+      "learning_rate": 3.547698465643763e-05,
+      "loss": 1.3889,
+      "step": 9682
+    },
+    {
+      "epoch": 8.938841449342258,
+      "grad_norm": 3.3671875,
+      "learning_rate": 3.547031354236158e-05,
+      "loss": 1.0891,
+      "step": 9683
+    },
+    {
+      "epoch": 8.939764597276714,
+      "grad_norm": 3.400390625,
+      "learning_rate": 3.5463642428285525e-05,
+      "loss": 1.3259,
+      "step": 9684
+    },
+    {
+      "epoch": 8.94068774521117,
+      "grad_norm": 3.62109375,
+      "learning_rate": 3.545697131420948e-05,
+      "loss": 1.2058,
+      "step": 9685
+    },
+    {
+      "epoch": 8.941610893145626,
+      "grad_norm": 2.986328125,
+      "learning_rate": 3.545030020013342e-05,
+      "loss": 1.3774,
+      "step": 9686
+    },
+    {
+      "epoch": 8.942534041080084,
+      "grad_norm": 3.306640625,
+      "learning_rate": 3.5443629086057375e-05,
+      "loss": 1.4124,
+      "step": 9687
+    },
+    {
+      "epoch": 8.94345718901454,
+      "grad_norm": 3.4921875,
+      "learning_rate": 3.543695797198132e-05,
+      "loss": 1.3147,
+      "step": 9688
+    },
+    {
+      "epoch": 8.944380336948996,
+      "grad_norm": 5.26171875,
+      "learning_rate": 3.543028685790527e-05,
+      "loss": 1.3425,
+      "step": 9689
+    },
+    {
+      "epoch": 8.945303484883453,
+      "grad_norm": 3.17578125,
+      "learning_rate": 3.542361574382922e-05,
+      "loss": 1.4473,
+      "step": 9690
+    },
+    {
+      "epoch": 8.94622663281791,
+      "grad_norm": 3.736328125,
+      "learning_rate": 3.541694462975317e-05,
+      "loss": 1.1156,
+      "step": 9691
+    },
+    {
+      "epoch": 8.947149780752365,
+      "grad_norm": 3.60546875,
+      "learning_rate": 3.5410273515677116e-05,
+      "loss": 1.3804,
+      "step": 9692
+    },
+    {
+      "epoch": 8.948072928686821,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.5403602401601075e-05,
+      "loss": 1.4219,
+      "step": 9693
+    },
+    {
+      "epoch": 8.94899607662128,
+      "grad_norm": 3.412109375,
+      "learning_rate": 3.539693128752502e-05,
+      "loss": 1.3499,
+      "step": 9694
+    },
+    {
+      "epoch": 8.949919224555735,
+      "grad_norm": 3.630859375,
+      "learning_rate": 3.539026017344897e-05,
+      "loss": 1.4167,
+      "step": 9695
+    },
+    {
+      "epoch": 8.950842372490191,
+      "grad_norm": 3.25390625,
+      "learning_rate": 3.538358905937292e-05,
+      "loss": 1.4346,
+      "step": 9696
+    },
+    {
+      "epoch": 8.951765520424647,
+      "grad_norm": 3.44921875,
+      "learning_rate": 3.537691794529687e-05,
+      "loss": 1.347,
+      "step": 9697
+    },
+    {
+      "epoch": 8.952688668359105,
+      "grad_norm": 4.26953125,
+      "learning_rate": 3.5370246831220815e-05,
+      "loss": 1.4136,
+      "step": 9698
+    },
+    {
+      "epoch": 8.953611816293561,
+      "grad_norm": 3.369140625,
+      "learning_rate": 3.536357571714477e-05,
+      "loss": 1.3555,
+      "step": 9699
+    },
+    {
+      "epoch": 8.954534964228017,
+      "grad_norm": 3.306640625,
+      "learning_rate": 3.535690460306871e-05,
+      "loss": 1.3132,
+      "step": 9700
+    },
+    {
+      "epoch": 8.955458112162473,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.5350233488992665e-05,
+      "loss": 0.9782,
+      "step": 9701
+    },
+    {
+      "epoch": 8.956381260096931,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.534356237491661e-05,
+      "loss": 1.1868,
+      "step": 9702
+    },
+    {
+      "epoch": 8.957304408031387,
+      "grad_norm": 3.529296875,
+      "learning_rate": 3.533689126084056e-05,
+      "loss": 1.4778,
+      "step": 9703
+    },
+    {
+      "epoch": 8.958227555965843,
+      "grad_norm": 3.310546875,
+      "learning_rate": 3.533022014676451e-05,
+      "loss": 1.2234,
+      "step": 9704
+    },
+    {
+      "epoch": 8.959150703900299,
+      "grad_norm": 3.66796875,
+      "learning_rate": 3.532354903268846e-05,
+      "loss": 1.3359,
+      "step": 9705
+    },
+    {
+      "epoch": 8.960073851834757,
+      "grad_norm": 2.580078125,
+      "learning_rate": 3.531687791861241e-05,
+      "loss": 1.0629,
+      "step": 9706
+    },
+    {
+      "epoch": 8.960996999769213,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.531020680453636e-05,
+      "loss": 1.3972,
+      "step": 9707
+    },
+    {
+      "epoch": 8.961920147703669,
+      "grad_norm": 3.279296875,
+      "learning_rate": 3.530353569046031e-05,
+      "loss": 1.3711,
+      "step": 9708
+    },
+    {
+      "epoch": 8.962843295638127,
+      "grad_norm": 3.529296875,
+      "learning_rate": 3.5296864576384256e-05,
+      "loss": 1.436,
+      "step": 9709
+    },
+    {
+      "epoch": 8.963766443572583,
+      "grad_norm": 3.212890625,
+      "learning_rate": 3.529019346230821e-05,
+      "loss": 1.0059,
+      "step": 9710
+    },
+    {
+      "epoch": 8.964689591507039,
+      "grad_norm": 3.2578125,
+      "learning_rate": 3.5283522348232154e-05,
+      "loss": 1.0505,
+      "step": 9711
+    },
+    {
+      "epoch": 8.965612739441495,
+      "grad_norm": 2.775390625,
+      "learning_rate": 3.5276851234156106e-05,
+      "loss": 1.3545,
+      "step": 9712
+    },
+    {
+      "epoch": 8.966535887375953,
+      "grad_norm": 3.166015625,
+      "learning_rate": 3.527018012008005e-05,
+      "loss": 0.9938,
+      "step": 9713
+    },
+    {
+      "epoch": 8.967459035310409,
+      "grad_norm": 2.716796875,
+      "learning_rate": 3.5263509006004e-05,
+      "loss": 1.4114,
+      "step": 9714
+    },
+    {
+      "epoch": 8.968382183244865,
+      "grad_norm": 2.853515625,
+      "learning_rate": 3.525683789192795e-05,
+      "loss": 1.1671,
+      "step": 9715
+    },
+    {
+      "epoch": 8.96930533117932,
+      "grad_norm": 3.779296875,
+      "learning_rate": 3.52501667778519e-05,
+      "loss": 1.1624,
+      "step": 9716
+    },
+    {
+      "epoch": 8.970228479113779,
+      "grad_norm": 3.892578125,
+      "learning_rate": 3.5243495663775846e-05,
+      "loss": 1.4868,
+      "step": 9717
+    },
+    {
+      "epoch": 8.971151627048235,
+      "grad_norm": 2.974609375,
+      "learning_rate": 3.5236824549699805e-05,
+      "loss": 1.3889,
+      "step": 9718
+    },
+    {
+      "epoch": 8.97207477498269,
+      "grad_norm": 3.876953125,
+      "learning_rate": 3.523015343562375e-05,
+      "loss": 1.0913,
+      "step": 9719
+    },
+    {
+      "epoch": 8.972997922917148,
+      "grad_norm": 2.974609375,
+      "learning_rate": 3.52234823215477e-05,
+      "loss": 1.1079,
+      "step": 9720
+    },
+    {
+      "epoch": 8.973921070851604,
+      "grad_norm": 3.6640625,
+      "learning_rate": 3.521681120747165e-05,
+      "loss": 1.1975,
+      "step": 9721
+    },
+    {
+      "epoch": 8.97484421878606,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.52101400933956e-05,
+      "loss": 1.2391,
+      "step": 9722
+    },
+    {
+      "epoch": 8.975767366720516,
+      "grad_norm": 2.701171875,
+      "learning_rate": 3.5203468979319546e-05,
+      "loss": 1.2803,
+      "step": 9723
+    },
+    {
+      "epoch": 8.976690514654974,
+      "grad_norm": 3.56640625,
+      "learning_rate": 3.51967978652435e-05,
+      "loss": 1.4437,
+      "step": 9724
+    },
+    {
+      "epoch": 8.97761366258943,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.5190126751167444e-05,
+      "loss": 1.479,
+      "step": 9725
+    },
+    {
+      "epoch": 8.978536810523886,
+      "grad_norm": 3.716796875,
+      "learning_rate": 3.5183455637091396e-05,
+      "loss": 1.4114,
+      "step": 9726
+    },
+    {
+      "epoch": 8.979459958458342,
+      "grad_norm": 3.279296875,
+      "learning_rate": 3.517678452301534e-05,
+      "loss": 1.0663,
+      "step": 9727
+    },
+    {
+      "epoch": 8.9803831063928,
+      "grad_norm": 3.248046875,
+      "learning_rate": 3.5170113408939294e-05,
+      "loss": 1.3644,
+      "step": 9728
+    },
+    {
+      "epoch": 8.981306254327256,
+      "grad_norm": 3.1640625,
+      "learning_rate": 3.516344229486324e-05,
+      "loss": 1.4927,
+      "step": 9729
+    },
+    {
+      "epoch": 8.982229402261712,
+      "grad_norm": 2.6484375,
+      "learning_rate": 3.51567711807872e-05,
+      "loss": 0.9014,
+      "step": 9730
+    },
+    {
+      "epoch": 8.983152550196168,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.5150100066711143e-05,
+      "loss": 0.869,
+      "step": 9731
+    },
+    {
+      "epoch": 8.984075698130626,
+      "grad_norm": 3.0859375,
+      "learning_rate": 3.5143428952635096e-05,
+      "loss": 0.8497,
+      "step": 9732
+    },
+    {
+      "epoch": 8.984998846065082,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.513675783855904e-05,
+      "loss": 1.1528,
+      "step": 9733
+    },
+    {
+      "epoch": 8.985921993999538,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.513008672448299e-05,
+      "loss": 1.3081,
+      "step": 9734
+    },
+    {
+      "epoch": 8.986845141933994,
+      "grad_norm": 3.033203125,
+      "learning_rate": 3.512341561040694e-05,
+      "loss": 1.405,
+      "step": 9735
+    },
+    {
+      "epoch": 8.987768289868452,
+      "grad_norm": 3.28515625,
+      "learning_rate": 3.511674449633089e-05,
+      "loss": 1.2081,
+      "step": 9736
+    },
+    {
+      "epoch": 8.988691437802908,
+      "grad_norm": 3.0859375,
+      "learning_rate": 3.5110073382254836e-05,
+      "loss": 1.2876,
+      "step": 9737
+    },
+    {
+      "epoch": 8.989614585737364,
+      "grad_norm": 3.64453125,
+      "learning_rate": 3.510340226817879e-05,
+      "loss": 1.3906,
+      "step": 9738
+    },
+    {
+      "epoch": 8.990537733671822,
+      "grad_norm": 3.76953125,
+      "learning_rate": 3.5096731154102734e-05,
+      "loss": 1.4032,
+      "step": 9739
+    },
+    {
+      "epoch": 8.991460881606278,
+      "grad_norm": 3.21484375,
+      "learning_rate": 3.5090060040026686e-05,
+      "loss": 1.0171,
+      "step": 9740
+    },
+    {
+      "epoch": 8.992384029540734,
+      "grad_norm": 3.2734375,
+      "learning_rate": 3.508338892595063e-05,
+      "loss": 1.3396,
+      "step": 9741
+    },
+    {
+      "epoch": 8.99330717747519,
+      "grad_norm": 3.48046875,
+      "learning_rate": 3.5076717811874584e-05,
+      "loss": 1.2943,
+      "step": 9742
+    },
+    {
+      "epoch": 8.994230325409648,
+      "grad_norm": 3.935546875,
+      "learning_rate": 3.5070046697798536e-05,
+      "loss": 1.1647,
+      "step": 9743
+    },
+    {
+      "epoch": 8.995153473344104,
+      "grad_norm": 2.857421875,
+      "learning_rate": 3.506337558372249e-05,
+      "loss": 1.3938,
+      "step": 9744
+    },
+    {
+      "epoch": 8.99607662127856,
+      "grad_norm": 3.572265625,
+      "learning_rate": 3.5056704469646434e-05,
+      "loss": 1.3542,
+      "step": 9745
+    },
+    {
+      "epoch": 8.996999769213016,
+      "grad_norm": 2.90234375,
+      "learning_rate": 3.5050033355570386e-05,
+      "loss": 1.2273,
+      "step": 9746
+    },
+    {
+      "epoch": 8.997922917147473,
+      "grad_norm": 3.5390625,
+      "learning_rate": 3.504336224149433e-05,
+      "loss": 1.4111,
+      "step": 9747
+    },
+    {
+      "epoch": 8.99884606508193,
+      "grad_norm": 3.13671875,
+      "learning_rate": 3.5036691127418284e-05,
+      "loss": 1.3672,
+      "step": 9748
+    },
+    {
+      "epoch": 8.999769213016386,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.503002001334223e-05,
+      "loss": 1.2236,
+      "step": 9749
+    },
+    {
+      "epoch": 9.000692360950842,
+      "grad_norm": 3.59765625,
+      "learning_rate": 3.502334889926618e-05,
+      "loss": 0.8718,
+      "step": 9750
+    },
+    {
+      "epoch": 9.0016155088853,
+      "grad_norm": 2.94921875,
+      "learning_rate": 3.5016677785190127e-05,
+      "loss": 1.0251,
+      "step": 9751
+    },
+    {
+      "epoch": 9.002538656819755,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.501000667111408e-05,
+      "loss": 1.2261,
+      "step": 9752
+    },
+    {
+      "epoch": 9.003461804754211,
+      "grad_norm": 3.1953125,
+      "learning_rate": 3.5003335557038024e-05,
+      "loss": 1.3271,
+      "step": 9753
+    },
+    {
+      "epoch": 9.00438495268867,
+      "grad_norm": 2.947265625,
+      "learning_rate": 3.4996664442961976e-05,
+      "loss": 1.2726,
+      "step": 9754
+    },
+    {
+      "epoch": 9.005308100623125,
+      "grad_norm": 3.767578125,
+      "learning_rate": 3.498999332888593e-05,
+      "loss": 1.2209,
+      "step": 9755
+    },
+    {
+      "epoch": 9.006231248557581,
+      "grad_norm": 3.486328125,
+      "learning_rate": 3.4983322214809874e-05,
+      "loss": 1.4116,
+      "step": 9756
+    },
+    {
+      "epoch": 9.007154396492037,
+      "grad_norm": 3.142578125,
+      "learning_rate": 3.4976651100733826e-05,
+      "loss": 1.2783,
+      "step": 9757
+    },
+    {
+      "epoch": 9.008077544426495,
+      "grad_norm": 3.166015625,
+      "learning_rate": 3.496997998665777e-05,
+      "loss": 1.4382,
+      "step": 9758
+    },
+    {
+      "epoch": 9.009000692360951,
+      "grad_norm": 2.892578125,
+      "learning_rate": 3.4963308872581724e-05,
+      "loss": 1.325,
+      "step": 9759
+    },
+    {
+      "epoch": 9.009923840295407,
+      "grad_norm": 3.416015625,
+      "learning_rate": 3.495663775850567e-05,
+      "loss": 1.5073,
+      "step": 9760
+    },
+    {
+      "epoch": 9.010846988229863,
+      "grad_norm": 3.26171875,
+      "learning_rate": 3.494996664442962e-05,
+      "loss": 1.0754,
+      "step": 9761
+    },
+    {
+      "epoch": 9.011770136164321,
+      "grad_norm": 3.705078125,
+      "learning_rate": 3.494329553035357e-05,
+      "loss": 1.4199,
+      "step": 9762
+    },
+    {
+      "epoch": 9.012693284098777,
+      "grad_norm": 3.48046875,
+      "learning_rate": 3.493662441627752e-05,
+      "loss": 1.428,
+      "step": 9763
+    },
+    {
+      "epoch": 9.013616432033233,
+      "grad_norm": 3.509765625,
+      "learning_rate": 3.4929953302201465e-05,
+      "loss": 1.1069,
+      "step": 9764
+    },
+    {
+      "epoch": 9.014539579967689,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.492328218812542e-05,
+      "loss": 1.199,
+      "step": 9765
+    },
+    {
+      "epoch": 9.015462727902147,
+      "grad_norm": 3.345703125,
+      "learning_rate": 3.491661107404937e-05,
+      "loss": 0.9938,
+      "step": 9766
+    },
+    {
+      "epoch": 9.016385875836603,
+      "grad_norm": 3.189453125,
+      "learning_rate": 3.4909939959973314e-05,
+      "loss": 1.2003,
+      "step": 9767
+    },
+    {
+      "epoch": 9.017309023771059,
+      "grad_norm": 3.279296875,
+      "learning_rate": 3.490326884589727e-05,
+      "loss": 1.26,
+      "step": 9768
+    },
+    {
+      "epoch": 9.018232171705515,
+      "grad_norm": 3.61328125,
+      "learning_rate": 3.489659773182122e-05,
+      "loss": 1.1166,
+      "step": 9769
+    },
+    {
+      "epoch": 9.019155319639973,
+      "grad_norm": 3.08203125,
+      "learning_rate": 3.4889926617745164e-05,
+      "loss": 1.4224,
+      "step": 9770
+    },
+    {
+      "epoch": 9.020078467574429,
+      "grad_norm": 4.08203125,
+      "learning_rate": 3.4883255503669117e-05,
+      "loss": 1.1372,
+      "step": 9771
+    },
+    {
+      "epoch": 9.021001615508885,
+      "grad_norm": 2.98046875,
+      "learning_rate": 3.487658438959306e-05,
+      "loss": 1.1732,
+      "step": 9772
+    },
+    {
+      "epoch": 9.021924763443343,
+      "grad_norm": 3.080078125,
+      "learning_rate": 3.4869913275517014e-05,
+      "loss": 1.3855,
+      "step": 9773
+    },
+    {
+      "epoch": 9.022847911377799,
+      "grad_norm": 3.6328125,
+      "learning_rate": 3.486324216144096e-05,
+      "loss": 1.3652,
+      "step": 9774
+    },
+    {
+      "epoch": 9.023771059312255,
+      "grad_norm": 3.1796875,
+      "learning_rate": 3.485657104736491e-05,
+      "loss": 1.3132,
+      "step": 9775
+    },
+    {
+      "epoch": 9.02469420724671,
+      "grad_norm": 3.33203125,
+      "learning_rate": 3.484989993328886e-05,
+      "loss": 1.3901,
+      "step": 9776
+    },
+    {
+      "epoch": 9.025617355181168,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.484322881921281e-05,
+      "loss": 1.2959,
+      "step": 9777
+    },
+    {
+      "epoch": 9.026540503115625,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.4836557705136755e-05,
+      "loss": 1.2814,
+      "step": 9778
+    },
+    {
+      "epoch": 9.02746365105008,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.482988659106071e-05,
+      "loss": 1.4939,
+      "step": 9779
+    },
+    {
+      "epoch": 9.028386798984537,
+      "grad_norm": 3.755859375,
+      "learning_rate": 3.482321547698466e-05,
+      "loss": 1.406,
+      "step": 9780
+    },
+    {
+      "epoch": 9.029309946918994,
+      "grad_norm": 3.1640625,
+      "learning_rate": 3.481654436290861e-05,
+      "loss": 1.193,
+      "step": 9781
+    },
+    {
+      "epoch": 9.03023309485345,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.480987324883256e-05,
+      "loss": 1.1359,
+      "step": 9782
+    },
+    {
+      "epoch": 9.031156242787906,
+      "grad_norm": 3.3203125,
+      "learning_rate": 3.480320213475651e-05,
+      "loss": 1.4775,
+      "step": 9783
+    },
+    {
+      "epoch": 9.032079390722362,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.4796531020680455e-05,
+      "loss": 1.3804,
+      "step": 9784
+    },
+    {
+      "epoch": 9.03300253865682,
+      "grad_norm": 3.1796875,
+      "learning_rate": 3.478985990660441e-05,
+      "loss": 1.3379,
+      "step": 9785
+    },
+    {
+      "epoch": 9.033925686591276,
+      "grad_norm": 3.0,
+      "learning_rate": 3.478318879252835e-05,
+      "loss": 1.3906,
+      "step": 9786
+    },
+    {
+      "epoch": 9.034848834525732,
+      "grad_norm": 3.376953125,
+      "learning_rate": 3.4776517678452304e-05,
+      "loss": 1.2697,
+      "step": 9787
+    },
+    {
+      "epoch": 9.03577198246019,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.476984656437625e-05,
+      "loss": 1.4243,
+      "step": 9788
+    },
+    {
+      "epoch": 9.036695130394646,
+      "grad_norm": 2.583984375,
+      "learning_rate": 3.47631754503002e-05,
+      "loss": 1.0687,
+      "step": 9789
+    },
+    {
+      "epoch": 9.037618278329102,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.475650433622415e-05,
+      "loss": 1.3911,
+      "step": 9790
+    },
+    {
+      "epoch": 9.038541426263558,
+      "grad_norm": 3.55859375,
+      "learning_rate": 3.47498332221481e-05,
+      "loss": 1.0894,
+      "step": 9791
+    },
+    {
+      "epoch": 9.039464574198016,
+      "grad_norm": 3.599609375,
+      "learning_rate": 3.4743162108072045e-05,
+      "loss": 1.2844,
+      "step": 9792
+    },
+    {
+      "epoch": 9.040387722132472,
+      "grad_norm": 3.1015625,
+      "learning_rate": 3.4736490993996004e-05,
+      "loss": 1.365,
+      "step": 9793
+    },
+    {
+      "epoch": 9.041310870066928,
+      "grad_norm": 3.759765625,
+      "learning_rate": 3.472981987991995e-05,
+      "loss": 1.3733,
+      "step": 9794
+    },
+    {
+      "epoch": 9.042234018001384,
+      "grad_norm": 3.2890625,
+      "learning_rate": 3.47231487658439e-05,
+      "loss": 1.3079,
+      "step": 9795
+    },
+    {
+      "epoch": 9.043157165935842,
+      "grad_norm": 3.47265625,
+      "learning_rate": 3.471647765176785e-05,
+      "loss": 1.325,
+      "step": 9796
+    },
+    {
+      "epoch": 9.044080313870298,
+      "grad_norm": 2.798828125,
+      "learning_rate": 3.47098065376918e-05,
+      "loss": 1.3538,
+      "step": 9797
+    },
+    {
+      "epoch": 9.045003461804754,
+      "grad_norm": 3.029296875,
+      "learning_rate": 3.4703135423615745e-05,
+      "loss": 1.2546,
+      "step": 9798
+    },
+    {
+      "epoch": 9.04592660973921,
+      "grad_norm": 3.189453125,
+      "learning_rate": 3.46964643095397e-05,
+      "loss": 1.2029,
+      "step": 9799
+    },
+    {
+      "epoch": 9.046849757673668,
+      "grad_norm": 3.302734375,
+      "learning_rate": 3.468979319546364e-05,
+      "loss": 1.2606,
+      "step": 9800
+    },
+    {
+      "epoch": 9.047772905608124,
+      "grad_norm": 3.53515625,
+      "learning_rate": 3.4683122081387595e-05,
+      "loss": 1.3452,
+      "step": 9801
+    },
+    {
+      "epoch": 9.04869605354258,
+      "grad_norm": 3.220703125,
+      "learning_rate": 3.467645096731154e-05,
+      "loss": 1.2455,
+      "step": 9802
+    },
+    {
+      "epoch": 9.049619201477038,
+      "grad_norm": 3.302734375,
+      "learning_rate": 3.466977985323549e-05,
+      "loss": 1.272,
+      "step": 9803
+    },
+    {
+      "epoch": 9.050542349411494,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.466310873915944e-05,
+      "loss": 1.132,
+      "step": 9804
+    },
+    {
+      "epoch": 9.05146549734595,
+      "grad_norm": 3.107421875,
+      "learning_rate": 3.46564376250834e-05,
+      "loss": 1.4531,
+      "step": 9805
+    },
+    {
+      "epoch": 9.052388645280406,
+      "grad_norm": 3.96484375,
+      "learning_rate": 3.464976651100734e-05,
+      "loss": 1.4368,
+      "step": 9806
+    },
+    {
+      "epoch": 9.053311793214863,
+      "grad_norm": 3.49609375,
+      "learning_rate": 3.4643095396931294e-05,
+      "loss": 1.2441,
+      "step": 9807
+    },
+    {
+      "epoch": 9.05423494114932,
+      "grad_norm": 3.5234375,
+      "learning_rate": 3.463642428285524e-05,
+      "loss": 1.3281,
+      "step": 9808
+    },
+    {
+      "epoch": 9.055158089083776,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.462975316877919e-05,
+      "loss": 1.2126,
+      "step": 9809
+    },
+    {
+      "epoch": 9.056081237018232,
+      "grad_norm": 3.076171875,
+      "learning_rate": 3.462308205470314e-05,
+      "loss": 1.2679,
+      "step": 9810
+    },
+    {
+      "epoch": 9.05700438495269,
+      "grad_norm": 3.513671875,
+      "learning_rate": 3.461641094062709e-05,
+      "loss": 1.1128,
+      "step": 9811
+    },
+    {
+      "epoch": 9.057927532887145,
+      "grad_norm": 2.7578125,
+      "learning_rate": 3.4609739826551035e-05,
+      "loss": 1.4426,
+      "step": 9812
+    },
+    {
+      "epoch": 9.058850680821601,
+      "grad_norm": 3.54296875,
+      "learning_rate": 3.460306871247499e-05,
+      "loss": 1.1086,
+      "step": 9813
+    },
+    {
+      "epoch": 9.059773828756057,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.459639759839893e-05,
+      "loss": 0.9858,
+      "step": 9814
+    },
+    {
+      "epoch": 9.060696976690515,
+      "grad_norm": 2.8671875,
+      "learning_rate": 3.4589726484322885e-05,
+      "loss": 1.0265,
+      "step": 9815
+    },
+    {
+      "epoch": 9.061620124624971,
+      "grad_norm": 2.61328125,
+      "learning_rate": 3.458305537024683e-05,
+      "loss": 1.1912,
+      "step": 9816
+    },
+    {
+      "epoch": 9.062543272559427,
+      "grad_norm": 3.150390625,
+      "learning_rate": 3.457638425617078e-05,
+      "loss": 1.4172,
+      "step": 9817
+    },
+    {
+      "epoch": 9.063466420493883,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.4569713142094735e-05,
+      "loss": 1.3779,
+      "step": 9818
+    },
+    {
+      "epoch": 9.064389568428341,
+      "grad_norm": 3.388671875,
+      "learning_rate": 3.456304202801868e-05,
+      "loss": 1.2491,
+      "step": 9819
+    },
+    {
+      "epoch": 9.065312716362797,
+      "grad_norm": 3.662109375,
+      "learning_rate": 3.455637091394263e-05,
+      "loss": 1.2418,
+      "step": 9820
+    },
+    {
+      "epoch": 9.066235864297253,
+      "grad_norm": 3.52734375,
+      "learning_rate": 3.454969979986658e-05,
+      "loss": 1.1799,
+      "step": 9821
+    },
+    {
+      "epoch": 9.067159012231711,
+      "grad_norm": 3.884765625,
+      "learning_rate": 3.454302868579053e-05,
+      "loss": 1.2777,
+      "step": 9822
+    },
+    {
+      "epoch": 9.068082160166167,
+      "grad_norm": 3.033203125,
+      "learning_rate": 3.4536357571714475e-05,
+      "loss": 1.3918,
+      "step": 9823
+    },
+    {
+      "epoch": 9.069005308100623,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.452968645763843e-05,
+      "loss": 1.4167,
+      "step": 9824
+    },
+    {
+      "epoch": 9.069928456035079,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.452301534356237e-05,
+      "loss": 1.4739,
+      "step": 9825
+    },
+    {
+      "epoch": 9.070851603969537,
+      "grad_norm": 3.392578125,
+      "learning_rate": 3.4516344229486325e-05,
+      "loss": 1.1415,
+      "step": 9826
+    },
+    {
+      "epoch": 9.071774751903993,
+      "grad_norm": 2.708984375,
+      "learning_rate": 3.450967311541027e-05,
+      "loss": 1.2468,
+      "step": 9827
+    },
+    {
+      "epoch": 9.072697899838449,
+      "grad_norm": 3.61328125,
+      "learning_rate": 3.450300200133422e-05,
+      "loss": 1.728,
+      "step": 9828
+    },
+    {
+      "epoch": 9.073621047772905,
+      "grad_norm": 3.171875,
+      "learning_rate": 3.449633088725817e-05,
+      "loss": 1.1692,
+      "step": 9829
+    },
+    {
+      "epoch": 9.074544195707363,
+      "grad_norm": 3.376953125,
+      "learning_rate": 3.448965977318213e-05,
+      "loss": 1.3096,
+      "step": 9830
+    },
+    {
+      "epoch": 9.075467343641819,
+      "grad_norm": 3.369140625,
+      "learning_rate": 3.448298865910607e-05,
+      "loss": 1.5205,
+      "step": 9831
+    },
+    {
+      "epoch": 9.076390491576275,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.4476317545030025e-05,
+      "loss": 1.3301,
+      "step": 9832
+    },
+    {
+      "epoch": 9.07731363951073,
+      "grad_norm": 3.349609375,
+      "learning_rate": 3.446964643095397e-05,
+      "loss": 1.2976,
+      "step": 9833
+    },
+    {
+      "epoch": 9.078236787445189,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.446297531687792e-05,
+      "loss": 1.407,
+      "step": 9834
+    },
+    {
+      "epoch": 9.079159935379645,
+      "grad_norm": 3.6640625,
+      "learning_rate": 3.445630420280187e-05,
+      "loss": 1.5105,
+      "step": 9835
+    },
+    {
+      "epoch": 9.0800830833141,
+      "grad_norm": 3.384765625,
+      "learning_rate": 3.444963308872582e-05,
+      "loss": 1.3262,
+      "step": 9836
+    },
+    {
+      "epoch": 9.081006231248558,
+      "grad_norm": 3.439453125,
+      "learning_rate": 3.4442961974649766e-05,
+      "loss": 1.2891,
+      "step": 9837
+    },
+    {
+      "epoch": 9.081929379183014,
+      "grad_norm": 3.4296875,
+      "learning_rate": 3.443629086057372e-05,
+      "loss": 1.5071,
+      "step": 9838
+    },
+    {
+      "epoch": 9.08285252711747,
+      "grad_norm": 3.72265625,
+      "learning_rate": 3.442961974649766e-05,
+      "loss": 1.2158,
+      "step": 9839
+    },
+    {
+      "epoch": 9.083775675051927,
+      "grad_norm": 3.5390625,
+      "learning_rate": 3.4422948632421615e-05,
+      "loss": 1.1993,
+      "step": 9840
+    },
+    {
+      "epoch": 9.084698822986384,
+      "grad_norm": 3.2578125,
+      "learning_rate": 3.441627751834556e-05,
+      "loss": 1.3875,
+      "step": 9841
+    },
+    {
+      "epoch": 9.08562197092084,
+      "grad_norm": 3.09765625,
+      "learning_rate": 3.440960640426951e-05,
+      "loss": 1.3826,
+      "step": 9842
+    },
+    {
+      "epoch": 9.086545118855296,
+      "grad_norm": 3.490234375,
+      "learning_rate": 3.4402935290193465e-05,
+      "loss": 1.1475,
+      "step": 9843
+    },
+    {
+      "epoch": 9.087468266789752,
+      "grad_norm": 3.1328125,
+      "learning_rate": 3.439626417611742e-05,
+      "loss": 1.3337,
+      "step": 9844
+    },
+    {
+      "epoch": 9.08839141472421,
+      "grad_norm": 3.5546875,
+      "learning_rate": 3.438959306204136e-05,
+      "loss": 1.1008,
+      "step": 9845
+    },
+    {
+      "epoch": 9.089314562658666,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.4382921947965315e-05,
+      "loss": 1.3483,
+      "step": 9846
+    },
+    {
+      "epoch": 9.090237710593122,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.437625083388926e-05,
+      "loss": 1.1882,
+      "step": 9847
+    },
+    {
+      "epoch": 9.091160858527578,
+      "grad_norm": 3.26953125,
+      "learning_rate": 3.436957971981321e-05,
+      "loss": 1.2178,
+      "step": 9848
+    },
+    {
+      "epoch": 9.092084006462036,
+      "grad_norm": 3.240234375,
+      "learning_rate": 3.436290860573716e-05,
+      "loss": 1.356,
+      "step": 9849
+    },
+    {
+      "epoch": 9.093007154396492,
+      "grad_norm": 3.04296875,
+      "learning_rate": 3.435623749166111e-05,
+      "loss": 1.4812,
+      "step": 9850
+    },
+    {
+      "epoch": 9.093930302330948,
+      "grad_norm": 3.470703125,
+      "learning_rate": 3.4349566377585056e-05,
+      "loss": 1.4487,
+      "step": 9851
+    },
+    {
+      "epoch": 9.094853450265404,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.434289526350901e-05,
+      "loss": 1.2761,
+      "step": 9852
+    },
+    {
+      "epoch": 9.095776598199862,
+      "grad_norm": 3.07421875,
+      "learning_rate": 3.4336224149432954e-05,
+      "loss": 1.4177,
+      "step": 9853
+    },
+    {
+      "epoch": 9.096699746134318,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.4329553035356906e-05,
+      "loss": 1.0283,
+      "step": 9854
+    },
+    {
+      "epoch": 9.097622894068774,
+      "grad_norm": 4.00390625,
+      "learning_rate": 3.432288192128085e-05,
+      "loss": 1.3179,
+      "step": 9855
+    },
+    {
+      "epoch": 9.098546042003232,
+      "grad_norm": 3.708984375,
+      "learning_rate": 3.431621080720481e-05,
+      "loss": 1.3545,
+      "step": 9856
+    },
+    {
+      "epoch": 9.099469189937688,
+      "grad_norm": 3.06640625,
+      "learning_rate": 3.4309539693128756e-05,
+      "loss": 1.2144,
+      "step": 9857
+    },
+    {
+      "epoch": 9.100392337872144,
+      "grad_norm": 3.88671875,
+      "learning_rate": 3.430286857905271e-05,
+      "loss": 1.1766,
+      "step": 9858
+    },
+    {
+      "epoch": 9.1013154858066,
+      "grad_norm": 2.99609375,
+      "learning_rate": 3.429619746497665e-05,
+      "loss": 1.3632,
+      "step": 9859
+    },
+    {
+      "epoch": 9.102238633741058,
+      "grad_norm": 3.955078125,
+      "learning_rate": 3.4289526350900605e-05,
+      "loss": 1.3394,
+      "step": 9860
+    },
+    {
+      "epoch": 9.103161781675514,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.428285523682455e-05,
+      "loss": 0.9429,
+      "step": 9861
+    },
+    {
+      "epoch": 9.10408492960997,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.42761841227485e-05,
+      "loss": 1.2296,
+      "step": 9862
+    },
+    {
+      "epoch": 9.105008077544426,
+      "grad_norm": 3.166015625,
+      "learning_rate": 3.426951300867245e-05,
+      "loss": 1.2321,
+      "step": 9863
+    },
+    {
+      "epoch": 9.105931225478884,
+      "grad_norm": 3.0703125,
+      "learning_rate": 3.42628418945964e-05,
+      "loss": 1.3655,
+      "step": 9864
+    },
+    {
+      "epoch": 9.10685437341334,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.4256170780520346e-05,
+      "loss": 1.3235,
+      "step": 9865
+    },
+    {
+      "epoch": 9.107777521347796,
+      "grad_norm": 3.650390625,
+      "learning_rate": 3.42494996664443e-05,
+      "loss": 1.4318,
+      "step": 9866
+    },
+    {
+      "epoch": 9.108700669282252,
+      "grad_norm": 3.10546875,
+      "learning_rate": 3.4242828552368244e-05,
+      "loss": 1.2595,
+      "step": 9867
+    },
+    {
+      "epoch": 9.10962381721671,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.4236157438292196e-05,
+      "loss": 1.2837,
+      "step": 9868
+    },
+    {
+      "epoch": 9.110546965151165,
+      "grad_norm": 3.314453125,
+      "learning_rate": 3.422948632421615e-05,
+      "loss": 1.229,
+      "step": 9869
+    },
+    {
+      "epoch": 9.111470113085621,
+      "grad_norm": 3.962890625,
+      "learning_rate": 3.4222815210140094e-05,
+      "loss": 1.5282,
+      "step": 9870
+    },
+    {
+      "epoch": 9.11239326102008,
+      "grad_norm": 2.951171875,
+      "learning_rate": 3.4216144096064046e-05,
+      "loss": 1.3616,
+      "step": 9871
+    },
+    {
+      "epoch": 9.113316408954535,
+      "grad_norm": 3.19921875,
+      "learning_rate": 3.420947298198799e-05,
+      "loss": 1.425,
+      "step": 9872
+    },
+    {
+      "epoch": 9.114239556888991,
+      "grad_norm": 3.73828125,
+      "learning_rate": 3.4202801867911943e-05,
+      "loss": 1.0688,
+      "step": 9873
+    },
+    {
+      "epoch": 9.115162704823447,
+      "grad_norm": 2.533203125,
+      "learning_rate": 3.419613075383589e-05,
+      "loss": 0.9504,
+      "step": 9874
+    },
+    {
+      "epoch": 9.116085852757905,
+      "grad_norm": 3.251953125,
+      "learning_rate": 3.418945963975984e-05,
+      "loss": 1.2445,
+      "step": 9875
+    },
+    {
+      "epoch": 9.117009000692361,
+      "grad_norm": 3.41796875,
+      "learning_rate": 3.4182788525683787e-05,
+      "loss": 1.3245,
+      "step": 9876
+    },
+    {
+      "epoch": 9.117932148626817,
+      "grad_norm": 3.615234375,
+      "learning_rate": 3.417611741160774e-05,
+      "loss": 1.4077,
+      "step": 9877
+    },
+    {
+      "epoch": 9.118855296561273,
+      "grad_norm": 3.134765625,
+      "learning_rate": 3.416944629753169e-05,
+      "loss": 1.4026,
+      "step": 9878
+    },
+    {
+      "epoch": 9.119778444495731,
+      "grad_norm": 3.6171875,
+      "learning_rate": 3.4162775183455636e-05,
+      "loss": 1.1953,
+      "step": 9879
+    },
+    {
+      "epoch": 9.120701592430187,
+      "grad_norm": 3.5859375,
+      "learning_rate": 3.415610406937959e-05,
+      "loss": 1.4214,
+      "step": 9880
+    },
+    {
+      "epoch": 9.121624740364643,
+      "grad_norm": 3.5703125,
+      "learning_rate": 3.414943295530354e-05,
+      "loss": 1.2892,
+      "step": 9881
+    },
+    {
+      "epoch": 9.1225478882991,
+      "grad_norm": 3.05859375,
+      "learning_rate": 3.4142761841227486e-05,
+      "loss": 1.4661,
+      "step": 9882
+    },
+    {
+      "epoch": 9.123471036233557,
+      "grad_norm": 3.06640625,
+      "learning_rate": 3.413609072715144e-05,
+      "loss": 1.2935,
+      "step": 9883
+    },
+    {
+      "epoch": 9.124394184168013,
+      "grad_norm": 3.26953125,
+      "learning_rate": 3.4129419613075384e-05,
+      "loss": 1.3425,
+      "step": 9884
+    },
+    {
+      "epoch": 9.125317332102469,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.4122748498999336e-05,
+      "loss": 1.2883,
+      "step": 9885
+    },
+    {
+      "epoch": 9.126240480036927,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.411607738492328e-05,
+      "loss": 1.4929,
+      "step": 9886
+    },
+    {
+      "epoch": 9.127163627971383,
+      "grad_norm": 3.22265625,
+      "learning_rate": 3.4109406270847234e-05,
+      "loss": 1.2358,
+      "step": 9887
+    },
+    {
+      "epoch": 9.128086775905839,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.410273515677118e-05,
+      "loss": 1.3755,
+      "step": 9888
+    },
+    {
+      "epoch": 9.129009923840295,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.409606404269513e-05,
+      "loss": 1.3406,
+      "step": 9889
+    },
+    {
+      "epoch": 9.129933071774753,
+      "grad_norm": 3.29296875,
+      "learning_rate": 3.408939292861908e-05,
+      "loss": 1.2308,
+      "step": 9890
+    },
+    {
+      "epoch": 9.130856219709209,
+      "grad_norm": 3.56640625,
+      "learning_rate": 3.408272181454303e-05,
+      "loss": 1.2799,
+      "step": 9891
+    },
+    {
+      "epoch": 9.131779367643665,
+      "grad_norm": 3.357421875,
+      "learning_rate": 3.4076050700466974e-05,
+      "loss": 1.4102,
+      "step": 9892
+    },
+    {
+      "epoch": 9.13270251557812,
+      "grad_norm": 3.462890625,
+      "learning_rate": 3.406937958639093e-05,
+      "loss": 1.356,
+      "step": 9893
+    },
+    {
+      "epoch": 9.133625663512579,
+      "grad_norm": 3.439453125,
+      "learning_rate": 3.406270847231488e-05,
+      "loss": 1.0419,
+      "step": 9894
+    },
+    {
+      "epoch": 9.134548811447035,
+      "grad_norm": 3.3046875,
+      "learning_rate": 3.405603735823883e-05,
+      "loss": 1.4814,
+      "step": 9895
+    },
+    {
+      "epoch": 9.13547195938149,
+      "grad_norm": 3.18359375,
+      "learning_rate": 3.4049366244162776e-05,
+      "loss": 1.2479,
+      "step": 9896
+    },
+    {
+      "epoch": 9.136395107315947,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.404269513008673e-05,
+      "loss": 0.8142,
+      "step": 9897
+    },
+    {
+      "epoch": 9.137318255250404,
+      "grad_norm": 3.142578125,
+      "learning_rate": 3.4036024016010674e-05,
+      "loss": 0.9505,
+      "step": 9898
+    },
+    {
+      "epoch": 9.13824140318486,
+      "grad_norm": 3.080078125,
+      "learning_rate": 3.4029352901934626e-05,
+      "loss": 1.5339,
+      "step": 9899
+    },
+    {
+      "epoch": 9.139164551119316,
+      "grad_norm": 3.634765625,
+      "learning_rate": 3.402268178785857e-05,
+      "loss": 1.3126,
+      "step": 9900
+    },
+    {
+      "epoch": 9.140087699053772,
+      "grad_norm": 3.732421875,
+      "learning_rate": 3.4016010673782524e-05,
+      "loss": 1.1025,
+      "step": 9901
+    },
+    {
+      "epoch": 9.14101084698823,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.400933955970647e-05,
+      "loss": 1.1604,
+      "step": 9902
+    },
+    {
+      "epoch": 9.141933994922686,
+      "grad_norm": 3.23828125,
+      "learning_rate": 3.400266844563042e-05,
+      "loss": 1.4446,
+      "step": 9903
+    },
+    {
+      "epoch": 9.142857142857142,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.399599733155437e-05,
+      "loss": 0.9368,
+      "step": 9904
+    },
+    {
+      "epoch": 9.1437802907916,
+      "grad_norm": 4.296875,
+      "learning_rate": 3.398932621747832e-05,
+      "loss": 1.4446,
+      "step": 9905
+    },
+    {
+      "epoch": 9.144703438726056,
+      "grad_norm": 3.498046875,
+      "learning_rate": 3.398265510340227e-05,
+      "loss": 1.3196,
+      "step": 9906
+    },
+    {
+      "epoch": 9.145626586660512,
+      "grad_norm": 2.861328125,
+      "learning_rate": 3.3975983989326224e-05,
+      "loss": 1.467,
+      "step": 9907
+    },
+    {
+      "epoch": 9.146549734594968,
+      "grad_norm": 3.75390625,
+      "learning_rate": 3.396931287525017e-05,
+      "loss": 1.1243,
+      "step": 9908
+    },
+    {
+      "epoch": 9.147472882529426,
+      "grad_norm": 3.162109375,
+      "learning_rate": 3.396264176117412e-05,
+      "loss": 1.29,
+      "step": 9909
+    },
+    {
+      "epoch": 9.148396030463882,
+      "grad_norm": 3.595703125,
+      "learning_rate": 3.395597064709807e-05,
+      "loss": 1.4148,
+      "step": 9910
+    },
+    {
+      "epoch": 9.149319178398338,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.394929953302202e-05,
+      "loss": 1.6472,
+      "step": 9911
+    },
+    {
+      "epoch": 9.150242326332794,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.3942628418945964e-05,
+      "loss": 1.2781,
+      "step": 9912
+    },
+    {
+      "epoch": 9.151165474267252,
+      "grad_norm": 3.375,
+      "learning_rate": 3.3935957304869917e-05,
+      "loss": 1.2808,
+      "step": 9913
+    },
+    {
+      "epoch": 9.152088622201708,
+      "grad_norm": 3.63671875,
+      "learning_rate": 3.392928619079386e-05,
+      "loss": 1.3696,
+      "step": 9914
+    },
+    {
+      "epoch": 9.153011770136164,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.3922615076717814e-05,
+      "loss": 1.1093,
+      "step": 9915
+    },
+    {
+      "epoch": 9.15393491807062,
+      "grad_norm": 3.162109375,
+      "learning_rate": 3.391594396264176e-05,
+      "loss": 1.2498,
+      "step": 9916
+    },
+    {
+      "epoch": 9.154858066005078,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.390927284856571e-05,
+      "loss": 1.1561,
+      "step": 9917
+    },
+    {
+      "epoch": 9.155781213939534,
+      "grad_norm": 3.083984375,
+      "learning_rate": 3.3902601734489664e-05,
+      "loss": 0.6537,
+      "step": 9918
+    },
+    {
+      "epoch": 9.15670436187399,
+      "grad_norm": 3.330078125,
+      "learning_rate": 3.3895930620413616e-05,
+      "loss": 1.1622,
+      "step": 9919
+    },
+    {
+      "epoch": 9.157627509808448,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.388925950633756e-05,
+      "loss": 1.2393,
+      "step": 9920
+    },
+    {
+      "epoch": 9.158550657742904,
+      "grad_norm": 3.283203125,
+      "learning_rate": 3.3882588392261514e-05,
+      "loss": 1.3354,
+      "step": 9921
+    },
+    {
+      "epoch": 9.15947380567736,
+      "grad_norm": 3.591796875,
+      "learning_rate": 3.387591727818546e-05,
+      "loss": 1.1743,
+      "step": 9922
+    },
+    {
+      "epoch": 9.160396953611816,
+      "grad_norm": 3.119140625,
+      "learning_rate": 3.386924616410941e-05,
+      "loss": 1.3083,
+      "step": 9923
+    },
+    {
+      "epoch": 9.161320101546274,
+      "grad_norm": 3.697265625,
+      "learning_rate": 3.386257505003336e-05,
+      "loss": 1.4617,
+      "step": 9924
+    },
+    {
+      "epoch": 9.16224324948073,
+      "grad_norm": 3.27734375,
+      "learning_rate": 3.385590393595731e-05,
+      "loss": 1.1608,
+      "step": 9925
+    },
+    {
+      "epoch": 9.163166397415186,
+      "grad_norm": 3.201171875,
+      "learning_rate": 3.3849232821881255e-05,
+      "loss": 1.2032,
+      "step": 9926
+    },
+    {
+      "epoch": 9.164089545349642,
+      "grad_norm": 3.78515625,
+      "learning_rate": 3.384256170780521e-05,
+      "loss": 1.189,
+      "step": 9927
+    },
+    {
+      "epoch": 9.1650126932841,
+      "grad_norm": 2.962890625,
+      "learning_rate": 3.383589059372915e-05,
+      "loss": 1.0826,
+      "step": 9928
+    },
+    {
+      "epoch": 9.165935841218555,
+      "grad_norm": 3.126953125,
+      "learning_rate": 3.3829219479653104e-05,
+      "loss": 1.3962,
+      "step": 9929
+    },
+    {
+      "epoch": 9.166858989153011,
+      "grad_norm": 3.21484375,
+      "learning_rate": 3.382254836557705e-05,
+      "loss": 1.4363,
+      "step": 9930
+    },
+    {
+      "epoch": 9.167782137087467,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.3815877251501e-05,
+      "loss": 1.4656,
+      "step": 9931
+    },
+    {
+      "epoch": 9.168705285021925,
+      "grad_norm": 3.56640625,
+      "learning_rate": 3.3809206137424954e-05,
+      "loss": 1.1129,
+      "step": 9932
+    },
+    {
+      "epoch": 9.169628432956381,
+      "grad_norm": 3.107421875,
+      "learning_rate": 3.38025350233489e-05,
+      "loss": 1.2817,
+      "step": 9933
+    },
+    {
+      "epoch": 9.170551580890837,
+      "grad_norm": 2.716796875,
+      "learning_rate": 3.379586390927285e-05,
+      "loss": 0.9811,
+      "step": 9934
+    },
+    {
+      "epoch": 9.171474728825295,
+      "grad_norm": 3.7109375,
+      "learning_rate": 3.37891927951968e-05,
+      "loss": 1.0991,
+      "step": 9935
+    },
+    {
+      "epoch": 9.172397876759751,
+      "grad_norm": 3.240234375,
+      "learning_rate": 3.378252168112075e-05,
+      "loss": 1.2112,
+      "step": 9936
+    },
+    {
+      "epoch": 9.173321024694207,
+      "grad_norm": 3.505859375,
+      "learning_rate": 3.3775850567044695e-05,
+      "loss": 1.218,
+      "step": 9937
+    },
+    {
+      "epoch": 9.174244172628663,
+      "grad_norm": 3.177734375,
+      "learning_rate": 3.376917945296865e-05,
+      "loss": 1.3132,
+      "step": 9938
+    },
+    {
+      "epoch": 9.175167320563121,
+      "grad_norm": 3.15234375,
+      "learning_rate": 3.376250833889259e-05,
+      "loss": 1.321,
+      "step": 9939
+    },
+    {
+      "epoch": 9.176090468497577,
+      "grad_norm": 2.861328125,
+      "learning_rate": 3.3755837224816545e-05,
+      "loss": 1.4854,
+      "step": 9940
+    },
+    {
+      "epoch": 9.177013616432033,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.374916611074049e-05,
+      "loss": 1.1768,
+      "step": 9941
+    },
+    {
+      "epoch": 9.177936764366489,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.374249499666444e-05,
+      "loss": 1.0935,
+      "step": 9942
+    },
+    {
+      "epoch": 9.178859912300947,
+      "grad_norm": 2.978515625,
+      "learning_rate": 3.3735823882588395e-05,
+      "loss": 1.1324,
+      "step": 9943
+    },
+    {
+      "epoch": 9.179783060235403,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.372915276851235e-05,
+      "loss": 1.3477,
+      "step": 9944
+    },
+    {
+      "epoch": 9.180706208169859,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.372248165443629e-05,
+      "loss": 1.4268,
+      "step": 9945
+    },
+    {
+      "epoch": 9.181629356104315,
+      "grad_norm": 3.37109375,
+      "learning_rate": 3.3715810540360244e-05,
+      "loss": 1.3308,
+      "step": 9946
+    },
+    {
+      "epoch": 9.182552504038773,
+      "grad_norm": 3.017578125,
+      "learning_rate": 3.370913942628419e-05,
+      "loss": 1.4075,
+      "step": 9947
+    },
+    {
+      "epoch": 9.183475651973229,
+      "grad_norm": 3.3359375,
+      "learning_rate": 3.370246831220814e-05,
+      "loss": 1.1898,
+      "step": 9948
+    },
+    {
+      "epoch": 9.184398799907685,
+      "grad_norm": 2.994140625,
+      "learning_rate": 3.369579719813209e-05,
+      "loss": 1.4038,
+      "step": 9949
+    },
+    {
+      "epoch": 9.18532194784214,
+      "grad_norm": 3.642578125,
+      "learning_rate": 3.368912608405604e-05,
+      "loss": 1.436,
+      "step": 9950
+    },
+    {
+      "epoch": 9.186245095776599,
+      "grad_norm": 3.146484375,
+      "learning_rate": 3.3682454969979985e-05,
+      "loss": 1.2939,
+      "step": 9951
+    },
+    {
+      "epoch": 9.187168243711055,
+      "grad_norm": 4.30859375,
+      "learning_rate": 3.367578385590394e-05,
+      "loss": 1.1447,
+      "step": 9952
+    },
+    {
+      "epoch": 9.18809139164551,
+      "grad_norm": 3.712890625,
+      "learning_rate": 3.366911274182788e-05,
+      "loss": 1.3687,
+      "step": 9953
+    },
+    {
+      "epoch": 9.189014539579968,
+      "grad_norm": 3.498046875,
+      "learning_rate": 3.3662441627751835e-05,
+      "loss": 1.1479,
+      "step": 9954
+    },
+    {
+      "epoch": 9.189937687514425,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.365577051367578e-05,
+      "loss": 1.2328,
+      "step": 9955
+    },
+    {
+      "epoch": 9.19086083544888,
+      "grad_norm": 3.4453125,
+      "learning_rate": 3.364909939959974e-05,
+      "loss": 1.0538,
+      "step": 9956
+    },
+    {
+      "epoch": 9.191783983383337,
+      "grad_norm": 3.041015625,
+      "learning_rate": 3.3642428285523685e-05,
+      "loss": 1.5715,
+      "step": 9957
+    },
+    {
+      "epoch": 9.192707131317794,
+      "grad_norm": 2.8359375,
+      "learning_rate": 3.363575717144764e-05,
+      "loss": 1.1743,
+      "step": 9958
+    },
+    {
+      "epoch": 9.19363027925225,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.362908605737158e-05,
+      "loss": 1.1184,
+      "step": 9959
+    },
+    {
+      "epoch": 9.194553427186706,
+      "grad_norm": 3.365234375,
+      "learning_rate": 3.3622414943295535e-05,
+      "loss": 1.0542,
+      "step": 9960
+    },
+    {
+      "epoch": 9.195476575121162,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.361574382921948e-05,
+      "loss": 1.0345,
+      "step": 9961
+    },
+    {
+      "epoch": 9.19639972305562,
+      "grad_norm": 3.01171875,
+      "learning_rate": 3.360907271514343e-05,
+      "loss": 1.4595,
+      "step": 9962
+    },
+    {
+      "epoch": 9.197322870990076,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.360240160106738e-05,
+      "loss": 1.2147,
+      "step": 9963
+    },
+    {
+      "epoch": 9.198246018924532,
+      "grad_norm": 3.05078125,
+      "learning_rate": 3.359573048699133e-05,
+      "loss": 1.1255,
+      "step": 9964
+    },
+    {
+      "epoch": 9.199169166858988,
+      "grad_norm": 3.384765625,
+      "learning_rate": 3.3589059372915275e-05,
+      "loss": 0.988,
+      "step": 9965
+    },
+    {
+      "epoch": 9.200092314793446,
+      "grad_norm": 3.951171875,
+      "learning_rate": 3.358238825883923e-05,
+      "loss": 1.2812,
+      "step": 9966
+    },
+    {
+      "epoch": 9.201015462727902,
+      "grad_norm": 3.29296875,
+      "learning_rate": 3.357571714476317e-05,
+      "loss": 1.2887,
+      "step": 9967
+    },
+    {
+      "epoch": 9.201938610662358,
+      "grad_norm": 2.998046875,
+      "learning_rate": 3.356904603068713e-05,
+      "loss": 1.2146,
+      "step": 9968
+    },
+    {
+      "epoch": 9.202861758596816,
+      "grad_norm": 3.44921875,
+      "learning_rate": 3.356237491661108e-05,
+      "loss": 1.1975,
+      "step": 9969
+    },
+    {
+      "epoch": 9.203784906531272,
+      "grad_norm": 3.458984375,
+      "learning_rate": 3.355570380253503e-05,
+      "loss": 1.3569,
+      "step": 9970
+    },
+    {
+      "epoch": 9.204708054465728,
+      "grad_norm": 3.736328125,
+      "learning_rate": 3.3549032688458975e-05,
+      "loss": 1.4156,
+      "step": 9971
+    },
+    {
+      "epoch": 9.205631202400184,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.354236157438293e-05,
+      "loss": 1.3256,
+      "step": 9972
+    },
+    {
+      "epoch": 9.206554350334642,
+      "grad_norm": 3.29296875,
+      "learning_rate": 3.353569046030687e-05,
+      "loss": 1.1895,
+      "step": 9973
+    },
+    {
+      "epoch": 9.207477498269098,
+      "grad_norm": 3.755859375,
+      "learning_rate": 3.3529019346230825e-05,
+      "loss": 1.3245,
+      "step": 9974
+    },
+    {
+      "epoch": 9.208400646203554,
+      "grad_norm": 4.00390625,
+      "learning_rate": 3.352234823215477e-05,
+      "loss": 1.2129,
+      "step": 9975
+    },
+    {
+      "epoch": 9.20932379413801,
+      "grad_norm": 2.99609375,
+      "learning_rate": 3.351567711807872e-05,
+      "loss": 0.9438,
+      "step": 9976
+    },
+    {
+      "epoch": 9.210246942072468,
+      "grad_norm": 2.9765625,
+      "learning_rate": 3.350900600400267e-05,
+      "loss": 1.2753,
+      "step": 9977
+    },
+    {
+      "epoch": 9.211170090006924,
+      "grad_norm": 3.48046875,
+      "learning_rate": 3.350233488992662e-05,
+      "loss": 1.4443,
+      "step": 9978
+    },
+    {
+      "epoch": 9.21209323794138,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.3495663775850566e-05,
+      "loss": 1.4714,
+      "step": 9979
+    },
+    {
+      "epoch": 9.213016385875836,
+      "grad_norm": 2.896484375,
+      "learning_rate": 3.348899266177452e-05,
+      "loss": 0.9738,
+      "step": 9980
+    },
+    {
+      "epoch": 9.213939533810294,
+      "grad_norm": 2.900390625,
+      "learning_rate": 3.348232154769847e-05,
+      "loss": 1.2893,
+      "step": 9981
+    },
+    {
+      "epoch": 9.21486268174475,
+      "grad_norm": 3.43359375,
+      "learning_rate": 3.3475650433622416e-05,
+      "loss": 1.2676,
+      "step": 9982
+    },
+    {
+      "epoch": 9.215785829679206,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.346897931954637e-05,
+      "loss": 1.2375,
+      "step": 9983
+    },
+    {
+      "epoch": 9.216708977613663,
+      "grad_norm": 3.20703125,
+      "learning_rate": 3.346230820547031e-05,
+      "loss": 1.3506,
+      "step": 9984
+    },
+    {
+      "epoch": 9.21763212554812,
+      "grad_norm": 3.037109375,
+      "learning_rate": 3.3455637091394265e-05,
+      "loss": 1.3247,
+      "step": 9985
+    },
+    {
+      "epoch": 9.218555273482576,
+      "grad_norm": 3.50390625,
+      "learning_rate": 3.344896597731821e-05,
+      "loss": 1.2966,
+      "step": 9986
+    },
+    {
+      "epoch": 9.219478421417032,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.344229486324216e-05,
+      "loss": 1.0785,
+      "step": 9987
+    },
+    {
+      "epoch": 9.22040156935149,
+      "grad_norm": 3.388671875,
+      "learning_rate": 3.343562374916611e-05,
+      "loss": 1.1489,
+      "step": 9988
+    },
+    {
+      "epoch": 9.221324717285945,
+      "grad_norm": 3.787109375,
+      "learning_rate": 3.342895263509006e-05,
+      "loss": 1.1963,
+      "step": 9989
+    },
+    {
+      "epoch": 9.222247865220401,
+      "grad_norm": 2.974609375,
+      "learning_rate": 3.342228152101401e-05,
+      "loss": 1.238,
+      "step": 9990
+    },
+    {
+      "epoch": 9.223171013154857,
+      "grad_norm": 3.88671875,
+      "learning_rate": 3.341561040693796e-05,
+      "loss": 1.374,
+      "step": 9991
+    },
+    {
+      "epoch": 9.224094161089315,
+      "grad_norm": 3.197265625,
+      "learning_rate": 3.340893929286191e-05,
+      "loss": 1.4858,
+      "step": 9992
+    },
+    {
+      "epoch": 9.225017309023771,
+      "grad_norm": 3.66015625,
+      "learning_rate": 3.340226817878586e-05,
+      "loss": 1.3176,
+      "step": 9993
+    },
+    {
+      "epoch": 9.225940456958227,
+      "grad_norm": 3.75,
+      "learning_rate": 3.339559706470981e-05,
+      "loss": 1.3076,
+      "step": 9994
+    },
+    {
+      "epoch": 9.226863604892683,
+      "grad_norm": 3.39453125,
+      "learning_rate": 3.338892595063376e-05,
+      "loss": 1.5239,
+      "step": 9995
+    },
+    {
+      "epoch": 9.227786752827141,
+      "grad_norm": 3.41015625,
+      "learning_rate": 3.3382254836557706e-05,
+      "loss": 1.2087,
+      "step": 9996
+    },
+    {
+      "epoch": 9.228709900761597,
+      "grad_norm": 3.80859375,
+      "learning_rate": 3.337558372248166e-05,
+      "loss": 1.1439,
+      "step": 9997
+    },
+    {
+      "epoch": 9.229633048696053,
+      "grad_norm": 3.142578125,
+      "learning_rate": 3.33689126084056e-05,
+      "loss": 1.4795,
+      "step": 9998
+    },
+    {
+      "epoch": 9.23055619663051,
+      "grad_norm": 2.99609375,
+      "learning_rate": 3.3362241494329556e-05,
+      "loss": 1.1331,
+      "step": 9999
+    },
+    {
+      "epoch": 9.231479344564967,
+      "grad_norm": 2.966796875,
+      "learning_rate": 3.33555703802535e-05,
+      "loss": 1.2925,
+      "step": 10000
+    },
+    {
+      "epoch": 9.231479344564967,
+      "eval_loss": 0.876953125,
+      "eval_runtime": 23.3655,
+      "eval_samples_per_second": 975.84,
+      "eval_steps_per_second": 61.03,
+      "step": 10000
+    },
+    {
+      "epoch": 9.232402492499423,
+      "grad_norm": 4.08203125,
+      "learning_rate": 3.334889926617745e-05,
+      "loss": 1.3262,
+      "step": 10001
+    },
+    {
+      "epoch": 9.233325640433879,
+      "grad_norm": 3.1484375,
+      "learning_rate": 3.33422281521014e-05,
+      "loss": 1.2871,
+      "step": 10002
+    },
+    {
+      "epoch": 9.234248788368337,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.333555703802535e-05,
+      "loss": 1.3254,
+      "step": 10003
+    },
+    {
+      "epoch": 9.235171936302793,
+      "grad_norm": 3.0546875,
+      "learning_rate": 3.3328885923949296e-05,
+      "loss": 1.4587,
+      "step": 10004
+    },
+    {
+      "epoch": 9.236095084237249,
+      "grad_norm": 3.365234375,
+      "learning_rate": 3.332221480987325e-05,
+      "loss": 1.5042,
+      "step": 10005
+    },
+    {
+      "epoch": 9.237018232171705,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.33155436957972e-05,
+      "loss": 1.1497,
+      "step": 10006
+    },
+    {
+      "epoch": 9.237941380106163,
+      "grad_norm": 3.658203125,
+      "learning_rate": 3.330887258172115e-05,
+      "loss": 1.3528,
+      "step": 10007
+    },
+    {
+      "epoch": 9.238864528040619,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.33022014676451e-05,
+      "loss": 1.2462,
+      "step": 10008
+    },
+    {
+      "epoch": 9.239787675975075,
+      "grad_norm": 3.119140625,
+      "learning_rate": 3.329553035356905e-05,
+      "loss": 1.4082,
+      "step": 10009
+    },
+    {
+      "epoch": 9.24071082390953,
+      "grad_norm": 3.775390625,
+      "learning_rate": 3.3288859239492996e-05,
+      "loss": 1.2649,
+      "step": 10010
+    },
+    {
+      "epoch": 9.241633971843989,
+      "grad_norm": 3.2578125,
+      "learning_rate": 3.328218812541695e-05,
+      "loss": 1.3867,
+      "step": 10011
+    },
+    {
+      "epoch": 9.242557119778445,
+      "grad_norm": 3.75,
+      "learning_rate": 3.3275517011340894e-05,
+      "loss": 1.1576,
+      "step": 10012
+    },
+    {
+      "epoch": 9.2434802677129,
+      "grad_norm": 3.6328125,
+      "learning_rate": 3.3268845897264846e-05,
+      "loss": 1.2686,
+      "step": 10013
+    },
+    {
+      "epoch": 9.244403415647357,
+      "grad_norm": 2.9921875,
+      "learning_rate": 3.326217478318879e-05,
+      "loss": 1.4429,
+      "step": 10014
+    },
+    {
+      "epoch": 9.245326563581814,
+      "grad_norm": 3.60546875,
+      "learning_rate": 3.3255503669112743e-05,
+      "loss": 1.365,
+      "step": 10015
+    },
+    {
+      "epoch": 9.24624971151627,
+      "grad_norm": 3.48828125,
+      "learning_rate": 3.324883255503669e-05,
+      "loss": 1.3491,
+      "step": 10016
+    },
+    {
+      "epoch": 9.247172859450727,
+      "grad_norm": 3.904296875,
+      "learning_rate": 3.324216144096064e-05,
+      "loss": 1.2463,
+      "step": 10017
+    },
+    {
+      "epoch": 9.248096007385184,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.323549032688459e-05,
+      "loss": 1.1884,
+      "step": 10018
+    },
+    {
+      "epoch": 9.24901915531964,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.3228819212808546e-05,
+      "loss": 1.2703,
+      "step": 10019
+    },
+    {
+      "epoch": 9.249942303254096,
+      "grad_norm": 3.333984375,
+      "learning_rate": 3.322214809873249e-05,
+      "loss": 1.2854,
+      "step": 10020
+    },
+    {
+      "epoch": 9.250865451188552,
+      "grad_norm": 2.94921875,
+      "learning_rate": 3.321547698465644e-05,
+      "loss": 1.4885,
+      "step": 10021
+    },
+    {
+      "epoch": 9.25178859912301,
+      "grad_norm": 3.240234375,
+      "learning_rate": 3.320880587058039e-05,
+      "loss": 1.2651,
+      "step": 10022
+    },
+    {
+      "epoch": 9.252711747057466,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.320213475650434e-05,
+      "loss": 1.4275,
+      "step": 10023
+    },
+    {
+      "epoch": 9.253634894991922,
+      "grad_norm": 3.943359375,
+      "learning_rate": 3.3195463642428286e-05,
+      "loss": 1.1444,
+      "step": 10024
+    },
+    {
+      "epoch": 9.254558042926378,
+      "grad_norm": 3.001953125,
+      "learning_rate": 3.318879252835224e-05,
+      "loss": 1.3875,
+      "step": 10025
+    },
+    {
+      "epoch": 9.255481190860836,
+      "grad_norm": 3.41796875,
+      "learning_rate": 3.3182121414276184e-05,
+      "loss": 1.2174,
+      "step": 10026
+    },
+    {
+      "epoch": 9.256404338795292,
+      "grad_norm": 3.6953125,
+      "learning_rate": 3.3175450300200136e-05,
+      "loss": 1.2906,
+      "step": 10027
+    },
+    {
+      "epoch": 9.257327486729748,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.316877918612408e-05,
+      "loss": 1.4951,
+      "step": 10028
+    },
+    {
+      "epoch": 9.258250634664204,
+      "grad_norm": 3.205078125,
+      "learning_rate": 3.3162108072048034e-05,
+      "loss": 1.2095,
+      "step": 10029
+    },
+    {
+      "epoch": 9.259173782598662,
+      "grad_norm": 3.009765625,
+      "learning_rate": 3.315543695797198e-05,
+      "loss": 1.3115,
+      "step": 10030
+    },
+    {
+      "epoch": 9.260096930533118,
+      "grad_norm": 3.333984375,
+      "learning_rate": 3.314876584389594e-05,
+      "loss": 1.3564,
+      "step": 10031
+    },
+    {
+      "epoch": 9.261020078467574,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.3142094729819884e-05,
+      "loss": 1.2458,
+      "step": 10032
+    },
+    {
+      "epoch": 9.261943226402032,
+      "grad_norm": 3.3984375,
+      "learning_rate": 3.3135423615743836e-05,
+      "loss": 1.3459,
+      "step": 10033
+    },
+    {
+      "epoch": 9.262866374336488,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.312875250166778e-05,
+      "loss": 1.637,
+      "step": 10034
+    },
+    {
+      "epoch": 9.263789522270944,
+      "grad_norm": 3.537109375,
+      "learning_rate": 3.3122081387591733e-05,
+      "loss": 1.2493,
+      "step": 10035
+    },
+    {
+      "epoch": 9.2647126702054,
+      "grad_norm": 3.44921875,
+      "learning_rate": 3.311541027351568e-05,
+      "loss": 1.4458,
+      "step": 10036
+    },
+    {
+      "epoch": 9.265635818139858,
+      "grad_norm": 3.330078125,
+      "learning_rate": 3.310873915943963e-05,
+      "loss": 1.0496,
+      "step": 10037
+    },
+    {
+      "epoch": 9.266558966074314,
+      "grad_norm": 3.005859375,
+      "learning_rate": 3.3102068045363576e-05,
+      "loss": 0.754,
+      "step": 10038
+    },
+    {
+      "epoch": 9.26748211400877,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.309539693128753e-05,
+      "loss": 1.2566,
+      "step": 10039
+    },
+    {
+      "epoch": 9.268405261943226,
+      "grad_norm": 2.466796875,
+      "learning_rate": 3.3088725817211474e-05,
+      "loss": 1.0248,
+      "step": 10040
+    },
+    {
+      "epoch": 9.269328409877684,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.3082054703135426e-05,
+      "loss": 1.1609,
+      "step": 10041
+    },
+    {
+      "epoch": 9.27025155781214,
+      "grad_norm": 3.927734375,
+      "learning_rate": 3.307538358905937e-05,
+      "loss": 1.4558,
+      "step": 10042
+    },
+    {
+      "epoch": 9.271174705746596,
+      "grad_norm": 3.626953125,
+      "learning_rate": 3.3068712474983324e-05,
+      "loss": 1.2478,
+      "step": 10043
+    },
+    {
+      "epoch": 9.272097853681052,
+      "grad_norm": 3.625,
+      "learning_rate": 3.3062041360907276e-05,
+      "loss": 1.1934,
+      "step": 10044
+    },
+    {
+      "epoch": 9.27302100161551,
+      "grad_norm": 3.615234375,
+      "learning_rate": 3.305537024683122e-05,
+      "loss": 1.4426,
+      "step": 10045
+    },
+    {
+      "epoch": 9.273944149549965,
+      "grad_norm": 2.71484375,
+      "learning_rate": 3.3048699132755174e-05,
+      "loss": 1.3063,
+      "step": 10046
+    },
+    {
+      "epoch": 9.274867297484422,
+      "grad_norm": 4.203125,
+      "learning_rate": 3.304202801867912e-05,
+      "loss": 1.5071,
+      "step": 10047
+    },
+    {
+      "epoch": 9.275790445418878,
+      "grad_norm": 3.068359375,
+      "learning_rate": 3.303535690460307e-05,
+      "loss": 1.3135,
+      "step": 10048
+    },
+    {
+      "epoch": 9.276713593353335,
+      "grad_norm": 3.4453125,
+      "learning_rate": 3.302868579052702e-05,
+      "loss": 1.4282,
+      "step": 10049
+    },
+    {
+      "epoch": 9.277636741287791,
+      "grad_norm": 3.5078125,
+      "learning_rate": 3.302201467645097e-05,
+      "loss": 1.5276,
+      "step": 10050
+    },
+    {
+      "epoch": 9.278559889222247,
+      "grad_norm": 3.111328125,
+      "learning_rate": 3.3015343562374914e-05,
+      "loss": 1.3673,
+      "step": 10051
+    },
+    {
+      "epoch": 9.279483037156705,
+      "grad_norm": 3.28515625,
+      "learning_rate": 3.300867244829887e-05,
+      "loss": 1.0756,
+      "step": 10052
+    },
+    {
+      "epoch": 9.280406185091161,
+      "grad_norm": 3.802734375,
+      "learning_rate": 3.300200133422281e-05,
+      "loss": 1.3082,
+      "step": 10053
+    },
+    {
+      "epoch": 9.281329333025617,
+      "grad_norm": 3.228515625,
+      "learning_rate": 3.2995330220146764e-05,
+      "loss": 1.4873,
+      "step": 10054
+    },
+    {
+      "epoch": 9.282252480960073,
+      "grad_norm": 2.81640625,
+      "learning_rate": 3.298865910607071e-05,
+      "loss": 1.3412,
+      "step": 10055
+    },
+    {
+      "epoch": 9.283175628894531,
+      "grad_norm": 3.08984375,
+      "learning_rate": 3.298198799199467e-05,
+      "loss": 1.3906,
+      "step": 10056
+    },
+    {
+      "epoch": 9.284098776828987,
+      "grad_norm": 2.9140625,
+      "learning_rate": 3.2975316877918614e-05,
+      "loss": 1.2925,
+      "step": 10057
+    },
+    {
+      "epoch": 9.285021924763443,
+      "grad_norm": 3.96484375,
+      "learning_rate": 3.2968645763842566e-05,
+      "loss": 1.501,
+      "step": 10058
+    },
+    {
+      "epoch": 9.2859450726979,
+      "grad_norm": 3.6328125,
+      "learning_rate": 3.296197464976651e-05,
+      "loss": 1.3442,
+      "step": 10059
+    },
+    {
+      "epoch": 9.286868220632357,
+      "grad_norm": 3.18359375,
+      "learning_rate": 3.2955303535690464e-05,
+      "loss": 1.0965,
+      "step": 10060
+    },
+    {
+      "epoch": 9.287791368566813,
+      "grad_norm": 3.080078125,
+      "learning_rate": 3.294863242161441e-05,
+      "loss": 1.3279,
+      "step": 10061
+    },
+    {
+      "epoch": 9.288714516501269,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.294196130753836e-05,
+      "loss": 1.2865,
+      "step": 10062
+    },
+    {
+      "epoch": 9.289637664435725,
+      "grad_norm": 3.74609375,
+      "learning_rate": 3.293529019346231e-05,
+      "loss": 1.0793,
+      "step": 10063
+    },
+    {
+      "epoch": 9.290560812370183,
+      "grad_norm": 3.0703125,
+      "learning_rate": 3.292861907938626e-05,
+      "loss": 1.3557,
+      "step": 10064
+    },
+    {
+      "epoch": 9.291483960304639,
+      "grad_norm": 3.537109375,
+      "learning_rate": 3.2921947965310205e-05,
+      "loss": 1.2625,
+      "step": 10065
+    },
+    {
+      "epoch": 9.292407108239095,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.291527685123416e-05,
+      "loss": 0.9575,
+      "step": 10066
+    },
+    {
+      "epoch": 9.293330256173551,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.29086057371581e-05,
+      "loss": 0.9678,
+      "step": 10067
+    },
+    {
+      "epoch": 9.294253404108009,
+      "grad_norm": 3.791015625,
+      "learning_rate": 3.2901934623082055e-05,
+      "loss": 1.3835,
+      "step": 10068
+    },
+    {
+      "epoch": 9.295176552042465,
+      "grad_norm": 3.51171875,
+      "learning_rate": 3.289526350900601e-05,
+      "loss": 1.3374,
+      "step": 10069
+    },
+    {
+      "epoch": 9.29609969997692,
+      "grad_norm": 3.1015625,
+      "learning_rate": 3.288859239492996e-05,
+      "loss": 1.22,
+      "step": 10070
+    },
+    {
+      "epoch": 9.297022847911379,
+      "grad_norm": 3.666015625,
+      "learning_rate": 3.2881921280853904e-05,
+      "loss": 1.1511,
+      "step": 10071
+    },
+    {
+      "epoch": 9.297945995845835,
+      "grad_norm": 3.072265625,
+      "learning_rate": 3.287525016677786e-05,
+      "loss": 1.2253,
+      "step": 10072
+    },
+    {
+      "epoch": 9.29886914378029,
+      "grad_norm": 3.966796875,
+      "learning_rate": 3.28685790527018e-05,
+      "loss": 1.2014,
+      "step": 10073
+    },
+    {
+      "epoch": 9.299792291714747,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.2861907938625754e-05,
+      "loss": 1.2605,
+      "step": 10074
+    },
+    {
+      "epoch": 9.300715439649204,
+      "grad_norm": 2.974609375,
+      "learning_rate": 3.28552368245497e-05,
+      "loss": 1.2087,
+      "step": 10075
+    },
+    {
+      "epoch": 9.30163858758366,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.284856571047365e-05,
+      "loss": 1.396,
+      "step": 10076
+    },
+    {
+      "epoch": 9.302561735518116,
+      "grad_norm": 3.150390625,
+      "learning_rate": 3.28418945963976e-05,
+      "loss": 1.3311,
+      "step": 10077
+    },
+    {
+      "epoch": 9.303484883452573,
+      "grad_norm": 3.701171875,
+      "learning_rate": 3.283522348232155e-05,
+      "loss": 1.311,
+      "step": 10078
+    },
+    {
+      "epoch": 9.30440803138703,
+      "grad_norm": 3.533203125,
+      "learning_rate": 3.2828552368245495e-05,
+      "loss": 1.0643,
+      "step": 10079
+    },
+    {
+      "epoch": 9.305331179321486,
+      "grad_norm": 2.82421875,
+      "learning_rate": 3.282188125416945e-05,
+      "loss": 1.2576,
+      "step": 10080
+    },
+    {
+      "epoch": 9.306254327255942,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.28152101400934e-05,
+      "loss": 1.364,
+      "step": 10081
+    },
+    {
+      "epoch": 9.3071774751904,
+      "grad_norm": 3.84765625,
+      "learning_rate": 3.280853902601735e-05,
+      "loss": 1.2474,
+      "step": 10082
+    },
+    {
+      "epoch": 9.308100623124856,
+      "grad_norm": 3.0546875,
+      "learning_rate": 3.28018679119413e-05,
+      "loss": 1.3324,
+      "step": 10083
+    },
+    {
+      "epoch": 9.309023771059312,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.279519679786525e-05,
+      "loss": 1.4087,
+      "step": 10084
+    },
+    {
+      "epoch": 9.309946918993768,
+      "grad_norm": 3.14453125,
+      "learning_rate": 3.2788525683789195e-05,
+      "loss": 1.2026,
+      "step": 10085
+    },
+    {
+      "epoch": 9.310870066928226,
+      "grad_norm": 3.60546875,
+      "learning_rate": 3.278185456971315e-05,
+      "loss": 1.2465,
+      "step": 10086
+    },
+    {
+      "epoch": 9.311793214862682,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.277518345563709e-05,
+      "loss": 1.1318,
+      "step": 10087
+    },
+    {
+      "epoch": 9.312716362797138,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.2768512341561044e-05,
+      "loss": 1.1743,
+      "step": 10088
+    },
+    {
+      "epoch": 9.313639510731594,
+      "grad_norm": 3.419921875,
+      "learning_rate": 3.276184122748499e-05,
+      "loss": 1.0897,
+      "step": 10089
+    },
+    {
+      "epoch": 9.314562658666052,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.275517011340894e-05,
+      "loss": 1.2008,
+      "step": 10090
+    },
+    {
+      "epoch": 9.315485806600508,
+      "grad_norm": 3.0703125,
+      "learning_rate": 3.274849899933289e-05,
+      "loss": 1.3896,
+      "step": 10091
+    },
+    {
+      "epoch": 9.316408954534964,
+      "grad_norm": 3.357421875,
+      "learning_rate": 3.274182788525684e-05,
+      "loss": 1.416,
+      "step": 10092
+    },
+    {
+      "epoch": 9.31733210246942,
+      "grad_norm": 3.080078125,
+      "learning_rate": 3.2735156771180785e-05,
+      "loss": 1.3618,
+      "step": 10093
+    },
+    {
+      "epoch": 9.318255250403878,
+      "grad_norm": 3.70703125,
+      "learning_rate": 3.272848565710474e-05,
+      "loss": 1.3616,
+      "step": 10094
+    },
+    {
+      "epoch": 9.319178398338334,
+      "grad_norm": 3.162109375,
+      "learning_rate": 3.272181454302869e-05,
+      "loss": 1.1882,
+      "step": 10095
+    },
+    {
+      "epoch": 9.32010154627279,
+      "grad_norm": 3.1015625,
+      "learning_rate": 3.2715143428952635e-05,
+      "loss": 1.2661,
+      "step": 10096
+    },
+    {
+      "epoch": 9.321024694207246,
+      "grad_norm": 3.599609375,
+      "learning_rate": 3.270847231487659e-05,
+      "loss": 1.2316,
+      "step": 10097
+    },
+    {
+      "epoch": 9.321947842141704,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.270180120080053e-05,
+      "loss": 1.4241,
+      "step": 10098
+    },
+    {
+      "epoch": 9.32287099007616,
+      "grad_norm": 3.529296875,
+      "learning_rate": 3.2695130086724485e-05,
+      "loss": 0.9774,
+      "step": 10099
+    },
+    {
+      "epoch": 9.323794138010616,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.268845897264843e-05,
+      "loss": 1.302,
+      "step": 10100
+    },
+    {
+      "epoch": 9.324717285945074,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.268178785857238e-05,
+      "loss": 1.3645,
+      "step": 10101
+    },
+    {
+      "epoch": 9.32564043387953,
+      "grad_norm": 3.375,
+      "learning_rate": 3.267511674449633e-05,
+      "loss": 1.3923,
+      "step": 10102
+    },
+    {
+      "epoch": 9.326563581813986,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.266844563042028e-05,
+      "loss": 1.4954,
+      "step": 10103
+    },
+    {
+      "epoch": 9.327486729748442,
+      "grad_norm": 3.20703125,
+      "learning_rate": 3.266177451634423e-05,
+      "loss": 1.3445,
+      "step": 10104
+    },
+    {
+      "epoch": 9.3284098776829,
+      "grad_norm": 3.396484375,
+      "learning_rate": 3.265510340226818e-05,
+      "loss": 1.3313,
+      "step": 10105
+    },
+    {
+      "epoch": 9.329333025617355,
+      "grad_norm": 3.212890625,
+      "learning_rate": 3.264843228819213e-05,
+      "loss": 1.2246,
+      "step": 10106
+    },
+    {
+      "epoch": 9.330256173551811,
+      "grad_norm": 2.96484375,
+      "learning_rate": 3.264176117411608e-05,
+      "loss": 1.4161,
+      "step": 10107
+    },
+    {
+      "epoch": 9.331179321486267,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.263509006004003e-05,
+      "loss": 1.2095,
+      "step": 10108
+    },
+    {
+      "epoch": 9.332102469420725,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.262841894596398e-05,
+      "loss": 1.5208,
+      "step": 10109
+    },
+    {
+      "epoch": 9.333025617355181,
+      "grad_norm": 3.337890625,
+      "learning_rate": 3.2621747831887925e-05,
+      "loss": 1.5286,
+      "step": 10110
+    },
+    {
+      "epoch": 9.333948765289637,
+      "grad_norm": 3.037109375,
+      "learning_rate": 3.261507671781188e-05,
+      "loss": 1.4343,
+      "step": 10111
+    },
+    {
+      "epoch": 9.334871913224093,
+      "grad_norm": 3.486328125,
+      "learning_rate": 3.260840560373582e-05,
+      "loss": 1.5232,
+      "step": 10112
+    },
+    {
+      "epoch": 9.335795061158551,
+      "grad_norm": 3.201171875,
+      "learning_rate": 3.2601734489659775e-05,
+      "loss": 1.3081,
+      "step": 10113
+    },
+    {
+      "epoch": 9.336718209093007,
+      "grad_norm": 3.60546875,
+      "learning_rate": 3.259506337558372e-05,
+      "loss": 1.2954,
+      "step": 10114
+    },
+    {
+      "epoch": 9.337641357027463,
+      "grad_norm": 2.974609375,
+      "learning_rate": 3.258839226150767e-05,
+      "loss": 1.1666,
+      "step": 10115
+    },
+    {
+      "epoch": 9.33856450496192,
+      "grad_norm": 3.263671875,
+      "learning_rate": 3.258172114743162e-05,
+      "loss": 1.2612,
+      "step": 10116
+    },
+    {
+      "epoch": 9.339487652896377,
+      "grad_norm": 3.5703125,
+      "learning_rate": 3.257505003335557e-05,
+      "loss": 1.3728,
+      "step": 10117
+    },
+    {
+      "epoch": 9.340410800830833,
+      "grad_norm": 3.443359375,
+      "learning_rate": 3.2568378919279516e-05,
+      "loss": 1.1826,
+      "step": 10118
+    },
+    {
+      "epoch": 9.341333948765289,
+      "grad_norm": 3.044921875,
+      "learning_rate": 3.2561707805203475e-05,
+      "loss": 1.4368,
+      "step": 10119
+    },
+    {
+      "epoch": 9.342257096699747,
+      "grad_norm": 4.15625,
+      "learning_rate": 3.255503669112742e-05,
+      "loss": 1.3354,
+      "step": 10120
+    },
+    {
+      "epoch": 9.343180244634203,
+      "grad_norm": 3.00390625,
+      "learning_rate": 3.254836557705137e-05,
+      "loss": 1.314,
+      "step": 10121
+    },
+    {
+      "epoch": 9.344103392568659,
+      "grad_norm": 3.1640625,
+      "learning_rate": 3.254169446297532e-05,
+      "loss": 1.3304,
+      "step": 10122
+    },
+    {
+      "epoch": 9.345026540503115,
+      "grad_norm": 3.486328125,
+      "learning_rate": 3.253502334889927e-05,
+      "loss": 1.2627,
+      "step": 10123
+    },
+    {
+      "epoch": 9.345949688437573,
+      "grad_norm": 2.904296875,
+      "learning_rate": 3.2528352234823216e-05,
+      "loss": 1.3906,
+      "step": 10124
+    },
+    {
+      "epoch": 9.346872836372029,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.252168112074717e-05,
+      "loss": 1.3541,
+      "step": 10125
+    },
+    {
+      "epoch": 9.347795984306485,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.251501000667111e-05,
+      "loss": 1.172,
+      "step": 10126
+    },
+    {
+      "epoch": 9.34871913224094,
+      "grad_norm": 3.505859375,
+      "learning_rate": 3.2508338892595065e-05,
+      "loss": 1.0758,
+      "step": 10127
+    },
+    {
+      "epoch": 9.349642280175399,
+      "grad_norm": 3.095703125,
+      "learning_rate": 3.250166777851901e-05,
+      "loss": 1.025,
+      "step": 10128
+    },
+    {
+      "epoch": 9.350565428109855,
+      "grad_norm": 2.798828125,
+      "learning_rate": 3.249499666444296e-05,
+      "loss": 1.4668,
+      "step": 10129
+    },
+    {
+      "epoch": 9.35148857604431,
+      "grad_norm": 3.611328125,
+      "learning_rate": 3.248832555036691e-05,
+      "loss": 1.4548,
+      "step": 10130
+    },
+    {
+      "epoch": 9.352411723978767,
+      "grad_norm": 3.103515625,
+      "learning_rate": 3.248165443629087e-05,
+      "loss": 1.3171,
+      "step": 10131
+    },
+    {
+      "epoch": 9.353334871913225,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.247498332221481e-05,
+      "loss": 1.325,
+      "step": 10132
+    },
+    {
+      "epoch": 9.35425801984768,
+      "grad_norm": 3.25,
+      "learning_rate": 3.2468312208138765e-05,
+      "loss": 1.5083,
+      "step": 10133
+    },
+    {
+      "epoch": 9.355181167782137,
+      "grad_norm": 3.181640625,
+      "learning_rate": 3.246164109406271e-05,
+      "loss": 1.4158,
+      "step": 10134
+    },
+    {
+      "epoch": 9.356104315716594,
+      "grad_norm": 4.27734375,
+      "learning_rate": 3.245496997998666e-05,
+      "loss": 1.1077,
+      "step": 10135
+    },
+    {
+      "epoch": 9.35702746365105,
+      "grad_norm": 3.349609375,
+      "learning_rate": 3.244829886591061e-05,
+      "loss": 1.2852,
+      "step": 10136
+    },
+    {
+      "epoch": 9.357950611585506,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.244162775183456e-05,
+      "loss": 1.1343,
+      "step": 10137
+    },
+    {
+      "epoch": 9.358873759519962,
+      "grad_norm": 3.482421875,
+      "learning_rate": 3.2434956637758506e-05,
+      "loss": 1.2618,
+      "step": 10138
+    },
+    {
+      "epoch": 9.35979690745442,
+      "grad_norm": 3.03515625,
+      "learning_rate": 3.242828552368246e-05,
+      "loss": 1.3746,
+      "step": 10139
+    },
+    {
+      "epoch": 9.360720055388876,
+      "grad_norm": 3.73046875,
+      "learning_rate": 3.2421614409606403e-05,
+      "loss": 1.3169,
+      "step": 10140
+    },
+    {
+      "epoch": 9.361643203323332,
+      "grad_norm": 3.365234375,
+      "learning_rate": 3.2414943295530356e-05,
+      "loss": 1.3267,
+      "step": 10141
+    },
+    {
+      "epoch": 9.362566351257788,
+      "grad_norm": 3.4765625,
+      "learning_rate": 3.24082721814543e-05,
+      "loss": 1.0725,
+      "step": 10142
+    },
+    {
+      "epoch": 9.363489499192246,
+      "grad_norm": 3.310546875,
+      "learning_rate": 3.240160106737825e-05,
+      "loss": 1.6035,
+      "step": 10143
+    },
+    {
+      "epoch": 9.364412647126702,
+      "grad_norm": 3.634765625,
+      "learning_rate": 3.2394929953302205e-05,
+      "loss": 1.322,
+      "step": 10144
+    },
+    {
+      "epoch": 9.365335795061158,
+      "grad_norm": 2.923828125,
+      "learning_rate": 3.238825883922616e-05,
+      "loss": 1.5288,
+      "step": 10145
+    },
+    {
+      "epoch": 9.366258942995614,
+      "grad_norm": 3.66796875,
+      "learning_rate": 3.23815877251501e-05,
+      "loss": 1.2568,
+      "step": 10146
+    },
+    {
+      "epoch": 9.367182090930072,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.2374916611074055e-05,
+      "loss": 1.2988,
+      "step": 10147
+    },
+    {
+      "epoch": 9.368105238864528,
+      "grad_norm": 3.771484375,
+      "learning_rate": 3.2368245496998e-05,
+      "loss": 1.5205,
+      "step": 10148
+    },
+    {
+      "epoch": 9.369028386798984,
+      "grad_norm": 3.09765625,
+      "learning_rate": 3.236157438292195e-05,
+      "loss": 1.415,
+      "step": 10149
+    },
+    {
+      "epoch": 9.369951534733442,
+      "grad_norm": 3.353515625,
+      "learning_rate": 3.23549032688459e-05,
+      "loss": 0.8268,
+      "step": 10150
+    },
+    {
+      "epoch": 9.370874682667898,
+      "grad_norm": 3.818359375,
+      "learning_rate": 3.234823215476985e-05,
+      "loss": 1.0433,
+      "step": 10151
+    },
+    {
+      "epoch": 9.371797830602354,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.2341561040693796e-05,
+      "loss": 1.0842,
+      "step": 10152
+    },
+    {
+      "epoch": 9.37272097853681,
+      "grad_norm": 4.125,
+      "learning_rate": 3.233488992661775e-05,
+      "loss": 1.2219,
+      "step": 10153
+    },
+    {
+      "epoch": 9.373644126471268,
+      "grad_norm": 3.419921875,
+      "learning_rate": 3.2328218812541694e-05,
+      "loss": 1.2435,
+      "step": 10154
+    },
+    {
+      "epoch": 9.374567274405724,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.2321547698465646e-05,
+      "loss": 1.2352,
+      "step": 10155
+    },
+    {
+      "epoch": 9.37549042234018,
+      "grad_norm": 2.9765625,
+      "learning_rate": 3.23148765843896e-05,
+      "loss": 1.3567,
+      "step": 10156
+    },
+    {
+      "epoch": 9.376413570274636,
+      "grad_norm": 3.576171875,
+      "learning_rate": 3.2308205470313543e-05,
+      "loss": 1.2322,
+      "step": 10157
+    },
+    {
+      "epoch": 9.377336718209094,
+      "grad_norm": 2.779296875,
+      "learning_rate": 3.2301534356237496e-05,
+      "loss": 1.2648,
+      "step": 10158
+    },
+    {
+      "epoch": 9.37825986614355,
+      "grad_norm": 3.73828125,
+      "learning_rate": 3.229486324216144e-05,
+      "loss": 1.3423,
+      "step": 10159
+    },
+    {
+      "epoch": 9.379183014078006,
+      "grad_norm": 3.20703125,
+      "learning_rate": 3.228819212808539e-05,
+      "loss": 1.2001,
+      "step": 10160
+    },
+    {
+      "epoch": 9.380106162012462,
+      "grad_norm": 3.1796875,
+      "learning_rate": 3.228152101400934e-05,
+      "loss": 1.2404,
+      "step": 10161
+    },
+    {
+      "epoch": 9.38102930994692,
+      "grad_norm": 3.013671875,
+      "learning_rate": 3.227484989993329e-05,
+      "loss": 1.3672,
+      "step": 10162
+    },
+    {
+      "epoch": 9.381952457881376,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.2268178785857236e-05,
+      "loss": 1.2823,
+      "step": 10163
+    },
+    {
+      "epoch": 9.382875605815832,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.226150767178119e-05,
+      "loss": 1.4087,
+      "step": 10164
+    },
+    {
+      "epoch": 9.383798753750288,
+      "grad_norm": 3.224609375,
+      "learning_rate": 3.2254836557705134e-05,
+      "loss": 1.3684,
+      "step": 10165
+    },
+    {
+      "epoch": 9.384721901684745,
+      "grad_norm": 3.02734375,
+      "learning_rate": 3.2248165443629086e-05,
+      "loss": 1.1664,
+      "step": 10166
+    },
+    {
+      "epoch": 9.385645049619201,
+      "grad_norm": 2.982421875,
+      "learning_rate": 3.224149432955303e-05,
+      "loss": 1.3335,
+      "step": 10167
+    },
+    {
+      "epoch": 9.386568197553657,
+      "grad_norm": 3.48828125,
+      "learning_rate": 3.2234823215476984e-05,
+      "loss": 1.0268,
+      "step": 10168
+    },
+    {
+      "epoch": 9.387491345488115,
+      "grad_norm": 3.333984375,
+      "learning_rate": 3.2228152101400936e-05,
+      "loss": 1.4458,
+      "step": 10169
+    },
+    {
+      "epoch": 9.388414493422571,
+      "grad_norm": 3.44921875,
+      "learning_rate": 3.222148098732489e-05,
+      "loss": 1.15,
+      "step": 10170
+    },
+    {
+      "epoch": 9.389337641357027,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.2214809873248834e-05,
+      "loss": 1.147,
+      "step": 10171
+    },
+    {
+      "epoch": 9.390260789291483,
+      "grad_norm": 3.142578125,
+      "learning_rate": 3.2208138759172786e-05,
+      "loss": 1.3242,
+      "step": 10172
+    },
+    {
+      "epoch": 9.391183937225941,
+      "grad_norm": 3.302734375,
+      "learning_rate": 3.220146764509673e-05,
+      "loss": 1.3384,
+      "step": 10173
+    },
+    {
+      "epoch": 9.392107085160397,
+      "grad_norm": 3.861328125,
+      "learning_rate": 3.2194796531020684e-05,
+      "loss": 1.2729,
+      "step": 10174
+    },
+    {
+      "epoch": 9.393030233094853,
+      "grad_norm": 3.34765625,
+      "learning_rate": 3.218812541694463e-05,
+      "loss": 1.3972,
+      "step": 10175
+    },
+    {
+      "epoch": 9.39395338102931,
+      "grad_norm": 3.1796875,
+      "learning_rate": 3.218145430286858e-05,
+      "loss": 1.292,
+      "step": 10176
+    },
+    {
+      "epoch": 9.394876528963767,
+      "grad_norm": 2.923828125,
+      "learning_rate": 3.217478318879253e-05,
+      "loss": 1.2278,
+      "step": 10177
+    },
+    {
+      "epoch": 9.395799676898223,
+      "grad_norm": 2.994140625,
+      "learning_rate": 3.216811207471648e-05,
+      "loss": 1.2095,
+      "step": 10178
+    },
+    {
+      "epoch": 9.396722824832679,
+      "grad_norm": 3.052734375,
+      "learning_rate": 3.2161440960640424e-05,
+      "loss": 1.5186,
+      "step": 10179
+    },
+    {
+      "epoch": 9.397645972767135,
+      "grad_norm": 4.04296875,
+      "learning_rate": 3.2154769846564376e-05,
+      "loss": 1.3704,
+      "step": 10180
+    },
+    {
+      "epoch": 9.398569120701593,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.214809873248833e-05,
+      "loss": 1.3721,
+      "step": 10181
+    },
+    {
+      "epoch": 9.399492268636049,
+      "grad_norm": 3.431640625,
+      "learning_rate": 3.214142761841228e-05,
+      "loss": 1.0566,
+      "step": 10182
+    },
+    {
+      "epoch": 9.400415416570505,
+      "grad_norm": 3.42578125,
+      "learning_rate": 3.2134756504336226e-05,
+      "loss": 0.9873,
+      "step": 10183
+    },
+    {
+      "epoch": 9.401338564504963,
+      "grad_norm": 3.23046875,
+      "learning_rate": 3.212808539026018e-05,
+      "loss": 1.1445,
+      "step": 10184
+    },
+    {
+      "epoch": 9.402261712439419,
+      "grad_norm": 3.6796875,
+      "learning_rate": 3.2121414276184124e-05,
+      "loss": 1.2646,
+      "step": 10185
+    },
+    {
+      "epoch": 9.403184860373875,
+      "grad_norm": 3.638671875,
+      "learning_rate": 3.2114743162108076e-05,
+      "loss": 1.5598,
+      "step": 10186
+    },
+    {
+      "epoch": 9.40410800830833,
+      "grad_norm": 3.4140625,
+      "learning_rate": 3.210807204803202e-05,
+      "loss": 1.3542,
+      "step": 10187
+    },
+    {
+      "epoch": 9.405031156242789,
+      "grad_norm": 3.208984375,
+      "learning_rate": 3.2101400933955974e-05,
+      "loss": 1.2574,
+      "step": 10188
+    },
+    {
+      "epoch": 9.405954304177245,
+      "grad_norm": 2.919921875,
+      "learning_rate": 3.209472981987992e-05,
+      "loss": 1.3623,
+      "step": 10189
+    },
+    {
+      "epoch": 9.4068774521117,
+      "grad_norm": 3.29296875,
+      "learning_rate": 3.208805870580387e-05,
+      "loss": 1.4038,
+      "step": 10190
+    },
+    {
+      "epoch": 9.407800600046157,
+      "grad_norm": 3.07421875,
+      "learning_rate": 3.208138759172782e-05,
+      "loss": 1.2361,
+      "step": 10191
+    },
+    {
+      "epoch": 9.408723747980615,
+      "grad_norm": 3.61328125,
+      "learning_rate": 3.207471647765177e-05,
+      "loss": 1.1263,
+      "step": 10192
+    },
+    {
+      "epoch": 9.40964689591507,
+      "grad_norm": 3.14453125,
+      "learning_rate": 3.2068045363575715e-05,
+      "loss": 1.311,
+      "step": 10193
+    },
+    {
+      "epoch": 9.410570043849527,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.2061374249499673e-05,
+      "loss": 1.2869,
+      "step": 10194
+    },
+    {
+      "epoch": 9.411493191783983,
+      "grad_norm": 3.61328125,
+      "learning_rate": 3.205470313542362e-05,
+      "loss": 1.3088,
+      "step": 10195
+    },
+    {
+      "epoch": 9.41241633971844,
+      "grad_norm": 3.369140625,
+      "learning_rate": 3.204803202134757e-05,
+      "loss": 1.1549,
+      "step": 10196
+    },
+    {
+      "epoch": 9.413339487652896,
+      "grad_norm": 2.86328125,
+      "learning_rate": 3.2041360907271517e-05,
+      "loss": 1.0671,
+      "step": 10197
+    },
+    {
+      "epoch": 9.414262635587352,
+      "grad_norm": 3.951171875,
+      "learning_rate": 3.203468979319547e-05,
+      "loss": 1.4602,
+      "step": 10198
+    },
+    {
+      "epoch": 9.41518578352181,
+      "grad_norm": 3.443359375,
+      "learning_rate": 3.2028018679119414e-05,
+      "loss": 1.2275,
+      "step": 10199
+    },
+    {
+      "epoch": 9.416108931456266,
+      "grad_norm": 3.3984375,
+      "learning_rate": 3.2021347565043366e-05,
+      "loss": 0.9073,
+      "step": 10200
+    },
+    {
+      "epoch": 9.417032079390722,
+      "grad_norm": 3.3671875,
+      "learning_rate": 3.201467645096731e-05,
+      "loss": 1.05,
+      "step": 10201
+    },
+    {
+      "epoch": 9.417955227325178,
+      "grad_norm": 3.263671875,
+      "learning_rate": 3.2008005336891264e-05,
+      "loss": 1.0206,
+      "step": 10202
+    },
+    {
+      "epoch": 9.418878375259636,
+      "grad_norm": 3.31640625,
+      "learning_rate": 3.200133422281521e-05,
+      "loss": 1.1237,
+      "step": 10203
+    },
+    {
+      "epoch": 9.419801523194092,
+      "grad_norm": 3.083984375,
+      "learning_rate": 3.199466310873916e-05,
+      "loss": 1.3171,
+      "step": 10204
+    },
+    {
+      "epoch": 9.420724671128548,
+      "grad_norm": 3.185546875,
+      "learning_rate": 3.198799199466311e-05,
+      "loss": 1.3689,
+      "step": 10205
+    },
+    {
+      "epoch": 9.421647819063004,
+      "grad_norm": 3.318359375,
+      "learning_rate": 3.198132088058706e-05,
+      "loss": 1.2771,
+      "step": 10206
+    },
+    {
+      "epoch": 9.422570966997462,
+      "grad_norm": 3.2890625,
+      "learning_rate": 3.197464976651101e-05,
+      "loss": 1.2894,
+      "step": 10207
+    },
+    {
+      "epoch": 9.423494114931918,
+      "grad_norm": 3.572265625,
+      "learning_rate": 3.196797865243496e-05,
+      "loss": 1.1843,
+      "step": 10208
+    },
+    {
+      "epoch": 9.424417262866374,
+      "grad_norm": 3.01953125,
+      "learning_rate": 3.196130753835891e-05,
+      "loss": 1.1102,
+      "step": 10209
+    },
+    {
+      "epoch": 9.42534041080083,
+      "grad_norm": 2.80078125,
+      "learning_rate": 3.1954636424282855e-05,
+      "loss": 1.243,
+      "step": 10210
+    },
+    {
+      "epoch": 9.426263558735288,
+      "grad_norm": 3.126953125,
+      "learning_rate": 3.194796531020681e-05,
+      "loss": 1.3513,
+      "step": 10211
+    },
+    {
+      "epoch": 9.427186706669744,
+      "grad_norm": 3.34765625,
+      "learning_rate": 3.194129419613075e-05,
+      "loss": 1.46,
+      "step": 10212
+    },
+    {
+      "epoch": 9.4281098546042,
+      "grad_norm": 3.1328125,
+      "learning_rate": 3.1934623082054704e-05,
+      "loss": 1.3271,
+      "step": 10213
+    },
+    {
+      "epoch": 9.429033002538656,
+      "grad_norm": 3.3203125,
+      "learning_rate": 3.192795196797865e-05,
+      "loss": 1.1321,
+      "step": 10214
+    },
+    {
+      "epoch": 9.429956150473114,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.19212808539026e-05,
+      "loss": 1.3958,
+      "step": 10215
+    },
+    {
+      "epoch": 9.43087929840757,
+      "grad_norm": 3.04296875,
+      "learning_rate": 3.1914609739826554e-05,
+      "loss": 1.2307,
+      "step": 10216
+    },
+    {
+      "epoch": 9.431802446342026,
+      "grad_norm": 2.57421875,
+      "learning_rate": 3.19079386257505e-05,
+      "loss": 1.1276,
+      "step": 10217
+    },
+    {
+      "epoch": 9.432725594276484,
+      "grad_norm": 3.375,
+      "learning_rate": 3.190126751167445e-05,
+      "loss": 1.4375,
+      "step": 10218
+    },
+    {
+      "epoch": 9.43364874221094,
+      "grad_norm": 3.880859375,
+      "learning_rate": 3.1894596397598404e-05,
+      "loss": 1.2107,
+      "step": 10219
+    },
+    {
+      "epoch": 9.434571890145396,
+      "grad_norm": 3.837890625,
+      "learning_rate": 3.188792528352235e-05,
+      "loss": 1.4751,
+      "step": 10220
+    },
+    {
+      "epoch": 9.435495038079852,
+      "grad_norm": 3.087890625,
+      "learning_rate": 3.18812541694463e-05,
+      "loss": 1.3907,
+      "step": 10221
+    },
+    {
+      "epoch": 9.43641818601431,
+      "grad_norm": 2.908203125,
+      "learning_rate": 3.187458305537025e-05,
+      "loss": 1.4863,
+      "step": 10222
+    },
+    {
+      "epoch": 9.437341333948766,
+      "grad_norm": 3.07421875,
+      "learning_rate": 3.18679119412942e-05,
+      "loss": 1.0564,
+      "step": 10223
+    },
+    {
+      "epoch": 9.438264481883222,
+      "grad_norm": 3.05078125,
+      "learning_rate": 3.1861240827218145e-05,
+      "loss": 1.0101,
+      "step": 10224
+    },
+    {
+      "epoch": 9.439187629817678,
+      "grad_norm": 4.0703125,
+      "learning_rate": 3.18545697131421e-05,
+      "loss": 1.3328,
+      "step": 10225
+    },
+    {
+      "epoch": 9.440110777752135,
+      "grad_norm": 3.4453125,
+      "learning_rate": 3.184789859906604e-05,
+      "loss": 1.3496,
+      "step": 10226
+    },
+    {
+      "epoch": 9.441033925686591,
+      "grad_norm": 3.3203125,
+      "learning_rate": 3.1841227484989995e-05,
+      "loss": 1.3551,
+      "step": 10227
+    },
+    {
+      "epoch": 9.441957073621047,
+      "grad_norm": 3.6015625,
+      "learning_rate": 3.183455637091394e-05,
+      "loss": 1.4658,
+      "step": 10228
+    },
+    {
+      "epoch": 9.442880221555503,
+      "grad_norm": 3.21484375,
+      "learning_rate": 3.182788525683789e-05,
+      "loss": 1.3235,
+      "step": 10229
+    },
+    {
+      "epoch": 9.443803369489961,
+      "grad_norm": 3.134765625,
+      "learning_rate": 3.182121414276184e-05,
+      "loss": 1.8013,
+      "step": 10230
+    },
+    {
+      "epoch": 9.444726517424417,
+      "grad_norm": 3.251953125,
+      "learning_rate": 3.181454302868579e-05,
+      "loss": 0.9828,
+      "step": 10231
+    },
+    {
+      "epoch": 9.445649665358873,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.180787191460974e-05,
+      "loss": 1.191,
+      "step": 10232
+    },
+    {
+      "epoch": 9.446572813293331,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.1801200800533694e-05,
+      "loss": 1.3855,
+      "step": 10233
+    },
+    {
+      "epoch": 9.447495961227787,
+      "grad_norm": 2.990234375,
+      "learning_rate": 3.179452968645764e-05,
+      "loss": 1.1166,
+      "step": 10234
+    },
+    {
+      "epoch": 9.448419109162243,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.178785857238159e-05,
+      "loss": 1.3003,
+      "step": 10235
+    },
+    {
+      "epoch": 9.4493422570967,
+      "grad_norm": 3.447265625,
+      "learning_rate": 3.178118745830554e-05,
+      "loss": 1.3611,
+      "step": 10236
+    },
+    {
+      "epoch": 9.450265405031157,
+      "grad_norm": 3.4765625,
+      "learning_rate": 3.177451634422949e-05,
+      "loss": 1.418,
+      "step": 10237
+    },
+    {
+      "epoch": 9.451188552965613,
+      "grad_norm": 3.380859375,
+      "learning_rate": 3.1767845230153435e-05,
+      "loss": 1.2283,
+      "step": 10238
+    },
+    {
+      "epoch": 9.452111700900069,
+      "grad_norm": 3.330078125,
+      "learning_rate": 3.176117411607739e-05,
+      "loss": 1.6538,
+      "step": 10239
+    },
+    {
+      "epoch": 9.453034848834525,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.175450300200133e-05,
+      "loss": 1.1548,
+      "step": 10240
+    },
+    {
+      "epoch": 9.453957996768983,
+      "grad_norm": 2.958984375,
+      "learning_rate": 3.1747831887925285e-05,
+      "loss": 1.5239,
+      "step": 10241
+    },
+    {
+      "epoch": 9.454881144703439,
+      "grad_norm": 3.51953125,
+      "learning_rate": 3.174116077384923e-05,
+      "loss": 1.2255,
+      "step": 10242
+    },
+    {
+      "epoch": 9.455804292637895,
+      "grad_norm": 4.09765625,
+      "learning_rate": 3.173448965977318e-05,
+      "loss": 1.3142,
+      "step": 10243
+    },
+    {
+      "epoch": 9.456727440572351,
+      "grad_norm": 3.25,
+      "learning_rate": 3.1727818545697135e-05,
+      "loss": 0.8662,
+      "step": 10244
+    },
+    {
+      "epoch": 9.457650588506809,
+      "grad_norm": 3.50390625,
+      "learning_rate": 3.172114743162109e-05,
+      "loss": 1.3707,
+      "step": 10245
+    },
+    {
+      "epoch": 9.458573736441265,
+      "grad_norm": 3.32421875,
+      "learning_rate": 3.171447631754503e-05,
+      "loss": 1.1434,
+      "step": 10246
+    },
+    {
+      "epoch": 9.45949688437572,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.1707805203468985e-05,
+      "loss": 1.0762,
+      "step": 10247
+    },
+    {
+      "epoch": 9.460420032310179,
+      "grad_norm": 2.982421875,
+      "learning_rate": 3.170113408939293e-05,
+      "loss": 1.4038,
+      "step": 10248
+    },
+    {
+      "epoch": 9.461343180244635,
+      "grad_norm": 3.55859375,
+      "learning_rate": 3.169446297531688e-05,
+      "loss": 1.261,
+      "step": 10249
+    },
+    {
+      "epoch": 9.46226632817909,
+      "grad_norm": 3.86328125,
+      "learning_rate": 3.168779186124083e-05,
+      "loss": 1.4412,
+      "step": 10250
+    },
+    {
+      "epoch": 9.463189476113547,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.168112074716478e-05,
+      "loss": 1.0714,
+      "step": 10251
+    },
+    {
+      "epoch": 9.464112624048004,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.1674449633088725e-05,
+      "loss": 1.4456,
+      "step": 10252
+    },
+    {
+      "epoch": 9.46503577198246,
+      "grad_norm": 3.625,
+      "learning_rate": 3.166777851901268e-05,
+      "loss": 1.1832,
+      "step": 10253
+    },
+    {
+      "epoch": 9.465958919916917,
+      "grad_norm": 3.07421875,
+      "learning_rate": 3.166110740493662e-05,
+      "loss": 1.1543,
+      "step": 10254
+    },
+    {
+      "epoch": 9.466882067851373,
+      "grad_norm": 3.115234375,
+      "learning_rate": 3.1654436290860575e-05,
+      "loss": 0.9807,
+      "step": 10255
+    },
+    {
+      "epoch": 9.46780521578583,
+      "grad_norm": 2.951171875,
+      "learning_rate": 3.164776517678452e-05,
+      "loss": 1.1966,
+      "step": 10256
+    },
+    {
+      "epoch": 9.468728363720286,
+      "grad_norm": 3.556640625,
+      "learning_rate": 3.164109406270848e-05,
+      "loss": 1.1851,
+      "step": 10257
+    },
+    {
+      "epoch": 9.469651511654742,
+      "grad_norm": 3.494140625,
+      "learning_rate": 3.1634422948632425e-05,
+      "loss": 1.1237,
+      "step": 10258
+    },
+    {
+      "epoch": 9.470574659589198,
+      "grad_norm": 3.005859375,
+      "learning_rate": 3.162775183455638e-05,
+      "loss": 1.1458,
+      "step": 10259
+    },
+    {
+      "epoch": 9.471497807523656,
+      "grad_norm": 3.69921875,
+      "learning_rate": 3.162108072048032e-05,
+      "loss": 1.3284,
+      "step": 10260
+    },
+    {
+      "epoch": 9.472420955458112,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.1614409606404275e-05,
+      "loss": 1.1923,
+      "step": 10261
+    },
+    {
+      "epoch": 9.473344103392568,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.160773849232822e-05,
+      "loss": 1.1267,
+      "step": 10262
+    },
+    {
+      "epoch": 9.474267251327024,
+      "grad_norm": 3.16015625,
+      "learning_rate": 3.160106737825217e-05,
+      "loss": 1.3523,
+      "step": 10263
+    },
+    {
+      "epoch": 9.475190399261482,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.159439626417612e-05,
+      "loss": 1.2616,
+      "step": 10264
+    },
+    {
+      "epoch": 9.476113547195938,
+      "grad_norm": 3.095703125,
+      "learning_rate": 3.158772515010007e-05,
+      "loss": 1.176,
+      "step": 10265
+    },
+    {
+      "epoch": 9.477036695130394,
+      "grad_norm": 3.216796875,
+      "learning_rate": 3.1581054036024016e-05,
+      "loss": 1.542,
+      "step": 10266
+    },
+    {
+      "epoch": 9.477959843064852,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.157438292194797e-05,
+      "loss": 1.1499,
+      "step": 10267
+    },
+    {
+      "epoch": 9.478882990999308,
+      "grad_norm": 4.29296875,
+      "learning_rate": 3.156771180787191e-05,
+      "loss": 1.3396,
+      "step": 10268
+    },
+    {
+      "epoch": 9.479806138933764,
+      "grad_norm": 2.931640625,
+      "learning_rate": 3.1561040693795865e-05,
+      "loss": 1.4751,
+      "step": 10269
+    },
+    {
+      "epoch": 9.48072928686822,
+      "grad_norm": 3.51953125,
+      "learning_rate": 3.155436957971982e-05,
+      "loss": 1.4294,
+      "step": 10270
+    },
+    {
+      "epoch": 9.481652434802678,
+      "grad_norm": 2.908203125,
+      "learning_rate": 3.154769846564376e-05,
+      "loss": 0.9788,
+      "step": 10271
+    },
+    {
+      "epoch": 9.482575582737134,
+      "grad_norm": 3.193359375,
+      "learning_rate": 3.1541027351567715e-05,
+      "loss": 1.2071,
+      "step": 10272
+    },
+    {
+      "epoch": 9.48349873067159,
+      "grad_norm": 3.125,
+      "learning_rate": 3.153435623749166e-05,
+      "loss": 1.4651,
+      "step": 10273
+    },
+    {
+      "epoch": 9.484421878606046,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.152768512341561e-05,
+      "loss": 1.7043,
+      "step": 10274
+    },
+    {
+      "epoch": 9.485345026540504,
+      "grad_norm": 3.224609375,
+      "learning_rate": 3.152101400933956e-05,
+      "loss": 1.0105,
+      "step": 10275
+    },
+    {
+      "epoch": 9.48626817447496,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.151434289526351e-05,
+      "loss": 1.0752,
+      "step": 10276
+    },
+    {
+      "epoch": 9.487191322409416,
+      "grad_norm": 3.177734375,
+      "learning_rate": 3.1507671781187456e-05,
+      "loss": 1.1355,
+      "step": 10277
+    },
+    {
+      "epoch": 9.488114470343872,
+      "grad_norm": 3.64453125,
+      "learning_rate": 3.150100066711141e-05,
+      "loss": 1.4116,
+      "step": 10278
+    },
+    {
+      "epoch": 9.48903761827833,
+      "grad_norm": 3.84765625,
+      "learning_rate": 3.1494329553035354e-05,
+      "loss": 1.3878,
+      "step": 10279
+    },
+    {
+      "epoch": 9.489960766212786,
+      "grad_norm": 3.37109375,
+      "learning_rate": 3.1487658438959306e-05,
+      "loss": 1.2681,
+      "step": 10280
+    },
+    {
+      "epoch": 9.490883914147242,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.148098732488325e-05,
+      "loss": 1.2407,
+      "step": 10281
+    },
+    {
+      "epoch": 9.4918070620817,
+      "grad_norm": 3.19921875,
+      "learning_rate": 3.147431621080721e-05,
+      "loss": 1.2747,
+      "step": 10282
+    },
+    {
+      "epoch": 9.492730210016155,
+      "grad_norm": 3.2890625,
+      "learning_rate": 3.1467645096731156e-05,
+      "loss": 1.2419,
+      "step": 10283
+    },
+    {
+      "epoch": 9.493653357950611,
+      "grad_norm": 2.677734375,
+      "learning_rate": 3.146097398265511e-05,
+      "loss": 1.3293,
+      "step": 10284
+    },
+    {
+      "epoch": 9.494576505885068,
+      "grad_norm": 2.92578125,
+      "learning_rate": 3.145430286857905e-05,
+      "loss": 1.4722,
+      "step": 10285
+    },
+    {
+      "epoch": 9.495499653819525,
+      "grad_norm": 3.3828125,
+      "learning_rate": 3.1447631754503005e-05,
+      "loss": 1.3936,
+      "step": 10286
+    },
+    {
+      "epoch": 9.496422801753981,
+      "grad_norm": 3.2890625,
+      "learning_rate": 3.144096064042695e-05,
+      "loss": 1.1785,
+      "step": 10287
+    },
+    {
+      "epoch": 9.497345949688437,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.14342895263509e-05,
+      "loss": 0.9052,
+      "step": 10288
+    },
+    {
+      "epoch": 9.498269097622893,
+      "grad_norm": 3.43359375,
+      "learning_rate": 3.142761841227485e-05,
+      "loss": 1.1483,
+      "step": 10289
+    },
+    {
+      "epoch": 9.499192245557351,
+      "grad_norm": 3.451171875,
+      "learning_rate": 3.14209472981988e-05,
+      "loss": 1.3506,
+      "step": 10290
+    },
+    {
+      "epoch": 9.500115393491807,
+      "grad_norm": 3.2265625,
+      "learning_rate": 3.1414276184122746e-05,
+      "loss": 1.1937,
+      "step": 10291
+    },
+    {
+      "epoch": 9.501038541426263,
+      "grad_norm": 2.74609375,
+      "learning_rate": 3.14076050700467e-05,
+      "loss": 1.2579,
+      "step": 10292
+    },
+    {
+      "epoch": 9.50196168936072,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.1400933955970644e-05,
+      "loss": 1.1653,
+      "step": 10293
+    },
+    {
+      "epoch": 9.502884837295177,
+      "grad_norm": 3.1015625,
+      "learning_rate": 3.13942628418946e-05,
+      "loss": 1.4409,
+      "step": 10294
+    },
+    {
+      "epoch": 9.503807985229633,
+      "grad_norm": 3.98046875,
+      "learning_rate": 3.138759172781855e-05,
+      "loss": 1.3542,
+      "step": 10295
+    },
+    {
+      "epoch": 9.50473113316409,
+      "grad_norm": 3.150390625,
+      "learning_rate": 3.13809206137425e-05,
+      "loss": 1.3989,
+      "step": 10296
+    },
+    {
+      "epoch": 9.505654281098547,
+      "grad_norm": 3.173828125,
+      "learning_rate": 3.1374249499666446e-05,
+      "loss": 1.0402,
+      "step": 10297
+    },
+    {
+      "epoch": 9.506577429033003,
+      "grad_norm": 4.11328125,
+      "learning_rate": 3.13675783855904e-05,
+      "loss": 1.3174,
+      "step": 10298
+    },
+    {
+      "epoch": 9.507500576967459,
+      "grad_norm": 3.107421875,
+      "learning_rate": 3.1360907271514343e-05,
+      "loss": 1.0896,
+      "step": 10299
+    },
+    {
+      "epoch": 9.508423724901915,
+      "grad_norm": 2.994140625,
+      "learning_rate": 3.1354236157438296e-05,
+      "loss": 1.3621,
+      "step": 10300
+    },
+    {
+      "epoch": 9.509346872836373,
+      "grad_norm": 3.76171875,
+      "learning_rate": 3.134756504336224e-05,
+      "loss": 1.3081,
+      "step": 10301
+    },
+    {
+      "epoch": 9.510270020770829,
+      "grad_norm": 3.28515625,
+      "learning_rate": 3.134089392928619e-05,
+      "loss": 1.0826,
+      "step": 10302
+    },
+    {
+      "epoch": 9.511193168705285,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.133422281521014e-05,
+      "loss": 1.434,
+      "step": 10303
+    },
+    {
+      "epoch": 9.51211631663974,
+      "grad_norm": 3.435546875,
+      "learning_rate": 3.132755170113409e-05,
+      "loss": 1.1682,
+      "step": 10304
+    },
+    {
+      "epoch": 9.513039464574199,
+      "grad_norm": 3.576171875,
+      "learning_rate": 3.1320880587058036e-05,
+      "loss": 1.3267,
+      "step": 10305
+    },
+    {
+      "epoch": 9.513962612508655,
+      "grad_norm": 3.509765625,
+      "learning_rate": 3.131420947298199e-05,
+      "loss": 1.3264,
+      "step": 10306
+    },
+    {
+      "epoch": 9.51488576044311,
+      "grad_norm": 3.763671875,
+      "learning_rate": 3.130753835890594e-05,
+      "loss": 1.2758,
+      "step": 10307
+    },
+    {
+      "epoch": 9.515808908377567,
+      "grad_norm": 3.9453125,
+      "learning_rate": 3.130086724482989e-05,
+      "loss": 1.4216,
+      "step": 10308
+    },
+    {
+      "epoch": 9.516732056312025,
+      "grad_norm": 3.18359375,
+      "learning_rate": 3.129419613075384e-05,
+      "loss": 1.488,
+      "step": 10309
+    },
+    {
+      "epoch": 9.51765520424648,
+      "grad_norm": 3.2890625,
+      "learning_rate": 3.128752501667779e-05,
+      "loss": 1.4294,
+      "step": 10310
+    },
+    {
+      "epoch": 9.518578352180937,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.1280853902601736e-05,
+      "loss": 1.257,
+      "step": 10311
+    },
+    {
+      "epoch": 9.519501500115393,
+      "grad_norm": 3.517578125,
+      "learning_rate": 3.127418278852569e-05,
+      "loss": 1.3672,
+      "step": 10312
+    },
+    {
+      "epoch": 9.52042464804985,
+      "grad_norm": 3.369140625,
+      "learning_rate": 3.1267511674449634e-05,
+      "loss": 1.2839,
+      "step": 10313
+    },
+    {
+      "epoch": 9.521347795984306,
+      "grad_norm": 3.251953125,
+      "learning_rate": 3.1260840560373586e-05,
+      "loss": 1.199,
+      "step": 10314
+    },
+    {
+      "epoch": 9.522270943918762,
+      "grad_norm": 3.181640625,
+      "learning_rate": 3.125416944629753e-05,
+      "loss": 1.2385,
+      "step": 10315
+    },
+    {
+      "epoch": 9.52319409185322,
+      "grad_norm": 3.892578125,
+      "learning_rate": 3.1247498332221484e-05,
+      "loss": 1.4387,
+      "step": 10316
+    },
+    {
+      "epoch": 9.524117239787676,
+      "grad_norm": 2.97265625,
+      "learning_rate": 3.124082721814543e-05,
+      "loss": 1.1929,
+      "step": 10317
+    },
+    {
+      "epoch": 9.525040387722132,
+      "grad_norm": 4.44140625,
+      "learning_rate": 3.123415610406938e-05,
+      "loss": 0.8823,
+      "step": 10318
+    },
+    {
+      "epoch": 9.525963535656588,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.1227484989993333e-05,
+      "loss": 1.179,
+      "step": 10319
+    },
+    {
+      "epoch": 9.526886683591046,
+      "grad_norm": 3.134765625,
+      "learning_rate": 3.122081387591728e-05,
+      "loss": 1.5234,
+      "step": 10320
+    },
+    {
+      "epoch": 9.527809831525502,
+      "grad_norm": 2.708984375,
+      "learning_rate": 3.121414276184123e-05,
+      "loss": 1.0807,
+      "step": 10321
+    },
+    {
+      "epoch": 9.528732979459958,
+      "grad_norm": 3.248046875,
+      "learning_rate": 3.1207471647765176e-05,
+      "loss": 1.53,
+      "step": 10322
+    },
+    {
+      "epoch": 9.529656127394414,
+      "grad_norm": 3.357421875,
+      "learning_rate": 3.120080053368913e-05,
+      "loss": 1.4126,
+      "step": 10323
+    },
+    {
+      "epoch": 9.530579275328872,
+      "grad_norm": 3.583984375,
+      "learning_rate": 3.1194129419613074e-05,
+      "loss": 1.1735,
+      "step": 10324
+    },
+    {
+      "epoch": 9.531502423263328,
+      "grad_norm": 3.681640625,
+      "learning_rate": 3.1187458305537026e-05,
+      "loss": 1.3364,
+      "step": 10325
+    },
+    {
+      "epoch": 9.532425571197784,
+      "grad_norm": 3.392578125,
+      "learning_rate": 3.118078719146097e-05,
+      "loss": 1.2332,
+      "step": 10326
+    },
+    {
+      "epoch": 9.53334871913224,
+      "grad_norm": 3.552734375,
+      "learning_rate": 3.1174116077384924e-05,
+      "loss": 1.3611,
+      "step": 10327
+    },
+    {
+      "epoch": 9.534271867066698,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.116744496330887e-05,
+      "loss": 1.3103,
+      "step": 10328
+    },
+    {
+      "epoch": 9.535195015001154,
+      "grad_norm": 3.322265625,
+      "learning_rate": 3.116077384923282e-05,
+      "loss": 1.5527,
+      "step": 10329
+    },
+    {
+      "epoch": 9.53611816293561,
+      "grad_norm": 2.943359375,
+      "learning_rate": 3.1154102735156774e-05,
+      "loss": 1.519,
+      "step": 10330
+    },
+    {
+      "epoch": 9.537041310870066,
+      "grad_norm": 3.34765625,
+      "learning_rate": 3.114743162108072e-05,
+      "loss": 1.3667,
+      "step": 10331
+    },
+    {
+      "epoch": 9.537964458804524,
+      "grad_norm": 3.37109375,
+      "learning_rate": 3.114076050700467e-05,
+      "loss": 1.3002,
+      "step": 10332
+    },
+    {
+      "epoch": 9.53888760673898,
+      "grad_norm": 3.470703125,
+      "learning_rate": 3.1134089392928624e-05,
+      "loss": 1.4165,
+      "step": 10333
+    },
+    {
+      "epoch": 9.539810754673436,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.112741827885257e-05,
+      "loss": 1.1667,
+      "step": 10334
+    },
+    {
+      "epoch": 9.540733902607894,
+      "grad_norm": 3.474609375,
+      "learning_rate": 3.112074716477652e-05,
+      "loss": 0.9045,
+      "step": 10335
+    },
+    {
+      "epoch": 9.54165705054235,
+      "grad_norm": 3.01953125,
+      "learning_rate": 3.111407605070047e-05,
+      "loss": 1.3008,
+      "step": 10336
+    },
+    {
+      "epoch": 9.542580198476806,
+      "grad_norm": 3.134765625,
+      "learning_rate": 3.110740493662442e-05,
+      "loss": 1.3015,
+      "step": 10337
+    },
+    {
+      "epoch": 9.543503346411262,
+      "grad_norm": 3.017578125,
+      "learning_rate": 3.1100733822548364e-05,
+      "loss": 1.2842,
+      "step": 10338
+    },
+    {
+      "epoch": 9.54442649434572,
+      "grad_norm": 3.541015625,
+      "learning_rate": 3.1094062708472317e-05,
+      "loss": 1.1027,
+      "step": 10339
+    },
+    {
+      "epoch": 9.545349642280176,
+      "grad_norm": 3.375,
+      "learning_rate": 3.108739159439626e-05,
+      "loss": 1.105,
+      "step": 10340
+    },
+    {
+      "epoch": 9.546272790214632,
+      "grad_norm": 3.30859375,
+      "learning_rate": 3.1080720480320214e-05,
+      "loss": 1.4402,
+      "step": 10341
+    },
+    {
+      "epoch": 9.547195938149088,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.107404936624416e-05,
+      "loss": 1.313,
+      "step": 10342
+    },
+    {
+      "epoch": 9.548119086083545,
+      "grad_norm": 3.22265625,
+      "learning_rate": 3.106737825216811e-05,
+      "loss": 1.3242,
+      "step": 10343
+    },
+    {
+      "epoch": 9.549042234018001,
+      "grad_norm": 3.259765625,
+      "learning_rate": 3.1060707138092064e-05,
+      "loss": 1.4871,
+      "step": 10344
+    },
+    {
+      "epoch": 9.549965381952457,
+      "grad_norm": 3.0703125,
+      "learning_rate": 3.1054036024016016e-05,
+      "loss": 1.4089,
+      "step": 10345
+    },
+    {
+      "epoch": 9.550888529886915,
+      "grad_norm": 3.23828125,
+      "learning_rate": 3.104736490993996e-05,
+      "loss": 1.3704,
+      "step": 10346
+    },
+    {
+      "epoch": 9.551811677821371,
+      "grad_norm": 3.427734375,
+      "learning_rate": 3.1040693795863914e-05,
+      "loss": 0.9114,
+      "step": 10347
+    },
+    {
+      "epoch": 9.552734825755827,
+      "grad_norm": 3.38671875,
+      "learning_rate": 3.103402268178786e-05,
+      "loss": 1.3481,
+      "step": 10348
+    },
+    {
+      "epoch": 9.553657973690283,
+      "grad_norm": 3.271484375,
+      "learning_rate": 3.102735156771181e-05,
+      "loss": 1.1482,
+      "step": 10349
+    },
+    {
+      "epoch": 9.554581121624741,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.102068045363576e-05,
+      "loss": 1.3757,
+      "step": 10350
+    },
+    {
+      "epoch": 9.555504269559197,
+      "grad_norm": 3.50390625,
+      "learning_rate": 3.101400933955971e-05,
+      "loss": 1.304,
+      "step": 10351
+    },
+    {
+      "epoch": 9.556427417493653,
+      "grad_norm": 2.830078125,
+      "learning_rate": 3.1007338225483655e-05,
+      "loss": 1.248,
+      "step": 10352
+    },
+    {
+      "epoch": 9.55735056542811,
+      "grad_norm": 3.431640625,
+      "learning_rate": 3.100066711140761e-05,
+      "loss": 1.2793,
+      "step": 10353
+    },
+    {
+      "epoch": 9.558273713362567,
+      "grad_norm": 3.560546875,
+      "learning_rate": 3.099399599733155e-05,
+      "loss": 1.3652,
+      "step": 10354
+    },
+    {
+      "epoch": 9.559196861297023,
+      "grad_norm": 3.6640625,
+      "learning_rate": 3.0987324883255504e-05,
+      "loss": 1.2262,
+      "step": 10355
+    },
+    {
+      "epoch": 9.560120009231479,
+      "grad_norm": 3.279296875,
+      "learning_rate": 3.098065376917945e-05,
+      "loss": 1.2671,
+      "step": 10356
+    },
+    {
+      "epoch": 9.561043157165935,
+      "grad_norm": 3.73046875,
+      "learning_rate": 3.097398265510341e-05,
+      "loss": 1.1511,
+      "step": 10357
+    },
+    {
+      "epoch": 9.561966305100393,
+      "grad_norm": 2.896484375,
+      "learning_rate": 3.0967311541027354e-05,
+      "loss": 1.3569,
+      "step": 10358
+    },
+    {
+      "epoch": 9.562889453034849,
+      "grad_norm": 2.76171875,
+      "learning_rate": 3.0960640426951306e-05,
+      "loss": 0.8487,
+      "step": 10359
+    },
+    {
+      "epoch": 9.563812600969305,
+      "grad_norm": 3.458984375,
+      "learning_rate": 3.095396931287525e-05,
+      "loss": 1.4417,
+      "step": 10360
+    },
+    {
+      "epoch": 9.564735748903761,
+      "grad_norm": 3.142578125,
+      "learning_rate": 3.0947298198799204e-05,
+      "loss": 1.1902,
+      "step": 10361
+    },
+    {
+      "epoch": 9.565658896838219,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.094062708472315e-05,
+      "loss": 0.8538,
+      "step": 10362
+    },
+    {
+      "epoch": 9.566582044772675,
+      "grad_norm": 3.6015625,
+      "learning_rate": 3.09339559706471e-05,
+      "loss": 0.9404,
+      "step": 10363
+    },
+    {
+      "epoch": 9.56750519270713,
+      "grad_norm": 3.30078125,
+      "learning_rate": 3.092728485657105e-05,
+      "loss": 1.3264,
+      "step": 10364
+    },
+    {
+      "epoch": 9.568428340641589,
+      "grad_norm": 3.67578125,
+      "learning_rate": 3.0920613742495e-05,
+      "loss": 1.2228,
+      "step": 10365
+    },
+    {
+      "epoch": 9.569351488576045,
+      "grad_norm": 3.375,
+      "learning_rate": 3.0913942628418945e-05,
+      "loss": 1.1086,
+      "step": 10366
+    },
+    {
+      "epoch": 9.5702746365105,
+      "grad_norm": 3.041015625,
+      "learning_rate": 3.09072715143429e-05,
+      "loss": 1.489,
+      "step": 10367
+    },
+    {
+      "epoch": 9.571197784444957,
+      "grad_norm": 2.974609375,
+      "learning_rate": 3.090060040026684e-05,
+      "loss": 1.282,
+      "step": 10368
+    },
+    {
+      "epoch": 9.572120932379415,
+      "grad_norm": 3.478515625,
+      "learning_rate": 3.08939292861908e-05,
+      "loss": 1.3828,
+      "step": 10369
+    },
+    {
+      "epoch": 9.57304408031387,
+      "grad_norm": 3.427734375,
+      "learning_rate": 3.088725817211475e-05,
+      "loss": 1.3247,
+      "step": 10370
+    },
+    {
+      "epoch": 9.573967228248327,
+      "grad_norm": 3.25,
+      "learning_rate": 3.08805870580387e-05,
+      "loss": 1.3767,
+      "step": 10371
+    },
+    {
+      "epoch": 9.574890376182783,
+      "grad_norm": 3.509765625,
+      "learning_rate": 3.0873915943962645e-05,
+      "loss": 1.2422,
+      "step": 10372
+    },
+    {
+      "epoch": 9.57581352411724,
+      "grad_norm": 3.572265625,
+      "learning_rate": 3.08672448298866e-05,
+      "loss": 1.4033,
+      "step": 10373
+    },
+    {
+      "epoch": 9.576736672051696,
+      "grad_norm": 3.205078125,
+      "learning_rate": 3.086057371581054e-05,
+      "loss": 1.3037,
+      "step": 10374
+    },
+    {
+      "epoch": 9.577659819986152,
+      "grad_norm": 3.3671875,
+      "learning_rate": 3.0853902601734494e-05,
+      "loss": 1.3441,
+      "step": 10375
+    },
+    {
+      "epoch": 9.578582967920608,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.084723148765844e-05,
+      "loss": 1.1576,
+      "step": 10376
+    },
+    {
+      "epoch": 9.579506115855066,
+      "grad_norm": 4.05078125,
+      "learning_rate": 3.084056037358239e-05,
+      "loss": 1.2656,
+      "step": 10377
+    },
+    {
+      "epoch": 9.580429263789522,
+      "grad_norm": 3.107421875,
+      "learning_rate": 3.083388925950634e-05,
+      "loss": 1.4265,
+      "step": 10378
+    },
+    {
+      "epoch": 9.581352411723978,
+      "grad_norm": 3.62109375,
+      "learning_rate": 3.082721814543029e-05,
+      "loss": 1.4727,
+      "step": 10379
+    },
+    {
+      "epoch": 9.582275559658434,
+      "grad_norm": 3.353515625,
+      "learning_rate": 3.0820547031354235e-05,
+      "loss": 1.3433,
+      "step": 10380
+    },
+    {
+      "epoch": 9.583198707592892,
+      "grad_norm": 2.95703125,
+      "learning_rate": 3.081387591727819e-05,
+      "loss": 1.1613,
+      "step": 10381
+    },
+    {
+      "epoch": 9.584121855527348,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.080720480320214e-05,
+      "loss": 1.3665,
+      "step": 10382
+    },
+    {
+      "epoch": 9.585045003461804,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.0800533689126085e-05,
+      "loss": 1.0139,
+      "step": 10383
+    },
+    {
+      "epoch": 9.585968151396262,
+      "grad_norm": 3.45703125,
+      "learning_rate": 3.079386257505004e-05,
+      "loss": 1.2996,
+      "step": 10384
+    },
+    {
+      "epoch": 9.586891299330718,
+      "grad_norm": 2.974609375,
+      "learning_rate": 3.078719146097398e-05,
+      "loss": 1.5322,
+      "step": 10385
+    },
+    {
+      "epoch": 9.587814447265174,
+      "grad_norm": 3.728515625,
+      "learning_rate": 3.0780520346897935e-05,
+      "loss": 1.6167,
+      "step": 10386
+    },
+    {
+      "epoch": 9.58873759519963,
+      "grad_norm": 3.6796875,
+      "learning_rate": 3.077384923282188e-05,
+      "loss": 1.1941,
+      "step": 10387
+    },
+    {
+      "epoch": 9.589660743134088,
+      "grad_norm": 3.291015625,
+      "learning_rate": 3.076717811874583e-05,
+      "loss": 1.4705,
+      "step": 10388
+    },
+    {
+      "epoch": 9.590583891068544,
+      "grad_norm": 3.865234375,
+      "learning_rate": 3.076050700466978e-05,
+      "loss": 1.4399,
+      "step": 10389
+    },
+    {
+      "epoch": 9.591507039003,
+      "grad_norm": 3.087890625,
+      "learning_rate": 3.075383589059373e-05,
+      "loss": 1.215,
+      "step": 10390
+    },
+    {
+      "epoch": 9.592430186937456,
+      "grad_norm": 3.248046875,
+      "learning_rate": 3.0747164776517675e-05,
+      "loss": 1.3281,
+      "step": 10391
+    },
+    {
+      "epoch": 9.593353334871914,
+      "grad_norm": 3.705078125,
+      "learning_rate": 3.074049366244163e-05,
+      "loss": 1.1124,
+      "step": 10392
+    },
+    {
+      "epoch": 9.59427648280637,
+      "grad_norm": 2.966796875,
+      "learning_rate": 3.073382254836557e-05,
+      "loss": 1.1637,
+      "step": 10393
+    },
+    {
+      "epoch": 9.595199630740826,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.072715143428953e-05,
+      "loss": 1.1296,
+      "step": 10394
+    },
+    {
+      "epoch": 9.596122778675284,
+      "grad_norm": 3.19140625,
+      "learning_rate": 3.072048032021348e-05,
+      "loss": 1.1425,
+      "step": 10395
+    },
+    {
+      "epoch": 9.59704592660974,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.071380920613743e-05,
+      "loss": 1.1947,
+      "step": 10396
+    },
+    {
+      "epoch": 9.597969074544196,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.0707138092061375e-05,
+      "loss": 1.2424,
+      "step": 10397
+    },
+    {
+      "epoch": 9.598892222478652,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.070046697798533e-05,
+      "loss": 1.4399,
+      "step": 10398
+    },
+    {
+      "epoch": 9.59981537041311,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.069379586390927e-05,
+      "loss": 1.4485,
+      "step": 10399
+    },
+    {
+      "epoch": 9.600738518347566,
+      "grad_norm": 3.1796875,
+      "learning_rate": 3.0687124749833225e-05,
+      "loss": 1.4766,
+      "step": 10400
+    },
+    {
+      "epoch": 9.601661666282022,
+      "grad_norm": 3.67578125,
+      "learning_rate": 3.068045363575717e-05,
+      "loss": 1.4222,
+      "step": 10401
+    },
+    {
+      "epoch": 9.602584814216478,
+      "grad_norm": 3.466796875,
+      "learning_rate": 3.067378252168112e-05,
+      "loss": 1.1292,
+      "step": 10402
+    },
+    {
+      "epoch": 9.603507962150935,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.066711140760507e-05,
+      "loss": 1.3237,
+      "step": 10403
+    },
+    {
+      "epoch": 9.604431110085391,
+      "grad_norm": 3.1953125,
+      "learning_rate": 3.066044029352902e-05,
+      "loss": 1.0739,
+      "step": 10404
+    },
+    {
+      "epoch": 9.605354258019847,
+      "grad_norm": 3.46484375,
+      "learning_rate": 3.0653769179452966e-05,
+      "loss": 1.0332,
+      "step": 10405
+    },
+    {
+      "epoch": 9.606277405954303,
+      "grad_norm": 3.490234375,
+      "learning_rate": 3.064709806537692e-05,
+      "loss": 1.2349,
+      "step": 10406
+    },
+    {
+      "epoch": 9.607200553888761,
+      "grad_norm": 3.22265625,
+      "learning_rate": 3.064042695130087e-05,
+      "loss": 1.3777,
+      "step": 10407
+    },
+    {
+      "epoch": 9.608123701823217,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.063375583722482e-05,
+      "loss": 1.2939,
+      "step": 10408
+    },
+    {
+      "epoch": 9.609046849757673,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.062708472314877e-05,
+      "loss": 1.3132,
+      "step": 10409
+    },
+    {
+      "epoch": 9.60996999769213,
+      "grad_norm": 3.33203125,
+      "learning_rate": 3.062041360907272e-05,
+      "loss": 1.3269,
+      "step": 10410
+    },
+    {
+      "epoch": 9.610893145626587,
+      "grad_norm": 8.46875,
+      "learning_rate": 3.0613742494996665e-05,
+      "loss": 1.113,
+      "step": 10411
+    },
+    {
+      "epoch": 9.611816293561043,
+      "grad_norm": 2.962890625,
+      "learning_rate": 3.060707138092062e-05,
+      "loss": 1.3594,
+      "step": 10412
+    },
+    {
+      "epoch": 9.6127394414955,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.060040026684456e-05,
+      "loss": 1.3289,
+      "step": 10413
+    },
+    {
+      "epoch": 9.613662589429957,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.0593729152768515e-05,
+      "loss": 1.5291,
+      "step": 10414
+    },
+    {
+      "epoch": 9.614585737364413,
+      "grad_norm": 3.650390625,
+      "learning_rate": 3.058705803869246e-05,
+      "loss": 1.3508,
+      "step": 10415
+    },
+    {
+      "epoch": 9.615508885298869,
+      "grad_norm": 3.0859375,
+      "learning_rate": 3.058038692461641e-05,
+      "loss": 1.2866,
+      "step": 10416
+    },
+    {
+      "epoch": 9.616432033233325,
+      "grad_norm": 3.404296875,
+      "learning_rate": 3.057371581054036e-05,
+      "loss": 1.1571,
+      "step": 10417
+    },
+    {
+      "epoch": 9.617355181167783,
+      "grad_norm": 3.00390625,
+      "learning_rate": 3.056704469646431e-05,
+      "loss": 1.276,
+      "step": 10418
+    },
+    {
+      "epoch": 9.618278329102239,
+      "grad_norm": 3.560546875,
+      "learning_rate": 3.0560373582388256e-05,
+      "loss": 1.27,
+      "step": 10419
+    },
+    {
+      "epoch": 9.619201477036695,
+      "grad_norm": 3.595703125,
+      "learning_rate": 3.0553702468312215e-05,
+      "loss": 1.2727,
+      "step": 10420
+    },
+    {
+      "epoch": 9.620124624971151,
+      "grad_norm": 3.517578125,
+      "learning_rate": 3.054703135423616e-05,
+      "loss": 1.4147,
+      "step": 10421
+    },
+    {
+      "epoch": 9.621047772905609,
+      "grad_norm": 3.689453125,
+      "learning_rate": 3.054036024016011e-05,
+      "loss": 1.5137,
+      "step": 10422
+    },
+    {
+      "epoch": 9.621970920840065,
+      "grad_norm": 3.310546875,
+      "learning_rate": 3.053368912608406e-05,
+      "loss": 1.2075,
+      "step": 10423
+    },
+    {
+      "epoch": 9.62289406877452,
+      "grad_norm": 3.212890625,
+      "learning_rate": 3.052701801200801e-05,
+      "loss": 1.2874,
+      "step": 10424
+    },
+    {
+      "epoch": 9.623817216708977,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.0520346897931956e-05,
+      "loss": 1.5002,
+      "step": 10425
+    },
+    {
+      "epoch": 9.624740364643435,
+      "grad_norm": 3.48046875,
+      "learning_rate": 3.0513675783855904e-05,
+      "loss": 1.2144,
+      "step": 10426
+    },
+    {
+      "epoch": 9.62566351257789,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.0507004669779853e-05,
+      "loss": 1.4368,
+      "step": 10427
+    },
+    {
+      "epoch": 9.626586660512347,
+      "grad_norm": 3.1015625,
+      "learning_rate": 3.0500333555703802e-05,
+      "loss": 1.4126,
+      "step": 10428
+    },
+    {
+      "epoch": 9.627509808446803,
+      "grad_norm": 3.880859375,
+      "learning_rate": 3.049366244162775e-05,
+      "loss": 1.3844,
+      "step": 10429
+    },
+    {
+      "epoch": 9.62843295638126,
+      "grad_norm": 3.15234375,
+      "learning_rate": 3.04869913275517e-05,
+      "loss": 1.0627,
+      "step": 10430
+    },
+    {
+      "epoch": 9.629356104315717,
+      "grad_norm": 3.10546875,
+      "learning_rate": 3.048032021347565e-05,
+      "loss": 1.3625,
+      "step": 10431
+    },
+    {
+      "epoch": 9.630279252250173,
+      "grad_norm": 3.986328125,
+      "learning_rate": 3.0473649099399604e-05,
+      "loss": 1.3016,
+      "step": 10432
+    },
+    {
+      "epoch": 9.63120240018463,
+      "grad_norm": 2.591796875,
+      "learning_rate": 3.0466977985323553e-05,
+      "loss": 0.9712,
+      "step": 10433
+    },
+    {
+      "epoch": 9.632125548119086,
+      "grad_norm": 3.306640625,
+      "learning_rate": 3.0460306871247502e-05,
+      "loss": 1.3259,
+      "step": 10434
+    },
+    {
+      "epoch": 9.633048696053542,
+      "grad_norm": 3.158203125,
+      "learning_rate": 3.045363575717145e-05,
+      "loss": 1.4692,
+      "step": 10435
+    },
+    {
+      "epoch": 9.633971843987998,
+      "grad_norm": 3.92578125,
+      "learning_rate": 3.04469646430954e-05,
+      "loss": 1.3157,
+      "step": 10436
+    },
+    {
+      "epoch": 9.634894991922456,
+      "grad_norm": 3.244140625,
+      "learning_rate": 3.0440293529019348e-05,
+      "loss": 1.1305,
+      "step": 10437
+    },
+    {
+      "epoch": 9.635818139856912,
+      "grad_norm": 3.1328125,
+      "learning_rate": 3.0433622414943297e-05,
+      "loss": 1.52,
+      "step": 10438
+    },
+    {
+      "epoch": 9.636741287791368,
+      "grad_norm": 3.009765625,
+      "learning_rate": 3.0426951300867246e-05,
+      "loss": 1.4146,
+      "step": 10439
+    },
+    {
+      "epoch": 9.637664435725824,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.0420280186791195e-05,
+      "loss": 1.3997,
+      "step": 10440
+    },
+    {
+      "epoch": 9.638587583660282,
+      "grad_norm": 3.439453125,
+      "learning_rate": 3.0413609072715144e-05,
+      "loss": 1.2545,
+      "step": 10441
+    },
+    {
+      "epoch": 9.639510731594738,
+      "grad_norm": 2.892578125,
+      "learning_rate": 3.0406937958639092e-05,
+      "loss": 1.3091,
+      "step": 10442
+    },
+    {
+      "epoch": 9.640433879529194,
+      "grad_norm": 3.435546875,
+      "learning_rate": 3.040026684456304e-05,
+      "loss": 1.3365,
+      "step": 10443
+    },
+    {
+      "epoch": 9.641357027463652,
+      "grad_norm": 2.955078125,
+      "learning_rate": 3.039359573048699e-05,
+      "loss": 1.1403,
+      "step": 10444
+    },
+    {
+      "epoch": 9.642280175398108,
+      "grad_norm": 3.04296875,
+      "learning_rate": 3.0386924616410946e-05,
+      "loss": 1.2526,
+      "step": 10445
+    },
+    {
+      "epoch": 9.643203323332564,
+      "grad_norm": 3.1328125,
+      "learning_rate": 3.0380253502334894e-05,
+      "loss": 1.2048,
+      "step": 10446
+    },
+    {
+      "epoch": 9.64412647126702,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.0373582388258843e-05,
+      "loss": 1.2433,
+      "step": 10447
+    },
+    {
+      "epoch": 9.645049619201478,
+      "grad_norm": 2.697265625,
+      "learning_rate": 3.0366911274182792e-05,
+      "loss": 1.1035,
+      "step": 10448
+    },
+    {
+      "epoch": 9.645972767135934,
+      "grad_norm": 3.98046875,
+      "learning_rate": 3.036024016010674e-05,
+      "loss": 1.2177,
+      "step": 10449
+    },
+    {
+      "epoch": 9.64689591507039,
+      "grad_norm": 3.634765625,
+      "learning_rate": 3.035356904603069e-05,
+      "loss": 0.9143,
+      "step": 10450
+    },
+    {
+      "epoch": 9.647819063004846,
+      "grad_norm": 2.98046875,
+      "learning_rate": 3.034689793195464e-05,
+      "loss": 1.3416,
+      "step": 10451
+    },
+    {
+      "epoch": 9.648742210939304,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.0340226817878587e-05,
+      "loss": 1.3054,
+      "step": 10452
+    },
+    {
+      "epoch": 9.64966535887376,
+      "grad_norm": 3.01171875,
+      "learning_rate": 3.0333555703802536e-05,
+      "loss": 0.9551,
+      "step": 10453
+    },
+    {
+      "epoch": 9.650588506808216,
+      "grad_norm": 3.697265625,
+      "learning_rate": 3.0326884589726485e-05,
+      "loss": 1.4561,
+      "step": 10454
+    },
+    {
+      "epoch": 9.651511654742672,
+      "grad_norm": 3.0078125,
+      "learning_rate": 3.0320213475650434e-05,
+      "loss": 1.4543,
+      "step": 10455
+    },
+    {
+      "epoch": 9.65243480267713,
+      "grad_norm": 3.5,
+      "learning_rate": 3.0313542361574383e-05,
+      "loss": 1.2023,
+      "step": 10456
+    },
+    {
+      "epoch": 9.653357950611586,
+      "grad_norm": 2.93359375,
+      "learning_rate": 3.0306871247498335e-05,
+      "loss": 1.1631,
+      "step": 10457
+    },
+    {
+      "epoch": 9.654281098546042,
+      "grad_norm": 3.419921875,
+      "learning_rate": 3.0300200133422284e-05,
+      "loss": 0.8075,
+      "step": 10458
+    },
+    {
+      "epoch": 9.655204246480498,
+      "grad_norm": 2.966796875,
+      "learning_rate": 3.0293529019346232e-05,
+      "loss": 1.3605,
+      "step": 10459
+    },
+    {
+      "epoch": 9.656127394414955,
+      "grad_norm": 3.435546875,
+      "learning_rate": 3.028685790527018e-05,
+      "loss": 1.4495,
+      "step": 10460
+    },
+    {
+      "epoch": 9.657050542349412,
+      "grad_norm": 3.41796875,
+      "learning_rate": 3.028018679119413e-05,
+      "loss": 1.2185,
+      "step": 10461
+    },
+    {
+      "epoch": 9.657973690283868,
+      "grad_norm": 3.677734375,
+      "learning_rate": 3.0273515677118082e-05,
+      "loss": 1.3069,
+      "step": 10462
+    },
+    {
+      "epoch": 9.658896838218325,
+      "grad_norm": 3.3671875,
+      "learning_rate": 3.026684456304203e-05,
+      "loss": 1.5291,
+      "step": 10463
+    },
+    {
+      "epoch": 9.659819986152781,
+      "grad_norm": 3.12890625,
+      "learning_rate": 3.026017344896598e-05,
+      "loss": 1.2375,
+      "step": 10464
+    },
+    {
+      "epoch": 9.660743134087237,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.025350233488993e-05,
+      "loss": 1.3105,
+      "step": 10465
+    },
+    {
+      "epoch": 9.661666282021693,
+      "grad_norm": 3.337890625,
+      "learning_rate": 3.0246831220813878e-05,
+      "loss": 1.3694,
+      "step": 10466
+    },
+    {
+      "epoch": 9.662589429956151,
+      "grad_norm": 3.380859375,
+      "learning_rate": 3.0240160106737826e-05,
+      "loss": 1.2537,
+      "step": 10467
+    },
+    {
+      "epoch": 9.663512577890607,
+      "grad_norm": 3.556640625,
+      "learning_rate": 3.0233488992661775e-05,
+      "loss": 1.016,
+      "step": 10468
+    },
+    {
+      "epoch": 9.664435725825063,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.0226817878585724e-05,
+      "loss": 1.3793,
+      "step": 10469
+    },
+    {
+      "epoch": 9.66535887375952,
+      "grad_norm": 3.7109375,
+      "learning_rate": 3.0220146764509676e-05,
+      "loss": 1.4106,
+      "step": 10470
+    },
+    {
+      "epoch": 9.666282021693977,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.0213475650433625e-05,
+      "loss": 1.5281,
+      "step": 10471
+    },
+    {
+      "epoch": 9.667205169628433,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.0206804536357574e-05,
+      "loss": 1.3589,
+      "step": 10472
+    },
+    {
+      "epoch": 9.66812831756289,
+      "grad_norm": 3.396484375,
+      "learning_rate": 3.0200133422281523e-05,
+      "loss": 1.3081,
+      "step": 10473
+    },
+    {
+      "epoch": 9.669051465497345,
+      "grad_norm": 3.34765625,
+      "learning_rate": 3.019346230820547e-05,
+      "loss": 1.0297,
+      "step": 10474
+    },
+    {
+      "epoch": 9.669974613431803,
+      "grad_norm": 3.04296875,
+      "learning_rate": 3.018679119412942e-05,
+      "loss": 1.4329,
+      "step": 10475
+    },
+    {
+      "epoch": 9.670897761366259,
+      "grad_norm": 2.9609375,
+      "learning_rate": 3.018012008005337e-05,
+      "loss": 1.4409,
+      "step": 10476
+    },
+    {
+      "epoch": 9.671820909300715,
+      "grad_norm": 2.80078125,
+      "learning_rate": 3.0173448965977318e-05,
+      "loss": 1.3174,
+      "step": 10477
+    },
+    {
+      "epoch": 9.672744057235171,
+      "grad_norm": 3.55859375,
+      "learning_rate": 3.0166777851901267e-05,
+      "loss": 1.2522,
+      "step": 10478
+    },
+    {
+      "epoch": 9.673667205169629,
+      "grad_norm": 3.021484375,
+      "learning_rate": 3.0160106737825216e-05,
+      "loss": 1.2971,
+      "step": 10479
+    },
+    {
+      "epoch": 9.674590353104085,
+      "grad_norm": 3.41796875,
+      "learning_rate": 3.0153435623749164e-05,
+      "loss": 1.2362,
+      "step": 10480
+    },
+    {
+      "epoch": 9.675513501038541,
+      "grad_norm": 3.26953125,
+      "learning_rate": 3.0146764509673113e-05,
+      "loss": 1.1245,
+      "step": 10481
+    },
+    {
+      "epoch": 9.676436648972999,
+      "grad_norm": 4.04296875,
+      "learning_rate": 3.014009339559707e-05,
+      "loss": 1.3481,
+      "step": 10482
+    },
+    {
+      "epoch": 9.677359796907455,
+      "grad_norm": 3.310546875,
+      "learning_rate": 3.0133422281521018e-05,
+      "loss": 1.3469,
+      "step": 10483
+    },
+    {
+      "epoch": 9.67828294484191,
+      "grad_norm": 3.283203125,
+      "learning_rate": 3.0126751167444966e-05,
+      "loss": 1.2266,
+      "step": 10484
+    },
+    {
+      "epoch": 9.679206092776367,
+      "grad_norm": 3.017578125,
+      "learning_rate": 3.0120080053368915e-05,
+      "loss": 1.447,
+      "step": 10485
+    },
+    {
+      "epoch": 9.680129240710825,
+      "grad_norm": 3.4453125,
+      "learning_rate": 3.0113408939292864e-05,
+      "loss": 1.3533,
+      "step": 10486
+    },
+    {
+      "epoch": 9.68105238864528,
+      "grad_norm": 3.1640625,
+      "learning_rate": 3.0106737825216813e-05,
+      "loss": 1.3928,
+      "step": 10487
+    },
+    {
+      "epoch": 9.681975536579737,
+      "grad_norm": 3.654296875,
+      "learning_rate": 3.0100066711140762e-05,
+      "loss": 1.2625,
+      "step": 10488
+    },
+    {
+      "epoch": 9.682898684514193,
+      "grad_norm": 3.353515625,
+      "learning_rate": 3.009339559706471e-05,
+      "loss": 1.1178,
+      "step": 10489
+    },
+    {
+      "epoch": 9.68382183244865,
+      "grad_norm": 3.431640625,
+      "learning_rate": 3.008672448298866e-05,
+      "loss": 1.4216,
+      "step": 10490
+    },
+    {
+      "epoch": 9.684744980383106,
+      "grad_norm": 3.052734375,
+      "learning_rate": 3.0080053368912608e-05,
+      "loss": 1.5896,
+      "step": 10491
+    },
+    {
+      "epoch": 9.685668128317563,
+      "grad_norm": 3.66015625,
+      "learning_rate": 3.0073382254836557e-05,
+      "loss": 1.3928,
+      "step": 10492
+    },
+    {
+      "epoch": 9.68659127625202,
+      "grad_norm": 3.298828125,
+      "learning_rate": 3.0066711140760506e-05,
+      "loss": 1.5593,
+      "step": 10493
+    },
+    {
+      "epoch": 9.687514424186476,
+      "grad_norm": 3.517578125,
+      "learning_rate": 3.0060040026684455e-05,
+      "loss": 1.2589,
+      "step": 10494
+    },
+    {
+      "epoch": 9.688437572120932,
+      "grad_norm": 3.166015625,
+      "learning_rate": 3.005336891260841e-05,
+      "loss": 1.3892,
+      "step": 10495
+    },
+    {
+      "epoch": 9.689360720055388,
+      "grad_norm": 3.130859375,
+      "learning_rate": 3.004669779853236e-05,
+      "loss": 1.4049,
+      "step": 10496
+    },
+    {
+      "epoch": 9.690283867989846,
+      "grad_norm": 3.4921875,
+      "learning_rate": 3.0040026684456308e-05,
+      "loss": 1.2794,
+      "step": 10497
+    },
+    {
+      "epoch": 9.691207015924302,
+      "grad_norm": 3.462890625,
+      "learning_rate": 3.0033355570380257e-05,
+      "loss": 1.0001,
+      "step": 10498
+    },
+    {
+      "epoch": 9.692130163858758,
+      "grad_norm": 4.09765625,
+      "learning_rate": 3.0026684456304205e-05,
+      "loss": 1.3311,
+      "step": 10499
+    },
+    {
+      "epoch": 9.693053311793214,
+      "grad_norm": 3.259765625,
+      "learning_rate": 3.0020013342228154e-05,
+      "loss": 1.1838,
+      "step": 10500
+    },
+    {
+      "epoch": 9.693976459727672,
+      "grad_norm": 3.400390625,
+      "learning_rate": 3.0013342228152103e-05,
+      "loss": 1.1421,
+      "step": 10501
+    },
+    {
+      "epoch": 9.694899607662128,
+      "grad_norm": 3.40234375,
+      "learning_rate": 3.0006671114076052e-05,
+      "loss": 1.3064,
+      "step": 10502
+    },
+    {
+      "epoch": 9.695822755596584,
+      "grad_norm": 2.96484375,
+      "learning_rate": 3e-05,
+      "loss": 1.2419,
+      "step": 10503
+    },
+    {
+      "epoch": 9.69674590353104,
+      "grad_norm": 3.68359375,
+      "learning_rate": 2.999332888592395e-05,
+      "loss": 1.3418,
+      "step": 10504
+    },
+    {
+      "epoch": 9.697669051465498,
+      "grad_norm": 3.431640625,
+      "learning_rate": 2.99866577718479e-05,
+      "loss": 1.1926,
+      "step": 10505
+    },
+    {
+      "epoch": 9.698592199399954,
+      "grad_norm": 3.220703125,
+      "learning_rate": 2.9979986657771847e-05,
+      "loss": 1.2516,
+      "step": 10506
+    },
+    {
+      "epoch": 9.69951534733441,
+      "grad_norm": 3.173828125,
+      "learning_rate": 2.9973315543695803e-05,
+      "loss": 1.2195,
+      "step": 10507
+    },
+    {
+      "epoch": 9.700438495268866,
+      "grad_norm": 3.17578125,
+      "learning_rate": 2.996664442961975e-05,
+      "loss": 1.321,
+      "step": 10508
+    },
+    {
+      "epoch": 9.701361643203324,
+      "grad_norm": 3.375,
+      "learning_rate": 2.99599733155437e-05,
+      "loss": 1.101,
+      "step": 10509
+    },
+    {
+      "epoch": 9.70228479113778,
+      "grad_norm": 3.09765625,
+      "learning_rate": 2.995330220146765e-05,
+      "loss": 1.4487,
+      "step": 10510
+    },
+    {
+      "epoch": 9.703207939072236,
+      "grad_norm": 3.404296875,
+      "learning_rate": 2.9946631087391598e-05,
+      "loss": 1.5684,
+      "step": 10511
+    },
+    {
+      "epoch": 9.704131087006694,
+      "grad_norm": 3.349609375,
+      "learning_rate": 2.9939959973315547e-05,
+      "loss": 1.2703,
+      "step": 10512
+    },
+    {
+      "epoch": 9.70505423494115,
+      "grad_norm": 2.66015625,
+      "learning_rate": 2.9933288859239496e-05,
+      "loss": 1.395,
+      "step": 10513
+    },
+    {
+      "epoch": 9.705977382875606,
+      "grad_norm": 3.443359375,
+      "learning_rate": 2.9926617745163445e-05,
+      "loss": 1.3606,
+      "step": 10514
+    },
+    {
+      "epoch": 9.706900530810062,
+      "grad_norm": 3.052734375,
+      "learning_rate": 2.9919946631087393e-05,
+      "loss": 1.096,
+      "step": 10515
+    },
+    {
+      "epoch": 9.70782367874452,
+      "grad_norm": 3.263671875,
+      "learning_rate": 2.9913275517011342e-05,
+      "loss": 1.4578,
+      "step": 10516
+    },
+    {
+      "epoch": 9.708746826678976,
+      "grad_norm": 3.400390625,
+      "learning_rate": 2.990660440293529e-05,
+      "loss": 1.2455,
+      "step": 10517
+    },
+    {
+      "epoch": 9.709669974613432,
+      "grad_norm": 3.43359375,
+      "learning_rate": 2.989993328885924e-05,
+      "loss": 1.3054,
+      "step": 10518
+    },
+    {
+      "epoch": 9.710593122547888,
+      "grad_norm": 3.755859375,
+      "learning_rate": 2.989326217478319e-05,
+      "loss": 1.4219,
+      "step": 10519
+    },
+    {
+      "epoch": 9.711516270482345,
+      "grad_norm": 3.5703125,
+      "learning_rate": 2.988659106070714e-05,
+      "loss": 1.448,
+      "step": 10520
+    },
+    {
+      "epoch": 9.712439418416801,
+      "grad_norm": 3.306640625,
+      "learning_rate": 2.987991994663109e-05,
+      "loss": 1.1965,
+      "step": 10521
+    },
+    {
+      "epoch": 9.713362566351257,
+      "grad_norm": 2.96484375,
+      "learning_rate": 2.987324883255504e-05,
+      "loss": 1.5933,
+      "step": 10522
+    },
+    {
+      "epoch": 9.714285714285714,
+      "grad_norm": 3.498046875,
+      "learning_rate": 2.9866577718478987e-05,
+      "loss": 1.4209,
+      "step": 10523
+    },
+    {
+      "epoch": 9.715208862220171,
+      "grad_norm": 3.474609375,
+      "learning_rate": 2.9859906604402936e-05,
+      "loss": 1.2529,
+      "step": 10524
+    },
+    {
+      "epoch": 9.716132010154627,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9853235490326885e-05,
+      "loss": 1.425,
+      "step": 10525
+    },
+    {
+      "epoch": 9.717055158089083,
+      "grad_norm": 3.44921875,
+      "learning_rate": 2.9846564376250834e-05,
+      "loss": 1.196,
+      "step": 10526
+    },
+    {
+      "epoch": 9.71797830602354,
+      "grad_norm": 3.25390625,
+      "learning_rate": 2.9839893262174783e-05,
+      "loss": 1.5027,
+      "step": 10527
+    },
+    {
+      "epoch": 9.718901453957997,
+      "grad_norm": 3.572265625,
+      "learning_rate": 2.983322214809873e-05,
+      "loss": 1.3286,
+      "step": 10528
+    },
+    {
+      "epoch": 9.719824601892453,
+      "grad_norm": 3.087890625,
+      "learning_rate": 2.982655103402268e-05,
+      "loss": 1.2736,
+      "step": 10529
+    },
+    {
+      "epoch": 9.72074774982691,
+      "grad_norm": 3.82421875,
+      "learning_rate": 2.981987991994663e-05,
+      "loss": 1.3794,
+      "step": 10530
+    },
+    {
+      "epoch": 9.721670897761367,
+      "grad_norm": 3.3984375,
+      "learning_rate": 2.9813208805870578e-05,
+      "loss": 1.3634,
+      "step": 10531
+    },
+    {
+      "epoch": 9.722594045695823,
+      "grad_norm": 2.869140625,
+      "learning_rate": 2.9806537691794533e-05,
+      "loss": 1.2496,
+      "step": 10532
+    },
+    {
+      "epoch": 9.723517193630279,
+      "grad_norm": 3.24609375,
+      "learning_rate": 2.9799866577718482e-05,
+      "loss": 1.5657,
+      "step": 10533
+    },
+    {
+      "epoch": 9.724440341564735,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.979319546364243e-05,
+      "loss": 1.2881,
+      "step": 10534
+    },
+    {
+      "epoch": 9.725363489499193,
+      "grad_norm": 3.458984375,
+      "learning_rate": 2.978652434956638e-05,
+      "loss": 1.2407,
+      "step": 10535
+    },
+    {
+      "epoch": 9.726286637433649,
+      "grad_norm": 3.0703125,
+      "learning_rate": 2.977985323549033e-05,
+      "loss": 1.2876,
+      "step": 10536
+    },
+    {
+      "epoch": 9.727209785368105,
+      "grad_norm": 3.583984375,
+      "learning_rate": 2.9773182121414278e-05,
+      "loss": 1.3455,
+      "step": 10537
+    },
+    {
+      "epoch": 9.728132933302561,
+      "grad_norm": 3.02734375,
+      "learning_rate": 2.9766511007338226e-05,
+      "loss": 1.2366,
+      "step": 10538
+    },
+    {
+      "epoch": 9.729056081237019,
+      "grad_norm": 4.55078125,
+      "learning_rate": 2.9759839893262175e-05,
+      "loss": 1.1852,
+      "step": 10539
+    },
+    {
+      "epoch": 9.729979229171475,
+      "grad_norm": 2.986328125,
+      "learning_rate": 2.9753168779186124e-05,
+      "loss": 1.416,
+      "step": 10540
+    },
+    {
+      "epoch": 9.73090237710593,
+      "grad_norm": 3.73046875,
+      "learning_rate": 2.9746497665110073e-05,
+      "loss": 1.4243,
+      "step": 10541
+    },
+    {
+      "epoch": 9.731825525040387,
+      "grad_norm": 4.11328125,
+      "learning_rate": 2.973982655103402e-05,
+      "loss": 1.1874,
+      "step": 10542
+    },
+    {
+      "epoch": 9.732748672974845,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.973315543695797e-05,
+      "loss": 1.3047,
+      "step": 10543
+    },
+    {
+      "epoch": 9.7336718209093,
+      "grad_norm": 3.3671875,
+      "learning_rate": 2.972648432288192e-05,
+      "loss": 1.0719,
+      "step": 10544
+    },
+    {
+      "epoch": 9.734594968843757,
+      "grad_norm": 3.294921875,
+      "learning_rate": 2.9719813208805875e-05,
+      "loss": 1.4146,
+      "step": 10545
+    },
+    {
+      "epoch": 9.735518116778213,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.9713142094729824e-05,
+      "loss": 1.3469,
+      "step": 10546
+    },
+    {
+      "epoch": 9.73644126471267,
+      "grad_norm": 3.982421875,
+      "learning_rate": 2.9706470980653772e-05,
+      "loss": 1.1808,
+      "step": 10547
+    },
+    {
+      "epoch": 9.737364412647127,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.969979986657772e-05,
+      "loss": 1.1669,
+      "step": 10548
+    },
+    {
+      "epoch": 9.738287560581583,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.969312875250167e-05,
+      "loss": 1.1948,
+      "step": 10549
+    },
+    {
+      "epoch": 9.73921070851604,
+      "grad_norm": 3.63671875,
+      "learning_rate": 2.968645763842562e-05,
+      "loss": 1.3582,
+      "step": 10550
+    },
+    {
+      "epoch": 9.740133856450496,
+      "grad_norm": 3.423828125,
+      "learning_rate": 2.9679786524349568e-05,
+      "loss": 1.3484,
+      "step": 10551
+    },
+    {
+      "epoch": 9.741057004384952,
+      "grad_norm": 3.044921875,
+      "learning_rate": 2.9673115410273517e-05,
+      "loss": 1.0756,
+      "step": 10552
+    },
+    {
+      "epoch": 9.741980152319409,
+      "grad_norm": 2.939453125,
+      "learning_rate": 2.9666444296197465e-05,
+      "loss": 1.2574,
+      "step": 10553
+    },
+    {
+      "epoch": 9.742903300253866,
+      "grad_norm": 2.9453125,
+      "learning_rate": 2.9659773182121414e-05,
+      "loss": 0.8226,
+      "step": 10554
+    },
+    {
+      "epoch": 9.743826448188322,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9653102068045363e-05,
+      "loss": 1.2115,
+      "step": 10555
+    },
+    {
+      "epoch": 9.744749596122778,
+      "grad_norm": 3.17578125,
+      "learning_rate": 2.9646430953969312e-05,
+      "loss": 1.1974,
+      "step": 10556
+    },
+    {
+      "epoch": 9.745672744057234,
+      "grad_norm": 3.025390625,
+      "learning_rate": 2.9639759839893267e-05,
+      "loss": 1.5444,
+      "step": 10557
+    },
+    {
+      "epoch": 9.746595891991692,
+      "grad_norm": 3.501953125,
+      "learning_rate": 2.9633088725817216e-05,
+      "loss": 1.1169,
+      "step": 10558
+    },
+    {
+      "epoch": 9.747519039926148,
+      "grad_norm": 3.607421875,
+      "learning_rate": 2.9626417611741165e-05,
+      "loss": 1.3337,
+      "step": 10559
+    },
+    {
+      "epoch": 9.748442187860604,
+      "grad_norm": 3.068359375,
+      "learning_rate": 2.9619746497665114e-05,
+      "loss": 1.1698,
+      "step": 10560
+    },
+    {
+      "epoch": 9.749365335795062,
+      "grad_norm": 3.228515625,
+      "learning_rate": 2.9613075383589063e-05,
+      "loss": 1.2886,
+      "step": 10561
+    },
+    {
+      "epoch": 9.750288483729518,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.960640426951301e-05,
+      "loss": 1.2543,
+      "step": 10562
+    },
+    {
+      "epoch": 9.751211631663974,
+      "grad_norm": 3.99609375,
+      "learning_rate": 2.959973315543696e-05,
+      "loss": 1.4905,
+      "step": 10563
+    },
+    {
+      "epoch": 9.75213477959843,
+      "grad_norm": 3.599609375,
+      "learning_rate": 2.959306204136091e-05,
+      "loss": 1.4424,
+      "step": 10564
+    },
+    {
+      "epoch": 9.753057927532888,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9586390927284858e-05,
+      "loss": 1.3765,
+      "step": 10565
+    },
+    {
+      "epoch": 9.753981075467344,
+      "grad_norm": 3.673828125,
+      "learning_rate": 2.9579719813208807e-05,
+      "loss": 1.502,
+      "step": 10566
+    },
+    {
+      "epoch": 9.7549042234018,
+      "grad_norm": 3.435546875,
+      "learning_rate": 2.9573048699132756e-05,
+      "loss": 1.2673,
+      "step": 10567
+    },
+    {
+      "epoch": 9.755827371336256,
+      "grad_norm": 2.958984375,
+      "learning_rate": 2.9566377585056704e-05,
+      "loss": 1.6318,
+      "step": 10568
+    },
+    {
+      "epoch": 9.756750519270714,
+      "grad_norm": 4.37109375,
+      "learning_rate": 2.9559706470980653e-05,
+      "loss": 1.2932,
+      "step": 10569
+    },
+    {
+      "epoch": 9.75767366720517,
+      "grad_norm": 3.37109375,
+      "learning_rate": 2.9553035356904605e-05,
+      "loss": 1.282,
+      "step": 10570
+    },
+    {
+      "epoch": 9.758596815139626,
+      "grad_norm": 3.486328125,
+      "learning_rate": 2.9546364242828554e-05,
+      "loss": 0.926,
+      "step": 10571
+    },
+    {
+      "epoch": 9.759519963074082,
+      "grad_norm": 2.9921875,
+      "learning_rate": 2.9539693128752503e-05,
+      "loss": 1.1957,
+      "step": 10572
+    },
+    {
+      "epoch": 9.76044311100854,
+      "grad_norm": 3.044921875,
+      "learning_rate": 2.9533022014676452e-05,
+      "loss": 1.418,
+      "step": 10573
+    },
+    {
+      "epoch": 9.761366258942996,
+      "grad_norm": 3.408203125,
+      "learning_rate": 2.95263509006004e-05,
+      "loss": 1.2571,
+      "step": 10574
+    },
+    {
+      "epoch": 9.762289406877452,
+      "grad_norm": 3.267578125,
+      "learning_rate": 2.9519679786524353e-05,
+      "loss": 1.1942,
+      "step": 10575
+    },
+    {
+      "epoch": 9.763212554811908,
+      "grad_norm": 3.36328125,
+      "learning_rate": 2.9513008672448302e-05,
+      "loss": 1.2244,
+      "step": 10576
+    },
+    {
+      "epoch": 9.764135702746366,
+      "grad_norm": 3.697265625,
+      "learning_rate": 2.950633755837225e-05,
+      "loss": 1.2214,
+      "step": 10577
+    },
+    {
+      "epoch": 9.765058850680822,
+      "grad_norm": 3.392578125,
+      "learning_rate": 2.94996664442962e-05,
+      "loss": 1.4832,
+      "step": 10578
+    },
+    {
+      "epoch": 9.765981998615278,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.9492995330220148e-05,
+      "loss": 1.2938,
+      "step": 10579
+    },
+    {
+      "epoch": 9.766905146549735,
+      "grad_norm": 4.10546875,
+      "learning_rate": 2.9486324216144097e-05,
+      "loss": 1.3499,
+      "step": 10580
+    },
+    {
+      "epoch": 9.767828294484191,
+      "grad_norm": 3.11328125,
+      "learning_rate": 2.9479653102068046e-05,
+      "loss": 1.151,
+      "step": 10581
+    },
+    {
+      "epoch": 9.768751442418647,
+      "grad_norm": 3.44140625,
+      "learning_rate": 2.9472981987991998e-05,
+      "loss": 1.4392,
+      "step": 10582
+    },
+    {
+      "epoch": 9.769674590353103,
+      "grad_norm": 3.2578125,
+      "learning_rate": 2.9466310873915947e-05,
+      "loss": 1.3871,
+      "step": 10583
+    },
+    {
+      "epoch": 9.770597738287561,
+      "grad_norm": 3.033203125,
+      "learning_rate": 2.9459639759839896e-05,
+      "loss": 1.2891,
+      "step": 10584
+    },
+    {
+      "epoch": 9.771520886222017,
+      "grad_norm": 3.021484375,
+      "learning_rate": 2.9452968645763845e-05,
+      "loss": 1.3442,
+      "step": 10585
+    },
+    {
+      "epoch": 9.772444034156473,
+      "grad_norm": 3.263671875,
+      "learning_rate": 2.9446297531687793e-05,
+      "loss": 1.3743,
+      "step": 10586
+    },
+    {
+      "epoch": 9.77336718209093,
+      "grad_norm": 3.478515625,
+      "learning_rate": 2.9439626417611742e-05,
+      "loss": 1.3904,
+      "step": 10587
+    },
+    {
+      "epoch": 9.774290330025387,
+      "grad_norm": 3.43359375,
+      "learning_rate": 2.943295530353569e-05,
+      "loss": 0.9132,
+      "step": 10588
+    },
+    {
+      "epoch": 9.775213477959843,
+      "grad_norm": 2.99609375,
+      "learning_rate": 2.942628418945964e-05,
+      "loss": 1.3529,
+      "step": 10589
+    },
+    {
+      "epoch": 9.7761366258943,
+      "grad_norm": 3.037109375,
+      "learning_rate": 2.941961307538359e-05,
+      "loss": 1.1532,
+      "step": 10590
+    },
+    {
+      "epoch": 9.777059773828755,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.9412941961307537e-05,
+      "loss": 1.3892,
+      "step": 10591
+    },
+    {
+      "epoch": 9.777982921763213,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.9406270847231486e-05,
+      "loss": 1.3936,
+      "step": 10592
+    },
+    {
+      "epoch": 9.778906069697669,
+      "grad_norm": 3.3984375,
+      "learning_rate": 2.9399599733155435e-05,
+      "loss": 1.2551,
+      "step": 10593
+    },
+    {
+      "epoch": 9.779829217632125,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9392928619079384e-05,
+      "loss": 1.4578,
+      "step": 10594
+    },
+    {
+      "epoch": 9.780752365566581,
+      "grad_norm": 3.318359375,
+      "learning_rate": 2.938625750500334e-05,
+      "loss": 1.196,
+      "step": 10595
+    },
+    {
+      "epoch": 9.781675513501039,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.937958639092729e-05,
+      "loss": 1.2729,
+      "step": 10596
+    },
+    {
+      "epoch": 9.782598661435495,
+      "grad_norm": 3.052734375,
+      "learning_rate": 2.9372915276851237e-05,
+      "loss": 1.1567,
+      "step": 10597
+    },
+    {
+      "epoch": 9.783521809369951,
+      "grad_norm": 3.05078125,
+      "learning_rate": 2.9366244162775186e-05,
+      "loss": 1.2976,
+      "step": 10598
+    },
+    {
+      "epoch": 9.784444957304409,
+      "grad_norm": 2.998046875,
+      "learning_rate": 2.9359573048699135e-05,
+      "loss": 1.4172,
+      "step": 10599
+    },
+    {
+      "epoch": 9.785368105238865,
+      "grad_norm": 3.556640625,
+      "learning_rate": 2.9352901934623084e-05,
+      "loss": 1.5312,
+      "step": 10600
+    },
+    {
+      "epoch": 9.78629125317332,
+      "grad_norm": 2.73828125,
+      "learning_rate": 2.9346230820547032e-05,
+      "loss": 1.3933,
+      "step": 10601
+    },
+    {
+      "epoch": 9.787214401107777,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.933955970647098e-05,
+      "loss": 1.2092,
+      "step": 10602
+    },
+    {
+      "epoch": 9.788137549042235,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.933288859239493e-05,
+      "loss": 1.5505,
+      "step": 10603
+    },
+    {
+      "epoch": 9.78906069697669,
+      "grad_norm": 3.517578125,
+      "learning_rate": 2.932621747831888e-05,
+      "loss": 1.384,
+      "step": 10604
+    },
+    {
+      "epoch": 9.789983844911147,
+      "grad_norm": 3.275390625,
+      "learning_rate": 2.9319546364242828e-05,
+      "loss": 1.2198,
+      "step": 10605
+    },
+    {
+      "epoch": 9.790906992845603,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9312875250166777e-05,
+      "loss": 1.1766,
+      "step": 10606
+    },
+    {
+      "epoch": 9.79183014078006,
+      "grad_norm": 2.818359375,
+      "learning_rate": 2.9306204136090725e-05,
+      "loss": 0.9867,
+      "step": 10607
+    },
+    {
+      "epoch": 9.792753288714517,
+      "grad_norm": 3.208984375,
+      "learning_rate": 2.929953302201468e-05,
+      "loss": 1.3804,
+      "step": 10608
+    },
+    {
+      "epoch": 9.793676436648973,
+      "grad_norm": 3.39453125,
+      "learning_rate": 2.929286190793863e-05,
+      "loss": 1.3252,
+      "step": 10609
+    },
+    {
+      "epoch": 9.79459958458343,
+      "grad_norm": 3.875,
+      "learning_rate": 2.928619079386258e-05,
+      "loss": 1.4299,
+      "step": 10610
+    },
+    {
+      "epoch": 9.795522732517886,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.9279519679786527e-05,
+      "loss": 1.3982,
+      "step": 10611
+    },
+    {
+      "epoch": 9.796445880452342,
+      "grad_norm": 3.669921875,
+      "learning_rate": 2.9272848565710476e-05,
+      "loss": 1.1816,
+      "step": 10612
+    },
+    {
+      "epoch": 9.797369028386798,
+      "grad_norm": 3.423828125,
+      "learning_rate": 2.9266177451634425e-05,
+      "loss": 1.5286,
+      "step": 10613
+    },
+    {
+      "epoch": 9.798292176321256,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.9259506337558374e-05,
+      "loss": 1.2192,
+      "step": 10614
+    },
+    {
+      "epoch": 9.799215324255712,
+      "grad_norm": 3.416015625,
+      "learning_rate": 2.9252835223482323e-05,
+      "loss": 1.3049,
+      "step": 10615
+    },
+    {
+      "epoch": 9.800138472190168,
+      "grad_norm": 3.201171875,
+      "learning_rate": 2.924616410940627e-05,
+      "loss": 1.2513,
+      "step": 10616
+    },
+    {
+      "epoch": 9.801061620124624,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.923949299533022e-05,
+      "loss": 1.6326,
+      "step": 10617
+    },
+    {
+      "epoch": 9.801984768059082,
+      "grad_norm": 3.462890625,
+      "learning_rate": 2.923282188125417e-05,
+      "loss": 0.9916,
+      "step": 10618
+    },
+    {
+      "epoch": 9.802907915993538,
+      "grad_norm": 3.447265625,
+      "learning_rate": 2.9226150767178118e-05,
+      "loss": 1.3777,
+      "step": 10619
+    },
+    {
+      "epoch": 9.803831063927994,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9219479653102074e-05,
+      "loss": 1.2073,
+      "step": 10620
+    },
+    {
+      "epoch": 9.80475421186245,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9212808539026022e-05,
+      "loss": 1.2988,
+      "step": 10621
+    },
+    {
+      "epoch": 9.805677359796908,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.920613742494997e-05,
+      "loss": 1.2729,
+      "step": 10622
+    },
+    {
+      "epoch": 9.806600507731364,
+      "grad_norm": 3.142578125,
+      "learning_rate": 2.919946631087392e-05,
+      "loss": 1.24,
+      "step": 10623
+    },
+    {
+      "epoch": 9.80752365566582,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.919279519679787e-05,
+      "loss": 1.4767,
+      "step": 10624
+    },
+    {
+      "epoch": 9.808446803600276,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.9186124082721818e-05,
+      "loss": 1.6089,
+      "step": 10625
+    },
+    {
+      "epoch": 9.809369951534734,
+      "grad_norm": 3.521484375,
+      "learning_rate": 2.9179452968645766e-05,
+      "loss": 1.1287,
+      "step": 10626
+    },
+    {
+      "epoch": 9.81029309946919,
+      "grad_norm": 3.076171875,
+      "learning_rate": 2.9172781854569715e-05,
+      "loss": 1.2551,
+      "step": 10627
+    },
+    {
+      "epoch": 9.811216247403646,
+      "grad_norm": 3.181640625,
+      "learning_rate": 2.9166110740493664e-05,
+      "loss": 1.4102,
+      "step": 10628
+    },
+    {
+      "epoch": 9.812139395338104,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.9159439626417613e-05,
+      "loss": 1.3865,
+      "step": 10629
+    },
+    {
+      "epoch": 9.81306254327256,
+      "grad_norm": 3.076171875,
+      "learning_rate": 2.9152768512341562e-05,
+      "loss": 1.283,
+      "step": 10630
+    },
+    {
+      "epoch": 9.813985691207016,
+      "grad_norm": 3.283203125,
+      "learning_rate": 2.914609739826551e-05,
+      "loss": 1.5667,
+      "step": 10631
+    },
+    {
+      "epoch": 9.814908839141472,
+      "grad_norm": 3.384765625,
+      "learning_rate": 2.913942628418946e-05,
+      "loss": 1.3386,
+      "step": 10632
+    },
+    {
+      "epoch": 9.81583198707593,
+      "grad_norm": 4.19140625,
+      "learning_rate": 2.913275517011341e-05,
+      "loss": 1.2041,
+      "step": 10633
+    },
+    {
+      "epoch": 9.816755135010386,
+      "grad_norm": 3.0234375,
+      "learning_rate": 2.912608405603736e-05,
+      "loss": 1.3479,
+      "step": 10634
+    },
+    {
+      "epoch": 9.817678282944842,
+      "grad_norm": 2.87890625,
+      "learning_rate": 2.911941294196131e-05,
+      "loss": 1.2981,
+      "step": 10635
+    },
+    {
+      "epoch": 9.818601430879298,
+      "grad_norm": 3.775390625,
+      "learning_rate": 2.9112741827885258e-05,
+      "loss": 1.3179,
+      "step": 10636
+    },
+    {
+      "epoch": 9.819524578813756,
+      "grad_norm": 3.09765625,
+      "learning_rate": 2.9106070713809207e-05,
+      "loss": 1.5222,
+      "step": 10637
+    },
+    {
+      "epoch": 9.820447726748212,
+      "grad_norm": 3.302734375,
+      "learning_rate": 2.9099399599733156e-05,
+      "loss": 1.3356,
+      "step": 10638
+    },
+    {
+      "epoch": 9.821370874682668,
+      "grad_norm": 3.49609375,
+      "learning_rate": 2.9092728485657104e-05,
+      "loss": 1.4324,
+      "step": 10639
+    },
+    {
+      "epoch": 9.822294022617124,
+      "grad_norm": 3.16796875,
+      "learning_rate": 2.9086057371581053e-05,
+      "loss": 1.5571,
+      "step": 10640
+    },
+    {
+      "epoch": 9.823217170551581,
+      "grad_norm": 2.783203125,
+      "learning_rate": 2.9079386257505002e-05,
+      "loss": 1.2356,
+      "step": 10641
+    },
+    {
+      "epoch": 9.824140318486037,
+      "grad_norm": 3.0078125,
+      "learning_rate": 2.907271514342895e-05,
+      "loss": 1.3831,
+      "step": 10642
+    },
+    {
+      "epoch": 9.825063466420493,
+      "grad_norm": 3.02734375,
+      "learning_rate": 2.90660440293529e-05,
+      "loss": 1.3716,
+      "step": 10643
+    },
+    {
+      "epoch": 9.82598661435495,
+      "grad_norm": 3.654296875,
+      "learning_rate": 2.905937291527685e-05,
+      "loss": 1.4194,
+      "step": 10644
+    },
+    {
+      "epoch": 9.826909762289407,
+      "grad_norm": 3.259765625,
+      "learning_rate": 2.9052701801200804e-05,
+      "loss": 1.1296,
+      "step": 10645
+    },
+    {
+      "epoch": 9.827832910223863,
+      "grad_norm": 3.353515625,
+      "learning_rate": 2.9046030687124753e-05,
+      "loss": 1.2582,
+      "step": 10646
+    },
+    {
+      "epoch": 9.82875605815832,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.9039359573048702e-05,
+      "loss": 1.3616,
+      "step": 10647
+    },
+    {
+      "epoch": 9.829679206092777,
+      "grad_norm": 3.271484375,
+      "learning_rate": 2.903268845897265e-05,
+      "loss": 1.3152,
+      "step": 10648
+    },
+    {
+      "epoch": 9.830602354027233,
+      "grad_norm": 3.1796875,
+      "learning_rate": 2.90260173448966e-05,
+      "loss": 1.3534,
+      "step": 10649
+    },
+    {
+      "epoch": 9.83152550196169,
+      "grad_norm": 3.087890625,
+      "learning_rate": 2.9019346230820548e-05,
+      "loss": 1.4359,
+      "step": 10650
+    },
+    {
+      "epoch": 9.832448649896145,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.9012675116744497e-05,
+      "loss": 1.0852,
+      "step": 10651
+    },
+    {
+      "epoch": 9.833371797830603,
+      "grad_norm": 3.35546875,
+      "learning_rate": 2.9006004002668446e-05,
+      "loss": 1.3479,
+      "step": 10652
+    },
+    {
+      "epoch": 9.834294945765059,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.8999332888592395e-05,
+      "loss": 1.4172,
+      "step": 10653
+    },
+    {
+      "epoch": 9.835218093699515,
+      "grad_norm": 3.7109375,
+      "learning_rate": 2.8992661774516344e-05,
+      "loss": 1.5461,
+      "step": 10654
+    },
+    {
+      "epoch": 9.836141241633971,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8985990660440292e-05,
+      "loss": 1.2288,
+      "step": 10655
+    },
+    {
+      "epoch": 9.837064389568429,
+      "grad_norm": 3.36328125,
+      "learning_rate": 2.897931954636424e-05,
+      "loss": 1.2633,
+      "step": 10656
+    },
+    {
+      "epoch": 9.837987537502885,
+      "grad_norm": 3.2265625,
+      "learning_rate": 2.897264843228819e-05,
+      "loss": 1.2689,
+      "step": 10657
+    },
+    {
+      "epoch": 9.838910685437341,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.8965977318212146e-05,
+      "loss": 1.2954,
+      "step": 10658
+    },
+    {
+      "epoch": 9.839833833371799,
+      "grad_norm": 3.982421875,
+      "learning_rate": 2.8959306204136094e-05,
+      "loss": 1.1726,
+      "step": 10659
+    },
+    {
+      "epoch": 9.840756981306255,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8952635090060043e-05,
+      "loss": 1.1863,
+      "step": 10660
+    },
+    {
+      "epoch": 9.84168012924071,
+      "grad_norm": 3.3515625,
+      "learning_rate": 2.8945963975983992e-05,
+      "loss": 1.3035,
+      "step": 10661
+    },
+    {
+      "epoch": 9.842603277175167,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.893929286190794e-05,
+      "loss": 1.2827,
+      "step": 10662
+    },
+    {
+      "epoch": 9.843526425109625,
+      "grad_norm": 3.361328125,
+      "learning_rate": 2.893262174783189e-05,
+      "loss": 0.994,
+      "step": 10663
+    },
+    {
+      "epoch": 9.84444957304408,
+      "grad_norm": 3.26171875,
+      "learning_rate": 2.892595063375584e-05,
+      "loss": 1.3536,
+      "step": 10664
+    },
+    {
+      "epoch": 9.845372720978537,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.8919279519679787e-05,
+      "loss": 1.6453,
+      "step": 10665
+    },
+    {
+      "epoch": 9.846295868912993,
+      "grad_norm": 3.35546875,
+      "learning_rate": 2.8912608405603736e-05,
+      "loss": 1.1885,
+      "step": 10666
+    },
+    {
+      "epoch": 9.84721901684745,
+      "grad_norm": 3.076171875,
+      "learning_rate": 2.8905937291527685e-05,
+      "loss": 1.4138,
+      "step": 10667
+    },
+    {
+      "epoch": 9.848142164781907,
+      "grad_norm": 3.986328125,
+      "learning_rate": 2.8899266177451634e-05,
+      "loss": 1.4941,
+      "step": 10668
+    },
+    {
+      "epoch": 9.849065312716363,
+      "grad_norm": 3.39453125,
+      "learning_rate": 2.8892595063375583e-05,
+      "loss": 1.2352,
+      "step": 10669
+    },
+    {
+      "epoch": 9.849988460650819,
+      "grad_norm": 3.634765625,
+      "learning_rate": 2.8885923949299538e-05,
+      "loss": 1.334,
+      "step": 10670
+    },
+    {
+      "epoch": 9.850911608585276,
+      "grad_norm": 2.966796875,
+      "learning_rate": 2.8879252835223487e-05,
+      "loss": 1.3579,
+      "step": 10671
+    },
+    {
+      "epoch": 9.851834756519732,
+      "grad_norm": 3.93359375,
+      "learning_rate": 2.8872581721147436e-05,
+      "loss": 1.6565,
+      "step": 10672
+    },
+    {
+      "epoch": 9.852757904454188,
+      "grad_norm": 2.892578125,
+      "learning_rate": 2.8865910607071385e-05,
+      "loss": 1.4114,
+      "step": 10673
+    },
+    {
+      "epoch": 9.853681052388644,
+      "grad_norm": 3.3046875,
+      "learning_rate": 2.8859239492995333e-05,
+      "loss": 1.3188,
+      "step": 10674
+    },
+    {
+      "epoch": 9.854604200323102,
+      "grad_norm": 3.259765625,
+      "learning_rate": 2.8852568378919282e-05,
+      "loss": 1.3442,
+      "step": 10675
+    },
+    {
+      "epoch": 9.855527348257558,
+      "grad_norm": 3.5078125,
+      "learning_rate": 2.884589726484323e-05,
+      "loss": 1.0809,
+      "step": 10676
+    },
+    {
+      "epoch": 9.856450496192014,
+      "grad_norm": 3.365234375,
+      "learning_rate": 2.883922615076718e-05,
+      "loss": 1.1855,
+      "step": 10677
+    },
+    {
+      "epoch": 9.857373644126472,
+      "grad_norm": 3.150390625,
+      "learning_rate": 2.883255503669113e-05,
+      "loss": 1.5234,
+      "step": 10678
+    },
+    {
+      "epoch": 9.858296792060928,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.8825883922615078e-05,
+      "loss": 1.4872,
+      "step": 10679
+    },
+    {
+      "epoch": 9.859219939995384,
+      "grad_norm": 3.259765625,
+      "learning_rate": 2.8819212808539026e-05,
+      "loss": 1.025,
+      "step": 10680
+    },
+    {
+      "epoch": 9.86014308792984,
+      "grad_norm": 3.12109375,
+      "learning_rate": 2.8812541694462975e-05,
+      "loss": 1.4651,
+      "step": 10681
+    },
+    {
+      "epoch": 9.861066235864298,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.8805870580386924e-05,
+      "loss": 1.2625,
+      "step": 10682
+    },
+    {
+      "epoch": 9.861989383798754,
+      "grad_norm": 3.52734375,
+      "learning_rate": 2.8799199466310876e-05,
+      "loss": 1.0496,
+      "step": 10683
+    },
+    {
+      "epoch": 9.86291253173321,
+      "grad_norm": 2.841796875,
+      "learning_rate": 2.8792528352234825e-05,
+      "loss": 1.2468,
+      "step": 10684
+    },
+    {
+      "epoch": 9.863835679667666,
+      "grad_norm": 3.462890625,
+      "learning_rate": 2.8785857238158774e-05,
+      "loss": 1.1108,
+      "step": 10685
+    },
+    {
+      "epoch": 9.864758827602124,
+      "grad_norm": 3.916015625,
+      "learning_rate": 2.8779186124082723e-05,
+      "loss": 1.1624,
+      "step": 10686
+    },
+    {
+      "epoch": 9.86568197553658,
+      "grad_norm": 2.6171875,
+      "learning_rate": 2.877251501000667e-05,
+      "loss": 1.1395,
+      "step": 10687
+    },
+    {
+      "epoch": 9.866605123471036,
+      "grad_norm": 3.287109375,
+      "learning_rate": 2.8765843895930624e-05,
+      "loss": 1.301,
+      "step": 10688
+    },
+    {
+      "epoch": 9.867528271405492,
+      "grad_norm": 2.982421875,
+      "learning_rate": 2.8759172781854573e-05,
+      "loss": 1.2754,
+      "step": 10689
+    },
+    {
+      "epoch": 9.86845141933995,
+      "grad_norm": 3.25390625,
+      "learning_rate": 2.875250166777852e-05,
+      "loss": 0.8112,
+      "step": 10690
+    },
+    {
+      "epoch": 9.869374567274406,
+      "grad_norm": 3.283203125,
+      "learning_rate": 2.874583055370247e-05,
+      "loss": 1.4355,
+      "step": 10691
+    },
+    {
+      "epoch": 9.870297715208862,
+      "grad_norm": 3.853515625,
+      "learning_rate": 2.873915943962642e-05,
+      "loss": 1.2925,
+      "step": 10692
+    },
+    {
+      "epoch": 9.871220863143318,
+      "grad_norm": 3.7890625,
+      "learning_rate": 2.8732488325550368e-05,
+      "loss": 1.126,
+      "step": 10693
+    },
+    {
+      "epoch": 9.872144011077776,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8725817211474317e-05,
+      "loss": 1.1843,
+      "step": 10694
+    },
+    {
+      "epoch": 9.873067159012232,
+      "grad_norm": 3.34765625,
+      "learning_rate": 2.871914609739827e-05,
+      "loss": 1.3279,
+      "step": 10695
+    },
+    {
+      "epoch": 9.873990306946688,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.8712474983322218e-05,
+      "loss": 1.4414,
+      "step": 10696
+    },
+    {
+      "epoch": 9.874913454881145,
+      "grad_norm": 2.888671875,
+      "learning_rate": 2.8705803869246166e-05,
+      "loss": 1.355,
+      "step": 10697
+    },
+    {
+      "epoch": 9.875836602815601,
+      "grad_norm": 3.26171875,
+      "learning_rate": 2.8699132755170115e-05,
+      "loss": 1.1846,
+      "step": 10698
+    },
+    {
+      "epoch": 9.876759750750058,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.8692461641094064e-05,
+      "loss": 1.3372,
+      "step": 10699
+    },
+    {
+      "epoch": 9.877682898684514,
+      "grad_norm": 2.70703125,
+      "learning_rate": 2.8685790527018013e-05,
+      "loss": 1.2514,
+      "step": 10700
+    },
+    {
+      "epoch": 9.878606046618971,
+      "grad_norm": 3.240234375,
+      "learning_rate": 2.8679119412941962e-05,
+      "loss": 1.3291,
+      "step": 10701
+    },
+    {
+      "epoch": 9.879529194553427,
+      "grad_norm": 3.095703125,
+      "learning_rate": 2.867244829886591e-05,
+      "loss": 1.249,
+      "step": 10702
+    },
+    {
+      "epoch": 9.880452342487883,
+      "grad_norm": 3.708984375,
+      "learning_rate": 2.866577718478986e-05,
+      "loss": 1.136,
+      "step": 10703
+    },
+    {
+      "epoch": 9.88137549042234,
+      "grad_norm": 2.80078125,
+      "learning_rate": 2.8659106070713808e-05,
+      "loss": 1.1733,
+      "step": 10704
+    },
+    {
+      "epoch": 9.882298638356797,
+      "grad_norm": 3.576171875,
+      "learning_rate": 2.8652434956637757e-05,
+      "loss": 1.4216,
+      "step": 10705
+    },
+    {
+      "epoch": 9.883221786291253,
+      "grad_norm": 3.05859375,
+      "learning_rate": 2.8645763842561706e-05,
+      "loss": 1.3511,
+      "step": 10706
+    },
+    {
+      "epoch": 9.88414493422571,
+      "grad_norm": 3.1796875,
+      "learning_rate": 2.8639092728485655e-05,
+      "loss": 1.3496,
+      "step": 10707
+    },
+    {
+      "epoch": 9.885068082160167,
+      "grad_norm": 3.3984375,
+      "learning_rate": 2.863242161440961e-05,
+      "loss": 1.0078,
+      "step": 10708
+    },
+    {
+      "epoch": 9.885991230094623,
+      "grad_norm": 3.287109375,
+      "learning_rate": 2.862575050033356e-05,
+      "loss": 1.4001,
+      "step": 10709
+    },
+    {
+      "epoch": 9.88691437802908,
+      "grad_norm": 3.681640625,
+      "learning_rate": 2.8619079386257508e-05,
+      "loss": 1.3005,
+      "step": 10710
+    },
+    {
+      "epoch": 9.887837525963535,
+      "grad_norm": 3.45703125,
+      "learning_rate": 2.8612408272181457e-05,
+      "loss": 1.1804,
+      "step": 10711
+    },
+    {
+      "epoch": 9.888760673897993,
+      "grad_norm": 3.279296875,
+      "learning_rate": 2.8605737158105406e-05,
+      "loss": 1.4673,
+      "step": 10712
+    },
+    {
+      "epoch": 9.889683821832449,
+      "grad_norm": 3.45703125,
+      "learning_rate": 2.8599066044029354e-05,
+      "loss": 1.438,
+      "step": 10713
+    },
+    {
+      "epoch": 9.890606969766905,
+      "grad_norm": 3.990234375,
+      "learning_rate": 2.8592394929953303e-05,
+      "loss": 1.1173,
+      "step": 10714
+    },
+    {
+      "epoch": 9.891530117701361,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.8585723815877252e-05,
+      "loss": 1.209,
+      "step": 10715
+    },
+    {
+      "epoch": 9.892453265635819,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.85790527018012e-05,
+      "loss": 1.2759,
+      "step": 10716
+    },
+    {
+      "epoch": 9.893376413570275,
+      "grad_norm": 2.541015625,
+      "learning_rate": 2.857238158772515e-05,
+      "loss": 0.9716,
+      "step": 10717
+    },
+    {
+      "epoch": 9.894299561504731,
+      "grad_norm": 3.416015625,
+      "learning_rate": 2.85657104736491e-05,
+      "loss": 1.0847,
+      "step": 10718
+    },
+    {
+      "epoch": 9.895222709439187,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8559039359573047e-05,
+      "loss": 1.0444,
+      "step": 10719
+    },
+    {
+      "epoch": 9.896145857373645,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8552368245497003e-05,
+      "loss": 1.5332,
+      "step": 10720
+    },
+    {
+      "epoch": 9.8970690053081,
+      "grad_norm": 3.91796875,
+      "learning_rate": 2.854569713142095e-05,
+      "loss": 1.1555,
+      "step": 10721
+    },
+    {
+      "epoch": 9.897992153242557,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.85390260173449e-05,
+      "loss": 1.3789,
+      "step": 10722
+    },
+    {
+      "epoch": 9.898915301177013,
+      "grad_norm": 3.103515625,
+      "learning_rate": 2.853235490326885e-05,
+      "loss": 1.5073,
+      "step": 10723
+    },
+    {
+      "epoch": 9.89983844911147,
+      "grad_norm": 3.3671875,
+      "learning_rate": 2.8525683789192798e-05,
+      "loss": 1.1874,
+      "step": 10724
+    },
+    {
+      "epoch": 9.900761597045927,
+      "grad_norm": 3.294921875,
+      "learning_rate": 2.8519012675116747e-05,
+      "loss": 1.3306,
+      "step": 10725
+    },
+    {
+      "epoch": 9.901684744980383,
+      "grad_norm": 3.115234375,
+      "learning_rate": 2.8512341561040696e-05,
+      "loss": 1.0997,
+      "step": 10726
+    },
+    {
+      "epoch": 9.90260789291484,
+      "grad_norm": 3.095703125,
+      "learning_rate": 2.8505670446964645e-05,
+      "loss": 1.4548,
+      "step": 10727
+    },
+    {
+      "epoch": 9.903531040849296,
+      "grad_norm": 3.25390625,
+      "learning_rate": 2.8498999332888593e-05,
+      "loss": 1.3931,
+      "step": 10728
+    },
+    {
+      "epoch": 9.904454188783752,
+      "grad_norm": 3.529296875,
+      "learning_rate": 2.8492328218812542e-05,
+      "loss": 1.3574,
+      "step": 10729
+    },
+    {
+      "epoch": 9.905377336718209,
+      "grad_norm": 3.693359375,
+      "learning_rate": 2.848565710473649e-05,
+      "loss": 1.4443,
+      "step": 10730
+    },
+    {
+      "epoch": 9.906300484652666,
+      "grad_norm": 2.986328125,
+      "learning_rate": 2.847898599066044e-05,
+      "loss": 1.1919,
+      "step": 10731
+    },
+    {
+      "epoch": 9.907223632587122,
+      "grad_norm": 2.8359375,
+      "learning_rate": 2.847231487658439e-05,
+      "loss": 1.4338,
+      "step": 10732
+    },
+    {
+      "epoch": 9.908146780521578,
+      "grad_norm": 3.025390625,
+      "learning_rate": 2.8465643762508344e-05,
+      "loss": 1.21,
+      "step": 10733
+    },
+    {
+      "epoch": 9.909069928456034,
+      "grad_norm": 3.248046875,
+      "learning_rate": 2.8458972648432293e-05,
+      "loss": 1.4136,
+      "step": 10734
+    },
+    {
+      "epoch": 9.909993076390492,
+      "grad_norm": 3.1171875,
+      "learning_rate": 2.8452301534356242e-05,
+      "loss": 1.1208,
+      "step": 10735
+    },
+    {
+      "epoch": 9.910916224324948,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.844563042028019e-05,
+      "loss": 1.2991,
+      "step": 10736
+    },
+    {
+      "epoch": 9.911839372259404,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.843895930620414e-05,
+      "loss": 1.5205,
+      "step": 10737
+    },
+    {
+      "epoch": 9.91276252019386,
+      "grad_norm": 3.60546875,
+      "learning_rate": 2.843228819212809e-05,
+      "loss": 1.1686,
+      "step": 10738
+    },
+    {
+      "epoch": 9.913685668128318,
+      "grad_norm": 3.55078125,
+      "learning_rate": 2.8425617078052037e-05,
+      "loss": 1.0787,
+      "step": 10739
+    },
+    {
+      "epoch": 9.914608816062774,
+      "grad_norm": 3.62109375,
+      "learning_rate": 2.8418945963975986e-05,
+      "loss": 1.4692,
+      "step": 10740
+    },
+    {
+      "epoch": 9.91553196399723,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.8412274849899935e-05,
+      "loss": 1.1532,
+      "step": 10741
+    },
+    {
+      "epoch": 9.916455111931686,
+      "grad_norm": 3.173828125,
+      "learning_rate": 2.8405603735823884e-05,
+      "loss": 1.2917,
+      "step": 10742
+    },
+    {
+      "epoch": 9.917378259866144,
+      "grad_norm": 3.517578125,
+      "learning_rate": 2.8398932621747832e-05,
+      "loss": 1.12,
+      "step": 10743
+    },
+    {
+      "epoch": 9.9183014078006,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.839226150767178e-05,
+      "loss": 1.1697,
+      "step": 10744
+    },
+    {
+      "epoch": 9.919224555735056,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8385590393595733e-05,
+      "loss": 1.0472,
+      "step": 10745
+    },
+    {
+      "epoch": 9.920147703669514,
+      "grad_norm": 3.8828125,
+      "learning_rate": 2.8378919279519682e-05,
+      "loss": 1.3713,
+      "step": 10746
+    },
+    {
+      "epoch": 9.92107085160397,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.837224816544363e-05,
+      "loss": 1.3286,
+      "step": 10747
+    },
+    {
+      "epoch": 9.921993999538426,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.836557705136758e-05,
+      "loss": 1.1255,
+      "step": 10748
+    },
+    {
+      "epoch": 9.922917147472882,
+      "grad_norm": 3.505859375,
+      "learning_rate": 2.835890593729153e-05,
+      "loss": 1.2086,
+      "step": 10749
+    },
+    {
+      "epoch": 9.92384029540734,
+      "grad_norm": 3.166015625,
+      "learning_rate": 2.8352234823215478e-05,
+      "loss": 1.4421,
+      "step": 10750
+    },
+    {
+      "epoch": 9.924763443341796,
+      "grad_norm": 3.287109375,
+      "learning_rate": 2.8345563709139426e-05,
+      "loss": 1.2598,
+      "step": 10751
+    },
+    {
+      "epoch": 9.925686591276252,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8338892595063375e-05,
+      "loss": 1.1741,
+      "step": 10752
+    },
+    {
+      "epoch": 9.926609739210708,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.8332221480987324e-05,
+      "loss": 1.0748,
+      "step": 10753
+    },
+    {
+      "epoch": 9.927532887145166,
+      "grad_norm": 3.009765625,
+      "learning_rate": 2.8325550366911273e-05,
+      "loss": 1.3246,
+      "step": 10754
+    },
+    {
+      "epoch": 9.928456035079622,
+      "grad_norm": 3.19140625,
+      "learning_rate": 2.831887925283522e-05,
+      "loss": 1.0494,
+      "step": 10755
+    },
+    {
+      "epoch": 9.929379183014078,
+      "grad_norm": 3.892578125,
+      "learning_rate": 2.831220813875917e-05,
+      "loss": 1.4653,
+      "step": 10756
+    },
+    {
+      "epoch": 9.930302330948535,
+      "grad_norm": 3.490234375,
+      "learning_rate": 2.830553702468312e-05,
+      "loss": 1.0979,
+      "step": 10757
+    },
+    {
+      "epoch": 9.931225478882991,
+      "grad_norm": 3.14453125,
+      "learning_rate": 2.8298865910607075e-05,
+      "loss": 1.3015,
+      "step": 10758
+    },
+    {
+      "epoch": 9.932148626817447,
+      "grad_norm": 3.0859375,
+      "learning_rate": 2.8292194796531024e-05,
+      "loss": 1.4543,
+      "step": 10759
+    },
+    {
+      "epoch": 9.933071774751904,
+      "grad_norm": 3.404296875,
+      "learning_rate": 2.8285523682454973e-05,
+      "loss": 1.3821,
+      "step": 10760
+    },
+    {
+      "epoch": 9.933994922686361,
+      "grad_norm": 3.23828125,
+      "learning_rate": 2.827885256837892e-05,
+      "loss": 1.0956,
+      "step": 10761
+    },
+    {
+      "epoch": 9.934918070620817,
+      "grad_norm": 3.3203125,
+      "learning_rate": 2.827218145430287e-05,
+      "loss": 0.9308,
+      "step": 10762
+    },
+    {
+      "epoch": 9.935841218555273,
+      "grad_norm": 3.107421875,
+      "learning_rate": 2.826551034022682e-05,
+      "loss": 1.4468,
+      "step": 10763
+    },
+    {
+      "epoch": 9.93676436648973,
+      "grad_norm": 3.197265625,
+      "learning_rate": 2.8258839226150768e-05,
+      "loss": 1.3816,
+      "step": 10764
+    },
+    {
+      "epoch": 9.937687514424187,
+      "grad_norm": 3.69921875,
+      "learning_rate": 2.8252168112074717e-05,
+      "loss": 1.0288,
+      "step": 10765
+    },
+    {
+      "epoch": 9.938610662358643,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8245496997998665e-05,
+      "loss": 1.1242,
+      "step": 10766
+    },
+    {
+      "epoch": 9.9395338102931,
+      "grad_norm": 2.939453125,
+      "learning_rate": 2.8238825883922614e-05,
+      "loss": 1.3138,
+      "step": 10767
+    },
+    {
+      "epoch": 9.940456958227555,
+      "grad_norm": 3.04296875,
+      "learning_rate": 2.8232154769846563e-05,
+      "loss": 1.1825,
+      "step": 10768
+    },
+    {
+      "epoch": 9.941380106162013,
+      "grad_norm": 3.185546875,
+      "learning_rate": 2.8225483655770512e-05,
+      "loss": 1.1401,
+      "step": 10769
+    },
+    {
+      "epoch": 9.942303254096469,
+      "grad_norm": 3.369140625,
+      "learning_rate": 2.8218812541694467e-05,
+      "loss": 1.3618,
+      "step": 10770
+    },
+    {
+      "epoch": 9.943226402030925,
+      "grad_norm": 4.33984375,
+      "learning_rate": 2.8212141427618416e-05,
+      "loss": 1.2493,
+      "step": 10771
+    },
+    {
+      "epoch": 9.944149549965381,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8205470313542365e-05,
+      "loss": 1.3228,
+      "step": 10772
+    },
+    {
+      "epoch": 9.945072697899839,
+      "grad_norm": 3.1484375,
+      "learning_rate": 2.8198799199466314e-05,
+      "loss": 1.1788,
+      "step": 10773
+    },
+    {
+      "epoch": 9.945995845834295,
+      "grad_norm": 3.10546875,
+      "learning_rate": 2.8192128085390263e-05,
+      "loss": 1.3181,
+      "step": 10774
+    },
+    {
+      "epoch": 9.946918993768751,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.818545697131421e-05,
+      "loss": 1.2343,
+      "step": 10775
+    },
+    {
+      "epoch": 9.947842141703209,
+      "grad_norm": 3.525390625,
+      "learning_rate": 2.817878585723816e-05,
+      "loss": 1.3086,
+      "step": 10776
+    },
+    {
+      "epoch": 9.948765289637665,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.817211474316211e-05,
+      "loss": 1.3301,
+      "step": 10777
+    },
+    {
+      "epoch": 9.94968843757212,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8165443629086058e-05,
+      "loss": 1.2504,
+      "step": 10778
+    },
+    {
+      "epoch": 9.950611585506577,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8158772515010007e-05,
+      "loss": 1.1101,
+      "step": 10779
+    },
+    {
+      "epoch": 9.951534733441035,
+      "grad_norm": 3.4140625,
+      "learning_rate": 2.8152101400933956e-05,
+      "loss": 1.2295,
+      "step": 10780
+    },
+    {
+      "epoch": 9.95245788137549,
+      "grad_norm": 3.146484375,
+      "learning_rate": 2.8145430286857904e-05,
+      "loss": 1.1829,
+      "step": 10781
+    },
+    {
+      "epoch": 9.953381029309947,
+      "grad_norm": 3.376953125,
+      "learning_rate": 2.8138759172781853e-05,
+      "loss": 1.3857,
+      "step": 10782
+    },
+    {
+      "epoch": 9.954304177244403,
+      "grad_norm": 3.1953125,
+      "learning_rate": 2.813208805870581e-05,
+      "loss": 1.2834,
+      "step": 10783
+    },
+    {
+      "epoch": 9.95522732517886,
+      "grad_norm": 3.38671875,
+      "learning_rate": 2.8125416944629758e-05,
+      "loss": 1.2681,
+      "step": 10784
+    },
+    {
+      "epoch": 9.956150473113317,
+      "grad_norm": 3.060546875,
+      "learning_rate": 2.8118745830553707e-05,
+      "loss": 1.3127,
+      "step": 10785
+    },
+    {
+      "epoch": 9.957073621047773,
+      "grad_norm": 3.314453125,
+      "learning_rate": 2.8112074716477655e-05,
+      "loss": 1.2327,
+      "step": 10786
+    },
+    {
+      "epoch": 9.957996768982229,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.8105403602401604e-05,
+      "loss": 1.4663,
+      "step": 10787
+    },
+    {
+      "epoch": 9.958919916916686,
+      "grad_norm": 4.01171875,
+      "learning_rate": 2.8098732488325553e-05,
+      "loss": 1.3862,
+      "step": 10788
+    },
+    {
+      "epoch": 9.959843064851142,
+      "grad_norm": 2.919921875,
+      "learning_rate": 2.8092061374249502e-05,
+      "loss": 1.3633,
+      "step": 10789
+    },
+    {
+      "epoch": 9.960766212785598,
+      "grad_norm": 3.6796875,
+      "learning_rate": 2.808539026017345e-05,
+      "loss": 1.22,
+      "step": 10790
+    },
+    {
+      "epoch": 9.961689360720055,
+      "grad_norm": 3.01953125,
+      "learning_rate": 2.80787191460974e-05,
+      "loss": 1.3314,
+      "step": 10791
+    },
+    {
+      "epoch": 9.962612508654512,
+      "grad_norm": 3.166015625,
+      "learning_rate": 2.8072048032021348e-05,
+      "loss": 1.2889,
+      "step": 10792
+    },
+    {
+      "epoch": 9.963535656588968,
+      "grad_norm": 3.7109375,
+      "learning_rate": 2.8065376917945297e-05,
+      "loss": 1.4243,
+      "step": 10793
+    },
+    {
+      "epoch": 9.964458804523424,
+      "grad_norm": 3.177734375,
+      "learning_rate": 2.8058705803869246e-05,
+      "loss": 1.2224,
+      "step": 10794
+    },
+    {
+      "epoch": 9.965381952457882,
+      "grad_norm": 3.251953125,
+      "learning_rate": 2.8052034689793195e-05,
+      "loss": 1.3453,
+      "step": 10795
+    },
+    {
+      "epoch": 9.966305100392338,
+      "grad_norm": 3.025390625,
+      "learning_rate": 2.8045363575717147e-05,
+      "loss": 1.1987,
+      "step": 10796
+    },
+    {
+      "epoch": 9.967228248326794,
+      "grad_norm": 3.38671875,
+      "learning_rate": 2.8038692461641096e-05,
+      "loss": 1.5261,
+      "step": 10797
+    },
+    {
+      "epoch": 9.96815139626125,
+      "grad_norm": 3.720703125,
+      "learning_rate": 2.8032021347565045e-05,
+      "loss": 1.2014,
+      "step": 10798
+    },
+    {
+      "epoch": 9.969074544195708,
+      "grad_norm": 3.341796875,
+      "learning_rate": 2.8025350233488993e-05,
+      "loss": 1.2963,
+      "step": 10799
+    },
+    {
+      "epoch": 9.969997692130164,
+      "grad_norm": 2.8359375,
+      "learning_rate": 2.8018679119412942e-05,
+      "loss": 0.98,
+      "step": 10800
+    },
+    {
+      "epoch": 9.97092084006462,
+      "grad_norm": 3.138671875,
+      "learning_rate": 2.8012008005336894e-05,
+      "loss": 1.3085,
+      "step": 10801
+    },
+    {
+      "epoch": 9.971843987999076,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8005336891260843e-05,
+      "loss": 1.2341,
+      "step": 10802
+    },
+    {
+      "epoch": 9.972767135933534,
+      "grad_norm": 3.42578125,
+      "learning_rate": 2.7998665777184792e-05,
+      "loss": 1.2002,
+      "step": 10803
+    },
+    {
+      "epoch": 9.97369028386799,
+      "grad_norm": 3.458984375,
+      "learning_rate": 2.799199466310874e-05,
+      "loss": 1.2987,
+      "step": 10804
+    },
+    {
+      "epoch": 9.974613431802446,
+      "grad_norm": 3.22265625,
+      "learning_rate": 2.798532354903269e-05,
+      "loss": 1.469,
+      "step": 10805
+    },
+    {
+      "epoch": 9.975536579736904,
+      "grad_norm": 3.220703125,
+      "learning_rate": 2.797865243495664e-05,
+      "loss": 1.4844,
+      "step": 10806
+    },
+    {
+      "epoch": 9.97645972767136,
+      "grad_norm": 3.384765625,
+      "learning_rate": 2.7971981320880587e-05,
+      "loss": 1.2629,
+      "step": 10807
+    },
+    {
+      "epoch": 9.977382875605816,
+      "grad_norm": 3.513671875,
+      "learning_rate": 2.796531020680454e-05,
+      "loss": 1.2738,
+      "step": 10808
+    },
+    {
+      "epoch": 9.978306023540272,
+      "grad_norm": 3.263671875,
+      "learning_rate": 2.795863909272849e-05,
+      "loss": 0.9812,
+      "step": 10809
+    },
+    {
+      "epoch": 9.97922917147473,
+      "grad_norm": 3.51171875,
+      "learning_rate": 2.7951967978652437e-05,
+      "loss": 1.2695,
+      "step": 10810
+    },
+    {
+      "epoch": 9.980152319409186,
+      "grad_norm": 3.76953125,
+      "learning_rate": 2.7945296864576386e-05,
+      "loss": 1.6055,
+      "step": 10811
+    },
+    {
+      "epoch": 9.981075467343642,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.7938625750500335e-05,
+      "loss": 1.2748,
+      "step": 10812
+    },
+    {
+      "epoch": 9.981998615278098,
+      "grad_norm": 3.580078125,
+      "learning_rate": 2.7931954636424284e-05,
+      "loss": 1.4199,
+      "step": 10813
+    },
+    {
+      "epoch": 9.982921763212556,
+      "grad_norm": 3.318359375,
+      "learning_rate": 2.7925283522348232e-05,
+      "loss": 1.3967,
+      "step": 10814
+    },
+    {
+      "epoch": 9.983844911147012,
+      "grad_norm": 3.00390625,
+      "learning_rate": 2.791861240827218e-05,
+      "loss": 1.2817,
+      "step": 10815
+    },
+    {
+      "epoch": 9.984768059081468,
+      "grad_norm": 3.35546875,
+      "learning_rate": 2.791194129419613e-05,
+      "loss": 1.2255,
+      "step": 10816
+    },
+    {
+      "epoch": 9.985691207015924,
+      "grad_norm": 3.201171875,
+      "learning_rate": 2.790527018012008e-05,
+      "loss": 1.2657,
+      "step": 10817
+    },
+    {
+      "epoch": 9.986614354950381,
+      "grad_norm": 3.736328125,
+      "learning_rate": 2.7898599066044028e-05,
+      "loss": 1.1053,
+      "step": 10818
+    },
+    {
+      "epoch": 9.987537502884837,
+      "grad_norm": 3.689453125,
+      "learning_rate": 2.7891927951967977e-05,
+      "loss": 1.3953,
+      "step": 10819
+    },
+    {
+      "epoch": 9.988460650819293,
+      "grad_norm": 3.4453125,
+      "learning_rate": 2.7885256837891925e-05,
+      "loss": 1.3845,
+      "step": 10820
+    },
+    {
+      "epoch": 9.98938379875375,
+      "grad_norm": 3.517578125,
+      "learning_rate": 2.787858572381588e-05,
+      "loss": 1.322,
+      "step": 10821
+    },
+    {
+      "epoch": 9.990306946688207,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.787191460973983e-05,
+      "loss": 1.3586,
+      "step": 10822
+    },
+    {
+      "epoch": 9.991230094622663,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.786524349566378e-05,
+      "loss": 1.3462,
+      "step": 10823
+    },
+    {
+      "epoch": 9.99215324255712,
+      "grad_norm": 3.205078125,
+      "learning_rate": 2.7858572381587727e-05,
+      "loss": 1.4741,
+      "step": 10824
+    },
+    {
+      "epoch": 9.993076390491577,
+      "grad_norm": 3.701171875,
+      "learning_rate": 2.7851901267511676e-05,
+      "loss": 1.3833,
+      "step": 10825
+    },
+    {
+      "epoch": 9.993999538426033,
+      "grad_norm": 2.98046875,
+      "learning_rate": 2.7845230153435625e-05,
+      "loss": 1.5188,
+      "step": 10826
+    },
+    {
+      "epoch": 9.99492268636049,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.7838559039359574e-05,
+      "loss": 1.043,
+      "step": 10827
+    },
+    {
+      "epoch": 9.995845834294945,
+      "grad_norm": 3.14453125,
+      "learning_rate": 2.7831887925283523e-05,
+      "loss": 1.2474,
+      "step": 10828
+    },
+    {
+      "epoch": 9.996768982229403,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.782521681120747e-05,
+      "loss": 1.2456,
+      "step": 10829
+    },
+    {
+      "epoch": 9.997692130163859,
+      "grad_norm": 2.986328125,
+      "learning_rate": 2.781854569713142e-05,
+      "loss": 1.2914,
+      "step": 10830
+    },
+    {
+      "epoch": 9.998615278098315,
+      "grad_norm": 3.333984375,
+      "learning_rate": 2.781187458305537e-05,
+      "loss": 1.296,
+      "step": 10831
+    },
+    {
+      "epoch": 9.999538426032771,
+      "grad_norm": 3.009765625,
+      "learning_rate": 2.7805203468979318e-05,
+      "loss": 1.2443,
+      "step": 10832
+    },
+    {
+      "epoch": 10.000461573967229,
+      "grad_norm": 5.08984375,
+      "learning_rate": 2.7798532354903274e-05,
+      "loss": 1.2217,
+      "step": 10833
+    },
+    {
+      "epoch": 10.001384721901685,
+      "grad_norm": 3.462890625,
+      "learning_rate": 2.7791861240827222e-05,
+      "loss": 1.262,
+      "step": 10834
+    },
+    {
+      "epoch": 10.002307869836141,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.778519012675117e-05,
+      "loss": 1.2539,
+      "step": 10835
+    },
+    {
+      "epoch": 10.003231017770597,
+      "grad_norm": 3.806640625,
+      "learning_rate": 2.777851901267512e-05,
+      "loss": 1.1824,
+      "step": 10836
+    },
+    {
+      "epoch": 10.004154165705055,
+      "grad_norm": 3.5,
+      "learning_rate": 2.777184789859907e-05,
+      "loss": 1.3191,
+      "step": 10837
+    },
+    {
+      "epoch": 10.00507731363951,
+      "grad_norm": 2.970703125,
+      "learning_rate": 2.7765176784523018e-05,
+      "loss": 1.3291,
+      "step": 10838
+    },
+    {
+      "epoch": 10.006000461573967,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.7758505670446966e-05,
+      "loss": 1.2988,
+      "step": 10839
+    },
+    {
+      "epoch": 10.006923609508423,
+      "grad_norm": 3.107421875,
+      "learning_rate": 2.7751834556370915e-05,
+      "loss": 1.3203,
+      "step": 10840
+    },
+    {
+      "epoch": 10.00784675744288,
+      "grad_norm": 2.994140625,
+      "learning_rate": 2.7745163442294864e-05,
+      "loss": 1.3573,
+      "step": 10841
+    },
+    {
+      "epoch": 10.008769905377337,
+      "grad_norm": 2.916015625,
+      "learning_rate": 2.7738492328218813e-05,
+      "loss": 1.1449,
+      "step": 10842
+    },
+    {
+      "epoch": 10.009693053311793,
+      "grad_norm": 3.18359375,
+      "learning_rate": 2.7731821214142762e-05,
+      "loss": 1.0464,
+      "step": 10843
+    },
+    {
+      "epoch": 10.01061620124625,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.772515010006671e-05,
+      "loss": 1.1146,
+      "step": 10844
+    },
+    {
+      "epoch": 10.011539349180707,
+      "grad_norm": 3.279296875,
+      "learning_rate": 2.771847898599066e-05,
+      "loss": 1.3879,
+      "step": 10845
+    },
+    {
+      "epoch": 10.012462497115163,
+      "grad_norm": 3.001953125,
+      "learning_rate": 2.7711807871914615e-05,
+      "loss": 1.1678,
+      "step": 10846
+    },
+    {
+      "epoch": 10.013385645049619,
+      "grad_norm": 3.2265625,
+      "learning_rate": 2.7705136757838564e-05,
+      "loss": 1.1982,
+      "step": 10847
+    },
+    {
+      "epoch": 10.014308792984076,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.7698465643762513e-05,
+      "loss": 1.1157,
+      "step": 10848
+    },
+    {
+      "epoch": 10.015231940918532,
+      "grad_norm": 3.404296875,
+      "learning_rate": 2.769179452968646e-05,
+      "loss": 1.4526,
+      "step": 10849
+    },
+    {
+      "epoch": 10.016155088852988,
+      "grad_norm": 3.34765625,
+      "learning_rate": 2.768512341561041e-05,
+      "loss": 1.0066,
+      "step": 10850
+    },
+    {
+      "epoch": 10.017078236787444,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.767845230153436e-05,
+      "loss": 1.1068,
+      "step": 10851
+    },
+    {
+      "epoch": 10.018001384721902,
+      "grad_norm": 3.525390625,
+      "learning_rate": 2.7671781187458308e-05,
+      "loss": 1.347,
+      "step": 10852
+    },
+    {
+      "epoch": 10.018924532656358,
+      "grad_norm": 3.58203125,
+      "learning_rate": 2.7665110073382257e-05,
+      "loss": 1.0853,
+      "step": 10853
+    },
+    {
+      "epoch": 10.019847680590814,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.7658438959306206e-05,
+      "loss": 1.1831,
+      "step": 10854
+    },
+    {
+      "epoch": 10.02077082852527,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7651767845230154e-05,
+      "loss": 1.3147,
+      "step": 10855
+    },
+    {
+      "epoch": 10.021693976459728,
+      "grad_norm": 3.169921875,
+      "learning_rate": 2.7645096731154103e-05,
+      "loss": 1.2308,
+      "step": 10856
+    },
+    {
+      "epoch": 10.022617124394184,
+      "grad_norm": 3.01171875,
+      "learning_rate": 2.7638425617078052e-05,
+      "loss": 1.3114,
+      "step": 10857
+    },
+    {
+      "epoch": 10.02354027232864,
+      "grad_norm": 3.138671875,
+      "learning_rate": 2.7631754503002004e-05,
+      "loss": 1.4937,
+      "step": 10858
+    },
+    {
+      "epoch": 10.024463420263098,
+      "grad_norm": 3.353515625,
+      "learning_rate": 2.7625083388925953e-05,
+      "loss": 1.3162,
+      "step": 10859
+    },
+    {
+      "epoch": 10.025386568197554,
+      "grad_norm": 2.943359375,
+      "learning_rate": 2.7618412274849902e-05,
+      "loss": 1.3455,
+      "step": 10860
+    },
+    {
+      "epoch": 10.02630971613201,
+      "grad_norm": 3.474609375,
+      "learning_rate": 2.761174116077385e-05,
+      "loss": 1.2197,
+      "step": 10861
+    },
+    {
+      "epoch": 10.027232864066466,
+      "grad_norm": 3.169921875,
+      "learning_rate": 2.76050700466978e-05,
+      "loss": 1.1654,
+      "step": 10862
+    },
+    {
+      "epoch": 10.028156012000924,
+      "grad_norm": 2.9765625,
+      "learning_rate": 2.7598398932621748e-05,
+      "loss": 1.2458,
+      "step": 10863
+    },
+    {
+      "epoch": 10.02907915993538,
+      "grad_norm": 3.587890625,
+      "learning_rate": 2.7591727818545697e-05,
+      "loss": 1.2656,
+      "step": 10864
+    },
+    {
+      "epoch": 10.030002307869836,
+      "grad_norm": 2.900390625,
+      "learning_rate": 2.7585056704469646e-05,
+      "loss": 1.3196,
+      "step": 10865
+    },
+    {
+      "epoch": 10.030925455804292,
+      "grad_norm": 3.427734375,
+      "learning_rate": 2.7578385590393595e-05,
+      "loss": 1.2437,
+      "step": 10866
+    },
+    {
+      "epoch": 10.03184860373875,
+      "grad_norm": 3.0390625,
+      "learning_rate": 2.7571714476317544e-05,
+      "loss": 1.3882,
+      "step": 10867
+    },
+    {
+      "epoch": 10.032771751673206,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.7565043362241492e-05,
+      "loss": 1.292,
+      "step": 10868
+    },
+    {
+      "epoch": 10.033694899607662,
+      "grad_norm": 3.42578125,
+      "learning_rate": 2.755837224816544e-05,
+      "loss": 1.3459,
+      "step": 10869
+    },
+    {
+      "epoch": 10.034618047542118,
+      "grad_norm": 3.126953125,
+      "learning_rate": 2.755170113408939e-05,
+      "loss": 1.004,
+      "step": 10870
+    },
+    {
+      "epoch": 10.035541195476576,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.7545030020013346e-05,
+      "loss": 1.0841,
+      "step": 10871
+    },
+    {
+      "epoch": 10.036464343411032,
+      "grad_norm": 3.490234375,
+      "learning_rate": 2.7538358905937294e-05,
+      "loss": 1.4314,
+      "step": 10872
+    },
+    {
+      "epoch": 10.037387491345488,
+      "grad_norm": 2.77734375,
+      "learning_rate": 2.7531687791861243e-05,
+      "loss": 1.3921,
+      "step": 10873
+    },
+    {
+      "epoch": 10.038310639279945,
+      "grad_norm": 3.673828125,
+      "learning_rate": 2.7525016677785192e-05,
+      "loss": 1.0985,
+      "step": 10874
+    },
+    {
+      "epoch": 10.039233787214402,
+      "grad_norm": 2.8828125,
+      "learning_rate": 2.751834556370914e-05,
+      "loss": 1.3223,
+      "step": 10875
+    },
+    {
+      "epoch": 10.040156935148858,
+      "grad_norm": 3.095703125,
+      "learning_rate": 2.751167444963309e-05,
+      "loss": 1.0845,
+      "step": 10876
+    },
+    {
+      "epoch": 10.041080083083314,
+      "grad_norm": 2.94140625,
+      "learning_rate": 2.750500333555704e-05,
+      "loss": 1.3879,
+      "step": 10877
+    },
+    {
+      "epoch": 10.042003231017771,
+      "grad_norm": 3.65234375,
+      "learning_rate": 2.7498332221480987e-05,
+      "loss": 1.2825,
+      "step": 10878
+    },
+    {
+      "epoch": 10.042926378952227,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.7491661107404936e-05,
+      "loss": 1.0104,
+      "step": 10879
+    },
+    {
+      "epoch": 10.043849526886683,
+      "grad_norm": 2.994140625,
+      "learning_rate": 2.7484989993328885e-05,
+      "loss": 1.0714,
+      "step": 10880
+    },
+    {
+      "epoch": 10.04477267482114,
+      "grad_norm": 3.1796875,
+      "learning_rate": 2.7478318879252834e-05,
+      "loss": 1.2426,
+      "step": 10881
+    },
+    {
+      "epoch": 10.045695822755597,
+      "grad_norm": 3.236328125,
+      "learning_rate": 2.7471647765176783e-05,
+      "loss": 1.2717,
+      "step": 10882
+    },
+    {
+      "epoch": 10.046618970690053,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7464976651100738e-05,
+      "loss": 1.2343,
+      "step": 10883
+    },
+    {
+      "epoch": 10.04754211862451,
+      "grad_norm": 3.599609375,
+      "learning_rate": 2.7458305537024687e-05,
+      "loss": 1.217,
+      "step": 10884
+    },
+    {
+      "epoch": 10.048465266558965,
+      "grad_norm": 3.048828125,
+      "learning_rate": 2.7451634422948636e-05,
+      "loss": 1.395,
+      "step": 10885
+    },
+    {
+      "epoch": 10.049388414493423,
+      "grad_norm": 3.126953125,
+      "learning_rate": 2.7444963308872585e-05,
+      "loss": 1.3171,
+      "step": 10886
+    },
+    {
+      "epoch": 10.05031156242788,
+      "grad_norm": 3.337890625,
+      "learning_rate": 2.7438292194796533e-05,
+      "loss": 1.0693,
+      "step": 10887
+    },
+    {
+      "epoch": 10.051234710362335,
+      "grad_norm": 3.5234375,
+      "learning_rate": 2.7431621080720482e-05,
+      "loss": 1.1869,
+      "step": 10888
+    },
+    {
+      "epoch": 10.052157858296791,
+      "grad_norm": 3.490234375,
+      "learning_rate": 2.742494996664443e-05,
+      "loss": 1.2388,
+      "step": 10889
+    },
+    {
+      "epoch": 10.053081006231249,
+      "grad_norm": 3.75390625,
+      "learning_rate": 2.741827885256838e-05,
+      "loss": 1.3823,
+      "step": 10890
+    },
+    {
+      "epoch": 10.054004154165705,
+      "grad_norm": 2.681640625,
+      "learning_rate": 2.741160773849233e-05,
+      "loss": 1.0879,
+      "step": 10891
+    },
+    {
+      "epoch": 10.054927302100161,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.7404936624416278e-05,
+      "loss": 1.3909,
+      "step": 10892
+    },
+    {
+      "epoch": 10.055850450034619,
+      "grad_norm": 3.41015625,
+      "learning_rate": 2.7398265510340226e-05,
+      "loss": 1.2393,
+      "step": 10893
+    },
+    {
+      "epoch": 10.056773597969075,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.7391594396264175e-05,
+      "loss": 1.3691,
+      "step": 10894
+    },
+    {
+      "epoch": 10.057696745903531,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.7384923282188124e-05,
+      "loss": 0.9154,
+      "step": 10895
+    },
+    {
+      "epoch": 10.058619893837987,
+      "grad_norm": 2.955078125,
+      "learning_rate": 2.737825216811208e-05,
+      "loss": 1.2527,
+      "step": 10896
+    },
+    {
+      "epoch": 10.059543041772445,
+      "grad_norm": 2.955078125,
+      "learning_rate": 2.737158105403603e-05,
+      "loss": 1.3096,
+      "step": 10897
+    },
+    {
+      "epoch": 10.0604661897069,
+      "grad_norm": 3.517578125,
+      "learning_rate": 2.7364909939959977e-05,
+      "loss": 1.4043,
+      "step": 10898
+    },
+    {
+      "epoch": 10.061389337641357,
+      "grad_norm": 3.111328125,
+      "learning_rate": 2.7358238825883926e-05,
+      "loss": 1.4358,
+      "step": 10899
+    },
+    {
+      "epoch": 10.062312485575813,
+      "grad_norm": 3.58203125,
+      "learning_rate": 2.7351567711807875e-05,
+      "loss": 1.3459,
+      "step": 10900
+    },
+    {
+      "epoch": 10.06323563351027,
+      "grad_norm": 4.29296875,
+      "learning_rate": 2.7344896597731824e-05,
+      "loss": 0.8462,
+      "step": 10901
+    },
+    {
+      "epoch": 10.064158781444727,
+      "grad_norm": 2.744140625,
+      "learning_rate": 2.7338225483655773e-05,
+      "loss": 1.4028,
+      "step": 10902
+    },
+    {
+      "epoch": 10.065081929379183,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.733155436957972e-05,
+      "loss": 1.271,
+      "step": 10903
+    },
+    {
+      "epoch": 10.066005077313639,
+      "grad_norm": 3.1171875,
+      "learning_rate": 2.732488325550367e-05,
+      "loss": 1.3123,
+      "step": 10904
+    },
+    {
+      "epoch": 10.066928225248096,
+      "grad_norm": 4.01171875,
+      "learning_rate": 2.731821214142762e-05,
+      "loss": 1.3276,
+      "step": 10905
+    },
+    {
+      "epoch": 10.067851373182553,
+      "grad_norm": 3.080078125,
+      "learning_rate": 2.7311541027351568e-05,
+      "loss": 1.4836,
+      "step": 10906
+    },
+    {
+      "epoch": 10.068774521117009,
+      "grad_norm": 3.5078125,
+      "learning_rate": 2.7304869913275517e-05,
+      "loss": 1.1223,
+      "step": 10907
+    },
+    {
+      "epoch": 10.069697669051466,
+      "grad_norm": 3.1015625,
+      "learning_rate": 2.729819879919947e-05,
+      "loss": 1.4026,
+      "step": 10908
+    },
+    {
+      "epoch": 10.070620816985922,
+      "grad_norm": 3.1484375,
+      "learning_rate": 2.7291527685123418e-05,
+      "loss": 1.1455,
+      "step": 10909
+    },
+    {
+      "epoch": 10.071543964920378,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.7284856571047366e-05,
+      "loss": 1.406,
+      "step": 10910
+    },
+    {
+      "epoch": 10.072467112854834,
+      "grad_norm": 3.15234375,
+      "learning_rate": 2.7278185456971315e-05,
+      "loss": 1.4285,
+      "step": 10911
+    },
+    {
+      "epoch": 10.073390260789292,
+      "grad_norm": 3.236328125,
+      "learning_rate": 2.7271514342895264e-05,
+      "loss": 1.4985,
+      "step": 10912
+    },
+    {
+      "epoch": 10.074313408723748,
+      "grad_norm": 3.78515625,
+      "learning_rate": 2.7264843228819213e-05,
+      "loss": 1.4312,
+      "step": 10913
+    },
+    {
+      "epoch": 10.075236556658204,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.7258172114743165e-05,
+      "loss": 1.229,
+      "step": 10914
+    },
+    {
+      "epoch": 10.07615970459266,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.7251501000667114e-05,
+      "loss": 1.0577,
+      "step": 10915
+    },
+    {
+      "epoch": 10.077082852527118,
+      "grad_norm": 3.04296875,
+      "learning_rate": 2.7244829886591063e-05,
+      "loss": 1.2371,
+      "step": 10916
+    },
+    {
+      "epoch": 10.078006000461574,
+      "grad_norm": 3.001953125,
+      "learning_rate": 2.723815877251501e-05,
+      "loss": 1.2723,
+      "step": 10917
+    },
+    {
+      "epoch": 10.07892914839603,
+      "grad_norm": 3.072265625,
+      "learning_rate": 2.723148765843896e-05,
+      "loss": 1.1809,
+      "step": 10918
+    },
+    {
+      "epoch": 10.079852296330486,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.722481654436291e-05,
+      "loss": 1.5308,
+      "step": 10919
+    },
+    {
+      "epoch": 10.080775444264944,
+      "grad_norm": 3.380859375,
+      "learning_rate": 2.7218145430286858e-05,
+      "loss": 1.3196,
+      "step": 10920
+    },
+    {
+      "epoch": 10.0816985921994,
+      "grad_norm": 3.482421875,
+      "learning_rate": 2.721147431621081e-05,
+      "loss": 1.3911,
+      "step": 10921
+    },
+    {
+      "epoch": 10.082621740133856,
+      "grad_norm": 3.416015625,
+      "learning_rate": 2.720480320213476e-05,
+      "loss": 1.5359,
+      "step": 10922
+    },
+    {
+      "epoch": 10.083544888068314,
+      "grad_norm": 3.2578125,
+      "learning_rate": 2.7198132088058708e-05,
+      "loss": 1.0441,
+      "step": 10923
+    },
+    {
+      "epoch": 10.08446803600277,
+      "grad_norm": 3.666015625,
+      "learning_rate": 2.7191460973982657e-05,
+      "loss": 1.2574,
+      "step": 10924
+    },
+    {
+      "epoch": 10.085391183937226,
+      "grad_norm": 3.173828125,
+      "learning_rate": 2.7184789859906606e-05,
+      "loss": 1.2007,
+      "step": 10925
+    },
+    {
+      "epoch": 10.086314331871682,
+      "grad_norm": 3.412109375,
+      "learning_rate": 2.7178118745830554e-05,
+      "loss": 1.3733,
+      "step": 10926
+    },
+    {
+      "epoch": 10.08723747980614,
+      "grad_norm": 3.275390625,
+      "learning_rate": 2.7171447631754503e-05,
+      "loss": 1.1111,
+      "step": 10927
+    },
+    {
+      "epoch": 10.088160627740596,
+      "grad_norm": 3.2578125,
+      "learning_rate": 2.7164776517678452e-05,
+      "loss": 1.1251,
+      "step": 10928
+    },
+    {
+      "epoch": 10.089083775675052,
+      "grad_norm": 2.900390625,
+      "learning_rate": 2.71581054036024e-05,
+      "loss": 1.3302,
+      "step": 10929
+    },
+    {
+      "epoch": 10.090006923609508,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.715143428952635e-05,
+      "loss": 1.3535,
+      "step": 10930
+    },
+    {
+      "epoch": 10.090930071543966,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.71447631754503e-05,
+      "loss": 1.2788,
+      "step": 10931
+    },
+    {
+      "epoch": 10.091853219478422,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.7138092061374247e-05,
+      "loss": 1.1101,
+      "step": 10932
+    },
+    {
+      "epoch": 10.092776367412878,
+      "grad_norm": 3.244140625,
+      "learning_rate": 2.7131420947298203e-05,
+      "loss": 0.9569,
+      "step": 10933
+    },
+    {
+      "epoch": 10.093699515347334,
+      "grad_norm": 3.40234375,
+      "learning_rate": 2.712474983322215e-05,
+      "loss": 1.2355,
+      "step": 10934
+    },
+    {
+      "epoch": 10.094622663281791,
+      "grad_norm": 3.091796875,
+      "learning_rate": 2.71180787191461e-05,
+      "loss": 1.3427,
+      "step": 10935
+    },
+    {
+      "epoch": 10.095545811216248,
+      "grad_norm": 3.724609375,
+      "learning_rate": 2.711140760507005e-05,
+      "loss": 1.2183,
+      "step": 10936
+    },
+    {
+      "epoch": 10.096468959150704,
+      "grad_norm": 3.185546875,
+      "learning_rate": 2.7104736490993998e-05,
+      "loss": 1.3173,
+      "step": 10937
+    },
+    {
+      "epoch": 10.09739210708516,
+      "grad_norm": 3.3359375,
+      "learning_rate": 2.7098065376917947e-05,
+      "loss": 1.2235,
+      "step": 10938
+    },
+    {
+      "epoch": 10.098315255019617,
+      "grad_norm": 3.455078125,
+      "learning_rate": 2.7091394262841896e-05,
+      "loss": 1.3428,
+      "step": 10939
+    },
+    {
+      "epoch": 10.099238402954073,
+      "grad_norm": 3.2421875,
+      "learning_rate": 2.7084723148765845e-05,
+      "loss": 1.3494,
+      "step": 10940
+    },
+    {
+      "epoch": 10.10016155088853,
+      "grad_norm": 3.181640625,
+      "learning_rate": 2.7078052034689793e-05,
+      "loss": 1.019,
+      "step": 10941
+    },
+    {
+      "epoch": 10.101084698822987,
+      "grad_norm": 2.49609375,
+      "learning_rate": 2.7071380920613742e-05,
+      "loss": 0.9955,
+      "step": 10942
+    },
+    {
+      "epoch": 10.102007846757443,
+      "grad_norm": 2.919921875,
+      "learning_rate": 2.706470980653769e-05,
+      "loss": 1.2303,
+      "step": 10943
+    },
+    {
+      "epoch": 10.1029309946919,
+      "grad_norm": 3.10546875,
+      "learning_rate": 2.705803869246164e-05,
+      "loss": 1.127,
+      "step": 10944
+    },
+    {
+      "epoch": 10.103854142626355,
+      "grad_norm": 3.3046875,
+      "learning_rate": 2.705136757838559e-05,
+      "loss": 1.3999,
+      "step": 10945
+    },
+    {
+      "epoch": 10.104777290560813,
+      "grad_norm": 3.373046875,
+      "learning_rate": 2.7044696464309544e-05,
+      "loss": 1.2949,
+      "step": 10946
+    },
+    {
+      "epoch": 10.105700438495269,
+      "grad_norm": 3.384765625,
+      "learning_rate": 2.7038025350233493e-05,
+      "loss": 1.3579,
+      "step": 10947
+    },
+    {
+      "epoch": 10.106623586429725,
+      "grad_norm": 3.49609375,
+      "learning_rate": 2.7031354236157442e-05,
+      "loss": 1.2046,
+      "step": 10948
+    },
+    {
+      "epoch": 10.107546734364181,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.702468312208139e-05,
+      "loss": 1.3177,
+      "step": 10949
+    },
+    {
+      "epoch": 10.108469882298639,
+      "grad_norm": 3.06640625,
+      "learning_rate": 2.701801200800534e-05,
+      "loss": 1.2424,
+      "step": 10950
+    },
+    {
+      "epoch": 10.109393030233095,
+      "grad_norm": 3.033203125,
+      "learning_rate": 2.701134089392929e-05,
+      "loss": 0.8742,
+      "step": 10951
+    },
+    {
+      "epoch": 10.110316178167551,
+      "grad_norm": 2.81640625,
+      "learning_rate": 2.7004669779853237e-05,
+      "loss": 1.0515,
+      "step": 10952
+    },
+    {
+      "epoch": 10.111239326102007,
+      "grad_norm": 3.55859375,
+      "learning_rate": 2.6997998665777186e-05,
+      "loss": 1.0576,
+      "step": 10953
+    },
+    {
+      "epoch": 10.112162474036465,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.6991327551701135e-05,
+      "loss": 1.2112,
+      "step": 10954
+    },
+    {
+      "epoch": 10.11308562197092,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.6984656437625084e-05,
+      "loss": 1.1702,
+      "step": 10955
+    },
+    {
+      "epoch": 10.114008769905377,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.6977985323549032e-05,
+      "loss": 1.2434,
+      "step": 10956
+    },
+    {
+      "epoch": 10.114931917839835,
+      "grad_norm": 3.435546875,
+      "learning_rate": 2.697131420947298e-05,
+      "loss": 1.1224,
+      "step": 10957
+    },
+    {
+      "epoch": 10.11585506577429,
+      "grad_norm": 3.0390625,
+      "learning_rate": 2.6964643095396937e-05,
+      "loss": 1.2122,
+      "step": 10958
+    },
+    {
+      "epoch": 10.116778213708747,
+      "grad_norm": 3.462890625,
+      "learning_rate": 2.6957971981320886e-05,
+      "loss": 1.2832,
+      "step": 10959
+    },
+    {
+      "epoch": 10.117701361643203,
+      "grad_norm": 3.03515625,
+      "learning_rate": 2.6951300867244835e-05,
+      "loss": 1.2891,
+      "step": 10960
+    },
+    {
+      "epoch": 10.11862450957766,
+      "grad_norm": 3.626953125,
+      "learning_rate": 2.6944629753168783e-05,
+      "loss": 1.2515,
+      "step": 10961
+    },
+    {
+      "epoch": 10.119547657512117,
+      "grad_norm": 3.599609375,
+      "learning_rate": 2.6937958639092732e-05,
+      "loss": 1.427,
+      "step": 10962
+    },
+    {
+      "epoch": 10.120470805446573,
+      "grad_norm": 3.037109375,
+      "learning_rate": 2.693128752501668e-05,
+      "loss": 1.2146,
+      "step": 10963
+    },
+    {
+      "epoch": 10.121393953381029,
+      "grad_norm": 3.15234375,
+      "learning_rate": 2.692461641094063e-05,
+      "loss": 1.272,
+      "step": 10964
+    },
+    {
+      "epoch": 10.122317101315486,
+      "grad_norm": 3.427734375,
+      "learning_rate": 2.691794529686458e-05,
+      "loss": 1.3796,
+      "step": 10965
+    },
+    {
+      "epoch": 10.123240249249942,
+      "grad_norm": 3.248046875,
+      "learning_rate": 2.6911274182788527e-05,
+      "loss": 1.4517,
+      "step": 10966
+    },
+    {
+      "epoch": 10.124163397184399,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.6904603068712476e-05,
+      "loss": 0.8477,
+      "step": 10967
+    },
+    {
+      "epoch": 10.125086545118855,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.6897931954636425e-05,
+      "loss": 1.3311,
+      "step": 10968
+    },
+    {
+      "epoch": 10.126009693053312,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.6891260840560374e-05,
+      "loss": 1.3257,
+      "step": 10969
+    },
+    {
+      "epoch": 10.126932840987768,
+      "grad_norm": 3.0390625,
+      "learning_rate": 2.6884589726484323e-05,
+      "loss": 1.2832,
+      "step": 10970
+    },
+    {
+      "epoch": 10.127855988922224,
+      "grad_norm": 3.126953125,
+      "learning_rate": 2.6877918612408275e-05,
+      "loss": 1.4048,
+      "step": 10971
+    },
+    {
+      "epoch": 10.12877913685668,
+      "grad_norm": 3.189453125,
+      "learning_rate": 2.6871247498332224e-05,
+      "loss": 1.1671,
+      "step": 10972
+    },
+    {
+      "epoch": 10.129702284791138,
+      "grad_norm": 3.400390625,
+      "learning_rate": 2.6864576384256173e-05,
+      "loss": 1.2053,
+      "step": 10973
+    },
+    {
+      "epoch": 10.130625432725594,
+      "grad_norm": 2.826171875,
+      "learning_rate": 2.685790527018012e-05,
+      "loss": 1.07,
+      "step": 10974
+    },
+    {
+      "epoch": 10.13154858066005,
+      "grad_norm": 3.228515625,
+      "learning_rate": 2.685123415610407e-05,
+      "loss": 1.511,
+      "step": 10975
+    },
+    {
+      "epoch": 10.132471728594508,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.684456304202802e-05,
+      "loss": 1.5005,
+      "step": 10976
+    },
+    {
+      "epoch": 10.133394876528964,
+      "grad_norm": 2.94140625,
+      "learning_rate": 2.6837891927951968e-05,
+      "loss": 1.3733,
+      "step": 10977
+    },
+    {
+      "epoch": 10.13431802446342,
+      "grad_norm": 3.6015625,
+      "learning_rate": 2.6831220813875917e-05,
+      "loss": 1.3611,
+      "step": 10978
+    },
+    {
+      "epoch": 10.135241172397876,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.6824549699799865e-05,
+      "loss": 1.2153,
+      "step": 10979
+    },
+    {
+      "epoch": 10.136164320332334,
+      "grad_norm": 2.931640625,
+      "learning_rate": 2.6817878585723814e-05,
+      "loss": 0.9444,
+      "step": 10980
+    },
+    {
+      "epoch": 10.13708746826679,
+      "grad_norm": 3.42578125,
+      "learning_rate": 2.6811207471647763e-05,
+      "loss": 1.3434,
+      "step": 10981
+    },
+    {
+      "epoch": 10.138010616201246,
+      "grad_norm": 3.384765625,
+      "learning_rate": 2.6804536357571712e-05,
+      "loss": 1.2056,
+      "step": 10982
+    },
+    {
+      "epoch": 10.138933764135702,
+      "grad_norm": 3.251953125,
+      "learning_rate": 2.679786524349566e-05,
+      "loss": 1.2971,
+      "step": 10983
+    },
+    {
+      "epoch": 10.13985691207016,
+      "grad_norm": 3.1328125,
+      "learning_rate": 2.6791194129419616e-05,
+      "loss": 1.1326,
+      "step": 10984
+    },
+    {
+      "epoch": 10.140780060004616,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.6784523015343565e-05,
+      "loss": 1.1893,
+      "step": 10985
+    },
+    {
+      "epoch": 10.141703207939072,
+      "grad_norm": 3.232421875,
+      "learning_rate": 2.6777851901267514e-05,
+      "loss": 1.145,
+      "step": 10986
+    },
+    {
+      "epoch": 10.142626355873528,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.6771180787191463e-05,
+      "loss": 1.2905,
+      "step": 10987
+    },
+    {
+      "epoch": 10.143549503807986,
+      "grad_norm": 3.009765625,
+      "learning_rate": 2.676450967311541e-05,
+      "loss": 1.269,
+      "step": 10988
+    },
+    {
+      "epoch": 10.144472651742442,
+      "grad_norm": 3.841796875,
+      "learning_rate": 2.675783855903936e-05,
+      "loss": 1.4163,
+      "step": 10989
+    },
+    {
+      "epoch": 10.145395799676898,
+      "grad_norm": 3.10546875,
+      "learning_rate": 2.675116744496331e-05,
+      "loss": 1.1326,
+      "step": 10990
+    },
+    {
+      "epoch": 10.146318947611356,
+      "grad_norm": 3.013671875,
+      "learning_rate": 2.6744496330887258e-05,
+      "loss": 1.3033,
+      "step": 10991
+    },
+    {
+      "epoch": 10.147242095545812,
+      "grad_norm": 3.255859375,
+      "learning_rate": 2.6737825216811207e-05,
+      "loss": 1.3171,
+      "step": 10992
+    },
+    {
+      "epoch": 10.148165243480268,
+      "grad_norm": 3.708984375,
+      "learning_rate": 2.6731154102735156e-05,
+      "loss": 1.2068,
+      "step": 10993
+    },
+    {
+      "epoch": 10.149088391414724,
+      "grad_norm": 3.087890625,
+      "learning_rate": 2.6724482988659105e-05,
+      "loss": 1.4133,
+      "step": 10994
+    },
+    {
+      "epoch": 10.150011539349181,
+      "grad_norm": 3.626953125,
+      "learning_rate": 2.6717811874583053e-05,
+      "loss": 1.1499,
+      "step": 10995
+    },
+    {
+      "epoch": 10.150934687283637,
+      "grad_norm": 3.255859375,
+      "learning_rate": 2.671114076050701e-05,
+      "loss": 1.4189,
+      "step": 10996
+    },
+    {
+      "epoch": 10.151857835218093,
+      "grad_norm": 2.939453125,
+      "learning_rate": 2.6704469646430958e-05,
+      "loss": 1.2664,
+      "step": 10997
+    },
+    {
+      "epoch": 10.15278098315255,
+      "grad_norm": 2.556640625,
+      "learning_rate": 2.6697798532354907e-05,
+      "loss": 0.9904,
+      "step": 10998
+    },
+    {
+      "epoch": 10.153704131087007,
+      "grad_norm": 3.677734375,
+      "learning_rate": 2.6691127418278855e-05,
+      "loss": 1.3474,
+      "step": 10999
+    },
+    {
+      "epoch": 10.154627279021463,
+      "grad_norm": 3.00390625,
+      "learning_rate": 2.6684456304202804e-05,
+      "loss": 1.4656,
+      "step": 11000
+    },
+    {
+      "epoch": 10.154627279021463,
+      "eval_loss": 0.8759765625,
+      "eval_runtime": 23.3787,
+      "eval_samples_per_second": 975.29,
+      "eval_steps_per_second": 60.996,
+      "step": 11000
+    },
+    {
+      "epoch": 10.15555042695592,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.6677785190126753e-05,
+      "loss": 1.0533,
+      "step": 11001
+    },
+    {
+      "epoch": 10.156473574890375,
+      "grad_norm": 3.189453125,
+      "learning_rate": 2.6671114076050702e-05,
+      "loss": 1.1573,
+      "step": 11002
+    },
+    {
+      "epoch": 10.157396722824833,
+      "grad_norm": 3.4140625,
+      "learning_rate": 2.666444296197465e-05,
+      "loss": 1.2249,
+      "step": 11003
+    },
+    {
+      "epoch": 10.15831987075929,
+      "grad_norm": 3.37890625,
+      "learning_rate": 2.66577718478986e-05,
+      "loss": 1.4675,
+      "step": 11004
+    },
+    {
+      "epoch": 10.159243018693745,
+      "grad_norm": 3.583984375,
+      "learning_rate": 2.6651100733822548e-05,
+      "loss": 1.3075,
+      "step": 11005
+    },
+    {
+      "epoch": 10.160166166628201,
+      "grad_norm": 3.240234375,
+      "learning_rate": 2.6644429619746497e-05,
+      "loss": 1.1882,
+      "step": 11006
+    },
+    {
+      "epoch": 10.161089314562659,
+      "grad_norm": 3.294921875,
+      "learning_rate": 2.6637758505670446e-05,
+      "loss": 1.1886,
+      "step": 11007
+    },
+    {
+      "epoch": 10.162012462497115,
+      "grad_norm": 3.462890625,
+      "learning_rate": 2.6631087391594395e-05,
+      "loss": 1.3103,
+      "step": 11008
+    },
+    {
+      "epoch": 10.162935610431571,
+      "grad_norm": 3.318359375,
+      "learning_rate": 2.662441627751835e-05,
+      "loss": 1.4336,
+      "step": 11009
+    },
+    {
+      "epoch": 10.163858758366029,
+      "grad_norm": 3.369140625,
+      "learning_rate": 2.66177451634423e-05,
+      "loss": 1.2698,
+      "step": 11010
+    },
+    {
+      "epoch": 10.164781906300485,
+      "grad_norm": 3.912109375,
+      "learning_rate": 2.6611074049366248e-05,
+      "loss": 1.4529,
+      "step": 11011
+    },
+    {
+      "epoch": 10.165705054234941,
+      "grad_norm": 3.35546875,
+      "learning_rate": 2.6604402935290197e-05,
+      "loss": 1.049,
+      "step": 11012
+    },
+    {
+      "epoch": 10.166628202169397,
+      "grad_norm": 4.1796875,
+      "learning_rate": 2.6597731821214146e-05,
+      "loss": 1.1143,
+      "step": 11013
+    },
+    {
+      "epoch": 10.167551350103855,
+      "grad_norm": 3.7265625,
+      "learning_rate": 2.6591060707138094e-05,
+      "loss": 1.4624,
+      "step": 11014
+    },
+    {
+      "epoch": 10.16847449803831,
+      "grad_norm": 3.3046875,
+      "learning_rate": 2.6584389593062043e-05,
+      "loss": 1.1818,
+      "step": 11015
+    },
+    {
+      "epoch": 10.169397645972767,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.6577718478985992e-05,
+      "loss": 1.3721,
+      "step": 11016
+    },
+    {
+      "epoch": 10.170320793907223,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.657104736490994e-05,
+      "loss": 1.2249,
+      "step": 11017
+    },
+    {
+      "epoch": 10.17124394184168,
+      "grad_norm": 3.263671875,
+      "learning_rate": 2.656437625083389e-05,
+      "loss": 0.9724,
+      "step": 11018
+    },
+    {
+      "epoch": 10.172167089776137,
+      "grad_norm": 3.470703125,
+      "learning_rate": 2.655770513675784e-05,
+      "loss": 1.1067,
+      "step": 11019
+    },
+    {
+      "epoch": 10.173090237710593,
+      "grad_norm": 3.412109375,
+      "learning_rate": 2.6551034022681787e-05,
+      "loss": 1.2017,
+      "step": 11020
+    },
+    {
+      "epoch": 10.174013385645049,
+      "grad_norm": 3.287109375,
+      "learning_rate": 2.654436290860574e-05,
+      "loss": 1.3711,
+      "step": 11021
+    },
+    {
+      "epoch": 10.174936533579507,
+      "grad_norm": 3.611328125,
+      "learning_rate": 2.653769179452969e-05,
+      "loss": 1.1437,
+      "step": 11022
+    },
+    {
+      "epoch": 10.175859681513963,
+      "grad_norm": 2.970703125,
+      "learning_rate": 2.6531020680453637e-05,
+      "loss": 1.4846,
+      "step": 11023
+    },
+    {
+      "epoch": 10.176782829448419,
+      "grad_norm": 2.580078125,
+      "learning_rate": 2.6524349566377586e-05,
+      "loss": 0.8703,
+      "step": 11024
+    },
+    {
+      "epoch": 10.177705977382876,
+      "grad_norm": 3.236328125,
+      "learning_rate": 2.6517678452301535e-05,
+      "loss": 1.385,
+      "step": 11025
+    },
+    {
+      "epoch": 10.178629125317332,
+      "grad_norm": 2.962890625,
+      "learning_rate": 2.6511007338225484e-05,
+      "loss": 1.3328,
+      "step": 11026
+    },
+    {
+      "epoch": 10.179552273251788,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.6504336224149436e-05,
+      "loss": 1.2402,
+      "step": 11027
+    },
+    {
+      "epoch": 10.180475421186244,
+      "grad_norm": 3.326171875,
+      "learning_rate": 2.6497665110073385e-05,
+      "loss": 1.2765,
+      "step": 11028
+    },
+    {
+      "epoch": 10.181398569120702,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.6490993995997333e-05,
+      "loss": 1.5239,
+      "step": 11029
+    },
+    {
+      "epoch": 10.182321717055158,
+      "grad_norm": 3.486328125,
+      "learning_rate": 2.6484322881921282e-05,
+      "loss": 1.4246,
+      "step": 11030
+    },
+    {
+      "epoch": 10.183244864989614,
+      "grad_norm": 2.6796875,
+      "learning_rate": 2.647765176784523e-05,
+      "loss": 1.0001,
+      "step": 11031
+    },
+    {
+      "epoch": 10.18416801292407,
+      "grad_norm": 3.568359375,
+      "learning_rate": 2.647098065376918e-05,
+      "loss": 1.408,
+      "step": 11032
+    },
+    {
+      "epoch": 10.185091160858528,
+      "grad_norm": 3.69921875,
+      "learning_rate": 2.646430953969313e-05,
+      "loss": 1.2045,
+      "step": 11033
+    },
+    {
+      "epoch": 10.186014308792984,
+      "grad_norm": 3.1328125,
+      "learning_rate": 2.645763842561708e-05,
+      "loss": 1.2678,
+      "step": 11034
+    },
+    {
+      "epoch": 10.18693745672744,
+      "grad_norm": 2.927734375,
+      "learning_rate": 2.645096731154103e-05,
+      "loss": 1.1119,
+      "step": 11035
+    },
+    {
+      "epoch": 10.187860604661896,
+      "grad_norm": 3.080078125,
+      "learning_rate": 2.644429619746498e-05,
+      "loss": 1.249,
+      "step": 11036
+    },
+    {
+      "epoch": 10.188783752596354,
+      "grad_norm": 3.615234375,
+      "learning_rate": 2.6437625083388927e-05,
+      "loss": 0.9753,
+      "step": 11037
+    },
+    {
+      "epoch": 10.18970690053081,
+      "grad_norm": 3.314453125,
+      "learning_rate": 2.6430953969312876e-05,
+      "loss": 1.2714,
+      "step": 11038
+    },
+    {
+      "epoch": 10.190630048465266,
+      "grad_norm": 3.55078125,
+      "learning_rate": 2.6424282855236825e-05,
+      "loss": 1.1392,
+      "step": 11039
+    },
+    {
+      "epoch": 10.191553196399724,
+      "grad_norm": 3.439453125,
+      "learning_rate": 2.6417611741160774e-05,
+      "loss": 1.177,
+      "step": 11040
+    },
+    {
+      "epoch": 10.19247634433418,
+      "grad_norm": 3.310546875,
+      "learning_rate": 2.6410940627084723e-05,
+      "loss": 1.3367,
+      "step": 11041
+    },
+    {
+      "epoch": 10.193399492268636,
+      "grad_norm": 3.248046875,
+      "learning_rate": 2.640426951300867e-05,
+      "loss": 1.2585,
+      "step": 11042
+    },
+    {
+      "epoch": 10.194322640203092,
+      "grad_norm": 3.064453125,
+      "learning_rate": 2.639759839893262e-05,
+      "loss": 1.2999,
+      "step": 11043
+    },
+    {
+      "epoch": 10.19524578813755,
+      "grad_norm": 3.115234375,
+      "learning_rate": 2.639092728485657e-05,
+      "loss": 1.425,
+      "step": 11044
+    },
+    {
+      "epoch": 10.196168936072006,
+      "grad_norm": 3.6328125,
+      "learning_rate": 2.6384256170780518e-05,
+      "loss": 1.6479,
+      "step": 11045
+    },
+    {
+      "epoch": 10.197092084006462,
+      "grad_norm": 2.998046875,
+      "learning_rate": 2.6377585056704474e-05,
+      "loss": 1.3909,
+      "step": 11046
+    },
+    {
+      "epoch": 10.198015231940918,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.6370913942628422e-05,
+      "loss": 1.4058,
+      "step": 11047
+    },
+    {
+      "epoch": 10.198938379875376,
+      "grad_norm": 3.310546875,
+      "learning_rate": 2.636424282855237e-05,
+      "loss": 1.5244,
+      "step": 11048
+    },
+    {
+      "epoch": 10.199861527809832,
+      "grad_norm": 3.31640625,
+      "learning_rate": 2.635757171447632e-05,
+      "loss": 1.4419,
+      "step": 11049
+    },
+    {
+      "epoch": 10.200784675744288,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.635090060040027e-05,
+      "loss": 1.5676,
+      "step": 11050
+    },
+    {
+      "epoch": 10.201707823678744,
+      "grad_norm": 3.60546875,
+      "learning_rate": 2.6344229486324218e-05,
+      "loss": 1.2957,
+      "step": 11051
+    },
+    {
+      "epoch": 10.202630971613202,
+      "grad_norm": 3.388671875,
+      "learning_rate": 2.6337558372248166e-05,
+      "loss": 1.2661,
+      "step": 11052
+    },
+    {
+      "epoch": 10.203554119547658,
+      "grad_norm": 3.224609375,
+      "learning_rate": 2.6330887258172115e-05,
+      "loss": 0.9463,
+      "step": 11053
+    },
+    {
+      "epoch": 10.204477267482114,
+      "grad_norm": 3.3984375,
+      "learning_rate": 2.6324216144096064e-05,
+      "loss": 1.4998,
+      "step": 11054
+    },
+    {
+      "epoch": 10.20540041541657,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.6317545030020013e-05,
+      "loss": 1.395,
+      "step": 11055
+    },
+    {
+      "epoch": 10.206323563351027,
+      "grad_norm": 3.08203125,
+      "learning_rate": 2.6310873915943962e-05,
+      "loss": 1.3499,
+      "step": 11056
+    },
+    {
+      "epoch": 10.207246711285483,
+      "grad_norm": 2.966796875,
+      "learning_rate": 2.630420280186791e-05,
+      "loss": 1.2429,
+      "step": 11057
+    },
+    {
+      "epoch": 10.20816985921994,
+      "grad_norm": 3.474609375,
+      "learning_rate": 2.629753168779186e-05,
+      "loss": 1.3088,
+      "step": 11058
+    },
+    {
+      "epoch": 10.209093007154397,
+      "grad_norm": 3.353515625,
+      "learning_rate": 2.6290860573715815e-05,
+      "loss": 1.2849,
+      "step": 11059
+    },
+    {
+      "epoch": 10.210016155088853,
+      "grad_norm": 2.556640625,
+      "learning_rate": 2.6284189459639764e-05,
+      "loss": 1.0679,
+      "step": 11060
+    },
+    {
+      "epoch": 10.21093930302331,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.6277518345563713e-05,
+      "loss": 1.3098,
+      "step": 11061
+    },
+    {
+      "epoch": 10.211862450957765,
+      "grad_norm": 3.3203125,
+      "learning_rate": 2.627084723148766e-05,
+      "loss": 1.2408,
+      "step": 11062
+    },
+    {
+      "epoch": 10.212785598892223,
+      "grad_norm": 3.212890625,
+      "learning_rate": 2.626417611741161e-05,
+      "loss": 1.2922,
+      "step": 11063
+    },
+    {
+      "epoch": 10.21370874682668,
+      "grad_norm": 3.43359375,
+      "learning_rate": 2.625750500333556e-05,
+      "loss": 1.3359,
+      "step": 11064
+    },
+    {
+      "epoch": 10.214631894761135,
+      "grad_norm": 3.591796875,
+      "learning_rate": 2.6250833889259508e-05,
+      "loss": 1.2622,
+      "step": 11065
+    },
+    {
+      "epoch": 10.215555042695591,
+      "grad_norm": 3.2421875,
+      "learning_rate": 2.6244162775183457e-05,
+      "loss": 1.3326,
+      "step": 11066
+    },
+    {
+      "epoch": 10.216478190630049,
+      "grad_norm": 3.55078125,
+      "learning_rate": 2.6237491661107406e-05,
+      "loss": 1.0919,
+      "step": 11067
+    },
+    {
+      "epoch": 10.217401338564505,
+      "grad_norm": 3.36328125,
+      "learning_rate": 2.6230820547031354e-05,
+      "loss": 1.2976,
+      "step": 11068
+    },
+    {
+      "epoch": 10.218324486498961,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.6224149432955303e-05,
+      "loss": 1.2665,
+      "step": 11069
+    },
+    {
+      "epoch": 10.219247634433417,
+      "grad_norm": 3.083984375,
+      "learning_rate": 2.6217478318879252e-05,
+      "loss": 1.4395,
+      "step": 11070
+    },
+    {
+      "epoch": 10.220170782367875,
+      "grad_norm": 3.138671875,
+      "learning_rate": 2.6210807204803208e-05,
+      "loss": 1.3009,
+      "step": 11071
+    },
+    {
+      "epoch": 10.221093930302331,
+      "grad_norm": 2.982421875,
+      "learning_rate": 2.6204136090727156e-05,
+      "loss": 1.3833,
+      "step": 11072
+    },
+    {
+      "epoch": 10.222017078236787,
+      "grad_norm": 3.5546875,
+      "learning_rate": 2.6197464976651105e-05,
+      "loss": 1.4061,
+      "step": 11073
+    },
+    {
+      "epoch": 10.222940226171245,
+      "grad_norm": 3.505859375,
+      "learning_rate": 2.6190793862575054e-05,
+      "loss": 1.2332,
+      "step": 11074
+    },
+    {
+      "epoch": 10.2238633741057,
+      "grad_norm": 3.12890625,
+      "learning_rate": 2.6184122748499003e-05,
+      "loss": 1.3237,
+      "step": 11075
+    },
+    {
+      "epoch": 10.224786522040157,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.617745163442295e-05,
+      "loss": 1.395,
+      "step": 11076
+    },
+    {
+      "epoch": 10.225709669974613,
+      "grad_norm": 3.197265625,
+      "learning_rate": 2.61707805203469e-05,
+      "loss": 1.2644,
+      "step": 11077
+    },
+    {
+      "epoch": 10.22663281790907,
+      "grad_norm": 2.814453125,
+      "learning_rate": 2.616410940627085e-05,
+      "loss": 1.0914,
+      "step": 11078
+    },
+    {
+      "epoch": 10.227555965843527,
+      "grad_norm": 3.564453125,
+      "learning_rate": 2.6157438292194798e-05,
+      "loss": 0.844,
+      "step": 11079
+    },
+    {
+      "epoch": 10.228479113777983,
+      "grad_norm": 3.7265625,
+      "learning_rate": 2.6150767178118747e-05,
+      "loss": 1.1167,
+      "step": 11080
+    },
+    {
+      "epoch": 10.229402261712439,
+      "grad_norm": 3.244140625,
+      "learning_rate": 2.6144096064042696e-05,
+      "loss": 1.3547,
+      "step": 11081
+    },
+    {
+      "epoch": 10.230325409646897,
+      "grad_norm": 3.2734375,
+      "learning_rate": 2.6137424949966645e-05,
+      "loss": 1.4136,
+      "step": 11082
+    },
+    {
+      "epoch": 10.231248557581353,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.6130753835890593e-05,
+      "loss": 1.3018,
+      "step": 11083
+    },
+    {
+      "epoch": 10.232171705515809,
+      "grad_norm": 3.458984375,
+      "learning_rate": 2.6124082721814546e-05,
+      "loss": 1.386,
+      "step": 11084
+    },
+    {
+      "epoch": 10.233094853450265,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.6117411607738494e-05,
+      "loss": 1.4121,
+      "step": 11085
+    },
+    {
+      "epoch": 10.234018001384722,
+      "grad_norm": 2.998046875,
+      "learning_rate": 2.6110740493662443e-05,
+      "loss": 1.1053,
+      "step": 11086
+    },
+    {
+      "epoch": 10.234941149319178,
+      "grad_norm": 3.279296875,
+      "learning_rate": 2.6104069379586392e-05,
+      "loss": 1.321,
+      "step": 11087
+    },
+    {
+      "epoch": 10.235864297253634,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.609739826551034e-05,
+      "loss": 1.275,
+      "step": 11088
+    },
+    {
+      "epoch": 10.236787445188092,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.609072715143429e-05,
+      "loss": 1.3696,
+      "step": 11089
+    },
+    {
+      "epoch": 10.237710593122548,
+      "grad_norm": 3.89453125,
+      "learning_rate": 2.608405603735824e-05,
+      "loss": 1.3755,
+      "step": 11090
+    },
+    {
+      "epoch": 10.238633741057004,
+      "grad_norm": 3.0546875,
+      "learning_rate": 2.6077384923282187e-05,
+      "loss": 1.2437,
+      "step": 11091
+    },
+    {
+      "epoch": 10.23955688899146,
+      "grad_norm": 3.673828125,
+      "learning_rate": 2.6070713809206136e-05,
+      "loss": 1.2844,
+      "step": 11092
+    },
+    {
+      "epoch": 10.240480036925918,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.6064042695130085e-05,
+      "loss": 1.3022,
+      "step": 11093
+    },
+    {
+      "epoch": 10.241403184860374,
+      "grad_norm": 3.201171875,
+      "learning_rate": 2.6057371581054034e-05,
+      "loss": 1.5859,
+      "step": 11094
+    },
+    {
+      "epoch": 10.24232633279483,
+      "grad_norm": 3.439453125,
+      "learning_rate": 2.6050700466977983e-05,
+      "loss": 1.2283,
+      "step": 11095
+    },
+    {
+      "epoch": 10.243249480729286,
+      "grad_norm": 3.2265625,
+      "learning_rate": 2.6044029352901938e-05,
+      "loss": 1.2458,
+      "step": 11096
+    },
+    {
+      "epoch": 10.244172628663744,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.6037358238825887e-05,
+      "loss": 1.2175,
+      "step": 11097
+    },
+    {
+      "epoch": 10.2450957765982,
+      "grad_norm": 3.908203125,
+      "learning_rate": 2.6030687124749836e-05,
+      "loss": 1.5845,
+      "step": 11098
+    },
+    {
+      "epoch": 10.246018924532656,
+      "grad_norm": 3.05859375,
+      "learning_rate": 2.6024016010673785e-05,
+      "loss": 1.2336,
+      "step": 11099
+    },
+    {
+      "epoch": 10.246942072467112,
+      "grad_norm": 3.302734375,
+      "learning_rate": 2.6017344896597734e-05,
+      "loss": 1.2135,
+      "step": 11100
+    },
+    {
+      "epoch": 10.24786522040157,
+      "grad_norm": 3.193359375,
+      "learning_rate": 2.6010673782521682e-05,
+      "loss": 1.1918,
+      "step": 11101
+    },
+    {
+      "epoch": 10.248788368336026,
+      "grad_norm": 2.75,
+      "learning_rate": 2.600400266844563e-05,
+      "loss": 1.4236,
+      "step": 11102
+    },
+    {
+      "epoch": 10.249711516270482,
+      "grad_norm": 3.88671875,
+      "learning_rate": 2.599733155436958e-05,
+      "loss": 1.231,
+      "step": 11103
+    },
+    {
+      "epoch": 10.250634664204938,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.599066044029353e-05,
+      "loss": 1.5681,
+      "step": 11104
+    },
+    {
+      "epoch": 10.251557812139396,
+      "grad_norm": 3.181640625,
+      "learning_rate": 2.5983989326217478e-05,
+      "loss": 1.3833,
+      "step": 11105
+    },
+    {
+      "epoch": 10.252480960073852,
+      "grad_norm": 12.75,
+      "learning_rate": 2.5977318212141426e-05,
+      "loss": 1.4242,
+      "step": 11106
+    },
+    {
+      "epoch": 10.253404108008308,
+      "grad_norm": 3.3515625,
+      "learning_rate": 2.5970647098065375e-05,
+      "loss": 1.365,
+      "step": 11107
+    },
+    {
+      "epoch": 10.254327255942766,
+      "grad_norm": 3.44140625,
+      "learning_rate": 2.5963975983989324e-05,
+      "loss": 1.2545,
+      "step": 11108
+    },
+    {
+      "epoch": 10.255250403877222,
+      "grad_norm": 3.27734375,
+      "learning_rate": 2.595730486991328e-05,
+      "loss": 1.2869,
+      "step": 11109
+    },
+    {
+      "epoch": 10.256173551811678,
+      "grad_norm": 2.8828125,
+      "learning_rate": 2.595063375583723e-05,
+      "loss": 0.8874,
+      "step": 11110
+    },
+    {
+      "epoch": 10.257096699746134,
+      "grad_norm": 3.52734375,
+      "learning_rate": 2.5943962641761177e-05,
+      "loss": 1.4287,
+      "step": 11111
+    },
+    {
+      "epoch": 10.258019847680591,
+      "grad_norm": 3.125,
+      "learning_rate": 2.5937291527685126e-05,
+      "loss": 1.0909,
+      "step": 11112
+    },
+    {
+      "epoch": 10.258942995615048,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.5930620413609075e-05,
+      "loss": 1.1224,
+      "step": 11113
+    },
+    {
+      "epoch": 10.259866143549504,
+      "grad_norm": 3.16796875,
+      "learning_rate": 2.5923949299533024e-05,
+      "loss": 1.3445,
+      "step": 11114
+    },
+    {
+      "epoch": 10.26078929148396,
+      "grad_norm": 3.732421875,
+      "learning_rate": 2.5917278185456973e-05,
+      "loss": 1.2764,
+      "step": 11115
+    },
+    {
+      "epoch": 10.261712439418417,
+      "grad_norm": 3.673828125,
+      "learning_rate": 2.591060707138092e-05,
+      "loss": 1.1123,
+      "step": 11116
+    },
+    {
+      "epoch": 10.262635587352873,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.590393595730487e-05,
+      "loss": 1.2058,
+      "step": 11117
+    },
+    {
+      "epoch": 10.26355873528733,
+      "grad_norm": 3.244140625,
+      "learning_rate": 2.589726484322882e-05,
+      "loss": 1.2222,
+      "step": 11118
+    },
+    {
+      "epoch": 10.264481883221785,
+      "grad_norm": 3.248046875,
+      "learning_rate": 2.5890593729152768e-05,
+      "loss": 1.361,
+      "step": 11119
+    },
+    {
+      "epoch": 10.265405031156243,
+      "grad_norm": 2.955078125,
+      "learning_rate": 2.5883922615076717e-05,
+      "loss": 1.2388,
+      "step": 11120
+    },
+    {
+      "epoch": 10.2663281790907,
+      "grad_norm": 3.28515625,
+      "learning_rate": 2.5877251501000672e-05,
+      "loss": 1.0913,
+      "step": 11121
+    },
+    {
+      "epoch": 10.267251327025155,
+      "grad_norm": 3.0078125,
+      "learning_rate": 2.587058038692462e-05,
+      "loss": 1.2463,
+      "step": 11122
+    },
+    {
+      "epoch": 10.268174474959613,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.586390927284857e-05,
+      "loss": 1.0969,
+      "step": 11123
+    },
+    {
+      "epoch": 10.26909762289407,
+      "grad_norm": 3.654296875,
+      "learning_rate": 2.585723815877252e-05,
+      "loss": 1.1973,
+      "step": 11124
+    },
+    {
+      "epoch": 10.270020770828525,
+      "grad_norm": 3.599609375,
+      "learning_rate": 2.5850567044696468e-05,
+      "loss": 1.1396,
+      "step": 11125
+    },
+    {
+      "epoch": 10.270943918762981,
+      "grad_norm": 2.787109375,
+      "learning_rate": 2.5843895930620416e-05,
+      "loss": 1.4531,
+      "step": 11126
+    },
+    {
+      "epoch": 10.271867066697439,
+      "grad_norm": 3.478515625,
+      "learning_rate": 2.5837224816544365e-05,
+      "loss": 1.1971,
+      "step": 11127
+    },
+    {
+      "epoch": 10.272790214631895,
+      "grad_norm": 3.03515625,
+      "learning_rate": 2.5830553702468314e-05,
+      "loss": 1.1262,
+      "step": 11128
+    },
+    {
+      "epoch": 10.273713362566351,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.5823882588392263e-05,
+      "loss": 1.2725,
+      "step": 11129
+    },
+    {
+      "epoch": 10.274636510500807,
+      "grad_norm": 3.240234375,
+      "learning_rate": 2.581721147431621e-05,
+      "loss": 1.1958,
+      "step": 11130
+    },
+    {
+      "epoch": 10.275559658435265,
+      "grad_norm": 3.08984375,
+      "learning_rate": 2.581054036024016e-05,
+      "loss": 1.2603,
+      "step": 11131
+    },
+    {
+      "epoch": 10.276482806369721,
+      "grad_norm": 3.068359375,
+      "learning_rate": 2.580386924616411e-05,
+      "loss": 1.1895,
+      "step": 11132
+    },
+    {
+      "epoch": 10.277405954304177,
+      "grad_norm": 3.189453125,
+      "learning_rate": 2.5797198132088058e-05,
+      "loss": 1.2892,
+      "step": 11133
+    },
+    {
+      "epoch": 10.278329102238633,
+      "grad_norm": 3.38671875,
+      "learning_rate": 2.579052701801201e-05,
+      "loss": 1.2703,
+      "step": 11134
+    },
+    {
+      "epoch": 10.27925225017309,
+      "grad_norm": 3.599609375,
+      "learning_rate": 2.578385590393596e-05,
+      "loss": 1.4211,
+      "step": 11135
+    },
+    {
+      "epoch": 10.280175398107547,
+      "grad_norm": 3.541015625,
+      "learning_rate": 2.5777184789859908e-05,
+      "loss": 1.03,
+      "step": 11136
+    },
+    {
+      "epoch": 10.281098546042003,
+      "grad_norm": 2.7421875,
+      "learning_rate": 2.5770513675783857e-05,
+      "loss": 1.4536,
+      "step": 11137
+    },
+    {
+      "epoch": 10.28202169397646,
+      "grad_norm": 2.9921875,
+      "learning_rate": 2.5763842561707806e-05,
+      "loss": 0.8906,
+      "step": 11138
+    },
+    {
+      "epoch": 10.282944841910917,
+      "grad_norm": 3.55859375,
+      "learning_rate": 2.5757171447631754e-05,
+      "loss": 1.397,
+      "step": 11139
+    },
+    {
+      "epoch": 10.283867989845373,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.5750500333555707e-05,
+      "loss": 1.3533,
+      "step": 11140
+    },
+    {
+      "epoch": 10.284791137779829,
+      "grad_norm": 3.599609375,
+      "learning_rate": 2.5743829219479655e-05,
+      "loss": 1.2295,
+      "step": 11141
+    },
+    {
+      "epoch": 10.285714285714286,
+      "grad_norm": 2.966796875,
+      "learning_rate": 2.5737158105403604e-05,
+      "loss": 1.2474,
+      "step": 11142
+    },
+    {
+      "epoch": 10.286637433648743,
+      "grad_norm": 3.189453125,
+      "learning_rate": 2.5730486991327553e-05,
+      "loss": 1.4048,
+      "step": 11143
+    },
+    {
+      "epoch": 10.287560581583199,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.5723815877251502e-05,
+      "loss": 1.0936,
+      "step": 11144
+    },
+    {
+      "epoch": 10.288483729517655,
+      "grad_norm": 3.0859375,
+      "learning_rate": 2.571714476317545e-05,
+      "loss": 1.5095,
+      "step": 11145
+    },
+    {
+      "epoch": 10.289406877452112,
+      "grad_norm": 3.154296875,
+      "learning_rate": 2.5710473649099403e-05,
+      "loss": 1.1917,
+      "step": 11146
+    },
+    {
+      "epoch": 10.290330025386568,
+      "grad_norm": 3.83203125,
+      "learning_rate": 2.5703802535023352e-05,
+      "loss": 1.3057,
+      "step": 11147
+    },
+    {
+      "epoch": 10.291253173321024,
+      "grad_norm": 3.232421875,
+      "learning_rate": 2.56971314209473e-05,
+      "loss": 1.1227,
+      "step": 11148
+    },
+    {
+      "epoch": 10.29217632125548,
+      "grad_norm": 3.458984375,
+      "learning_rate": 2.569046030687125e-05,
+      "loss": 1.2186,
+      "step": 11149
+    },
+    {
+      "epoch": 10.293099469189938,
+      "grad_norm": 3.189453125,
+      "learning_rate": 2.5683789192795198e-05,
+      "loss": 0.9989,
+      "step": 11150
+    },
+    {
+      "epoch": 10.294022617124394,
+      "grad_norm": 2.25,
+      "learning_rate": 2.5677118078719147e-05,
+      "loss": 1.1084,
+      "step": 11151
+    },
+    {
+      "epoch": 10.29494576505885,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.5670446964643096e-05,
+      "loss": 1.501,
+      "step": 11152
+    },
+    {
+      "epoch": 10.295868912993306,
+      "grad_norm": 3.419921875,
+      "learning_rate": 2.5663775850567045e-05,
+      "loss": 1.1975,
+      "step": 11153
+    },
+    {
+      "epoch": 10.296792060927764,
+      "grad_norm": 3.27734375,
+      "learning_rate": 2.5657104736490993e-05,
+      "loss": 1.2749,
+      "step": 11154
+    },
+    {
+      "epoch": 10.29771520886222,
+      "grad_norm": 3.19921875,
+      "learning_rate": 2.5650433622414942e-05,
+      "loss": 1.348,
+      "step": 11155
+    },
+    {
+      "epoch": 10.298638356796676,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.564376250833889e-05,
+      "loss": 1.3806,
+      "step": 11156
+    },
+    {
+      "epoch": 10.299561504731134,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.563709139426284e-05,
+      "loss": 1.3086,
+      "step": 11157
+    },
+    {
+      "epoch": 10.30048465266559,
+      "grad_norm": 3.02734375,
+      "learning_rate": 2.563042028018679e-05,
+      "loss": 1.2191,
+      "step": 11158
+    },
+    {
+      "epoch": 10.301407800600046,
+      "grad_norm": 3.240234375,
+      "learning_rate": 2.5623749166110744e-05,
+      "loss": 1.2183,
+      "step": 11159
+    },
+    {
+      "epoch": 10.302330948534502,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.5617078052034693e-05,
+      "loss": 1.3042,
+      "step": 11160
+    },
+    {
+      "epoch": 10.30325409646896,
+      "grad_norm": 3.041015625,
+      "learning_rate": 2.5610406937958642e-05,
+      "loss": 1.2551,
+      "step": 11161
+    },
+    {
+      "epoch": 10.304177244403416,
+      "grad_norm": 3.095703125,
+      "learning_rate": 2.560373582388259e-05,
+      "loss": 1.2329,
+      "step": 11162
+    },
+    {
+      "epoch": 10.305100392337872,
+      "grad_norm": 2.83203125,
+      "learning_rate": 2.559706470980654e-05,
+      "loss": 1.4124,
+      "step": 11163
+    },
+    {
+      "epoch": 10.306023540272328,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.559039359573049e-05,
+      "loss": 1.2356,
+      "step": 11164
+    },
+    {
+      "epoch": 10.306946688206786,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.5583722481654437e-05,
+      "loss": 1.2791,
+      "step": 11165
+    },
+    {
+      "epoch": 10.307869836141242,
+      "grad_norm": 3.037109375,
+      "learning_rate": 2.5577051367578386e-05,
+      "loss": 1.2571,
+      "step": 11166
+    },
+    {
+      "epoch": 10.308792984075698,
+      "grad_norm": 3.00390625,
+      "learning_rate": 2.5570380253502335e-05,
+      "loss": 1.3921,
+      "step": 11167
+    },
+    {
+      "epoch": 10.309716132010154,
+      "grad_norm": 3.310546875,
+      "learning_rate": 2.5563709139426284e-05,
+      "loss": 1.6399,
+      "step": 11168
+    },
+    {
+      "epoch": 10.310639279944612,
+      "grad_norm": 3.15234375,
+      "learning_rate": 2.5557038025350232e-05,
+      "loss": 1.2628,
+      "step": 11169
+    },
+    {
+      "epoch": 10.311562427879068,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.555036691127418e-05,
+      "loss": 1.2483,
+      "step": 11170
+    },
+    {
+      "epoch": 10.312485575813524,
+      "grad_norm": 3.189453125,
+      "learning_rate": 2.554369579719813e-05,
+      "loss": 1.1601,
+      "step": 11171
+    },
+    {
+      "epoch": 10.313408723747981,
+      "grad_norm": 3.26953125,
+      "learning_rate": 2.5537024683122086e-05,
+      "loss": 1.4741,
+      "step": 11172
+    },
+    {
+      "epoch": 10.314331871682437,
+      "grad_norm": 3.3203125,
+      "learning_rate": 2.5530353569046035e-05,
+      "loss": 1.3464,
+      "step": 11173
+    },
+    {
+      "epoch": 10.315255019616894,
+      "grad_norm": 3.02734375,
+      "learning_rate": 2.5523682454969983e-05,
+      "loss": 1.2886,
+      "step": 11174
+    },
+    {
+      "epoch": 10.31617816755135,
+      "grad_norm": 3.373046875,
+      "learning_rate": 2.5517011340893932e-05,
+      "loss": 1.3967,
+      "step": 11175
+    },
+    {
+      "epoch": 10.317101315485807,
+      "grad_norm": 3.57421875,
+      "learning_rate": 2.551034022681788e-05,
+      "loss": 1.3003,
+      "step": 11176
+    },
+    {
+      "epoch": 10.318024463420263,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.550366911274183e-05,
+      "loss": 1.3657,
+      "step": 11177
+    },
+    {
+      "epoch": 10.31894761135472,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.549699799866578e-05,
+      "loss": 1.3439,
+      "step": 11178
+    },
+    {
+      "epoch": 10.319870759289175,
+      "grad_norm": 3.072265625,
+      "learning_rate": 2.5490326884589727e-05,
+      "loss": 1.145,
+      "step": 11179
+    },
+    {
+      "epoch": 10.320793907223633,
+      "grad_norm": 3.3671875,
+      "learning_rate": 2.5483655770513676e-05,
+      "loss": 1.354,
+      "step": 11180
+    },
+    {
+      "epoch": 10.32171705515809,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.5476984656437625e-05,
+      "loss": 1.3093,
+      "step": 11181
+    },
+    {
+      "epoch": 10.322640203092545,
+      "grad_norm": 3.072265625,
+      "learning_rate": 2.5470313542361574e-05,
+      "loss": 1.3823,
+      "step": 11182
+    },
+    {
+      "epoch": 10.323563351027001,
+      "grad_norm": 3.6171875,
+      "learning_rate": 2.5463642428285523e-05,
+      "loss": 1.3113,
+      "step": 11183
+    },
+    {
+      "epoch": 10.324486498961459,
+      "grad_norm": 3.388671875,
+      "learning_rate": 2.545697131420948e-05,
+      "loss": 1.5215,
+      "step": 11184
+    },
+    {
+      "epoch": 10.325409646895915,
+      "grad_norm": 3.5078125,
+      "learning_rate": 2.5450300200133427e-05,
+      "loss": 1.3296,
+      "step": 11185
+    },
+    {
+      "epoch": 10.326332794830371,
+      "grad_norm": 3.2890625,
+      "learning_rate": 2.5443629086057376e-05,
+      "loss": 1.3079,
+      "step": 11186
+    },
+    {
+      "epoch": 10.327255942764829,
+      "grad_norm": 3.4921875,
+      "learning_rate": 2.5436957971981325e-05,
+      "loss": 1.3091,
+      "step": 11187
+    },
+    {
+      "epoch": 10.328179090699285,
+      "grad_norm": 2.95703125,
+      "learning_rate": 2.5430286857905274e-05,
+      "loss": 1.4485,
+      "step": 11188
+    },
+    {
+      "epoch": 10.329102238633741,
+      "grad_norm": 3.943359375,
+      "learning_rate": 2.5423615743829222e-05,
+      "loss": 1.1064,
+      "step": 11189
+    },
+    {
+      "epoch": 10.330025386568197,
+      "grad_norm": 2.849609375,
+      "learning_rate": 2.541694462975317e-05,
+      "loss": 1.5173,
+      "step": 11190
+    },
+    {
+      "epoch": 10.330948534502655,
+      "grad_norm": 3.287109375,
+      "learning_rate": 2.541027351567712e-05,
+      "loss": 1.4292,
+      "step": 11191
+    },
+    {
+      "epoch": 10.33187168243711,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.540360240160107e-05,
+      "loss": 1.3464,
+      "step": 11192
+    },
+    {
+      "epoch": 10.332794830371567,
+      "grad_norm": 3.392578125,
+      "learning_rate": 2.5396931287525018e-05,
+      "loss": 1.4561,
+      "step": 11193
+    },
+    {
+      "epoch": 10.333717978306023,
+      "grad_norm": 3.12109375,
+      "learning_rate": 2.5390260173448967e-05,
+      "loss": 1.3069,
+      "step": 11194
+    },
+    {
+      "epoch": 10.33464112624048,
+      "grad_norm": 3.6953125,
+      "learning_rate": 2.5383589059372915e-05,
+      "loss": 1.3677,
+      "step": 11195
+    },
+    {
+      "epoch": 10.335564274174937,
+      "grad_norm": 3.611328125,
+      "learning_rate": 2.5376917945296864e-05,
+      "loss": 1.2356,
+      "step": 11196
+    },
+    {
+      "epoch": 10.336487422109393,
+      "grad_norm": 3.162109375,
+      "learning_rate": 2.5370246831220816e-05,
+      "loss": 1.1798,
+      "step": 11197
+    },
+    {
+      "epoch": 10.337410570043849,
+      "grad_norm": 3.388671875,
+      "learning_rate": 2.5363575717144765e-05,
+      "loss": 1.0211,
+      "step": 11198
+    },
+    {
+      "epoch": 10.338333717978307,
+      "grad_norm": 3.111328125,
+      "learning_rate": 2.5356904603068714e-05,
+      "loss": 1.3208,
+      "step": 11199
+    },
+    {
+      "epoch": 10.339256865912763,
+      "grad_norm": 3.091796875,
+      "learning_rate": 2.5350233488992663e-05,
+      "loss": 1.4819,
+      "step": 11200
+    },
+    {
+      "epoch": 10.340180013847219,
+      "grad_norm": 3.236328125,
+      "learning_rate": 2.534356237491661e-05,
+      "loss": 1.3918,
+      "step": 11201
+    },
+    {
+      "epoch": 10.341103161781675,
+      "grad_norm": 3.205078125,
+      "learning_rate": 2.533689126084056e-05,
+      "loss": 1.1722,
+      "step": 11202
+    },
+    {
+      "epoch": 10.342026309716132,
+      "grad_norm": 3.2265625,
+      "learning_rate": 2.533022014676451e-05,
+      "loss": 1.0527,
+      "step": 11203
+    },
+    {
+      "epoch": 10.342949457650588,
+      "grad_norm": 3.412109375,
+      "learning_rate": 2.5323549032688458e-05,
+      "loss": 1.3679,
+      "step": 11204
+    },
+    {
+      "epoch": 10.343872605585045,
+      "grad_norm": 3.68359375,
+      "learning_rate": 2.5316877918612407e-05,
+      "loss": 1.2777,
+      "step": 11205
+    },
+    {
+      "epoch": 10.344795753519502,
+      "grad_norm": 3.31640625,
+      "learning_rate": 2.5310206804536356e-05,
+      "loss": 1.1906,
+      "step": 11206
+    },
+    {
+      "epoch": 10.345718901453958,
+      "grad_norm": 3.134765625,
+      "learning_rate": 2.5303535690460305e-05,
+      "loss": 1.3688,
+      "step": 11207
+    },
+    {
+      "epoch": 10.346642049388414,
+      "grad_norm": 3.3515625,
+      "learning_rate": 2.5296864576384253e-05,
+      "loss": 1.1913,
+      "step": 11208
+    },
+    {
+      "epoch": 10.34756519732287,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.529019346230821e-05,
+      "loss": 1.2094,
+      "step": 11209
+    },
+    {
+      "epoch": 10.348488345257328,
+      "grad_norm": 3.595703125,
+      "learning_rate": 2.5283522348232158e-05,
+      "loss": 1.2373,
+      "step": 11210
+    },
+    {
+      "epoch": 10.349411493191784,
+      "grad_norm": 3.64453125,
+      "learning_rate": 2.5276851234156107e-05,
+      "loss": 1.2432,
+      "step": 11211
+    },
+    {
+      "epoch": 10.35033464112624,
+      "grad_norm": 3.095703125,
+      "learning_rate": 2.5270180120080055e-05,
+      "loss": 1.3235,
+      "step": 11212
+    },
+    {
+      "epoch": 10.351257789060696,
+      "grad_norm": 3.7265625,
+      "learning_rate": 2.5263509006004004e-05,
+      "loss": 1.2817,
+      "step": 11213
+    },
+    {
+      "epoch": 10.352180936995154,
+      "grad_norm": 3.76953125,
+      "learning_rate": 2.5256837891927953e-05,
+      "loss": 1.4631,
+      "step": 11214
+    },
+    {
+      "epoch": 10.35310408492961,
+      "grad_norm": 3.373046875,
+      "learning_rate": 2.5250166777851902e-05,
+      "loss": 1.183,
+      "step": 11215
+    },
+    {
+      "epoch": 10.354027232864066,
+      "grad_norm": 3.05859375,
+      "learning_rate": 2.524349566377585e-05,
+      "loss": 1.3191,
+      "step": 11216
+    },
+    {
+      "epoch": 10.354950380798522,
+      "grad_norm": 2.970703125,
+      "learning_rate": 2.52368245496998e-05,
+      "loss": 1.4656,
+      "step": 11217
+    },
+    {
+      "epoch": 10.35587352873298,
+      "grad_norm": 3.236328125,
+      "learning_rate": 2.523015343562375e-05,
+      "loss": 1.1306,
+      "step": 11218
+    },
+    {
+      "epoch": 10.356796676667436,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.5223482321547697e-05,
+      "loss": 1.2844,
+      "step": 11219
+    },
+    {
+      "epoch": 10.357719824601892,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.5216811207471646e-05,
+      "loss": 1.386,
+      "step": 11220
+    },
+    {
+      "epoch": 10.35864297253635,
+      "grad_norm": 3.10546875,
+      "learning_rate": 2.5210140093395595e-05,
+      "loss": 1.22,
+      "step": 11221
+    },
+    {
+      "epoch": 10.359566120470806,
+      "grad_norm": 3.478515625,
+      "learning_rate": 2.520346897931955e-05,
+      "loss": 1.2249,
+      "step": 11222
+    },
+    {
+      "epoch": 10.360489268405262,
+      "grad_norm": 2.931640625,
+      "learning_rate": 2.51967978652435e-05,
+      "loss": 1.2881,
+      "step": 11223
+    },
+    {
+      "epoch": 10.361412416339718,
+      "grad_norm": 3.3203125,
+      "learning_rate": 2.5190126751167448e-05,
+      "loss": 1.0548,
+      "step": 11224
+    },
+    {
+      "epoch": 10.362335564274176,
+      "grad_norm": 3.748046875,
+      "learning_rate": 2.5183455637091397e-05,
+      "loss": 1.199,
+      "step": 11225
+    },
+    {
+      "epoch": 10.363258712208632,
+      "grad_norm": 3.5078125,
+      "learning_rate": 2.5176784523015346e-05,
+      "loss": 1.314,
+      "step": 11226
+    },
+    {
+      "epoch": 10.364181860143088,
+      "grad_norm": 2.880859375,
+      "learning_rate": 2.5170113408939294e-05,
+      "loss": 1.3669,
+      "step": 11227
+    },
+    {
+      "epoch": 10.365105008077544,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.5163442294863243e-05,
+      "loss": 1.1293,
+      "step": 11228
+    },
+    {
+      "epoch": 10.366028156012002,
+      "grad_norm": 3.341796875,
+      "learning_rate": 2.5156771180787192e-05,
+      "loss": 1.4934,
+      "step": 11229
+    },
+    {
+      "epoch": 10.366951303946458,
+      "grad_norm": 3.03515625,
+      "learning_rate": 2.515010006671114e-05,
+      "loss": 1.1571,
+      "step": 11230
+    },
+    {
+      "epoch": 10.367874451880914,
+      "grad_norm": 3.744140625,
+      "learning_rate": 2.514342895263509e-05,
+      "loss": 1.4834,
+      "step": 11231
+    },
+    {
+      "epoch": 10.36879759981537,
+      "grad_norm": 2.9609375,
+      "learning_rate": 2.513675783855904e-05,
+      "loss": 1.3406,
+      "step": 11232
+    },
+    {
+      "epoch": 10.369720747749827,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.5130086724482987e-05,
+      "loss": 1.2218,
+      "step": 11233
+    },
+    {
+      "epoch": 10.370643895684283,
+      "grad_norm": 2.595703125,
+      "learning_rate": 2.5123415610406943e-05,
+      "loss": 1.4517,
+      "step": 11234
+    },
+    {
+      "epoch": 10.37156704361874,
+      "grad_norm": 3.912109375,
+      "learning_rate": 2.5116744496330892e-05,
+      "loss": 1.4797,
+      "step": 11235
+    },
+    {
+      "epoch": 10.372490191553197,
+      "grad_norm": 3.654296875,
+      "learning_rate": 2.511007338225484e-05,
+      "loss": 1.4553,
+      "step": 11236
+    },
+    {
+      "epoch": 10.373413339487653,
+      "grad_norm": 3.857421875,
+      "learning_rate": 2.510340226817879e-05,
+      "loss": 1.2334,
+      "step": 11237
+    },
+    {
+      "epoch": 10.37433648742211,
+      "grad_norm": 3.00390625,
+      "learning_rate": 2.5096731154102738e-05,
+      "loss": 1.4597,
+      "step": 11238
+    },
+    {
+      "epoch": 10.375259635356565,
+      "grad_norm": 3.416015625,
+      "learning_rate": 2.5090060040026687e-05,
+      "loss": 1.3284,
+      "step": 11239
+    },
+    {
+      "epoch": 10.376182783291023,
+      "grad_norm": 3.2890625,
+      "learning_rate": 2.5083388925950636e-05,
+      "loss": 1.2219,
+      "step": 11240
+    },
+    {
+      "epoch": 10.37710593122548,
+      "grad_norm": 3.88671875,
+      "learning_rate": 2.5076717811874585e-05,
+      "loss": 1.2954,
+      "step": 11241
+    },
+    {
+      "epoch": 10.378029079159935,
+      "grad_norm": 3.326171875,
+      "learning_rate": 2.5070046697798534e-05,
+      "loss": 1.1353,
+      "step": 11242
+    },
+    {
+      "epoch": 10.378952227094391,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.5063375583722482e-05,
+      "loss": 1.1244,
+      "step": 11243
+    },
+    {
+      "epoch": 10.379875375028849,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.505670446964643e-05,
+      "loss": 1.3169,
+      "step": 11244
+    },
+    {
+      "epoch": 10.380798522963305,
+      "grad_norm": 3.18359375,
+      "learning_rate": 2.505003335557038e-05,
+      "loss": 1.3674,
+      "step": 11245
+    },
+    {
+      "epoch": 10.381721670897761,
+      "grad_norm": 3.01171875,
+      "learning_rate": 2.504336224149433e-05,
+      "loss": 1.1549,
+      "step": 11246
+    },
+    {
+      "epoch": 10.382644818832217,
+      "grad_norm": 3.353515625,
+      "learning_rate": 2.503669112741828e-05,
+      "loss": 1.1873,
+      "step": 11247
+    },
+    {
+      "epoch": 10.383567966766675,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.503002001334223e-05,
+      "loss": 1.2371,
+      "step": 11248
+    },
+    {
+      "epoch": 10.384491114701131,
+      "grad_norm": 3.2578125,
+      "learning_rate": 2.502334889926618e-05,
+      "loss": 1.3176,
+      "step": 11249
+    },
+    {
+      "epoch": 10.385414262635587,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.5016677785190127e-05,
+      "loss": 0.9227,
+      "step": 11250
+    },
+    {
+      "epoch": 10.386337410570043,
+      "grad_norm": 3.435546875,
+      "learning_rate": 2.5010006671114076e-05,
+      "loss": 1.1494,
+      "step": 11251
+    },
+    {
+      "epoch": 10.3872605585045,
+      "grad_norm": 2.4140625,
+      "learning_rate": 2.5003335557038025e-05,
+      "loss": 0.8924,
+      "step": 11252
+    },
+    {
+      "epoch": 10.388183706438957,
+      "grad_norm": 3.466796875,
+      "learning_rate": 2.4996664442961977e-05,
+      "loss": 1.4044,
+      "step": 11253
+    },
+    {
+      "epoch": 10.389106854373413,
+      "grad_norm": 2.935546875,
+      "learning_rate": 2.4989993328885926e-05,
+      "loss": 1.356,
+      "step": 11254
+    },
+    {
+      "epoch": 10.39003000230787,
+      "grad_norm": 3.458984375,
+      "learning_rate": 2.4983322214809875e-05,
+      "loss": 1.2856,
+      "step": 11255
+    },
+    {
+      "epoch": 10.390953150242327,
+      "grad_norm": 3.142578125,
+      "learning_rate": 2.4976651100733824e-05,
+      "loss": 1.2091,
+      "step": 11256
+    },
+    {
+      "epoch": 10.391876298176783,
+      "grad_norm": 3.662109375,
+      "learning_rate": 2.4969979986657773e-05,
+      "loss": 1.3198,
+      "step": 11257
+    },
+    {
+      "epoch": 10.392799446111239,
+      "grad_norm": 2.91015625,
+      "learning_rate": 2.496330887258172e-05,
+      "loss": 1.3806,
+      "step": 11258
+    },
+    {
+      "epoch": 10.393722594045697,
+      "grad_norm": 2.990234375,
+      "learning_rate": 2.495663775850567e-05,
+      "loss": 1.3051,
+      "step": 11259
+    },
+    {
+      "epoch": 10.394645741980153,
+      "grad_norm": 3.154296875,
+      "learning_rate": 2.494996664442962e-05,
+      "loss": 1.2548,
+      "step": 11260
+    },
+    {
+      "epoch": 10.395568889914609,
+      "grad_norm": 3.279296875,
+      "learning_rate": 2.4943295530353568e-05,
+      "loss": 1.2798,
+      "step": 11261
+    },
+    {
+      "epoch": 10.396492037849065,
+      "grad_norm": 3.27734375,
+      "learning_rate": 2.493662441627752e-05,
+      "loss": 1.4341,
+      "step": 11262
+    },
+    {
+      "epoch": 10.397415185783522,
+      "grad_norm": 3.064453125,
+      "learning_rate": 2.492995330220147e-05,
+      "loss": 1.5049,
+      "step": 11263
+    },
+    {
+      "epoch": 10.398338333717978,
+      "grad_norm": 3.333984375,
+      "learning_rate": 2.4923282188125418e-05,
+      "loss": 1.4727,
+      "step": 11264
+    },
+    {
+      "epoch": 10.399261481652434,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.4916611074049367e-05,
+      "loss": 1.3865,
+      "step": 11265
+    },
+    {
+      "epoch": 10.40018462958689,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.4909939959973315e-05,
+      "loss": 1.1976,
+      "step": 11266
+    },
+    {
+      "epoch": 10.401107777521348,
+      "grad_norm": 3.6171875,
+      "learning_rate": 2.4903268845897264e-05,
+      "loss": 0.9832,
+      "step": 11267
+    },
+    {
+      "epoch": 10.402030925455804,
+      "grad_norm": 3.126953125,
+      "learning_rate": 2.4896597731821213e-05,
+      "loss": 1.1395,
+      "step": 11268
+    },
+    {
+      "epoch": 10.40295407339026,
+      "grad_norm": 3.86328125,
+      "learning_rate": 2.4889926617745165e-05,
+      "loss": 1.3401,
+      "step": 11269
+    },
+    {
+      "epoch": 10.403877221324716,
+      "grad_norm": 2.87890625,
+      "learning_rate": 2.4883255503669114e-05,
+      "loss": 1.3674,
+      "step": 11270
+    },
+    {
+      "epoch": 10.404800369259174,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.4876584389593063e-05,
+      "loss": 1.2013,
+      "step": 11271
+    },
+    {
+      "epoch": 10.40572351719363,
+      "grad_norm": 2.970703125,
+      "learning_rate": 2.486991327551701e-05,
+      "loss": 1.2751,
+      "step": 11272
+    },
+    {
+      "epoch": 10.406646665128086,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.486324216144096e-05,
+      "loss": 1.344,
+      "step": 11273
+    },
+    {
+      "epoch": 10.407569813062544,
+      "grad_norm": 2.9921875,
+      "learning_rate": 2.485657104736491e-05,
+      "loss": 1.4124,
+      "step": 11274
+    },
+    {
+      "epoch": 10.408492960997,
+      "grad_norm": 3.375,
+      "learning_rate": 2.484989993328886e-05,
+      "loss": 1.4009,
+      "step": 11275
+    },
+    {
+      "epoch": 10.409416108931456,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.484322881921281e-05,
+      "loss": 1.3372,
+      "step": 11276
+    },
+    {
+      "epoch": 10.410339256865912,
+      "grad_norm": 3.37109375,
+      "learning_rate": 2.483655770513676e-05,
+      "loss": 1.3225,
+      "step": 11277
+    },
+    {
+      "epoch": 10.41126240480037,
+      "grad_norm": 3.228515625,
+      "learning_rate": 2.4829886591060708e-05,
+      "loss": 1.3049,
+      "step": 11278
+    },
+    {
+      "epoch": 10.412185552734826,
+      "grad_norm": 3.52734375,
+      "learning_rate": 2.4823215476984657e-05,
+      "loss": 1.0901,
+      "step": 11279
+    },
+    {
+      "epoch": 10.413108700669282,
+      "grad_norm": 3.083984375,
+      "learning_rate": 2.4816544362908606e-05,
+      "loss": 1.1342,
+      "step": 11280
+    },
+    {
+      "epoch": 10.414031848603738,
+      "grad_norm": 2.90234375,
+      "learning_rate": 2.4809873248832558e-05,
+      "loss": 1.4333,
+      "step": 11281
+    },
+    {
+      "epoch": 10.414954996538196,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.4803202134756507e-05,
+      "loss": 1.201,
+      "step": 11282
+    },
+    {
+      "epoch": 10.415878144472652,
+      "grad_norm": 3.501953125,
+      "learning_rate": 2.4796531020680455e-05,
+      "loss": 1.3066,
+      "step": 11283
+    },
+    {
+      "epoch": 10.416801292407108,
+      "grad_norm": 3.10546875,
+      "learning_rate": 2.4789859906604404e-05,
+      "loss": 1.342,
+      "step": 11284
+    },
+    {
+      "epoch": 10.417724440341566,
+      "grad_norm": 3.43359375,
+      "learning_rate": 2.4783188792528353e-05,
+      "loss": 0.7357,
+      "step": 11285
+    },
+    {
+      "epoch": 10.418647588276022,
+      "grad_norm": 2.93359375,
+      "learning_rate": 2.4776517678452302e-05,
+      "loss": 1.2151,
+      "step": 11286
+    },
+    {
+      "epoch": 10.419570736210478,
+      "grad_norm": 3.291015625,
+      "learning_rate": 2.4769846564376254e-05,
+      "loss": 1.1755,
+      "step": 11287
+    },
+    {
+      "epoch": 10.420493884144934,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.4763175450300203e-05,
+      "loss": 1.0718,
+      "step": 11288
+    },
+    {
+      "epoch": 10.421417032079392,
+      "grad_norm": 3.169921875,
+      "learning_rate": 2.4756504336224152e-05,
+      "loss": 1.2107,
+      "step": 11289
+    },
+    {
+      "epoch": 10.422340180013848,
+      "grad_norm": 3.271484375,
+      "learning_rate": 2.47498332221481e-05,
+      "loss": 1.3884,
+      "step": 11290
+    },
+    {
+      "epoch": 10.423263327948304,
+      "grad_norm": 3.240234375,
+      "learning_rate": 2.474316210807205e-05,
+      "loss": 1.2446,
+      "step": 11291
+    },
+    {
+      "epoch": 10.42418647588276,
+      "grad_norm": 3.931640625,
+      "learning_rate": 2.4736490993995998e-05,
+      "loss": 1.2192,
+      "step": 11292
+    },
+    {
+      "epoch": 10.425109623817217,
+      "grad_norm": 3.318359375,
+      "learning_rate": 2.4729819879919947e-05,
+      "loss": 1.4219,
+      "step": 11293
+    },
+    {
+      "epoch": 10.426032771751673,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.47231487658439e-05,
+      "loss": 1.3688,
+      "step": 11294
+    },
+    {
+      "epoch": 10.42695591968613,
+      "grad_norm": 3.798828125,
+      "learning_rate": 2.4716477651767848e-05,
+      "loss": 1.1292,
+      "step": 11295
+    },
+    {
+      "epoch": 10.427879067620585,
+      "grad_norm": 3.064453125,
+      "learning_rate": 2.4709806537691797e-05,
+      "loss": 1.1738,
+      "step": 11296
+    },
+    {
+      "epoch": 10.428802215555043,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.4703135423615746e-05,
+      "loss": 1.3274,
+      "step": 11297
+    },
+    {
+      "epoch": 10.4297253634895,
+      "grad_norm": 3.447265625,
+      "learning_rate": 2.4696464309539694e-05,
+      "loss": 1.2153,
+      "step": 11298
+    },
+    {
+      "epoch": 10.430648511423955,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.4689793195463643e-05,
+      "loss": 1.5291,
+      "step": 11299
+    },
+    {
+      "epoch": 10.431571659358411,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.4683122081387595e-05,
+      "loss": 1.1148,
+      "step": 11300
+    },
+    {
+      "epoch": 10.43249480729287,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.4676450967311544e-05,
+      "loss": 1.1082,
+      "step": 11301
+    },
+    {
+      "epoch": 10.433417955227325,
+      "grad_norm": 2.92578125,
+      "learning_rate": 2.4669779853235493e-05,
+      "loss": 1.3716,
+      "step": 11302
+    },
+    {
+      "epoch": 10.434341103161781,
+      "grad_norm": 3.169921875,
+      "learning_rate": 2.4663108739159442e-05,
+      "loss": 1.3142,
+      "step": 11303
+    },
+    {
+      "epoch": 10.435264251096239,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.465643762508339e-05,
+      "loss": 1.0353,
+      "step": 11304
+    },
+    {
+      "epoch": 10.436187399030695,
+      "grad_norm": 2.9296875,
+      "learning_rate": 2.464976651100734e-05,
+      "loss": 1.269,
+      "step": 11305
+    },
+    {
+      "epoch": 10.437110546965151,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.464309539693129e-05,
+      "loss": 1.3079,
+      "step": 11306
+    },
+    {
+      "epoch": 10.438033694899607,
+      "grad_norm": 3.771484375,
+      "learning_rate": 2.4636424282855237e-05,
+      "loss": 1.3955,
+      "step": 11307
+    },
+    {
+      "epoch": 10.438956842834065,
+      "grad_norm": 3.625,
+      "learning_rate": 2.4629753168779186e-05,
+      "loss": 1.2815,
+      "step": 11308
+    },
+    {
+      "epoch": 10.439879990768521,
+      "grad_norm": 3.044921875,
+      "learning_rate": 2.4623082054703138e-05,
+      "loss": 1.3676,
+      "step": 11309
+    },
+    {
+      "epoch": 10.440803138702977,
+      "grad_norm": 2.95703125,
+      "learning_rate": 2.4616410940627087e-05,
+      "loss": 1.2778,
+      "step": 11310
+    },
+    {
+      "epoch": 10.441726286637433,
+      "grad_norm": 3.375,
+      "learning_rate": 2.4609739826551036e-05,
+      "loss": 1.1737,
+      "step": 11311
+    },
+    {
+      "epoch": 10.44264943457189,
+      "grad_norm": 3.0703125,
+      "learning_rate": 2.4603068712474985e-05,
+      "loss": 1.2174,
+      "step": 11312
+    },
+    {
+      "epoch": 10.443572582506347,
+      "grad_norm": 3.322265625,
+      "learning_rate": 2.4596397598398934e-05,
+      "loss": 1.2764,
+      "step": 11313
+    },
+    {
+      "epoch": 10.444495730440803,
+      "grad_norm": 3.525390625,
+      "learning_rate": 2.4589726484322882e-05,
+      "loss": 1.4619,
+      "step": 11314
+    },
+    {
+      "epoch": 10.445418878375259,
+      "grad_norm": 3.1328125,
+      "learning_rate": 2.458305537024683e-05,
+      "loss": 1.1841,
+      "step": 11315
+    },
+    {
+      "epoch": 10.446342026309717,
+      "grad_norm": 4.10546875,
+      "learning_rate": 2.457638425617078e-05,
+      "loss": 1.3496,
+      "step": 11316
+    },
+    {
+      "epoch": 10.447265174244173,
+      "grad_norm": 3.603515625,
+      "learning_rate": 2.456971314209473e-05,
+      "loss": 1.3169,
+      "step": 11317
+    },
+    {
+      "epoch": 10.448188322178629,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.4563042028018678e-05,
+      "loss": 1.3516,
+      "step": 11318
+    },
+    {
+      "epoch": 10.449111470113085,
+      "grad_norm": 3.62109375,
+      "learning_rate": 2.455637091394263e-05,
+      "loss": 1.2224,
+      "step": 11319
+    },
+    {
+      "epoch": 10.450034618047543,
+      "grad_norm": 3.42578125,
+      "learning_rate": 2.454969979986658e-05,
+      "loss": 1.1705,
+      "step": 11320
+    },
+    {
+      "epoch": 10.450957765981999,
+      "grad_norm": 3.060546875,
+      "learning_rate": 2.4543028685790527e-05,
+      "loss": 1.2531,
+      "step": 11321
+    },
+    {
+      "epoch": 10.451880913916455,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.4536357571714476e-05,
+      "loss": 1.1912,
+      "step": 11322
+    },
+    {
+      "epoch": 10.452804061850912,
+      "grad_norm": 2.92578125,
+      "learning_rate": 2.4529686457638425e-05,
+      "loss": 1.1074,
+      "step": 11323
+    },
+    {
+      "epoch": 10.453727209785368,
+      "grad_norm": 3.279296875,
+      "learning_rate": 2.4523015343562374e-05,
+      "loss": 1.2814,
+      "step": 11324
+    },
+    {
+      "epoch": 10.454650357719824,
+      "grad_norm": 3.388671875,
+      "learning_rate": 2.4516344229486326e-05,
+      "loss": 1.2141,
+      "step": 11325
+    },
+    {
+      "epoch": 10.45557350565428,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.4509673115410275e-05,
+      "loss": 1.3257,
+      "step": 11326
+    },
+    {
+      "epoch": 10.456496653588738,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.4503002001334224e-05,
+      "loss": 1.4385,
+      "step": 11327
+    },
+    {
+      "epoch": 10.457419801523194,
+      "grad_norm": 3.62109375,
+      "learning_rate": 2.4496330887258173e-05,
+      "loss": 1.2722,
+      "step": 11328
+    },
+    {
+      "epoch": 10.45834294945765,
+      "grad_norm": 3.51953125,
+      "learning_rate": 2.448965977318212e-05,
+      "loss": 0.9625,
+      "step": 11329
+    },
+    {
+      "epoch": 10.459266097392106,
+      "grad_norm": 3.42578125,
+      "learning_rate": 2.448298865910607e-05,
+      "loss": 1.1614,
+      "step": 11330
+    },
+    {
+      "epoch": 10.460189245326564,
+      "grad_norm": 3.056640625,
+      "learning_rate": 2.4476317545030022e-05,
+      "loss": 1.1421,
+      "step": 11331
+    },
+    {
+      "epoch": 10.46111239326102,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.446964643095397e-05,
+      "loss": 1.2086,
+      "step": 11332
+    },
+    {
+      "epoch": 10.462035541195476,
+      "grad_norm": 2.9296875,
+      "learning_rate": 2.446297531687792e-05,
+      "loss": 1.3774,
+      "step": 11333
+    },
+    {
+      "epoch": 10.462958689129934,
+      "grad_norm": 3.623046875,
+      "learning_rate": 2.445630420280187e-05,
+      "loss": 1.3179,
+      "step": 11334
+    },
+    {
+      "epoch": 10.46388183706439,
+      "grad_norm": 3.646484375,
+      "learning_rate": 2.4449633088725818e-05,
+      "loss": 1.3762,
+      "step": 11335
+    },
+    {
+      "epoch": 10.464804984998846,
+      "grad_norm": 3.0546875,
+      "learning_rate": 2.4442961974649767e-05,
+      "loss": 1.1962,
+      "step": 11336
+    },
+    {
+      "epoch": 10.465728132933302,
+      "grad_norm": 3.05078125,
+      "learning_rate": 2.4436290860573715e-05,
+      "loss": 1.1904,
+      "step": 11337
+    },
+    {
+      "epoch": 10.46665128086776,
+      "grad_norm": 3.498046875,
+      "learning_rate": 2.4429619746497668e-05,
+      "loss": 1.3354,
+      "step": 11338
+    },
+    {
+      "epoch": 10.467574428802216,
+      "grad_norm": 2.646484375,
+      "learning_rate": 2.4422948632421616e-05,
+      "loss": 1.1466,
+      "step": 11339
+    },
+    {
+      "epoch": 10.468497576736672,
+      "grad_norm": 3.126953125,
+      "learning_rate": 2.4416277518345565e-05,
+      "loss": 1.2904,
+      "step": 11340
+    },
+    {
+      "epoch": 10.469420724671128,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.4409606404269514e-05,
+      "loss": 1.3979,
+      "step": 11341
+    },
+    {
+      "epoch": 10.470343872605586,
+      "grad_norm": 3.78515625,
+      "learning_rate": 2.4402935290193463e-05,
+      "loss": 1.4683,
+      "step": 11342
+    },
+    {
+      "epoch": 10.471267020540042,
+      "grad_norm": 3.85546875,
+      "learning_rate": 2.439626417611741e-05,
+      "loss": 0.9562,
+      "step": 11343
+    },
+    {
+      "epoch": 10.472190168474498,
+      "grad_norm": 2.83203125,
+      "learning_rate": 2.4389593062041364e-05,
+      "loss": 1.2349,
+      "step": 11344
+    },
+    {
+      "epoch": 10.473113316408954,
+      "grad_norm": 3.25,
+      "learning_rate": 2.4382921947965313e-05,
+      "loss": 1.4365,
+      "step": 11345
+    },
+    {
+      "epoch": 10.474036464343412,
+      "grad_norm": 3.0078125,
+      "learning_rate": 2.437625083388926e-05,
+      "loss": 1.4004,
+      "step": 11346
+    },
+    {
+      "epoch": 10.474959612277868,
+      "grad_norm": 2.966796875,
+      "learning_rate": 2.436957971981321e-05,
+      "loss": 1.3538,
+      "step": 11347
+    },
+    {
+      "epoch": 10.475882760212324,
+      "grad_norm": 3.189453125,
+      "learning_rate": 2.436290860573716e-05,
+      "loss": 1.2727,
+      "step": 11348
+    },
+    {
+      "epoch": 10.47680590814678,
+      "grad_norm": 3.5,
+      "learning_rate": 2.4356237491661108e-05,
+      "loss": 1.1851,
+      "step": 11349
+    },
+    {
+      "epoch": 10.477729056081238,
+      "grad_norm": 3.326171875,
+      "learning_rate": 2.434956637758506e-05,
+      "loss": 1.1442,
+      "step": 11350
+    },
+    {
+      "epoch": 10.478652204015694,
+      "grad_norm": 3.67578125,
+      "learning_rate": 2.434289526350901e-05,
+      "loss": 1.2678,
+      "step": 11351
+    },
+    {
+      "epoch": 10.47957535195015,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.4336224149432958e-05,
+      "loss": 1.4233,
+      "step": 11352
+    },
+    {
+      "epoch": 10.480498499884607,
+      "grad_norm": 3.52734375,
+      "learning_rate": 2.4329553035356907e-05,
+      "loss": 0.9303,
+      "step": 11353
+    },
+    {
+      "epoch": 10.481421647819063,
+      "grad_norm": 2.7109375,
+      "learning_rate": 2.4322881921280855e-05,
+      "loss": 1.2642,
+      "step": 11354
+    },
+    {
+      "epoch": 10.48234479575352,
+      "grad_norm": 3.611328125,
+      "learning_rate": 2.4316210807204804e-05,
+      "loss": 1.488,
+      "step": 11355
+    },
+    {
+      "epoch": 10.483267943687975,
+      "grad_norm": 2.7734375,
+      "learning_rate": 2.4309539693128756e-05,
+      "loss": 1.449,
+      "step": 11356
+    },
+    {
+      "epoch": 10.484191091622433,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.4302868579052705e-05,
+      "loss": 1.4417,
+      "step": 11357
+    },
+    {
+      "epoch": 10.48511423955689,
+      "grad_norm": 3.404296875,
+      "learning_rate": 2.4296197464976654e-05,
+      "loss": 1.2751,
+      "step": 11358
+    },
+    {
+      "epoch": 10.486037387491345,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.4289526350900603e-05,
+      "loss": 1.364,
+      "step": 11359
+    },
+    {
+      "epoch": 10.486960535425801,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.4282855236824552e-05,
+      "loss": 1.4172,
+      "step": 11360
+    },
+    {
+      "epoch": 10.48788368336026,
+      "grad_norm": 3.33203125,
+      "learning_rate": 2.42761841227485e-05,
+      "loss": 1.1787,
+      "step": 11361
+    },
+    {
+      "epoch": 10.488806831294715,
+      "grad_norm": 3.6328125,
+      "learning_rate": 2.426951300867245e-05,
+      "loss": 1.2944,
+      "step": 11362
+    },
+    {
+      "epoch": 10.489729979229171,
+      "grad_norm": 3.376953125,
+      "learning_rate": 2.4262841894596398e-05,
+      "loss": 1.0002,
+      "step": 11363
+    },
+    {
+      "epoch": 10.490653127163627,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.4256170780520347e-05,
+      "loss": 1.4102,
+      "step": 11364
+    },
+    {
+      "epoch": 10.491576275098085,
+      "grad_norm": 3.48046875,
+      "learning_rate": 2.4249499666444296e-05,
+      "loss": 1.3221,
+      "step": 11365
+    },
+    {
+      "epoch": 10.492499423032541,
+      "grad_norm": 3.08203125,
+      "learning_rate": 2.4242828552368248e-05,
+      "loss": 1.1497,
+      "step": 11366
+    },
+    {
+      "epoch": 10.493422570966997,
+      "grad_norm": 3.080078125,
+      "learning_rate": 2.4236157438292197e-05,
+      "loss": 1.3606,
+      "step": 11367
+    },
+    {
+      "epoch": 10.494345718901453,
+      "grad_norm": 2.916015625,
+      "learning_rate": 2.4229486324216146e-05,
+      "loss": 1.0236,
+      "step": 11368
+    },
+    {
+      "epoch": 10.49526886683591,
+      "grad_norm": 3.5078125,
+      "learning_rate": 2.4222815210140094e-05,
+      "loss": 1.2927,
+      "step": 11369
+    },
+    {
+      "epoch": 10.496192014770367,
+      "grad_norm": 3.1015625,
+      "learning_rate": 2.4216144096064043e-05,
+      "loss": 1.4333,
+      "step": 11370
+    },
+    {
+      "epoch": 10.497115162704823,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.4209472981987992e-05,
+      "loss": 1.2258,
+      "step": 11371
+    },
+    {
+      "epoch": 10.49803831063928,
+      "grad_norm": 3.009765625,
+      "learning_rate": 2.420280186791194e-05,
+      "loss": 1.2456,
+      "step": 11372
+    },
+    {
+      "epoch": 10.498961458573737,
+      "grad_norm": 3.291015625,
+      "learning_rate": 2.419613075383589e-05,
+      "loss": 1.4268,
+      "step": 11373
+    },
+    {
+      "epoch": 10.499884606508193,
+      "grad_norm": 3.4453125,
+      "learning_rate": 2.418945963975984e-05,
+      "loss": 1.2178,
+      "step": 11374
+    },
+    {
+      "epoch": 10.500807754442649,
+      "grad_norm": 3.193359375,
+      "learning_rate": 2.418278852568379e-05,
+      "loss": 1.3052,
+      "step": 11375
+    },
+    {
+      "epoch": 10.501730902377107,
+      "grad_norm": 3.068359375,
+      "learning_rate": 2.417611741160774e-05,
+      "loss": 1.3271,
+      "step": 11376
+    },
+    {
+      "epoch": 10.502654050311563,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.416944629753169e-05,
+      "loss": 1.325,
+      "step": 11377
+    },
+    {
+      "epoch": 10.503577198246019,
+      "grad_norm": 3.58984375,
+      "learning_rate": 2.4162775183455637e-05,
+      "loss": 1.4548,
+      "step": 11378
+    },
+    {
+      "epoch": 10.504500346180475,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.4156104069379586e-05,
+      "loss": 1.2192,
+      "step": 11379
+    },
+    {
+      "epoch": 10.505423494114932,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.4149432955303535e-05,
+      "loss": 1.1716,
+      "step": 11380
+    },
+    {
+      "epoch": 10.506346642049389,
+      "grad_norm": 3.1015625,
+      "learning_rate": 2.4142761841227487e-05,
+      "loss": 1.2537,
+      "step": 11381
+    },
+    {
+      "epoch": 10.507269789983845,
+      "grad_norm": 2.982421875,
+      "learning_rate": 2.4136090727151436e-05,
+      "loss": 1.3689,
+      "step": 11382
+    },
+    {
+      "epoch": 10.508192937918302,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.4129419613075385e-05,
+      "loss": 1.1859,
+      "step": 11383
+    },
+    {
+      "epoch": 10.509116085852758,
+      "grad_norm": 2.720703125,
+      "learning_rate": 2.4122748498999334e-05,
+      "loss": 1.2297,
+      "step": 11384
+    },
+    {
+      "epoch": 10.510039233787214,
+      "grad_norm": 3.302734375,
+      "learning_rate": 2.4116077384923282e-05,
+      "loss": 1.4092,
+      "step": 11385
+    },
+    {
+      "epoch": 10.51096238172167,
+      "grad_norm": 2.8515625,
+      "learning_rate": 2.410940627084723e-05,
+      "loss": 1.3154,
+      "step": 11386
+    },
+    {
+      "epoch": 10.511885529656128,
+      "grad_norm": 2.91015625,
+      "learning_rate": 2.410273515677118e-05,
+      "loss": 1.396,
+      "step": 11387
+    },
+    {
+      "epoch": 10.512808677590584,
+      "grad_norm": 3.076171875,
+      "learning_rate": 2.4096064042695132e-05,
+      "loss": 1.1726,
+      "step": 11388
+    },
+    {
+      "epoch": 10.51373182552504,
+      "grad_norm": 3.34765625,
+      "learning_rate": 2.408939292861908e-05,
+      "loss": 1.3962,
+      "step": 11389
+    },
+    {
+      "epoch": 10.514654973459496,
+      "grad_norm": 3.376953125,
+      "learning_rate": 2.408272181454303e-05,
+      "loss": 1.3982,
+      "step": 11390
+    },
+    {
+      "epoch": 10.515578121393954,
+      "grad_norm": 3.2578125,
+      "learning_rate": 2.407605070046698e-05,
+      "loss": 1.3511,
+      "step": 11391
+    },
+    {
+      "epoch": 10.51650126932841,
+      "grad_norm": 3.623046875,
+      "learning_rate": 2.4069379586390927e-05,
+      "loss": 1.4058,
+      "step": 11392
+    },
+    {
+      "epoch": 10.517424417262866,
+      "grad_norm": 3.654296875,
+      "learning_rate": 2.4062708472314876e-05,
+      "loss": 1.3772,
+      "step": 11393
+    },
+    {
+      "epoch": 10.518347565197322,
+      "grad_norm": 3.470703125,
+      "learning_rate": 2.405603735823883e-05,
+      "loss": 1.1281,
+      "step": 11394
+    },
+    {
+      "epoch": 10.51927071313178,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.4049366244162777e-05,
+      "loss": 1.2198,
+      "step": 11395
+    },
+    {
+      "epoch": 10.520193861066236,
+      "grad_norm": 3.486328125,
+      "learning_rate": 2.4042695130086726e-05,
+      "loss": 1.1897,
+      "step": 11396
+    },
+    {
+      "epoch": 10.521117009000692,
+      "grad_norm": 3.185546875,
+      "learning_rate": 2.4036024016010675e-05,
+      "loss": 1.2656,
+      "step": 11397
+    },
+    {
+      "epoch": 10.522040156935148,
+      "grad_norm": 3.18359375,
+      "learning_rate": 2.4029352901934624e-05,
+      "loss": 1.1985,
+      "step": 11398
+    },
+    {
+      "epoch": 10.522963304869606,
+      "grad_norm": 3.4140625,
+      "learning_rate": 2.4022681787858573e-05,
+      "loss": 1.1488,
+      "step": 11399
+    },
+    {
+      "epoch": 10.523886452804062,
+      "grad_norm": 2.931640625,
+      "learning_rate": 2.4016010673782525e-05,
+      "loss": 1.3406,
+      "step": 11400
+    },
+    {
+      "epoch": 10.524809600738518,
+      "grad_norm": 2.896484375,
+      "learning_rate": 2.4009339559706474e-05,
+      "loss": 1.2812,
+      "step": 11401
+    },
+    {
+      "epoch": 10.525732748672976,
+      "grad_norm": 3.498046875,
+      "learning_rate": 2.4002668445630422e-05,
+      "loss": 1.1454,
+      "step": 11402
+    },
+    {
+      "epoch": 10.526655896607432,
+      "grad_norm": 3.138671875,
+      "learning_rate": 2.399599733155437e-05,
+      "loss": 1.3711,
+      "step": 11403
+    },
+    {
+      "epoch": 10.527579044541888,
+      "grad_norm": 3.126953125,
+      "learning_rate": 2.398932621747832e-05,
+      "loss": 1.2791,
+      "step": 11404
+    },
+    {
+      "epoch": 10.528502192476344,
+      "grad_norm": 3.2421875,
+      "learning_rate": 2.398265510340227e-05,
+      "loss": 1.4319,
+      "step": 11405
+    },
+    {
+      "epoch": 10.529425340410802,
+      "grad_norm": 3.3203125,
+      "learning_rate": 2.3975983989326218e-05,
+      "loss": 1.2887,
+      "step": 11406
+    },
+    {
+      "epoch": 10.530348488345258,
+      "grad_norm": 3.623046875,
+      "learning_rate": 2.396931287525017e-05,
+      "loss": 1.2253,
+      "step": 11407
+    },
+    {
+      "epoch": 10.531271636279714,
+      "grad_norm": 3.884765625,
+      "learning_rate": 2.396264176117412e-05,
+      "loss": 1.1915,
+      "step": 11408
+    },
+    {
+      "epoch": 10.53219478421417,
+      "grad_norm": 4.2890625,
+      "learning_rate": 2.3955970647098068e-05,
+      "loss": 1.3591,
+      "step": 11409
+    },
+    {
+      "epoch": 10.533117932148627,
+      "grad_norm": 3.55859375,
+      "learning_rate": 2.3949299533022016e-05,
+      "loss": 1.028,
+      "step": 11410
+    },
+    {
+      "epoch": 10.534041080083083,
+      "grad_norm": 3.291015625,
+      "learning_rate": 2.3942628418945965e-05,
+      "loss": 1.478,
+      "step": 11411
+    },
+    {
+      "epoch": 10.53496422801754,
+      "grad_norm": 3.08203125,
+      "learning_rate": 2.3935957304869914e-05,
+      "loss": 1.4646,
+      "step": 11412
+    },
+    {
+      "epoch": 10.535887375951996,
+      "grad_norm": 3.013671875,
+      "learning_rate": 2.3929286190793866e-05,
+      "loss": 1.2003,
+      "step": 11413
+    },
+    {
+      "epoch": 10.536810523886453,
+      "grad_norm": 3.427734375,
+      "learning_rate": 2.3922615076717815e-05,
+      "loss": 1.2673,
+      "step": 11414
+    },
+    {
+      "epoch": 10.53773367182091,
+      "grad_norm": 3.1484375,
+      "learning_rate": 2.3915943962641764e-05,
+      "loss": 1.3145,
+      "step": 11415
+    },
+    {
+      "epoch": 10.538656819755365,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.3909272848565713e-05,
+      "loss": 1.2192,
+      "step": 11416
+    },
+    {
+      "epoch": 10.539579967689821,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.390260173448966e-05,
+      "loss": 1.1255,
+      "step": 11417
+    },
+    {
+      "epoch": 10.54050311562428,
+      "grad_norm": 3.216796875,
+      "learning_rate": 2.389593062041361e-05,
+      "loss": 1.2343,
+      "step": 11418
+    },
+    {
+      "epoch": 10.541426263558735,
+      "grad_norm": 3.28515625,
+      "learning_rate": 2.388925950633756e-05,
+      "loss": 1.184,
+      "step": 11419
+    },
+    {
+      "epoch": 10.542349411493191,
+      "grad_norm": 3.76171875,
+      "learning_rate": 2.3882588392261508e-05,
+      "loss": 1.4558,
+      "step": 11420
+    },
+    {
+      "epoch": 10.543272559427649,
+      "grad_norm": 2.900390625,
+      "learning_rate": 2.3875917278185457e-05,
+      "loss": 1.1414,
+      "step": 11421
+    },
+    {
+      "epoch": 10.544195707362105,
+      "grad_norm": 3.232421875,
+      "learning_rate": 2.386924616410941e-05,
+      "loss": 1.2319,
+      "step": 11422
+    },
+    {
+      "epoch": 10.545118855296561,
+      "grad_norm": 3.341796875,
+      "learning_rate": 2.3862575050033358e-05,
+      "loss": 1.3066,
+      "step": 11423
+    },
+    {
+      "epoch": 10.546042003231017,
+      "grad_norm": 3.134765625,
+      "learning_rate": 2.3855903935957307e-05,
+      "loss": 1.2559,
+      "step": 11424
+    },
+    {
+      "epoch": 10.546965151165475,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.3849232821881255e-05,
+      "loss": 1.2375,
+      "step": 11425
+    },
+    {
+      "epoch": 10.547888299099931,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.3842561707805204e-05,
+      "loss": 1.1484,
+      "step": 11426
+    },
+    {
+      "epoch": 10.548811447034387,
+      "grad_norm": 3.396484375,
+      "learning_rate": 2.3835890593729153e-05,
+      "loss": 1.3826,
+      "step": 11427
+    },
+    {
+      "epoch": 10.549734594968843,
+      "grad_norm": 2.68359375,
+      "learning_rate": 2.3829219479653102e-05,
+      "loss": 1.0017,
+      "step": 11428
+    },
+    {
+      "epoch": 10.5506577429033,
+      "grad_norm": 3.673828125,
+      "learning_rate": 2.382254836557705e-05,
+      "loss": 1.4556,
+      "step": 11429
+    },
+    {
+      "epoch": 10.551580890837757,
+      "grad_norm": 3.20703125,
+      "learning_rate": 2.3815877251501e-05,
+      "loss": 1.1346,
+      "step": 11430
+    },
+    {
+      "epoch": 10.552504038772213,
+      "grad_norm": 3.369140625,
+      "learning_rate": 2.380920613742495e-05,
+      "loss": 1.3779,
+      "step": 11431
+    },
+    {
+      "epoch": 10.55342718670667,
+      "grad_norm": 3.115234375,
+      "learning_rate": 2.38025350233489e-05,
+      "loss": 1.5024,
+      "step": 11432
+    },
+    {
+      "epoch": 10.554350334641127,
+      "grad_norm": 3.52734375,
+      "learning_rate": 2.379586390927285e-05,
+      "loss": 1.0701,
+      "step": 11433
+    },
+    {
+      "epoch": 10.555273482575583,
+      "grad_norm": 3.306640625,
+      "learning_rate": 2.3789192795196798e-05,
+      "loss": 1.2881,
+      "step": 11434
+    },
+    {
+      "epoch": 10.556196630510039,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.3782521681120747e-05,
+      "loss": 1.2549,
+      "step": 11435
+    },
+    {
+      "epoch": 10.557119778444497,
+      "grad_norm": 3.15234375,
+      "learning_rate": 2.3775850567044696e-05,
+      "loss": 1.1746,
+      "step": 11436
+    },
+    {
+      "epoch": 10.558042926378953,
+      "grad_norm": 3.1953125,
+      "learning_rate": 2.3769179452968645e-05,
+      "loss": 1.3379,
+      "step": 11437
+    },
+    {
+      "epoch": 10.558966074313409,
+      "grad_norm": 3.27734375,
+      "learning_rate": 2.3762508338892597e-05,
+      "loss": 1.4355,
+      "step": 11438
+    },
+    {
+      "epoch": 10.559889222247865,
+      "grad_norm": 3.408203125,
+      "learning_rate": 2.3755837224816546e-05,
+      "loss": 1.5278,
+      "step": 11439
+    },
+    {
+      "epoch": 10.560812370182322,
+      "grad_norm": 3.275390625,
+      "learning_rate": 2.3749166110740494e-05,
+      "loss": 1.3828,
+      "step": 11440
+    },
+    {
+      "epoch": 10.561735518116778,
+      "grad_norm": 3.427734375,
+      "learning_rate": 2.3742494996664443e-05,
+      "loss": 1.3589,
+      "step": 11441
+    },
+    {
+      "epoch": 10.562658666051234,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.3735823882588392e-05,
+      "loss": 1.3145,
+      "step": 11442
+    },
+    {
+      "epoch": 10.56358181398569,
+      "grad_norm": 3.857421875,
+      "learning_rate": 2.372915276851234e-05,
+      "loss": 1.3582,
+      "step": 11443
+    },
+    {
+      "epoch": 10.564504961920148,
+      "grad_norm": 3.314453125,
+      "learning_rate": 2.3722481654436293e-05,
+      "loss": 0.9499,
+      "step": 11444
+    },
+    {
+      "epoch": 10.565428109854604,
+      "grad_norm": 3.662109375,
+      "learning_rate": 2.3715810540360242e-05,
+      "loss": 1.3628,
+      "step": 11445
+    },
+    {
+      "epoch": 10.56635125778906,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.370913942628419e-05,
+      "loss": 1.2186,
+      "step": 11446
+    },
+    {
+      "epoch": 10.567274405723516,
+      "grad_norm": 3.37109375,
+      "learning_rate": 2.370246831220814e-05,
+      "loss": 1.3884,
+      "step": 11447
+    },
+    {
+      "epoch": 10.568197553657974,
+      "grad_norm": 3.806640625,
+      "learning_rate": 2.369579719813209e-05,
+      "loss": 1.3911,
+      "step": 11448
+    },
+    {
+      "epoch": 10.56912070159243,
+      "grad_norm": 2.84765625,
+      "learning_rate": 2.3689126084056037e-05,
+      "loss": 1.2443,
+      "step": 11449
+    },
+    {
+      "epoch": 10.570043849526886,
+      "grad_norm": 3.263671875,
+      "learning_rate": 2.368245496997999e-05,
+      "loss": 1.0383,
+      "step": 11450
+    },
+    {
+      "epoch": 10.570966997461344,
+      "grad_norm": 3.611328125,
+      "learning_rate": 2.3675783855903938e-05,
+      "loss": 1.193,
+      "step": 11451
+    },
+    {
+      "epoch": 10.5718901453958,
+      "grad_norm": 3.26953125,
+      "learning_rate": 2.3669112741827887e-05,
+      "loss": 1.2518,
+      "step": 11452
+    },
+    {
+      "epoch": 10.572813293330256,
+      "grad_norm": 3.275390625,
+      "learning_rate": 2.3662441627751836e-05,
+      "loss": 1.2892,
+      "step": 11453
+    },
+    {
+      "epoch": 10.573736441264712,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.3655770513675785e-05,
+      "loss": 1.2319,
+      "step": 11454
+    },
+    {
+      "epoch": 10.57465958919917,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.3649099399599734e-05,
+      "loss": 1.2128,
+      "step": 11455
+    },
+    {
+      "epoch": 10.575582737133626,
+      "grad_norm": 3.154296875,
+      "learning_rate": 2.3642428285523682e-05,
+      "loss": 1.1139,
+      "step": 11456
+    },
+    {
+      "epoch": 10.576505885068082,
+      "grad_norm": 2.826171875,
+      "learning_rate": 2.3635757171447635e-05,
+      "loss": 1.438,
+      "step": 11457
+    },
+    {
+      "epoch": 10.577429033002538,
+      "grad_norm": 3.693359375,
+      "learning_rate": 2.3629086057371583e-05,
+      "loss": 0.9181,
+      "step": 11458
+    },
+    {
+      "epoch": 10.578352180936996,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.3622414943295532e-05,
+      "loss": 1.2668,
+      "step": 11459
+    },
+    {
+      "epoch": 10.579275328871452,
+      "grad_norm": 3.29296875,
+      "learning_rate": 2.361574382921948e-05,
+      "loss": 1.2571,
+      "step": 11460
+    },
+    {
+      "epoch": 10.580198476805908,
+      "grad_norm": 2.884765625,
+      "learning_rate": 2.360907271514343e-05,
+      "loss": 1.2004,
+      "step": 11461
+    },
+    {
+      "epoch": 10.581121624740364,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.360240160106738e-05,
+      "loss": 1.3647,
+      "step": 11462
+    },
+    {
+      "epoch": 10.582044772674822,
+      "grad_norm": 3.51953125,
+      "learning_rate": 2.359573048699133e-05,
+      "loss": 1.2101,
+      "step": 11463
+    },
+    {
+      "epoch": 10.582967920609278,
+      "grad_norm": 2.8203125,
+      "learning_rate": 2.358905937291528e-05,
+      "loss": 1.2146,
+      "step": 11464
+    },
+    {
+      "epoch": 10.583891068543734,
+      "grad_norm": 2.888671875,
+      "learning_rate": 2.358238825883923e-05,
+      "loss": 1.0441,
+      "step": 11465
+    },
+    {
+      "epoch": 10.58481421647819,
+      "grad_norm": 2.919921875,
+      "learning_rate": 2.3575717144763177e-05,
+      "loss": 0.9725,
+      "step": 11466
+    },
+    {
+      "epoch": 10.585737364412648,
+      "grad_norm": 3.49609375,
+      "learning_rate": 2.3569046030687126e-05,
+      "loss": 1.3945,
+      "step": 11467
+    },
+    {
+      "epoch": 10.586660512347104,
+      "grad_norm": 3.177734375,
+      "learning_rate": 2.3562374916611075e-05,
+      "loss": 1.3704,
+      "step": 11468
+    },
+    {
+      "epoch": 10.58758366028156,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.3555703802535027e-05,
+      "loss": 1.1089,
+      "step": 11469
+    },
+    {
+      "epoch": 10.588506808216017,
+      "grad_norm": 3.482421875,
+      "learning_rate": 2.3549032688458976e-05,
+      "loss": 1.312,
+      "step": 11470
+    },
+    {
+      "epoch": 10.589429956150473,
+      "grad_norm": 3.212890625,
+      "learning_rate": 2.3542361574382925e-05,
+      "loss": 1.397,
+      "step": 11471
+    },
+    {
+      "epoch": 10.59035310408493,
+      "grad_norm": 3.361328125,
+      "learning_rate": 2.3535690460306874e-05,
+      "loss": 1.177,
+      "step": 11472
+    },
+    {
+      "epoch": 10.591276252019385,
+      "grad_norm": 3.0703125,
+      "learning_rate": 2.3529019346230822e-05,
+      "loss": 1.0551,
+      "step": 11473
+    },
+    {
+      "epoch": 10.592199399953843,
+      "grad_norm": 3.552734375,
+      "learning_rate": 2.352234823215477e-05,
+      "loss": 1.2837,
+      "step": 11474
+    },
+    {
+      "epoch": 10.5931225478883,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.351567711807872e-05,
+      "loss": 1.1803,
+      "step": 11475
+    },
+    {
+      "epoch": 10.594045695822755,
+      "grad_norm": 3.384765625,
+      "learning_rate": 2.350900600400267e-05,
+      "loss": 1.2273,
+      "step": 11476
+    },
+    {
+      "epoch": 10.594968843757211,
+      "grad_norm": 3.103515625,
+      "learning_rate": 2.3502334889926618e-05,
+      "loss": 1.0092,
+      "step": 11477
+    },
+    {
+      "epoch": 10.59589199169167,
+      "grad_norm": 2.8671875,
+      "learning_rate": 2.3495663775850567e-05,
+      "loss": 1.5232,
+      "step": 11478
+    },
+    {
+      "epoch": 10.596815139626125,
+      "grad_norm": 3.794921875,
+      "learning_rate": 2.348899266177452e-05,
+      "loss": 1.1591,
+      "step": 11479
+    },
+    {
+      "epoch": 10.597738287560581,
+      "grad_norm": 3.22265625,
+      "learning_rate": 2.3482321547698468e-05,
+      "loss": 1.3896,
+      "step": 11480
+    },
+    {
+      "epoch": 10.598661435495037,
+      "grad_norm": 2.912109375,
+      "learning_rate": 2.3475650433622416e-05,
+      "loss": 1.1382,
+      "step": 11481
+    },
+    {
+      "epoch": 10.599584583429495,
+      "grad_norm": 3.794921875,
+      "learning_rate": 2.3468979319546365e-05,
+      "loss": 1.2642,
+      "step": 11482
+    },
+    {
+      "epoch": 10.600507731363951,
+      "grad_norm": 3.29296875,
+      "learning_rate": 2.3462308205470314e-05,
+      "loss": 1.1091,
+      "step": 11483
+    },
+    {
+      "epoch": 10.601430879298407,
+      "grad_norm": 3.23828125,
+      "learning_rate": 2.3455637091394263e-05,
+      "loss": 1.2108,
+      "step": 11484
+    },
+    {
+      "epoch": 10.602354027232863,
+      "grad_norm": 3.34765625,
+      "learning_rate": 2.344896597731821e-05,
+      "loss": 1.4725,
+      "step": 11485
+    },
+    {
+      "epoch": 10.603277175167321,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.344229486324216e-05,
+      "loss": 1.0951,
+      "step": 11486
+    },
+    {
+      "epoch": 10.604200323101777,
+      "grad_norm": 3.208984375,
+      "learning_rate": 2.343562374916611e-05,
+      "loss": 1.2739,
+      "step": 11487
+    },
+    {
+      "epoch": 10.605123471036233,
+      "grad_norm": 3.115234375,
+      "learning_rate": 2.342895263509006e-05,
+      "loss": 1.438,
+      "step": 11488
+    },
+    {
+      "epoch": 10.60604661897069,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.342228152101401e-05,
+      "loss": 1.3816,
+      "step": 11489
+    },
+    {
+      "epoch": 10.606969766905147,
+      "grad_norm": 3.38671875,
+      "learning_rate": 2.341561040693796e-05,
+      "loss": 1.2373,
+      "step": 11490
+    },
+    {
+      "epoch": 10.607892914839603,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.3408939292861908e-05,
+      "loss": 1.7314,
+      "step": 11491
+    },
+    {
+      "epoch": 10.608816062774059,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.3402268178785857e-05,
+      "loss": 1.2498,
+      "step": 11492
+    },
+    {
+      "epoch": 10.609739210708517,
+      "grad_norm": 3.5546875,
+      "learning_rate": 2.3395597064709806e-05,
+      "loss": 1.3818,
+      "step": 11493
+    },
+    {
+      "epoch": 10.610662358642973,
+      "grad_norm": 3.294921875,
+      "learning_rate": 2.3388925950633758e-05,
+      "loss": 1.3577,
+      "step": 11494
+    },
+    {
+      "epoch": 10.611585506577429,
+      "grad_norm": 2.951171875,
+      "learning_rate": 2.3382254836557707e-05,
+      "loss": 1.448,
+      "step": 11495
+    },
+    {
+      "epoch": 10.612508654511885,
+      "grad_norm": 3.61328125,
+      "learning_rate": 2.3375583722481655e-05,
+      "loss": 1.1665,
+      "step": 11496
+    },
+    {
+      "epoch": 10.613431802446343,
+      "grad_norm": 3.083984375,
+      "learning_rate": 2.3368912608405604e-05,
+      "loss": 0.7676,
+      "step": 11497
+    },
+    {
+      "epoch": 10.614354950380799,
+      "grad_norm": 3.384765625,
+      "learning_rate": 2.3362241494329553e-05,
+      "loss": 1.3691,
+      "step": 11498
+    },
+    {
+      "epoch": 10.615278098315255,
+      "grad_norm": 3.61328125,
+      "learning_rate": 2.3355570380253502e-05,
+      "loss": 1.1326,
+      "step": 11499
+    },
+    {
+      "epoch": 10.616201246249712,
+      "grad_norm": 3.177734375,
+      "learning_rate": 2.334889926617745e-05,
+      "loss": 1.4814,
+      "step": 11500
+    },
+    {
+      "epoch": 10.617124394184168,
+      "grad_norm": 3.16796875,
+      "learning_rate": 2.3342228152101403e-05,
+      "loss": 1.4446,
+      "step": 11501
+    },
+    {
+      "epoch": 10.618047542118624,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.3335557038025352e-05,
+      "loss": 1.2009,
+      "step": 11502
+    },
+    {
+      "epoch": 10.61897069005308,
+      "grad_norm": 3.07421875,
+      "learning_rate": 2.33288859239493e-05,
+      "loss": 1.3352,
+      "step": 11503
+    },
+    {
+      "epoch": 10.619893837987538,
+      "grad_norm": 3.57421875,
+      "learning_rate": 2.332221480987325e-05,
+      "loss": 1.2379,
+      "step": 11504
+    },
+    {
+      "epoch": 10.620816985921994,
+      "grad_norm": 3.595703125,
+      "learning_rate": 2.3315543695797198e-05,
+      "loss": 1.368,
+      "step": 11505
+    },
+    {
+      "epoch": 10.62174013385645,
+      "grad_norm": 3.931640625,
+      "learning_rate": 2.3308872581721147e-05,
+      "loss": 1.3208,
+      "step": 11506
+    },
+    {
+      "epoch": 10.622663281790906,
+      "grad_norm": 3.103515625,
+      "learning_rate": 2.33022014676451e-05,
+      "loss": 1.2714,
+      "step": 11507
+    },
+    {
+      "epoch": 10.623586429725364,
+      "grad_norm": 3.236328125,
+      "learning_rate": 2.3295530353569048e-05,
+      "loss": 1.3711,
+      "step": 11508
+    },
+    {
+      "epoch": 10.62450957765982,
+      "grad_norm": 3.107421875,
+      "learning_rate": 2.3288859239492997e-05,
+      "loss": 1.5481,
+      "step": 11509
+    },
+    {
+      "epoch": 10.625432725594276,
+      "grad_norm": 3.408203125,
+      "learning_rate": 2.3282188125416946e-05,
+      "loss": 1.261,
+      "step": 11510
+    },
+    {
+      "epoch": 10.626355873528732,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.3275517011340894e-05,
+      "loss": 1.2463,
+      "step": 11511
+    },
+    {
+      "epoch": 10.62727902146319,
+      "grad_norm": 3.1796875,
+      "learning_rate": 2.3268845897264843e-05,
+      "loss": 1.3394,
+      "step": 11512
+    },
+    {
+      "epoch": 10.628202169397646,
+      "grad_norm": 3.2265625,
+      "learning_rate": 2.3262174783188796e-05,
+      "loss": 1.2217,
+      "step": 11513
+    },
+    {
+      "epoch": 10.629125317332102,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.3255503669112744e-05,
+      "loss": 1.1372,
+      "step": 11514
+    },
+    {
+      "epoch": 10.630048465266558,
+      "grad_norm": 3.205078125,
+      "learning_rate": 2.3248832555036693e-05,
+      "loss": 1.4285,
+      "step": 11515
+    },
+    {
+      "epoch": 10.630971613201016,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.3242161440960642e-05,
+      "loss": 1.3333,
+      "step": 11516
+    },
+    {
+      "epoch": 10.631894761135472,
+      "grad_norm": 3.025390625,
+      "learning_rate": 2.323549032688459e-05,
+      "loss": 1.5247,
+      "step": 11517
+    },
+    {
+      "epoch": 10.632817909069928,
+      "grad_norm": 3.6484375,
+      "learning_rate": 2.322881921280854e-05,
+      "loss": 1.157,
+      "step": 11518
+    },
+    {
+      "epoch": 10.633741057004386,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.3222148098732492e-05,
+      "loss": 1.3232,
+      "step": 11519
+    },
+    {
+      "epoch": 10.634664204938842,
+      "grad_norm": 3.294921875,
+      "learning_rate": 2.321547698465644e-05,
+      "loss": 0.8589,
+      "step": 11520
+    },
+    {
+      "epoch": 10.635587352873298,
+      "grad_norm": 3.568359375,
+      "learning_rate": 2.320880587058039e-05,
+      "loss": 1.1436,
+      "step": 11521
+    },
+    {
+      "epoch": 10.636510500807754,
+      "grad_norm": 3.1484375,
+      "learning_rate": 2.3202134756504338e-05,
+      "loss": 1.3887,
+      "step": 11522
+    },
+    {
+      "epoch": 10.637433648742212,
+      "grad_norm": 2.943359375,
+      "learning_rate": 2.3195463642428287e-05,
+      "loss": 1.3455,
+      "step": 11523
+    },
+    {
+      "epoch": 10.638356796676668,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.3188792528352236e-05,
+      "loss": 1.314,
+      "step": 11524
+    },
+    {
+      "epoch": 10.639279944611124,
+      "grad_norm": 2.810546875,
+      "learning_rate": 2.3182121414276185e-05,
+      "loss": 0.8979,
+      "step": 11525
+    },
+    {
+      "epoch": 10.64020309254558,
+      "grad_norm": 3.162109375,
+      "learning_rate": 2.3175450300200137e-05,
+      "loss": 1.3767,
+      "step": 11526
+    },
+    {
+      "epoch": 10.641126240480038,
+      "grad_norm": 3.337890625,
+      "learning_rate": 2.3168779186124086e-05,
+      "loss": 1.2529,
+      "step": 11527
+    },
+    {
+      "epoch": 10.642049388414494,
+      "grad_norm": 3.408203125,
+      "learning_rate": 2.3162108072048035e-05,
+      "loss": 1.1614,
+      "step": 11528
+    },
+    {
+      "epoch": 10.64297253634895,
+      "grad_norm": 3.84765625,
+      "learning_rate": 2.3155436957971983e-05,
+      "loss": 1.161,
+      "step": 11529
+    },
+    {
+      "epoch": 10.643895684283406,
+      "grad_norm": 3.1796875,
+      "learning_rate": 2.3148765843895932e-05,
+      "loss": 1.5164,
+      "step": 11530
+    },
+    {
+      "epoch": 10.644818832217863,
+      "grad_norm": 2.833984375,
+      "learning_rate": 2.314209472981988e-05,
+      "loss": 1.3013,
+      "step": 11531
+    },
+    {
+      "epoch": 10.64574198015232,
+      "grad_norm": 3.2421875,
+      "learning_rate": 2.313542361574383e-05,
+      "loss": 1.3228,
+      "step": 11532
+    },
+    {
+      "epoch": 10.646665128086775,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.312875250166778e-05,
+      "loss": 1.0823,
+      "step": 11533
+    },
+    {
+      "epoch": 10.647588276021231,
+      "grad_norm": 3.060546875,
+      "learning_rate": 2.3122081387591727e-05,
+      "loss": 1.4453,
+      "step": 11534
+    },
+    {
+      "epoch": 10.64851142395569,
+      "grad_norm": 3.44140625,
+      "learning_rate": 2.311541027351568e-05,
+      "loss": 1.259,
+      "step": 11535
+    },
+    {
+      "epoch": 10.649434571890145,
+      "grad_norm": 3.2421875,
+      "learning_rate": 2.310873915943963e-05,
+      "loss": 1.4497,
+      "step": 11536
+    },
+    {
+      "epoch": 10.650357719824601,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.3102068045363577e-05,
+      "loss": 1.1755,
+      "step": 11537
+    },
+    {
+      "epoch": 10.65128086775906,
+      "grad_norm": 3.177734375,
+      "learning_rate": 2.3095396931287526e-05,
+      "loss": 1.3091,
+      "step": 11538
+    },
+    {
+      "epoch": 10.652204015693515,
+      "grad_norm": 3.025390625,
+      "learning_rate": 2.3088725817211475e-05,
+      "loss": 1.1827,
+      "step": 11539
+    },
+    {
+      "epoch": 10.653127163627971,
+      "grad_norm": 3.20703125,
+      "learning_rate": 2.3082054703135424e-05,
+      "loss": 1.2645,
+      "step": 11540
+    },
+    {
+      "epoch": 10.654050311562427,
+      "grad_norm": 3.232421875,
+      "learning_rate": 2.3075383589059373e-05,
+      "loss": 1.4158,
+      "step": 11541
+    },
+    {
+      "epoch": 10.654973459496885,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.306871247498332e-05,
+      "loss": 1.5005,
+      "step": 11542
+    },
+    {
+      "epoch": 10.655896607431341,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.306204136090727e-05,
+      "loss": 1.4832,
+      "step": 11543
+    },
+    {
+      "epoch": 10.656819755365797,
+      "grad_norm": 3.873046875,
+      "learning_rate": 2.3055370246831222e-05,
+      "loss": 1.4036,
+      "step": 11544
+    },
+    {
+      "epoch": 10.657742903300253,
+      "grad_norm": 3.236328125,
+      "learning_rate": 2.304869913275517e-05,
+      "loss": 1.2378,
+      "step": 11545
+    },
+    {
+      "epoch": 10.658666051234711,
+      "grad_norm": 2.806640625,
+      "learning_rate": 2.304202801867912e-05,
+      "loss": 1.2844,
+      "step": 11546
+    },
+    {
+      "epoch": 10.659589199169167,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.303535690460307e-05,
+      "loss": 1.1471,
+      "step": 11547
+    },
+    {
+      "epoch": 10.660512347103623,
+      "grad_norm": 2.990234375,
+      "learning_rate": 2.3028685790527018e-05,
+      "loss": 1.2483,
+      "step": 11548
+    },
+    {
+      "epoch": 10.66143549503808,
+      "grad_norm": 3.107421875,
+      "learning_rate": 2.3022014676450967e-05,
+      "loss": 1.2888,
+      "step": 11549
+    },
+    {
+      "epoch": 10.662358642972537,
+      "grad_norm": 2.98828125,
+      "learning_rate": 2.3015343562374915e-05,
+      "loss": 1.2244,
+      "step": 11550
+    },
+    {
+      "epoch": 10.663281790906993,
+      "grad_norm": 3.177734375,
+      "learning_rate": 2.3008672448298868e-05,
+      "loss": 1.1027,
+      "step": 11551
+    },
+    {
+      "epoch": 10.664204938841449,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.3002001334222816e-05,
+      "loss": 1.2063,
+      "step": 11552
+    },
+    {
+      "epoch": 10.665128086775907,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.2995330220146765e-05,
+      "loss": 1.4163,
+      "step": 11553
+    },
+    {
+      "epoch": 10.666051234710363,
+      "grad_norm": 3.54296875,
+      "learning_rate": 2.2988659106070714e-05,
+      "loss": 1.3059,
+      "step": 11554
+    },
+    {
+      "epoch": 10.666974382644819,
+      "grad_norm": 3.779296875,
+      "learning_rate": 2.2981987991994663e-05,
+      "loss": 1.2271,
+      "step": 11555
+    },
+    {
+      "epoch": 10.667897530579275,
+      "grad_norm": 3.396484375,
+      "learning_rate": 2.297531687791861e-05,
+      "loss": 0.9359,
+      "step": 11556
+    },
+    {
+      "epoch": 10.668820678513733,
+      "grad_norm": 2.9609375,
+      "learning_rate": 2.2968645763842564e-05,
+      "loss": 1.3032,
+      "step": 11557
+    },
+    {
+      "epoch": 10.669743826448189,
+      "grad_norm": 3.146484375,
+      "learning_rate": 2.2961974649766513e-05,
+      "loss": 1.061,
+      "step": 11558
+    },
+    {
+      "epoch": 10.670666974382645,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.295530353569046e-05,
+      "loss": 1.3042,
+      "step": 11559
+    },
+    {
+      "epoch": 10.6715901223171,
+      "grad_norm": 2.9921875,
+      "learning_rate": 2.294863242161441e-05,
+      "loss": 1.7078,
+      "step": 11560
+    },
+    {
+      "epoch": 10.672513270251558,
+      "grad_norm": 3.7734375,
+      "learning_rate": 2.294196130753836e-05,
+      "loss": 1.1543,
+      "step": 11561
+    },
+    {
+      "epoch": 10.673436418186014,
+      "grad_norm": 3.216796875,
+      "learning_rate": 2.2935290193462308e-05,
+      "loss": 1.2706,
+      "step": 11562
+    },
+    {
+      "epoch": 10.67435956612047,
+      "grad_norm": 3.197265625,
+      "learning_rate": 2.292861907938626e-05,
+      "loss": 1.1935,
+      "step": 11563
+    },
+    {
+      "epoch": 10.675282714054926,
+      "grad_norm": 3.0859375,
+      "learning_rate": 2.292194796531021e-05,
+      "loss": 1.2397,
+      "step": 11564
+    },
+    {
+      "epoch": 10.676205861989384,
+      "grad_norm": 3.1328125,
+      "learning_rate": 2.2915276851234158e-05,
+      "loss": 1.2847,
+      "step": 11565
+    },
+    {
+      "epoch": 10.67712900992384,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.2908605737158107e-05,
+      "loss": 1.4241,
+      "step": 11566
+    },
+    {
+      "epoch": 10.678052157858296,
+      "grad_norm": 3.69921875,
+      "learning_rate": 2.2901934623082055e-05,
+      "loss": 1.1553,
+      "step": 11567
+    },
+    {
+      "epoch": 10.678975305792754,
+      "grad_norm": 3.107421875,
+      "learning_rate": 2.2895263509006004e-05,
+      "loss": 1.1562,
+      "step": 11568
+    },
+    {
+      "epoch": 10.67989845372721,
+      "grad_norm": 3.08203125,
+      "learning_rate": 2.2888592394929956e-05,
+      "loss": 1.2025,
+      "step": 11569
+    },
+    {
+      "epoch": 10.680821601661666,
+      "grad_norm": 3.314453125,
+      "learning_rate": 2.2881921280853905e-05,
+      "loss": 1.3132,
+      "step": 11570
+    },
+    {
+      "epoch": 10.681744749596122,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.2875250166777854e-05,
+      "loss": 1.2206,
+      "step": 11571
+    },
+    {
+      "epoch": 10.68266789753058,
+      "grad_norm": 2.978515625,
+      "learning_rate": 2.2868579052701803e-05,
+      "loss": 1.2827,
+      "step": 11572
+    },
+    {
+      "epoch": 10.683591045465036,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.2861907938625752e-05,
+      "loss": 1.417,
+      "step": 11573
+    },
+    {
+      "epoch": 10.684514193399492,
+      "grad_norm": 3.5,
+      "learning_rate": 2.28552368245497e-05,
+      "loss": 1.0909,
+      "step": 11574
+    },
+    {
+      "epoch": 10.685437341333948,
+      "grad_norm": 3.7109375,
+      "learning_rate": 2.284856571047365e-05,
+      "loss": 1.328,
+      "step": 11575
+    },
+    {
+      "epoch": 10.686360489268406,
+      "grad_norm": 3.2890625,
+      "learning_rate": 2.28418945963976e-05,
+      "loss": 1.2803,
+      "step": 11576
+    },
+    {
+      "epoch": 10.687283637202862,
+      "grad_norm": 3.498046875,
+      "learning_rate": 2.283522348232155e-05,
+      "loss": 1.194,
+      "step": 11577
+    },
+    {
+      "epoch": 10.688206785137318,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.28285523682455e-05,
+      "loss": 1.4062,
+      "step": 11578
+    },
+    {
+      "epoch": 10.689129933071774,
+      "grad_norm": 3.306640625,
+      "learning_rate": 2.2821881254169448e-05,
+      "loss": 1.5193,
+      "step": 11579
+    },
+    {
+      "epoch": 10.690053081006232,
+      "grad_norm": 3.458984375,
+      "learning_rate": 2.2815210140093397e-05,
+      "loss": 1.3113,
+      "step": 11580
+    },
+    {
+      "epoch": 10.690976228940688,
+      "grad_norm": 3.17578125,
+      "learning_rate": 2.2808539026017346e-05,
+      "loss": 1.4424,
+      "step": 11581
+    },
+    {
+      "epoch": 10.691899376875144,
+      "grad_norm": 3.349609375,
+      "learning_rate": 2.2801867911941298e-05,
+      "loss": 1.1165,
+      "step": 11582
+    },
+    {
+      "epoch": 10.6928225248096,
+      "grad_norm": 3.08984375,
+      "learning_rate": 2.2795196797865247e-05,
+      "loss": 1.1108,
+      "step": 11583
+    },
+    {
+      "epoch": 10.693745672744058,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.2788525683789196e-05,
+      "loss": 1.4131,
+      "step": 11584
+    },
+    {
+      "epoch": 10.694668820678514,
+      "grad_norm": 3.197265625,
+      "learning_rate": 2.2781854569713144e-05,
+      "loss": 1.3237,
+      "step": 11585
+    },
+    {
+      "epoch": 10.69559196861297,
+      "grad_norm": 3.451171875,
+      "learning_rate": 2.2775183455637093e-05,
+      "loss": 1.5432,
+      "step": 11586
+    },
+    {
+      "epoch": 10.696515116547427,
+      "grad_norm": 3.365234375,
+      "learning_rate": 2.2768512341561042e-05,
+      "loss": 1.1552,
+      "step": 11587
+    },
+    {
+      "epoch": 10.697438264481884,
+      "grad_norm": 3.228515625,
+      "learning_rate": 2.276184122748499e-05,
+      "loss": 1.1954,
+      "step": 11588
+    },
+    {
+      "epoch": 10.69836141241634,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.275517011340894e-05,
+      "loss": 1.1609,
+      "step": 11589
+    },
+    {
+      "epoch": 10.699284560350796,
+      "grad_norm": 4.1640625,
+      "learning_rate": 2.274849899933289e-05,
+      "loss": 1.3213,
+      "step": 11590
+    },
+    {
+      "epoch": 10.700207708285253,
+      "grad_norm": 3.39453125,
+      "learning_rate": 2.2741827885256837e-05,
+      "loss": 1.2183,
+      "step": 11591
+    },
+    {
+      "epoch": 10.70113085621971,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.273515677118079e-05,
+      "loss": 1.2056,
+      "step": 11592
+    },
+    {
+      "epoch": 10.702054004154165,
+      "grad_norm": 3.416015625,
+      "learning_rate": 2.2728485657104738e-05,
+      "loss": 1.1477,
+      "step": 11593
+    },
+    {
+      "epoch": 10.702977152088621,
+      "grad_norm": 3.052734375,
+      "learning_rate": 2.2721814543028687e-05,
+      "loss": 1.1295,
+      "step": 11594
+    },
+    {
+      "epoch": 10.70390030002308,
+      "grad_norm": 3.201171875,
+      "learning_rate": 2.2715143428952636e-05,
+      "loss": 1.4661,
+      "step": 11595
+    },
+    {
+      "epoch": 10.704823447957535,
+      "grad_norm": 3.154296875,
+      "learning_rate": 2.2708472314876585e-05,
+      "loss": 1.3064,
+      "step": 11596
+    },
+    {
+      "epoch": 10.705746595891991,
+      "grad_norm": 3.2421875,
+      "learning_rate": 2.2701801200800534e-05,
+      "loss": 1.5603,
+      "step": 11597
+    },
+    {
+      "epoch": 10.706669743826449,
+      "grad_norm": 3.439453125,
+      "learning_rate": 2.2695130086724482e-05,
+      "loss": 1.4937,
+      "step": 11598
+    },
+    {
+      "epoch": 10.707592891760905,
+      "grad_norm": 3.400390625,
+      "learning_rate": 2.268845897264843e-05,
+      "loss": 1.3029,
+      "step": 11599
+    },
+    {
+      "epoch": 10.708516039695361,
+      "grad_norm": 3.29296875,
+      "learning_rate": 2.268178785857238e-05,
+      "loss": 1.199,
+      "step": 11600
+    },
+    {
+      "epoch": 10.709439187629817,
+      "grad_norm": 3.25390625,
+      "learning_rate": 2.2675116744496332e-05,
+      "loss": 1.2815,
+      "step": 11601
+    },
+    {
+      "epoch": 10.710362335564275,
+      "grad_norm": 3.0546875,
+      "learning_rate": 2.266844563042028e-05,
+      "loss": 1.3101,
+      "step": 11602
+    },
+    {
+      "epoch": 10.711285483498731,
+      "grad_norm": 3.107421875,
+      "learning_rate": 2.266177451634423e-05,
+      "loss": 1.2563,
+      "step": 11603
+    },
+    {
+      "epoch": 10.712208631433187,
+      "grad_norm": 3.83984375,
+      "learning_rate": 2.265510340226818e-05,
+      "loss": 1.0583,
+      "step": 11604
+    },
+    {
+      "epoch": 10.713131779367643,
+      "grad_norm": 3.318359375,
+      "learning_rate": 2.2648432288192127e-05,
+      "loss": 1.3884,
+      "step": 11605
+    },
+    {
+      "epoch": 10.7140549273021,
+      "grad_norm": 3.521484375,
+      "learning_rate": 2.2641761174116076e-05,
+      "loss": 1.0859,
+      "step": 11606
+    },
+    {
+      "epoch": 10.714978075236557,
+      "grad_norm": 2.951171875,
+      "learning_rate": 2.263509006004003e-05,
+      "loss": 1.365,
+      "step": 11607
+    },
+    {
+      "epoch": 10.715901223171013,
+      "grad_norm": 3.439453125,
+      "learning_rate": 2.2628418945963977e-05,
+      "loss": 1.5293,
+      "step": 11608
+    },
+    {
+      "epoch": 10.716824371105469,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.2621747831887926e-05,
+      "loss": 1.2019,
+      "step": 11609
+    },
+    {
+      "epoch": 10.717747519039927,
+      "grad_norm": 3.3984375,
+      "learning_rate": 2.2615076717811875e-05,
+      "loss": 1.4248,
+      "step": 11610
+    },
+    {
+      "epoch": 10.718670666974383,
+      "grad_norm": 3.486328125,
+      "learning_rate": 2.2608405603735824e-05,
+      "loss": 1.257,
+      "step": 11611
+    },
+    {
+      "epoch": 10.719593814908839,
+      "grad_norm": 3.994140625,
+      "learning_rate": 2.2601734489659773e-05,
+      "loss": 1.397,
+      "step": 11612
+    },
+    {
+      "epoch": 10.720516962843295,
+      "grad_norm": 2.9609375,
+      "learning_rate": 2.2595063375583725e-05,
+      "loss": 1.2601,
+      "step": 11613
+    },
+    {
+      "epoch": 10.721440110777753,
+      "grad_norm": 2.873046875,
+      "learning_rate": 2.2588392261507674e-05,
+      "loss": 1.2582,
+      "step": 11614
+    },
+    {
+      "epoch": 10.722363258712209,
+      "grad_norm": 3.44140625,
+      "learning_rate": 2.2581721147431622e-05,
+      "loss": 1.4512,
+      "step": 11615
+    },
+    {
+      "epoch": 10.723286406646665,
+      "grad_norm": 3.142578125,
+      "learning_rate": 2.257505003335557e-05,
+      "loss": 1.3701,
+      "step": 11616
+    },
+    {
+      "epoch": 10.724209554581122,
+      "grad_norm": 3.1484375,
+      "learning_rate": 2.256837891927952e-05,
+      "loss": 1.1605,
+      "step": 11617
+    },
+    {
+      "epoch": 10.725132702515578,
+      "grad_norm": 3.041015625,
+      "learning_rate": 2.256170780520347e-05,
+      "loss": 1.436,
+      "step": 11618
+    },
+    {
+      "epoch": 10.726055850450035,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.2555036691127418e-05,
+      "loss": 1.3372,
+      "step": 11619
+    },
+    {
+      "epoch": 10.72697899838449,
+      "grad_norm": 3.408203125,
+      "learning_rate": 2.254836557705137e-05,
+      "loss": 1.387,
+      "step": 11620
+    },
+    {
+      "epoch": 10.727902146318948,
+      "grad_norm": 3.29296875,
+      "learning_rate": 2.254169446297532e-05,
+      "loss": 1.1211,
+      "step": 11621
+    },
+    {
+      "epoch": 10.728825294253404,
+      "grad_norm": 3.2265625,
+      "learning_rate": 2.2535023348899268e-05,
+      "loss": 1.1131,
+      "step": 11622
+    },
+    {
+      "epoch": 10.72974844218786,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.2528352234823216e-05,
+      "loss": 1.3257,
+      "step": 11623
+    },
+    {
+      "epoch": 10.730671590122316,
+      "grad_norm": 3.5390625,
+      "learning_rate": 2.2521681120747165e-05,
+      "loss": 1.2595,
+      "step": 11624
+    },
+    {
+      "epoch": 10.731594738056774,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.2515010006671114e-05,
+      "loss": 1.0894,
+      "step": 11625
+    },
+    {
+      "epoch": 10.73251788599123,
+      "grad_norm": 3.24609375,
+      "learning_rate": 2.2508338892595066e-05,
+      "loss": 1.1411,
+      "step": 11626
+    },
+    {
+      "epoch": 10.733441033925686,
+      "grad_norm": 3.083984375,
+      "learning_rate": 2.2501667778519015e-05,
+      "loss": 0.8615,
+      "step": 11627
+    },
+    {
+      "epoch": 10.734364181860142,
+      "grad_norm": 3.1796875,
+      "learning_rate": 2.2494996664442964e-05,
+      "loss": 1.118,
+      "step": 11628
+    },
+    {
+      "epoch": 10.7352873297946,
+      "grad_norm": 3.591796875,
+      "learning_rate": 2.2488325550366913e-05,
+      "loss": 1.0752,
+      "step": 11629
+    },
+    {
+      "epoch": 10.736210477729056,
+      "grad_norm": 3.37890625,
+      "learning_rate": 2.248165443629086e-05,
+      "loss": 1.4729,
+      "step": 11630
+    },
+    {
+      "epoch": 10.737133625663512,
+      "grad_norm": 2.880859375,
+      "learning_rate": 2.247498332221481e-05,
+      "loss": 1.0686,
+      "step": 11631
+    },
+    {
+      "epoch": 10.738056773597968,
+      "grad_norm": 3.6328125,
+      "learning_rate": 2.2468312208138763e-05,
+      "loss": 1.3452,
+      "step": 11632
+    },
+    {
+      "epoch": 10.738979921532426,
+      "grad_norm": 3.451171875,
+      "learning_rate": 2.246164109406271e-05,
+      "loss": 1.4297,
+      "step": 11633
+    },
+    {
+      "epoch": 10.739903069466882,
+      "grad_norm": 3.431640625,
+      "learning_rate": 2.245496997998666e-05,
+      "loss": 1.1802,
+      "step": 11634
+    },
+    {
+      "epoch": 10.740826217401338,
+      "grad_norm": 3.337890625,
+      "learning_rate": 2.244829886591061e-05,
+      "loss": 1.3097,
+      "step": 11635
+    },
+    {
+      "epoch": 10.741749365335796,
+      "grad_norm": 3.005859375,
+      "learning_rate": 2.2441627751834558e-05,
+      "loss": 1.0526,
+      "step": 11636
+    },
+    {
+      "epoch": 10.742672513270252,
+      "grad_norm": 2.97265625,
+      "learning_rate": 2.2434956637758507e-05,
+      "loss": 1.3315,
+      "step": 11637
+    },
+    {
+      "epoch": 10.743595661204708,
+      "grad_norm": 3.193359375,
+      "learning_rate": 2.242828552368246e-05,
+      "loss": 1.0355,
+      "step": 11638
+    },
+    {
+      "epoch": 10.744518809139164,
+      "grad_norm": 3.01171875,
+      "learning_rate": 2.2421614409606408e-05,
+      "loss": 0.9027,
+      "step": 11639
+    },
+    {
+      "epoch": 10.745441957073622,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.2414943295530356e-05,
+      "loss": 1.2698,
+      "step": 11640
+    },
+    {
+      "epoch": 10.746365105008078,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.2408272181454305e-05,
+      "loss": 1.1738,
+      "step": 11641
+    },
+    {
+      "epoch": 10.747288252942534,
+      "grad_norm": 3.341796875,
+      "learning_rate": 2.2401601067378254e-05,
+      "loss": 1.4351,
+      "step": 11642
+    },
+    {
+      "epoch": 10.74821140087699,
+      "grad_norm": 3.333984375,
+      "learning_rate": 2.2394929953302203e-05,
+      "loss": 1.4771,
+      "step": 11643
+    },
+    {
+      "epoch": 10.749134548811448,
+      "grad_norm": 3.49609375,
+      "learning_rate": 2.2388258839226152e-05,
+      "loss": 1.2656,
+      "step": 11644
+    },
+    {
+      "epoch": 10.750057696745904,
+      "grad_norm": 3.03515625,
+      "learning_rate": 2.23815877251501e-05,
+      "loss": 1.4119,
+      "step": 11645
+    },
+    {
+      "epoch": 10.75098084468036,
+      "grad_norm": 2.892578125,
+      "learning_rate": 2.237491661107405e-05,
+      "loss": 1.1219,
+      "step": 11646
+    },
+    {
+      "epoch": 10.751903992614817,
+      "grad_norm": 3.3046875,
+      "learning_rate": 2.2368245496997998e-05,
+      "loss": 1.3918,
+      "step": 11647
+    },
+    {
+      "epoch": 10.752827140549273,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.236157438292195e-05,
+      "loss": 1.397,
+      "step": 11648
+    },
+    {
+      "epoch": 10.75375028848373,
+      "grad_norm": 3.197265625,
+      "learning_rate": 2.23549032688459e-05,
+      "loss": 1.2583,
+      "step": 11649
+    },
+    {
+      "epoch": 10.754673436418186,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.2348232154769848e-05,
+      "loss": 1.2174,
+      "step": 11650
+    },
+    {
+      "epoch": 10.755596584352643,
+      "grad_norm": 3.232421875,
+      "learning_rate": 2.2341561040693797e-05,
+      "loss": 1.3142,
+      "step": 11651
+    },
+    {
+      "epoch": 10.7565197322871,
+      "grad_norm": 3.482421875,
+      "learning_rate": 2.2334889926617746e-05,
+      "loss": 1.3014,
+      "step": 11652
+    },
+    {
+      "epoch": 10.757442880221555,
+      "grad_norm": 3.412109375,
+      "learning_rate": 2.2328218812541695e-05,
+      "loss": 1.4075,
+      "step": 11653
+    },
+    {
+      "epoch": 10.758366028156011,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.2321547698465643e-05,
+      "loss": 1.3962,
+      "step": 11654
+    },
+    {
+      "epoch": 10.75928917609047,
+      "grad_norm": 3.486328125,
+      "learning_rate": 2.2314876584389592e-05,
+      "loss": 1.1694,
+      "step": 11655
+    },
+    {
+      "epoch": 10.760212324024925,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.230820547031354e-05,
+      "loss": 1.2581,
+      "step": 11656
+    },
+    {
+      "epoch": 10.761135471959381,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.2301534356237493e-05,
+      "loss": 1.2183,
+      "step": 11657
+    },
+    {
+      "epoch": 10.762058619893837,
+      "grad_norm": 2.8203125,
+      "learning_rate": 2.2294863242161442e-05,
+      "loss": 1.1251,
+      "step": 11658
+    },
+    {
+      "epoch": 10.762981767828295,
+      "grad_norm": 3.2578125,
+      "learning_rate": 2.228819212808539e-05,
+      "loss": 1.1077,
+      "step": 11659
+    },
+    {
+      "epoch": 10.763904915762751,
+      "grad_norm": 3.337890625,
+      "learning_rate": 2.228152101400934e-05,
+      "loss": 1.4226,
+      "step": 11660
+    },
+    {
+      "epoch": 10.764828063697207,
+      "grad_norm": 3.375,
+      "learning_rate": 2.227484989993329e-05,
+      "loss": 1.141,
+      "step": 11661
+    },
+    {
+      "epoch": 10.765751211631663,
+      "grad_norm": 3.05078125,
+      "learning_rate": 2.2268178785857237e-05,
+      "loss": 1.0342,
+      "step": 11662
+    },
+    {
+      "epoch": 10.766674359566121,
+      "grad_norm": 3.271484375,
+      "learning_rate": 2.226150767178119e-05,
+      "loss": 1.6582,
+      "step": 11663
+    },
+    {
+      "epoch": 10.767597507500577,
+      "grad_norm": 3.337890625,
+      "learning_rate": 2.2254836557705138e-05,
+      "loss": 1.1446,
+      "step": 11664
+    },
+    {
+      "epoch": 10.768520655435033,
+      "grad_norm": 2.67578125,
+      "learning_rate": 2.2248165443629087e-05,
+      "loss": 1.4653,
+      "step": 11665
+    },
+    {
+      "epoch": 10.76944380336949,
+      "grad_norm": 3.427734375,
+      "learning_rate": 2.2241494329553036e-05,
+      "loss": 1.1677,
+      "step": 11666
+    },
+    {
+      "epoch": 10.770366951303947,
+      "grad_norm": 3.466796875,
+      "learning_rate": 2.2234823215476985e-05,
+      "loss": 1.2363,
+      "step": 11667
+    },
+    {
+      "epoch": 10.771290099238403,
+      "grad_norm": 3.462890625,
+      "learning_rate": 2.2228152101400934e-05,
+      "loss": 1.1678,
+      "step": 11668
+    },
+    {
+      "epoch": 10.772213247172859,
+      "grad_norm": 3.3359375,
+      "learning_rate": 2.2221480987324882e-05,
+      "loss": 1.0583,
+      "step": 11669
+    },
+    {
+      "epoch": 10.773136395107317,
+      "grad_norm": 3.2109375,
+      "learning_rate": 2.2214809873248835e-05,
+      "loss": 1.3555,
+      "step": 11670
+    },
+    {
+      "epoch": 10.774059543041773,
+      "grad_norm": 3.0,
+      "learning_rate": 2.2208138759172783e-05,
+      "loss": 1.5215,
+      "step": 11671
+    },
+    {
+      "epoch": 10.774982690976229,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.2201467645096732e-05,
+      "loss": 1.1952,
+      "step": 11672
+    },
+    {
+      "epoch": 10.775905838910685,
+      "grad_norm": 3.1953125,
+      "learning_rate": 2.219479653102068e-05,
+      "loss": 1.2888,
+      "step": 11673
+    },
+    {
+      "epoch": 10.776828986845143,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.218812541694463e-05,
+      "loss": 1.3916,
+      "step": 11674
+    },
+    {
+      "epoch": 10.777752134779599,
+      "grad_norm": 3.2265625,
+      "learning_rate": 2.218145430286858e-05,
+      "loss": 1.2941,
+      "step": 11675
+    },
+    {
+      "epoch": 10.778675282714055,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.217478318879253e-05,
+      "loss": 1.2928,
+      "step": 11676
+    },
+    {
+      "epoch": 10.77959843064851,
+      "grad_norm": 3.693359375,
+      "learning_rate": 2.216811207471648e-05,
+      "loss": 1.4758,
+      "step": 11677
+    },
+    {
+      "epoch": 10.780521578582968,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.216144096064043e-05,
+      "loss": 1.3372,
+      "step": 11678
+    },
+    {
+      "epoch": 10.781444726517424,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.2154769846564377e-05,
+      "loss": 1.0721,
+      "step": 11679
+    },
+    {
+      "epoch": 10.78236787445188,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.2148098732488326e-05,
+      "loss": 1.4707,
+      "step": 11680
+    },
+    {
+      "epoch": 10.783291022386337,
+      "grad_norm": 3.248046875,
+      "learning_rate": 2.2141427618412275e-05,
+      "loss": 1.2747,
+      "step": 11681
+    },
+    {
+      "epoch": 10.784214170320794,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.2134756504336227e-05,
+      "loss": 1.2573,
+      "step": 11682
+    },
+    {
+      "epoch": 10.78513731825525,
+      "grad_norm": 3.521484375,
+      "learning_rate": 2.2128085390260176e-05,
+      "loss": 1.2324,
+      "step": 11683
+    },
+    {
+      "epoch": 10.786060466189706,
+      "grad_norm": 3.361328125,
+      "learning_rate": 2.2121414276184125e-05,
+      "loss": 1.3904,
+      "step": 11684
+    },
+    {
+      "epoch": 10.786983614124164,
+      "grad_norm": 3.9140625,
+      "learning_rate": 2.2114743162108074e-05,
+      "loss": 1.4924,
+      "step": 11685
+    },
+    {
+      "epoch": 10.78790676205862,
+      "grad_norm": 3.1796875,
+      "learning_rate": 2.2108072048032022e-05,
+      "loss": 1.2407,
+      "step": 11686
+    },
+    {
+      "epoch": 10.788829909993076,
+      "grad_norm": 3.728515625,
+      "learning_rate": 2.210140093395597e-05,
+      "loss": 1.2968,
+      "step": 11687
+    },
+    {
+      "epoch": 10.789753057927532,
+      "grad_norm": 3.755859375,
+      "learning_rate": 2.209472981987992e-05,
+      "loss": 1.2181,
+      "step": 11688
+    },
+    {
+      "epoch": 10.79067620586199,
+      "grad_norm": 3.19921875,
+      "learning_rate": 2.2088058705803872e-05,
+      "loss": 1.3843,
+      "step": 11689
+    },
+    {
+      "epoch": 10.791599353796446,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.208138759172782e-05,
+      "loss": 1.2717,
+      "step": 11690
+    },
+    {
+      "epoch": 10.792522501730902,
+      "grad_norm": 3.01953125,
+      "learning_rate": 2.207471647765177e-05,
+      "loss": 1.3591,
+      "step": 11691
+    },
+    {
+      "epoch": 10.793445649665358,
+      "grad_norm": 3.064453125,
+      "learning_rate": 2.206804536357572e-05,
+      "loss": 1.1954,
+      "step": 11692
+    },
+    {
+      "epoch": 10.794368797599816,
+      "grad_norm": 3.599609375,
+      "learning_rate": 2.2061374249499668e-05,
+      "loss": 1.205,
+      "step": 11693
+    },
+    {
+      "epoch": 10.795291945534272,
+      "grad_norm": 3.166015625,
+      "learning_rate": 2.2054703135423616e-05,
+      "loss": 1.2751,
+      "step": 11694
+    },
+    {
+      "epoch": 10.796215093468728,
+      "grad_norm": 2.677734375,
+      "learning_rate": 2.204803202134757e-05,
+      "loss": 1.3633,
+      "step": 11695
+    },
+    {
+      "epoch": 10.797138241403186,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.2041360907271517e-05,
+      "loss": 1.2168,
+      "step": 11696
+    },
+    {
+      "epoch": 10.798061389337642,
+      "grad_norm": 5.48046875,
+      "learning_rate": 2.2034689793195466e-05,
+      "loss": 1.2665,
+      "step": 11697
+    },
+    {
+      "epoch": 10.798984537272098,
+      "grad_norm": 3.125,
+      "learning_rate": 2.2028018679119415e-05,
+      "loss": 1.3542,
+      "step": 11698
+    },
+    {
+      "epoch": 10.799907685206554,
+      "grad_norm": 3.087890625,
+      "learning_rate": 2.2021347565043364e-05,
+      "loss": 1.196,
+      "step": 11699
+    },
+    {
+      "epoch": 10.800830833141012,
+      "grad_norm": 3.099609375,
+      "learning_rate": 2.2014676450967313e-05,
+      "loss": 1.3132,
+      "step": 11700
+    },
+    {
+      "epoch": 10.801753981075468,
+      "grad_norm": 2.9921875,
+      "learning_rate": 2.200800533689126e-05,
+      "loss": 1.0165,
+      "step": 11701
+    },
+    {
+      "epoch": 10.802677129009924,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.200133422281521e-05,
+      "loss": 1.2653,
+      "step": 11702
+    },
+    {
+      "epoch": 10.80360027694438,
+      "grad_norm": 3.470703125,
+      "learning_rate": 2.199466310873916e-05,
+      "loss": 1.396,
+      "step": 11703
+    },
+    {
+      "epoch": 10.804523424878838,
+      "grad_norm": 3.10546875,
+      "learning_rate": 2.1987991994663108e-05,
+      "loss": 1.2533,
+      "step": 11704
+    },
+    {
+      "epoch": 10.805446572813294,
+      "grad_norm": 3.576171875,
+      "learning_rate": 2.198132088058706e-05,
+      "loss": 0.9552,
+      "step": 11705
+    },
+    {
+      "epoch": 10.80636972074775,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.197464976651101e-05,
+      "loss": 1.1423,
+      "step": 11706
+    },
+    {
+      "epoch": 10.807292868682206,
+      "grad_norm": 3.31640625,
+      "learning_rate": 2.1967978652434958e-05,
+      "loss": 1.2417,
+      "step": 11707
+    },
+    {
+      "epoch": 10.808216016616663,
+      "grad_norm": 3.361328125,
+      "learning_rate": 2.1961307538358907e-05,
+      "loss": 1.6528,
+      "step": 11708
+    },
+    {
+      "epoch": 10.80913916455112,
+      "grad_norm": 4.01953125,
+      "learning_rate": 2.1954636424282855e-05,
+      "loss": 1.3892,
+      "step": 11709
+    },
+    {
+      "epoch": 10.810062312485575,
+      "grad_norm": 2.763671875,
+      "learning_rate": 2.1947965310206804e-05,
+      "loss": 0.9917,
+      "step": 11710
+    },
+    {
+      "epoch": 10.810985460420032,
+      "grad_norm": 3.291015625,
+      "learning_rate": 2.1941294196130753e-05,
+      "loss": 1.3557,
+      "step": 11711
+    },
+    {
+      "epoch": 10.81190860835449,
+      "grad_norm": 3.60546875,
+      "learning_rate": 2.1934623082054702e-05,
+      "loss": 1.2102,
+      "step": 11712
+    },
+    {
+      "epoch": 10.812831756288945,
+      "grad_norm": 3.396484375,
+      "learning_rate": 2.192795196797865e-05,
+      "loss": 1.3137,
+      "step": 11713
+    },
+    {
+      "epoch": 10.813754904223401,
+      "grad_norm": 3.103515625,
+      "learning_rate": 2.1921280853902603e-05,
+      "loss": 0.9944,
+      "step": 11714
+    },
+    {
+      "epoch": 10.81467805215786,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.1914609739826552e-05,
+      "loss": 1.261,
+      "step": 11715
+    },
+    {
+      "epoch": 10.815601200092315,
+      "grad_norm": 3.283203125,
+      "learning_rate": 2.19079386257505e-05,
+      "loss": 1.2083,
+      "step": 11716
+    },
+    {
+      "epoch": 10.816524348026771,
+      "grad_norm": 3.19140625,
+      "learning_rate": 2.190126751167445e-05,
+      "loss": 1.3572,
+      "step": 11717
+    },
+    {
+      "epoch": 10.817447495961227,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.1894596397598398e-05,
+      "loss": 1.3936,
+      "step": 11718
+    },
+    {
+      "epoch": 10.818370643895685,
+      "grad_norm": 3.244140625,
+      "learning_rate": 2.1887925283522347e-05,
+      "loss": 1.2139,
+      "step": 11719
+    },
+    {
+      "epoch": 10.819293791830141,
+      "grad_norm": 3.42578125,
+      "learning_rate": 2.18812541694463e-05,
+      "loss": 1.2644,
+      "step": 11720
+    },
+    {
+      "epoch": 10.820216939764597,
+      "grad_norm": 2.98046875,
+      "learning_rate": 2.1874583055370248e-05,
+      "loss": 1.0603,
+      "step": 11721
+    },
+    {
+      "epoch": 10.821140087699053,
+      "grad_norm": 3.314453125,
+      "learning_rate": 2.1867911941294197e-05,
+      "loss": 1.2263,
+      "step": 11722
+    },
+    {
+      "epoch": 10.822063235633511,
+      "grad_norm": 3.314453125,
+      "learning_rate": 2.1861240827218146e-05,
+      "loss": 1.3154,
+      "step": 11723
+    },
+    {
+      "epoch": 10.822986383567967,
+      "grad_norm": 2.93359375,
+      "learning_rate": 2.1854569713142095e-05,
+      "loss": 1.0146,
+      "step": 11724
+    },
+    {
+      "epoch": 10.823909531502423,
+      "grad_norm": 3.123046875,
+      "learning_rate": 2.1847898599066043e-05,
+      "loss": 0.9249,
+      "step": 11725
+    },
+    {
+      "epoch": 10.824832679436879,
+      "grad_norm": 3.146484375,
+      "learning_rate": 2.1841227484989996e-05,
+      "loss": 1.2085,
+      "step": 11726
+    },
+    {
+      "epoch": 10.825755827371337,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.1834556370913944e-05,
+      "loss": 1.3013,
+      "step": 11727
+    },
+    {
+      "epoch": 10.826678975305793,
+      "grad_norm": 3.611328125,
+      "learning_rate": 2.1827885256837893e-05,
+      "loss": 1.0271,
+      "step": 11728
+    },
+    {
+      "epoch": 10.827602123240249,
+      "grad_norm": 3.408203125,
+      "learning_rate": 2.1821214142761842e-05,
+      "loss": 1.3943,
+      "step": 11729
+    },
+    {
+      "epoch": 10.828525271174705,
+      "grad_norm": 3.46484375,
+      "learning_rate": 2.181454302868579e-05,
+      "loss": 1.4121,
+      "step": 11730
+    },
+    {
+      "epoch": 10.829448419109163,
+      "grad_norm": 3.20703125,
+      "learning_rate": 2.180787191460974e-05,
+      "loss": 1.489,
+      "step": 11731
+    },
+    {
+      "epoch": 10.830371567043619,
+      "grad_norm": 3.3515625,
+      "learning_rate": 2.1801200800533692e-05,
+      "loss": 1.6382,
+      "step": 11732
+    },
+    {
+      "epoch": 10.831294714978075,
+      "grad_norm": 3.337890625,
+      "learning_rate": 2.179452968645764e-05,
+      "loss": 1.3911,
+      "step": 11733
+    },
+    {
+      "epoch": 10.832217862912533,
+      "grad_norm": 3.150390625,
+      "learning_rate": 2.178785857238159e-05,
+      "loss": 1.1897,
+      "step": 11734
+    },
+    {
+      "epoch": 10.833141010846989,
+      "grad_norm": 3.4921875,
+      "learning_rate": 2.1781187458305538e-05,
+      "loss": 1.3311,
+      "step": 11735
+    },
+    {
+      "epoch": 10.834064158781445,
+      "grad_norm": 2.833984375,
+      "learning_rate": 2.1774516344229487e-05,
+      "loss": 1.2915,
+      "step": 11736
+    },
+    {
+      "epoch": 10.8349873067159,
+      "grad_norm": 3.13671875,
+      "learning_rate": 2.1767845230153436e-05,
+      "loss": 1.3459,
+      "step": 11737
+    },
+    {
+      "epoch": 10.835910454650358,
+      "grad_norm": 2.79296875,
+      "learning_rate": 2.1761174116077385e-05,
+      "loss": 1.2998,
+      "step": 11738
+    },
+    {
+      "epoch": 10.836833602584814,
+      "grad_norm": 3.11328125,
+      "learning_rate": 2.1754503002001337e-05,
+      "loss": 1.1694,
+      "step": 11739
+    },
+    {
+      "epoch": 10.83775675051927,
+      "grad_norm": 3.26171875,
+      "learning_rate": 2.1747831887925286e-05,
+      "loss": 1.4111,
+      "step": 11740
+    },
+    {
+      "epoch": 10.838679898453726,
+      "grad_norm": 3.01953125,
+      "learning_rate": 2.1741160773849235e-05,
+      "loss": 1.1941,
+      "step": 11741
+    },
+    {
+      "epoch": 10.839603046388184,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.1734489659773183e-05,
+      "loss": 1.3555,
+      "step": 11742
+    },
+    {
+      "epoch": 10.84052619432264,
+      "grad_norm": 3.302734375,
+      "learning_rate": 2.1727818545697132e-05,
+      "loss": 1.4297,
+      "step": 11743
+    },
+    {
+      "epoch": 10.841449342257096,
+      "grad_norm": 3.5078125,
+      "learning_rate": 2.172114743162108e-05,
+      "loss": 1.6182,
+      "step": 11744
+    },
+    {
+      "epoch": 10.842372490191554,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.1714476317545033e-05,
+      "loss": 1.3887,
+      "step": 11745
+    },
+    {
+      "epoch": 10.84329563812601,
+      "grad_norm": 3.267578125,
+      "learning_rate": 2.1707805203468982e-05,
+      "loss": 1.2859,
+      "step": 11746
+    },
+    {
+      "epoch": 10.844218786060466,
+      "grad_norm": 3.24609375,
+      "learning_rate": 2.170113408939293e-05,
+      "loss": 1.2137,
+      "step": 11747
+    },
+    {
+      "epoch": 10.845141933994922,
+      "grad_norm": 3.322265625,
+      "learning_rate": 2.169446297531688e-05,
+      "loss": 1.3787,
+      "step": 11748
+    },
+    {
+      "epoch": 10.84606508192938,
+      "grad_norm": 3.212890625,
+      "learning_rate": 2.168779186124083e-05,
+      "loss": 1.194,
+      "step": 11749
+    },
+    {
+      "epoch": 10.846988229863836,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.1681120747164777e-05,
+      "loss": 1.2834,
+      "step": 11750
+    },
+    {
+      "epoch": 10.847911377798292,
+      "grad_norm": 2.8046875,
+      "learning_rate": 2.167444963308873e-05,
+      "loss": 1.3643,
+      "step": 11751
+    },
+    {
+      "epoch": 10.848834525732748,
+      "grad_norm": 3.056640625,
+      "learning_rate": 2.166777851901268e-05,
+      "loss": 1.0068,
+      "step": 11752
+    },
+    {
+      "epoch": 10.849757673667206,
+      "grad_norm": 3.2578125,
+      "learning_rate": 2.1661107404936627e-05,
+      "loss": 1.2343,
+      "step": 11753
+    },
+    {
+      "epoch": 10.850680821601662,
+      "grad_norm": 3.85546875,
+      "learning_rate": 2.1654436290860576e-05,
+      "loss": 1.3921,
+      "step": 11754
+    },
+    {
+      "epoch": 10.851603969536118,
+      "grad_norm": 3.048828125,
+      "learning_rate": 2.1647765176784525e-05,
+      "loss": 1.3774,
+      "step": 11755
+    },
+    {
+      "epoch": 10.852527117470574,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.1641094062708474e-05,
+      "loss": 1.5535,
+      "step": 11756
+    },
+    {
+      "epoch": 10.853450265405032,
+      "grad_norm": 3.17578125,
+      "learning_rate": 2.1634422948632422e-05,
+      "loss": 1.2819,
+      "step": 11757
+    },
+    {
+      "epoch": 10.854373413339488,
+      "grad_norm": 2.958984375,
+      "learning_rate": 2.162775183455637e-05,
+      "loss": 1.2561,
+      "step": 11758
+    },
+    {
+      "epoch": 10.855296561273944,
+      "grad_norm": 3.052734375,
+      "learning_rate": 2.162108072048032e-05,
+      "loss": 1.1281,
+      "step": 11759
+    },
+    {
+      "epoch": 10.8562197092084,
+      "grad_norm": 3.1953125,
+      "learning_rate": 2.161440960640427e-05,
+      "loss": 1.2549,
+      "step": 11760
+    },
+    {
+      "epoch": 10.857142857142858,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.160773849232822e-05,
+      "loss": 1.0502,
+      "step": 11761
+    },
+    {
+      "epoch": 10.858066005077314,
+      "grad_norm": 3.03515625,
+      "learning_rate": 2.160106737825217e-05,
+      "loss": 1.4153,
+      "step": 11762
+    },
+    {
+      "epoch": 10.85898915301177,
+      "grad_norm": 3.3046875,
+      "learning_rate": 2.159439626417612e-05,
+      "loss": 1.2031,
+      "step": 11763
+    },
+    {
+      "epoch": 10.859912300946228,
+      "grad_norm": 2.96484375,
+      "learning_rate": 2.1587725150100068e-05,
+      "loss": 1.1313,
+      "step": 11764
+    },
+    {
+      "epoch": 10.860835448880684,
+      "grad_norm": 3.369140625,
+      "learning_rate": 2.1581054036024016e-05,
+      "loss": 1.2966,
+      "step": 11765
+    },
+    {
+      "epoch": 10.86175859681514,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.1574382921947965e-05,
+      "loss": 1.375,
+      "step": 11766
+    },
+    {
+      "epoch": 10.862681744749596,
+      "grad_norm": 3.47265625,
+      "learning_rate": 2.1567711807871914e-05,
+      "loss": 1.2908,
+      "step": 11767
+    },
+    {
+      "epoch": 10.863604892684053,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.1561040693795863e-05,
+      "loss": 1.3958,
+      "step": 11768
+    },
+    {
+      "epoch": 10.86452804061851,
+      "grad_norm": 3.529296875,
+      "learning_rate": 2.155436957971981e-05,
+      "loss": 1.2861,
+      "step": 11769
+    },
+    {
+      "epoch": 10.865451188552965,
+      "grad_norm": 3.06640625,
+      "learning_rate": 2.1547698465643764e-05,
+      "loss": 1.4749,
+      "step": 11770
+    },
+    {
+      "epoch": 10.866374336487421,
+      "grad_norm": 3.052734375,
+      "learning_rate": 2.1541027351567713e-05,
+      "loss": 1.1248,
+      "step": 11771
+    },
+    {
+      "epoch": 10.86729748442188,
+      "grad_norm": 3.376953125,
+      "learning_rate": 2.153435623749166e-05,
+      "loss": 1.2732,
+      "step": 11772
+    },
+    {
+      "epoch": 10.868220632356335,
+      "grad_norm": 3.57421875,
+      "learning_rate": 2.152768512341561e-05,
+      "loss": 1.2915,
+      "step": 11773
+    },
+    {
+      "epoch": 10.869143780290791,
+      "grad_norm": 3.4453125,
+      "learning_rate": 2.152101400933956e-05,
+      "loss": 1.2642,
+      "step": 11774
+    },
+    {
+      "epoch": 10.870066928225247,
+      "grad_norm": 3.44140625,
+      "learning_rate": 2.1514342895263508e-05,
+      "loss": 1.3177,
+      "step": 11775
+    },
+    {
+      "epoch": 10.870990076159705,
+      "grad_norm": 3.001953125,
+      "learning_rate": 2.150767178118746e-05,
+      "loss": 1.2234,
+      "step": 11776
+    },
+    {
+      "epoch": 10.871913224094161,
+      "grad_norm": 2.98046875,
+      "learning_rate": 2.150100066711141e-05,
+      "loss": 1.3921,
+      "step": 11777
+    },
+    {
+      "epoch": 10.872836372028617,
+      "grad_norm": 3.693359375,
+      "learning_rate": 2.1494329553035358e-05,
+      "loss": 1.3806,
+      "step": 11778
+    },
+    {
+      "epoch": 10.873759519963073,
+      "grad_norm": 2.986328125,
+      "learning_rate": 2.1487658438959307e-05,
+      "loss": 1.1655,
+      "step": 11779
+    },
+    {
+      "epoch": 10.874682667897531,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.1480987324883255e-05,
+      "loss": 1.2683,
+      "step": 11780
+    },
+    {
+      "epoch": 10.875605815831987,
+      "grad_norm": 3.212890625,
+      "learning_rate": 2.1474316210807204e-05,
+      "loss": 0.8947,
+      "step": 11781
+    },
+    {
+      "epoch": 10.876528963766443,
+      "grad_norm": 2.822265625,
+      "learning_rate": 2.1467645096731153e-05,
+      "loss": 1.3574,
+      "step": 11782
+    },
+    {
+      "epoch": 10.8774521117009,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.1460973982655105e-05,
+      "loss": 1.2957,
+      "step": 11783
+    },
+    {
+      "epoch": 10.878375259635357,
+      "grad_norm": 3.498046875,
+      "learning_rate": 2.1454302868579054e-05,
+      "loss": 1.2426,
+      "step": 11784
+    },
+    {
+      "epoch": 10.879298407569813,
+      "grad_norm": 3.4921875,
+      "learning_rate": 2.1447631754503003e-05,
+      "loss": 1.24,
+      "step": 11785
+    },
+    {
+      "epoch": 10.880221555504269,
+      "grad_norm": 3.646484375,
+      "learning_rate": 2.1440960640426952e-05,
+      "loss": 1.2762,
+      "step": 11786
+    },
+    {
+      "epoch": 10.881144703438727,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.14342895263509e-05,
+      "loss": 1.344,
+      "step": 11787
+    },
+    {
+      "epoch": 10.882067851373183,
+      "grad_norm": 3.251953125,
+      "learning_rate": 2.142761841227485e-05,
+      "loss": 1.238,
+      "step": 11788
+    },
+    {
+      "epoch": 10.882990999307639,
+      "grad_norm": 3.5234375,
+      "learning_rate": 2.14209472981988e-05,
+      "loss": 1.5442,
+      "step": 11789
+    },
+    {
+      "epoch": 10.883914147242095,
+      "grad_norm": 3.3828125,
+      "learning_rate": 2.141427618412275e-05,
+      "loss": 1.3159,
+      "step": 11790
+    },
+    {
+      "epoch": 10.884837295176553,
+      "grad_norm": 3.74609375,
+      "learning_rate": 2.14076050700467e-05,
+      "loss": 1.1392,
+      "step": 11791
+    },
+    {
+      "epoch": 10.885760443111009,
+      "grad_norm": 3.658203125,
+      "learning_rate": 2.1400933955970648e-05,
+      "loss": 1.1216,
+      "step": 11792
+    },
+    {
+      "epoch": 10.886683591045465,
+      "grad_norm": 3.0546875,
+      "learning_rate": 2.1394262841894597e-05,
+      "loss": 1.1206,
+      "step": 11793
+    },
+    {
+      "epoch": 10.88760673897992,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.1387591727818546e-05,
+      "loss": 1.3198,
+      "step": 11794
+    },
+    {
+      "epoch": 10.888529886914379,
+      "grad_norm": 3.091796875,
+      "learning_rate": 2.1380920613742498e-05,
+      "loss": 1.4778,
+      "step": 11795
+    },
+    {
+      "epoch": 10.889453034848835,
+      "grad_norm": 3.169921875,
+      "learning_rate": 2.1374249499666447e-05,
+      "loss": 1.2162,
+      "step": 11796
+    },
+    {
+      "epoch": 10.89037618278329,
+      "grad_norm": 3.447265625,
+      "learning_rate": 2.1367578385590396e-05,
+      "loss": 1.427,
+      "step": 11797
+    },
+    {
+      "epoch": 10.891299330717747,
+      "grad_norm": 3.373046875,
+      "learning_rate": 2.1360907271514344e-05,
+      "loss": 1.5481,
+      "step": 11798
+    },
+    {
+      "epoch": 10.892222478652204,
+      "grad_norm": 3.115234375,
+      "learning_rate": 2.1354236157438293e-05,
+      "loss": 1.228,
+      "step": 11799
+    },
+    {
+      "epoch": 10.89314562658666,
+      "grad_norm": 3.904296875,
+      "learning_rate": 2.1347565043362242e-05,
+      "loss": 0.9677,
+      "step": 11800
+    },
+    {
+      "epoch": 10.894068774521116,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.1340893929286194e-05,
+      "loss": 1.2413,
+      "step": 11801
+    },
+    {
+      "epoch": 10.894991922455574,
+      "grad_norm": 3.27734375,
+      "learning_rate": 2.1334222815210143e-05,
+      "loss": 1.3463,
+      "step": 11802
+    },
+    {
+      "epoch": 10.89591507039003,
+      "grad_norm": 3.193359375,
+      "learning_rate": 2.1327551701134092e-05,
+      "loss": 1.2141,
+      "step": 11803
+    },
+    {
+      "epoch": 10.896838218324486,
+      "grad_norm": 3.302734375,
+      "learning_rate": 2.132088058705804e-05,
+      "loss": 1.3342,
+      "step": 11804
+    },
+    {
+      "epoch": 10.897761366258942,
+      "grad_norm": 3.279296875,
+      "learning_rate": 2.131420947298199e-05,
+      "loss": 1.4297,
+      "step": 11805
+    },
+    {
+      "epoch": 10.8986845141934,
+      "grad_norm": 3.173828125,
+      "learning_rate": 2.130753835890594e-05,
+      "loss": 1.1951,
+      "step": 11806
+    },
+    {
+      "epoch": 10.899607662127856,
+      "grad_norm": 3.361328125,
+      "learning_rate": 2.1300867244829887e-05,
+      "loss": 1.0554,
+      "step": 11807
+    },
+    {
+      "epoch": 10.900530810062312,
+      "grad_norm": 3.564453125,
+      "learning_rate": 2.129419613075384e-05,
+      "loss": 1.1428,
+      "step": 11808
+    },
+    {
+      "epoch": 10.901453957996768,
+      "grad_norm": 3.056640625,
+      "learning_rate": 2.1287525016677788e-05,
+      "loss": 1.321,
+      "step": 11809
+    },
+    {
+      "epoch": 10.902377105931226,
+      "grad_norm": 3.462890625,
+      "learning_rate": 2.1280853902601737e-05,
+      "loss": 1.1122,
+      "step": 11810
+    },
+    {
+      "epoch": 10.903300253865682,
+      "grad_norm": 2.9140625,
+      "learning_rate": 2.1274182788525686e-05,
+      "loss": 1.3826,
+      "step": 11811
+    },
+    {
+      "epoch": 10.904223401800138,
+      "grad_norm": 3.310546875,
+      "learning_rate": 2.1267511674449635e-05,
+      "loss": 1.4846,
+      "step": 11812
+    },
+    {
+      "epoch": 10.905146549734596,
+      "grad_norm": 3.51171875,
+      "learning_rate": 2.1260840560373583e-05,
+      "loss": 1.1078,
+      "step": 11813
+    },
+    {
+      "epoch": 10.906069697669052,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.1254169446297532e-05,
+      "loss": 0.8727,
+      "step": 11814
+    },
+    {
+      "epoch": 10.906992845603508,
+      "grad_norm": 3.849609375,
+      "learning_rate": 2.124749833222148e-05,
+      "loss": 1.2141,
+      "step": 11815
+    },
+    {
+      "epoch": 10.907915993537964,
+      "grad_norm": 3.77734375,
+      "learning_rate": 2.124082721814543e-05,
+      "loss": 1.4702,
+      "step": 11816
+    },
+    {
+      "epoch": 10.908839141472422,
+      "grad_norm": 3.36328125,
+      "learning_rate": 2.123415610406938e-05,
+      "loss": 1.2512,
+      "step": 11817
+    },
+    {
+      "epoch": 10.909762289406878,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.122748498999333e-05,
+      "loss": 1.3079,
+      "step": 11818
+    },
+    {
+      "epoch": 10.910685437341334,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.122081387591728e-05,
+      "loss": 1.2554,
+      "step": 11819
+    },
+    {
+      "epoch": 10.91160858527579,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.121414276184123e-05,
+      "loss": 1.3845,
+      "step": 11820
+    },
+    {
+      "epoch": 10.912531733210248,
+      "grad_norm": 3.90234375,
+      "learning_rate": 2.1207471647765177e-05,
+      "loss": 1.2242,
+      "step": 11821
+    },
+    {
+      "epoch": 10.913454881144704,
+      "grad_norm": 3.162109375,
+      "learning_rate": 2.1200800533689126e-05,
+      "loss": 1.4512,
+      "step": 11822
+    },
+    {
+      "epoch": 10.91437802907916,
+      "grad_norm": 3.24609375,
+      "learning_rate": 2.1194129419613075e-05,
+      "loss": 1.2131,
+      "step": 11823
+    },
+    {
+      "epoch": 10.915301177013616,
+      "grad_norm": 3.349609375,
+      "learning_rate": 2.1187458305537024e-05,
+      "loss": 1.3369,
+      "step": 11824
+    },
+    {
+      "epoch": 10.916224324948073,
+      "grad_norm": 3.41015625,
+      "learning_rate": 2.1180787191460973e-05,
+      "loss": 1.5181,
+      "step": 11825
+    },
+    {
+      "epoch": 10.91714747288253,
+      "grad_norm": 3.08984375,
+      "learning_rate": 2.1174116077384925e-05,
+      "loss": 1.4062,
+      "step": 11826
+    },
+    {
+      "epoch": 10.918070620816986,
+      "grad_norm": 3.35546875,
+      "learning_rate": 2.1167444963308874e-05,
+      "loss": 1.4375,
+      "step": 11827
+    },
+    {
+      "epoch": 10.918993768751442,
+      "grad_norm": 3.392578125,
+      "learning_rate": 2.1160773849232822e-05,
+      "loss": 1.2388,
+      "step": 11828
+    },
+    {
+      "epoch": 10.9199169166859,
+      "grad_norm": 3.025390625,
+      "learning_rate": 2.115410273515677e-05,
+      "loss": 1.3958,
+      "step": 11829
+    },
+    {
+      "epoch": 10.920840064620355,
+      "grad_norm": 3.45703125,
+      "learning_rate": 2.114743162108072e-05,
+      "loss": 1.4971,
+      "step": 11830
+    },
+    {
+      "epoch": 10.921763212554811,
+      "grad_norm": 3.501953125,
+      "learning_rate": 2.114076050700467e-05,
+      "loss": 1.3271,
+      "step": 11831
+    },
+    {
+      "epoch": 10.92268636048927,
+      "grad_norm": 3.36328125,
+      "learning_rate": 2.1134089392928618e-05,
+      "loss": 1.0172,
+      "step": 11832
+    },
+    {
+      "epoch": 10.923609508423725,
+      "grad_norm": 3.509765625,
+      "learning_rate": 2.112741827885257e-05,
+      "loss": 1.2131,
+      "step": 11833
+    },
+    {
+      "epoch": 10.924532656358181,
+      "grad_norm": 3.4921875,
+      "learning_rate": 2.112074716477652e-05,
+      "loss": 1.3726,
+      "step": 11834
+    },
+    {
+      "epoch": 10.925455804292637,
+      "grad_norm": 3.162109375,
+      "learning_rate": 2.1114076050700468e-05,
+      "loss": 1.0148,
+      "step": 11835
+    },
+    {
+      "epoch": 10.926378952227095,
+      "grad_norm": 3.853515625,
+      "learning_rate": 2.1107404936624416e-05,
+      "loss": 1.2621,
+      "step": 11836
+    },
+    {
+      "epoch": 10.927302100161551,
+      "grad_norm": 3.240234375,
+      "learning_rate": 2.1100733822548365e-05,
+      "loss": 1.2069,
+      "step": 11837
+    },
+    {
+      "epoch": 10.928225248096007,
+      "grad_norm": 3.498046875,
+      "learning_rate": 2.1094062708472314e-05,
+      "loss": 1.2626,
+      "step": 11838
+    },
+    {
+      "epoch": 10.929148396030463,
+      "grad_norm": 3.369140625,
+      "learning_rate": 2.1087391594396266e-05,
+      "loss": 1.4521,
+      "step": 11839
+    },
+    {
+      "epoch": 10.930071543964921,
+      "grad_norm": 3.548828125,
+      "learning_rate": 2.1080720480320215e-05,
+      "loss": 1.6177,
+      "step": 11840
+    },
+    {
+      "epoch": 10.930994691899377,
+      "grad_norm": 3.06640625,
+      "learning_rate": 2.1074049366244164e-05,
+      "loss": 1.1195,
+      "step": 11841
+    },
+    {
+      "epoch": 10.931917839833833,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.1067378252168113e-05,
+      "loss": 1.0664,
+      "step": 11842
+    },
+    {
+      "epoch": 10.932840987768289,
+      "grad_norm": 3.654296875,
+      "learning_rate": 2.106070713809206e-05,
+      "loss": 1.3804,
+      "step": 11843
+    },
+    {
+      "epoch": 10.933764135702747,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.105403602401601e-05,
+      "loss": 1.228,
+      "step": 11844
+    },
+    {
+      "epoch": 10.934687283637203,
+      "grad_norm": 3.5234375,
+      "learning_rate": 2.1047364909939963e-05,
+      "loss": 0.8151,
+      "step": 11845
+    },
+    {
+      "epoch": 10.935610431571659,
+      "grad_norm": 3.162109375,
+      "learning_rate": 2.104069379586391e-05,
+      "loss": 1.3093,
+      "step": 11846
+    },
+    {
+      "epoch": 10.936533579506115,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.103402268178786e-05,
+      "loss": 1.4023,
+      "step": 11847
+    },
+    {
+      "epoch": 10.937456727440573,
+      "grad_norm": 3.00390625,
+      "learning_rate": 2.102735156771181e-05,
+      "loss": 1.1599,
+      "step": 11848
+    },
+    {
+      "epoch": 10.938379875375029,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.1020680453635758e-05,
+      "loss": 1.3206,
+      "step": 11849
+    },
+    {
+      "epoch": 10.939303023309485,
+      "grad_norm": 3.2890625,
+      "learning_rate": 2.1014009339559707e-05,
+      "loss": 1.2957,
+      "step": 11850
+    },
+    {
+      "epoch": 10.940226171243943,
+      "grad_norm": 3.212890625,
+      "learning_rate": 2.100733822548366e-05,
+      "loss": 1.3232,
+      "step": 11851
+    },
+    {
+      "epoch": 10.941149319178399,
+      "grad_norm": 3.16796875,
+      "learning_rate": 2.1000667111407608e-05,
+      "loss": 1.1837,
+      "step": 11852
+    },
+    {
+      "epoch": 10.942072467112855,
+      "grad_norm": 3.603515625,
+      "learning_rate": 2.0993995997331557e-05,
+      "loss": 1.3157,
+      "step": 11853
+    },
+    {
+      "epoch": 10.94299561504731,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.0987324883255505e-05,
+      "loss": 1.264,
+      "step": 11854
+    },
+    {
+      "epoch": 10.943918762981768,
+      "grad_norm": 3.048828125,
+      "learning_rate": 2.0980653769179454e-05,
+      "loss": 1.2944,
+      "step": 11855
+    },
+    {
+      "epoch": 10.944841910916224,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.0973982655103403e-05,
+      "loss": 1.2776,
+      "step": 11856
+    },
+    {
+      "epoch": 10.94576505885068,
+      "grad_norm": 3.220703125,
+      "learning_rate": 2.0967311541027352e-05,
+      "loss": 1.152,
+      "step": 11857
+    },
+    {
+      "epoch": 10.946688206785137,
+      "grad_norm": 3.11328125,
+      "learning_rate": 2.0960640426951304e-05,
+      "loss": 1.2441,
+      "step": 11858
+    },
+    {
+      "epoch": 10.947611354719594,
+      "grad_norm": 3.09765625,
+      "learning_rate": 2.0953969312875253e-05,
+      "loss": 1.448,
+      "step": 11859
+    },
+    {
+      "epoch": 10.94853450265405,
+      "grad_norm": 3.177734375,
+      "learning_rate": 2.09472981987992e-05,
+      "loss": 0.8215,
+      "step": 11860
+    },
+    {
+      "epoch": 10.949457650588506,
+      "grad_norm": 3.36328125,
+      "learning_rate": 2.094062708472315e-05,
+      "loss": 1.2893,
+      "step": 11861
+    },
+    {
+      "epoch": 10.950380798522964,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.09339559706471e-05,
+      "loss": 1.291,
+      "step": 11862
+    },
+    {
+      "epoch": 10.95130394645742,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.0927284856571048e-05,
+      "loss": 0.9159,
+      "step": 11863
+    },
+    {
+      "epoch": 10.952227094391876,
+      "grad_norm": 2.693359375,
+      "learning_rate": 2.0920613742495e-05,
+      "loss": 1.3868,
+      "step": 11864
+    },
+    {
+      "epoch": 10.953150242326332,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.091394262841895e-05,
+      "loss": 1.2468,
+      "step": 11865
+    },
+    {
+      "epoch": 10.95407339026079,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.0907271514342898e-05,
+      "loss": 0.9841,
+      "step": 11866
+    },
+    {
+      "epoch": 10.954996538195246,
+      "grad_norm": 3.208984375,
+      "learning_rate": 2.0900600400266847e-05,
+      "loss": 1.1569,
+      "step": 11867
+    },
+    {
+      "epoch": 10.955919686129702,
+      "grad_norm": 3.244140625,
+      "learning_rate": 2.0893929286190796e-05,
+      "loss": 1.2238,
+      "step": 11868
+    },
+    {
+      "epoch": 10.956842834064158,
+      "grad_norm": 3.5234375,
+      "learning_rate": 2.0887258172114744e-05,
+      "loss": 1.3013,
+      "step": 11869
+    },
+    {
+      "epoch": 10.957765981998616,
+      "grad_norm": 3.513671875,
+      "learning_rate": 2.0880587058038693e-05,
+      "loss": 1.0427,
+      "step": 11870
+    },
+    {
+      "epoch": 10.958689129933072,
+      "grad_norm": 3.314453125,
+      "learning_rate": 2.0873915943962642e-05,
+      "loss": 1.5168,
+      "step": 11871
+    },
+    {
+      "epoch": 10.959612277867528,
+      "grad_norm": 3.353515625,
+      "learning_rate": 2.086724482988659e-05,
+      "loss": 1.0638,
+      "step": 11872
+    },
+    {
+      "epoch": 10.960535425801984,
+      "grad_norm": 3.3671875,
+      "learning_rate": 2.086057371581054e-05,
+      "loss": 1.251,
+      "step": 11873
+    },
+    {
+      "epoch": 10.961458573736442,
+      "grad_norm": 2.9140625,
+      "learning_rate": 2.0853902601734492e-05,
+      "loss": 1.3815,
+      "step": 11874
+    },
+    {
+      "epoch": 10.962381721670898,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.084723148765844e-05,
+      "loss": 1.3738,
+      "step": 11875
+    },
+    {
+      "epoch": 10.963304869605354,
+      "grad_norm": 3.04296875,
+      "learning_rate": 2.084056037358239e-05,
+      "loss": 0.8865,
+      "step": 11876
+    },
+    {
+      "epoch": 10.96422801753981,
+      "grad_norm": 3.14453125,
+      "learning_rate": 2.083388925950634e-05,
+      "loss": 1.2817,
+      "step": 11877
+    },
+    {
+      "epoch": 10.965151165474268,
+      "grad_norm": 3.662109375,
+      "learning_rate": 2.0827218145430287e-05,
+      "loss": 1.5139,
+      "step": 11878
+    },
+    {
+      "epoch": 10.966074313408724,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.0820547031354236e-05,
+      "loss": 1.1771,
+      "step": 11879
+    },
+    {
+      "epoch": 10.96699746134318,
+      "grad_norm": 3.4140625,
+      "learning_rate": 2.0813875917278185e-05,
+      "loss": 1.1199,
+      "step": 11880
+    },
+    {
+      "epoch": 10.967920609277638,
+      "grad_norm": 3.724609375,
+      "learning_rate": 2.0807204803202134e-05,
+      "loss": 1.2795,
+      "step": 11881
+    },
+    {
+      "epoch": 10.968843757212094,
+      "grad_norm": 3.5703125,
+      "learning_rate": 2.0800533689126082e-05,
+      "loss": 1.45,
+      "step": 11882
+    },
+    {
+      "epoch": 10.96976690514655,
+      "grad_norm": 2.77734375,
+      "learning_rate": 2.0793862575050035e-05,
+      "loss": 1.4165,
+      "step": 11883
+    },
+    {
+      "epoch": 10.970690053081006,
+      "grad_norm": 3.0390625,
+      "learning_rate": 2.0787191460973983e-05,
+      "loss": 1.1255,
+      "step": 11884
+    },
+    {
+      "epoch": 10.971613201015463,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.0780520346897932e-05,
+      "loss": 1.1223,
+      "step": 11885
+    },
+    {
+      "epoch": 10.97253634894992,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.077384923282188e-05,
+      "loss": 1.1904,
+      "step": 11886
+    },
+    {
+      "epoch": 10.973459496884375,
+      "grad_norm": 3.080078125,
+      "learning_rate": 2.076717811874583e-05,
+      "loss": 1.3181,
+      "step": 11887
+    },
+    {
+      "epoch": 10.974382644818832,
+      "grad_norm": 3.455078125,
+      "learning_rate": 2.076050700466978e-05,
+      "loss": 1.3774,
+      "step": 11888
+    },
+    {
+      "epoch": 10.97530579275329,
+      "grad_norm": 3.3984375,
+      "learning_rate": 2.075383589059373e-05,
+      "loss": 1.4496,
+      "step": 11889
+    },
+    {
+      "epoch": 10.976228940687745,
+      "grad_norm": 3.78515625,
+      "learning_rate": 2.074716477651768e-05,
+      "loss": 1.3914,
+      "step": 11890
+    },
+    {
+      "epoch": 10.977152088622201,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.074049366244163e-05,
+      "loss": 1.1615,
+      "step": 11891
+    },
+    {
+      "epoch": 10.978075236556657,
+      "grad_norm": 3.23828125,
+      "learning_rate": 2.0733822548365577e-05,
+      "loss": 1.2079,
+      "step": 11892
+    },
+    {
+      "epoch": 10.978998384491115,
+      "grad_norm": 3.26171875,
+      "learning_rate": 2.0727151434289526e-05,
+      "loss": 1.2979,
+      "step": 11893
+    },
+    {
+      "epoch": 10.979921532425571,
+      "grad_norm": 3.166015625,
+      "learning_rate": 2.0720480320213475e-05,
+      "loss": 1.2733,
+      "step": 11894
+    },
+    {
+      "epoch": 10.980844680360027,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0713809206137427e-05,
+      "loss": 1.2144,
+      "step": 11895
+    },
+    {
+      "epoch": 10.981767828294483,
+      "grad_norm": 3.12109375,
+      "learning_rate": 2.0707138092061376e-05,
+      "loss": 1.2795,
+      "step": 11896
+    },
+    {
+      "epoch": 10.982690976228941,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.0700466977985325e-05,
+      "loss": 1.3054,
+      "step": 11897
+    },
+    {
+      "epoch": 10.983614124163397,
+      "grad_norm": 3.30859375,
+      "learning_rate": 2.0693795863909274e-05,
+      "loss": 1.3987,
+      "step": 11898
+    },
+    {
+      "epoch": 10.984537272097853,
+      "grad_norm": 2.98828125,
+      "learning_rate": 2.0687124749833222e-05,
+      "loss": 1.2128,
+      "step": 11899
+    },
+    {
+      "epoch": 10.985460420032311,
+      "grad_norm": 3.490234375,
+      "learning_rate": 2.068045363575717e-05,
+      "loss": 1.3188,
+      "step": 11900
+    },
+    {
+      "epoch": 10.986383567966767,
+      "grad_norm": 3.345703125,
+      "learning_rate": 2.067378252168112e-05,
+      "loss": 1.2539,
+      "step": 11901
+    },
+    {
+      "epoch": 10.987306715901223,
+      "grad_norm": 3.02734375,
+      "learning_rate": 2.0667111407605072e-05,
+      "loss": 1.4065,
+      "step": 11902
+    },
+    {
+      "epoch": 10.988229863835679,
+      "grad_norm": 3.259765625,
+      "learning_rate": 2.066044029352902e-05,
+      "loss": 1.1517,
+      "step": 11903
+    },
+    {
+      "epoch": 10.989153011770137,
+      "grad_norm": 3.494140625,
+      "learning_rate": 2.065376917945297e-05,
+      "loss": 1.3126,
+      "step": 11904
+    },
+    {
+      "epoch": 10.990076159704593,
+      "grad_norm": 2.791015625,
+      "learning_rate": 2.064709806537692e-05,
+      "loss": 1.3142,
+      "step": 11905
+    },
+    {
+      "epoch": 10.990999307639049,
+      "grad_norm": 3.4921875,
+      "learning_rate": 2.0640426951300868e-05,
+      "loss": 1.3706,
+      "step": 11906
+    },
+    {
+      "epoch": 10.991922455573505,
+      "grad_norm": 3.029296875,
+      "learning_rate": 2.0633755837224816e-05,
+      "loss": 1.2028,
+      "step": 11907
+    },
+    {
+      "epoch": 10.992845603507963,
+      "grad_norm": 3.01953125,
+      "learning_rate": 2.062708472314877e-05,
+      "loss": 1.436,
+      "step": 11908
+    },
+    {
+      "epoch": 10.993768751442419,
+      "grad_norm": 3.072265625,
+      "learning_rate": 2.0620413609072717e-05,
+      "loss": 1.395,
+      "step": 11909
+    },
+    {
+      "epoch": 10.994691899376875,
+      "grad_norm": 3.298828125,
+      "learning_rate": 2.0613742494996666e-05,
+      "loss": 1.342,
+      "step": 11910
+    },
+    {
+      "epoch": 10.995615047311333,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.0607071380920615e-05,
+      "loss": 1.3411,
+      "step": 11911
+    },
+    {
+      "epoch": 10.996538195245789,
+      "grad_norm": 3.380859375,
+      "learning_rate": 2.0600400266844564e-05,
+      "loss": 1.1311,
+      "step": 11912
+    },
+    {
+      "epoch": 10.997461343180245,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.0593729152768513e-05,
+      "loss": 1.0826,
+      "step": 11913
+    },
+    {
+      "epoch": 10.9983844911147,
+      "grad_norm": 3.0859375,
+      "learning_rate": 2.0587058038692465e-05,
+      "loss": 1.3909,
+      "step": 11914
+    },
+    {
+      "epoch": 10.999307639049158,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.0580386924616414e-05,
+      "loss": 1.1206,
+      "step": 11915
+    },
+    {
+      "epoch": 11.000230786983614,
+      "grad_norm": 3.154296875,
+      "learning_rate": 2.0573715810540363e-05,
+      "loss": 0.9806,
+      "step": 11916
+    },
+    {
+      "epoch": 11.00115393491807,
+      "grad_norm": 3.55859375,
+      "learning_rate": 2.056704469646431e-05,
+      "loss": 1.0386,
+      "step": 11917
+    },
+    {
+      "epoch": 11.002077082852527,
+      "grad_norm": 3.22265625,
+      "learning_rate": 2.056037358238826e-05,
+      "loss": 1.1223,
+      "step": 11918
+    },
+    {
+      "epoch": 11.003000230786984,
+      "grad_norm": 3.294921875,
+      "learning_rate": 2.055370246831221e-05,
+      "loss": 1.0216,
+      "step": 11919
+    },
+    {
+      "epoch": 11.00392337872144,
+      "grad_norm": 3.1171875,
+      "learning_rate": 2.054703135423616e-05,
+      "loss": 0.9466,
+      "step": 11920
+    },
+    {
+      "epoch": 11.004846526655896,
+      "grad_norm": 2.97265625,
+      "learning_rate": 2.054036024016011e-05,
+      "loss": 1.366,
+      "step": 11921
+    },
+    {
+      "epoch": 11.005769674590352,
+      "grad_norm": 3.29296875,
+      "learning_rate": 2.053368912608406e-05,
+      "loss": 1.2004,
+      "step": 11922
+    },
+    {
+      "epoch": 11.00669282252481,
+      "grad_norm": 3.369140625,
+      "learning_rate": 2.0527018012008008e-05,
+      "loss": 1.2456,
+      "step": 11923
+    },
+    {
+      "epoch": 11.007615970459266,
+      "grad_norm": 3.19140625,
+      "learning_rate": 2.0520346897931957e-05,
+      "loss": 0.9125,
+      "step": 11924
+    },
+    {
+      "epoch": 11.008539118393722,
+      "grad_norm": 3.09765625,
+      "learning_rate": 2.0513675783855905e-05,
+      "loss": 1.3586,
+      "step": 11925
+    },
+    {
+      "epoch": 11.009462266328178,
+      "grad_norm": 3.244140625,
+      "learning_rate": 2.0507004669779854e-05,
+      "loss": 1.4421,
+      "step": 11926
+    },
+    {
+      "epoch": 11.010385414262636,
+      "grad_norm": 2.884765625,
+      "learning_rate": 2.0500333555703803e-05,
+      "loss": 1.4114,
+      "step": 11927
+    },
+    {
+      "epoch": 11.011308562197092,
+      "grad_norm": 3.00390625,
+      "learning_rate": 2.0493662441627752e-05,
+      "loss": 1.4766,
+      "step": 11928
+    },
+    {
+      "epoch": 11.012231710131548,
+      "grad_norm": 3.83203125,
+      "learning_rate": 2.04869913275517e-05,
+      "loss": 1.146,
+      "step": 11929
+    },
+    {
+      "epoch": 11.013154858066006,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.048032021347565e-05,
+      "loss": 1.1279,
+      "step": 11930
+    },
+    {
+      "epoch": 11.014078006000462,
+      "grad_norm": 3.42578125,
+      "learning_rate": 2.04736490993996e-05,
+      "loss": 1.231,
+      "step": 11931
+    },
+    {
+      "epoch": 11.015001153934918,
+      "grad_norm": 3.154296875,
+      "learning_rate": 2.046697798532355e-05,
+      "loss": 1.2473,
+      "step": 11932
+    },
+    {
+      "epoch": 11.015924301869374,
+      "grad_norm": 3.27734375,
+      "learning_rate": 2.04603068712475e-05,
+      "loss": 1.1923,
+      "step": 11933
+    },
+    {
+      "epoch": 11.016847449803832,
+      "grad_norm": 3.072265625,
+      "learning_rate": 2.0453635757171448e-05,
+      "loss": 1.1292,
+      "step": 11934
+    },
+    {
+      "epoch": 11.017770597738288,
+      "grad_norm": 3.638671875,
+      "learning_rate": 2.0446964643095397e-05,
+      "loss": 0.6793,
+      "step": 11935
+    },
+    {
+      "epoch": 11.018693745672744,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.0440293529019346e-05,
+      "loss": 1.1075,
+      "step": 11936
+    },
+    {
+      "epoch": 11.0196168936072,
+      "grad_norm": 3.564453125,
+      "learning_rate": 2.0433622414943295e-05,
+      "loss": 1.2917,
+      "step": 11937
+    },
+    {
+      "epoch": 11.020540041541658,
+      "grad_norm": 3.17578125,
+      "learning_rate": 2.0426951300867243e-05,
+      "loss": 1.3152,
+      "step": 11938
+    },
+    {
+      "epoch": 11.021463189476114,
+      "grad_norm": 3.23828125,
+      "learning_rate": 2.0420280186791196e-05,
+      "loss": 1.2815,
+      "step": 11939
+    },
+    {
+      "epoch": 11.02238633741057,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.0413609072715144e-05,
+      "loss": 0.8814,
+      "step": 11940
+    },
+    {
+      "epoch": 11.023309485345026,
+      "grad_norm": 3.818359375,
+      "learning_rate": 2.0406937958639093e-05,
+      "loss": 1.0398,
+      "step": 11941
+    },
+    {
+      "epoch": 11.024232633279484,
+      "grad_norm": 3.181640625,
+      "learning_rate": 2.0400266844563042e-05,
+      "loss": 1.5784,
+      "step": 11942
+    },
+    {
+      "epoch": 11.02515578121394,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.039359573048699e-05,
+      "loss": 1.3499,
+      "step": 11943
+    },
+    {
+      "epoch": 11.026078929148396,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.038692461641094e-05,
+      "loss": 1.5286,
+      "step": 11944
+    },
+    {
+      "epoch": 11.027002077082853,
+      "grad_norm": 4.0078125,
+      "learning_rate": 2.0380253502334892e-05,
+      "loss": 1.2126,
+      "step": 11945
+    },
+    {
+      "epoch": 11.02792522501731,
+      "grad_norm": 3.123046875,
+      "learning_rate": 2.037358238825884e-05,
+      "loss": 1.3325,
+      "step": 11946
+    },
+    {
+      "epoch": 11.028848372951765,
+      "grad_norm": 3.6015625,
+      "learning_rate": 2.036691127418279e-05,
+      "loss": 1.1243,
+      "step": 11947
+    },
+    {
+      "epoch": 11.029771520886221,
+      "grad_norm": 3.193359375,
+      "learning_rate": 2.036024016010674e-05,
+      "loss": 1.2209,
+      "step": 11948
+    },
+    {
+      "epoch": 11.03069466882068,
+      "grad_norm": 3.20703125,
+      "learning_rate": 2.0353569046030687e-05,
+      "loss": 1.2847,
+      "step": 11949
+    },
+    {
+      "epoch": 11.031617816755135,
+      "grad_norm": 3.1640625,
+      "learning_rate": 2.0346897931954636e-05,
+      "loss": 1.2104,
+      "step": 11950
+    },
+    {
+      "epoch": 11.032540964689591,
+      "grad_norm": 2.9453125,
+      "learning_rate": 2.0340226817878585e-05,
+      "loss": 1.2195,
+      "step": 11951
+    },
+    {
+      "epoch": 11.033464112624047,
+      "grad_norm": 2.94921875,
+      "learning_rate": 2.0333555703802537e-05,
+      "loss": 1.3628,
+      "step": 11952
+    },
+    {
+      "epoch": 11.034387260558505,
+      "grad_norm": 3.9609375,
+      "learning_rate": 2.0326884589726486e-05,
+      "loss": 1.4229,
+      "step": 11953
+    },
+    {
+      "epoch": 11.035310408492961,
+      "grad_norm": 3.52734375,
+      "learning_rate": 2.0320213475650435e-05,
+      "loss": 1.0366,
+      "step": 11954
+    },
+    {
+      "epoch": 11.036233556427417,
+      "grad_norm": 3.251953125,
+      "learning_rate": 2.0313542361574383e-05,
+      "loss": 1.2896,
+      "step": 11955
+    },
+    {
+      "epoch": 11.037156704361873,
+      "grad_norm": 2.94140625,
+      "learning_rate": 2.0306871247498332e-05,
+      "loss": 1.3481,
+      "step": 11956
+    },
+    {
+      "epoch": 11.038079852296331,
+      "grad_norm": 3.15234375,
+      "learning_rate": 2.030020013342228e-05,
+      "loss": 1.037,
+      "step": 11957
+    },
+    {
+      "epoch": 11.039003000230787,
+      "grad_norm": 2.8515625,
+      "learning_rate": 2.0293529019346233e-05,
+      "loss": 1.3286,
+      "step": 11958
+    },
+    {
+      "epoch": 11.039926148165243,
+      "grad_norm": 3.58203125,
+      "learning_rate": 2.0286857905270182e-05,
+      "loss": 1.4299,
+      "step": 11959
+    },
+    {
+      "epoch": 11.0408492960997,
+      "grad_norm": 3.490234375,
+      "learning_rate": 2.028018679119413e-05,
+      "loss": 1.1995,
+      "step": 11960
+    },
+    {
+      "epoch": 11.041772444034157,
+      "grad_norm": 3.1171875,
+      "learning_rate": 2.027351567711808e-05,
+      "loss": 1.0337,
+      "step": 11961
+    },
+    {
+      "epoch": 11.042695591968613,
+      "grad_norm": 3.166015625,
+      "learning_rate": 2.026684456304203e-05,
+      "loss": 1.2959,
+      "step": 11962
+    },
+    {
+      "epoch": 11.043618739903069,
+      "grad_norm": 3.85546875,
+      "learning_rate": 2.0260173448965977e-05,
+      "loss": 1.1462,
+      "step": 11963
+    },
+    {
+      "epoch": 11.044541887837527,
+      "grad_norm": 3.353515625,
+      "learning_rate": 2.025350233488993e-05,
+      "loss": 1.0693,
+      "step": 11964
+    },
+    {
+      "epoch": 11.045465035771983,
+      "grad_norm": 3.248046875,
+      "learning_rate": 2.024683122081388e-05,
+      "loss": 1.179,
+      "step": 11965
+    },
+    {
+      "epoch": 11.046388183706439,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.0240160106737827e-05,
+      "loss": 1.1976,
+      "step": 11966
+    },
+    {
+      "epoch": 11.047311331640895,
+      "grad_norm": 2.935546875,
+      "learning_rate": 2.0233488992661776e-05,
+      "loss": 1.3474,
+      "step": 11967
+    },
+    {
+      "epoch": 11.048234479575353,
+      "grad_norm": 2.55859375,
+      "learning_rate": 2.0226817878585725e-05,
+      "loss": 1.0047,
+      "step": 11968
+    },
+    {
+      "epoch": 11.049157627509809,
+      "grad_norm": 3.9765625,
+      "learning_rate": 2.0220146764509674e-05,
+      "loss": 1.1898,
+      "step": 11969
+    },
+    {
+      "epoch": 11.050080775444265,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.0213475650433622e-05,
+      "loss": 1.2871,
+      "step": 11970
+    },
+    {
+      "epoch": 11.05100392337872,
+      "grad_norm": 3.099609375,
+      "learning_rate": 2.0206804536357575e-05,
+      "loss": 1.4351,
+      "step": 11971
+    },
+    {
+      "epoch": 11.051927071313179,
+      "grad_norm": 2.90234375,
+      "learning_rate": 2.0200133422281524e-05,
+      "loss": 1.4666,
+      "step": 11972
+    },
+    {
+      "epoch": 11.052850219247635,
+      "grad_norm": 3.55859375,
+      "learning_rate": 2.0193462308205472e-05,
+      "loss": 1.0667,
+      "step": 11973
+    },
+    {
+      "epoch": 11.05377336718209,
+      "grad_norm": 3.248046875,
+      "learning_rate": 2.018679119412942e-05,
+      "loss": 1.3999,
+      "step": 11974
+    },
+    {
+      "epoch": 11.054696515116547,
+      "grad_norm": 2.962890625,
+      "learning_rate": 2.018012008005337e-05,
+      "loss": 1.281,
+      "step": 11975
+    },
+    {
+      "epoch": 11.055619663051004,
+      "grad_norm": 3.275390625,
+      "learning_rate": 2.017344896597732e-05,
+      "loss": 1.2915,
+      "step": 11976
+    },
+    {
+      "epoch": 11.05654281098546,
+      "grad_norm": 2.935546875,
+      "learning_rate": 2.016677785190127e-05,
+      "loss": 1.4553,
+      "step": 11977
+    },
+    {
+      "epoch": 11.057465958919916,
+      "grad_norm": 3.22265625,
+      "learning_rate": 2.016010673782522e-05,
+      "loss": 1.5212,
+      "step": 11978
+    },
+    {
+      "epoch": 11.058389106854374,
+      "grad_norm": 3.7109375,
+      "learning_rate": 2.015343562374917e-05,
+      "loss": 1.186,
+      "step": 11979
+    },
+    {
+      "epoch": 11.05931225478883,
+      "grad_norm": 3.34765625,
+      "learning_rate": 2.0146764509673117e-05,
+      "loss": 1.2002,
+      "step": 11980
+    },
+    {
+      "epoch": 11.060235402723286,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.0140093395597066e-05,
+      "loss": 1.4287,
+      "step": 11981
+    },
+    {
+      "epoch": 11.061158550657742,
+      "grad_norm": 3.498046875,
+      "learning_rate": 2.0133422281521015e-05,
+      "loss": 1.3486,
+      "step": 11982
+    },
+    {
+      "epoch": 11.0620816985922,
+      "grad_norm": 2.982421875,
+      "learning_rate": 2.0126751167444964e-05,
+      "loss": 1.2887,
+      "step": 11983
+    },
+    {
+      "epoch": 11.063004846526656,
+      "grad_norm": 2.94921875,
+      "learning_rate": 2.0120080053368913e-05,
+      "loss": 1.3623,
+      "step": 11984
+    },
+    {
+      "epoch": 11.063927994461112,
+      "grad_norm": 3.375,
+      "learning_rate": 2.011340893929286e-05,
+      "loss": 1.4574,
+      "step": 11985
+    },
+    {
+      "epoch": 11.064851142395568,
+      "grad_norm": 3.228515625,
+      "learning_rate": 2.010673782521681e-05,
+      "loss": 1.2628,
+      "step": 11986
+    },
+    {
+      "epoch": 11.065774290330026,
+      "grad_norm": 3.111328125,
+      "learning_rate": 2.0100066711140763e-05,
+      "loss": 1.2122,
+      "step": 11987
+    },
+    {
+      "epoch": 11.066697438264482,
+      "grad_norm": 3.083984375,
+      "learning_rate": 2.009339559706471e-05,
+      "loss": 1.4197,
+      "step": 11988
+    },
+    {
+      "epoch": 11.067620586198938,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.008672448298866e-05,
+      "loss": 1.2932,
+      "step": 11989
+    },
+    {
+      "epoch": 11.068543734133394,
+      "grad_norm": 3.439453125,
+      "learning_rate": 2.008005336891261e-05,
+      "loss": 1.397,
+      "step": 11990
+    },
+    {
+      "epoch": 11.069466882067852,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.0073382254836558e-05,
+      "loss": 1.0407,
+      "step": 11991
+    },
+    {
+      "epoch": 11.070390030002308,
+      "grad_norm": 3.55859375,
+      "learning_rate": 2.0066711140760507e-05,
+      "loss": 1.2314,
+      "step": 11992
+    },
+    {
+      "epoch": 11.071313177936764,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.0060040026684455e-05,
+      "loss": 1.1648,
+      "step": 11993
+    },
+    {
+      "epoch": 11.07223632587122,
+      "grad_norm": 3.177734375,
+      "learning_rate": 2.0053368912608404e-05,
+      "loss": 1.3907,
+      "step": 11994
+    },
+    {
+      "epoch": 11.073159473805678,
+      "grad_norm": 3.51953125,
+      "learning_rate": 2.0046697798532353e-05,
+      "loss": 1.1289,
+      "step": 11995
+    },
+    {
+      "epoch": 11.074082621740134,
+      "grad_norm": 3.064453125,
+      "learning_rate": 2.0040026684456305e-05,
+      "loss": 1.2883,
+      "step": 11996
+    },
+    {
+      "epoch": 11.07500576967459,
+      "grad_norm": 2.931640625,
+      "learning_rate": 2.0033355570380254e-05,
+      "loss": 1.0844,
+      "step": 11997
+    },
+    {
+      "epoch": 11.075928917609048,
+      "grad_norm": 3.16015625,
+      "learning_rate": 2.0026684456304203e-05,
+      "loss": 1.3003,
+      "step": 11998
+    },
+    {
+      "epoch": 11.076852065543504,
+      "grad_norm": 2.89453125,
+      "learning_rate": 2.0020013342228152e-05,
+      "loss": 1.2749,
+      "step": 11999
+    },
+    {
+      "epoch": 11.07777521347796,
+      "grad_norm": 3.11328125,
+      "learning_rate": 2.00133422281521e-05,
+      "loss": 1.1154,
+      "step": 12000
+    },
+    {
+      "epoch": 11.07777521347796,
+      "eval_loss": 0.87255859375,
+      "eval_runtime": 23.3471,
+      "eval_samples_per_second": 976.611,
+      "eval_steps_per_second": 61.078,
+      "step": 12000
+    },
+    {
+      "epoch": 11.078698361412416,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.000667111407605e-05,
+      "loss": 1.3613,
+      "step": 12001
+    },
+    {
+      "epoch": 11.079621509346874,
+      "grad_norm": 3.4296875,
+      "learning_rate": 2e-05,
+      "loss": 1.1259,
+      "step": 12002
+    },
+    {
+      "epoch": 11.08054465728133,
+      "grad_norm": 3.263671875,
+      "learning_rate": 1.999332888592395e-05,
+      "loss": 1.3452,
+      "step": 12003
+    },
+    {
+      "epoch": 11.081467805215786,
+      "grad_norm": 3.19921875,
+      "learning_rate": 1.99866577718479e-05,
+      "loss": 1.3547,
+      "step": 12004
+    },
+    {
+      "epoch": 11.082390953150242,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.9979986657771848e-05,
+      "loss": 1.4727,
+      "step": 12005
+    },
+    {
+      "epoch": 11.0833141010847,
+      "grad_norm": 3.330078125,
+      "learning_rate": 1.9973315543695797e-05,
+      "loss": 1.1278,
+      "step": 12006
+    },
+    {
+      "epoch": 11.084237249019155,
+      "grad_norm": 2.8359375,
+      "learning_rate": 1.9966644429619746e-05,
+      "loss": 1.0103,
+      "step": 12007
+    },
+    {
+      "epoch": 11.085160396953611,
+      "grad_norm": 2.935546875,
+      "learning_rate": 1.9959973315543698e-05,
+      "loss": 1.2388,
+      "step": 12008
+    },
+    {
+      "epoch": 11.086083544888067,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.9953302201467647e-05,
+      "loss": 1.2126,
+      "step": 12009
+    },
+    {
+      "epoch": 11.087006692822525,
+      "grad_norm": 2.6484375,
+      "learning_rate": 1.9946631087391596e-05,
+      "loss": 0.9457,
+      "step": 12010
+    },
+    {
+      "epoch": 11.087929840756981,
+      "grad_norm": 3.685546875,
+      "learning_rate": 1.9939959973315544e-05,
+      "loss": 1.416,
+      "step": 12011
+    },
+    {
+      "epoch": 11.088852988691437,
+      "grad_norm": 3.5703125,
+      "learning_rate": 1.9933288859239493e-05,
+      "loss": 1.1775,
+      "step": 12012
+    },
+    {
+      "epoch": 11.089776136625895,
+      "grad_norm": 3.55078125,
+      "learning_rate": 1.9926617745163442e-05,
+      "loss": 1.4492,
+      "step": 12013
+    },
+    {
+      "epoch": 11.090699284560351,
+      "grad_norm": 3.3515625,
+      "learning_rate": 1.9919946631087394e-05,
+      "loss": 1.1666,
+      "step": 12014
+    },
+    {
+      "epoch": 11.091622432494807,
+      "grad_norm": 3.275390625,
+      "learning_rate": 1.9913275517011343e-05,
+      "loss": 1.1561,
+      "step": 12015
+    },
+    {
+      "epoch": 11.092545580429263,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.9906604402935292e-05,
+      "loss": 1.3024,
+      "step": 12016
+    },
+    {
+      "epoch": 11.093468728363721,
+      "grad_norm": 3.42578125,
+      "learning_rate": 1.989993328885924e-05,
+      "loss": 1.3606,
+      "step": 12017
+    },
+    {
+      "epoch": 11.094391876298177,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.989326217478319e-05,
+      "loss": 1.3608,
+      "step": 12018
+    },
+    {
+      "epoch": 11.095315024232633,
+      "grad_norm": 3.0,
+      "learning_rate": 1.988659106070714e-05,
+      "loss": 1.2727,
+      "step": 12019
+    },
+    {
+      "epoch": 11.096238172167089,
+      "grad_norm": 3.6015625,
+      "learning_rate": 1.9879919946631087e-05,
+      "loss": 1.3491,
+      "step": 12020
+    },
+    {
+      "epoch": 11.097161320101547,
+      "grad_norm": 3.427734375,
+      "learning_rate": 1.987324883255504e-05,
+      "loss": 1.5415,
+      "step": 12021
+    },
+    {
+      "epoch": 11.098084468036003,
+      "grad_norm": 3.357421875,
+      "learning_rate": 1.9866577718478988e-05,
+      "loss": 1.3163,
+      "step": 12022
+    },
+    {
+      "epoch": 11.099007615970459,
+      "grad_norm": 3.39453125,
+      "learning_rate": 1.9859906604402937e-05,
+      "loss": 0.9879,
+      "step": 12023
+    },
+    {
+      "epoch": 11.099930763904915,
+      "grad_norm": 3.17578125,
+      "learning_rate": 1.9853235490326886e-05,
+      "loss": 1.4719,
+      "step": 12024
+    },
+    {
+      "epoch": 11.100853911839373,
+      "grad_norm": 3.369140625,
+      "learning_rate": 1.9846564376250835e-05,
+      "loss": 1.1321,
+      "step": 12025
+    },
+    {
+      "epoch": 11.101777059773829,
+      "grad_norm": 3.361328125,
+      "learning_rate": 1.9839893262174783e-05,
+      "loss": 1.3696,
+      "step": 12026
+    },
+    {
+      "epoch": 11.102700207708285,
+      "grad_norm": 3.498046875,
+      "learning_rate": 1.9833222148098736e-05,
+      "loss": 1.3628,
+      "step": 12027
+    },
+    {
+      "epoch": 11.103623355642743,
+      "grad_norm": 3.349609375,
+      "learning_rate": 1.9826551034022684e-05,
+      "loss": 1.3503,
+      "step": 12028
+    },
+    {
+      "epoch": 11.104546503577199,
+      "grad_norm": 2.857421875,
+      "learning_rate": 1.9819879919946633e-05,
+      "loss": 1.0384,
+      "step": 12029
+    },
+    {
+      "epoch": 11.105469651511655,
+      "grad_norm": 3.75390625,
+      "learning_rate": 1.9813208805870582e-05,
+      "loss": 0.8689,
+      "step": 12030
+    },
+    {
+      "epoch": 11.10639279944611,
+      "grad_norm": 3.2421875,
+      "learning_rate": 1.980653769179453e-05,
+      "loss": 1.1299,
+      "step": 12031
+    },
+    {
+      "epoch": 11.107315947380568,
+      "grad_norm": 3.509765625,
+      "learning_rate": 1.979986657771848e-05,
+      "loss": 1.4434,
+      "step": 12032
+    },
+    {
+      "epoch": 11.108239095315025,
+      "grad_norm": 2.923828125,
+      "learning_rate": 1.9793195463642432e-05,
+      "loss": 0.9911,
+      "step": 12033
+    },
+    {
+      "epoch": 11.10916224324948,
+      "grad_norm": 3.375,
+      "learning_rate": 1.978652434956638e-05,
+      "loss": 1.4521,
+      "step": 12034
+    },
+    {
+      "epoch": 11.110085391183937,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.977985323549033e-05,
+      "loss": 1.2939,
+      "step": 12035
+    },
+    {
+      "epoch": 11.111008539118394,
+      "grad_norm": 3.59765625,
+      "learning_rate": 1.977318212141428e-05,
+      "loss": 1.1824,
+      "step": 12036
+    },
+    {
+      "epoch": 11.11193168705285,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.9766511007338227e-05,
+      "loss": 1.2183,
+      "step": 12037
+    },
+    {
+      "epoch": 11.112854834987306,
+      "grad_norm": 3.1484375,
+      "learning_rate": 1.9759839893262176e-05,
+      "loss": 1.2417,
+      "step": 12038
+    },
+    {
+      "epoch": 11.113777982921762,
+      "grad_norm": 3.07421875,
+      "learning_rate": 1.9753168779186125e-05,
+      "loss": 1.2975,
+      "step": 12039
+    },
+    {
+      "epoch": 11.11470113085622,
+      "grad_norm": 3.50390625,
+      "learning_rate": 1.9746497665110074e-05,
+      "loss": 1.2266,
+      "step": 12040
+    },
+    {
+      "epoch": 11.115624278790676,
+      "grad_norm": 3.205078125,
+      "learning_rate": 1.9739826551034023e-05,
+      "loss": 1.3499,
+      "step": 12041
+    },
+    {
+      "epoch": 11.116547426725132,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.973315543695797e-05,
+      "loss": 1.2865,
+      "step": 12042
+    },
+    {
+      "epoch": 11.117470574659588,
+      "grad_norm": 3.326171875,
+      "learning_rate": 1.972648432288192e-05,
+      "loss": 1.3046,
+      "step": 12043
+    },
+    {
+      "epoch": 11.118393722594046,
+      "grad_norm": 3.91015625,
+      "learning_rate": 1.9719813208805872e-05,
+      "loss": 1.2324,
+      "step": 12044
+    },
+    {
+      "epoch": 11.119316870528502,
+      "grad_norm": 3.072265625,
+      "learning_rate": 1.971314209472982e-05,
+      "loss": 1.1622,
+      "step": 12045
+    },
+    {
+      "epoch": 11.120240018462958,
+      "grad_norm": 3.130859375,
+      "learning_rate": 1.970647098065377e-05,
+      "loss": 1.2805,
+      "step": 12046
+    },
+    {
+      "epoch": 11.121163166397416,
+      "grad_norm": 3.01953125,
+      "learning_rate": 1.969979986657772e-05,
+      "loss": 1.1884,
+      "step": 12047
+    },
+    {
+      "epoch": 11.122086314331872,
+      "grad_norm": 3.052734375,
+      "learning_rate": 1.9693128752501668e-05,
+      "loss": 1.314,
+      "step": 12048
+    },
+    {
+      "epoch": 11.123009462266328,
+      "grad_norm": 3.26171875,
+      "learning_rate": 1.9686457638425616e-05,
+      "loss": 1.0831,
+      "step": 12049
+    },
+    {
+      "epoch": 11.123932610200784,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.9679786524349565e-05,
+      "loss": 1.2717,
+      "step": 12050
+    },
+    {
+      "epoch": 11.124855758135242,
+      "grad_norm": 3.005859375,
+      "learning_rate": 1.9673115410273514e-05,
+      "loss": 1.2327,
+      "step": 12051
+    },
+    {
+      "epoch": 11.125778906069698,
+      "grad_norm": 3.763671875,
+      "learning_rate": 1.9666444296197466e-05,
+      "loss": 1.2531,
+      "step": 12052
+    },
+    {
+      "epoch": 11.126702054004154,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.9659773182121415e-05,
+      "loss": 1.4829,
+      "step": 12053
+    },
+    {
+      "epoch": 11.12762520193861,
+      "grad_norm": 3.068359375,
+      "learning_rate": 1.9653102068045364e-05,
+      "loss": 1.2446,
+      "step": 12054
+    },
+    {
+      "epoch": 11.128548349873068,
+      "grad_norm": 3.12109375,
+      "learning_rate": 1.9646430953969313e-05,
+      "loss": 1.1958,
+      "step": 12055
+    },
+    {
+      "epoch": 11.129471497807524,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.963975983989326e-05,
+      "loss": 1.3464,
+      "step": 12056
+    },
+    {
+      "epoch": 11.13039464574198,
+      "grad_norm": 3.607421875,
+      "learning_rate": 1.963308872581721e-05,
+      "loss": 1.1417,
+      "step": 12057
+    },
+    {
+      "epoch": 11.131317793676436,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.9626417611741163e-05,
+      "loss": 1.3423,
+      "step": 12058
+    },
+    {
+      "epoch": 11.132240941610894,
+      "grad_norm": 3.51171875,
+      "learning_rate": 1.961974649766511e-05,
+      "loss": 1.4934,
+      "step": 12059
+    },
+    {
+      "epoch": 11.13316408954535,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.961307538358906e-05,
+      "loss": 1.4541,
+      "step": 12060
+    },
+    {
+      "epoch": 11.134087237479806,
+      "grad_norm": 3.583984375,
+      "learning_rate": 1.960640426951301e-05,
+      "loss": 1.3435,
+      "step": 12061
+    },
+    {
+      "epoch": 11.135010385414263,
+      "grad_norm": 3.51953125,
+      "learning_rate": 1.9599733155436958e-05,
+      "loss": 1.3745,
+      "step": 12062
+    },
+    {
+      "epoch": 11.13593353334872,
+      "grad_norm": 3.05859375,
+      "learning_rate": 1.9593062041360907e-05,
+      "loss": 1.3552,
+      "step": 12063
+    },
+    {
+      "epoch": 11.136856681283176,
+      "grad_norm": 3.75390625,
+      "learning_rate": 1.9586390927284855e-05,
+      "loss": 1.2828,
+      "step": 12064
+    },
+    {
+      "epoch": 11.137779829217632,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.9579719813208808e-05,
+      "loss": 1.3142,
+      "step": 12065
+    },
+    {
+      "epoch": 11.13870297715209,
+      "grad_norm": 3.193359375,
+      "learning_rate": 1.9573048699132757e-05,
+      "loss": 1.2047,
+      "step": 12066
+    },
+    {
+      "epoch": 11.139626125086545,
+      "grad_norm": 2.66015625,
+      "learning_rate": 1.9566377585056705e-05,
+      "loss": 1.2725,
+      "step": 12067
+    },
+    {
+      "epoch": 11.140549273021001,
+      "grad_norm": 3.48046875,
+      "learning_rate": 1.9559706470980654e-05,
+      "loss": 1.0858,
+      "step": 12068
+    },
+    {
+      "epoch": 11.141472420955457,
+      "grad_norm": 3.44140625,
+      "learning_rate": 1.9553035356904603e-05,
+      "loss": 1.4333,
+      "step": 12069
+    },
+    {
+      "epoch": 11.142395568889915,
+      "grad_norm": 2.955078125,
+      "learning_rate": 1.9546364242828552e-05,
+      "loss": 1.3477,
+      "step": 12070
+    },
+    {
+      "epoch": 11.143318716824371,
+      "grad_norm": 4.91015625,
+      "learning_rate": 1.9539693128752504e-05,
+      "loss": 0.9988,
+      "step": 12071
+    },
+    {
+      "epoch": 11.144241864758827,
+      "grad_norm": 3.67578125,
+      "learning_rate": 1.9533022014676453e-05,
+      "loss": 1.2238,
+      "step": 12072
+    },
+    {
+      "epoch": 11.145165012693283,
+      "grad_norm": 3.525390625,
+      "learning_rate": 1.95263509006004e-05,
+      "loss": 1.342,
+      "step": 12073
+    },
+    {
+      "epoch": 11.146088160627741,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.951967978652435e-05,
+      "loss": 1.2726,
+      "step": 12074
+    },
+    {
+      "epoch": 11.147011308562197,
+      "grad_norm": 3.423828125,
+      "learning_rate": 1.95130086724483e-05,
+      "loss": 1.3005,
+      "step": 12075
+    },
+    {
+      "epoch": 11.147934456496653,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.9506337558372248e-05,
+      "loss": 1.4233,
+      "step": 12076
+    },
+    {
+      "epoch": 11.148857604431111,
+      "grad_norm": 3.599609375,
+      "learning_rate": 1.94996664442962e-05,
+      "loss": 1.3799,
+      "step": 12077
+    },
+    {
+      "epoch": 11.149780752365567,
+      "grad_norm": 3.060546875,
+      "learning_rate": 1.949299533022015e-05,
+      "loss": 1.3336,
+      "step": 12078
+    },
+    {
+      "epoch": 11.150703900300023,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.9486324216144098e-05,
+      "loss": 1.0259,
+      "step": 12079
+    },
+    {
+      "epoch": 11.151627048234479,
+      "grad_norm": 3.287109375,
+      "learning_rate": 1.9479653102068047e-05,
+      "loss": 0.9997,
+      "step": 12080
+    },
+    {
+      "epoch": 11.152550196168937,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.9472981987991996e-05,
+      "loss": 0.9928,
+      "step": 12081
+    },
+    {
+      "epoch": 11.153473344103393,
+      "grad_norm": 3.08984375,
+      "learning_rate": 1.9466310873915944e-05,
+      "loss": 1.0502,
+      "step": 12082
+    },
+    {
+      "epoch": 11.154396492037849,
+      "grad_norm": 3.1640625,
+      "learning_rate": 1.9459639759839897e-05,
+      "loss": 1.3894,
+      "step": 12083
+    },
+    {
+      "epoch": 11.155319639972305,
+      "grad_norm": 3.23828125,
+      "learning_rate": 1.9452968645763845e-05,
+      "loss": 1.2559,
+      "step": 12084
+    },
+    {
+      "epoch": 11.156242787906763,
+      "grad_norm": 3.4296875,
+      "learning_rate": 1.9446297531687794e-05,
+      "loss": 1.4232,
+      "step": 12085
+    },
+    {
+      "epoch": 11.157165935841219,
+      "grad_norm": 3.63671875,
+      "learning_rate": 1.9439626417611743e-05,
+      "loss": 1.3218,
+      "step": 12086
+    },
+    {
+      "epoch": 11.158089083775675,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.9432955303535692e-05,
+      "loss": 1.2661,
+      "step": 12087
+    },
+    {
+      "epoch": 11.15901223171013,
+      "grad_norm": 3.416015625,
+      "learning_rate": 1.942628418945964e-05,
+      "loss": 1.4097,
+      "step": 12088
+    },
+    {
+      "epoch": 11.159935379644589,
+      "grad_norm": 3.458984375,
+      "learning_rate": 1.941961307538359e-05,
+      "loss": 1.2661,
+      "step": 12089
+    },
+    {
+      "epoch": 11.160858527579045,
+      "grad_norm": 2.9140625,
+      "learning_rate": 1.9412941961307542e-05,
+      "loss": 1.3826,
+      "step": 12090
+    },
+    {
+      "epoch": 11.1617816755135,
+      "grad_norm": 3.037109375,
+      "learning_rate": 1.940627084723149e-05,
+      "loss": 1.2021,
+      "step": 12091
+    },
+    {
+      "epoch": 11.162704823447957,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.939959973315544e-05,
+      "loss": 1.3081,
+      "step": 12092
+    },
+    {
+      "epoch": 11.163627971382414,
+      "grad_norm": 3.380859375,
+      "learning_rate": 1.9392928619079388e-05,
+      "loss": 1.1362,
+      "step": 12093
+    },
+    {
+      "epoch": 11.16455111931687,
+      "grad_norm": 3.49609375,
+      "learning_rate": 1.9386257505003337e-05,
+      "loss": 1.3035,
+      "step": 12094
+    },
+    {
+      "epoch": 11.165474267251327,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.9379586390927286e-05,
+      "loss": 1.2578,
+      "step": 12095
+    },
+    {
+      "epoch": 11.166397415185784,
+      "grad_norm": 3.166015625,
+      "learning_rate": 1.9372915276851235e-05,
+      "loss": 1.2942,
+      "step": 12096
+    },
+    {
+      "epoch": 11.16732056312024,
+      "grad_norm": 2.728515625,
+      "learning_rate": 1.9366244162775183e-05,
+      "loss": 1.1776,
+      "step": 12097
+    },
+    {
+      "epoch": 11.168243711054696,
+      "grad_norm": 3.705078125,
+      "learning_rate": 1.9359573048699132e-05,
+      "loss": 1.3372,
+      "step": 12098
+    },
+    {
+      "epoch": 11.169166858989152,
+      "grad_norm": 2.98046875,
+      "learning_rate": 1.935290193462308e-05,
+      "loss": 1.0737,
+      "step": 12099
+    },
+    {
+      "epoch": 11.17009000692361,
+      "grad_norm": 3.62109375,
+      "learning_rate": 1.9346230820547033e-05,
+      "loss": 1.1169,
+      "step": 12100
+    },
+    {
+      "epoch": 11.171013154858066,
+      "grad_norm": 3.0234375,
+      "learning_rate": 1.9339559706470982e-05,
+      "loss": 1.3441,
+      "step": 12101
+    },
+    {
+      "epoch": 11.171936302792522,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.933288859239493e-05,
+      "loss": 1.0482,
+      "step": 12102
+    },
+    {
+      "epoch": 11.172859450726978,
+      "grad_norm": 3.14453125,
+      "learning_rate": 1.932621747831888e-05,
+      "loss": 1.2417,
+      "step": 12103
+    },
+    {
+      "epoch": 11.173782598661436,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.931954636424283e-05,
+      "loss": 1.2418,
+      "step": 12104
+    },
+    {
+      "epoch": 11.174705746595892,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.9312875250166777e-05,
+      "loss": 1.2773,
+      "step": 12105
+    },
+    {
+      "epoch": 11.175628894530348,
+      "grad_norm": 3.24609375,
+      "learning_rate": 1.9306204136090726e-05,
+      "loss": 1.2345,
+      "step": 12106
+    },
+    {
+      "epoch": 11.176552042464804,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.9299533022014675e-05,
+      "loss": 1.3386,
+      "step": 12107
+    },
+    {
+      "epoch": 11.177475190399262,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.9292861907938627e-05,
+      "loss": 1.4404,
+      "step": 12108
+    },
+    {
+      "epoch": 11.178398338333718,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.9286190793862576e-05,
+      "loss": 1.3418,
+      "step": 12109
+    },
+    {
+      "epoch": 11.179321486268174,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.9279519679786525e-05,
+      "loss": 1.3735,
+      "step": 12110
+    },
+    {
+      "epoch": 11.180244634202632,
+      "grad_norm": 2.99609375,
+      "learning_rate": 1.9272848565710474e-05,
+      "loss": 1.2834,
+      "step": 12111
+    },
+    {
+      "epoch": 11.181167782137088,
+      "grad_norm": 3.376953125,
+      "learning_rate": 1.9266177451634423e-05,
+      "loss": 1.2537,
+      "step": 12112
+    },
+    {
+      "epoch": 11.182090930071544,
+      "grad_norm": 3.048828125,
+      "learning_rate": 1.925950633755837e-05,
+      "loss": 1.3501,
+      "step": 12113
+    },
+    {
+      "epoch": 11.183014078006,
+      "grad_norm": 3.05078125,
+      "learning_rate": 1.925283522348232e-05,
+      "loss": 1.1926,
+      "step": 12114
+    },
+    {
+      "epoch": 11.183937225940458,
+      "grad_norm": 3.00390625,
+      "learning_rate": 1.9246164109406272e-05,
+      "loss": 1.3306,
+      "step": 12115
+    },
+    {
+      "epoch": 11.184860373874914,
+      "grad_norm": 3.40234375,
+      "learning_rate": 1.923949299533022e-05,
+      "loss": 1.1748,
+      "step": 12116
+    },
+    {
+      "epoch": 11.18578352180937,
+      "grad_norm": 2.833984375,
+      "learning_rate": 1.923282188125417e-05,
+      "loss": 1.4192,
+      "step": 12117
+    },
+    {
+      "epoch": 11.186706669743826,
+      "grad_norm": 3.25,
+      "learning_rate": 1.922615076717812e-05,
+      "loss": 1.3857,
+      "step": 12118
+    },
+    {
+      "epoch": 11.187629817678284,
+      "grad_norm": 3.48046875,
+      "learning_rate": 1.9219479653102068e-05,
+      "loss": 1.1588,
+      "step": 12119
+    },
+    {
+      "epoch": 11.18855296561274,
+      "grad_norm": 2.94921875,
+      "learning_rate": 1.9212808539026016e-05,
+      "loss": 1.4998,
+      "step": 12120
+    },
+    {
+      "epoch": 11.189476113547196,
+      "grad_norm": 3.408203125,
+      "learning_rate": 1.920613742494997e-05,
+      "loss": 1.0144,
+      "step": 12121
+    },
+    {
+      "epoch": 11.190399261481652,
+      "grad_norm": 3.083984375,
+      "learning_rate": 1.9199466310873917e-05,
+      "loss": 1.5181,
+      "step": 12122
+    },
+    {
+      "epoch": 11.19132240941611,
+      "grad_norm": 3.443359375,
+      "learning_rate": 1.9192795196797866e-05,
+      "loss": 1.1702,
+      "step": 12123
+    },
+    {
+      "epoch": 11.192245557350565,
+      "grad_norm": 3.013671875,
+      "learning_rate": 1.9186124082721815e-05,
+      "loss": 1.4949,
+      "step": 12124
+    },
+    {
+      "epoch": 11.193168705285022,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.9179452968645764e-05,
+      "loss": 0.978,
+      "step": 12125
+    },
+    {
+      "epoch": 11.19409185321948,
+      "grad_norm": 4.25,
+      "learning_rate": 1.9172781854569713e-05,
+      "loss": 1.2783,
+      "step": 12126
+    },
+    {
+      "epoch": 11.195015001153935,
+      "grad_norm": 4.08203125,
+      "learning_rate": 1.9166110740493665e-05,
+      "loss": 1.1151,
+      "step": 12127
+    },
+    {
+      "epoch": 11.195938149088391,
+      "grad_norm": 3.5078125,
+      "learning_rate": 1.9159439626417614e-05,
+      "loss": 1.2876,
+      "step": 12128
+    },
+    {
+      "epoch": 11.196861297022847,
+      "grad_norm": 3.708984375,
+      "learning_rate": 1.9152768512341563e-05,
+      "loss": 1.3645,
+      "step": 12129
+    },
+    {
+      "epoch": 11.197784444957305,
+      "grad_norm": 3.68359375,
+      "learning_rate": 1.914609739826551e-05,
+      "loss": 1.2395,
+      "step": 12130
+    },
+    {
+      "epoch": 11.198707592891761,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.913942628418946e-05,
+      "loss": 1.1963,
+      "step": 12131
+    },
+    {
+      "epoch": 11.199630740826217,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.913275517011341e-05,
+      "loss": 1.2915,
+      "step": 12132
+    },
+    {
+      "epoch": 11.200553888760673,
+      "grad_norm": 2.96484375,
+      "learning_rate": 1.912608405603736e-05,
+      "loss": 1.1342,
+      "step": 12133
+    },
+    {
+      "epoch": 11.201477036695131,
+      "grad_norm": 3.44921875,
+      "learning_rate": 1.911941294196131e-05,
+      "loss": 1.1832,
+      "step": 12134
+    },
+    {
+      "epoch": 11.202400184629587,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.911274182788526e-05,
+      "loss": 1.4077,
+      "step": 12135
+    },
+    {
+      "epoch": 11.203323332564043,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.9106070713809208e-05,
+      "loss": 1.2129,
+      "step": 12136
+    },
+    {
+      "epoch": 11.2042464804985,
+      "grad_norm": 2.798828125,
+      "learning_rate": 1.9099399599733157e-05,
+      "loss": 1.4418,
+      "step": 12137
+    },
+    {
+      "epoch": 11.205169628432957,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.9092728485657105e-05,
+      "loss": 1.4204,
+      "step": 12138
+    },
+    {
+      "epoch": 11.206092776367413,
+      "grad_norm": 3.208984375,
+      "learning_rate": 1.9086057371581054e-05,
+      "loss": 1.2083,
+      "step": 12139
+    },
+    {
+      "epoch": 11.207015924301869,
+      "grad_norm": 3.48046875,
+      "learning_rate": 1.9079386257505006e-05,
+      "loss": 1.2837,
+      "step": 12140
+    },
+    {
+      "epoch": 11.207939072236325,
+      "grad_norm": 3.103515625,
+      "learning_rate": 1.9072715143428955e-05,
+      "loss": 1.3657,
+      "step": 12141
+    },
+    {
+      "epoch": 11.208862220170783,
+      "grad_norm": 3.380859375,
+      "learning_rate": 1.9066044029352904e-05,
+      "loss": 1.3567,
+      "step": 12142
+    },
+    {
+      "epoch": 11.209785368105239,
+      "grad_norm": 3.56640625,
+      "learning_rate": 1.9059372915276853e-05,
+      "loss": 1.3608,
+      "step": 12143
+    },
+    {
+      "epoch": 11.210708516039695,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.90527018012008e-05,
+      "loss": 0.9895,
+      "step": 12144
+    },
+    {
+      "epoch": 11.211631663974153,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.904603068712475e-05,
+      "loss": 1.2646,
+      "step": 12145
+    },
+    {
+      "epoch": 11.212554811908609,
+      "grad_norm": 3.0703125,
+      "learning_rate": 1.9039359573048703e-05,
+      "loss": 1.3032,
+      "step": 12146
+    },
+    {
+      "epoch": 11.213477959843065,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.903268845897265e-05,
+      "loss": 1.2238,
+      "step": 12147
+    },
+    {
+      "epoch": 11.21440110777752,
+      "grad_norm": 3.3359375,
+      "learning_rate": 1.90260173448966e-05,
+      "loss": 1.2891,
+      "step": 12148
+    },
+    {
+      "epoch": 11.215324255711979,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.901934623082055e-05,
+      "loss": 1.1569,
+      "step": 12149
+    },
+    {
+      "epoch": 11.216247403646435,
+      "grad_norm": 3.2265625,
+      "learning_rate": 1.9012675116744498e-05,
+      "loss": 1.4263,
+      "step": 12150
+    },
+    {
+      "epoch": 11.21717055158089,
+      "grad_norm": 3.025390625,
+      "learning_rate": 1.9006004002668447e-05,
+      "loss": 1.2446,
+      "step": 12151
+    },
+    {
+      "epoch": 11.218093699515347,
+      "grad_norm": 3.0546875,
+      "learning_rate": 1.8999332888592396e-05,
+      "loss": 1.127,
+      "step": 12152
+    },
+    {
+      "epoch": 11.219016847449804,
+      "grad_norm": 3.603515625,
+      "learning_rate": 1.8992661774516344e-05,
+      "loss": 0.8407,
+      "step": 12153
+    },
+    {
+      "epoch": 11.21993999538426,
+      "grad_norm": 2.919921875,
+      "learning_rate": 1.8985990660440293e-05,
+      "loss": 1.2278,
+      "step": 12154
+    },
+    {
+      "epoch": 11.220863143318716,
+      "grad_norm": 3.4609375,
+      "learning_rate": 1.8979319546364242e-05,
+      "loss": 1.4734,
+      "step": 12155
+    },
+    {
+      "epoch": 11.221786291253173,
+      "grad_norm": 3.416015625,
+      "learning_rate": 1.897264843228819e-05,
+      "loss": 1.022,
+      "step": 12156
+    },
+    {
+      "epoch": 11.22270943918763,
+      "grad_norm": 3.55859375,
+      "learning_rate": 1.8965977318212143e-05,
+      "loss": 1.1305,
+      "step": 12157
+    },
+    {
+      "epoch": 11.223632587122086,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.8959306204136092e-05,
+      "loss": 1.4048,
+      "step": 12158
+    },
+    {
+      "epoch": 11.224555735056542,
+      "grad_norm": 2.998046875,
+      "learning_rate": 1.895263509006004e-05,
+      "loss": 1.1299,
+      "step": 12159
+    },
+    {
+      "epoch": 11.225478882991,
+      "grad_norm": 3.12109375,
+      "learning_rate": 1.894596397598399e-05,
+      "loss": 1.2799,
+      "step": 12160
+    },
+    {
+      "epoch": 11.226402030925456,
+      "grad_norm": 3.0078125,
+      "learning_rate": 1.893929286190794e-05,
+      "loss": 1.1476,
+      "step": 12161
+    },
+    {
+      "epoch": 11.227325178859912,
+      "grad_norm": 3.41796875,
+      "learning_rate": 1.8932621747831887e-05,
+      "loss": 1.1436,
+      "step": 12162
+    },
+    {
+      "epoch": 11.228248326794368,
+      "grad_norm": 3.587890625,
+      "learning_rate": 1.8925950633755836e-05,
+      "loss": 1.4905,
+      "step": 12163
+    },
+    {
+      "epoch": 11.229171474728826,
+      "grad_norm": 3.166015625,
+      "learning_rate": 1.8919279519679785e-05,
+      "loss": 1.3435,
+      "step": 12164
+    },
+    {
+      "epoch": 11.230094622663282,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.8912608405603737e-05,
+      "loss": 1.2178,
+      "step": 12165
+    },
+    {
+      "epoch": 11.231017770597738,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.8905937291527686e-05,
+      "loss": 1.2473,
+      "step": 12166
+    },
+    {
+      "epoch": 11.231940918532194,
+      "grad_norm": 3.2265625,
+      "learning_rate": 1.8899266177451635e-05,
+      "loss": 1.2673,
+      "step": 12167
+    },
+    {
+      "epoch": 11.232864066466652,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.8892595063375583e-05,
+      "loss": 1.3445,
+      "step": 12168
+    },
+    {
+      "epoch": 11.233787214401108,
+      "grad_norm": 4.09765625,
+      "learning_rate": 1.8885923949299532e-05,
+      "loss": 1.251,
+      "step": 12169
+    },
+    {
+      "epoch": 11.234710362335564,
+      "grad_norm": 2.865234375,
+      "learning_rate": 1.887925283522348e-05,
+      "loss": 1.3813,
+      "step": 12170
+    },
+    {
+      "epoch": 11.23563351027002,
+      "grad_norm": 3.189453125,
+      "learning_rate": 1.8872581721147433e-05,
+      "loss": 1.1111,
+      "step": 12171
+    },
+    {
+      "epoch": 11.236556658204478,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.8865910607071382e-05,
+      "loss": 1.3718,
+      "step": 12172
+    },
+    {
+      "epoch": 11.237479806138934,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.885923949299533e-05,
+      "loss": 1.0995,
+      "step": 12173
+    },
+    {
+      "epoch": 11.23840295407339,
+      "grad_norm": 3.45703125,
+      "learning_rate": 1.885256837891928e-05,
+      "loss": 1.1836,
+      "step": 12174
+    },
+    {
+      "epoch": 11.239326102007848,
+      "grad_norm": 2.947265625,
+      "learning_rate": 1.884589726484323e-05,
+      "loss": 1.2739,
+      "step": 12175
+    },
+    {
+      "epoch": 11.240249249942304,
+      "grad_norm": 3.611328125,
+      "learning_rate": 1.8839226150767177e-05,
+      "loss": 1.1848,
+      "step": 12176
+    },
+    {
+      "epoch": 11.24117239787676,
+      "grad_norm": 3.5703125,
+      "learning_rate": 1.883255503669113e-05,
+      "loss": 1.3252,
+      "step": 12177
+    },
+    {
+      "epoch": 11.242095545811216,
+      "grad_norm": 3.2265625,
+      "learning_rate": 1.882588392261508e-05,
+      "loss": 0.9614,
+      "step": 12178
+    },
+    {
+      "epoch": 11.243018693745674,
+      "grad_norm": 3.12109375,
+      "learning_rate": 1.8819212808539027e-05,
+      "loss": 0.68,
+      "step": 12179
+    },
+    {
+      "epoch": 11.24394184168013,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.8812541694462976e-05,
+      "loss": 0.8383,
+      "step": 12180
+    },
+    {
+      "epoch": 11.244864989614586,
+      "grad_norm": 3.326171875,
+      "learning_rate": 1.8805870580386925e-05,
+      "loss": 1.447,
+      "step": 12181
+    },
+    {
+      "epoch": 11.245788137549042,
+      "grad_norm": 3.49609375,
+      "learning_rate": 1.8799199466310874e-05,
+      "loss": 1.2332,
+      "step": 12182
+    },
+    {
+      "epoch": 11.2467112854835,
+      "grad_norm": 3.6015625,
+      "learning_rate": 1.8792528352234823e-05,
+      "loss": 1.3264,
+      "step": 12183
+    },
+    {
+      "epoch": 11.247634433417955,
+      "grad_norm": 3.861328125,
+      "learning_rate": 1.8785857238158775e-05,
+      "loss": 1.2302,
+      "step": 12184
+    },
+    {
+      "epoch": 11.248557581352411,
+      "grad_norm": 3.4609375,
+      "learning_rate": 1.8779186124082724e-05,
+      "loss": 1.4426,
+      "step": 12185
+    },
+    {
+      "epoch": 11.249480729286867,
+      "grad_norm": 3.119140625,
+      "learning_rate": 1.8772515010006672e-05,
+      "loss": 1.1721,
+      "step": 12186
+    },
+    {
+      "epoch": 11.250403877221325,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.876584389593062e-05,
+      "loss": 1.2566,
+      "step": 12187
+    },
+    {
+      "epoch": 11.251327025155781,
+      "grad_norm": 3.36328125,
+      "learning_rate": 1.875917278185457e-05,
+      "loss": 1.2627,
+      "step": 12188
+    },
+    {
+      "epoch": 11.252250173090237,
+      "grad_norm": 3.4921875,
+      "learning_rate": 1.875250166777852e-05,
+      "loss": 1.4346,
+      "step": 12189
+    },
+    {
+      "epoch": 11.253173321024693,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.874583055370247e-05,
+      "loss": 1.3677,
+      "step": 12190
+    },
+    {
+      "epoch": 11.254096468959151,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.873915943962642e-05,
+      "loss": 1.3669,
+      "step": 12191
+    },
+    {
+      "epoch": 11.255019616893607,
+      "grad_norm": 2.95703125,
+      "learning_rate": 1.873248832555037e-05,
+      "loss": 1.4358,
+      "step": 12192
+    },
+    {
+      "epoch": 11.255942764828063,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.8725817211474317e-05,
+      "loss": 1.2944,
+      "step": 12193
+    },
+    {
+      "epoch": 11.256865912762521,
+      "grad_norm": 2.986328125,
+      "learning_rate": 1.8719146097398266e-05,
+      "loss": 1.4714,
+      "step": 12194
+    },
+    {
+      "epoch": 11.257789060696977,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.8712474983322215e-05,
+      "loss": 1.332,
+      "step": 12195
+    },
+    {
+      "epoch": 11.258712208631433,
+      "grad_norm": 2.48046875,
+      "learning_rate": 1.8705803869246167e-05,
+      "loss": 1.1538,
+      "step": 12196
+    },
+    {
+      "epoch": 11.259635356565889,
+      "grad_norm": 3.095703125,
+      "learning_rate": 1.8699132755170116e-05,
+      "loss": 1.3545,
+      "step": 12197
+    },
+    {
+      "epoch": 11.260558504500347,
+      "grad_norm": 3.736328125,
+      "learning_rate": 1.8692461641094065e-05,
+      "loss": 1.4158,
+      "step": 12198
+    },
+    {
+      "epoch": 11.261481652434803,
+      "grad_norm": 3.025390625,
+      "learning_rate": 1.8685790527018014e-05,
+      "loss": 1.4016,
+      "step": 12199
+    },
+    {
+      "epoch": 11.262404800369259,
+      "grad_norm": 3.31640625,
+      "learning_rate": 1.8679119412941963e-05,
+      "loss": 1.3257,
+      "step": 12200
+    },
+    {
+      "epoch": 11.263327948303715,
+      "grad_norm": 3.5,
+      "learning_rate": 1.867244829886591e-05,
+      "loss": 1.4167,
+      "step": 12201
+    },
+    {
+      "epoch": 11.264251096238173,
+      "grad_norm": 3.005859375,
+      "learning_rate": 1.8665777184789864e-05,
+      "loss": 1.2843,
+      "step": 12202
+    },
+    {
+      "epoch": 11.265174244172629,
+      "grad_norm": 2.951171875,
+      "learning_rate": 1.8659106070713812e-05,
+      "loss": 1.271,
+      "step": 12203
+    },
+    {
+      "epoch": 11.266097392107085,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.865243495663776e-05,
+      "loss": 1.304,
+      "step": 12204
+    },
+    {
+      "epoch": 11.26702054004154,
+      "grad_norm": 3.630859375,
+      "learning_rate": 1.864576384256171e-05,
+      "loss": 1.2051,
+      "step": 12205
+    },
+    {
+      "epoch": 11.267943687975999,
+      "grad_norm": 2.951171875,
+      "learning_rate": 1.863909272848566e-05,
+      "loss": 1.2214,
+      "step": 12206
+    },
+    {
+      "epoch": 11.268866835910455,
+      "grad_norm": 2.8828125,
+      "learning_rate": 1.8632421614409608e-05,
+      "loss": 1.3557,
+      "step": 12207
+    },
+    {
+      "epoch": 11.26978998384491,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.8625750500333557e-05,
+      "loss": 1.335,
+      "step": 12208
+    },
+    {
+      "epoch": 11.270713131779367,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.8619079386257505e-05,
+      "loss": 1.0385,
+      "step": 12209
+    },
+    {
+      "epoch": 11.271636279713825,
+      "grad_norm": 3.091796875,
+      "learning_rate": 1.8612408272181454e-05,
+      "loss": 1.1306,
+      "step": 12210
+    },
+    {
+      "epoch": 11.27255942764828,
+      "grad_norm": 3.451171875,
+      "learning_rate": 1.8605737158105403e-05,
+      "loss": 1.1527,
+      "step": 12211
+    },
+    {
+      "epoch": 11.273482575582737,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.8599066044029352e-05,
+      "loss": 1.491,
+      "step": 12212
+    },
+    {
+      "epoch": 11.274405723517194,
+      "grad_norm": 3.10546875,
+      "learning_rate": 1.8592394929953304e-05,
+      "loss": 1.2428,
+      "step": 12213
+    },
+    {
+      "epoch": 11.27532887145165,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.8585723815877253e-05,
+      "loss": 1.1187,
+      "step": 12214
+    },
+    {
+      "epoch": 11.276252019386106,
+      "grad_norm": 2.873046875,
+      "learning_rate": 1.85790527018012e-05,
+      "loss": 1.2479,
+      "step": 12215
+    },
+    {
+      "epoch": 11.277175167320562,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.857238158772515e-05,
+      "loss": 1.1665,
+      "step": 12216
+    },
+    {
+      "epoch": 11.27809831525502,
+      "grad_norm": 2.912109375,
+      "learning_rate": 1.85657104736491e-05,
+      "loss": 1.3364,
+      "step": 12217
+    },
+    {
+      "epoch": 11.279021463189476,
+      "grad_norm": 2.884765625,
+      "learning_rate": 1.8559039359573048e-05,
+      "loss": 0.8888,
+      "step": 12218
+    },
+    {
+      "epoch": 11.279944611123932,
+      "grad_norm": 3.416015625,
+      "learning_rate": 1.8552368245496997e-05,
+      "loss": 1.3105,
+      "step": 12219
+    },
+    {
+      "epoch": 11.280867759058388,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.8545697131420946e-05,
+      "loss": 1.4326,
+      "step": 12220
+    },
+    {
+      "epoch": 11.281790906992846,
+      "grad_norm": 4.25390625,
+      "learning_rate": 1.8539026017344898e-05,
+      "loss": 1.2151,
+      "step": 12221
+    },
+    {
+      "epoch": 11.282714054927302,
+      "grad_norm": 3.25,
+      "learning_rate": 1.8532354903268847e-05,
+      "loss": 0.9618,
+      "step": 12222
+    },
+    {
+      "epoch": 11.283637202861758,
+      "grad_norm": 3.630859375,
+      "learning_rate": 1.8525683789192796e-05,
+      "loss": 1.3643,
+      "step": 12223
+    },
+    {
+      "epoch": 11.284560350796216,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.8519012675116744e-05,
+      "loss": 1.1732,
+      "step": 12224
+    },
+    {
+      "epoch": 11.285483498730672,
+      "grad_norm": 3.4453125,
+      "learning_rate": 1.8512341561040693e-05,
+      "loss": 1.3096,
+      "step": 12225
+    },
+    {
+      "epoch": 11.286406646665128,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.8505670446964642e-05,
+      "loss": 1.2483,
+      "step": 12226
+    },
+    {
+      "epoch": 11.287329794599584,
+      "grad_norm": 2.986328125,
+      "learning_rate": 1.8498999332888594e-05,
+      "loss": 1.3672,
+      "step": 12227
+    },
+    {
+      "epoch": 11.288252942534042,
+      "grad_norm": 3.61328125,
+      "learning_rate": 1.8492328218812543e-05,
+      "loss": 1.2871,
+      "step": 12228
+    },
+    {
+      "epoch": 11.289176090468498,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.8485657104736492e-05,
+      "loss": 1.2461,
+      "step": 12229
+    },
+    {
+      "epoch": 11.290099238402954,
+      "grad_norm": 3.31640625,
+      "learning_rate": 1.847898599066044e-05,
+      "loss": 1.2703,
+      "step": 12230
+    },
+    {
+      "epoch": 11.29102238633741,
+      "grad_norm": 3.630859375,
+      "learning_rate": 1.847231487658439e-05,
+      "loss": 1.0883,
+      "step": 12231
+    },
+    {
+      "epoch": 11.291945534271868,
+      "grad_norm": 3.04296875,
+      "learning_rate": 1.846564376250834e-05,
+      "loss": 1.2375,
+      "step": 12232
+    },
+    {
+      "epoch": 11.292868682206324,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.8458972648432287e-05,
+      "loss": 1.3274,
+      "step": 12233
+    },
+    {
+      "epoch": 11.29379183014078,
+      "grad_norm": 3.349609375,
+      "learning_rate": 1.845230153435624e-05,
+      "loss": 1.1965,
+      "step": 12234
+    },
+    {
+      "epoch": 11.294714978075236,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.8445630420280188e-05,
+      "loss": 1.1097,
+      "step": 12235
+    },
+    {
+      "epoch": 11.295638126009694,
+      "grad_norm": 3.404296875,
+      "learning_rate": 1.8438959306204137e-05,
+      "loss": 1.1298,
+      "step": 12236
+    },
+    {
+      "epoch": 11.29656127394415,
+      "grad_norm": 3.314453125,
+      "learning_rate": 1.8432288192128086e-05,
+      "loss": 1.3003,
+      "step": 12237
+    },
+    {
+      "epoch": 11.297484421878606,
+      "grad_norm": 3.576171875,
+      "learning_rate": 1.8425617078052035e-05,
+      "loss": 1.1049,
+      "step": 12238
+    },
+    {
+      "epoch": 11.298407569813062,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.8418945963975983e-05,
+      "loss": 1.1157,
+      "step": 12239
+    },
+    {
+      "epoch": 11.29933071774752,
+      "grad_norm": 3.2421875,
+      "learning_rate": 1.8412274849899936e-05,
+      "loss": 1.5034,
+      "step": 12240
+    },
+    {
+      "epoch": 11.300253865681976,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.8405603735823884e-05,
+      "loss": 1.2866,
+      "step": 12241
+    },
+    {
+      "epoch": 11.301177013616432,
+      "grad_norm": 3.47265625,
+      "learning_rate": 1.8398932621747833e-05,
+      "loss": 1.4749,
+      "step": 12242
+    },
+    {
+      "epoch": 11.30210016155089,
+      "grad_norm": 3.333984375,
+      "learning_rate": 1.8392261507671782e-05,
+      "loss": 1.2954,
+      "step": 12243
+    },
+    {
+      "epoch": 11.303023309485345,
+      "grad_norm": 3.38671875,
+      "learning_rate": 1.838559039359573e-05,
+      "loss": 1.2758,
+      "step": 12244
+    },
+    {
+      "epoch": 11.303946457419801,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.837891927951968e-05,
+      "loss": 1.2842,
+      "step": 12245
+    },
+    {
+      "epoch": 11.304869605354257,
+      "grad_norm": 3.23828125,
+      "learning_rate": 1.8372248165443632e-05,
+      "loss": 1.4948,
+      "step": 12246
+    },
+    {
+      "epoch": 11.305792753288715,
+      "grad_norm": 3.64453125,
+      "learning_rate": 1.836557705136758e-05,
+      "loss": 1.3545,
+      "step": 12247
+    },
+    {
+      "epoch": 11.306715901223171,
+      "grad_norm": 2.96484375,
+      "learning_rate": 1.835890593729153e-05,
+      "loss": 1.2067,
+      "step": 12248
+    },
+    {
+      "epoch": 11.307639049157627,
+      "grad_norm": 3.5234375,
+      "learning_rate": 1.835223482321548e-05,
+      "loss": 1.3926,
+      "step": 12249
+    },
+    {
+      "epoch": 11.308562197092083,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.8345563709139427e-05,
+      "loss": 1.303,
+      "step": 12250
+    },
+    {
+      "epoch": 11.309485345026541,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8338892595063376e-05,
+      "loss": 1.1042,
+      "step": 12251
+    },
+    {
+      "epoch": 11.310408492960997,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.8332221480987325e-05,
+      "loss": 1.3887,
+      "step": 12252
+    },
+    {
+      "epoch": 11.311331640895453,
+      "grad_norm": 3.05078125,
+      "learning_rate": 1.8325550366911277e-05,
+      "loss": 1.335,
+      "step": 12253
+    },
+    {
+      "epoch": 11.31225478882991,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.8318879252835226e-05,
+      "loss": 1.3104,
+      "step": 12254
+    },
+    {
+      "epoch": 11.313177936764367,
+      "grad_norm": 3.126953125,
+      "learning_rate": 1.8312208138759175e-05,
+      "loss": 1.3123,
+      "step": 12255
+    },
+    {
+      "epoch": 11.314101084698823,
+      "grad_norm": 2.9765625,
+      "learning_rate": 1.8305537024683124e-05,
+      "loss": 1.2356,
+      "step": 12256
+    },
+    {
+      "epoch": 11.315024232633279,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.8298865910607072e-05,
+      "loss": 1.3018,
+      "step": 12257
+    },
+    {
+      "epoch": 11.315947380567735,
+      "grad_norm": 3.068359375,
+      "learning_rate": 1.829219479653102e-05,
+      "loss": 1.2054,
+      "step": 12258
+    },
+    {
+      "epoch": 11.316870528502193,
+      "grad_norm": 3.62109375,
+      "learning_rate": 1.8285523682454973e-05,
+      "loss": 1.146,
+      "step": 12259
+    },
+    {
+      "epoch": 11.317793676436649,
+      "grad_norm": 3.220703125,
+      "learning_rate": 1.8278852568378922e-05,
+      "loss": 1.3188,
+      "step": 12260
+    },
+    {
+      "epoch": 11.318716824371105,
+      "grad_norm": 3.0078125,
+      "learning_rate": 1.827218145430287e-05,
+      "loss": 1.1411,
+      "step": 12261
+    },
+    {
+      "epoch": 11.319639972305563,
+      "grad_norm": 3.126953125,
+      "learning_rate": 1.826551034022682e-05,
+      "loss": 1.3311,
+      "step": 12262
+    },
+    {
+      "epoch": 11.320563120240019,
+      "grad_norm": 3.142578125,
+      "learning_rate": 1.825883922615077e-05,
+      "loss": 1.335,
+      "step": 12263
+    },
+    {
+      "epoch": 11.321486268174475,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.8252168112074717e-05,
+      "loss": 1.3333,
+      "step": 12264
+    },
+    {
+      "epoch": 11.32240941610893,
+      "grad_norm": 3.04296875,
+      "learning_rate": 1.8245496997998666e-05,
+      "loss": 1.3906,
+      "step": 12265
+    },
+    {
+      "epoch": 11.323332564043389,
+      "grad_norm": 3.4140625,
+      "learning_rate": 1.8238825883922615e-05,
+      "loss": 1.104,
+      "step": 12266
+    },
+    {
+      "epoch": 11.324255711977845,
+      "grad_norm": 3.232421875,
+      "learning_rate": 1.8232154769846564e-05,
+      "loss": 1.1897,
+      "step": 12267
+    },
+    {
+      "epoch": 11.3251788599123,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.8225483655770513e-05,
+      "loss": 1.4482,
+      "step": 12268
+    },
+    {
+      "epoch": 11.326102007846757,
+      "grad_norm": 3.482421875,
+      "learning_rate": 1.821881254169446e-05,
+      "loss": 1.5146,
+      "step": 12269
+    },
+    {
+      "epoch": 11.327025155781214,
+      "grad_norm": 3.517578125,
+      "learning_rate": 1.8212141427618414e-05,
+      "loss": 1.2035,
+      "step": 12270
+    },
+    {
+      "epoch": 11.32794830371567,
+      "grad_norm": 2.95703125,
+      "learning_rate": 1.8205470313542363e-05,
+      "loss": 1.3284,
+      "step": 12271
+    },
+    {
+      "epoch": 11.328871451650127,
+      "grad_norm": 3.0703125,
+      "learning_rate": 1.819879919946631e-05,
+      "loss": 1.2008,
+      "step": 12272
+    },
+    {
+      "epoch": 11.329794599584583,
+      "grad_norm": 3.5,
+      "learning_rate": 1.819212808539026e-05,
+      "loss": 1.2694,
+      "step": 12273
+    },
+    {
+      "epoch": 11.33071774751904,
+      "grad_norm": 3.150390625,
+      "learning_rate": 1.818545697131421e-05,
+      "loss": 1.2552,
+      "step": 12274
+    },
+    {
+      "epoch": 11.331640895453496,
+      "grad_norm": 2.814453125,
+      "learning_rate": 1.8178785857238158e-05,
+      "loss": 1.1416,
+      "step": 12275
+    },
+    {
+      "epoch": 11.332564043387952,
+      "grad_norm": 3.345703125,
+      "learning_rate": 1.8172114743162107e-05,
+      "loss": 1.1891,
+      "step": 12276
+    },
+    {
+      "epoch": 11.33348719132241,
+      "grad_norm": 3.2578125,
+      "learning_rate": 1.8165443629086056e-05,
+      "loss": 1.231,
+      "step": 12277
+    },
+    {
+      "epoch": 11.334410339256866,
+      "grad_norm": 3.08203125,
+      "learning_rate": 1.8158772515010008e-05,
+      "loss": 1.1832,
+      "step": 12278
+    },
+    {
+      "epoch": 11.335333487191322,
+      "grad_norm": 2.935546875,
+      "learning_rate": 1.8152101400933957e-05,
+      "loss": 1.4019,
+      "step": 12279
+    },
+    {
+      "epoch": 11.336256635125778,
+      "grad_norm": 3.37890625,
+      "learning_rate": 1.8145430286857905e-05,
+      "loss": 1.088,
+      "step": 12280
+    },
+    {
+      "epoch": 11.337179783060236,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.8138759172781854e-05,
+      "loss": 1.1532,
+      "step": 12281
+    },
+    {
+      "epoch": 11.338102930994692,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.8132088058705803e-05,
+      "loss": 1.3291,
+      "step": 12282
+    },
+    {
+      "epoch": 11.339026078929148,
+      "grad_norm": 3.0,
+      "learning_rate": 1.8125416944629752e-05,
+      "loss": 1.1589,
+      "step": 12283
+    },
+    {
+      "epoch": 11.339949226863604,
+      "grad_norm": 4.08984375,
+      "learning_rate": 1.8118745830553704e-05,
+      "loss": 1.1218,
+      "step": 12284
+    },
+    {
+      "epoch": 11.340872374798062,
+      "grad_norm": 3.361328125,
+      "learning_rate": 1.8112074716477653e-05,
+      "loss": 1.3427,
+      "step": 12285
+    },
+    {
+      "epoch": 11.341795522732518,
+      "grad_norm": 3.083984375,
+      "learning_rate": 1.81054036024016e-05,
+      "loss": 1.2786,
+      "step": 12286
+    },
+    {
+      "epoch": 11.342718670666974,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.809873248832555e-05,
+      "loss": 1.1698,
+      "step": 12287
+    },
+    {
+      "epoch": 11.34364181860143,
+      "grad_norm": 3.056640625,
+      "learning_rate": 1.80920613742495e-05,
+      "loss": 1.1163,
+      "step": 12288
+    },
+    {
+      "epoch": 11.344564966535888,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.8085390260173448e-05,
+      "loss": 1.4021,
+      "step": 12289
+    },
+    {
+      "epoch": 11.345488114470344,
+      "grad_norm": 3.32421875,
+      "learning_rate": 1.80787191460974e-05,
+      "loss": 1.2657,
+      "step": 12290
+    },
+    {
+      "epoch": 11.3464112624048,
+      "grad_norm": 3.626953125,
+      "learning_rate": 1.807204803202135e-05,
+      "loss": 1.2756,
+      "step": 12291
+    },
+    {
+      "epoch": 11.347334410339258,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.8065376917945298e-05,
+      "loss": 1.1228,
+      "step": 12292
+    },
+    {
+      "epoch": 11.348257558273714,
+      "grad_norm": 3.33984375,
+      "learning_rate": 1.8058705803869247e-05,
+      "loss": 0.9924,
+      "step": 12293
+    },
+    {
+      "epoch": 11.34918070620817,
+      "grad_norm": 2.94140625,
+      "learning_rate": 1.8052034689793196e-05,
+      "loss": 1.3779,
+      "step": 12294
+    },
+    {
+      "epoch": 11.350103854142626,
+      "grad_norm": 3.01171875,
+      "learning_rate": 1.8045363575717144e-05,
+      "loss": 1.3325,
+      "step": 12295
+    },
+    {
+      "epoch": 11.351027002077084,
+      "grad_norm": 3.076171875,
+      "learning_rate": 1.8038692461641097e-05,
+      "loss": 1.3489,
+      "step": 12296
+    },
+    {
+      "epoch": 11.35195015001154,
+      "grad_norm": 3.0078125,
+      "learning_rate": 1.8032021347565045e-05,
+      "loss": 1.3452,
+      "step": 12297
+    },
+    {
+      "epoch": 11.352873297945996,
+      "grad_norm": 3.306640625,
+      "learning_rate": 1.8025350233488994e-05,
+      "loss": 1.2263,
+      "step": 12298
+    },
+    {
+      "epoch": 11.353796445880452,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.8018679119412943e-05,
+      "loss": 1.2485,
+      "step": 12299
+    },
+    {
+      "epoch": 11.35471959381491,
+      "grad_norm": 3.125,
+      "learning_rate": 1.8012008005336892e-05,
+      "loss": 1.3057,
+      "step": 12300
+    },
+    {
+      "epoch": 11.355642741749365,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.800533689126084e-05,
+      "loss": 1.2517,
+      "step": 12301
+    },
+    {
+      "epoch": 11.356565889683822,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.799866577718479e-05,
+      "loss": 1.2812,
+      "step": 12302
+    },
+    {
+      "epoch": 11.357489037618278,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.7991994663108742e-05,
+      "loss": 1.4751,
+      "step": 12303
+    },
+    {
+      "epoch": 11.358412185552735,
+      "grad_norm": 3.591796875,
+      "learning_rate": 1.798532354903269e-05,
+      "loss": 1.3979,
+      "step": 12304
+    },
+    {
+      "epoch": 11.359335333487191,
+      "grad_norm": 2.861328125,
+      "learning_rate": 1.797865243495664e-05,
+      "loss": 1.2449,
+      "step": 12305
+    },
+    {
+      "epoch": 11.360258481421647,
+      "grad_norm": 3.3515625,
+      "learning_rate": 1.7971981320880588e-05,
+      "loss": 1.542,
+      "step": 12306
+    },
+    {
+      "epoch": 11.361181629356103,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.7965310206804537e-05,
+      "loss": 1.137,
+      "step": 12307
+    },
+    {
+      "epoch": 11.362104777290561,
+      "grad_norm": 3.31640625,
+      "learning_rate": 1.7958639092728486e-05,
+      "loss": 1.0131,
+      "step": 12308
+    },
+    {
+      "epoch": 11.363027925225017,
+      "grad_norm": 3.599609375,
+      "learning_rate": 1.7951967978652438e-05,
+      "loss": 1.2078,
+      "step": 12309
+    },
+    {
+      "epoch": 11.363951073159473,
+      "grad_norm": 3.255859375,
+      "learning_rate": 1.7945296864576387e-05,
+      "loss": 1.3898,
+      "step": 12310
+    },
+    {
+      "epoch": 11.364874221093931,
+      "grad_norm": 3.509765625,
+      "learning_rate": 1.7938625750500336e-05,
+      "loss": 1.0936,
+      "step": 12311
+    },
+    {
+      "epoch": 11.365797369028387,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.7931954636424285e-05,
+      "loss": 1.1355,
+      "step": 12312
+    },
+    {
+      "epoch": 11.366720516962843,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.7925283522348233e-05,
+      "loss": 1.2063,
+      "step": 12313
+    },
+    {
+      "epoch": 11.3676436648973,
+      "grad_norm": 3.375,
+      "learning_rate": 1.7918612408272182e-05,
+      "loss": 1.1248,
+      "step": 12314
+    },
+    {
+      "epoch": 11.368566812831757,
+      "grad_norm": 3.298828125,
+      "learning_rate": 1.7911941294196134e-05,
+      "loss": 1.3015,
+      "step": 12315
+    },
+    {
+      "epoch": 11.369489960766213,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.7905270180120083e-05,
+      "loss": 1.4702,
+      "step": 12316
+    },
+    {
+      "epoch": 11.370413108700669,
+      "grad_norm": 3.037109375,
+      "learning_rate": 1.7898599066044032e-05,
+      "loss": 1.1217,
+      "step": 12317
+    },
+    {
+      "epoch": 11.371336256635125,
+      "grad_norm": 3.521484375,
+      "learning_rate": 1.789192795196798e-05,
+      "loss": 1.3528,
+      "step": 12318
+    },
+    {
+      "epoch": 11.372259404569583,
+      "grad_norm": 3.048828125,
+      "learning_rate": 1.788525683789193e-05,
+      "loss": 0.9651,
+      "step": 12319
+    },
+    {
+      "epoch": 11.373182552504039,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.787858572381588e-05,
+      "loss": 1.3479,
+      "step": 12320
+    },
+    {
+      "epoch": 11.374105700438495,
+      "grad_norm": 3.587890625,
+      "learning_rate": 1.7871914609739827e-05,
+      "loss": 1.0398,
+      "step": 12321
+    },
+    {
+      "epoch": 11.375028848372951,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.7865243495663776e-05,
+      "loss": 1.3132,
+      "step": 12322
+    },
+    {
+      "epoch": 11.375951996307409,
+      "grad_norm": 3.22265625,
+      "learning_rate": 1.7858572381587725e-05,
+      "loss": 1.3569,
+      "step": 12323
+    },
+    {
+      "epoch": 11.376875144241865,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.7851901267511674e-05,
+      "loss": 1.2786,
+      "step": 12324
+    },
+    {
+      "epoch": 11.37779829217632,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.7845230153435623e-05,
+      "loss": 1.3896,
+      "step": 12325
+    },
+    {
+      "epoch": 11.378721440110779,
+      "grad_norm": 3.591796875,
+      "learning_rate": 1.7838559039359575e-05,
+      "loss": 1.27,
+      "step": 12326
+    },
+    {
+      "epoch": 11.379644588045235,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.7831887925283524e-05,
+      "loss": 1.5466,
+      "step": 12327
+    },
+    {
+      "epoch": 11.38056773597969,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.7825216811207472e-05,
+      "loss": 1.5022,
+      "step": 12328
+    },
+    {
+      "epoch": 11.381490883914147,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.781854569713142e-05,
+      "loss": 1.3616,
+      "step": 12329
+    },
+    {
+      "epoch": 11.382414031848604,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.781187458305537e-05,
+      "loss": 1.1653,
+      "step": 12330
+    },
+    {
+      "epoch": 11.38333717978306,
+      "grad_norm": 2.94921875,
+      "learning_rate": 1.780520346897932e-05,
+      "loss": 1.2708,
+      "step": 12331
+    },
+    {
+      "epoch": 11.384260327717517,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.7798532354903268e-05,
+      "loss": 1.3137,
+      "step": 12332
+    },
+    {
+      "epoch": 11.385183475651973,
+      "grad_norm": 3.013671875,
+      "learning_rate": 1.7791861240827216e-05,
+      "loss": 1.2793,
+      "step": 12333
+    },
+    {
+      "epoch": 11.38610662358643,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.778519012675117e-05,
+      "loss": 1.1211,
+      "step": 12334
+    },
+    {
+      "epoch": 11.387029771520886,
+      "grad_norm": 3.01953125,
+      "learning_rate": 1.7778519012675117e-05,
+      "loss": 1.1956,
+      "step": 12335
+    },
+    {
+      "epoch": 11.387952919455342,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.7771847898599066e-05,
+      "loss": 1.2153,
+      "step": 12336
+    },
+    {
+      "epoch": 11.388876067389798,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.7765176784523015e-05,
+      "loss": 1.4927,
+      "step": 12337
+    },
+    {
+      "epoch": 11.389799215324256,
+      "grad_norm": 3.080078125,
+      "learning_rate": 1.7758505670446964e-05,
+      "loss": 1.3191,
+      "step": 12338
+    },
+    {
+      "epoch": 11.390722363258712,
+      "grad_norm": 3.5,
+      "learning_rate": 1.7751834556370913e-05,
+      "loss": 1.3096,
+      "step": 12339
+    },
+    {
+      "epoch": 11.391645511193168,
+      "grad_norm": 3.158203125,
+      "learning_rate": 1.7745163442294865e-05,
+      "loss": 1.3674,
+      "step": 12340
+    },
+    {
+      "epoch": 11.392568659127626,
+      "grad_norm": 3.3515625,
+      "learning_rate": 1.7738492328218814e-05,
+      "loss": 1.0881,
+      "step": 12341
+    },
+    {
+      "epoch": 11.393491807062082,
+      "grad_norm": 2.931640625,
+      "learning_rate": 1.7731821214142763e-05,
+      "loss": 1.3867,
+      "step": 12342
+    },
+    {
+      "epoch": 11.394414954996538,
+      "grad_norm": 2.970703125,
+      "learning_rate": 1.772515010006671e-05,
+      "loss": 1.4661,
+      "step": 12343
+    },
+    {
+      "epoch": 11.395338102930994,
+      "grad_norm": 3.2421875,
+      "learning_rate": 1.771847898599066e-05,
+      "loss": 1.1481,
+      "step": 12344
+    },
+    {
+      "epoch": 11.396261250865452,
+      "grad_norm": 2.876953125,
+      "learning_rate": 1.771180787191461e-05,
+      "loss": 1.3596,
+      "step": 12345
+    },
+    {
+      "epoch": 11.397184398799908,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.7705136757838558e-05,
+      "loss": 1.5315,
+      "step": 12346
+    },
+    {
+      "epoch": 11.398107546734364,
+      "grad_norm": 3.509765625,
+      "learning_rate": 1.769846564376251e-05,
+      "loss": 1.1604,
+      "step": 12347
+    },
+    {
+      "epoch": 11.39903069466882,
+      "grad_norm": 3.232421875,
+      "learning_rate": 1.769179452968646e-05,
+      "loss": 1.5593,
+      "step": 12348
+    },
+    {
+      "epoch": 11.399953842603278,
+      "grad_norm": 3.30078125,
+      "learning_rate": 1.7685123415610408e-05,
+      "loss": 1.4078,
+      "step": 12349
+    },
+    {
+      "epoch": 11.400876990537734,
+      "grad_norm": 3.37109375,
+      "learning_rate": 1.7678452301534357e-05,
+      "loss": 1.3347,
+      "step": 12350
+    },
+    {
+      "epoch": 11.40180013847219,
+      "grad_norm": 2.93359375,
+      "learning_rate": 1.7671781187458305e-05,
+      "loss": 1.0546,
+      "step": 12351
+    },
+    {
+      "epoch": 11.402723286406646,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.7665110073382254e-05,
+      "loss": 1.2974,
+      "step": 12352
+    },
+    {
+      "epoch": 11.403646434341104,
+      "grad_norm": 2.970703125,
+      "learning_rate": 1.7658438959306206e-05,
+      "loss": 1.3406,
+      "step": 12353
+    },
+    {
+      "epoch": 11.40456958227556,
+      "grad_norm": 3.22265625,
+      "learning_rate": 1.7651767845230155e-05,
+      "loss": 1.3987,
+      "step": 12354
+    },
+    {
+      "epoch": 11.405492730210016,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.7645096731154104e-05,
+      "loss": 1.2242,
+      "step": 12355
+    },
+    {
+      "epoch": 11.406415878144472,
+      "grad_norm": 3.255859375,
+      "learning_rate": 1.7638425617078053e-05,
+      "loss": 1.0775,
+      "step": 12356
+    },
+    {
+      "epoch": 11.40733902607893,
+      "grad_norm": 3.04296875,
+      "learning_rate": 1.7631754503002e-05,
+      "loss": 1.3264,
+      "step": 12357
+    },
+    {
+      "epoch": 11.408262174013386,
+      "grad_norm": 3.0859375,
+      "learning_rate": 1.762508338892595e-05,
+      "loss": 1.416,
+      "step": 12358
+    },
+    {
+      "epoch": 11.409185321947842,
+      "grad_norm": 2.76171875,
+      "learning_rate": 1.7618412274849903e-05,
+      "loss": 1.1993,
+      "step": 12359
+    },
+    {
+      "epoch": 11.4101084698823,
+      "grad_norm": 3.31640625,
+      "learning_rate": 1.761174116077385e-05,
+      "loss": 1.3677,
+      "step": 12360
+    },
+    {
+      "epoch": 11.411031617816755,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.76050700466978e-05,
+      "loss": 1.3501,
+      "step": 12361
+    },
+    {
+      "epoch": 11.411954765751211,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.759839893262175e-05,
+      "loss": 0.7393,
+      "step": 12362
+    },
+    {
+      "epoch": 11.412877913685668,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.7591727818545698e-05,
+      "loss": 1.2496,
+      "step": 12363
+    },
+    {
+      "epoch": 11.413801061620125,
+      "grad_norm": 2.955078125,
+      "learning_rate": 1.7585056704469647e-05,
+      "loss": 1.1997,
+      "step": 12364
+    },
+    {
+      "epoch": 11.414724209554581,
+      "grad_norm": 3.6953125,
+      "learning_rate": 1.75783855903936e-05,
+      "loss": 0.9947,
+      "step": 12365
+    },
+    {
+      "epoch": 11.415647357489037,
+      "grad_norm": 3.486328125,
+      "learning_rate": 1.7571714476317548e-05,
+      "loss": 1.3408,
+      "step": 12366
+    },
+    {
+      "epoch": 11.416570505423493,
+      "grad_norm": 3.576171875,
+      "learning_rate": 1.7565043362241497e-05,
+      "loss": 1.3191,
+      "step": 12367
+    },
+    {
+      "epoch": 11.417493653357951,
+      "grad_norm": 3.07421875,
+      "learning_rate": 1.7558372248165445e-05,
+      "loss": 1.1626,
+      "step": 12368
+    },
+    {
+      "epoch": 11.418416801292407,
+      "grad_norm": 3.529296875,
+      "learning_rate": 1.7551701134089394e-05,
+      "loss": 1.2845,
+      "step": 12369
+    },
+    {
+      "epoch": 11.419339949226863,
+      "grad_norm": 3.412109375,
+      "learning_rate": 1.7545030020013343e-05,
+      "loss": 1.0479,
+      "step": 12370
+    },
+    {
+      "epoch": 11.42026309716132,
+      "grad_norm": 3.2421875,
+      "learning_rate": 1.7538358905937292e-05,
+      "loss": 1.0963,
+      "step": 12371
+    },
+    {
+      "epoch": 11.421186245095777,
+      "grad_norm": 3.447265625,
+      "learning_rate": 1.7531687791861244e-05,
+      "loss": 1.0592,
+      "step": 12372
+    },
+    {
+      "epoch": 11.422109393030233,
+      "grad_norm": 3.15234375,
+      "learning_rate": 1.7525016677785193e-05,
+      "loss": 1.4146,
+      "step": 12373
+    },
+    {
+      "epoch": 11.42303254096469,
+      "grad_norm": 2.82421875,
+      "learning_rate": 1.7518345563709142e-05,
+      "loss": 1.3424,
+      "step": 12374
+    },
+    {
+      "epoch": 11.423955688899147,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.751167444963309e-05,
+      "loss": 1.4163,
+      "step": 12375
+    },
+    {
+      "epoch": 11.424878836833603,
+      "grad_norm": 3.248046875,
+      "learning_rate": 1.750500333555704e-05,
+      "loss": 1.2869,
+      "step": 12376
+    },
+    {
+      "epoch": 11.425801984768059,
+      "grad_norm": 3.560546875,
+      "learning_rate": 1.7498332221480988e-05,
+      "loss": 1.3191,
+      "step": 12377
+    },
+    {
+      "epoch": 11.426725132702515,
+      "grad_norm": 3.716796875,
+      "learning_rate": 1.7491661107404937e-05,
+      "loss": 1.1978,
+      "step": 12378
+    },
+    {
+      "epoch": 11.427648280636973,
+      "grad_norm": 3.37109375,
+      "learning_rate": 1.7484989993328886e-05,
+      "loss": 1.1786,
+      "step": 12379
+    },
+    {
+      "epoch": 11.428571428571429,
+      "grad_norm": 2.646484375,
+      "learning_rate": 1.7478318879252835e-05,
+      "loss": 1.0844,
+      "step": 12380
+    },
+    {
+      "epoch": 11.429494576505885,
+      "grad_norm": 3.69921875,
+      "learning_rate": 1.7471647765176783e-05,
+      "loss": 1.2917,
+      "step": 12381
+    },
+    {
+      "epoch": 11.43041772444034,
+      "grad_norm": 2.97265625,
+      "learning_rate": 1.7464976651100732e-05,
+      "loss": 1.1388,
+      "step": 12382
+    },
+    {
+      "epoch": 11.431340872374799,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.7458305537024685e-05,
+      "loss": 1.3474,
+      "step": 12383
+    },
+    {
+      "epoch": 11.432264020309255,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.7451634422948633e-05,
+      "loss": 1.342,
+      "step": 12384
+    },
+    {
+      "epoch": 11.43318716824371,
+      "grad_norm": 3.4609375,
+      "learning_rate": 1.7444963308872582e-05,
+      "loss": 1.2214,
+      "step": 12385
+    },
+    {
+      "epoch": 11.434110316178167,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.743829219479653e-05,
+      "loss": 1.425,
+      "step": 12386
+    },
+    {
+      "epoch": 11.435033464112625,
+      "grad_norm": 2.798828125,
+      "learning_rate": 1.743162108072048e-05,
+      "loss": 0.9753,
+      "step": 12387
+    },
+    {
+      "epoch": 11.43595661204708,
+      "grad_norm": 4.32421875,
+      "learning_rate": 1.742494996664443e-05,
+      "loss": 1.2827,
+      "step": 12388
+    },
+    {
+      "epoch": 11.436879759981537,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.7418278852568377e-05,
+      "loss": 1.4272,
+      "step": 12389
+    },
+    {
+      "epoch": 11.437802907915994,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.741160773849233e-05,
+      "loss": 1.3813,
+      "step": 12390
+    },
+    {
+      "epoch": 11.43872605585045,
+      "grad_norm": 3.099609375,
+      "learning_rate": 1.740493662441628e-05,
+      "loss": 1.2385,
+      "step": 12391
+    },
+    {
+      "epoch": 11.439649203784906,
+      "grad_norm": 2.646484375,
+      "learning_rate": 1.7398265510340227e-05,
+      "loss": 1.0756,
+      "step": 12392
+    },
+    {
+      "epoch": 11.440572351719362,
+      "grad_norm": 3.1640625,
+      "learning_rate": 1.7391594396264176e-05,
+      "loss": 1.301,
+      "step": 12393
+    },
+    {
+      "epoch": 11.44149549965382,
+      "grad_norm": 3.49609375,
+      "learning_rate": 1.7384923282188125e-05,
+      "loss": 1.2778,
+      "step": 12394
+    },
+    {
+      "epoch": 11.442418647588276,
+      "grad_norm": 3.162109375,
+      "learning_rate": 1.7378252168112074e-05,
+      "loss": 1.0735,
+      "step": 12395
+    },
+    {
+      "epoch": 11.443341795522732,
+      "grad_norm": 3.435546875,
+      "learning_rate": 1.7371581054036023e-05,
+      "loss": 1.0074,
+      "step": 12396
+    },
+    {
+      "epoch": 11.444264943457188,
+      "grad_norm": 3.3046875,
+      "learning_rate": 1.7364909939959975e-05,
+      "loss": 1.4773,
+      "step": 12397
+    },
+    {
+      "epoch": 11.445188091391646,
+      "grad_norm": 3.248046875,
+      "learning_rate": 1.7358238825883924e-05,
+      "loss": 1.3496,
+      "step": 12398
+    },
+    {
+      "epoch": 11.446111239326102,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.7351567711807872e-05,
+      "loss": 1.4419,
+      "step": 12399
+    },
+    {
+      "epoch": 11.447034387260558,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.734489659773182e-05,
+      "loss": 1.3733,
+      "step": 12400
+    },
+    {
+      "epoch": 11.447957535195014,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.733822548365577e-05,
+      "loss": 1.4404,
+      "step": 12401
+    },
+    {
+      "epoch": 11.448880683129472,
+      "grad_norm": 3.39453125,
+      "learning_rate": 1.733155436957972e-05,
+      "loss": 1.0283,
+      "step": 12402
+    },
+    {
+      "epoch": 11.449803831063928,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.732488325550367e-05,
+      "loss": 1.1755,
+      "step": 12403
+    },
+    {
+      "epoch": 11.450726978998384,
+      "grad_norm": 3.189453125,
+      "learning_rate": 1.731821214142762e-05,
+      "loss": 1.2537,
+      "step": 12404
+    },
+    {
+      "epoch": 11.45165012693284,
+      "grad_norm": 3.7890625,
+      "learning_rate": 1.731154102735157e-05,
+      "loss": 1.3989,
+      "step": 12405
+    },
+    {
+      "epoch": 11.452573274867298,
+      "grad_norm": 3.158203125,
+      "learning_rate": 1.7304869913275518e-05,
+      "loss": 1.1288,
+      "step": 12406
+    },
+    {
+      "epoch": 11.453496422801754,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.7298198799199466e-05,
+      "loss": 1.2975,
+      "step": 12407
+    },
+    {
+      "epoch": 11.45441957073621,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.7291527685123415e-05,
+      "loss": 1.0659,
+      "step": 12408
+    },
+    {
+      "epoch": 11.455342718670668,
+      "grad_norm": 3.45703125,
+      "learning_rate": 1.7284856571047367e-05,
+      "loss": 1.2401,
+      "step": 12409
+    },
+    {
+      "epoch": 11.456265866605124,
+      "grad_norm": 2.998046875,
+      "learning_rate": 1.7278185456971316e-05,
+      "loss": 1.3665,
+      "step": 12410
+    },
+    {
+      "epoch": 11.45718901453958,
+      "grad_norm": 3.6015625,
+      "learning_rate": 1.7271514342895265e-05,
+      "loss": 1.2607,
+      "step": 12411
+    },
+    {
+      "epoch": 11.458112162474036,
+      "grad_norm": 3.6015625,
+      "learning_rate": 1.7264843228819214e-05,
+      "loss": 1.2949,
+      "step": 12412
+    },
+    {
+      "epoch": 11.459035310408494,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.7258172114743163e-05,
+      "loss": 1.3174,
+      "step": 12413
+    },
+    {
+      "epoch": 11.45995845834295,
+      "grad_norm": 3.2109375,
+      "learning_rate": 1.725150100066711e-05,
+      "loss": 1.126,
+      "step": 12414
+    },
+    {
+      "epoch": 11.460881606277406,
+      "grad_norm": 3.4453125,
+      "learning_rate": 1.7244829886591064e-05,
+      "loss": 1.2383,
+      "step": 12415
+    },
+    {
+      "epoch": 11.461804754211862,
+      "grad_norm": 3.541015625,
+      "learning_rate": 1.7238158772515012e-05,
+      "loss": 1.2883,
+      "step": 12416
+    },
+    {
+      "epoch": 11.46272790214632,
+      "grad_norm": 3.779296875,
+      "learning_rate": 1.723148765843896e-05,
+      "loss": 1.0774,
+      "step": 12417
+    },
+    {
+      "epoch": 11.463651050080776,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.722481654436291e-05,
+      "loss": 1.3391,
+      "step": 12418
+    },
+    {
+      "epoch": 11.464574198015232,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.721814543028686e-05,
+      "loss": 1.1814,
+      "step": 12419
+    },
+    {
+      "epoch": 11.465497345949688,
+      "grad_norm": 3.6484375,
+      "learning_rate": 1.7211474316210808e-05,
+      "loss": 1.1755,
+      "step": 12420
+    },
+    {
+      "epoch": 11.466420493884145,
+      "grad_norm": 3.30078125,
+      "learning_rate": 1.7204803202134757e-05,
+      "loss": 1.4172,
+      "step": 12421
+    },
+    {
+      "epoch": 11.467343641818601,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.719813208805871e-05,
+      "loss": 1.1492,
+      "step": 12422
+    },
+    {
+      "epoch": 11.468266789753057,
+      "grad_norm": 2.9921875,
+      "learning_rate": 1.7191460973982658e-05,
+      "loss": 1.0339,
+      "step": 12423
+    },
+    {
+      "epoch": 11.469189937687515,
+      "grad_norm": 3.068359375,
+      "learning_rate": 1.7184789859906606e-05,
+      "loss": 0.604,
+      "step": 12424
+    },
+    {
+      "epoch": 11.470113085621971,
+      "grad_norm": 3.158203125,
+      "learning_rate": 1.7178118745830555e-05,
+      "loss": 1.3743,
+      "step": 12425
+    },
+    {
+      "epoch": 11.471036233556427,
+      "grad_norm": 3.55078125,
+      "learning_rate": 1.7171447631754504e-05,
+      "loss": 1.4065,
+      "step": 12426
+    },
+    {
+      "epoch": 11.471959381490883,
+      "grad_norm": 2.8359375,
+      "learning_rate": 1.7164776517678453e-05,
+      "loss": 1.1861,
+      "step": 12427
+    },
+    {
+      "epoch": 11.472882529425341,
+      "grad_norm": 3.26171875,
+      "learning_rate": 1.7158105403602405e-05,
+      "loss": 1.4219,
+      "step": 12428
+    },
+    {
+      "epoch": 11.473805677359797,
+      "grad_norm": 3.55859375,
+      "learning_rate": 1.7151434289526354e-05,
+      "loss": 1.158,
+      "step": 12429
+    },
+    {
+      "epoch": 11.474728825294253,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.7144763175450303e-05,
+      "loss": 1.0668,
+      "step": 12430
+    },
+    {
+      "epoch": 11.47565197322871,
+      "grad_norm": 3.59765625,
+      "learning_rate": 1.713809206137425e-05,
+      "loss": 1.3357,
+      "step": 12431
+    },
+    {
+      "epoch": 11.476575121163167,
+      "grad_norm": 2.66796875,
+      "learning_rate": 1.71314209472982e-05,
+      "loss": 1.2466,
+      "step": 12432
+    },
+    {
+      "epoch": 11.477498269097623,
+      "grad_norm": 3.287109375,
+      "learning_rate": 1.712474983322215e-05,
+      "loss": 1.1667,
+      "step": 12433
+    },
+    {
+      "epoch": 11.478421417032079,
+      "grad_norm": 3.16015625,
+      "learning_rate": 1.7118078719146098e-05,
+      "loss": 1.0554,
+      "step": 12434
+    },
+    {
+      "epoch": 11.479344564966535,
+      "grad_norm": 2.939453125,
+      "learning_rate": 1.7111407605070047e-05,
+      "loss": 1.2769,
+      "step": 12435
+    },
+    {
+      "epoch": 11.480267712900993,
+      "grad_norm": 2.591796875,
+      "learning_rate": 1.7104736490993996e-05,
+      "loss": 1.1085,
+      "step": 12436
+    },
+    {
+      "epoch": 11.481190860835449,
+      "grad_norm": 3.099609375,
+      "learning_rate": 1.7098065376917944e-05,
+      "loss": 1.1758,
+      "step": 12437
+    },
+    {
+      "epoch": 11.482114008769905,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.7091394262841893e-05,
+      "loss": 1.0135,
+      "step": 12438
+    },
+    {
+      "epoch": 11.483037156704363,
+      "grad_norm": 3.150390625,
+      "learning_rate": 1.7084723148765845e-05,
+      "loss": 1.2361,
+      "step": 12439
+    },
+    {
+      "epoch": 11.483960304638819,
+      "grad_norm": 3.373046875,
+      "learning_rate": 1.7078052034689794e-05,
+      "loss": 1.321,
+      "step": 12440
+    },
+    {
+      "epoch": 11.484883452573275,
+      "grad_norm": 3.138671875,
+      "learning_rate": 1.7071380920613743e-05,
+      "loss": 1.3193,
+      "step": 12441
+    },
+    {
+      "epoch": 11.48580660050773,
+      "grad_norm": 3.248046875,
+      "learning_rate": 1.7064709806537692e-05,
+      "loss": 1.3552,
+      "step": 12442
+    },
+    {
+      "epoch": 11.486729748442189,
+      "grad_norm": 3.427734375,
+      "learning_rate": 1.705803869246164e-05,
+      "loss": 1.0127,
+      "step": 12443
+    },
+    {
+      "epoch": 11.487652896376645,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.705136757838559e-05,
+      "loss": 1.3102,
+      "step": 12444
+    },
+    {
+      "epoch": 11.4885760443111,
+      "grad_norm": 3.65234375,
+      "learning_rate": 1.704469646430954e-05,
+      "loss": 1.226,
+      "step": 12445
+    },
+    {
+      "epoch": 11.489499192245557,
+      "grad_norm": 3.505859375,
+      "learning_rate": 1.7038025350233487e-05,
+      "loss": 1.0863,
+      "step": 12446
+    },
+    {
+      "epoch": 11.490422340180015,
+      "grad_norm": 3.45703125,
+      "learning_rate": 1.703135423615744e-05,
+      "loss": 0.7922,
+      "step": 12447
+    },
+    {
+      "epoch": 11.49134548811447,
+      "grad_norm": 4.12890625,
+      "learning_rate": 1.7024683122081388e-05,
+      "loss": 1.157,
+      "step": 12448
+    },
+    {
+      "epoch": 11.492268636048927,
+      "grad_norm": 3.013671875,
+      "learning_rate": 1.7018012008005337e-05,
+      "loss": 1.1135,
+      "step": 12449
+    },
+    {
+      "epoch": 11.493191783983383,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.7011340893929286e-05,
+      "loss": 1.2994,
+      "step": 12450
+    },
+    {
+      "epoch": 11.49411493191784,
+      "grad_norm": 2.998046875,
+      "learning_rate": 1.7004669779853235e-05,
+      "loss": 1.4216,
+      "step": 12451
+    },
+    {
+      "epoch": 11.495038079852296,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.6997998665777183e-05,
+      "loss": 1.1715,
+      "step": 12452
+    },
+    {
+      "epoch": 11.495961227786752,
+      "grad_norm": 3.541015625,
+      "learning_rate": 1.6991327551701136e-05,
+      "loss": 1.1841,
+      "step": 12453
+    },
+    {
+      "epoch": 11.496884375721208,
+      "grad_norm": 3.248046875,
+      "learning_rate": 1.6984656437625085e-05,
+      "loss": 1.3228,
+      "step": 12454
+    },
+    {
+      "epoch": 11.497807523655666,
+      "grad_norm": 2.994140625,
+      "learning_rate": 1.6977985323549033e-05,
+      "loss": 1.1665,
+      "step": 12455
+    },
+    {
+      "epoch": 11.498730671590122,
+      "grad_norm": 3.6015625,
+      "learning_rate": 1.6971314209472982e-05,
+      "loss": 1.1361,
+      "step": 12456
+    },
+    {
+      "epoch": 11.499653819524578,
+      "grad_norm": 2.962890625,
+      "learning_rate": 1.696464309539693e-05,
+      "loss": 1.2986,
+      "step": 12457
+    },
+    {
+      "epoch": 11.500576967459036,
+      "grad_norm": 3.1484375,
+      "learning_rate": 1.695797198132088e-05,
+      "loss": 1.2084,
+      "step": 12458
+    },
+    {
+      "epoch": 11.501500115393492,
+      "grad_norm": 3.001953125,
+      "learning_rate": 1.6951300867244832e-05,
+      "loss": 0.944,
+      "step": 12459
+    },
+    {
+      "epoch": 11.502423263327948,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.694462975316878e-05,
+      "loss": 1.3016,
+      "step": 12460
+    },
+    {
+      "epoch": 11.503346411262404,
+      "grad_norm": 3.134765625,
+      "learning_rate": 1.693795863909273e-05,
+      "loss": 1.2852,
+      "step": 12461
+    },
+    {
+      "epoch": 11.504269559196862,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.693128752501668e-05,
+      "loss": 1.3843,
+      "step": 12462
+    },
+    {
+      "epoch": 11.505192707131318,
+      "grad_norm": 3.2421875,
+      "learning_rate": 1.6924616410940627e-05,
+      "loss": 1.3933,
+      "step": 12463
+    },
+    {
+      "epoch": 11.506115855065774,
+      "grad_norm": 3.09765625,
+      "learning_rate": 1.6917945296864576e-05,
+      "loss": 1.4426,
+      "step": 12464
+    },
+    {
+      "epoch": 11.50703900300023,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.6911274182788525e-05,
+      "loss": 1.0194,
+      "step": 12465
+    },
+    {
+      "epoch": 11.507962150934688,
+      "grad_norm": 3.064453125,
+      "learning_rate": 1.6904603068712477e-05,
+      "loss": 1.1544,
+      "step": 12466
+    },
+    {
+      "epoch": 11.508885298869144,
+      "grad_norm": 3.13671875,
+      "learning_rate": 1.6897931954636426e-05,
+      "loss": 1.4456,
+      "step": 12467
+    },
+    {
+      "epoch": 11.5098084468036,
+      "grad_norm": 3.0234375,
+      "learning_rate": 1.6891260840560375e-05,
+      "loss": 1.4001,
+      "step": 12468
+    },
+    {
+      "epoch": 11.510731594738056,
+      "grad_norm": 3.080078125,
+      "learning_rate": 1.6884589726484324e-05,
+      "loss": 1.3542,
+      "step": 12469
+    },
+    {
+      "epoch": 11.511654742672514,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.6877918612408272e-05,
+      "loss": 1.0202,
+      "step": 12470
+    },
+    {
+      "epoch": 11.51257789060697,
+      "grad_norm": 3.013671875,
+      "learning_rate": 1.687124749833222e-05,
+      "loss": 1.3291,
+      "step": 12471
+    },
+    {
+      "epoch": 11.513501038541426,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.6864576384256173e-05,
+      "loss": 1.4922,
+      "step": 12472
+    },
+    {
+      "epoch": 11.514424186475882,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.6857905270180122e-05,
+      "loss": 1.3164,
+      "step": 12473
+    },
+    {
+      "epoch": 11.51534733441034,
+      "grad_norm": 3.263671875,
+      "learning_rate": 1.685123415610407e-05,
+      "loss": 1.184,
+      "step": 12474
+    },
+    {
+      "epoch": 11.516270482344796,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.684456304202802e-05,
+      "loss": 1.3018,
+      "step": 12475
+    },
+    {
+      "epoch": 11.517193630279252,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.683789192795197e-05,
+      "loss": 1.3525,
+      "step": 12476
+    },
+    {
+      "epoch": 11.51811677821371,
+      "grad_norm": 3.73828125,
+      "learning_rate": 1.6831220813875918e-05,
+      "loss": 0.8547,
+      "step": 12477
+    },
+    {
+      "epoch": 11.519039926148166,
+      "grad_norm": 2.435546875,
+      "learning_rate": 1.682454969979987e-05,
+      "loss": 1.0304,
+      "step": 12478
+    },
+    {
+      "epoch": 11.519963074082622,
+      "grad_norm": 3.482421875,
+      "learning_rate": 1.681787858572382e-05,
+      "loss": 1.1461,
+      "step": 12479
+    },
+    {
+      "epoch": 11.520886222017078,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.6811207471647767e-05,
+      "loss": 1.2568,
+      "step": 12480
+    },
+    {
+      "epoch": 11.521809369951535,
+      "grad_norm": 3.205078125,
+      "learning_rate": 1.6804536357571716e-05,
+      "loss": 1.3259,
+      "step": 12481
+    },
+    {
+      "epoch": 11.522732517885991,
+      "grad_norm": 3.306640625,
+      "learning_rate": 1.6797865243495665e-05,
+      "loss": 1.2944,
+      "step": 12482
+    },
+    {
+      "epoch": 11.523655665820447,
+      "grad_norm": 3.02734375,
+      "learning_rate": 1.6791194129419614e-05,
+      "loss": 1.3901,
+      "step": 12483
+    },
+    {
+      "epoch": 11.524578813754903,
+      "grad_norm": 3.62890625,
+      "learning_rate": 1.6784523015343566e-05,
+      "loss": 1.4348,
+      "step": 12484
+    },
+    {
+      "epoch": 11.525501961689361,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.6777851901267515e-05,
+      "loss": 0.8776,
+      "step": 12485
+    },
+    {
+      "epoch": 11.526425109623817,
+      "grad_norm": 3.533203125,
+      "learning_rate": 1.6771180787191464e-05,
+      "loss": 1.3008,
+      "step": 12486
+    },
+    {
+      "epoch": 11.527348257558273,
+      "grad_norm": 3.30078125,
+      "learning_rate": 1.6764509673115412e-05,
+      "loss": 1.312,
+      "step": 12487
+    },
+    {
+      "epoch": 11.528271405492731,
+      "grad_norm": 3.408203125,
+      "learning_rate": 1.675783855903936e-05,
+      "loss": 1.2827,
+      "step": 12488
+    },
+    {
+      "epoch": 11.529194553427187,
+      "grad_norm": 2.919921875,
+      "learning_rate": 1.675116744496331e-05,
+      "loss": 1.4524,
+      "step": 12489
+    },
+    {
+      "epoch": 11.530117701361643,
+      "grad_norm": 3.0,
+      "learning_rate": 1.674449633088726e-05,
+      "loss": 1.4263,
+      "step": 12490
+    },
+    {
+      "epoch": 11.5310408492961,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.6737825216811208e-05,
+      "loss": 0.9561,
+      "step": 12491
+    },
+    {
+      "epoch": 11.531963997230557,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.6731154102735157e-05,
+      "loss": 1.4961,
+      "step": 12492
+    },
+    {
+      "epoch": 11.532887145165013,
+      "grad_norm": 3.744140625,
+      "learning_rate": 1.6724482988659105e-05,
+      "loss": 1.2864,
+      "step": 12493
+    },
+    {
+      "epoch": 11.533810293099469,
+      "grad_norm": 3.3359375,
+      "learning_rate": 1.6717811874583054e-05,
+      "loss": 1.4006,
+      "step": 12494
+    },
+    {
+      "epoch": 11.534733441033925,
+      "grad_norm": 2.98046875,
+      "learning_rate": 1.6711140760507006e-05,
+      "loss": 1.2629,
+      "step": 12495
+    },
+    {
+      "epoch": 11.535656588968383,
+      "grad_norm": 3.3203125,
+      "learning_rate": 1.6704469646430955e-05,
+      "loss": 1.2043,
+      "step": 12496
+    },
+    {
+      "epoch": 11.536579736902839,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.6697798532354904e-05,
+      "loss": 1.2119,
+      "step": 12497
+    },
+    {
+      "epoch": 11.537502884837295,
+      "grad_norm": 2.9765625,
+      "learning_rate": 1.6691127418278853e-05,
+      "loss": 0.6103,
+      "step": 12498
+    },
+    {
+      "epoch": 11.538426032771751,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.66844563042028e-05,
+      "loss": 1.1019,
+      "step": 12499
+    },
+    {
+      "epoch": 11.539349180706209,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.667778519012675e-05,
+      "loss": 1.2534,
+      "step": 12500
+    },
+    {
+      "epoch": 11.540272328640665,
+      "grad_norm": 3.1015625,
+      "learning_rate": 1.66711140760507e-05,
+      "loss": 1.0472,
+      "step": 12501
+    },
+    {
+      "epoch": 11.54119547657512,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.6664442961974648e-05,
+      "loss": 1.269,
+      "step": 12502
+    },
+    {
+      "epoch": 11.542118624509577,
+      "grad_norm": 3.373046875,
+      "learning_rate": 1.66577718478986e-05,
+      "loss": 1.085,
+      "step": 12503
+    },
+    {
+      "epoch": 11.543041772444035,
+      "grad_norm": 2.955078125,
+      "learning_rate": 1.665110073382255e-05,
+      "loss": 1.2979,
+      "step": 12504
+    },
+    {
+      "epoch": 11.54396492037849,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.6644429619746498e-05,
+      "loss": 1.2872,
+      "step": 12505
+    },
+    {
+      "epoch": 11.544888068312947,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.6637758505670447e-05,
+      "loss": 1.2241,
+      "step": 12506
+    },
+    {
+      "epoch": 11.545811216247404,
+      "grad_norm": 2.908203125,
+      "learning_rate": 1.6631087391594396e-05,
+      "loss": 1.365,
+      "step": 12507
+    },
+    {
+      "epoch": 11.54673436418186,
+      "grad_norm": 2.9296875,
+      "learning_rate": 1.6624416277518344e-05,
+      "loss": 1.2845,
+      "step": 12508
+    },
+    {
+      "epoch": 11.547657512116317,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.6617745163442297e-05,
+      "loss": 1.2888,
+      "step": 12509
+    },
+    {
+      "epoch": 11.548580660050773,
+      "grad_norm": 3.025390625,
+      "learning_rate": 1.6611074049366245e-05,
+      "loss": 1.2812,
+      "step": 12510
+    },
+    {
+      "epoch": 11.54950380798523,
+      "grad_norm": 3.158203125,
+      "learning_rate": 1.6604402935290194e-05,
+      "loss": 1.3013,
+      "step": 12511
+    },
+    {
+      "epoch": 11.550426955919686,
+      "grad_norm": 2.865234375,
+      "learning_rate": 1.6597731821214143e-05,
+      "loss": 1.5327,
+      "step": 12512
+    },
+    {
+      "epoch": 11.551350103854142,
+      "grad_norm": 3.39453125,
+      "learning_rate": 1.6591060707138092e-05,
+      "loss": 1.3875,
+      "step": 12513
+    },
+    {
+      "epoch": 11.552273251788598,
+      "grad_norm": 3.9453125,
+      "learning_rate": 1.658438959306204e-05,
+      "loss": 1.5076,
+      "step": 12514
+    },
+    {
+      "epoch": 11.553196399723056,
+      "grad_norm": 3.30078125,
+      "learning_rate": 1.657771847898599e-05,
+      "loss": 1.2537,
+      "step": 12515
+    },
+    {
+      "epoch": 11.554119547657512,
+      "grad_norm": 3.8515625,
+      "learning_rate": 1.6571047364909942e-05,
+      "loss": 1.3857,
+      "step": 12516
+    },
+    {
+      "epoch": 11.555042695591968,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.656437625083389e-05,
+      "loss": 1.2079,
+      "step": 12517
+    },
+    {
+      "epoch": 11.555965843526424,
+      "grad_norm": 2.955078125,
+      "learning_rate": 1.655770513675784e-05,
+      "loss": 1.3691,
+      "step": 12518
+    },
+    {
+      "epoch": 11.556888991460882,
+      "grad_norm": 3.240234375,
+      "learning_rate": 1.6551034022681788e-05,
+      "loss": 1.1719,
+      "step": 12519
+    },
+    {
+      "epoch": 11.557812139395338,
+      "grad_norm": 3.486328125,
+      "learning_rate": 1.6544362908605737e-05,
+      "loss": 1.2737,
+      "step": 12520
+    },
+    {
+      "epoch": 11.558735287329794,
+      "grad_norm": 3.458984375,
+      "learning_rate": 1.6537691794529686e-05,
+      "loss": 1.0824,
+      "step": 12521
+    },
+    {
+      "epoch": 11.55965843526425,
+      "grad_norm": 2.88671875,
+      "learning_rate": 1.6531020680453638e-05,
+      "loss": 1.2242,
+      "step": 12522
+    },
+    {
+      "epoch": 11.560581583198708,
+      "grad_norm": 3.380859375,
+      "learning_rate": 1.6524349566377587e-05,
+      "loss": 1.2312,
+      "step": 12523
+    },
+    {
+      "epoch": 11.561504731133164,
+      "grad_norm": 4.2890625,
+      "learning_rate": 1.6517678452301536e-05,
+      "loss": 1.1151,
+      "step": 12524
+    },
+    {
+      "epoch": 11.56242787906762,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.6511007338225485e-05,
+      "loss": 1.1915,
+      "step": 12525
+    },
+    {
+      "epoch": 11.563351027002078,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.6504336224149433e-05,
+      "loss": 1.2402,
+      "step": 12526
+    },
+    {
+      "epoch": 11.564274174936534,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.6497665110073382e-05,
+      "loss": 1.0861,
+      "step": 12527
+    },
+    {
+      "epoch": 11.56519732287099,
+      "grad_norm": 3.29296875,
+      "learning_rate": 1.6490993995997334e-05,
+      "loss": 1.179,
+      "step": 12528
+    },
+    {
+      "epoch": 11.566120470805446,
+      "grad_norm": 2.951171875,
+      "learning_rate": 1.6484322881921283e-05,
+      "loss": 1.0929,
+      "step": 12529
+    },
+    {
+      "epoch": 11.567043618739904,
+      "grad_norm": 2.8984375,
+      "learning_rate": 1.6477651767845232e-05,
+      "loss": 1.375,
+      "step": 12530
+    },
+    {
+      "epoch": 11.56796676667436,
+      "grad_norm": 2.97265625,
+      "learning_rate": 1.647098065376918e-05,
+      "loss": 1.2488,
+      "step": 12531
+    },
+    {
+      "epoch": 11.568889914608816,
+      "grad_norm": 3.333984375,
+      "learning_rate": 1.646430953969313e-05,
+      "loss": 1.0542,
+      "step": 12532
+    },
+    {
+      "epoch": 11.569813062543272,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.645763842561708e-05,
+      "loss": 1.1223,
+      "step": 12533
+    },
+    {
+      "epoch": 11.57073621047773,
+      "grad_norm": 3.248046875,
+      "learning_rate": 1.6450967311541027e-05,
+      "loss": 1.0243,
+      "step": 12534
+    },
+    {
+      "epoch": 11.571659358412186,
+      "grad_norm": 2.826171875,
+      "learning_rate": 1.644429619746498e-05,
+      "loss": 1.2065,
+      "step": 12535
+    },
+    {
+      "epoch": 11.572582506346642,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.643762508338893e-05,
+      "loss": 1.2836,
+      "step": 12536
+    },
+    {
+      "epoch": 11.5735056542811,
+      "grad_norm": 2.8203125,
+      "learning_rate": 1.6430953969312877e-05,
+      "loss": 1.0582,
+      "step": 12537
+    },
+    {
+      "epoch": 11.574428802215555,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6424282855236826e-05,
+      "loss": 1.2402,
+      "step": 12538
+    },
+    {
+      "epoch": 11.575351950150012,
+      "grad_norm": 3.322265625,
+      "learning_rate": 1.6417611741160775e-05,
+      "loss": 1.2632,
+      "step": 12539
+    },
+    {
+      "epoch": 11.576275098084468,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.6410940627084724e-05,
+      "loss": 0.8791,
+      "step": 12540
+    },
+    {
+      "epoch": 11.577198246018925,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.6404269513008676e-05,
+      "loss": 1.2539,
+      "step": 12541
+    },
+    {
+      "epoch": 11.578121393953381,
+      "grad_norm": 3.73046875,
+      "learning_rate": 1.6397598398932625e-05,
+      "loss": 1.1345,
+      "step": 12542
+    },
+    {
+      "epoch": 11.579044541887837,
+      "grad_norm": 3.0859375,
+      "learning_rate": 1.6390927284856573e-05,
+      "loss": 0.9808,
+      "step": 12543
+    },
+    {
+      "epoch": 11.579967689822293,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.6384256170780522e-05,
+      "loss": 1.5569,
+      "step": 12544
+    },
+    {
+      "epoch": 11.580890837756751,
+      "grad_norm": 3.642578125,
+      "learning_rate": 1.637758505670447e-05,
+      "loss": 1.4338,
+      "step": 12545
+    },
+    {
+      "epoch": 11.581813985691207,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.637091394262842e-05,
+      "loss": 1.3381,
+      "step": 12546
+    },
+    {
+      "epoch": 11.582737133625663,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.636424282855237e-05,
+      "loss": 1.4148,
+      "step": 12547
+    },
+    {
+      "epoch": 11.58366028156012,
+      "grad_norm": 2.91015625,
+      "learning_rate": 1.6357571714476318e-05,
+      "loss": 1.003,
+      "step": 12548
+    },
+    {
+      "epoch": 11.584583429494577,
+      "grad_norm": 3.216796875,
+      "learning_rate": 1.6350900600400266e-05,
+      "loss": 1.2853,
+      "step": 12549
+    },
+    {
+      "epoch": 11.585506577429033,
+      "grad_norm": 3.119140625,
+      "learning_rate": 1.6344229486324215e-05,
+      "loss": 1.4048,
+      "step": 12550
+    },
+    {
+      "epoch": 11.58642972536349,
+      "grad_norm": 3.052734375,
+      "learning_rate": 1.6337558372248164e-05,
+      "loss": 1.264,
+      "step": 12551
+    },
+    {
+      "epoch": 11.587352873297945,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.6330887258172116e-05,
+      "loss": 1.0397,
+      "step": 12552
+    },
+    {
+      "epoch": 11.588276021232403,
+      "grad_norm": 3.10546875,
+      "learning_rate": 1.6324216144096065e-05,
+      "loss": 1.4202,
+      "step": 12553
+    },
+    {
+      "epoch": 11.589199169166859,
+      "grad_norm": 3.322265625,
+      "learning_rate": 1.6317545030020014e-05,
+      "loss": 1.1824,
+      "step": 12554
+    },
+    {
+      "epoch": 11.590122317101315,
+      "grad_norm": 3.50390625,
+      "learning_rate": 1.6310873915943963e-05,
+      "loss": 1.135,
+      "step": 12555
+    },
+    {
+      "epoch": 11.591045465035773,
+      "grad_norm": 3.41796875,
+      "learning_rate": 1.630420280186791e-05,
+      "loss": 1.3242,
+      "step": 12556
+    },
+    {
+      "epoch": 11.591968612970229,
+      "grad_norm": 3.189453125,
+      "learning_rate": 1.629753168779186e-05,
+      "loss": 1.3782,
+      "step": 12557
+    },
+    {
+      "epoch": 11.592891760904685,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.629086057371581e-05,
+      "loss": 1.2471,
+      "step": 12558
+    },
+    {
+      "epoch": 11.593814908839141,
+      "grad_norm": 3.4609375,
+      "learning_rate": 1.6284189459639758e-05,
+      "loss": 1.2637,
+      "step": 12559
+    },
+    {
+      "epoch": 11.594738056773599,
+      "grad_norm": 3.142578125,
+      "learning_rate": 1.627751834556371e-05,
+      "loss": 1.2015,
+      "step": 12560
+    },
+    {
+      "epoch": 11.595661204708055,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.627084723148766e-05,
+      "loss": 1.3977,
+      "step": 12561
+    },
+    {
+      "epoch": 11.59658435264251,
+      "grad_norm": 3.50390625,
+      "learning_rate": 1.6264176117411608e-05,
+      "loss": 1.5032,
+      "step": 12562
+    },
+    {
+      "epoch": 11.597507500576967,
+      "grad_norm": 3.05859375,
+      "learning_rate": 1.6257505003335557e-05,
+      "loss": 1.2917,
+      "step": 12563
+    },
+    {
+      "epoch": 11.598430648511425,
+      "grad_norm": 2.9140625,
+      "learning_rate": 1.6250833889259505e-05,
+      "loss": 1.3674,
+      "step": 12564
+    },
+    {
+      "epoch": 11.59935379644588,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.6244162775183454e-05,
+      "loss": 1.2725,
+      "step": 12565
+    },
+    {
+      "epoch": 11.600276944380337,
+      "grad_norm": 3.69140625,
+      "learning_rate": 1.6237491661107406e-05,
+      "loss": 1.302,
+      "step": 12566
+    },
+    {
+      "epoch": 11.601200092314793,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.6230820547031355e-05,
+      "loss": 1.321,
+      "step": 12567
+    },
+    {
+      "epoch": 11.60212324024925,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.6224149432955304e-05,
+      "loss": 1.4099,
+      "step": 12568
+    },
+    {
+      "epoch": 11.603046388183706,
+      "grad_norm": 3.2265625,
+      "learning_rate": 1.6217478318879253e-05,
+      "loss": 1.2004,
+      "step": 12569
+    },
+    {
+      "epoch": 11.603969536118163,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.6210807204803202e-05,
+      "loss": 1.412,
+      "step": 12570
+    },
+    {
+      "epoch": 11.604892684052619,
+      "grad_norm": 3.294921875,
+      "learning_rate": 1.620413609072715e-05,
+      "loss": 1.2183,
+      "step": 12571
+    },
+    {
+      "epoch": 11.605815831987076,
+      "grad_norm": 3.923828125,
+      "learning_rate": 1.6197464976651103e-05,
+      "loss": 1.1929,
+      "step": 12572
+    },
+    {
+      "epoch": 11.606738979921532,
+      "grad_norm": 3.408203125,
+      "learning_rate": 1.619079386257505e-05,
+      "loss": 1.2942,
+      "step": 12573
+    },
+    {
+      "epoch": 11.607662127855988,
+      "grad_norm": 2.880859375,
+      "learning_rate": 1.6184122748499e-05,
+      "loss": 1.4651,
+      "step": 12574
+    },
+    {
+      "epoch": 11.608585275790446,
+      "grad_norm": 3.29296875,
+      "learning_rate": 1.617745163442295e-05,
+      "loss": 1.2532,
+      "step": 12575
+    },
+    {
+      "epoch": 11.609508423724902,
+      "grad_norm": 2.8046875,
+      "learning_rate": 1.6170780520346898e-05,
+      "loss": 1.2499,
+      "step": 12576
+    },
+    {
+      "epoch": 11.610431571659358,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.6164109406270847e-05,
+      "loss": 0.9837,
+      "step": 12577
+    },
+    {
+      "epoch": 11.611354719593814,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.61574382921948e-05,
+      "loss": 1.2566,
+      "step": 12578
+    },
+    {
+      "epoch": 11.612277867528272,
+      "grad_norm": 2.8359375,
+      "learning_rate": 1.6150767178118748e-05,
+      "loss": 1.0992,
+      "step": 12579
+    },
+    {
+      "epoch": 11.613201015462728,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.6144096064042697e-05,
+      "loss": 1.3835,
+      "step": 12580
+    },
+    {
+      "epoch": 11.614124163397184,
+      "grad_norm": 3.50390625,
+      "learning_rate": 1.6137424949966645e-05,
+      "loss": 1.3044,
+      "step": 12581
+    },
+    {
+      "epoch": 11.61504731133164,
+      "grad_norm": 3.396484375,
+      "learning_rate": 1.6130753835890594e-05,
+      "loss": 1.3213,
+      "step": 12582
+    },
+    {
+      "epoch": 11.615970459266098,
+      "grad_norm": 3.439453125,
+      "learning_rate": 1.6124082721814543e-05,
+      "loss": 1.2183,
+      "step": 12583
+    },
+    {
+      "epoch": 11.616893607200554,
+      "grad_norm": 2.533203125,
+      "learning_rate": 1.6117411607738492e-05,
+      "loss": 0.9873,
+      "step": 12584
+    },
+    {
+      "epoch": 11.61781675513501,
+      "grad_norm": 2.8515625,
+      "learning_rate": 1.6110740493662444e-05,
+      "loss": 1.3066,
+      "step": 12585
+    },
+    {
+      "epoch": 11.618739903069468,
+      "grad_norm": 3.205078125,
+      "learning_rate": 1.6104069379586393e-05,
+      "loss": 1.3218,
+      "step": 12586
+    },
+    {
+      "epoch": 11.619663051003924,
+      "grad_norm": 3.107421875,
+      "learning_rate": 1.6097398265510342e-05,
+      "loss": 1.2776,
+      "step": 12587
+    },
+    {
+      "epoch": 11.62058619893838,
+      "grad_norm": 3.537109375,
+      "learning_rate": 1.609072715143429e-05,
+      "loss": 1.1196,
+      "step": 12588
+    },
+    {
+      "epoch": 11.621509346872836,
+      "grad_norm": 3.033203125,
+      "learning_rate": 1.608405603735824e-05,
+      "loss": 1.0895,
+      "step": 12589
+    },
+    {
+      "epoch": 11.622432494807294,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.6077384923282188e-05,
+      "loss": 1.1091,
+      "step": 12590
+    },
+    {
+      "epoch": 11.62335564274175,
+      "grad_norm": 3.158203125,
+      "learning_rate": 1.607071380920614e-05,
+      "loss": 1.188,
+      "step": 12591
+    },
+    {
+      "epoch": 11.624278790676206,
+      "grad_norm": 3.552734375,
+      "learning_rate": 1.606404269513009e-05,
+      "loss": 0.9813,
+      "step": 12592
+    },
+    {
+      "epoch": 11.625201938610662,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.6057371581054038e-05,
+      "loss": 1.1,
+      "step": 12593
+    },
+    {
+      "epoch": 11.62612508654512,
+      "grad_norm": 3.01953125,
+      "learning_rate": 1.6050700466977987e-05,
+      "loss": 1.3853,
+      "step": 12594
+    },
+    {
+      "epoch": 11.627048234479576,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.6044029352901936e-05,
+      "loss": 1.3127,
+      "step": 12595
+    },
+    {
+      "epoch": 11.627971382414032,
+      "grad_norm": 3.150390625,
+      "learning_rate": 1.6037358238825885e-05,
+      "loss": 1.1458,
+      "step": 12596
+    },
+    {
+      "epoch": 11.628894530348488,
+      "grad_norm": 3.134765625,
+      "learning_rate": 1.6030687124749837e-05,
+      "loss": 1.4138,
+      "step": 12597
+    },
+    {
+      "epoch": 11.629817678282945,
+      "grad_norm": 3.60546875,
+      "learning_rate": 1.6024016010673786e-05,
+      "loss": 1.1077,
+      "step": 12598
+    },
+    {
+      "epoch": 11.630740826217401,
+      "grad_norm": 3.134765625,
+      "learning_rate": 1.6017344896597734e-05,
+      "loss": 1.3403,
+      "step": 12599
+    },
+    {
+      "epoch": 11.631663974151857,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.6010673782521683e-05,
+      "loss": 1.5007,
+      "step": 12600
+    },
+    {
+      "epoch": 11.632587122086314,
+      "grad_norm": 3.72265625,
+      "learning_rate": 1.6004002668445632e-05,
+      "loss": 1.3202,
+      "step": 12601
+    },
+    {
+      "epoch": 11.633510270020771,
+      "grad_norm": 3.25,
+      "learning_rate": 1.599733155436958e-05,
+      "loss": 1.269,
+      "step": 12602
+    },
+    {
+      "epoch": 11.634433417955227,
+      "grad_norm": 2.9921875,
+      "learning_rate": 1.599066044029353e-05,
+      "loss": 1.4653,
+      "step": 12603
+    },
+    {
+      "epoch": 11.635356565889683,
+      "grad_norm": 3.39453125,
+      "learning_rate": 1.598398932621748e-05,
+      "loss": 1.1663,
+      "step": 12604
+    },
+    {
+      "epoch": 11.636279713824141,
+      "grad_norm": 3.646484375,
+      "learning_rate": 1.5977318212141427e-05,
+      "loss": 1.4556,
+      "step": 12605
+    },
+    {
+      "epoch": 11.637202861758597,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.5970647098065376e-05,
+      "loss": 1.4331,
+      "step": 12606
+    },
+    {
+      "epoch": 11.638126009693053,
+      "grad_norm": 3.49609375,
+      "learning_rate": 1.5963975983989325e-05,
+      "loss": 1.4421,
+      "step": 12607
+    },
+    {
+      "epoch": 11.63904915762751,
+      "grad_norm": 3.544921875,
+      "learning_rate": 1.5957304869913277e-05,
+      "loss": 1.2407,
+      "step": 12608
+    },
+    {
+      "epoch": 11.639972305561967,
+      "grad_norm": 3.708984375,
+      "learning_rate": 1.5950633755837226e-05,
+      "loss": 1.2744,
+      "step": 12609
+    },
+    {
+      "epoch": 11.640895453496423,
+      "grad_norm": 3.345703125,
+      "learning_rate": 1.5943962641761175e-05,
+      "loss": 1.3381,
+      "step": 12610
+    },
+    {
+      "epoch": 11.641818601430879,
+      "grad_norm": 3.267578125,
+      "learning_rate": 1.5937291527685124e-05,
+      "loss": 1.4116,
+      "step": 12611
+    },
+    {
+      "epoch": 11.642741749365335,
+      "grad_norm": 3.125,
+      "learning_rate": 1.5930620413609072e-05,
+      "loss": 1.0577,
+      "step": 12612
+    },
+    {
+      "epoch": 11.643664897299793,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.592394929953302e-05,
+      "loss": 1.3818,
+      "step": 12613
+    },
+    {
+      "epoch": 11.644588045234249,
+      "grad_norm": 3.2890625,
+      "learning_rate": 1.591727818545697e-05,
+      "loss": 1.3892,
+      "step": 12614
+    },
+    {
+      "epoch": 11.645511193168705,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.591060707138092e-05,
+      "loss": 1.1996,
+      "step": 12615
+    },
+    {
+      "epoch": 11.646434341103161,
+      "grad_norm": 3.279296875,
+      "learning_rate": 1.590393595730487e-05,
+      "loss": 1.2078,
+      "step": 12616
+    },
+    {
+      "epoch": 11.647357489037619,
+      "grad_norm": 3.01171875,
+      "learning_rate": 1.589726484322882e-05,
+      "loss": 1.4543,
+      "step": 12617
+    },
+    {
+      "epoch": 11.648280636972075,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.589059372915277e-05,
+      "loss": 1.1488,
+      "step": 12618
+    },
+    {
+      "epoch": 11.64920378490653,
+      "grad_norm": 3.287109375,
+      "learning_rate": 1.5883922615076718e-05,
+      "loss": 1.3477,
+      "step": 12619
+    },
+    {
+      "epoch": 11.650126932840987,
+      "grad_norm": 3.216796875,
+      "learning_rate": 1.5877251501000666e-05,
+      "loss": 1.0471,
+      "step": 12620
+    },
+    {
+      "epoch": 11.651050080775445,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.5870580386924615e-05,
+      "loss": 1.238,
+      "step": 12621
+    },
+    {
+      "epoch": 11.6519732287099,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.5863909272848567e-05,
+      "loss": 1.4507,
+      "step": 12622
+    },
+    {
+      "epoch": 11.652896376644357,
+      "grad_norm": 3.4921875,
+      "learning_rate": 1.5857238158772516e-05,
+      "loss": 1.3452,
+      "step": 12623
+    },
+    {
+      "epoch": 11.653819524578815,
+      "grad_norm": 2.9765625,
+      "learning_rate": 1.5850567044696465e-05,
+      "loss": 1.1686,
+      "step": 12624
+    },
+    {
+      "epoch": 11.65474267251327,
+      "grad_norm": 2.75,
+      "learning_rate": 1.5843895930620414e-05,
+      "loss": 1.1534,
+      "step": 12625
+    },
+    {
+      "epoch": 11.655665820447727,
+      "grad_norm": 3.412109375,
+      "learning_rate": 1.5837224816544363e-05,
+      "loss": 1.1089,
+      "step": 12626
+    },
+    {
+      "epoch": 11.656588968382183,
+      "grad_norm": 2.947265625,
+      "learning_rate": 1.583055370246831e-05,
+      "loss": 1.1772,
+      "step": 12627
+    },
+    {
+      "epoch": 11.65751211631664,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.582388258839226e-05,
+      "loss": 1.3374,
+      "step": 12628
+    },
+    {
+      "epoch": 11.658435264251096,
+      "grad_norm": 3.396484375,
+      "learning_rate": 1.5817211474316212e-05,
+      "loss": 1.5303,
+      "step": 12629
+    },
+    {
+      "epoch": 11.659358412185552,
+      "grad_norm": 3.302734375,
+      "learning_rate": 1.581054036024016e-05,
+      "loss": 1.4932,
+      "step": 12630
+    },
+    {
+      "epoch": 11.660281560120008,
+      "grad_norm": 3.103515625,
+      "learning_rate": 1.580386924616411e-05,
+      "loss": 1.2638,
+      "step": 12631
+    },
+    {
+      "epoch": 11.661204708054466,
+      "grad_norm": 3.705078125,
+      "learning_rate": 1.579719813208806e-05,
+      "loss": 1.2505,
+      "step": 12632
+    },
+    {
+      "epoch": 11.662127855988922,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.5790527018012008e-05,
+      "loss": 1.1917,
+      "step": 12633
+    },
+    {
+      "epoch": 11.663051003923378,
+      "grad_norm": 3.06640625,
+      "learning_rate": 1.5783855903935957e-05,
+      "loss": 1.355,
+      "step": 12634
+    },
+    {
+      "epoch": 11.663974151857836,
+      "grad_norm": 3.130859375,
+      "learning_rate": 1.577718478985991e-05,
+      "loss": 1.2572,
+      "step": 12635
+    },
+    {
+      "epoch": 11.664897299792292,
+      "grad_norm": 3.279296875,
+      "learning_rate": 1.5770513675783858e-05,
+      "loss": 1.3291,
+      "step": 12636
+    },
+    {
+      "epoch": 11.665820447726748,
+      "grad_norm": 3.46484375,
+      "learning_rate": 1.5763842561707806e-05,
+      "loss": 1.5513,
+      "step": 12637
+    },
+    {
+      "epoch": 11.666743595661204,
+      "grad_norm": 3.439453125,
+      "learning_rate": 1.5757171447631755e-05,
+      "loss": 0.9572,
+      "step": 12638
+    },
+    {
+      "epoch": 11.667666743595662,
+      "grad_norm": 3.23046875,
+      "learning_rate": 1.5750500333555704e-05,
+      "loss": 1.439,
+      "step": 12639
+    },
+    {
+      "epoch": 11.668589891530118,
+      "grad_norm": 3.603515625,
+      "learning_rate": 1.5743829219479653e-05,
+      "loss": 1.1951,
+      "step": 12640
+    },
+    {
+      "epoch": 11.669513039464574,
+      "grad_norm": 3.00390625,
+      "learning_rate": 1.5737158105403605e-05,
+      "loss": 1.2303,
+      "step": 12641
+    },
+    {
+      "epoch": 11.67043618739903,
+      "grad_norm": 3.14453125,
+      "learning_rate": 1.5730486991327554e-05,
+      "loss": 1.4375,
+      "step": 12642
+    },
+    {
+      "epoch": 11.671359335333488,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.5723815877251503e-05,
+      "loss": 1.3948,
+      "step": 12643
+    },
+    {
+      "epoch": 11.672282483267944,
+      "grad_norm": 3.208984375,
+      "learning_rate": 1.571714476317545e-05,
+      "loss": 1.3967,
+      "step": 12644
+    },
+    {
+      "epoch": 11.6732056312024,
+      "grad_norm": 3.55078125,
+      "learning_rate": 1.57104736490994e-05,
+      "loss": 1.2211,
+      "step": 12645
+    },
+    {
+      "epoch": 11.674128779136856,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.570380253502335e-05,
+      "loss": 1.2998,
+      "step": 12646
+    },
+    {
+      "epoch": 11.675051927071314,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.56971314209473e-05,
+      "loss": 1.2357,
+      "step": 12647
+    },
+    {
+      "epoch": 11.67597507500577,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.569046030687125e-05,
+      "loss": 1.3003,
+      "step": 12648
+    },
+    {
+      "epoch": 11.676898222940226,
+      "grad_norm": 3.38671875,
+      "learning_rate": 1.56837891927952e-05,
+      "loss": 1.2075,
+      "step": 12649
+    },
+    {
+      "epoch": 11.677821370874682,
+      "grad_norm": 3.048828125,
+      "learning_rate": 1.5677118078719148e-05,
+      "loss": 1.2517,
+      "step": 12650
+    },
+    {
+      "epoch": 11.67874451880914,
+      "grad_norm": 2.76953125,
+      "learning_rate": 1.5670446964643097e-05,
+      "loss": 1.2185,
+      "step": 12651
+    },
+    {
+      "epoch": 11.679667666743596,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.5663775850567045e-05,
+      "loss": 1.3474,
+      "step": 12652
+    },
+    {
+      "epoch": 11.680590814678052,
+      "grad_norm": 3.09765625,
+      "learning_rate": 1.5657104736490994e-05,
+      "loss": 1.2224,
+      "step": 12653
+    },
+    {
+      "epoch": 11.68151396261251,
+      "grad_norm": 4.015625,
+      "learning_rate": 1.5650433622414947e-05,
+      "loss": 1.3341,
+      "step": 12654
+    },
+    {
+      "epoch": 11.682437110546966,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.5643762508338895e-05,
+      "loss": 1.2301,
+      "step": 12655
+    },
+    {
+      "epoch": 11.683360258481422,
+      "grad_norm": 3.70703125,
+      "learning_rate": 1.5637091394262844e-05,
+      "loss": 1.2266,
+      "step": 12656
+    },
+    {
+      "epoch": 11.684283406415878,
+      "grad_norm": 3.6484375,
+      "learning_rate": 1.5630420280186793e-05,
+      "loss": 1.321,
+      "step": 12657
+    },
+    {
+      "epoch": 11.685206554350335,
+      "grad_norm": 3.966796875,
+      "learning_rate": 1.5623749166110742e-05,
+      "loss": 1.2827,
+      "step": 12658
+    },
+    {
+      "epoch": 11.686129702284791,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.561707805203469e-05,
+      "loss": 1.2136,
+      "step": 12659
+    },
+    {
+      "epoch": 11.687052850219247,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.561040693795864e-05,
+      "loss": 1.2162,
+      "step": 12660
+    },
+    {
+      "epoch": 11.687975998153703,
+      "grad_norm": 2.927734375,
+      "learning_rate": 1.5603735823882588e-05,
+      "loss": 1.4133,
+      "step": 12661
+    },
+    {
+      "epoch": 11.688899146088161,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.5597064709806537e-05,
+      "loss": 1.2354,
+      "step": 12662
+    },
+    {
+      "epoch": 11.689822294022617,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.5590393595730486e-05,
+      "loss": 1.3167,
+      "step": 12663
+    },
+    {
+      "epoch": 11.690745441957073,
+      "grad_norm": 2.912109375,
+      "learning_rate": 1.5583722481654435e-05,
+      "loss": 1.2478,
+      "step": 12664
+    },
+    {
+      "epoch": 11.69166858989153,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.5577051367578387e-05,
+      "loss": 1.3267,
+      "step": 12665
+    },
+    {
+      "epoch": 11.692591737825987,
+      "grad_norm": 3.146484375,
+      "learning_rate": 1.5570380253502336e-05,
+      "loss": 1.218,
+      "step": 12666
+    },
+    {
+      "epoch": 11.693514885760443,
+      "grad_norm": 2.865234375,
+      "learning_rate": 1.5563709139426285e-05,
+      "loss": 1.4121,
+      "step": 12667
+    },
+    {
+      "epoch": 11.6944380336949,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.5557038025350233e-05,
+      "loss": 1.0452,
+      "step": 12668
+    },
+    {
+      "epoch": 11.695361181629355,
+      "grad_norm": 3.263671875,
+      "learning_rate": 1.5550366911274182e-05,
+      "loss": 1.3403,
+      "step": 12669
+    },
+    {
+      "epoch": 11.696284329563813,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.554369579719813e-05,
+      "loss": 1.1414,
+      "step": 12670
+    },
+    {
+      "epoch": 11.697207477498269,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.553702468312208e-05,
+      "loss": 1.323,
+      "step": 12671
+    },
+    {
+      "epoch": 11.698130625432725,
+      "grad_norm": 3.7109375,
+      "learning_rate": 1.5530353569046032e-05,
+      "loss": 1.265,
+      "step": 12672
+    },
+    {
+      "epoch": 11.699053773367183,
+      "grad_norm": 3.16015625,
+      "learning_rate": 1.552368245496998e-05,
+      "loss": 1.1875,
+      "step": 12673
+    },
+    {
+      "epoch": 11.699976921301639,
+      "grad_norm": 3.45703125,
+      "learning_rate": 1.551701134089393e-05,
+      "loss": 1.3154,
+      "step": 12674
+    },
+    {
+      "epoch": 11.700900069236095,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.551034022681788e-05,
+      "loss": 1.1217,
+      "step": 12675
+    },
+    {
+      "epoch": 11.701823217170551,
+      "grad_norm": 2.994140625,
+      "learning_rate": 1.5503669112741827e-05,
+      "loss": 1.2347,
+      "step": 12676
+    },
+    {
+      "epoch": 11.702746365105009,
+      "grad_norm": 3.33984375,
+      "learning_rate": 1.5496997998665776e-05,
+      "loss": 1.1348,
+      "step": 12677
+    },
+    {
+      "epoch": 11.703669513039465,
+      "grad_norm": 3.10546875,
+      "learning_rate": 1.5490326884589725e-05,
+      "loss": 1.2238,
+      "step": 12678
+    },
+    {
+      "epoch": 11.70459266097392,
+      "grad_norm": 2.830078125,
+      "learning_rate": 1.5483655770513677e-05,
+      "loss": 1.0491,
+      "step": 12679
+    },
+    {
+      "epoch": 11.705515808908377,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.5476984656437626e-05,
+      "loss": 1.4026,
+      "step": 12680
+    },
+    {
+      "epoch": 11.706438956842835,
+      "grad_norm": 3.26953125,
+      "learning_rate": 1.5470313542361575e-05,
+      "loss": 1.2472,
+      "step": 12681
+    },
+    {
+      "epoch": 11.70736210477729,
+      "grad_norm": 3.091796875,
+      "learning_rate": 1.5463642428285524e-05,
+      "loss": 1.0245,
+      "step": 12682
+    },
+    {
+      "epoch": 11.708285252711747,
+      "grad_norm": 3.603515625,
+      "learning_rate": 1.5456971314209472e-05,
+      "loss": 1.0416,
+      "step": 12683
+    },
+    {
+      "epoch": 11.709208400646204,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.545030020013342e-05,
+      "loss": 0.97,
+      "step": 12684
+    },
+    {
+      "epoch": 11.71013154858066,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.5443629086057373e-05,
+      "loss": 1.3267,
+      "step": 12685
+    },
+    {
+      "epoch": 11.711054696515117,
+      "grad_norm": 3.107421875,
+      "learning_rate": 1.5436957971981322e-05,
+      "loss": 1.3308,
+      "step": 12686
+    },
+    {
+      "epoch": 11.711977844449573,
+      "grad_norm": 3.15234375,
+      "learning_rate": 1.543028685790527e-05,
+      "loss": 1.3271,
+      "step": 12687
+    },
+    {
+      "epoch": 11.712900992384029,
+      "grad_norm": 3.1171875,
+      "learning_rate": 1.542361574382922e-05,
+      "loss": 1.4419,
+      "step": 12688
+    },
+    {
+      "epoch": 11.713824140318486,
+      "grad_norm": 3.2265625,
+      "learning_rate": 1.541694462975317e-05,
+      "loss": 1.5039,
+      "step": 12689
+    },
+    {
+      "epoch": 11.714747288252942,
+      "grad_norm": 3.625,
+      "learning_rate": 1.5410273515677118e-05,
+      "loss": 1.0409,
+      "step": 12690
+    },
+    {
+      "epoch": 11.715670436187398,
+      "grad_norm": 3.12109375,
+      "learning_rate": 1.540360240160107e-05,
+      "loss": 0.981,
+      "step": 12691
+    },
+    {
+      "epoch": 11.716593584121856,
+      "grad_norm": 2.97265625,
+      "learning_rate": 1.539693128752502e-05,
+      "loss": 1.3784,
+      "step": 12692
+    },
+    {
+      "epoch": 11.717516732056312,
+      "grad_norm": 3.18359375,
+      "learning_rate": 1.5390260173448967e-05,
+      "loss": 1.1899,
+      "step": 12693
+    },
+    {
+      "epoch": 11.718439879990768,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.5383589059372916e-05,
+      "loss": 1.244,
+      "step": 12694
+    },
+    {
+      "epoch": 11.719363027925224,
+      "grad_norm": 3.29296875,
+      "learning_rate": 1.5376917945296865e-05,
+      "loss": 1.2964,
+      "step": 12695
+    },
+    {
+      "epoch": 11.720286175859682,
+      "grad_norm": 3.2578125,
+      "learning_rate": 1.5370246831220814e-05,
+      "loss": 1.4116,
+      "step": 12696
+    },
+    {
+      "epoch": 11.721209323794138,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.5363575717144766e-05,
+      "loss": 1.2545,
+      "step": 12697
+    },
+    {
+      "epoch": 11.722132471728594,
+      "grad_norm": 2.853515625,
+      "learning_rate": 1.5356904603068715e-05,
+      "loss": 0.9391,
+      "step": 12698
+    },
+    {
+      "epoch": 11.72305561966305,
+      "grad_norm": 3.25,
+      "learning_rate": 1.5350233488992664e-05,
+      "loss": 1.3438,
+      "step": 12699
+    },
+    {
+      "epoch": 11.723978767597508,
+      "grad_norm": 3.205078125,
+      "learning_rate": 1.5343562374916613e-05,
+      "loss": 1.3234,
+      "step": 12700
+    },
+    {
+      "epoch": 11.724901915531964,
+      "grad_norm": 3.345703125,
+      "learning_rate": 1.533689126084056e-05,
+      "loss": 1.3645,
+      "step": 12701
+    },
+    {
+      "epoch": 11.72582506346642,
+      "grad_norm": 2.99609375,
+      "learning_rate": 1.533022014676451e-05,
+      "loss": 1.0675,
+      "step": 12702
+    },
+    {
+      "epoch": 11.726748211400878,
+      "grad_norm": 3.404296875,
+      "learning_rate": 1.532354903268846e-05,
+      "loss": 1.3684,
+      "step": 12703
+    },
+    {
+      "epoch": 11.727671359335334,
+      "grad_norm": 3.0546875,
+      "learning_rate": 1.531687791861241e-05,
+      "loss": 1.3103,
+      "step": 12704
+    },
+    {
+      "epoch": 11.72859450726979,
+      "grad_norm": 3.107421875,
+      "learning_rate": 1.531020680453636e-05,
+      "loss": 1.0306,
+      "step": 12705
+    },
+    {
+      "epoch": 11.729517655204246,
+      "grad_norm": 3.42578125,
+      "learning_rate": 1.530353569046031e-05,
+      "loss": 1.248,
+      "step": 12706
+    },
+    {
+      "epoch": 11.730440803138704,
+      "grad_norm": 3.373046875,
+      "learning_rate": 1.5296864576384258e-05,
+      "loss": 1.4614,
+      "step": 12707
+    },
+    {
+      "epoch": 11.73136395107316,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.5290193462308206e-05,
+      "loss": 1.3735,
+      "step": 12708
+    },
+    {
+      "epoch": 11.732287099007616,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.5283522348232155e-05,
+      "loss": 1.2908,
+      "step": 12709
+    },
+    {
+      "epoch": 11.733210246942072,
+      "grad_norm": 2.96484375,
+      "learning_rate": 1.5276851234156107e-05,
+      "loss": 1.1943,
+      "step": 12710
+    },
+    {
+      "epoch": 11.73413339487653,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.5270180120080056e-05,
+      "loss": 1.4314,
+      "step": 12711
+    },
+    {
+      "epoch": 11.735056542810986,
+      "grad_norm": 3.08984375,
+      "learning_rate": 1.5263509006004005e-05,
+      "loss": 1.2734,
+      "step": 12712
+    },
+    {
+      "epoch": 11.735979690745442,
+      "grad_norm": 3.03515625,
+      "learning_rate": 1.5256837891927952e-05,
+      "loss": 1.2627,
+      "step": 12713
+    },
+    {
+      "epoch": 11.736902838679898,
+      "grad_norm": 3.2890625,
+      "learning_rate": 1.5250166777851901e-05,
+      "loss": 1.2551,
+      "step": 12714
+    },
+    {
+      "epoch": 11.737825986614356,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.524349566377585e-05,
+      "loss": 1.2404,
+      "step": 12715
+    },
+    {
+      "epoch": 11.738749134548812,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5236824549699802e-05,
+      "loss": 1.4397,
+      "step": 12716
+    },
+    {
+      "epoch": 11.739672282483268,
+      "grad_norm": 3.138671875,
+      "learning_rate": 1.5230153435623751e-05,
+      "loss": 1.4377,
+      "step": 12717
+    },
+    {
+      "epoch": 11.740595430417724,
+      "grad_norm": 3.14453125,
+      "learning_rate": 1.52234823215477e-05,
+      "loss": 1.3298,
+      "step": 12718
+    },
+    {
+      "epoch": 11.741518578352181,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.5216811207471649e-05,
+      "loss": 1.3284,
+      "step": 12719
+    },
+    {
+      "epoch": 11.742441726286637,
+      "grad_norm": 3.13671875,
+      "learning_rate": 1.5210140093395597e-05,
+      "loss": 1.3684,
+      "step": 12720
+    },
+    {
+      "epoch": 11.743364874221093,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.5203468979319546e-05,
+      "loss": 1.0081,
+      "step": 12721
+    },
+    {
+      "epoch": 11.744288022155551,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.5196797865243495e-05,
+      "loss": 1.3169,
+      "step": 12722
+    },
+    {
+      "epoch": 11.745211170090007,
+      "grad_norm": 2.8984375,
+      "learning_rate": 1.5190126751167447e-05,
+      "loss": 1.5049,
+      "step": 12723
+    },
+    {
+      "epoch": 11.746134318024463,
+      "grad_norm": 3.37109375,
+      "learning_rate": 1.5183455637091396e-05,
+      "loss": 1.2198,
+      "step": 12724
+    },
+    {
+      "epoch": 11.74705746595892,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.5176784523015345e-05,
+      "loss": 1.1952,
+      "step": 12725
+    },
+    {
+      "epoch": 11.747980613893377,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.5170113408939294e-05,
+      "loss": 1.0657,
+      "step": 12726
+    },
+    {
+      "epoch": 11.748903761827833,
+      "grad_norm": 3.248046875,
+      "learning_rate": 1.5163442294863242e-05,
+      "loss": 1.1699,
+      "step": 12727
+    },
+    {
+      "epoch": 11.74982690976229,
+      "grad_norm": 3.07421875,
+      "learning_rate": 1.5156771180787191e-05,
+      "loss": 1.201,
+      "step": 12728
+    },
+    {
+      "epoch": 11.750750057696745,
+      "grad_norm": 3.01953125,
+      "learning_rate": 1.5150100066711142e-05,
+      "loss": 1.1831,
+      "step": 12729
+    },
+    {
+      "epoch": 11.751673205631203,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.514342895263509e-05,
+      "loss": 1.1707,
+      "step": 12730
+    },
+    {
+      "epoch": 11.752596353565659,
+      "grad_norm": 3.376953125,
+      "learning_rate": 1.5136757838559041e-05,
+      "loss": 1.2933,
+      "step": 12731
+    },
+    {
+      "epoch": 11.753519501500115,
+      "grad_norm": 3.189453125,
+      "learning_rate": 1.513008672448299e-05,
+      "loss": 1.2603,
+      "step": 12732
+    },
+    {
+      "epoch": 11.754442649434571,
+      "grad_norm": 3.068359375,
+      "learning_rate": 1.5123415610406939e-05,
+      "loss": 1.2616,
+      "step": 12733
+    },
+    {
+      "epoch": 11.755365797369029,
+      "grad_norm": 3.37890625,
+      "learning_rate": 1.5116744496330888e-05,
+      "loss": 1.3398,
+      "step": 12734
+    },
+    {
+      "epoch": 11.756288945303485,
+      "grad_norm": 2.83984375,
+      "learning_rate": 1.5110073382254838e-05,
+      "loss": 1.3469,
+      "step": 12735
+    },
+    {
+      "epoch": 11.757212093237941,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.5103402268178787e-05,
+      "loss": 1.4785,
+      "step": 12736
+    },
+    {
+      "epoch": 11.758135241172397,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.5096731154102736e-05,
+      "loss": 1.2329,
+      "step": 12737
+    },
+    {
+      "epoch": 11.759058389106855,
+      "grad_norm": 3.51953125,
+      "learning_rate": 1.5090060040026685e-05,
+      "loss": 1.3442,
+      "step": 12738
+    },
+    {
+      "epoch": 11.75998153704131,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.5083388925950633e-05,
+      "loss": 1.4146,
+      "step": 12739
+    },
+    {
+      "epoch": 11.760904684975767,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.5076717811874582e-05,
+      "loss": 1.2346,
+      "step": 12740
+    },
+    {
+      "epoch": 11.761827832910225,
+      "grad_norm": 3.142578125,
+      "learning_rate": 1.5070046697798534e-05,
+      "loss": 1.2422,
+      "step": 12741
+    },
+    {
+      "epoch": 11.76275098084468,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.5063375583722483e-05,
+      "loss": 1.3254,
+      "step": 12742
+    },
+    {
+      "epoch": 11.763674128779137,
+      "grad_norm": 3.033203125,
+      "learning_rate": 1.5056704469646432e-05,
+      "loss": 1.2137,
+      "step": 12743
+    },
+    {
+      "epoch": 11.764597276713593,
+      "grad_norm": 3.0859375,
+      "learning_rate": 1.5050033355570381e-05,
+      "loss": 1.1133,
+      "step": 12744
+    },
+    {
+      "epoch": 11.76552042464805,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.504336224149433e-05,
+      "loss": 1.2266,
+      "step": 12745
+    },
+    {
+      "epoch": 11.766443572582507,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.5036691127418278e-05,
+      "loss": 1.5701,
+      "step": 12746
+    },
+    {
+      "epoch": 11.767366720516963,
+      "grad_norm": 3.3671875,
+      "learning_rate": 1.5030020013342227e-05,
+      "loss": 1.2683,
+      "step": 12747
+    },
+    {
+      "epoch": 11.768289868451419,
+      "grad_norm": 3.037109375,
+      "learning_rate": 1.502334889926618e-05,
+      "loss": 1.0287,
+      "step": 12748
+    },
+    {
+      "epoch": 11.769213016385876,
+      "grad_norm": 3.564453125,
+      "learning_rate": 1.5016677785190128e-05,
+      "loss": 1.4124,
+      "step": 12749
+    },
+    {
+      "epoch": 11.770136164320332,
+      "grad_norm": 3.4296875,
+      "learning_rate": 1.5010006671114077e-05,
+      "loss": 1.2148,
+      "step": 12750
+    },
+    {
+      "epoch": 11.771059312254788,
+      "grad_norm": 2.904296875,
+      "learning_rate": 1.5003335557038026e-05,
+      "loss": 1.3442,
+      "step": 12751
+    },
+    {
+      "epoch": 11.771982460189246,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.4996664442961975e-05,
+      "loss": 1.3098,
+      "step": 12752
+    },
+    {
+      "epoch": 11.772905608123702,
+      "grad_norm": 3.279296875,
+      "learning_rate": 1.4989993328885924e-05,
+      "loss": 1.0264,
+      "step": 12753
+    },
+    {
+      "epoch": 11.773828756058158,
+      "grad_norm": 3.986328125,
+      "learning_rate": 1.4983322214809876e-05,
+      "loss": 1.0953,
+      "step": 12754
+    },
+    {
+      "epoch": 11.774751903992614,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.4976651100733825e-05,
+      "loss": 1.4326,
+      "step": 12755
+    },
+    {
+      "epoch": 11.775675051927072,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.4969979986657773e-05,
+      "loss": 1.1771,
+      "step": 12756
+    },
+    {
+      "epoch": 11.776598199861528,
+      "grad_norm": 3.216796875,
+      "learning_rate": 1.4963308872581722e-05,
+      "loss": 1.2285,
+      "step": 12757
+    },
+    {
+      "epoch": 11.777521347795984,
+      "grad_norm": 3.36328125,
+      "learning_rate": 1.4956637758505671e-05,
+      "loss": 1.3196,
+      "step": 12758
+    },
+    {
+      "epoch": 11.77844449573044,
+      "grad_norm": 3.10546875,
+      "learning_rate": 1.494996664442962e-05,
+      "loss": 1.193,
+      "step": 12759
+    },
+    {
+      "epoch": 11.779367643664898,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.494329553035357e-05,
+      "loss": 1.2388,
+      "step": 12760
+    },
+    {
+      "epoch": 11.780290791599354,
+      "grad_norm": 2.978515625,
+      "learning_rate": 1.493662441627752e-05,
+      "loss": 1.2428,
+      "step": 12761
+    },
+    {
+      "epoch": 11.78121393953381,
+      "grad_norm": 3.056640625,
+      "learning_rate": 1.4929953302201468e-05,
+      "loss": 1.2606,
+      "step": 12762
+    },
+    {
+      "epoch": 11.782137087468266,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.4923282188125417e-05,
+      "loss": 1.3628,
+      "step": 12763
+    },
+    {
+      "epoch": 11.783060235402724,
+      "grad_norm": 2.95703125,
+      "learning_rate": 1.4916611074049366e-05,
+      "loss": 1.5376,
+      "step": 12764
+    },
+    {
+      "epoch": 11.78398338333718,
+      "grad_norm": 3.056640625,
+      "learning_rate": 1.4909939959973315e-05,
+      "loss": 1.4908,
+      "step": 12765
+    },
+    {
+      "epoch": 11.784906531271636,
+      "grad_norm": 3.416015625,
+      "learning_rate": 1.4903268845897267e-05,
+      "loss": 1.3923,
+      "step": 12766
+    },
+    {
+      "epoch": 11.785829679206092,
+      "grad_norm": 3.23046875,
+      "learning_rate": 1.4896597731821216e-05,
+      "loss": 1.3976,
+      "step": 12767
+    },
+    {
+      "epoch": 11.78675282714055,
+      "grad_norm": 3.7734375,
+      "learning_rate": 1.4889926617745164e-05,
+      "loss": 1.0991,
+      "step": 12768
+    },
+    {
+      "epoch": 11.787675975075006,
+      "grad_norm": 3.5390625,
+      "learning_rate": 1.4883255503669113e-05,
+      "loss": 1.3899,
+      "step": 12769
+    },
+    {
+      "epoch": 11.788599123009462,
+      "grad_norm": 3.037109375,
+      "learning_rate": 1.4876584389593062e-05,
+      "loss": 1.2112,
+      "step": 12770
+    },
+    {
+      "epoch": 11.78952227094392,
+      "grad_norm": 3.560546875,
+      "learning_rate": 1.486991327551701e-05,
+      "loss": 1.2535,
+      "step": 12771
+    },
+    {
+      "epoch": 11.790445418878376,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.486324216144096e-05,
+      "loss": 1.0316,
+      "step": 12772
+    },
+    {
+      "epoch": 11.791368566812832,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.4856571047364912e-05,
+      "loss": 1.5203,
+      "step": 12773
+    },
+    {
+      "epoch": 11.792291714747288,
+      "grad_norm": 3.3671875,
+      "learning_rate": 1.484989993328886e-05,
+      "loss": 1.3271,
+      "step": 12774
+    },
+    {
+      "epoch": 11.793214862681745,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.484322881921281e-05,
+      "loss": 1.2086,
+      "step": 12775
+    },
+    {
+      "epoch": 11.794138010616201,
+      "grad_norm": 3.5546875,
+      "learning_rate": 1.4836557705136758e-05,
+      "loss": 1.2839,
+      "step": 12776
+    },
+    {
+      "epoch": 11.795061158550658,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.4829886591060707e-05,
+      "loss": 1.2842,
+      "step": 12777
+    },
+    {
+      "epoch": 11.795984306485114,
+      "grad_norm": 2.802734375,
+      "learning_rate": 1.4823215476984656e-05,
+      "loss": 1.1143,
+      "step": 12778
+    },
+    {
+      "epoch": 11.796907454419571,
+      "grad_norm": 3.125,
+      "learning_rate": 1.4816544362908608e-05,
+      "loss": 1.3582,
+      "step": 12779
+    },
+    {
+      "epoch": 11.797830602354027,
+      "grad_norm": 3.375,
+      "learning_rate": 1.4809873248832557e-05,
+      "loss": 1.24,
+      "step": 12780
+    },
+    {
+      "epoch": 11.798753750288483,
+      "grad_norm": 3.123046875,
+      "learning_rate": 1.4803202134756506e-05,
+      "loss": 1.3879,
+      "step": 12781
+    },
+    {
+      "epoch": 11.79967689822294,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.4796531020680455e-05,
+      "loss": 1.1757,
+      "step": 12782
+    },
+    {
+      "epoch": 11.800600046157397,
+      "grad_norm": 3.044921875,
+      "learning_rate": 1.4789859906604403e-05,
+      "loss": 1.3855,
+      "step": 12783
+    },
+    {
+      "epoch": 11.801523194091853,
+      "grad_norm": 4.0546875,
+      "learning_rate": 1.4783188792528352e-05,
+      "loss": 1.1002,
+      "step": 12784
+    },
+    {
+      "epoch": 11.80244634202631,
+      "grad_norm": 3.73828125,
+      "learning_rate": 1.4776517678452303e-05,
+      "loss": 1.4644,
+      "step": 12785
+    },
+    {
+      "epoch": 11.803369489960765,
+      "grad_norm": 3.869140625,
+      "learning_rate": 1.4769846564376252e-05,
+      "loss": 1.384,
+      "step": 12786
+    },
+    {
+      "epoch": 11.804292637895223,
+      "grad_norm": 3.705078125,
+      "learning_rate": 1.47631754503002e-05,
+      "loss": 1.3494,
+      "step": 12787
+    },
+    {
+      "epoch": 11.80521578582968,
+      "grad_norm": 3.091796875,
+      "learning_rate": 1.4756504336224151e-05,
+      "loss": 1.2754,
+      "step": 12788
+    },
+    {
+      "epoch": 11.806138933764135,
+      "grad_norm": 3.123046875,
+      "learning_rate": 1.47498332221481e-05,
+      "loss": 1.0573,
+      "step": 12789
+    },
+    {
+      "epoch": 11.807062081698593,
+      "grad_norm": 3.14453125,
+      "learning_rate": 1.4743162108072049e-05,
+      "loss": 1.2544,
+      "step": 12790
+    },
+    {
+      "epoch": 11.807985229633049,
+      "grad_norm": 3.298828125,
+      "learning_rate": 1.4736490993995999e-05,
+      "loss": 1.2377,
+      "step": 12791
+    },
+    {
+      "epoch": 11.808908377567505,
+      "grad_norm": 3.345703125,
+      "learning_rate": 1.4729819879919948e-05,
+      "loss": 1.218,
+      "step": 12792
+    },
+    {
+      "epoch": 11.809831525501961,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.4723148765843897e-05,
+      "loss": 1.3779,
+      "step": 12793
+    },
+    {
+      "epoch": 11.810754673436419,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.4716477651767846e-05,
+      "loss": 1.2437,
+      "step": 12794
+    },
+    {
+      "epoch": 11.811677821370875,
+      "grad_norm": 2.96484375,
+      "learning_rate": 1.4709806537691794e-05,
+      "loss": 1.4099,
+      "step": 12795
+    },
+    {
+      "epoch": 11.81260096930533,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.4703135423615743e-05,
+      "loss": 1.139,
+      "step": 12796
+    },
+    {
+      "epoch": 11.813524117239787,
+      "grad_norm": 3.220703125,
+      "learning_rate": 1.4696464309539692e-05,
+      "loss": 1.3071,
+      "step": 12797
+    },
+    {
+      "epoch": 11.814447265174245,
+      "grad_norm": 3.44921875,
+      "learning_rate": 1.4689793195463644e-05,
+      "loss": 1.4509,
+      "step": 12798
+    },
+    {
+      "epoch": 11.8153704131087,
+      "grad_norm": 3.322265625,
+      "learning_rate": 1.4683122081387593e-05,
+      "loss": 1.2026,
+      "step": 12799
+    },
+    {
+      "epoch": 11.816293561043157,
+      "grad_norm": 3.435546875,
+      "learning_rate": 1.4676450967311542e-05,
+      "loss": 1.4316,
+      "step": 12800
+    },
+    {
+      "epoch": 11.817216708977615,
+      "grad_norm": 2.93359375,
+      "learning_rate": 1.466977985323549e-05,
+      "loss": 1.2496,
+      "step": 12801
+    },
+    {
+      "epoch": 11.81813985691207,
+      "grad_norm": 3.416015625,
+      "learning_rate": 1.466310873915944e-05,
+      "loss": 1.3711,
+      "step": 12802
+    },
+    {
+      "epoch": 11.819063004846527,
+      "grad_norm": 2.595703125,
+      "learning_rate": 1.4656437625083388e-05,
+      "loss": 0.9173,
+      "step": 12803
+    },
+    {
+      "epoch": 11.819986152780983,
+      "grad_norm": 3.25,
+      "learning_rate": 1.464976651100734e-05,
+      "loss": 1.4785,
+      "step": 12804
+    },
+    {
+      "epoch": 11.82090930071544,
+      "grad_norm": 2.904296875,
+      "learning_rate": 1.464309539693129e-05,
+      "loss": 1.2928,
+      "step": 12805
+    },
+    {
+      "epoch": 11.821832448649896,
+      "grad_norm": 2.955078125,
+      "learning_rate": 1.4636424282855238e-05,
+      "loss": 1.3335,
+      "step": 12806
+    },
+    {
+      "epoch": 11.822755596584352,
+      "grad_norm": 3.3515625,
+      "learning_rate": 1.4629753168779187e-05,
+      "loss": 1.4106,
+      "step": 12807
+    },
+    {
+      "epoch": 11.823678744518809,
+      "grad_norm": 3.2578125,
+      "learning_rate": 1.4623082054703136e-05,
+      "loss": 1.2104,
+      "step": 12808
+    },
+    {
+      "epoch": 11.824601892453266,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.4616410940627085e-05,
+      "loss": 1.4519,
+      "step": 12809
+    },
+    {
+      "epoch": 11.825525040387722,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.4609739826551037e-05,
+      "loss": 1.0784,
+      "step": 12810
+    },
+    {
+      "epoch": 11.826448188322178,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.4603068712474986e-05,
+      "loss": 1.1895,
+      "step": 12811
+    },
+    {
+      "epoch": 11.827371336256634,
+      "grad_norm": 3.46484375,
+      "learning_rate": 1.4596397598398934e-05,
+      "loss": 1.1172,
+      "step": 12812
+    },
+    {
+      "epoch": 11.828294484191092,
+      "grad_norm": 3.79296875,
+      "learning_rate": 1.4589726484322883e-05,
+      "loss": 1.178,
+      "step": 12813
+    },
+    {
+      "epoch": 11.829217632125548,
+      "grad_norm": 2.884765625,
+      "learning_rate": 1.4583055370246832e-05,
+      "loss": 1.1609,
+      "step": 12814
+    },
+    {
+      "epoch": 11.830140780060004,
+      "grad_norm": 2.73046875,
+      "learning_rate": 1.4576384256170781e-05,
+      "loss": 1.2927,
+      "step": 12815
+    },
+    {
+      "epoch": 11.83106392799446,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.456971314209473e-05,
+      "loss": 1.4421,
+      "step": 12816
+    },
+    {
+      "epoch": 11.831987075928918,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.456304202801868e-05,
+      "loss": 1.3391,
+      "step": 12817
+    },
+    {
+      "epoch": 11.832910223863374,
+      "grad_norm": 3.310546875,
+      "learning_rate": 1.4556370913942629e-05,
+      "loss": 1.4114,
+      "step": 12818
+    },
+    {
+      "epoch": 11.83383337179783,
+      "grad_norm": 2.63671875,
+      "learning_rate": 1.4549699799866578e-05,
+      "loss": 1.0754,
+      "step": 12819
+    },
+    {
+      "epoch": 11.834756519732288,
+      "grad_norm": 3.095703125,
+      "learning_rate": 1.4543028685790527e-05,
+      "loss": 1.1475,
+      "step": 12820
+    },
+    {
+      "epoch": 11.835679667666744,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.4536357571714475e-05,
+      "loss": 1.3469,
+      "step": 12821
+    },
+    {
+      "epoch": 11.8366028156012,
+      "grad_norm": 3.134765625,
+      "learning_rate": 1.4529686457638424e-05,
+      "loss": 1.2996,
+      "step": 12822
+    },
+    {
+      "epoch": 11.837525963535656,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.4523015343562376e-05,
+      "loss": 1.3965,
+      "step": 12823
+    },
+    {
+      "epoch": 11.838449111470114,
+      "grad_norm": 3.02734375,
+      "learning_rate": 1.4516344229486325e-05,
+      "loss": 1.3494,
+      "step": 12824
+    },
+    {
+      "epoch": 11.83937225940457,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.4509673115410274e-05,
+      "loss": 1.3835,
+      "step": 12825
+    },
+    {
+      "epoch": 11.840295407339026,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.4503002001334223e-05,
+      "loss": 1.2234,
+      "step": 12826
+    },
+    {
+      "epoch": 11.841218555273482,
+      "grad_norm": 3.57421875,
+      "learning_rate": 1.4496330887258172e-05,
+      "loss": 1.1039,
+      "step": 12827
+    },
+    {
+      "epoch": 11.84214170320794,
+      "grad_norm": 3.150390625,
+      "learning_rate": 1.448965977318212e-05,
+      "loss": 1.0812,
+      "step": 12828
+    },
+    {
+      "epoch": 11.843064851142396,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.4482988659106073e-05,
+      "loss": 1.355,
+      "step": 12829
+    },
+    {
+      "epoch": 11.843987999076852,
+      "grad_norm": 3.166015625,
+      "learning_rate": 1.4476317545030022e-05,
+      "loss": 1.4622,
+      "step": 12830
+    },
+    {
+      "epoch": 11.844911147011308,
+      "grad_norm": 2.994140625,
+      "learning_rate": 1.446964643095397e-05,
+      "loss": 1.3958,
+      "step": 12831
+    },
+    {
+      "epoch": 11.845834294945766,
+      "grad_norm": 3.1171875,
+      "learning_rate": 1.446297531687792e-05,
+      "loss": 1.2725,
+      "step": 12832
+    },
+    {
+      "epoch": 11.846757442880222,
+      "grad_norm": 3.2578125,
+      "learning_rate": 1.4456304202801868e-05,
+      "loss": 1.1843,
+      "step": 12833
+    },
+    {
+      "epoch": 11.847680590814678,
+      "grad_norm": 3.013671875,
+      "learning_rate": 1.4449633088725817e-05,
+      "loss": 1.4197,
+      "step": 12834
+    },
+    {
+      "epoch": 11.848603738749134,
+      "grad_norm": 3.55078125,
+      "learning_rate": 1.4442961974649769e-05,
+      "loss": 1.0692,
+      "step": 12835
+    },
+    {
+      "epoch": 11.849526886683591,
+      "grad_norm": 2.8671875,
+      "learning_rate": 1.4436290860573718e-05,
+      "loss": 1.2903,
+      "step": 12836
+    },
+    {
+      "epoch": 11.850450034618047,
+      "grad_norm": 3.646484375,
+      "learning_rate": 1.4429619746497667e-05,
+      "loss": 1.3767,
+      "step": 12837
+    },
+    {
+      "epoch": 11.851373182552503,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.4422948632421616e-05,
+      "loss": 1.2739,
+      "step": 12838
+    },
+    {
+      "epoch": 11.852296330486961,
+      "grad_norm": 2.849609375,
+      "learning_rate": 1.4416277518345564e-05,
+      "loss": 1.4177,
+      "step": 12839
+    },
+    {
+      "epoch": 11.853219478421417,
+      "grad_norm": 3.298828125,
+      "learning_rate": 1.4409606404269513e-05,
+      "loss": 1.3475,
+      "step": 12840
+    },
+    {
+      "epoch": 11.854142626355873,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.4402935290193462e-05,
+      "loss": 1.5835,
+      "step": 12841
+    },
+    {
+      "epoch": 11.85506577429033,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.4396264176117413e-05,
+      "loss": 1.1875,
+      "step": 12842
+    },
+    {
+      "epoch": 11.855988922224787,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.4389593062041361e-05,
+      "loss": 1.3821,
+      "step": 12843
+    },
+    {
+      "epoch": 11.856912070159243,
+      "grad_norm": 2.955078125,
+      "learning_rate": 1.4382921947965312e-05,
+      "loss": 1.1688,
+      "step": 12844
+    },
+    {
+      "epoch": 11.8578352180937,
+      "grad_norm": 3.3515625,
+      "learning_rate": 1.437625083388926e-05,
+      "loss": 1.2524,
+      "step": 12845
+    },
+    {
+      "epoch": 11.858758366028155,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.436957971981321e-05,
+      "loss": 1.2483,
+      "step": 12846
+    },
+    {
+      "epoch": 11.859681513962613,
+      "grad_norm": 2.923828125,
+      "learning_rate": 1.4362908605737158e-05,
+      "loss": 1.1172,
+      "step": 12847
+    },
+    {
+      "epoch": 11.860604661897069,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.4356237491661109e-05,
+      "loss": 1.3367,
+      "step": 12848
+    },
+    {
+      "epoch": 11.861527809831525,
+      "grad_norm": 2.998046875,
+      "learning_rate": 1.4349566377585058e-05,
+      "loss": 1.2156,
+      "step": 12849
+    },
+    {
+      "epoch": 11.862450957765983,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.4342895263509006e-05,
+      "loss": 1.2644,
+      "step": 12850
+    },
+    {
+      "epoch": 11.863374105700439,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.4336224149432955e-05,
+      "loss": 1.2905,
+      "step": 12851
+    },
+    {
+      "epoch": 11.864297253634895,
+      "grad_norm": 2.9609375,
+      "learning_rate": 1.4329553035356904e-05,
+      "loss": 1.395,
+      "step": 12852
+    },
+    {
+      "epoch": 11.865220401569351,
+      "grad_norm": 3.7578125,
+      "learning_rate": 1.4322881921280853e-05,
+      "loss": 1.2103,
+      "step": 12853
+    },
+    {
+      "epoch": 11.866143549503809,
+      "grad_norm": 3.1171875,
+      "learning_rate": 1.4316210807204805e-05,
+      "loss": 1.2583,
+      "step": 12854
+    },
+    {
+      "epoch": 11.867066697438265,
+      "grad_norm": 3.0859375,
+      "learning_rate": 1.4309539693128754e-05,
+      "loss": 1.1232,
+      "step": 12855
+    },
+    {
+      "epoch": 11.86798984537272,
+      "grad_norm": 3.220703125,
+      "learning_rate": 1.4302868579052703e-05,
+      "loss": 1.5784,
+      "step": 12856
+    },
+    {
+      "epoch": 11.868912993307177,
+      "grad_norm": 3.01171875,
+      "learning_rate": 1.4296197464976652e-05,
+      "loss": 1.2769,
+      "step": 12857
+    },
+    {
+      "epoch": 11.869836141241635,
+      "grad_norm": 3.427734375,
+      "learning_rate": 1.42895263509006e-05,
+      "loss": 1.2795,
+      "step": 12858
+    },
+    {
+      "epoch": 11.87075928917609,
+      "grad_norm": 3.052734375,
+      "learning_rate": 1.428285523682455e-05,
+      "loss": 1.3787,
+      "step": 12859
+    },
+    {
+      "epoch": 11.871682437110547,
+      "grad_norm": 3.12109375,
+      "learning_rate": 1.4276184122748501e-05,
+      "loss": 1.0781,
+      "step": 12860
+    },
+    {
+      "epoch": 11.872605585045003,
+      "grad_norm": 3.08984375,
+      "learning_rate": 1.426951300867245e-05,
+      "loss": 1.2611,
+      "step": 12861
+    },
+    {
+      "epoch": 11.87352873297946,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.4262841894596399e-05,
+      "loss": 1.0942,
+      "step": 12862
+    },
+    {
+      "epoch": 11.874451880913917,
+      "grad_norm": 3.02734375,
+      "learning_rate": 1.4256170780520348e-05,
+      "loss": 1.2837,
+      "step": 12863
+    },
+    {
+      "epoch": 11.875375028848373,
+      "grad_norm": 3.341796875,
+      "learning_rate": 1.4249499666444297e-05,
+      "loss": 1.4514,
+      "step": 12864
+    },
+    {
+      "epoch": 11.876298176782829,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.4242828552368246e-05,
+      "loss": 0.8336,
+      "step": 12865
+    },
+    {
+      "epoch": 11.877221324717286,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4236157438292194e-05,
+      "loss": 1.4236,
+      "step": 12866
+    },
+    {
+      "epoch": 11.878144472651742,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.4229486324216147e-05,
+      "loss": 1.1184,
+      "step": 12867
+    },
+    {
+      "epoch": 11.879067620586198,
+      "grad_norm": 3.1015625,
+      "learning_rate": 1.4222815210140095e-05,
+      "loss": 1.3623,
+      "step": 12868
+    },
+    {
+      "epoch": 11.879990768520656,
+      "grad_norm": 3.14453125,
+      "learning_rate": 1.4216144096064044e-05,
+      "loss": 1.1055,
+      "step": 12869
+    },
+    {
+      "epoch": 11.880913916455112,
+      "grad_norm": 3.376953125,
+      "learning_rate": 1.4209472981987993e-05,
+      "loss": 1.4292,
+      "step": 12870
+    },
+    {
+      "epoch": 11.881837064389568,
+      "grad_norm": 3.294921875,
+      "learning_rate": 1.4202801867911942e-05,
+      "loss": 1.3743,
+      "step": 12871
+    },
+    {
+      "epoch": 11.882760212324024,
+      "grad_norm": 3.4453125,
+      "learning_rate": 1.419613075383589e-05,
+      "loss": 1.2341,
+      "step": 12872
+    },
+    {
+      "epoch": 11.883683360258482,
+      "grad_norm": 2.916015625,
+      "learning_rate": 1.4189459639759841e-05,
+      "loss": 1.4272,
+      "step": 12873
+    },
+    {
+      "epoch": 11.884606508192938,
+      "grad_norm": 3.51171875,
+      "learning_rate": 1.418278852568379e-05,
+      "loss": 1.3474,
+      "step": 12874
+    },
+    {
+      "epoch": 11.885529656127394,
+      "grad_norm": 3.30859375,
+      "learning_rate": 1.4176117411607739e-05,
+      "loss": 1.3796,
+      "step": 12875
+    },
+    {
+      "epoch": 11.88645280406185,
+      "grad_norm": 3.322265625,
+      "learning_rate": 1.4169446297531688e-05,
+      "loss": 1.156,
+      "step": 12876
+    },
+    {
+      "epoch": 11.887375951996308,
+      "grad_norm": 2.791015625,
+      "learning_rate": 1.4162775183455636e-05,
+      "loss": 1.297,
+      "step": 12877
+    },
+    {
+      "epoch": 11.888299099930764,
+      "grad_norm": 3.486328125,
+      "learning_rate": 1.4156104069379585e-05,
+      "loss": 1.355,
+      "step": 12878
+    },
+    {
+      "epoch": 11.88922224786522,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.4149432955303537e-05,
+      "loss": 1.1132,
+      "step": 12879
+    },
+    {
+      "epoch": 11.890145395799676,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.4142761841227486e-05,
+      "loss": 1.287,
+      "step": 12880
+    },
+    {
+      "epoch": 11.891068543734134,
+      "grad_norm": 2.99609375,
+      "learning_rate": 1.4136090727151435e-05,
+      "loss": 1.2291,
+      "step": 12881
+    },
+    {
+      "epoch": 11.89199169166859,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.4129419613075384e-05,
+      "loss": 1.2398,
+      "step": 12882
+    },
+    {
+      "epoch": 11.892914839603046,
+      "grad_norm": 3.548828125,
+      "learning_rate": 1.4122748498999333e-05,
+      "loss": 1.2107,
+      "step": 12883
+    },
+    {
+      "epoch": 11.893837987537502,
+      "grad_norm": 3.166015625,
+      "learning_rate": 1.4116077384923282e-05,
+      "loss": 1.3167,
+      "step": 12884
+    },
+    {
+      "epoch": 11.89476113547196,
+      "grad_norm": 2.994140625,
+      "learning_rate": 1.4109406270847234e-05,
+      "loss": 1.2212,
+      "step": 12885
+    },
+    {
+      "epoch": 11.895684283406416,
+      "grad_norm": 3.095703125,
+      "learning_rate": 1.4102735156771183e-05,
+      "loss": 1.3411,
+      "step": 12886
+    },
+    {
+      "epoch": 11.896607431340872,
+      "grad_norm": 3.341796875,
+      "learning_rate": 1.4096064042695131e-05,
+      "loss": 1.1107,
+      "step": 12887
+    },
+    {
+      "epoch": 11.89753057927533,
+      "grad_norm": 3.384765625,
+      "learning_rate": 1.408939292861908e-05,
+      "loss": 1.3105,
+      "step": 12888
+    },
+    {
+      "epoch": 11.898453727209786,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.4082721814543029e-05,
+      "loss": 1.2169,
+      "step": 12889
+    },
+    {
+      "epoch": 11.899376875144242,
+      "grad_norm": 3.115234375,
+      "learning_rate": 1.4076050700466978e-05,
+      "loss": 1.1736,
+      "step": 12890
+    },
+    {
+      "epoch": 11.900300023078698,
+      "grad_norm": 3.650390625,
+      "learning_rate": 1.4069379586390927e-05,
+      "loss": 1.4834,
+      "step": 12891
+    },
+    {
+      "epoch": 11.901223171013156,
+      "grad_norm": 3.833984375,
+      "learning_rate": 1.4062708472314879e-05,
+      "loss": 1.1216,
+      "step": 12892
+    },
+    {
+      "epoch": 11.902146318947612,
+      "grad_norm": 3.0546875,
+      "learning_rate": 1.4056037358238828e-05,
+      "loss": 1.2037,
+      "step": 12893
+    },
+    {
+      "epoch": 11.903069466882068,
+      "grad_norm": 3.8046875,
+      "learning_rate": 1.4049366244162776e-05,
+      "loss": 1.3269,
+      "step": 12894
+    },
+    {
+      "epoch": 11.903992614816524,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.4042695130086725e-05,
+      "loss": 1.3792,
+      "step": 12895
+    },
+    {
+      "epoch": 11.904915762750981,
+      "grad_norm": 3.5703125,
+      "learning_rate": 1.4036024016010674e-05,
+      "loss": 1.1561,
+      "step": 12896
+    },
+    {
+      "epoch": 11.905838910685437,
+      "grad_norm": 3.517578125,
+      "learning_rate": 1.4029352901934623e-05,
+      "loss": 1.3892,
+      "step": 12897
+    },
+    {
+      "epoch": 11.906762058619893,
+      "grad_norm": 3.107421875,
+      "learning_rate": 1.4022681787858573e-05,
+      "loss": 1.2755,
+      "step": 12898
+    },
+    {
+      "epoch": 11.907685206554351,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.4016010673782522e-05,
+      "loss": 1.2292,
+      "step": 12899
+    },
+    {
+      "epoch": 11.908608354488807,
+      "grad_norm": 3.126953125,
+      "learning_rate": 1.4009339559706471e-05,
+      "loss": 1.3124,
+      "step": 12900
+    },
+    {
+      "epoch": 11.909531502423263,
+      "grad_norm": 3.982421875,
+      "learning_rate": 1.4002668445630422e-05,
+      "loss": 1.3281,
+      "step": 12901
+    },
+    {
+      "epoch": 11.91045465035772,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.399599733155437e-05,
+      "loss": 1.1584,
+      "step": 12902
+    },
+    {
+      "epoch": 11.911377798292177,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.398932621747832e-05,
+      "loss": 1.3345,
+      "step": 12903
+    },
+    {
+      "epoch": 11.912300946226633,
+      "grad_norm": 3.2109375,
+      "learning_rate": 1.398265510340227e-05,
+      "loss": 1.0518,
+      "step": 12904
+    },
+    {
+      "epoch": 11.91322409416109,
+      "grad_norm": 3.47265625,
+      "learning_rate": 1.3975983989326219e-05,
+      "loss": 1.2947,
+      "step": 12905
+    },
+    {
+      "epoch": 11.914147242095545,
+      "grad_norm": 3.6796875,
+      "learning_rate": 1.3969312875250167e-05,
+      "loss": 0.9403,
+      "step": 12906
+    },
+    {
+      "epoch": 11.915070390030003,
+      "grad_norm": 3.439453125,
+      "learning_rate": 1.3962641761174116e-05,
+      "loss": 1.1934,
+      "step": 12907
+    },
+    {
+      "epoch": 11.915993537964459,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.3955970647098065e-05,
+      "loss": 1.0971,
+      "step": 12908
+    },
+    {
+      "epoch": 11.916916685898915,
+      "grad_norm": 2.708984375,
+      "learning_rate": 1.3949299533022014e-05,
+      "loss": 1.2637,
+      "step": 12909
+    },
+    {
+      "epoch": 11.917839833833371,
+      "grad_norm": 2.869140625,
+      "learning_rate": 1.3942628418945963e-05,
+      "loss": 1.4692,
+      "step": 12910
+    },
+    {
+      "epoch": 11.918762981767829,
+      "grad_norm": 3.30078125,
+      "learning_rate": 1.3935957304869915e-05,
+      "loss": 1.0685,
+      "step": 12911
+    },
+    {
+      "epoch": 11.919686129702285,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.3929286190793864e-05,
+      "loss": 1.0844,
+      "step": 12912
+    },
+    {
+      "epoch": 11.920609277636741,
+      "grad_norm": 3.052734375,
+      "learning_rate": 1.3922615076717813e-05,
+      "loss": 1.1456,
+      "step": 12913
+    },
+    {
+      "epoch": 11.921532425571197,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.3915943962641761e-05,
+      "loss": 1.1033,
+      "step": 12914
+    },
+    {
+      "epoch": 11.922455573505655,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.390927284856571e-05,
+      "loss": 1.0804,
+      "step": 12915
+    },
+    {
+      "epoch": 11.92337872144011,
+      "grad_norm": 3.076171875,
+      "learning_rate": 1.3902601734489659e-05,
+      "loss": 0.8643,
+      "step": 12916
+    },
+    {
+      "epoch": 11.924301869374567,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.3895930620413611e-05,
+      "loss": 1.076,
+      "step": 12917
+    },
+    {
+      "epoch": 11.925225017309025,
+      "grad_norm": 2.97265625,
+      "learning_rate": 1.388925950633756e-05,
+      "loss": 1.0573,
+      "step": 12918
+    },
+    {
+      "epoch": 11.92614816524348,
+      "grad_norm": 2.96484375,
+      "learning_rate": 1.3882588392261509e-05,
+      "loss": 1.3433,
+      "step": 12919
+    },
+    {
+      "epoch": 11.927071313177937,
+      "grad_norm": 3.541015625,
+      "learning_rate": 1.3875917278185458e-05,
+      "loss": 1.3896,
+      "step": 12920
+    },
+    {
+      "epoch": 11.927994461112393,
+      "grad_norm": 3.5546875,
+      "learning_rate": 1.3869246164109406e-05,
+      "loss": 1.3447,
+      "step": 12921
+    },
+    {
+      "epoch": 11.92891760904685,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.3862575050033355e-05,
+      "loss": 1.2198,
+      "step": 12922
+    },
+    {
+      "epoch": 11.929840756981307,
+      "grad_norm": 3.451171875,
+      "learning_rate": 1.3855903935957307e-05,
+      "loss": 1.2684,
+      "step": 12923
+    },
+    {
+      "epoch": 11.930763904915763,
+      "grad_norm": 2.86328125,
+      "learning_rate": 1.3849232821881256e-05,
+      "loss": 1.2767,
+      "step": 12924
+    },
+    {
+      "epoch": 11.931687052850219,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.3842561707805205e-05,
+      "loss": 1.4355,
+      "step": 12925
+    },
+    {
+      "epoch": 11.932610200784676,
+      "grad_norm": 3.330078125,
+      "learning_rate": 1.3835890593729154e-05,
+      "loss": 1.188,
+      "step": 12926
+    },
+    {
+      "epoch": 11.933533348719132,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3829219479653103e-05,
+      "loss": 1.4053,
+      "step": 12927
+    },
+    {
+      "epoch": 11.934456496653588,
+      "grad_norm": 3.189453125,
+      "learning_rate": 1.3822548365577052e-05,
+      "loss": 1.3313,
+      "step": 12928
+    },
+    {
+      "epoch": 11.935379644588044,
+      "grad_norm": 3.36328125,
+      "learning_rate": 1.3815877251501002e-05,
+      "loss": 1.1089,
+      "step": 12929
+    },
+    {
+      "epoch": 11.936302792522502,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.3809206137424951e-05,
+      "loss": 1.3921,
+      "step": 12930
+    },
+    {
+      "epoch": 11.937225940456958,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.38025350233489e-05,
+      "loss": 1.3523,
+      "step": 12931
+    },
+    {
+      "epoch": 11.938149088391414,
+      "grad_norm": 3.146484375,
+      "learning_rate": 1.3795863909272849e-05,
+      "loss": 0.9852,
+      "step": 12932
+    },
+    {
+      "epoch": 11.93907223632587,
+      "grad_norm": 2.98828125,
+      "learning_rate": 1.3789192795196797e-05,
+      "loss": 1.1681,
+      "step": 12933
+    },
+    {
+      "epoch": 11.939995384260328,
+      "grad_norm": 3.119140625,
+      "learning_rate": 1.3782521681120746e-05,
+      "loss": 1.2966,
+      "step": 12934
+    },
+    {
+      "epoch": 11.940918532194784,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.3775850567044695e-05,
+      "loss": 1.3135,
+      "step": 12935
+    },
+    {
+      "epoch": 11.94184168012924,
+      "grad_norm": 3.0390625,
+      "learning_rate": 1.3769179452968647e-05,
+      "loss": 1.3472,
+      "step": 12936
+    },
+    {
+      "epoch": 11.942764828063698,
+      "grad_norm": 3.267578125,
+      "learning_rate": 1.3762508338892596e-05,
+      "loss": 1.196,
+      "step": 12937
+    },
+    {
+      "epoch": 11.943687975998154,
+      "grad_norm": 2.97265625,
+      "learning_rate": 1.3755837224816545e-05,
+      "loss": 1.1963,
+      "step": 12938
+    },
+    {
+      "epoch": 11.94461112393261,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.3749166110740494e-05,
+      "loss": 1.4846,
+      "step": 12939
+    },
+    {
+      "epoch": 11.945534271867066,
+      "grad_norm": 3.638671875,
+      "learning_rate": 1.3742494996664442e-05,
+      "loss": 1.1418,
+      "step": 12940
+    },
+    {
+      "epoch": 11.946457419801524,
+      "grad_norm": 2.8828125,
+      "learning_rate": 1.3735823882588391e-05,
+      "loss": 1.1792,
+      "step": 12941
+    },
+    {
+      "epoch": 11.94738056773598,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.3729152768512344e-05,
+      "loss": 1.5052,
+      "step": 12942
+    },
+    {
+      "epoch": 11.948303715670436,
+      "grad_norm": 3.681640625,
+      "learning_rate": 1.3722481654436292e-05,
+      "loss": 1.2997,
+      "step": 12943
+    },
+    {
+      "epoch": 11.949226863604892,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.3715810540360241e-05,
+      "loss": 1.3452,
+      "step": 12944
+    },
+    {
+      "epoch": 11.95015001153935,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.370913942628419e-05,
+      "loss": 1.3945,
+      "step": 12945
+    },
+    {
+      "epoch": 11.951073159473806,
+      "grad_norm": 3.369140625,
+      "learning_rate": 1.3702468312208139e-05,
+      "loss": 1.4773,
+      "step": 12946
+    },
+    {
+      "epoch": 11.951996307408262,
+      "grad_norm": 3.240234375,
+      "learning_rate": 1.3695797198132088e-05,
+      "loss": 0.7851,
+      "step": 12947
+    },
+    {
+      "epoch": 11.95291945534272,
+      "grad_norm": 3.263671875,
+      "learning_rate": 1.368912608405604e-05,
+      "loss": 1.5852,
+      "step": 12948
+    },
+    {
+      "epoch": 11.953842603277176,
+      "grad_norm": 3.005859375,
+      "learning_rate": 1.3682454969979989e-05,
+      "loss": 1.3499,
+      "step": 12949
+    },
+    {
+      "epoch": 11.954765751211632,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.3675783855903937e-05,
+      "loss": 1.4734,
+      "step": 12950
+    },
+    {
+      "epoch": 11.955688899146088,
+      "grad_norm": 3.33984375,
+      "learning_rate": 1.3669112741827886e-05,
+      "loss": 1.4097,
+      "step": 12951
+    },
+    {
+      "epoch": 11.956612047080545,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.3662441627751835e-05,
+      "loss": 1.5061,
+      "step": 12952
+    },
+    {
+      "epoch": 11.957535195015002,
+      "grad_norm": 3.10546875,
+      "learning_rate": 1.3655770513675784e-05,
+      "loss": 1.4617,
+      "step": 12953
+    },
+    {
+      "epoch": 11.958458342949458,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.3649099399599734e-05,
+      "loss": 1.1812,
+      "step": 12954
+    },
+    {
+      "epoch": 11.959381490883914,
+      "grad_norm": 3.1015625,
+      "learning_rate": 1.3642428285523683e-05,
+      "loss": 1.2139,
+      "step": 12955
+    },
+    {
+      "epoch": 11.960304638818371,
+      "grad_norm": 3.08203125,
+      "learning_rate": 1.3635757171447632e-05,
+      "loss": 1.4392,
+      "step": 12956
+    },
+    {
+      "epoch": 11.961227786752827,
+      "grad_norm": 3.22265625,
+      "learning_rate": 1.3629086057371583e-05,
+      "loss": 1.5247,
+      "step": 12957
+    },
+    {
+      "epoch": 11.962150934687283,
+      "grad_norm": 3.490234375,
+      "learning_rate": 1.3622414943295531e-05,
+      "loss": 1.0908,
+      "step": 12958
+    },
+    {
+      "epoch": 11.96307408262174,
+      "grad_norm": 3.341796875,
+      "learning_rate": 1.361574382921948e-05,
+      "loss": 1.3711,
+      "step": 12959
+    },
+    {
+      "epoch": 11.963997230556197,
+      "grad_norm": 3.06640625,
+      "learning_rate": 1.3609072715143429e-05,
+      "loss": 1.1437,
+      "step": 12960
+    },
+    {
+      "epoch": 11.964920378490653,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.360240160106738e-05,
+      "loss": 1.406,
+      "step": 12961
+    },
+    {
+      "epoch": 11.96584352642511,
+      "grad_norm": 3.201171875,
+      "learning_rate": 1.3595730486991328e-05,
+      "loss": 1.0682,
+      "step": 12962
+    },
+    {
+      "epoch": 11.966766674359565,
+      "grad_norm": 2.55078125,
+      "learning_rate": 1.3589059372915277e-05,
+      "loss": 1.0651,
+      "step": 12963
+    },
+    {
+      "epoch": 11.967689822294023,
+      "grad_norm": 3.458984375,
+      "learning_rate": 1.3582388258839226e-05,
+      "loss": 1.4121,
+      "step": 12964
+    },
+    {
+      "epoch": 11.96861297022848,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.3575717144763175e-05,
+      "loss": 1.2708,
+      "step": 12965
+    },
+    {
+      "epoch": 11.969536118162935,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3569046030687124e-05,
+      "loss": 0.9932,
+      "step": 12966
+    },
+    {
+      "epoch": 11.970459266097393,
+      "grad_norm": 3.353515625,
+      "learning_rate": 1.3562374916611076e-05,
+      "loss": 1.4102,
+      "step": 12967
+    },
+    {
+      "epoch": 11.971382414031849,
+      "grad_norm": 3.07421875,
+      "learning_rate": 1.3555703802535025e-05,
+      "loss": 1.2812,
+      "step": 12968
+    },
+    {
+      "epoch": 11.972305561966305,
+      "grad_norm": 3.3828125,
+      "learning_rate": 1.3549032688458973e-05,
+      "loss": 1.0414,
+      "step": 12969
+    },
+    {
+      "epoch": 11.973228709900761,
+      "grad_norm": 3.267578125,
+      "learning_rate": 1.3542361574382922e-05,
+      "loss": 1.3105,
+      "step": 12970
+    },
+    {
+      "epoch": 11.974151857835219,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.3535690460306871e-05,
+      "loss": 1.5486,
+      "step": 12971
+    },
+    {
+      "epoch": 11.975075005769675,
+      "grad_norm": 3.38671875,
+      "learning_rate": 1.352901934623082e-05,
+      "loss": 1.355,
+      "step": 12972
+    },
+    {
+      "epoch": 11.975998153704131,
+      "grad_norm": 3.439453125,
+      "learning_rate": 1.3522348232154772e-05,
+      "loss": 1.1488,
+      "step": 12973
+    },
+    {
+      "epoch": 11.976921301638587,
+      "grad_norm": 2.794921875,
+      "learning_rate": 1.3515677118078721e-05,
+      "loss": 1.0675,
+      "step": 12974
+    },
+    {
+      "epoch": 11.977844449573045,
+      "grad_norm": 3.36328125,
+      "learning_rate": 1.350900600400267e-05,
+      "loss": 1.2406,
+      "step": 12975
+    },
+    {
+      "epoch": 11.9787675975075,
+      "grad_norm": 3.0703125,
+      "learning_rate": 1.3502334889926619e-05,
+      "loss": 0.9124,
+      "step": 12976
+    },
+    {
+      "epoch": 11.979690745441957,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.3495663775850567e-05,
+      "loss": 1.2927,
+      "step": 12977
+    },
+    {
+      "epoch": 11.980613893376413,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.3488992661774516e-05,
+      "loss": 1.3442,
+      "step": 12978
+    },
+    {
+      "epoch": 11.98153704131087,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.3482321547698468e-05,
+      "loss": 1.2751,
+      "step": 12979
+    },
+    {
+      "epoch": 11.982460189245327,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.3475650433622417e-05,
+      "loss": 1.4648,
+      "step": 12980
+    },
+    {
+      "epoch": 11.983383337179783,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.3468979319546366e-05,
+      "loss": 1.2693,
+      "step": 12981
+    },
+    {
+      "epoch": 11.984306485114239,
+      "grad_norm": 3.0390625,
+      "learning_rate": 1.3462308205470315e-05,
+      "loss": 1.2654,
+      "step": 12982
+    },
+    {
+      "epoch": 11.985229633048696,
+      "grad_norm": 3.201171875,
+      "learning_rate": 1.3455637091394264e-05,
+      "loss": 1.3271,
+      "step": 12983
+    },
+    {
+      "epoch": 11.986152780983153,
+      "grad_norm": 2.912109375,
+      "learning_rate": 1.3448965977318213e-05,
+      "loss": 1.3325,
+      "step": 12984
+    },
+    {
+      "epoch": 11.987075928917609,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.3442294863242161e-05,
+      "loss": 1.2722,
+      "step": 12985
+    },
+    {
+      "epoch": 11.987999076852066,
+      "grad_norm": 2.98046875,
+      "learning_rate": 1.3435623749166112e-05,
+      "loss": 1.5305,
+      "step": 12986
+    },
+    {
+      "epoch": 11.988922224786522,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.342895263509006e-05,
+      "loss": 1.0529,
+      "step": 12987
+    },
+    {
+      "epoch": 11.989845372720978,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.342228152101401e-05,
+      "loss": 0.7326,
+      "step": 12988
+    },
+    {
+      "epoch": 11.990768520655434,
+      "grad_norm": 3.30859375,
+      "learning_rate": 1.3415610406937958e-05,
+      "loss": 1.2487,
+      "step": 12989
+    },
+    {
+      "epoch": 11.991691668589892,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.3408939292861907e-05,
+      "loss": 1.3811,
+      "step": 12990
+    },
+    {
+      "epoch": 11.992614816524348,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.3402268178785856e-05,
+      "loss": 1.2386,
+      "step": 12991
+    },
+    {
+      "epoch": 11.993537964458804,
+      "grad_norm": 3.44140625,
+      "learning_rate": 1.3395597064709808e-05,
+      "loss": 1.53,
+      "step": 12992
+    },
+    {
+      "epoch": 11.99446111239326,
+      "grad_norm": 2.919921875,
+      "learning_rate": 1.3388925950633757e-05,
+      "loss": 1.3384,
+      "step": 12993
+    },
+    {
+      "epoch": 11.995384260327718,
+      "grad_norm": 3.267578125,
+      "learning_rate": 1.3382254836557706e-05,
+      "loss": 1.4084,
+      "step": 12994
+    },
+    {
+      "epoch": 11.996307408262174,
+      "grad_norm": 2.89453125,
+      "learning_rate": 1.3375583722481655e-05,
+      "loss": 1.2231,
+      "step": 12995
+    },
+    {
+      "epoch": 11.99723055619663,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.3368912608405603e-05,
+      "loss": 1.3029,
+      "step": 12996
+    },
+    {
+      "epoch": 11.998153704131088,
+      "grad_norm": 2.89453125,
+      "learning_rate": 1.3362241494329552e-05,
+      "loss": 1.1041,
+      "step": 12997
+    },
+    {
+      "epoch": 11.999076852065544,
+      "grad_norm": 2.666015625,
+      "learning_rate": 1.3355570380253504e-05,
+      "loss": 1.459,
+      "step": 12998
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 3.376953125,
+      "learning_rate": 1.3348899266177453e-05,
+      "loss": 0.6998,
+      "step": 12999
+    },
+    {
+      "epoch": 12.000923147934456,
+      "grad_norm": 3.255859375,
+      "learning_rate": 1.3342228152101402e-05,
+      "loss": 1.2156,
+      "step": 13000
+    },
+    {
+      "epoch": 12.000923147934456,
+      "eval_loss": 0.86962890625,
+      "eval_runtime": 23.3966,
+      "eval_samples_per_second": 974.542,
+      "eval_steps_per_second": 60.949,
+      "step": 13000
+    },
+    {
+      "epoch": 12.001846295868914,
+      "grad_norm": 3.576171875,
+      "learning_rate": 1.3335557038025351e-05,
+      "loss": 1.3135,
+      "step": 13001
+    },
+    {
+      "epoch": 12.00276944380337,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.33288859239493e-05,
+      "loss": 1.1882,
+      "step": 13002
+    },
+    {
+      "epoch": 12.003692591737826,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.3322214809873249e-05,
+      "loss": 1.4897,
+      "step": 13003
+    },
+    {
+      "epoch": 12.004615739672282,
+      "grad_norm": 3.125,
+      "learning_rate": 1.3315543695797197e-05,
+      "loss": 1.3127,
+      "step": 13004
+    },
+    {
+      "epoch": 12.00553888760674,
+      "grad_norm": 3.69140625,
+      "learning_rate": 1.330887258172115e-05,
+      "loss": 1.2578,
+      "step": 13005
+    },
+    {
+      "epoch": 12.006462035541196,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.3302201467645098e-05,
+      "loss": 1.1646,
+      "step": 13006
+    },
+    {
+      "epoch": 12.007385183475652,
+      "grad_norm": 3.12890625,
+      "learning_rate": 1.3295530353569047e-05,
+      "loss": 1.4507,
+      "step": 13007
+    },
+    {
+      "epoch": 12.008308331410108,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.3288859239492996e-05,
+      "loss": 1.2783,
+      "step": 13008
+    },
+    {
+      "epoch": 12.009231479344566,
+      "grad_norm": 3.314453125,
+      "learning_rate": 1.3282188125416945e-05,
+      "loss": 1.1638,
+      "step": 13009
+    },
+    {
+      "epoch": 12.010154627279022,
+      "grad_norm": 3.130859375,
+      "learning_rate": 1.3275517011340894e-05,
+      "loss": 1.2659,
+      "step": 13010
+    },
+    {
+      "epoch": 12.011077775213478,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.3268845897264844e-05,
+      "loss": 1.3171,
+      "step": 13011
+    },
+    {
+      "epoch": 12.012000923147934,
+      "grad_norm": 3.037109375,
+      "learning_rate": 1.3262174783188793e-05,
+      "loss": 1.4592,
+      "step": 13012
+    },
+    {
+      "epoch": 12.012924071082391,
+      "grad_norm": 3.294921875,
+      "learning_rate": 1.3255503669112742e-05,
+      "loss": 1.2786,
+      "step": 13013
+    },
+    {
+      "epoch": 12.013847219016847,
+      "grad_norm": 2.998046875,
+      "learning_rate": 1.3248832555036692e-05,
+      "loss": 1.269,
+      "step": 13014
+    },
+    {
+      "epoch": 12.014770366951304,
+      "grad_norm": 3.1796875,
+      "learning_rate": 1.3242161440960641e-05,
+      "loss": 1.3542,
+      "step": 13015
+    },
+    {
+      "epoch": 12.015693514885761,
+      "grad_norm": 3.384765625,
+      "learning_rate": 1.323549032688459e-05,
+      "loss": 1.2168,
+      "step": 13016
+    },
+    {
+      "epoch": 12.016616662820217,
+      "grad_norm": 2.89453125,
+      "learning_rate": 1.322881921280854e-05,
+      "loss": 1.3403,
+      "step": 13017
+    },
+    {
+      "epoch": 12.017539810754673,
+      "grad_norm": 3.326171875,
+      "learning_rate": 1.322214809873249e-05,
+      "loss": 1.1718,
+      "step": 13018
+    },
+    {
+      "epoch": 12.01846295868913,
+      "grad_norm": 3.15234375,
+      "learning_rate": 1.3215476984656438e-05,
+      "loss": 0.8606,
+      "step": 13019
+    },
+    {
+      "epoch": 12.019386106623587,
+      "grad_norm": 3.40234375,
+      "learning_rate": 1.3208805870580387e-05,
+      "loss": 1.0659,
+      "step": 13020
+    },
+    {
+      "epoch": 12.020309254558043,
+      "grad_norm": 3.005859375,
+      "learning_rate": 1.3202134756504336e-05,
+      "loss": 0.9664,
+      "step": 13021
+    },
+    {
+      "epoch": 12.0212324024925,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.3195463642428285e-05,
+      "loss": 1.4346,
+      "step": 13022
+    },
+    {
+      "epoch": 12.022155550426955,
+      "grad_norm": 3.4296875,
+      "learning_rate": 1.3188792528352237e-05,
+      "loss": 1.2844,
+      "step": 13023
+    },
+    {
+      "epoch": 12.023078698361413,
+      "grad_norm": 3.646484375,
+      "learning_rate": 1.3182121414276186e-05,
+      "loss": 1.3656,
+      "step": 13024
+    },
+    {
+      "epoch": 12.024001846295869,
+      "grad_norm": 3.12890625,
+      "learning_rate": 1.3175450300200134e-05,
+      "loss": 1.3274,
+      "step": 13025
+    },
+    {
+      "epoch": 12.024924994230325,
+      "grad_norm": 2.853515625,
+      "learning_rate": 1.3168779186124083e-05,
+      "loss": 1.2627,
+      "step": 13026
+    },
+    {
+      "epoch": 12.025848142164781,
+      "grad_norm": 3.126953125,
+      "learning_rate": 1.3162108072048032e-05,
+      "loss": 1.1188,
+      "step": 13027
+    },
+    {
+      "epoch": 12.026771290099239,
+      "grad_norm": 3.1015625,
+      "learning_rate": 1.3155436957971981e-05,
+      "loss": 1.1777,
+      "step": 13028
+    },
+    {
+      "epoch": 12.027694438033695,
+      "grad_norm": 3.24609375,
+      "learning_rate": 1.314876584389593e-05,
+      "loss": 1.136,
+      "step": 13029
+    },
+    {
+      "epoch": 12.028617585968151,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.3142094729819882e-05,
+      "loss": 1.1162,
+      "step": 13030
+    },
+    {
+      "epoch": 12.029540733902607,
+      "grad_norm": 2.75390625,
+      "learning_rate": 1.313542361574383e-05,
+      "loss": 1.2808,
+      "step": 13031
+    },
+    {
+      "epoch": 12.030463881837065,
+      "grad_norm": 3.11328125,
+      "learning_rate": 1.312875250166778e-05,
+      "loss": 1.4036,
+      "step": 13032
+    },
+    {
+      "epoch": 12.03138702977152,
+      "grad_norm": 3.0,
+      "learning_rate": 1.3122081387591728e-05,
+      "loss": 1.3853,
+      "step": 13033
+    },
+    {
+      "epoch": 12.032310177705977,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.3115410273515677e-05,
+      "loss": 1.1102,
+      "step": 13034
+    },
+    {
+      "epoch": 12.033233325640435,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.3108739159439626e-05,
+      "loss": 1.2338,
+      "step": 13035
+    },
+    {
+      "epoch": 12.03415647357489,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.3102068045363578e-05,
+      "loss": 1.4282,
+      "step": 13036
+    },
+    {
+      "epoch": 12.035079621509347,
+      "grad_norm": 3.1640625,
+      "learning_rate": 1.3095396931287527e-05,
+      "loss": 1.2112,
+      "step": 13037
+    },
+    {
+      "epoch": 12.036002769443803,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.3088725817211476e-05,
+      "loss": 1.1274,
+      "step": 13038
+    },
+    {
+      "epoch": 12.03692591737826,
+      "grad_norm": 2.92578125,
+      "learning_rate": 1.3082054703135425e-05,
+      "loss": 1.2961,
+      "step": 13039
+    },
+    {
+      "epoch": 12.037849065312717,
+      "grad_norm": 4.1953125,
+      "learning_rate": 1.3075383589059373e-05,
+      "loss": 1.203,
+      "step": 13040
+    },
+    {
+      "epoch": 12.038772213247173,
+      "grad_norm": 3.126953125,
+      "learning_rate": 1.3068712474983322e-05,
+      "loss": 1.087,
+      "step": 13041
+    },
+    {
+      "epoch": 12.039695361181629,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.3062041360907273e-05,
+      "loss": 1.2156,
+      "step": 13042
+    },
+    {
+      "epoch": 12.040618509116086,
+      "grad_norm": 3.50390625,
+      "learning_rate": 1.3055370246831222e-05,
+      "loss": 1.2681,
+      "step": 13043
+    },
+    {
+      "epoch": 12.041541657050542,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.304869913275517e-05,
+      "loss": 1.4951,
+      "step": 13044
+    },
+    {
+      "epoch": 12.042464804984998,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.304202801867912e-05,
+      "loss": 1.1633,
+      "step": 13045
+    },
+    {
+      "epoch": 12.043387952919455,
+      "grad_norm": 3.58984375,
+      "learning_rate": 1.3035356904603068e-05,
+      "loss": 1.0667,
+      "step": 13046
+    },
+    {
+      "epoch": 12.044311100853912,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.3028685790527017e-05,
+      "loss": 1.0707,
+      "step": 13047
+    },
+    {
+      "epoch": 12.045234248788368,
+      "grad_norm": 2.69921875,
+      "learning_rate": 1.3022014676450969e-05,
+      "loss": 1.0925,
+      "step": 13048
+    },
+    {
+      "epoch": 12.046157396722824,
+      "grad_norm": 3.16796875,
+      "learning_rate": 1.3015343562374918e-05,
+      "loss": 1.4098,
+      "step": 13049
+    },
+    {
+      "epoch": 12.047080544657282,
+      "grad_norm": 3.373046875,
+      "learning_rate": 1.3008672448298867e-05,
+      "loss": 1.2158,
+      "step": 13050
+    },
+    {
+      "epoch": 12.048003692591738,
+      "grad_norm": 3.23828125,
+      "learning_rate": 1.3002001334222816e-05,
+      "loss": 1.036,
+      "step": 13051
+    },
+    {
+      "epoch": 12.048926840526194,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.2995330220146764e-05,
+      "loss": 1.2307,
+      "step": 13052
+    },
+    {
+      "epoch": 12.04984998846065,
+      "grad_norm": 3.3828125,
+      "learning_rate": 1.2988659106070713e-05,
+      "loss": 1.3589,
+      "step": 13053
+    },
+    {
+      "epoch": 12.050773136395108,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.2981987991994662e-05,
+      "loss": 1.2791,
+      "step": 13054
+    },
+    {
+      "epoch": 12.051696284329564,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.2975316877918614e-05,
+      "loss": 1.1394,
+      "step": 13055
+    },
+    {
+      "epoch": 12.05261943226402,
+      "grad_norm": 2.744140625,
+      "learning_rate": 1.2968645763842563e-05,
+      "loss": 1.0541,
+      "step": 13056
+    },
+    {
+      "epoch": 12.053542580198476,
+      "grad_norm": 2.966796875,
+      "learning_rate": 1.2961974649766512e-05,
+      "loss": 1.1812,
+      "step": 13057
+    },
+    {
+      "epoch": 12.054465728132934,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.295530353569046e-05,
+      "loss": 1.3665,
+      "step": 13058
+    },
+    {
+      "epoch": 12.05538887606739,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.294863242161441e-05,
+      "loss": 1.4275,
+      "step": 13059
+    },
+    {
+      "epoch": 12.056312024001846,
+      "grad_norm": 3.990234375,
+      "learning_rate": 1.2941961307538358e-05,
+      "loss": 1.1807,
+      "step": 13060
+    },
+    {
+      "epoch": 12.057235171936302,
+      "grad_norm": 3.279296875,
+      "learning_rate": 1.293529019346231e-05,
+      "loss": 1.396,
+      "step": 13061
+    },
+    {
+      "epoch": 12.05815831987076,
+      "grad_norm": 3.3359375,
+      "learning_rate": 1.292861907938626e-05,
+      "loss": 1.2444,
+      "step": 13062
+    },
+    {
+      "epoch": 12.059081467805216,
+      "grad_norm": 3.4453125,
+      "learning_rate": 1.2921947965310208e-05,
+      "loss": 1.2493,
+      "step": 13063
+    },
+    {
+      "epoch": 12.060004615739672,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.2915276851234157e-05,
+      "loss": 1.4187,
+      "step": 13064
+    },
+    {
+      "epoch": 12.06092776367413,
+      "grad_norm": 2.875,
+      "learning_rate": 1.2908605737158106e-05,
+      "loss": 1.3379,
+      "step": 13065
+    },
+    {
+      "epoch": 12.061850911608586,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.2901934623082055e-05,
+      "loss": 1.4854,
+      "step": 13066
+    },
+    {
+      "epoch": 12.062774059543042,
+      "grad_norm": 3.06640625,
+      "learning_rate": 1.2895263509006005e-05,
+      "loss": 1.3064,
+      "step": 13067
+    },
+    {
+      "epoch": 12.063697207477498,
+      "grad_norm": 3.134765625,
+      "learning_rate": 1.2888592394929954e-05,
+      "loss": 1.3752,
+      "step": 13068
+    },
+    {
+      "epoch": 12.064620355411956,
+      "grad_norm": 3.70703125,
+      "learning_rate": 1.2881921280853903e-05,
+      "loss": 1.2303,
+      "step": 13069
+    },
+    {
+      "epoch": 12.065543503346412,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2875250166777853e-05,
+      "loss": 1.3533,
+      "step": 13070
+    },
+    {
+      "epoch": 12.066466651280868,
+      "grad_norm": 3.79296875,
+      "learning_rate": 1.2868579052701802e-05,
+      "loss": 1.2444,
+      "step": 13071
+    },
+    {
+      "epoch": 12.067389799215324,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.2861907938625751e-05,
+      "loss": 1.4902,
+      "step": 13072
+    },
+    {
+      "epoch": 12.068312947149781,
+      "grad_norm": 3.669921875,
+      "learning_rate": 1.2855236824549701e-05,
+      "loss": 1.1331,
+      "step": 13073
+    },
+    {
+      "epoch": 12.069236095084237,
+      "grad_norm": 2.904296875,
+      "learning_rate": 1.284856571047365e-05,
+      "loss": 1.2859,
+      "step": 13074
+    },
+    {
+      "epoch": 12.070159243018693,
+      "grad_norm": 4.25,
+      "learning_rate": 1.2841894596397599e-05,
+      "loss": 1.1713,
+      "step": 13075
+    },
+    {
+      "epoch": 12.07108239095315,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.2835223482321548e-05,
+      "loss": 1.1658,
+      "step": 13076
+    },
+    {
+      "epoch": 12.072005538887607,
+      "grad_norm": 3.00390625,
+      "learning_rate": 1.2828552368245497e-05,
+      "loss": 1.2927,
+      "step": 13077
+    },
+    {
+      "epoch": 12.072928686822063,
+      "grad_norm": 3.001953125,
+      "learning_rate": 1.2821881254169446e-05,
+      "loss": 1.3785,
+      "step": 13078
+    },
+    {
+      "epoch": 12.07385183475652,
+      "grad_norm": 3.09765625,
+      "learning_rate": 1.2815210140093394e-05,
+      "loss": 1.2183,
+      "step": 13079
+    },
+    {
+      "epoch": 12.074774982690975,
+      "grad_norm": 3.0703125,
+      "learning_rate": 1.2808539026017347e-05,
+      "loss": 1.3423,
+      "step": 13080
+    },
+    {
+      "epoch": 12.075698130625433,
+      "grad_norm": 2.849609375,
+      "learning_rate": 1.2801867911941295e-05,
+      "loss": 1.4648,
+      "step": 13081
+    },
+    {
+      "epoch": 12.07662127855989,
+      "grad_norm": 3.162109375,
+      "learning_rate": 1.2795196797865244e-05,
+      "loss": 1.1692,
+      "step": 13082
+    },
+    {
+      "epoch": 12.077544426494345,
+      "grad_norm": 3.525390625,
+      "learning_rate": 1.2788525683789193e-05,
+      "loss": 1.3667,
+      "step": 13083
+    },
+    {
+      "epoch": 12.078467574428803,
+      "grad_norm": 3.365234375,
+      "learning_rate": 1.2781854569713142e-05,
+      "loss": 1.4202,
+      "step": 13084
+    },
+    {
+      "epoch": 12.079390722363259,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.277518345563709e-05,
+      "loss": 1.0709,
+      "step": 13085
+    },
+    {
+      "epoch": 12.080313870297715,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.2768512341561043e-05,
+      "loss": 0.9767,
+      "step": 13086
+    },
+    {
+      "epoch": 12.081237018232171,
+      "grad_norm": 3.19140625,
+      "learning_rate": 1.2761841227484992e-05,
+      "loss": 1.2087,
+      "step": 13087
+    },
+    {
+      "epoch": 12.082160166166629,
+      "grad_norm": 3.306640625,
+      "learning_rate": 1.275517011340894e-05,
+      "loss": 1.2092,
+      "step": 13088
+    },
+    {
+      "epoch": 12.083083314101085,
+      "grad_norm": 3.7421875,
+      "learning_rate": 1.274849899933289e-05,
+      "loss": 1.2825,
+      "step": 13089
+    },
+    {
+      "epoch": 12.084006462035541,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.2741827885256838e-05,
+      "loss": 1.1434,
+      "step": 13090
+    },
+    {
+      "epoch": 12.084929609969997,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.2735156771180787e-05,
+      "loss": 1.1624,
+      "step": 13091
+    },
+    {
+      "epoch": 12.085852757904455,
+      "grad_norm": 3.76171875,
+      "learning_rate": 1.272848565710474e-05,
+      "loss": 1.2646,
+      "step": 13092
+    },
+    {
+      "epoch": 12.08677590583891,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.2721814543028688e-05,
+      "loss": 1.5449,
+      "step": 13093
+    },
+    {
+      "epoch": 12.087699053773367,
+      "grad_norm": 2.9296875,
+      "learning_rate": 1.2715143428952637e-05,
+      "loss": 1.311,
+      "step": 13094
+    },
+    {
+      "epoch": 12.088622201707823,
+      "grad_norm": 3.560546875,
+      "learning_rate": 1.2708472314876586e-05,
+      "loss": 1.1577,
+      "step": 13095
+    },
+    {
+      "epoch": 12.08954534964228,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.2701801200800534e-05,
+      "loss": 1.0965,
+      "step": 13096
+    },
+    {
+      "epoch": 12.090468497576737,
+      "grad_norm": 3.61328125,
+      "learning_rate": 1.2695130086724483e-05,
+      "loss": 1.126,
+      "step": 13097
+    },
+    {
+      "epoch": 12.091391645511193,
+      "grad_norm": 3.23828125,
+      "learning_rate": 1.2688458972648432e-05,
+      "loss": 1.542,
+      "step": 13098
+    },
+    {
+      "epoch": 12.09231479344565,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.2681787858572383e-05,
+      "loss": 1.3242,
+      "step": 13099
+    },
+    {
+      "epoch": 12.093237941380107,
+      "grad_norm": 3.125,
+      "learning_rate": 1.2675116744496331e-05,
+      "loss": 0.9684,
+      "step": 13100
+    },
+    {
+      "epoch": 12.094161089314563,
+      "grad_norm": 2.75,
+      "learning_rate": 1.266844563042028e-05,
+      "loss": 1.3105,
+      "step": 13101
+    },
+    {
+      "epoch": 12.095084237249019,
+      "grad_norm": 3.12890625,
+      "learning_rate": 1.2661774516344229e-05,
+      "loss": 1.2537,
+      "step": 13102
+    },
+    {
+      "epoch": 12.096007385183476,
+      "grad_norm": 3.076171875,
+      "learning_rate": 1.2655103402268178e-05,
+      "loss": 1.3696,
+      "step": 13103
+    },
+    {
+      "epoch": 12.096930533117932,
+      "grad_norm": 3.0,
+      "learning_rate": 1.2648432288192127e-05,
+      "loss": 1.2346,
+      "step": 13104
+    },
+    {
+      "epoch": 12.097853681052388,
+      "grad_norm": 3.029296875,
+      "learning_rate": 1.2641761174116079e-05,
+      "loss": 1.3296,
+      "step": 13105
+    },
+    {
+      "epoch": 12.098776828986844,
+      "grad_norm": 3.287109375,
+      "learning_rate": 1.2635090060040028e-05,
+      "loss": 1.2877,
+      "step": 13106
+    },
+    {
+      "epoch": 12.099699976921302,
+      "grad_norm": 3.36328125,
+      "learning_rate": 1.2628418945963977e-05,
+      "loss": 1.0968,
+      "step": 13107
+    },
+    {
+      "epoch": 12.100623124855758,
+      "grad_norm": 3.099609375,
+      "learning_rate": 1.2621747831887925e-05,
+      "loss": 1.293,
+      "step": 13108
+    },
+    {
+      "epoch": 12.101546272790214,
+      "grad_norm": 3.52734375,
+      "learning_rate": 1.2615076717811874e-05,
+      "loss": 1.4124,
+      "step": 13109
+    },
+    {
+      "epoch": 12.10246942072467,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2608405603735823e-05,
+      "loss": 1.2151,
+      "step": 13110
+    },
+    {
+      "epoch": 12.103392568659128,
+      "grad_norm": 3.01171875,
+      "learning_rate": 1.2601734489659775e-05,
+      "loss": 1.363,
+      "step": 13111
+    },
+    {
+      "epoch": 12.104315716593584,
+      "grad_norm": 3.23828125,
+      "learning_rate": 1.2595063375583724e-05,
+      "loss": 1.04,
+      "step": 13112
+    },
+    {
+      "epoch": 12.10523886452804,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.2588392261507673e-05,
+      "loss": 1.4004,
+      "step": 13113
+    },
+    {
+      "epoch": 12.106162012462496,
+      "grad_norm": 3.30859375,
+      "learning_rate": 1.2581721147431622e-05,
+      "loss": 1.2531,
+      "step": 13114
+    },
+    {
+      "epoch": 12.107085160396954,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.257505003335557e-05,
+      "loss": 1.1404,
+      "step": 13115
+    },
+    {
+      "epoch": 12.10800830833141,
+      "grad_norm": 3.1796875,
+      "learning_rate": 1.256837891927952e-05,
+      "loss": 1.0865,
+      "step": 13116
+    },
+    {
+      "epoch": 12.108931456265866,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.2561707805203471e-05,
+      "loss": 1.1638,
+      "step": 13117
+    },
+    {
+      "epoch": 12.109854604200324,
+      "grad_norm": 3.16796875,
+      "learning_rate": 1.255503669112742e-05,
+      "loss": 1.3926,
+      "step": 13118
+    },
+    {
+      "epoch": 12.11077775213478,
+      "grad_norm": 3.431640625,
+      "learning_rate": 1.2548365577051369e-05,
+      "loss": 1.3433,
+      "step": 13119
+    },
+    {
+      "epoch": 12.111700900069236,
+      "grad_norm": 3.220703125,
+      "learning_rate": 1.2541694462975318e-05,
+      "loss": 1.3865,
+      "step": 13120
+    },
+    {
+      "epoch": 12.112624048003692,
+      "grad_norm": 2.994140625,
+      "learning_rate": 1.2535023348899267e-05,
+      "loss": 1.0791,
+      "step": 13121
+    },
+    {
+      "epoch": 12.11354719593815,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2528352234823216e-05,
+      "loss": 1.2114,
+      "step": 13122
+    },
+    {
+      "epoch": 12.114470343872606,
+      "grad_norm": 3.013671875,
+      "learning_rate": 1.2521681120747164e-05,
+      "loss": 1.2354,
+      "step": 13123
+    },
+    {
+      "epoch": 12.115393491807062,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.2515010006671115e-05,
+      "loss": 1.2085,
+      "step": 13124
+    },
+    {
+      "epoch": 12.116316639741518,
+      "grad_norm": 3.0078125,
+      "learning_rate": 1.2508338892595064e-05,
+      "loss": 1.3225,
+      "step": 13125
+    },
+    {
+      "epoch": 12.117239787675976,
+      "grad_norm": 3.1171875,
+      "learning_rate": 1.2501667778519013e-05,
+      "loss": 1.1997,
+      "step": 13126
+    },
+    {
+      "epoch": 12.118162935610432,
+      "grad_norm": 2.875,
+      "learning_rate": 1.2494996664442963e-05,
+      "loss": 1.2678,
+      "step": 13127
+    },
+    {
+      "epoch": 12.119086083544888,
+      "grad_norm": 3.0,
+      "learning_rate": 1.2488325550366912e-05,
+      "loss": 1.3303,
+      "step": 13128
+    },
+    {
+      "epoch": 12.120009231479344,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.248165443629086e-05,
+      "loss": 1.1836,
+      "step": 13129
+    },
+    {
+      "epoch": 12.120932379413802,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.247498332221481e-05,
+      "loss": 1.2473,
+      "step": 13130
+    },
+    {
+      "epoch": 12.121855527348258,
+      "grad_norm": 3.509765625,
+      "learning_rate": 1.246831220813876e-05,
+      "loss": 1.1282,
+      "step": 13131
+    },
+    {
+      "epoch": 12.122778675282714,
+      "grad_norm": 3.779296875,
+      "learning_rate": 1.2461641094062709e-05,
+      "loss": 1.5479,
+      "step": 13132
+    },
+    {
+      "epoch": 12.123701823217171,
+      "grad_norm": 3.794921875,
+      "learning_rate": 1.2454969979986658e-05,
+      "loss": 1.2225,
+      "step": 13133
+    },
+    {
+      "epoch": 12.124624971151627,
+      "grad_norm": 3.04296875,
+      "learning_rate": 1.2448298865910606e-05,
+      "loss": 1.4788,
+      "step": 13134
+    },
+    {
+      "epoch": 12.125548119086083,
+      "grad_norm": 3.482421875,
+      "learning_rate": 1.2441627751834557e-05,
+      "loss": 1.2664,
+      "step": 13135
+    },
+    {
+      "epoch": 12.12647126702054,
+      "grad_norm": 3.0703125,
+      "learning_rate": 1.2434956637758506e-05,
+      "loss": 1.2742,
+      "step": 13136
+    },
+    {
+      "epoch": 12.127394414954997,
+      "grad_norm": 2.783203125,
+      "learning_rate": 1.2428285523682455e-05,
+      "loss": 1.2814,
+      "step": 13137
+    },
+    {
+      "epoch": 12.128317562889453,
+      "grad_norm": 3.166015625,
+      "learning_rate": 1.2421614409606405e-05,
+      "loss": 1.4031,
+      "step": 13138
+    },
+    {
+      "epoch": 12.12924071082391,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2414943295530354e-05,
+      "loss": 1.1938,
+      "step": 13139
+    },
+    {
+      "epoch": 12.130163858758365,
+      "grad_norm": 2.951171875,
+      "learning_rate": 1.2408272181454303e-05,
+      "loss": 1.2948,
+      "step": 13140
+    },
+    {
+      "epoch": 12.131087006692823,
+      "grad_norm": 2.955078125,
+      "learning_rate": 1.2401601067378253e-05,
+      "loss": 0.9287,
+      "step": 13141
+    },
+    {
+      "epoch": 12.13201015462728,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.2394929953302202e-05,
+      "loss": 0.9321,
+      "step": 13142
+    },
+    {
+      "epoch": 12.132933302561735,
+      "grad_norm": 3.236328125,
+      "learning_rate": 1.2388258839226151e-05,
+      "loss": 1.2607,
+      "step": 13143
+    },
+    {
+      "epoch": 12.133856450496191,
+      "grad_norm": 3.107421875,
+      "learning_rate": 1.2381587725150101e-05,
+      "loss": 1.2931,
+      "step": 13144
+    },
+    {
+      "epoch": 12.134779598430649,
+      "grad_norm": 3.001953125,
+      "learning_rate": 1.237491661107405e-05,
+      "loss": 1.3479,
+      "step": 13145
+    },
+    {
+      "epoch": 12.135702746365105,
+      "grad_norm": 3.26953125,
+      "learning_rate": 1.2368245496997999e-05,
+      "loss": 1.1617,
+      "step": 13146
+    },
+    {
+      "epoch": 12.136625894299561,
+      "grad_norm": 3.392578125,
+      "learning_rate": 1.236157438292195e-05,
+      "loss": 1.381,
+      "step": 13147
+    },
+    {
+      "epoch": 12.137549042234019,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.2354903268845898e-05,
+      "loss": 1.1234,
+      "step": 13148
+    },
+    {
+      "epoch": 12.138472190168475,
+      "grad_norm": 3.064453125,
+      "learning_rate": 1.2348232154769847e-05,
+      "loss": 1.3015,
+      "step": 13149
+    },
+    {
+      "epoch": 12.139395338102931,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.2341561040693798e-05,
+      "loss": 1.1279,
+      "step": 13150
+    },
+    {
+      "epoch": 12.140318486037387,
+      "grad_norm": 3.48828125,
+      "learning_rate": 1.2334889926617747e-05,
+      "loss": 1.16,
+      "step": 13151
+    },
+    {
+      "epoch": 12.141241633971845,
+      "grad_norm": 3.767578125,
+      "learning_rate": 1.2328218812541695e-05,
+      "loss": 1.4111,
+      "step": 13152
+    },
+    {
+      "epoch": 12.1421647819063,
+      "grad_norm": 2.806640625,
+      "learning_rate": 1.2321547698465644e-05,
+      "loss": 1.4255,
+      "step": 13153
+    },
+    {
+      "epoch": 12.143087929840757,
+      "grad_norm": 3.0,
+      "learning_rate": 1.2314876584389593e-05,
+      "loss": 0.6659,
+      "step": 13154
+    },
+    {
+      "epoch": 12.144011077775213,
+      "grad_norm": 3.427734375,
+      "learning_rate": 1.2308205470313544e-05,
+      "loss": 1.3906,
+      "step": 13155
+    },
+    {
+      "epoch": 12.14493422570967,
+      "grad_norm": 3.400390625,
+      "learning_rate": 1.2301534356237492e-05,
+      "loss": 1.3799,
+      "step": 13156
+    },
+    {
+      "epoch": 12.145857373644127,
+      "grad_norm": 3.740234375,
+      "learning_rate": 1.2294863242161441e-05,
+      "loss": 1.3025,
+      "step": 13157
+    },
+    {
+      "epoch": 12.146780521578583,
+      "grad_norm": 3.04296875,
+      "learning_rate": 1.228819212808539e-05,
+      "loss": 1.3699,
+      "step": 13158
+    },
+    {
+      "epoch": 12.147703669513039,
+      "grad_norm": 3.435546875,
+      "learning_rate": 1.2281521014009339e-05,
+      "loss": 1.1572,
+      "step": 13159
+    },
+    {
+      "epoch": 12.148626817447497,
+      "grad_norm": 3.03515625,
+      "learning_rate": 1.227484989993329e-05,
+      "loss": 1.2468,
+      "step": 13160
+    },
+    {
+      "epoch": 12.149549965381953,
+      "grad_norm": 3.32421875,
+      "learning_rate": 1.2268178785857238e-05,
+      "loss": 1.2605,
+      "step": 13161
+    },
+    {
+      "epoch": 12.150473113316409,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2261507671781187e-05,
+      "loss": 1.163,
+      "step": 13162
+    },
+    {
+      "epoch": 12.151396261250865,
+      "grad_norm": 3.20703125,
+      "learning_rate": 1.2254836557705137e-05,
+      "loss": 1.3026,
+      "step": 13163
+    },
+    {
+      "epoch": 12.152319409185322,
+      "grad_norm": 3.326171875,
+      "learning_rate": 1.2248165443629086e-05,
+      "loss": 1.2261,
+      "step": 13164
+    },
+    {
+      "epoch": 12.153242557119778,
+      "grad_norm": 3.044921875,
+      "learning_rate": 1.2241494329553035e-05,
+      "loss": 1.2135,
+      "step": 13165
+    },
+    {
+      "epoch": 12.154165705054234,
+      "grad_norm": 3.521484375,
+      "learning_rate": 1.2234823215476986e-05,
+      "loss": 1.2385,
+      "step": 13166
+    },
+    {
+      "epoch": 12.155088852988692,
+      "grad_norm": 3.333984375,
+      "learning_rate": 1.2228152101400934e-05,
+      "loss": 1.0625,
+      "step": 13167
+    },
+    {
+      "epoch": 12.156012000923148,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.2221480987324883e-05,
+      "loss": 1.3572,
+      "step": 13168
+    },
+    {
+      "epoch": 12.156935148857604,
+      "grad_norm": 3.486328125,
+      "learning_rate": 1.2214809873248834e-05,
+      "loss": 1.2379,
+      "step": 13169
+    },
+    {
+      "epoch": 12.15785829679206,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2208138759172783e-05,
+      "loss": 1.3333,
+      "step": 13170
+    },
+    {
+      "epoch": 12.158781444726518,
+      "grad_norm": 3.134765625,
+      "learning_rate": 1.2201467645096731e-05,
+      "loss": 1.3938,
+      "step": 13171
+    },
+    {
+      "epoch": 12.159704592660974,
+      "grad_norm": 3.751953125,
+      "learning_rate": 1.2194796531020682e-05,
+      "loss": 1.007,
+      "step": 13172
+    },
+    {
+      "epoch": 12.16062774059543,
+      "grad_norm": 2.91015625,
+      "learning_rate": 1.218812541694463e-05,
+      "loss": 1.3508,
+      "step": 13173
+    },
+    {
+      "epoch": 12.161550888529886,
+      "grad_norm": 3.263671875,
+      "learning_rate": 1.218145430286858e-05,
+      "loss": 1.2964,
+      "step": 13174
+    },
+    {
+      "epoch": 12.162474036464344,
+      "grad_norm": 2.822265625,
+      "learning_rate": 1.217478318879253e-05,
+      "loss": 1.551,
+      "step": 13175
+    },
+    {
+      "epoch": 12.1633971843988,
+      "grad_norm": 3.2265625,
+      "learning_rate": 1.2168112074716479e-05,
+      "loss": 1.3201,
+      "step": 13176
+    },
+    {
+      "epoch": 12.164320332333256,
+      "grad_norm": 3.29296875,
+      "learning_rate": 1.2161440960640428e-05,
+      "loss": 1.2189,
+      "step": 13177
+    },
+    {
+      "epoch": 12.165243480267712,
+      "grad_norm": 3.095703125,
+      "learning_rate": 1.2154769846564378e-05,
+      "loss": 1.2411,
+      "step": 13178
+    },
+    {
+      "epoch": 12.16616662820217,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.2148098732488327e-05,
+      "loss": 1.28,
+      "step": 13179
+    },
+    {
+      "epoch": 12.167089776136626,
+      "grad_norm": 3.17578125,
+      "learning_rate": 1.2141427618412276e-05,
+      "loss": 1.3209,
+      "step": 13180
+    },
+    {
+      "epoch": 12.168012924071082,
+      "grad_norm": 3.4765625,
+      "learning_rate": 1.2134756504336225e-05,
+      "loss": 1.4419,
+      "step": 13181
+    },
+    {
+      "epoch": 12.16893607200554,
+      "grad_norm": 3.0546875,
+      "learning_rate": 1.2128085390260173e-05,
+      "loss": 1.0738,
+      "step": 13182
+    },
+    {
+      "epoch": 12.169859219939996,
+      "grad_norm": 3.4921875,
+      "learning_rate": 1.2121414276184124e-05,
+      "loss": 1.3347,
+      "step": 13183
+    },
+    {
+      "epoch": 12.170782367874452,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.2114743162108073e-05,
+      "loss": 1.3821,
+      "step": 13184
+    },
+    {
+      "epoch": 12.171705515808908,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.2108072048032022e-05,
+      "loss": 1.1466,
+      "step": 13185
+    },
+    {
+      "epoch": 12.172628663743366,
+      "grad_norm": 3.994140625,
+      "learning_rate": 1.210140093395597e-05,
+      "loss": 0.8485,
+      "step": 13186
+    },
+    {
+      "epoch": 12.173551811677822,
+      "grad_norm": 3.158203125,
+      "learning_rate": 1.209472981987992e-05,
+      "loss": 1.3994,
+      "step": 13187
+    },
+    {
+      "epoch": 12.174474959612278,
+      "grad_norm": 3.212890625,
+      "learning_rate": 1.208805870580387e-05,
+      "loss": 1.2786,
+      "step": 13188
+    },
+    {
+      "epoch": 12.175398107546734,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.2081387591727819e-05,
+      "loss": 1.1594,
+      "step": 13189
+    },
+    {
+      "epoch": 12.176321255481191,
+      "grad_norm": 2.552734375,
+      "learning_rate": 1.2074716477651767e-05,
+      "loss": 0.9623,
+      "step": 13190
+    },
+    {
+      "epoch": 12.177244403415648,
+      "grad_norm": 3.412109375,
+      "learning_rate": 1.2068045363575718e-05,
+      "loss": 1.1666,
+      "step": 13191
+    },
+    {
+      "epoch": 12.178167551350104,
+      "grad_norm": 3.408203125,
+      "learning_rate": 1.2061374249499667e-05,
+      "loss": 1.1685,
+      "step": 13192
+    },
+    {
+      "epoch": 12.17909069928456,
+      "grad_norm": 2.4765625,
+      "learning_rate": 1.2054703135423616e-05,
+      "loss": 1.0322,
+      "step": 13193
+    },
+    {
+      "epoch": 12.180013847219017,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2048032021347566e-05,
+      "loss": 1.2488,
+      "step": 13194
+    },
+    {
+      "epoch": 12.180936995153473,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.2041360907271515e-05,
+      "loss": 1.3527,
+      "step": 13195
+    },
+    {
+      "epoch": 12.18186014308793,
+      "grad_norm": 3.763671875,
+      "learning_rate": 1.2034689793195464e-05,
+      "loss": 1.1614,
+      "step": 13196
+    },
+    {
+      "epoch": 12.182783291022385,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.2028018679119414e-05,
+      "loss": 1.1824,
+      "step": 13197
+    },
+    {
+      "epoch": 12.183706438956843,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.2021347565043363e-05,
+      "loss": 1.0473,
+      "step": 13198
+    },
+    {
+      "epoch": 12.1846295868913,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.2014676450967312e-05,
+      "loss": 1.0932,
+      "step": 13199
+    },
+    {
+      "epoch": 12.185552734825755,
+      "grad_norm": 3.77734375,
+      "learning_rate": 1.2008005336891262e-05,
+      "loss": 1.335,
+      "step": 13200
+    },
+    {
+      "epoch": 12.186475882760213,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.2001334222815211e-05,
+      "loss": 1.2175,
+      "step": 13201
+    },
+    {
+      "epoch": 12.18739903069467,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.199466310873916e-05,
+      "loss": 1.2881,
+      "step": 13202
+    },
+    {
+      "epoch": 12.188322178629125,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.1987991994663109e-05,
+      "loss": 1.2687,
+      "step": 13203
+    },
+    {
+      "epoch": 12.189245326563581,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.198132088058706e-05,
+      "loss": 1.3837,
+      "step": 13204
+    },
+    {
+      "epoch": 12.190168474498039,
+      "grad_norm": 3.2890625,
+      "learning_rate": 1.1974649766511008e-05,
+      "loss": 1.2498,
+      "step": 13205
+    },
+    {
+      "epoch": 12.191091622432495,
+      "grad_norm": 3.201171875,
+      "learning_rate": 1.1967978652434957e-05,
+      "loss": 1.3229,
+      "step": 13206
+    },
+    {
+      "epoch": 12.192014770366951,
+      "grad_norm": 2.9453125,
+      "learning_rate": 1.1961307538358908e-05,
+      "loss": 1.1523,
+      "step": 13207
+    },
+    {
+      "epoch": 12.192937918301407,
+      "grad_norm": 3.37109375,
+      "learning_rate": 1.1954636424282856e-05,
+      "loss": 1.394,
+      "step": 13208
+    },
+    {
+      "epoch": 12.193861066235865,
+      "grad_norm": 3.505859375,
+      "learning_rate": 1.1947965310206805e-05,
+      "loss": 1.3569,
+      "step": 13209
+    },
+    {
+      "epoch": 12.19478421417032,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.1941294196130754e-05,
+      "loss": 1.0152,
+      "step": 13210
+    },
+    {
+      "epoch": 12.195707362104777,
+      "grad_norm": 3.361328125,
+      "learning_rate": 1.1934623082054704e-05,
+      "loss": 1.2435,
+      "step": 13211
+    },
+    {
+      "epoch": 12.196630510039233,
+      "grad_norm": 3.146484375,
+      "learning_rate": 1.1927951967978653e-05,
+      "loss": 1.3569,
+      "step": 13212
+    },
+    {
+      "epoch": 12.19755365797369,
+      "grad_norm": 4.375,
+      "learning_rate": 1.1921280853902602e-05,
+      "loss": 1.0859,
+      "step": 13213
+    },
+    {
+      "epoch": 12.198476805908147,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.1914609739826551e-05,
+      "loss": 1.407,
+      "step": 13214
+    },
+    {
+      "epoch": 12.199399953842603,
+      "grad_norm": 3.11328125,
+      "learning_rate": 1.19079386257505e-05,
+      "loss": 1.2351,
+      "step": 13215
+    },
+    {
+      "epoch": 12.20032310177706,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.190126751167445e-05,
+      "loss": 1.1051,
+      "step": 13216
+    },
+    {
+      "epoch": 12.201246249711517,
+      "grad_norm": 3.30859375,
+      "learning_rate": 1.1894596397598399e-05,
+      "loss": 1.3292,
+      "step": 13217
+    },
+    {
+      "epoch": 12.202169397645973,
+      "grad_norm": 3.10546875,
+      "learning_rate": 1.1887925283522348e-05,
+      "loss": 1.2208,
+      "step": 13218
+    },
+    {
+      "epoch": 12.203092545580429,
+      "grad_norm": 3.439453125,
+      "learning_rate": 1.1881254169446298e-05,
+      "loss": 1.0464,
+      "step": 13219
+    },
+    {
+      "epoch": 12.204015693514886,
+      "grad_norm": 3.205078125,
+      "learning_rate": 1.1874583055370247e-05,
+      "loss": 1.0504,
+      "step": 13220
+    },
+    {
+      "epoch": 12.204938841449342,
+      "grad_norm": 3.62890625,
+      "learning_rate": 1.1867911941294196e-05,
+      "loss": 1.0383,
+      "step": 13221
+    },
+    {
+      "epoch": 12.205861989383799,
+      "grad_norm": 3.302734375,
+      "learning_rate": 1.1861240827218147e-05,
+      "loss": 1.1689,
+      "step": 13222
+    },
+    {
+      "epoch": 12.206785137318255,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.1854569713142095e-05,
+      "loss": 1.2733,
+      "step": 13223
+    },
+    {
+      "epoch": 12.207708285252712,
+      "grad_norm": 3.369140625,
+      "learning_rate": 1.1847898599066044e-05,
+      "loss": 1.1875,
+      "step": 13224
+    },
+    {
+      "epoch": 12.208631433187168,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.1841227484989995e-05,
+      "loss": 1.1409,
+      "step": 13225
+    },
+    {
+      "epoch": 12.209554581121624,
+      "grad_norm": 3.341796875,
+      "learning_rate": 1.1834556370913944e-05,
+      "loss": 1.4193,
+      "step": 13226
+    },
+    {
+      "epoch": 12.21047772905608,
+      "grad_norm": 3.357421875,
+      "learning_rate": 1.1827885256837892e-05,
+      "loss": 1.4189,
+      "step": 13227
+    },
+    {
+      "epoch": 12.211400876990538,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.1821214142761841e-05,
+      "loss": 1.4524,
+      "step": 13228
+    },
+    {
+      "epoch": 12.212324024924994,
+      "grad_norm": 2.81640625,
+      "learning_rate": 1.1814543028685792e-05,
+      "loss": 1.4236,
+      "step": 13229
+    },
+    {
+      "epoch": 12.21324717285945,
+      "grad_norm": 3.5078125,
+      "learning_rate": 1.180787191460974e-05,
+      "loss": 1.3475,
+      "step": 13230
+    },
+    {
+      "epoch": 12.214170320793908,
+      "grad_norm": 2.92578125,
+      "learning_rate": 1.180120080053369e-05,
+      "loss": 1.3503,
+      "step": 13231
+    },
+    {
+      "epoch": 12.215093468728364,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.179452968645764e-05,
+      "loss": 1.1537,
+      "step": 13232
+    },
+    {
+      "epoch": 12.21601661666282,
+      "grad_norm": 2.74609375,
+      "learning_rate": 1.1787858572381589e-05,
+      "loss": 1.3772,
+      "step": 13233
+    },
+    {
+      "epoch": 12.216939764597276,
+      "grad_norm": 3.076171875,
+      "learning_rate": 1.1781187458305537e-05,
+      "loss": 1.4236,
+      "step": 13234
+    },
+    {
+      "epoch": 12.217862912531734,
+      "grad_norm": 3.17578125,
+      "learning_rate": 1.1774516344229488e-05,
+      "loss": 1.3362,
+      "step": 13235
+    },
+    {
+      "epoch": 12.21878606046619,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.1767845230153437e-05,
+      "loss": 1.235,
+      "step": 13236
+    },
+    {
+      "epoch": 12.219709208400646,
+      "grad_norm": 3.521484375,
+      "learning_rate": 1.1761174116077386e-05,
+      "loss": 1.3796,
+      "step": 13237
+    },
+    {
+      "epoch": 12.220632356335102,
+      "grad_norm": 3.533203125,
+      "learning_rate": 1.1754503002001334e-05,
+      "loss": 1.4467,
+      "step": 13238
+    },
+    {
+      "epoch": 12.22155550426956,
+      "grad_norm": 3.197265625,
+      "learning_rate": 1.1747831887925283e-05,
+      "loss": 1.3103,
+      "step": 13239
+    },
+    {
+      "epoch": 12.222478652204016,
+      "grad_norm": 3.44140625,
+      "learning_rate": 1.1741160773849234e-05,
+      "loss": 1.1753,
+      "step": 13240
+    },
+    {
+      "epoch": 12.223401800138472,
+      "grad_norm": 2.6015625,
+      "learning_rate": 1.1734489659773183e-05,
+      "loss": 0.8098,
+      "step": 13241
+    },
+    {
+      "epoch": 12.224324948072928,
+      "grad_norm": 2.974609375,
+      "learning_rate": 1.1727818545697131e-05,
+      "loss": 1.5525,
+      "step": 13242
+    },
+    {
+      "epoch": 12.225248096007386,
+      "grad_norm": 3.16796875,
+      "learning_rate": 1.172114743162108e-05,
+      "loss": 1.3354,
+      "step": 13243
+    },
+    {
+      "epoch": 12.226171243941842,
+      "grad_norm": 3.6171875,
+      "learning_rate": 1.171447631754503e-05,
+      "loss": 1.2211,
+      "step": 13244
+    },
+    {
+      "epoch": 12.227094391876298,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.170780520346898e-05,
+      "loss": 1.2915,
+      "step": 13245
+    },
+    {
+      "epoch": 12.228017539810754,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.1701134089392928e-05,
+      "loss": 1.3232,
+      "step": 13246
+    },
+    {
+      "epoch": 12.228940687745212,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1694462975316879e-05,
+      "loss": 1.3176,
+      "step": 13247
+    },
+    {
+      "epoch": 12.229863835679668,
+      "grad_norm": 3.21484375,
+      "learning_rate": 1.1687791861240828e-05,
+      "loss": 1.2932,
+      "step": 13248
+    },
+    {
+      "epoch": 12.230786983614124,
+      "grad_norm": 3.048828125,
+      "learning_rate": 1.1681120747164777e-05,
+      "loss": 1.2478,
+      "step": 13249
+    },
+    {
+      "epoch": 12.231710131548581,
+      "grad_norm": 2.986328125,
+      "learning_rate": 1.1674449633088725e-05,
+      "loss": 1.2971,
+      "step": 13250
+    },
+    {
+      "epoch": 12.232633279483037,
+      "grad_norm": 2.970703125,
+      "learning_rate": 1.1667778519012676e-05,
+      "loss": 1.3228,
+      "step": 13251
+    },
+    {
+      "epoch": 12.233556427417493,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.1661107404936625e-05,
+      "loss": 1.4824,
+      "step": 13252
+    },
+    {
+      "epoch": 12.23447957535195,
+      "grad_norm": 2.82421875,
+      "learning_rate": 1.1654436290860574e-05,
+      "loss": 1.3381,
+      "step": 13253
+    },
+    {
+      "epoch": 12.235402723286407,
+      "grad_norm": 3.345703125,
+      "learning_rate": 1.1647765176784524e-05,
+      "loss": 1.3633,
+      "step": 13254
+    },
+    {
+      "epoch": 12.236325871220863,
+      "grad_norm": 2.98828125,
+      "learning_rate": 1.1641094062708473e-05,
+      "loss": 1.0682,
+      "step": 13255
+    },
+    {
+      "epoch": 12.23724901915532,
+      "grad_norm": 2.748046875,
+      "learning_rate": 1.1634422948632422e-05,
+      "loss": 0.9392,
+      "step": 13256
+    },
+    {
+      "epoch": 12.238172167089775,
+      "grad_norm": 2.60546875,
+      "learning_rate": 1.1627751834556372e-05,
+      "loss": 1.0155,
+      "step": 13257
+    },
+    {
+      "epoch": 12.239095315024233,
+      "grad_norm": 3.96484375,
+      "learning_rate": 1.1621080720480321e-05,
+      "loss": 1.3525,
+      "step": 13258
+    },
+    {
+      "epoch": 12.24001846295869,
+      "grad_norm": 2.98046875,
+      "learning_rate": 1.161440960640427e-05,
+      "loss": 0.985,
+      "step": 13259
+    },
+    {
+      "epoch": 12.240941610893145,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.160773849232822e-05,
+      "loss": 1.4634,
+      "step": 13260
+    },
+    {
+      "epoch": 12.241864758827601,
+      "grad_norm": 3.263671875,
+      "learning_rate": 1.1601067378252169e-05,
+      "loss": 1.3604,
+      "step": 13261
+    },
+    {
+      "epoch": 12.242787906762059,
+      "grad_norm": 3.306640625,
+      "learning_rate": 1.1594396264176118e-05,
+      "loss": 1.2312,
+      "step": 13262
+    },
+    {
+      "epoch": 12.243711054696515,
+      "grad_norm": 3.373046875,
+      "learning_rate": 1.1587725150100068e-05,
+      "loss": 1.4893,
+      "step": 13263
+    },
+    {
+      "epoch": 12.244634202630971,
+      "grad_norm": 3.1484375,
+      "learning_rate": 1.1581054036024017e-05,
+      "loss": 1.427,
+      "step": 13264
+    },
+    {
+      "epoch": 12.245557350565429,
+      "grad_norm": 3.91796875,
+      "learning_rate": 1.1574382921947966e-05,
+      "loss": 1.2603,
+      "step": 13265
+    },
+    {
+      "epoch": 12.246480498499885,
+      "grad_norm": 3.4765625,
+      "learning_rate": 1.1567711807871915e-05,
+      "loss": 1.2183,
+      "step": 13266
+    },
+    {
+      "epoch": 12.247403646434341,
+      "grad_norm": 3.138671875,
+      "learning_rate": 1.1561040693795864e-05,
+      "loss": 1.224,
+      "step": 13267
+    },
+    {
+      "epoch": 12.248326794368797,
+      "grad_norm": 3.0546875,
+      "learning_rate": 1.1554369579719814e-05,
+      "loss": 0.9984,
+      "step": 13268
+    },
+    {
+      "epoch": 12.249249942303255,
+      "grad_norm": 3.341796875,
+      "learning_rate": 1.1547698465643763e-05,
+      "loss": 1.2495,
+      "step": 13269
+    },
+    {
+      "epoch": 12.25017309023771,
+      "grad_norm": 4.328125,
+      "learning_rate": 1.1541027351567712e-05,
+      "loss": 1.4243,
+      "step": 13270
+    },
+    {
+      "epoch": 12.251096238172167,
+      "grad_norm": 2.966796875,
+      "learning_rate": 1.153435623749166e-05,
+      "loss": 1.542,
+      "step": 13271
+    },
+    {
+      "epoch": 12.252019386106623,
+      "grad_norm": 2.962890625,
+      "learning_rate": 1.1527685123415611e-05,
+      "loss": 1.2415,
+      "step": 13272
+    },
+    {
+      "epoch": 12.25294253404108,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.152101400933956e-05,
+      "loss": 1.3059,
+      "step": 13273
+    },
+    {
+      "epoch": 12.253865681975537,
+      "grad_norm": 3.21484375,
+      "learning_rate": 1.1514342895263509e-05,
+      "loss": 1.3376,
+      "step": 13274
+    },
+    {
+      "epoch": 12.254788829909993,
+      "grad_norm": 3.91015625,
+      "learning_rate": 1.1507671781187458e-05,
+      "loss": 0.9275,
+      "step": 13275
+    },
+    {
+      "epoch": 12.255711977844449,
+      "grad_norm": 3.166015625,
+      "learning_rate": 1.1501000667111408e-05,
+      "loss": 1.3306,
+      "step": 13276
+    },
+    {
+      "epoch": 12.256635125778907,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.1494329553035357e-05,
+      "loss": 1.3424,
+      "step": 13277
+    },
+    {
+      "epoch": 12.257558273713363,
+      "grad_norm": 3.509765625,
+      "learning_rate": 1.1487658438959306e-05,
+      "loss": 1.1436,
+      "step": 13278
+    },
+    {
+      "epoch": 12.258481421647819,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.1480987324883256e-05,
+      "loss": 1.3813,
+      "step": 13279
+    },
+    {
+      "epoch": 12.259404569582276,
+      "grad_norm": 3.0078125,
+      "learning_rate": 1.1474316210807205e-05,
+      "loss": 1.3594,
+      "step": 13280
+    },
+    {
+      "epoch": 12.260327717516732,
+      "grad_norm": 3.35546875,
+      "learning_rate": 1.1467645096731154e-05,
+      "loss": 1.0104,
+      "step": 13281
+    },
+    {
+      "epoch": 12.261250865451188,
+      "grad_norm": 3.35546875,
+      "learning_rate": 1.1460973982655104e-05,
+      "loss": 1.124,
+      "step": 13282
+    },
+    {
+      "epoch": 12.262174013385644,
+      "grad_norm": 2.98046875,
+      "learning_rate": 1.1454302868579053e-05,
+      "loss": 1.3435,
+      "step": 13283
+    },
+    {
+      "epoch": 12.263097161320102,
+      "grad_norm": 2.94921875,
+      "learning_rate": 1.1447631754503002e-05,
+      "loss": 1.3357,
+      "step": 13284
+    },
+    {
+      "epoch": 12.264020309254558,
+      "grad_norm": 3.474609375,
+      "learning_rate": 1.1440960640426953e-05,
+      "loss": 1.2319,
+      "step": 13285
+    },
+    {
+      "epoch": 12.264943457189014,
+      "grad_norm": 3.029296875,
+      "learning_rate": 1.1434289526350901e-05,
+      "loss": 1.5004,
+      "step": 13286
+    },
+    {
+      "epoch": 12.26586660512347,
+      "grad_norm": 3.044921875,
+      "learning_rate": 1.142761841227485e-05,
+      "loss": 1.4077,
+      "step": 13287
+    },
+    {
+      "epoch": 12.266789753057928,
+      "grad_norm": 3.46484375,
+      "learning_rate": 1.14209472981988e-05,
+      "loss": 1.0803,
+      "step": 13288
+    },
+    {
+      "epoch": 12.267712900992384,
+      "grad_norm": 3.267578125,
+      "learning_rate": 1.141427618412275e-05,
+      "loss": 1.1602,
+      "step": 13289
+    },
+    {
+      "epoch": 12.26863604892684,
+      "grad_norm": 3.15234375,
+      "learning_rate": 1.1407605070046698e-05,
+      "loss": 1.4763,
+      "step": 13290
+    },
+    {
+      "epoch": 12.269559196861296,
+      "grad_norm": 2.837890625,
+      "learning_rate": 1.1400933955970649e-05,
+      "loss": 1.2888,
+      "step": 13291
+    },
+    {
+      "epoch": 12.270482344795754,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.1394262841894598e-05,
+      "loss": 1.2883,
+      "step": 13292
+    },
+    {
+      "epoch": 12.27140549273021,
+      "grad_norm": 3.07421875,
+      "learning_rate": 1.1387591727818547e-05,
+      "loss": 1.4116,
+      "step": 13293
+    },
+    {
+      "epoch": 12.272328640664666,
+      "grad_norm": 3.46484375,
+      "learning_rate": 1.1380920613742495e-05,
+      "loss": 1.1967,
+      "step": 13294
+    },
+    {
+      "epoch": 12.273251788599122,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.1374249499666444e-05,
+      "loss": 1.2372,
+      "step": 13295
+    },
+    {
+      "epoch": 12.27417493653358,
+      "grad_norm": 3.125,
+      "learning_rate": 1.1367578385590395e-05,
+      "loss": 1.3506,
+      "step": 13296
+    },
+    {
+      "epoch": 12.275098084468036,
+      "grad_norm": 4.015625,
+      "learning_rate": 1.1360907271514344e-05,
+      "loss": 1.4082,
+      "step": 13297
+    },
+    {
+      "epoch": 12.276021232402492,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.1354236157438292e-05,
+      "loss": 1.1597,
+      "step": 13298
+    },
+    {
+      "epoch": 12.27694438033695,
+      "grad_norm": 3.67578125,
+      "learning_rate": 1.1347565043362241e-05,
+      "loss": 1.0239,
+      "step": 13299
+    },
+    {
+      "epoch": 12.277867528271406,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.134089392928619e-05,
+      "loss": 1.4175,
+      "step": 13300
+    },
+    {
+      "epoch": 12.278790676205862,
+      "grad_norm": 2.943359375,
+      "learning_rate": 1.133422281521014e-05,
+      "loss": 0.8216,
+      "step": 13301
+    },
+    {
+      "epoch": 12.279713824140318,
+      "grad_norm": 3.25,
+      "learning_rate": 1.132755170113409e-05,
+      "loss": 1.092,
+      "step": 13302
+    },
+    {
+      "epoch": 12.280636972074776,
+      "grad_norm": 3.478515625,
+      "learning_rate": 1.1320880587058038e-05,
+      "loss": 1.4014,
+      "step": 13303
+    },
+    {
+      "epoch": 12.281560120009232,
+      "grad_norm": 3.537109375,
+      "learning_rate": 1.1314209472981989e-05,
+      "loss": 1.2035,
+      "step": 13304
+    },
+    {
+      "epoch": 12.282483267943688,
+      "grad_norm": 3.099609375,
+      "learning_rate": 1.1307538358905937e-05,
+      "loss": 1.2815,
+      "step": 13305
+    },
+    {
+      "epoch": 12.283406415878144,
+      "grad_norm": 2.931640625,
+      "learning_rate": 1.1300867244829886e-05,
+      "loss": 1.1682,
+      "step": 13306
+    },
+    {
+      "epoch": 12.284329563812602,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.1294196130753837e-05,
+      "loss": 1.1633,
+      "step": 13307
+    },
+    {
+      "epoch": 12.285252711747058,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.1287525016677786e-05,
+      "loss": 1.0776,
+      "step": 13308
+    },
+    {
+      "epoch": 12.286175859681514,
+      "grad_norm": 2.8984375,
+      "learning_rate": 1.1280853902601734e-05,
+      "loss": 1.5337,
+      "step": 13309
+    },
+    {
+      "epoch": 12.28709900761597,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.1274182788525685e-05,
+      "loss": 1.5144,
+      "step": 13310
+    },
+    {
+      "epoch": 12.288022155550427,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.1267511674449634e-05,
+      "loss": 1.2004,
+      "step": 13311
+    },
+    {
+      "epoch": 12.288945303484883,
+      "grad_norm": 3.4609375,
+      "learning_rate": 1.1260840560373583e-05,
+      "loss": 1.5217,
+      "step": 13312
+    },
+    {
+      "epoch": 12.28986845141934,
+      "grad_norm": 3.306640625,
+      "learning_rate": 1.1254169446297533e-05,
+      "loss": 1.3162,
+      "step": 13313
+    },
+    {
+      "epoch": 12.290791599353797,
+      "grad_norm": 3.005859375,
+      "learning_rate": 1.1247498332221482e-05,
+      "loss": 1.3162,
+      "step": 13314
+    },
+    {
+      "epoch": 12.291714747288253,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.124082721814543e-05,
+      "loss": 1.3617,
+      "step": 13315
+    },
+    {
+      "epoch": 12.29263789522271,
+      "grad_norm": 3.505859375,
+      "learning_rate": 1.1234156104069381e-05,
+      "loss": 1.1934,
+      "step": 13316
+    },
+    {
+      "epoch": 12.293561043157165,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.122748498999333e-05,
+      "loss": 1.2002,
+      "step": 13317
+    },
+    {
+      "epoch": 12.294484191091623,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.1220813875917279e-05,
+      "loss": 1.3159,
+      "step": 13318
+    },
+    {
+      "epoch": 12.29540733902608,
+      "grad_norm": 2.9921875,
+      "learning_rate": 1.121414276184123e-05,
+      "loss": 1.3838,
+      "step": 13319
+    },
+    {
+      "epoch": 12.296330486960535,
+      "grad_norm": 2.91796875,
+      "learning_rate": 1.1207471647765178e-05,
+      "loss": 1.2424,
+      "step": 13320
+    },
+    {
+      "epoch": 12.297253634894991,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.1200800533689127e-05,
+      "loss": 1.2063,
+      "step": 13321
+    },
+    {
+      "epoch": 12.298176782829449,
+      "grad_norm": 3.125,
+      "learning_rate": 1.1194129419613076e-05,
+      "loss": 0.8152,
+      "step": 13322
+    },
+    {
+      "epoch": 12.299099930763905,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.1187458305537025e-05,
+      "loss": 1.2021,
+      "step": 13323
+    },
+    {
+      "epoch": 12.300023078698361,
+      "grad_norm": 3.029296875,
+      "learning_rate": 1.1180787191460975e-05,
+      "loss": 1.2197,
+      "step": 13324
+    },
+    {
+      "epoch": 12.300946226632817,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.1174116077384924e-05,
+      "loss": 1.0994,
+      "step": 13325
+    },
+    {
+      "epoch": 12.301869374567275,
+      "grad_norm": 3.119140625,
+      "learning_rate": 1.1167444963308873e-05,
+      "loss": 1.1853,
+      "step": 13326
+    },
+    {
+      "epoch": 12.302792522501731,
+      "grad_norm": 3.7734375,
+      "learning_rate": 1.1160773849232822e-05,
+      "loss": 1.3611,
+      "step": 13327
+    },
+    {
+      "epoch": 12.303715670436187,
+      "grad_norm": 2.89453125,
+      "learning_rate": 1.115410273515677e-05,
+      "loss": 1.0327,
+      "step": 13328
+    },
+    {
+      "epoch": 12.304638818370645,
+      "grad_norm": 3.146484375,
+      "learning_rate": 1.1147431621080721e-05,
+      "loss": 1.3699,
+      "step": 13329
+    },
+    {
+      "epoch": 12.3055619663051,
+      "grad_norm": 2.861328125,
+      "learning_rate": 1.114076050700467e-05,
+      "loss": 1.0887,
+      "step": 13330
+    },
+    {
+      "epoch": 12.306485114239557,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.1134089392928619e-05,
+      "loss": 1.2047,
+      "step": 13331
+    },
+    {
+      "epoch": 12.307408262174013,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.1127418278852569e-05,
+      "loss": 1.0088,
+      "step": 13332
+    },
+    {
+      "epoch": 12.30833141010847,
+      "grad_norm": 3.13671875,
+      "learning_rate": 1.1120747164776518e-05,
+      "loss": 1.2418,
+      "step": 13333
+    },
+    {
+      "epoch": 12.309254558042927,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.1114076050700467e-05,
+      "loss": 1.322,
+      "step": 13334
+    },
+    {
+      "epoch": 12.310177705977383,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.1107404936624417e-05,
+      "loss": 1.179,
+      "step": 13335
+    },
+    {
+      "epoch": 12.311100853911839,
+      "grad_norm": 3.19921875,
+      "learning_rate": 1.1100733822548366e-05,
+      "loss": 1.3557,
+      "step": 13336
+    },
+    {
+      "epoch": 12.312024001846297,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.1094062708472315e-05,
+      "loss": 1.1396,
+      "step": 13337
+    },
+    {
+      "epoch": 12.312947149780753,
+      "grad_norm": 3.373046875,
+      "learning_rate": 1.1087391594396265e-05,
+      "loss": 1.0579,
+      "step": 13338
+    },
+    {
+      "epoch": 12.313870297715209,
+      "grad_norm": 3.21484375,
+      "learning_rate": 1.1080720480320214e-05,
+      "loss": 1.5,
+      "step": 13339
+    },
+    {
+      "epoch": 12.314793445649665,
+      "grad_norm": 3.357421875,
+      "learning_rate": 1.1074049366244163e-05,
+      "loss": 1.1453,
+      "step": 13340
+    },
+    {
+      "epoch": 12.315716593584122,
+      "grad_norm": 2.94921875,
+      "learning_rate": 1.1067378252168114e-05,
+      "loss": 1.0812,
+      "step": 13341
+    },
+    {
+      "epoch": 12.316639741518578,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.1060707138092062e-05,
+      "loss": 1.25,
+      "step": 13342
+    },
+    {
+      "epoch": 12.317562889453034,
+      "grad_norm": 3.29296875,
+      "learning_rate": 1.1054036024016011e-05,
+      "loss": 1.2493,
+      "step": 13343
+    },
+    {
+      "epoch": 12.31848603738749,
+      "grad_norm": 3.556640625,
+      "learning_rate": 1.104736490993996e-05,
+      "loss": 1.4658,
+      "step": 13344
+    },
+    {
+      "epoch": 12.319409185321948,
+      "grad_norm": 2.87890625,
+      "learning_rate": 1.104069379586391e-05,
+      "loss": 1.3308,
+      "step": 13345
+    },
+    {
+      "epoch": 12.320332333256404,
+      "grad_norm": 3.595703125,
+      "learning_rate": 1.103402268178786e-05,
+      "loss": 1.3665,
+      "step": 13346
+    },
+    {
+      "epoch": 12.32125548119086,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.1027351567711808e-05,
+      "loss": 1.212,
+      "step": 13347
+    },
+    {
+      "epoch": 12.322178629125318,
+      "grad_norm": 3.13671875,
+      "learning_rate": 1.1020680453635759e-05,
+      "loss": 1.269,
+      "step": 13348
+    },
+    {
+      "epoch": 12.323101777059774,
+      "grad_norm": 3.435546875,
+      "learning_rate": 1.1014009339559708e-05,
+      "loss": 1.2095,
+      "step": 13349
+    },
+    {
+      "epoch": 12.32402492499423,
+      "grad_norm": 3.037109375,
+      "learning_rate": 1.1007338225483656e-05,
+      "loss": 1.3652,
+      "step": 13350
+    },
+    {
+      "epoch": 12.324948072928686,
+      "grad_norm": 3.060546875,
+      "learning_rate": 1.1000667111407605e-05,
+      "loss": 1.0475,
+      "step": 13351
+    },
+    {
+      "epoch": 12.325871220863144,
+      "grad_norm": 3.357421875,
+      "learning_rate": 1.0993995997331554e-05,
+      "loss": 1.1866,
+      "step": 13352
+    },
+    {
+      "epoch": 12.3267943687976,
+      "grad_norm": 3.322265625,
+      "learning_rate": 1.0987324883255505e-05,
+      "loss": 1.0859,
+      "step": 13353
+    },
+    {
+      "epoch": 12.327717516732056,
+      "grad_norm": 3.572265625,
+      "learning_rate": 1.0980653769179453e-05,
+      "loss": 1.0355,
+      "step": 13354
+    },
+    {
+      "epoch": 12.328640664666512,
+      "grad_norm": 2.7734375,
+      "learning_rate": 1.0973982655103402e-05,
+      "loss": 1.3127,
+      "step": 13355
+    },
+    {
+      "epoch": 12.32956381260097,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.0967311541027351e-05,
+      "loss": 1.4194,
+      "step": 13356
+    },
+    {
+      "epoch": 12.330486960535426,
+      "grad_norm": 2.95703125,
+      "learning_rate": 1.0960640426951301e-05,
+      "loss": 1.3914,
+      "step": 13357
+    },
+    {
+      "epoch": 12.331410108469882,
+      "grad_norm": 2.9609375,
+      "learning_rate": 1.095396931287525e-05,
+      "loss": 1.1897,
+      "step": 13358
+    },
+    {
+      "epoch": 12.332333256404338,
+      "grad_norm": 3.59765625,
+      "learning_rate": 1.0947298198799199e-05,
+      "loss": 1.4814,
+      "step": 13359
+    },
+    {
+      "epoch": 12.333256404338796,
+      "grad_norm": 2.8515625,
+      "learning_rate": 1.094062708472315e-05,
+      "loss": 1.1345,
+      "step": 13360
+    },
+    {
+      "epoch": 12.334179552273252,
+      "grad_norm": 3.62890625,
+      "learning_rate": 1.0933955970647098e-05,
+      "loss": 1.2583,
+      "step": 13361
+    },
+    {
+      "epoch": 12.335102700207708,
+      "grad_norm": 3.4296875,
+      "learning_rate": 1.0927284856571047e-05,
+      "loss": 1.2991,
+      "step": 13362
+    },
+    {
+      "epoch": 12.336025848142166,
+      "grad_norm": 3.58203125,
+      "learning_rate": 1.0920613742494998e-05,
+      "loss": 1.2513,
+      "step": 13363
+    },
+    {
+      "epoch": 12.336948996076622,
+      "grad_norm": 2.806640625,
+      "learning_rate": 1.0913942628418947e-05,
+      "loss": 1.2576,
+      "step": 13364
+    },
+    {
+      "epoch": 12.337872144011078,
+      "grad_norm": 3.095703125,
+      "learning_rate": 1.0907271514342895e-05,
+      "loss": 1.3308,
+      "step": 13365
+    },
+    {
+      "epoch": 12.338795291945534,
+      "grad_norm": 3.033203125,
+      "learning_rate": 1.0900600400266846e-05,
+      "loss": 1.1759,
+      "step": 13366
+    },
+    {
+      "epoch": 12.339718439879992,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.0893929286190795e-05,
+      "loss": 1.2471,
+      "step": 13367
+    },
+    {
+      "epoch": 12.340641587814448,
+      "grad_norm": 3.2265625,
+      "learning_rate": 1.0887258172114744e-05,
+      "loss": 1.1443,
+      "step": 13368
+    },
+    {
+      "epoch": 12.341564735748904,
+      "grad_norm": 3.0859375,
+      "learning_rate": 1.0880587058038692e-05,
+      "loss": 1.3867,
+      "step": 13369
+    },
+    {
+      "epoch": 12.34248788368336,
+      "grad_norm": 3.205078125,
+      "learning_rate": 1.0873915943962643e-05,
+      "loss": 1.0657,
+      "step": 13370
+    },
+    {
+      "epoch": 12.343411031617817,
+      "grad_norm": 3.009765625,
+      "learning_rate": 1.0867244829886592e-05,
+      "loss": 1.2791,
+      "step": 13371
+    },
+    {
+      "epoch": 12.344334179552273,
+      "grad_norm": 2.9140625,
+      "learning_rate": 1.086057371581054e-05,
+      "loss": 1.3274,
+      "step": 13372
+    },
+    {
+      "epoch": 12.34525732748673,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.0853902601734491e-05,
+      "loss": 1.205,
+      "step": 13373
+    },
+    {
+      "epoch": 12.346180475421185,
+      "grad_norm": 2.958984375,
+      "learning_rate": 1.084723148765844e-05,
+      "loss": 1.1967,
+      "step": 13374
+    },
+    {
+      "epoch": 12.347103623355643,
+      "grad_norm": 3.1640625,
+      "learning_rate": 1.0840560373582389e-05,
+      "loss": 1.2279,
+      "step": 13375
+    },
+    {
+      "epoch": 12.3480267712901,
+      "grad_norm": 3.54296875,
+      "learning_rate": 1.083388925950634e-05,
+      "loss": 1.3818,
+      "step": 13376
+    },
+    {
+      "epoch": 12.348949919224555,
+      "grad_norm": 3.001953125,
+      "learning_rate": 1.0827218145430288e-05,
+      "loss": 1.3052,
+      "step": 13377
+    },
+    {
+      "epoch": 12.349873067159013,
+      "grad_norm": 3.158203125,
+      "learning_rate": 1.0820547031354237e-05,
+      "loss": 1.4189,
+      "step": 13378
+    },
+    {
+      "epoch": 12.35079621509347,
+      "grad_norm": 3.349609375,
+      "learning_rate": 1.0813875917278186e-05,
+      "loss": 1.3103,
+      "step": 13379
+    },
+    {
+      "epoch": 12.351719363027925,
+      "grad_norm": 3.166015625,
+      "learning_rate": 1.0807204803202134e-05,
+      "loss": 1.3645,
+      "step": 13380
+    },
+    {
+      "epoch": 12.352642510962381,
+      "grad_norm": 2.89453125,
+      "learning_rate": 1.0800533689126085e-05,
+      "loss": 1.4534,
+      "step": 13381
+    },
+    {
+      "epoch": 12.353565658896839,
+      "grad_norm": 3.740234375,
+      "learning_rate": 1.0793862575050034e-05,
+      "loss": 1.2732,
+      "step": 13382
+    },
+    {
+      "epoch": 12.354488806831295,
+      "grad_norm": 2.951171875,
+      "learning_rate": 1.0787191460973983e-05,
+      "loss": 1.3085,
+      "step": 13383
+    },
+    {
+      "epoch": 12.355411954765751,
+      "grad_norm": 3.365234375,
+      "learning_rate": 1.0780520346897931e-05,
+      "loss": 1.2122,
+      "step": 13384
+    },
+    {
+      "epoch": 12.356335102700207,
+      "grad_norm": 3.951171875,
+      "learning_rate": 1.0773849232821882e-05,
+      "loss": 1.1741,
+      "step": 13385
+    },
+    {
+      "epoch": 12.357258250634665,
+      "grad_norm": 3.3515625,
+      "learning_rate": 1.076717811874583e-05,
+      "loss": 1.1783,
+      "step": 13386
+    },
+    {
+      "epoch": 12.358181398569121,
+      "grad_norm": 3.16015625,
+      "learning_rate": 1.076050700466978e-05,
+      "loss": 1.439,
+      "step": 13387
+    },
+    {
+      "epoch": 12.359104546503577,
+      "grad_norm": 3.552734375,
+      "learning_rate": 1.075383589059373e-05,
+      "loss": 1.3861,
+      "step": 13388
+    },
+    {
+      "epoch": 12.360027694438033,
+      "grad_norm": 3.126953125,
+      "learning_rate": 1.0747164776517679e-05,
+      "loss": 1.3472,
+      "step": 13389
+    },
+    {
+      "epoch": 12.36095084237249,
+      "grad_norm": 2.91796875,
+      "learning_rate": 1.0740493662441628e-05,
+      "loss": 1.1772,
+      "step": 13390
+    },
+    {
+      "epoch": 12.361873990306947,
+      "grad_norm": 3.01171875,
+      "learning_rate": 1.0733822548365577e-05,
+      "loss": 1.3477,
+      "step": 13391
+    },
+    {
+      "epoch": 12.362797138241403,
+      "grad_norm": 2.8828125,
+      "learning_rate": 1.0727151434289527e-05,
+      "loss": 1.5447,
+      "step": 13392
+    },
+    {
+      "epoch": 12.363720286175859,
+      "grad_norm": 3.689453125,
+      "learning_rate": 1.0720480320213476e-05,
+      "loss": 1.3416,
+      "step": 13393
+    },
+    {
+      "epoch": 12.364643434110317,
+      "grad_norm": 3.65234375,
+      "learning_rate": 1.0713809206137425e-05,
+      "loss": 1.5715,
+      "step": 13394
+    },
+    {
+      "epoch": 12.365566582044773,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.0707138092061375e-05,
+      "loss": 1.4269,
+      "step": 13395
+    },
+    {
+      "epoch": 12.366489729979229,
+      "grad_norm": 3.41796875,
+      "learning_rate": 1.0700466977985324e-05,
+      "loss": 1.3416,
+      "step": 13396
+    },
+    {
+      "epoch": 12.367412877913686,
+      "grad_norm": 3.37109375,
+      "learning_rate": 1.0693795863909273e-05,
+      "loss": 1.2004,
+      "step": 13397
+    },
+    {
+      "epoch": 12.368336025848143,
+      "grad_norm": 3.279296875,
+      "learning_rate": 1.0687124749833223e-05,
+      "loss": 0.9618,
+      "step": 13398
+    },
+    {
+      "epoch": 12.369259173782599,
+      "grad_norm": 3.615234375,
+      "learning_rate": 1.0680453635757172e-05,
+      "loss": 1.3931,
+      "step": 13399
+    },
+    {
+      "epoch": 12.370182321717055,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.0673782521681121e-05,
+      "loss": 1.2372,
+      "step": 13400
+    },
+    {
+      "epoch": 12.371105469651512,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.0667111407605072e-05,
+      "loss": 1.1804,
+      "step": 13401
+    },
+    {
+      "epoch": 12.372028617585968,
+      "grad_norm": 3.2421875,
+      "learning_rate": 1.066044029352902e-05,
+      "loss": 1.277,
+      "step": 13402
+    },
+    {
+      "epoch": 12.372951765520424,
+      "grad_norm": 2.94921875,
+      "learning_rate": 1.065376917945297e-05,
+      "loss": 1.0851,
+      "step": 13403
+    },
+    {
+      "epoch": 12.37387491345488,
+      "grad_norm": 2.9765625,
+      "learning_rate": 1.064709806537692e-05,
+      "loss": 1.1562,
+      "step": 13404
+    },
+    {
+      "epoch": 12.374798061389338,
+      "grad_norm": 2.91015625,
+      "learning_rate": 1.0640426951300868e-05,
+      "loss": 1.2559,
+      "step": 13405
+    },
+    {
+      "epoch": 12.375721209323794,
+      "grad_norm": 4.00390625,
+      "learning_rate": 1.0633755837224817e-05,
+      "loss": 1.3079,
+      "step": 13406
+    },
+    {
+      "epoch": 12.37664435725825,
+      "grad_norm": 2.998046875,
+      "learning_rate": 1.0627084723148766e-05,
+      "loss": 1.2786,
+      "step": 13407
+    },
+    {
+      "epoch": 12.377567505192706,
+      "grad_norm": 2.9296875,
+      "learning_rate": 1.0620413609072715e-05,
+      "loss": 1.2712,
+      "step": 13408
+    },
+    {
+      "epoch": 12.378490653127164,
+      "grad_norm": 3.208984375,
+      "learning_rate": 1.0613742494996665e-05,
+      "loss": 1.2434,
+      "step": 13409
+    },
+    {
+      "epoch": 12.37941380106162,
+      "grad_norm": 2.97265625,
+      "learning_rate": 1.0607071380920614e-05,
+      "loss": 1.3097,
+      "step": 13410
+    },
+    {
+      "epoch": 12.380336948996076,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0600400266844563e-05,
+      "loss": 1.1598,
+      "step": 13411
+    },
+    {
+      "epoch": 12.381260096930534,
+      "grad_norm": 3.25,
+      "learning_rate": 1.0593729152768512e-05,
+      "loss": 1.2439,
+      "step": 13412
+    },
+    {
+      "epoch": 12.38218324486499,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0587058038692462e-05,
+      "loss": 1.3585,
+      "step": 13413
+    },
+    {
+      "epoch": 12.383106392799446,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.0580386924616411e-05,
+      "loss": 1.2136,
+      "step": 13414
+    },
+    {
+      "epoch": 12.384029540733902,
+      "grad_norm": 2.9453125,
+      "learning_rate": 1.057371581054036e-05,
+      "loss": 0.9531,
+      "step": 13415
+    },
+    {
+      "epoch": 12.38495268866836,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.0567044696464309e-05,
+      "loss": 1.0104,
+      "step": 13416
+    },
+    {
+      "epoch": 12.385875836602816,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.056037358238826e-05,
+      "loss": 1.1569,
+      "step": 13417
+    },
+    {
+      "epoch": 12.386798984537272,
+      "grad_norm": 3.4765625,
+      "learning_rate": 1.0553702468312208e-05,
+      "loss": 1.3088,
+      "step": 13418
+    },
+    {
+      "epoch": 12.387722132471728,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.0547031354236157e-05,
+      "loss": 1.3376,
+      "step": 13419
+    },
+    {
+      "epoch": 12.388645280406186,
+      "grad_norm": 3.12890625,
+      "learning_rate": 1.0540360240160108e-05,
+      "loss": 1.1426,
+      "step": 13420
+    },
+    {
+      "epoch": 12.389568428340642,
+      "grad_norm": 3.248046875,
+      "learning_rate": 1.0533689126084056e-05,
+      "loss": 1.2441,
+      "step": 13421
+    },
+    {
+      "epoch": 12.390491576275098,
+      "grad_norm": 3.24609375,
+      "learning_rate": 1.0527018012008005e-05,
+      "loss": 1.0931,
+      "step": 13422
+    },
+    {
+      "epoch": 12.391414724209554,
+      "grad_norm": 3.384765625,
+      "learning_rate": 1.0520346897931956e-05,
+      "loss": 1.0398,
+      "step": 13423
+    },
+    {
+      "epoch": 12.392337872144012,
+      "grad_norm": 2.958984375,
+      "learning_rate": 1.0513675783855905e-05,
+      "loss": 1.3225,
+      "step": 13424
+    },
+    {
+      "epoch": 12.393261020078468,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.0507004669779853e-05,
+      "loss": 1.4851,
+      "step": 13425
+    },
+    {
+      "epoch": 12.394184168012924,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.0500333555703804e-05,
+      "loss": 1.3469,
+      "step": 13426
+    },
+    {
+      "epoch": 12.395107315947381,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.0493662441627753e-05,
+      "loss": 1.1599,
+      "step": 13427
+    },
+    {
+      "epoch": 12.396030463881837,
+      "grad_norm": 3.40234375,
+      "learning_rate": 1.0486991327551701e-05,
+      "loss": 1.4578,
+      "step": 13428
+    },
+    {
+      "epoch": 12.396953611816294,
+      "grad_norm": 3.443359375,
+      "learning_rate": 1.0480320213475652e-05,
+      "loss": 1.4133,
+      "step": 13429
+    },
+    {
+      "epoch": 12.39787675975075,
+      "grad_norm": 3.3046875,
+      "learning_rate": 1.04736490993996e-05,
+      "loss": 1.2908,
+      "step": 13430
+    },
+    {
+      "epoch": 12.398799907685207,
+      "grad_norm": 3.896484375,
+      "learning_rate": 1.046697798532355e-05,
+      "loss": 1.1,
+      "step": 13431
+    },
+    {
+      "epoch": 12.399723055619663,
+      "grad_norm": 3.01953125,
+      "learning_rate": 1.04603068712475e-05,
+      "loss": 1.3879,
+      "step": 13432
+    },
+    {
+      "epoch": 12.40064620355412,
+      "grad_norm": 3.1015625,
+      "learning_rate": 1.0453635757171449e-05,
+      "loss": 1.3507,
+      "step": 13433
+    },
+    {
+      "epoch": 12.401569351488575,
+      "grad_norm": 2.95703125,
+      "learning_rate": 1.0446964643095398e-05,
+      "loss": 1.4055,
+      "step": 13434
+    },
+    {
+      "epoch": 12.402492499423033,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.0440293529019347e-05,
+      "loss": 1.0913,
+      "step": 13435
+    },
+    {
+      "epoch": 12.40341564735749,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.0433622414943295e-05,
+      "loss": 0.9265,
+      "step": 13436
+    },
+    {
+      "epoch": 12.404338795291945,
+      "grad_norm": 3.25,
+      "learning_rate": 1.0426951300867246e-05,
+      "loss": 1.0684,
+      "step": 13437
+    },
+    {
+      "epoch": 12.405261943226401,
+      "grad_norm": 2.94140625,
+      "learning_rate": 1.0420280186791195e-05,
+      "loss": 1.0764,
+      "step": 13438
+    },
+    {
+      "epoch": 12.406185091160859,
+      "grad_norm": 2.94921875,
+      "learning_rate": 1.0413609072715144e-05,
+      "loss": 1.4807,
+      "step": 13439
+    },
+    {
+      "epoch": 12.407108239095315,
+      "grad_norm": 3.36328125,
+      "learning_rate": 1.0406937958639092e-05,
+      "loss": 1.3406,
+      "step": 13440
+    },
+    {
+      "epoch": 12.408031387029771,
+      "grad_norm": 3.03515625,
+      "learning_rate": 1.0400266844563041e-05,
+      "loss": 1.1487,
+      "step": 13441
+    },
+    {
+      "epoch": 12.408954534964227,
+      "grad_norm": 3.3203125,
+      "learning_rate": 1.0393595730486992e-05,
+      "loss": 1.3116,
+      "step": 13442
+    },
+    {
+      "epoch": 12.409877682898685,
+      "grad_norm": 3.392578125,
+      "learning_rate": 1.038692461641094e-05,
+      "loss": 1.5142,
+      "step": 13443
+    },
+    {
+      "epoch": 12.410800830833141,
+      "grad_norm": 2.830078125,
+      "learning_rate": 1.038025350233489e-05,
+      "loss": 1.2106,
+      "step": 13444
+    },
+    {
+      "epoch": 12.411723978767597,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.037358238825884e-05,
+      "loss": 1.4038,
+      "step": 13445
+    },
+    {
+      "epoch": 12.412647126702055,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.0366911274182789e-05,
+      "loss": 1.3494,
+      "step": 13446
+    },
+    {
+      "epoch": 12.41357027463651,
+      "grad_norm": 3.12109375,
+      "learning_rate": 1.0360240160106738e-05,
+      "loss": 1.3169,
+      "step": 13447
+    },
+    {
+      "epoch": 12.414493422570967,
+      "grad_norm": 3.087890625,
+      "learning_rate": 1.0353569046030688e-05,
+      "loss": 1.3696,
+      "step": 13448
+    },
+    {
+      "epoch": 12.415416570505423,
+      "grad_norm": 2.904296875,
+      "learning_rate": 1.0346897931954637e-05,
+      "loss": 1.2272,
+      "step": 13449
+    },
+    {
+      "epoch": 12.41633971843988,
+      "grad_norm": 3.310546875,
+      "learning_rate": 1.0340226817878586e-05,
+      "loss": 0.9133,
+      "step": 13450
+    },
+    {
+      "epoch": 12.417262866374337,
+      "grad_norm": 3.22265625,
+      "learning_rate": 1.0333555703802536e-05,
+      "loss": 1.2909,
+      "step": 13451
+    },
+    {
+      "epoch": 12.418186014308793,
+      "grad_norm": 2.962890625,
+      "learning_rate": 1.0326884589726485e-05,
+      "loss": 1.2505,
+      "step": 13452
+    },
+    {
+      "epoch": 12.419109162243249,
+      "grad_norm": 3.18359375,
+      "learning_rate": 1.0320213475650434e-05,
+      "loss": 0.9138,
+      "step": 13453
+    },
+    {
+      "epoch": 12.420032310177707,
+      "grad_norm": 2.99609375,
+      "learning_rate": 1.0313542361574384e-05,
+      "loss": 1.2798,
+      "step": 13454
+    },
+    {
+      "epoch": 12.420955458112163,
+      "grad_norm": 3.275390625,
+      "learning_rate": 1.0306871247498333e-05,
+      "loss": 1.2827,
+      "step": 13455
+    },
+    {
+      "epoch": 12.421878606046619,
+      "grad_norm": 3.486328125,
+      "learning_rate": 1.0300200133422282e-05,
+      "loss": 1.3604,
+      "step": 13456
+    },
+    {
+      "epoch": 12.422801753981075,
+      "grad_norm": 3.326171875,
+      "learning_rate": 1.0293529019346232e-05,
+      "loss": 1.2742,
+      "step": 13457
+    },
+    {
+      "epoch": 12.423724901915532,
+      "grad_norm": 3.11328125,
+      "learning_rate": 1.0286857905270181e-05,
+      "loss": 1.3611,
+      "step": 13458
+    },
+    {
+      "epoch": 12.424648049849988,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.028018679119413e-05,
+      "loss": 1.311,
+      "step": 13459
+    },
+    {
+      "epoch": 12.425571197784445,
+      "grad_norm": 3.041015625,
+      "learning_rate": 1.027351567711808e-05,
+      "loss": 1.1644,
+      "step": 13460
+    },
+    {
+      "epoch": 12.4264943457189,
+      "grad_norm": 2.99609375,
+      "learning_rate": 1.026684456304203e-05,
+      "loss": 1.2979,
+      "step": 13461
+    },
+    {
+      "epoch": 12.427417493653358,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.0260173448965978e-05,
+      "loss": 1.2925,
+      "step": 13462
+    },
+    {
+      "epoch": 12.428340641587814,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.0253502334889927e-05,
+      "loss": 1.1927,
+      "step": 13463
+    },
+    {
+      "epoch": 12.42926378952227,
+      "grad_norm": 3.150390625,
+      "learning_rate": 1.0246831220813876e-05,
+      "loss": 1.1884,
+      "step": 13464
+    },
+    {
+      "epoch": 12.430186937456728,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.0240160106737825e-05,
+      "loss": 1.3071,
+      "step": 13465
+    },
+    {
+      "epoch": 12.431110085391184,
+      "grad_norm": 2.736328125,
+      "learning_rate": 1.0233488992661775e-05,
+      "loss": 1.3352,
+      "step": 13466
+    },
+    {
+      "epoch": 12.43203323332564,
+      "grad_norm": 3.412109375,
+      "learning_rate": 1.0226817878585724e-05,
+      "loss": 1.3391,
+      "step": 13467
+    },
+    {
+      "epoch": 12.432956381260096,
+      "grad_norm": 2.90234375,
+      "learning_rate": 1.0220146764509673e-05,
+      "loss": 1.301,
+      "step": 13468
+    },
+    {
+      "epoch": 12.433879529194554,
+      "grad_norm": 3.623046875,
+      "learning_rate": 1.0213475650433622e-05,
+      "loss": 1.1514,
+      "step": 13469
+    },
+    {
+      "epoch": 12.43480267712901,
+      "grad_norm": 3.490234375,
+      "learning_rate": 1.0206804536357572e-05,
+      "loss": 1.2542,
+      "step": 13470
+    },
+    {
+      "epoch": 12.435725825063466,
+      "grad_norm": 3.61328125,
+      "learning_rate": 1.0200133422281521e-05,
+      "loss": 1.1769,
+      "step": 13471
+    },
+    {
+      "epoch": 12.436648972997922,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.019346230820547e-05,
+      "loss": 1.0814,
+      "step": 13472
+    },
+    {
+      "epoch": 12.43757212093238,
+      "grad_norm": 3.146484375,
+      "learning_rate": 1.018679119412942e-05,
+      "loss": 1.1953,
+      "step": 13473
+    },
+    {
+      "epoch": 12.438495268866836,
+      "grad_norm": 3.01171875,
+      "learning_rate": 1.018012008005337e-05,
+      "loss": 1.2688,
+      "step": 13474
+    },
+    {
+      "epoch": 12.439418416801292,
+      "grad_norm": 3.236328125,
+      "learning_rate": 1.0173448965977318e-05,
+      "loss": 1.2668,
+      "step": 13475
+    },
+    {
+      "epoch": 12.44034156473575,
+      "grad_norm": 3.056640625,
+      "learning_rate": 1.0166777851901268e-05,
+      "loss": 1.1272,
+      "step": 13476
+    },
+    {
+      "epoch": 12.441264712670206,
+      "grad_norm": 2.923828125,
+      "learning_rate": 1.0160106737825217e-05,
+      "loss": 1.282,
+      "step": 13477
+    },
+    {
+      "epoch": 12.442187860604662,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.0153435623749166e-05,
+      "loss": 1.1353,
+      "step": 13478
+    },
+    {
+      "epoch": 12.443111008539118,
+      "grad_norm": 3.03515625,
+      "learning_rate": 1.0146764509673117e-05,
+      "loss": 1.3379,
+      "step": 13479
+    },
+    {
+      "epoch": 12.444034156473576,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0140093395597065e-05,
+      "loss": 1.2373,
+      "step": 13480
+    },
+    {
+      "epoch": 12.444957304408032,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.0133422281521014e-05,
+      "loss": 1.2303,
+      "step": 13481
+    },
+    {
+      "epoch": 12.445880452342488,
+      "grad_norm": 3.5703125,
+      "learning_rate": 1.0126751167444965e-05,
+      "loss": 1.238,
+      "step": 13482
+    },
+    {
+      "epoch": 12.446803600276944,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.0120080053368914e-05,
+      "loss": 1.3699,
+      "step": 13483
+    },
+    {
+      "epoch": 12.447726748211402,
+      "grad_norm": 3.87890625,
+      "learning_rate": 1.0113408939292862e-05,
+      "loss": 1.1713,
+      "step": 13484
+    },
+    {
+      "epoch": 12.448649896145858,
+      "grad_norm": 3.33984375,
+      "learning_rate": 1.0106737825216811e-05,
+      "loss": 1.3384,
+      "step": 13485
+    },
+    {
+      "epoch": 12.449573044080314,
+      "grad_norm": 3.306640625,
+      "learning_rate": 1.0100066711140762e-05,
+      "loss": 1.176,
+      "step": 13486
+    },
+    {
+      "epoch": 12.45049619201477,
+      "grad_norm": 3.853515625,
+      "learning_rate": 1.009339559706471e-05,
+      "loss": 1.3282,
+      "step": 13487
+    },
+    {
+      "epoch": 12.451419339949227,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.008672448298866e-05,
+      "loss": 1.1182,
+      "step": 13488
+    },
+    {
+      "epoch": 12.452342487883683,
+      "grad_norm": 3.232421875,
+      "learning_rate": 1.008005336891261e-05,
+      "loss": 1.4316,
+      "step": 13489
+    },
+    {
+      "epoch": 12.45326563581814,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.0073382254836559e-05,
+      "loss": 1.3317,
+      "step": 13490
+    },
+    {
+      "epoch": 12.454188783752596,
+      "grad_norm": 2.8671875,
+      "learning_rate": 1.0066711140760508e-05,
+      "loss": 1.2963,
+      "step": 13491
+    },
+    {
+      "epoch": 12.455111931687053,
+      "grad_norm": 3.16015625,
+      "learning_rate": 1.0060040026684456e-05,
+      "loss": 1.2324,
+      "step": 13492
+    },
+    {
+      "epoch": 12.45603507962151,
+      "grad_norm": 3.009765625,
+      "learning_rate": 1.0053368912608405e-05,
+      "loss": 1.0756,
+      "step": 13493
+    },
+    {
+      "epoch": 12.456958227555965,
+      "grad_norm": 3.15234375,
+      "learning_rate": 1.0046697798532356e-05,
+      "loss": 1.1057,
+      "step": 13494
+    },
+    {
+      "epoch": 12.457881375490423,
+      "grad_norm": 3.15234375,
+      "learning_rate": 1.0040026684456305e-05,
+      "loss": 1.2346,
+      "step": 13495
+    },
+    {
+      "epoch": 12.45880452342488,
+      "grad_norm": 3.255859375,
+      "learning_rate": 1.0033355570380253e-05,
+      "loss": 1.2059,
+      "step": 13496
+    },
+    {
+      "epoch": 12.459727671359335,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.0026684456304202e-05,
+      "loss": 1.1097,
+      "step": 13497
+    },
+    {
+      "epoch": 12.460650819293791,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.0020013342228153e-05,
+      "loss": 1.4031,
+      "step": 13498
+    },
+    {
+      "epoch": 12.461573967228249,
+      "grad_norm": 2.923828125,
+      "learning_rate": 1.0013342228152101e-05,
+      "loss": 1.5894,
+      "step": 13499
+    },
+    {
+      "epoch": 12.462497115162705,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.000667111407605e-05,
+      "loss": 1.1395,
+      "step": 13500
+    },
+    {
+      "epoch": 12.463420263097161,
+      "grad_norm": 2.525390625,
+      "learning_rate": 1e-05,
+      "loss": 0.9675,
+      "step": 13501
+    },
+    {
+      "epoch": 12.464343411031617,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.99332888592395e-06,
+      "loss": 1.3391,
+      "step": 13502
+    },
+    {
+      "epoch": 12.465266558966075,
+      "grad_norm": 3.208984375,
+      "learning_rate": 9.986657771847898e-06,
+      "loss": 1.1547,
+      "step": 13503
+    },
+    {
+      "epoch": 12.466189706900531,
+      "grad_norm": 3.416015625,
+      "learning_rate": 9.979986657771849e-06,
+      "loss": 1.2642,
+      "step": 13504
+    },
+    {
+      "epoch": 12.467112854834987,
+      "grad_norm": 2.98046875,
+      "learning_rate": 9.973315543695798e-06,
+      "loss": 0.9761,
+      "step": 13505
+    },
+    {
+      "epoch": 12.468036002769443,
+      "grad_norm": 3.2734375,
+      "learning_rate": 9.966644429619747e-06,
+      "loss": 1.2438,
+      "step": 13506
+    },
+    {
+      "epoch": 12.4689591507039,
+      "grad_norm": 3.087890625,
+      "learning_rate": 9.959973315543697e-06,
+      "loss": 1.4043,
+      "step": 13507
+    },
+    {
+      "epoch": 12.469882298638357,
+      "grad_norm": 3.177734375,
+      "learning_rate": 9.953302201467646e-06,
+      "loss": 1.2169,
+      "step": 13508
+    },
+    {
+      "epoch": 12.470805446572813,
+      "grad_norm": 3.216796875,
+      "learning_rate": 9.946631087391595e-06,
+      "loss": 1.2996,
+      "step": 13509
+    },
+    {
+      "epoch": 12.471728594507269,
+      "grad_norm": 3.625,
+      "learning_rate": 9.939959973315544e-06,
+      "loss": 1.3113,
+      "step": 13510
+    },
+    {
+      "epoch": 12.472651742441727,
+      "grad_norm": 2.990234375,
+      "learning_rate": 9.933288859239494e-06,
+      "loss": 1.1044,
+      "step": 13511
+    },
+    {
+      "epoch": 12.473574890376183,
+      "grad_norm": 3.119140625,
+      "learning_rate": 9.926617745163443e-06,
+      "loss": 1.1976,
+      "step": 13512
+    },
+    {
+      "epoch": 12.474498038310639,
+      "grad_norm": 3.564453125,
+      "learning_rate": 9.919946631087392e-06,
+      "loss": 1.1929,
+      "step": 13513
+    },
+    {
+      "epoch": 12.475421186245097,
+      "grad_norm": 3.349609375,
+      "learning_rate": 9.913275517011342e-06,
+      "loss": 1.1432,
+      "step": 13514
+    },
+    {
+      "epoch": 12.476344334179553,
+      "grad_norm": 3.330078125,
+      "learning_rate": 9.906604402935291e-06,
+      "loss": 1.3674,
+      "step": 13515
+    },
+    {
+      "epoch": 12.477267482114009,
+      "grad_norm": 2.89453125,
+      "learning_rate": 9.89993328885924e-06,
+      "loss": 1.1888,
+      "step": 13516
+    },
+    {
+      "epoch": 12.478190630048465,
+      "grad_norm": 3.287109375,
+      "learning_rate": 9.89326217478319e-06,
+      "loss": 1.1329,
+      "step": 13517
+    },
+    {
+      "epoch": 12.479113777982922,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.88659106070714e-06,
+      "loss": 0.9698,
+      "step": 13518
+    },
+    {
+      "epoch": 12.480036925917378,
+      "grad_norm": 3.0390625,
+      "learning_rate": 9.879919946631088e-06,
+      "loss": 1.1504,
+      "step": 13519
+    },
+    {
+      "epoch": 12.480960073851834,
+      "grad_norm": 2.92578125,
+      "learning_rate": 9.873248832555037e-06,
+      "loss": 1.1707,
+      "step": 13520
+    },
+    {
+      "epoch": 12.48188322178629,
+      "grad_norm": 3.314453125,
+      "learning_rate": 9.866577718478986e-06,
+      "loss": 1.1177,
+      "step": 13521
+    },
+    {
+      "epoch": 12.482806369720748,
+      "grad_norm": 3.0859375,
+      "learning_rate": 9.859906604402936e-06,
+      "loss": 1.2502,
+      "step": 13522
+    },
+    {
+      "epoch": 12.483729517655204,
+      "grad_norm": 2.66015625,
+      "learning_rate": 9.853235490326885e-06,
+      "loss": 1.0602,
+      "step": 13523
+    },
+    {
+      "epoch": 12.48465266558966,
+      "grad_norm": 3.326171875,
+      "learning_rate": 9.846564376250834e-06,
+      "loss": 1.1719,
+      "step": 13524
+    },
+    {
+      "epoch": 12.485575813524116,
+      "grad_norm": 3.498046875,
+      "learning_rate": 9.839893262174783e-06,
+      "loss": 1.1957,
+      "step": 13525
+    },
+    {
+      "epoch": 12.486498961458574,
+      "grad_norm": 3.462890625,
+      "learning_rate": 9.833222148098733e-06,
+      "loss": 0.9766,
+      "step": 13526
+    },
+    {
+      "epoch": 12.48742210939303,
+      "grad_norm": 3.22265625,
+      "learning_rate": 9.826551034022682e-06,
+      "loss": 1.207,
+      "step": 13527
+    },
+    {
+      "epoch": 12.488345257327486,
+      "grad_norm": 3.142578125,
+      "learning_rate": 9.81987991994663e-06,
+      "loss": 1.0842,
+      "step": 13528
+    },
+    {
+      "epoch": 12.489268405261944,
+      "grad_norm": 3.185546875,
+      "learning_rate": 9.813208805870581e-06,
+      "loss": 1.0936,
+      "step": 13529
+    },
+    {
+      "epoch": 12.4901915531964,
+      "grad_norm": 3.435546875,
+      "learning_rate": 9.80653769179453e-06,
+      "loss": 0.9104,
+      "step": 13530
+    },
+    {
+      "epoch": 12.491114701130856,
+      "grad_norm": 3.58203125,
+      "learning_rate": 9.799866577718479e-06,
+      "loss": 1.1274,
+      "step": 13531
+    },
+    {
+      "epoch": 12.492037849065312,
+      "grad_norm": 3.001953125,
+      "learning_rate": 9.793195463642428e-06,
+      "loss": 1.1377,
+      "step": 13532
+    },
+    {
+      "epoch": 12.49296099699977,
+      "grad_norm": 3.07421875,
+      "learning_rate": 9.786524349566378e-06,
+      "loss": 0.9835,
+      "step": 13533
+    },
+    {
+      "epoch": 12.493884144934226,
+      "grad_norm": 3.60546875,
+      "learning_rate": 9.779853235490327e-06,
+      "loss": 1.2808,
+      "step": 13534
+    },
+    {
+      "epoch": 12.494807292868682,
+      "grad_norm": 2.857421875,
+      "learning_rate": 9.773182121414276e-06,
+      "loss": 1.3743,
+      "step": 13535
+    },
+    {
+      "epoch": 12.495730440803138,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.766511007338226e-06,
+      "loss": 1.4446,
+      "step": 13536
+    },
+    {
+      "epoch": 12.496653588737596,
+      "grad_norm": 3.15234375,
+      "learning_rate": 9.759839893262175e-06,
+      "loss": 1.0726,
+      "step": 13537
+    },
+    {
+      "epoch": 12.497576736672052,
+      "grad_norm": 2.84765625,
+      "learning_rate": 9.753168779186124e-06,
+      "loss": 1.3452,
+      "step": 13538
+    },
+    {
+      "epoch": 12.498499884606508,
+      "grad_norm": 3.244140625,
+      "learning_rate": 9.746497665110075e-06,
+      "loss": 1.2795,
+      "step": 13539
+    },
+    {
+      "epoch": 12.499423032540964,
+      "grad_norm": 3.27734375,
+      "learning_rate": 9.739826551034023e-06,
+      "loss": 1.1661,
+      "step": 13540
+    },
+    {
+      "epoch": 12.500346180475422,
+      "grad_norm": 3.34765625,
+      "learning_rate": 9.733155436957972e-06,
+      "loss": 1.2053,
+      "step": 13541
+    },
+    {
+      "epoch": 12.501269328409878,
+      "grad_norm": 3.7890625,
+      "learning_rate": 9.726484322881923e-06,
+      "loss": 1.2462,
+      "step": 13542
+    },
+    {
+      "epoch": 12.502192476344334,
+      "grad_norm": 3.13671875,
+      "learning_rate": 9.719813208805872e-06,
+      "loss": 1.2432,
+      "step": 13543
+    },
+    {
+      "epoch": 12.503115624278792,
+      "grad_norm": 3.4296875,
+      "learning_rate": 9.71314209472982e-06,
+      "loss": 1.302,
+      "step": 13544
+    },
+    {
+      "epoch": 12.504038772213248,
+      "grad_norm": 3.201171875,
+      "learning_rate": 9.706470980653771e-06,
+      "loss": 1.158,
+      "step": 13545
+    },
+    {
+      "epoch": 12.504961920147704,
+      "grad_norm": 3.294921875,
+      "learning_rate": 9.69979986657772e-06,
+      "loss": 1.3149,
+      "step": 13546
+    },
+    {
+      "epoch": 12.50588506808216,
+      "grad_norm": 3.1015625,
+      "learning_rate": 9.693128752501669e-06,
+      "loss": 1.3037,
+      "step": 13547
+    },
+    {
+      "epoch": 12.506808216016617,
+      "grad_norm": 3.083984375,
+      "learning_rate": 9.686457638425617e-06,
+      "loss": 1.3884,
+      "step": 13548
+    },
+    {
+      "epoch": 12.507731363951073,
+      "grad_norm": 3.2890625,
+      "learning_rate": 9.679786524349566e-06,
+      "loss": 1.177,
+      "step": 13549
+    },
+    {
+      "epoch": 12.50865451188553,
+      "grad_norm": 3.09765625,
+      "learning_rate": 9.673115410273517e-06,
+      "loss": 1.2295,
+      "step": 13550
+    },
+    {
+      "epoch": 12.509577659819985,
+      "grad_norm": 2.935546875,
+      "learning_rate": 9.666444296197465e-06,
+      "loss": 1.2556,
+      "step": 13551
+    },
+    {
+      "epoch": 12.510500807754443,
+      "grad_norm": 2.615234375,
+      "learning_rate": 9.659773182121414e-06,
+      "loss": 1.0688,
+      "step": 13552
+    },
+    {
+      "epoch": 12.5114239556889,
+      "grad_norm": 3.1640625,
+      "learning_rate": 9.653102068045363e-06,
+      "loss": 1.4817,
+      "step": 13553
+    },
+    {
+      "epoch": 12.512347103623355,
+      "grad_norm": 3.4140625,
+      "learning_rate": 9.646430953969314e-06,
+      "loss": 1.5789,
+      "step": 13554
+    },
+    {
+      "epoch": 12.513270251557811,
+      "grad_norm": 3.845703125,
+      "learning_rate": 9.639759839893262e-06,
+      "loss": 1.2822,
+      "step": 13555
+    },
+    {
+      "epoch": 12.51419339949227,
+      "grad_norm": 2.8046875,
+      "learning_rate": 9.633088725817211e-06,
+      "loss": 1.1113,
+      "step": 13556
+    },
+    {
+      "epoch": 12.515116547426725,
+      "grad_norm": 3.0703125,
+      "learning_rate": 9.62641761174116e-06,
+      "loss": 0.883,
+      "step": 13557
+    },
+    {
+      "epoch": 12.516039695361181,
+      "grad_norm": 3.26953125,
+      "learning_rate": 9.61974649766511e-06,
+      "loss": 1.2959,
+      "step": 13558
+    },
+    {
+      "epoch": 12.516962843295637,
+      "grad_norm": 2.919921875,
+      "learning_rate": 9.61307538358906e-06,
+      "loss": 1.4148,
+      "step": 13559
+    },
+    {
+      "epoch": 12.517885991230095,
+      "grad_norm": 3.4921875,
+      "learning_rate": 9.606404269513008e-06,
+      "loss": 1.2528,
+      "step": 13560
+    },
+    {
+      "epoch": 12.518809139164551,
+      "grad_norm": 3.1953125,
+      "learning_rate": 9.599733155436959e-06,
+      "loss": 1.3333,
+      "step": 13561
+    },
+    {
+      "epoch": 12.519732287099007,
+      "grad_norm": 3.0703125,
+      "learning_rate": 9.593062041360908e-06,
+      "loss": 1.3359,
+      "step": 13562
+    },
+    {
+      "epoch": 12.520655435033465,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.586390927284856e-06,
+      "loss": 1.2223,
+      "step": 13563
+    },
+    {
+      "epoch": 12.521578582967921,
+      "grad_norm": 3.02734375,
+      "learning_rate": 9.579719813208807e-06,
+      "loss": 1.3367,
+      "step": 13564
+    },
+    {
+      "epoch": 12.522501730902377,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.573048699132756e-06,
+      "loss": 1.1871,
+      "step": 13565
+    },
+    {
+      "epoch": 12.523424878836833,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.566377585056705e-06,
+      "loss": 1.1917,
+      "step": 13566
+    },
+    {
+      "epoch": 12.52434802677129,
+      "grad_norm": 3.05078125,
+      "learning_rate": 9.559706470980655e-06,
+      "loss": 0.8967,
+      "step": 13567
+    },
+    {
+      "epoch": 12.525271174705747,
+      "grad_norm": 3.205078125,
+      "learning_rate": 9.553035356904604e-06,
+      "loss": 1.0698,
+      "step": 13568
+    },
+    {
+      "epoch": 12.526194322640203,
+      "grad_norm": 3.07421875,
+      "learning_rate": 9.546364242828553e-06,
+      "loss": 1.1085,
+      "step": 13569
+    },
+    {
+      "epoch": 12.527117470574659,
+      "grad_norm": 3.3046875,
+      "learning_rate": 9.539693128752503e-06,
+      "loss": 1.2452,
+      "step": 13570
+    },
+    {
+      "epoch": 12.528040618509117,
+      "grad_norm": 3.2734375,
+      "learning_rate": 9.533022014676452e-06,
+      "loss": 1.2878,
+      "step": 13571
+    },
+    {
+      "epoch": 12.528963766443573,
+      "grad_norm": 2.849609375,
+      "learning_rate": 9.5263509006004e-06,
+      "loss": 1.1973,
+      "step": 13572
+    },
+    {
+      "epoch": 12.529886914378029,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.519679786524351e-06,
+      "loss": 1.1772,
+      "step": 13573
+    },
+    {
+      "epoch": 12.530810062312487,
+      "grad_norm": 3.419921875,
+      "learning_rate": 9.5130086724483e-06,
+      "loss": 1.1974,
+      "step": 13574
+    },
+    {
+      "epoch": 12.531733210246943,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.506337558372249e-06,
+      "loss": 1.1621,
+      "step": 13575
+    },
+    {
+      "epoch": 12.532656358181399,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.499666444296198e-06,
+      "loss": 1.2714,
+      "step": 13576
+    },
+    {
+      "epoch": 12.533579506115855,
+      "grad_norm": 3.15234375,
+      "learning_rate": 9.492995330220147e-06,
+      "loss": 1.116,
+      "step": 13577
+    },
+    {
+      "epoch": 12.534502654050312,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.486324216144095e-06,
+      "loss": 1.2422,
+      "step": 13578
+    },
+    {
+      "epoch": 12.535425801984768,
+      "grad_norm": 3.935546875,
+      "learning_rate": 9.479653102068046e-06,
+      "loss": 1.26,
+      "step": 13579
+    },
+    {
+      "epoch": 12.536348949919224,
+      "grad_norm": 3.111328125,
+      "learning_rate": 9.472981987991995e-06,
+      "loss": 1.1997,
+      "step": 13580
+    },
+    {
+      "epoch": 12.53727209785368,
+      "grad_norm": 3.490234375,
+      "learning_rate": 9.466310873915944e-06,
+      "loss": 1.4238,
+      "step": 13581
+    },
+    {
+      "epoch": 12.538195245788138,
+      "grad_norm": 2.5390625,
+      "learning_rate": 9.459639759839892e-06,
+      "loss": 1.019,
+      "step": 13582
+    },
+    {
+      "epoch": 12.539118393722594,
+      "grad_norm": 3.876953125,
+      "learning_rate": 9.452968645763843e-06,
+      "loss": 1.2859,
+      "step": 13583
+    },
+    {
+      "epoch": 12.54004154165705,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.446297531687792e-06,
+      "loss": 1.0662,
+      "step": 13584
+    },
+    {
+      "epoch": 12.540964689591506,
+      "grad_norm": 3.208984375,
+      "learning_rate": 9.43962641761174e-06,
+      "loss": 1.4031,
+      "step": 13585
+    },
+    {
+      "epoch": 12.541887837525964,
+      "grad_norm": 3.029296875,
+      "learning_rate": 9.432955303535691e-06,
+      "loss": 1.3099,
+      "step": 13586
+    },
+    {
+      "epoch": 12.54281098546042,
+      "grad_norm": 3.408203125,
+      "learning_rate": 9.42628418945964e-06,
+      "loss": 1.2759,
+      "step": 13587
+    },
+    {
+      "epoch": 12.543734133394876,
+      "grad_norm": 2.72265625,
+      "learning_rate": 9.419613075383589e-06,
+      "loss": 0.7974,
+      "step": 13588
+    },
+    {
+      "epoch": 12.544657281329332,
+      "grad_norm": 3.294921875,
+      "learning_rate": 9.41294196130754e-06,
+      "loss": 1.3525,
+      "step": 13589
+    },
+    {
+      "epoch": 12.54558042926379,
+      "grad_norm": 3.650390625,
+      "learning_rate": 9.406270847231488e-06,
+      "loss": 1.0535,
+      "step": 13590
+    },
+    {
+      "epoch": 12.546503577198246,
+      "grad_norm": 3.3359375,
+      "learning_rate": 9.399599733155437e-06,
+      "loss": 1.2358,
+      "step": 13591
+    },
+    {
+      "epoch": 12.547426725132702,
+      "grad_norm": 3.18359375,
+      "learning_rate": 9.392928619079387e-06,
+      "loss": 1.2216,
+      "step": 13592
+    },
+    {
+      "epoch": 12.54834987306716,
+      "grad_norm": 3.326171875,
+      "learning_rate": 9.386257505003336e-06,
+      "loss": 1.0969,
+      "step": 13593
+    },
+    {
+      "epoch": 12.549273021001616,
+      "grad_norm": 3.158203125,
+      "learning_rate": 9.379586390927285e-06,
+      "loss": 1.332,
+      "step": 13594
+    },
+    {
+      "epoch": 12.550196168936072,
+      "grad_norm": 2.7578125,
+      "learning_rate": 9.372915276851236e-06,
+      "loss": 0.9892,
+      "step": 13595
+    },
+    {
+      "epoch": 12.551119316870528,
+      "grad_norm": 3.337890625,
+      "learning_rate": 9.366244162775184e-06,
+      "loss": 1.3091,
+      "step": 13596
+    },
+    {
+      "epoch": 12.552042464804986,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.359573048699133e-06,
+      "loss": 1.2914,
+      "step": 13597
+    },
+    {
+      "epoch": 12.552965612739442,
+      "grad_norm": 2.701171875,
+      "learning_rate": 9.352901934623084e-06,
+      "loss": 1.2429,
+      "step": 13598
+    },
+    {
+      "epoch": 12.553888760673898,
+      "grad_norm": 2.94921875,
+      "learning_rate": 9.346230820547032e-06,
+      "loss": 1.0874,
+      "step": 13599
+    },
+    {
+      "epoch": 12.554811908608354,
+      "grad_norm": 3.197265625,
+      "learning_rate": 9.339559706470981e-06,
+      "loss": 1.0778,
+      "step": 13600
+    },
+    {
+      "epoch": 12.555735056542812,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.332888592394932e-06,
+      "loss": 1.2141,
+      "step": 13601
+    },
+    {
+      "epoch": 12.556658204477268,
+      "grad_norm": 3.017578125,
+      "learning_rate": 9.32621747831888e-06,
+      "loss": 1.2056,
+      "step": 13602
+    },
+    {
+      "epoch": 12.557581352411724,
+      "grad_norm": 3.158203125,
+      "learning_rate": 9.31954636424283e-06,
+      "loss": 1.3762,
+      "step": 13603
+    },
+    {
+      "epoch": 12.55850450034618,
+      "grad_norm": 3.056640625,
+      "learning_rate": 9.312875250166778e-06,
+      "loss": 1.2686,
+      "step": 13604
+    },
+    {
+      "epoch": 12.559427648280638,
+      "grad_norm": 3.4921875,
+      "learning_rate": 9.306204136090727e-06,
+      "loss": 1.1294,
+      "step": 13605
+    },
+    {
+      "epoch": 12.560350796215094,
+      "grad_norm": 3.111328125,
+      "learning_rate": 9.299533022014676e-06,
+      "loss": 1.354,
+      "step": 13606
+    },
+    {
+      "epoch": 12.56127394414955,
+      "grad_norm": 3.11328125,
+      "learning_rate": 9.292861907938626e-06,
+      "loss": 1.108,
+      "step": 13607
+    },
+    {
+      "epoch": 12.562197092084006,
+      "grad_norm": 3.498046875,
+      "learning_rate": 9.286190793862575e-06,
+      "loss": 1.5425,
+      "step": 13608
+    },
+    {
+      "epoch": 12.563120240018463,
+      "grad_norm": 3.18359375,
+      "learning_rate": 9.279519679786524e-06,
+      "loss": 1.2231,
+      "step": 13609
+    },
+    {
+      "epoch": 12.56404338795292,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.272848565710473e-06,
+      "loss": 1.321,
+      "step": 13610
+    },
+    {
+      "epoch": 12.564966535887375,
+      "grad_norm": 3.0234375,
+      "learning_rate": 9.266177451634423e-06,
+      "loss": 1.3296,
+      "step": 13611
+    },
+    {
+      "epoch": 12.565889683821833,
+      "grad_norm": 3.2265625,
+      "learning_rate": 9.259506337558372e-06,
+      "loss": 1.1689,
+      "step": 13612
+    },
+    {
+      "epoch": 12.56681283175629,
+      "grad_norm": 2.9765625,
+      "learning_rate": 9.252835223482321e-06,
+      "loss": 1.1514,
+      "step": 13613
+    },
+    {
+      "epoch": 12.567735979690745,
+      "grad_norm": 3.001953125,
+      "learning_rate": 9.246164109406272e-06,
+      "loss": 1.4695,
+      "step": 13614
+    },
+    {
+      "epoch": 12.568659127625201,
+      "grad_norm": 3.369140625,
+      "learning_rate": 9.23949299533022e-06,
+      "loss": 1.4351,
+      "step": 13615
+    },
+    {
+      "epoch": 12.56958227555966,
+      "grad_norm": 3.10546875,
+      "learning_rate": 9.23282188125417e-06,
+      "loss": 1.2596,
+      "step": 13616
+    },
+    {
+      "epoch": 12.570505423494115,
+      "grad_norm": 3.220703125,
+      "learning_rate": 9.22615076717812e-06,
+      "loss": 1.2461,
+      "step": 13617
+    },
+    {
+      "epoch": 12.571428571428571,
+      "grad_norm": 3.35546875,
+      "learning_rate": 9.219479653102069e-06,
+      "loss": 0.9854,
+      "step": 13618
+    },
+    {
+      "epoch": 12.572351719363027,
+      "grad_norm": 3.056640625,
+      "learning_rate": 9.212808539026017e-06,
+      "loss": 1.3308,
+      "step": 13619
+    },
+    {
+      "epoch": 12.573274867297485,
+      "grad_norm": 2.98828125,
+      "learning_rate": 9.206137424949968e-06,
+      "loss": 1.4004,
+      "step": 13620
+    },
+    {
+      "epoch": 12.574198015231941,
+      "grad_norm": 3.26171875,
+      "learning_rate": 9.199466310873917e-06,
+      "loss": 1.3733,
+      "step": 13621
+    },
+    {
+      "epoch": 12.575121163166397,
+      "grad_norm": 3.685546875,
+      "learning_rate": 9.192795196797865e-06,
+      "loss": 1.2668,
+      "step": 13622
+    },
+    {
+      "epoch": 12.576044311100853,
+      "grad_norm": 3.125,
+      "learning_rate": 9.186124082721816e-06,
+      "loss": 1.2496,
+      "step": 13623
+    },
+    {
+      "epoch": 12.57696745903531,
+      "grad_norm": 3.302734375,
+      "learning_rate": 9.179452968645765e-06,
+      "loss": 1.1665,
+      "step": 13624
+    },
+    {
+      "epoch": 12.577890606969767,
+      "grad_norm": 2.939453125,
+      "learning_rate": 9.172781854569714e-06,
+      "loss": 1.3545,
+      "step": 13625
+    },
+    {
+      "epoch": 12.578813754904223,
+      "grad_norm": 3.634765625,
+      "learning_rate": 9.166110740493662e-06,
+      "loss": 0.9941,
+      "step": 13626
+    },
+    {
+      "epoch": 12.579736902838679,
+      "grad_norm": 3.025390625,
+      "learning_rate": 9.159439626417613e-06,
+      "loss": 1.2778,
+      "step": 13627
+    },
+    {
+      "epoch": 12.580660050773137,
+      "grad_norm": 2.9765625,
+      "learning_rate": 9.152768512341562e-06,
+      "loss": 1.2402,
+      "step": 13628
+    },
+    {
+      "epoch": 12.581583198707593,
+      "grad_norm": 2.943359375,
+      "learning_rate": 9.14609739826551e-06,
+      "loss": 1.3281,
+      "step": 13629
+    },
+    {
+      "epoch": 12.582506346642049,
+      "grad_norm": 3.33203125,
+      "learning_rate": 9.139426284189461e-06,
+      "loss": 1.0279,
+      "step": 13630
+    },
+    {
+      "epoch": 12.583429494576507,
+      "grad_norm": 3.478515625,
+      "learning_rate": 9.13275517011341e-06,
+      "loss": 1.3088,
+      "step": 13631
+    },
+    {
+      "epoch": 12.584352642510963,
+      "grad_norm": 3.576171875,
+      "learning_rate": 9.126084056037359e-06,
+      "loss": 1.4155,
+      "step": 13632
+    },
+    {
+      "epoch": 12.585275790445419,
+      "grad_norm": 3.333984375,
+      "learning_rate": 9.119412941961308e-06,
+      "loss": 1.2349,
+      "step": 13633
+    },
+    {
+      "epoch": 12.586198938379875,
+      "grad_norm": 3.05859375,
+      "learning_rate": 9.112741827885256e-06,
+      "loss": 1.1875,
+      "step": 13634
+    },
+    {
+      "epoch": 12.587122086314332,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.106070713809207e-06,
+      "loss": 1.3015,
+      "step": 13635
+    },
+    {
+      "epoch": 12.588045234248789,
+      "grad_norm": 3.248046875,
+      "learning_rate": 9.099399599733156e-06,
+      "loss": 0.9962,
+      "step": 13636
+    },
+    {
+      "epoch": 12.588968382183245,
+      "grad_norm": 3.32421875,
+      "learning_rate": 9.092728485657105e-06,
+      "loss": 1.0135,
+      "step": 13637
+    },
+    {
+      "epoch": 12.5898915301177,
+      "grad_norm": 3.103515625,
+      "learning_rate": 9.086057371581053e-06,
+      "loss": 1.4998,
+      "step": 13638
+    },
+    {
+      "epoch": 12.590814678052158,
+      "grad_norm": 3.025390625,
+      "learning_rate": 9.079386257505004e-06,
+      "loss": 1.179,
+      "step": 13639
+    },
+    {
+      "epoch": 12.591737825986614,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.072715143428953e-06,
+      "loss": 1.1948,
+      "step": 13640
+    },
+    {
+      "epoch": 12.59266097392107,
+      "grad_norm": 3.076171875,
+      "learning_rate": 9.066044029352902e-06,
+      "loss": 1.3174,
+      "step": 13641
+    },
+    {
+      "epoch": 12.593584121855528,
+      "grad_norm": 3.431640625,
+      "learning_rate": 9.059372915276852e-06,
+      "loss": 1.3223,
+      "step": 13642
+    },
+    {
+      "epoch": 12.594507269789984,
+      "grad_norm": 3.728515625,
+      "learning_rate": 9.0527018012008e-06,
+      "loss": 1.354,
+      "step": 13643
+    },
+    {
+      "epoch": 12.59543041772444,
+      "grad_norm": 3.291015625,
+      "learning_rate": 9.04603068712475e-06,
+      "loss": 1.3792,
+      "step": 13644
+    },
+    {
+      "epoch": 12.596353565658896,
+      "grad_norm": 3.201171875,
+      "learning_rate": 9.0393595730487e-06,
+      "loss": 1.3669,
+      "step": 13645
+    },
+    {
+      "epoch": 12.597276713593354,
+      "grad_norm": 3.26171875,
+      "learning_rate": 9.032688458972649e-06,
+      "loss": 1.2908,
+      "step": 13646
+    },
+    {
+      "epoch": 12.59819986152781,
+      "grad_norm": 2.9140625,
+      "learning_rate": 9.026017344896598e-06,
+      "loss": 1.2488,
+      "step": 13647
+    },
+    {
+      "epoch": 12.599123009462266,
+      "grad_norm": 3.228515625,
+      "learning_rate": 9.019346230820548e-06,
+      "loss": 1.0085,
+      "step": 13648
+    },
+    {
+      "epoch": 12.600046157396722,
+      "grad_norm": 3.072265625,
+      "learning_rate": 9.012675116744497e-06,
+      "loss": 1.255,
+      "step": 13649
+    },
+    {
+      "epoch": 12.60096930533118,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.006004002668446e-06,
+      "loss": 0.9324,
+      "step": 13650
+    },
+    {
+      "epoch": 12.601892453265636,
+      "grad_norm": 3.158203125,
+      "learning_rate": 8.999332888592395e-06,
+      "loss": 1.1467,
+      "step": 13651
+    },
+    {
+      "epoch": 12.602815601200092,
+      "grad_norm": 2.982421875,
+      "learning_rate": 8.992661774516345e-06,
+      "loss": 1.2218,
+      "step": 13652
+    },
+    {
+      "epoch": 12.603738749134548,
+      "grad_norm": 3.26953125,
+      "learning_rate": 8.985990660440294e-06,
+      "loss": 1.3795,
+      "step": 13653
+    },
+    {
+      "epoch": 12.604661897069006,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.979319546364243e-06,
+      "loss": 1.2739,
+      "step": 13654
+    },
+    {
+      "epoch": 12.605585045003462,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.972648432288193e-06,
+      "loss": 1.4729,
+      "step": 13655
+    },
+    {
+      "epoch": 12.606508192937918,
+      "grad_norm": 3.134765625,
+      "learning_rate": 8.965977318212142e-06,
+      "loss": 1.366,
+      "step": 13656
+    },
+    {
+      "epoch": 12.607431340872374,
+      "grad_norm": 3.375,
+      "learning_rate": 8.959306204136091e-06,
+      "loss": 1.0088,
+      "step": 13657
+    },
+    {
+      "epoch": 12.608354488806832,
+      "grad_norm": 3.0,
+      "learning_rate": 8.952635090060042e-06,
+      "loss": 1.293,
+      "step": 13658
+    },
+    {
+      "epoch": 12.609277636741288,
+      "grad_norm": 3.55078125,
+      "learning_rate": 8.94596397598399e-06,
+      "loss": 1.3035,
+      "step": 13659
+    },
+    {
+      "epoch": 12.610200784675744,
+      "grad_norm": 2.65234375,
+      "learning_rate": 8.93929286190794e-06,
+      "loss": 1.2517,
+      "step": 13660
+    },
+    {
+      "epoch": 12.611123932610202,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.932621747831888e-06,
+      "loss": 1.3638,
+      "step": 13661
+    },
+    {
+      "epoch": 12.612047080544658,
+      "grad_norm": 3.298828125,
+      "learning_rate": 8.925950633755837e-06,
+      "loss": 1.1831,
+      "step": 13662
+    },
+    {
+      "epoch": 12.612970228479114,
+      "grad_norm": 3.107421875,
+      "learning_rate": 8.919279519679787e-06,
+      "loss": 1.3198,
+      "step": 13663
+    },
+    {
+      "epoch": 12.61389337641357,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.912608405603736e-06,
+      "loss": 1.1593,
+      "step": 13664
+    },
+    {
+      "epoch": 12.614816524348027,
+      "grad_norm": 3.42578125,
+      "learning_rate": 8.905937291527685e-06,
+      "loss": 0.8595,
+      "step": 13665
+    },
+    {
+      "epoch": 12.615739672282483,
+      "grad_norm": 3.0078125,
+      "learning_rate": 8.899266177451634e-06,
+      "loss": 1.2664,
+      "step": 13666
+    },
+    {
+      "epoch": 12.61666282021694,
+      "grad_norm": 3.177734375,
+      "learning_rate": 8.892595063375584e-06,
+      "loss": 1.2649,
+      "step": 13667
+    },
+    {
+      "epoch": 12.617585968151396,
+      "grad_norm": 3.224609375,
+      "learning_rate": 8.885923949299533e-06,
+      "loss": 1.2047,
+      "step": 13668
+    },
+    {
+      "epoch": 12.618509116085853,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.879252835223482e-06,
+      "loss": 1.4001,
+      "step": 13669
+    },
+    {
+      "epoch": 12.61943226402031,
+      "grad_norm": 3.267578125,
+      "learning_rate": 8.872581721147432e-06,
+      "loss": 1.4373,
+      "step": 13670
+    },
+    {
+      "epoch": 12.620355411954765,
+      "grad_norm": 3.07421875,
+      "learning_rate": 8.865910607071381e-06,
+      "loss": 1.3708,
+      "step": 13671
+    },
+    {
+      "epoch": 12.621278559889221,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.85923949299533e-06,
+      "loss": 1.187,
+      "step": 13672
+    },
+    {
+      "epoch": 12.62220170782368,
+      "grad_norm": 3.216796875,
+      "learning_rate": 8.852568378919279e-06,
+      "loss": 1.3997,
+      "step": 13673
+    },
+    {
+      "epoch": 12.623124855758135,
+      "grad_norm": 3.033203125,
+      "learning_rate": 8.84589726484323e-06,
+      "loss": 1.2612,
+      "step": 13674
+    },
+    {
+      "epoch": 12.624048003692591,
+      "grad_norm": 3.087890625,
+      "learning_rate": 8.839226150767178e-06,
+      "loss": 1.1757,
+      "step": 13675
+    },
+    {
+      "epoch": 12.624971151627047,
+      "grad_norm": 3.482421875,
+      "learning_rate": 8.832555036691127e-06,
+      "loss": 1.2344,
+      "step": 13676
+    },
+    {
+      "epoch": 12.625894299561505,
+      "grad_norm": 3.2265625,
+      "learning_rate": 8.825883922615078e-06,
+      "loss": 1.4326,
+      "step": 13677
+    },
+    {
+      "epoch": 12.626817447495961,
+      "grad_norm": 3.2421875,
+      "learning_rate": 8.819212808539026e-06,
+      "loss": 1.2473,
+      "step": 13678
+    },
+    {
+      "epoch": 12.627740595430417,
+      "grad_norm": 3.396484375,
+      "learning_rate": 8.812541694462975e-06,
+      "loss": 1.2324,
+      "step": 13679
+    },
+    {
+      "epoch": 12.628663743364875,
+      "grad_norm": 3.091796875,
+      "learning_rate": 8.805870580386926e-06,
+      "loss": 1.1606,
+      "step": 13680
+    },
+    {
+      "epoch": 12.629586891299331,
+      "grad_norm": 2.89453125,
+      "learning_rate": 8.799199466310875e-06,
+      "loss": 1.2485,
+      "step": 13681
+    },
+    {
+      "epoch": 12.630510039233787,
+      "grad_norm": 3.041015625,
+      "learning_rate": 8.792528352234823e-06,
+      "loss": 1.2019,
+      "step": 13682
+    },
+    {
+      "epoch": 12.631433187168243,
+      "grad_norm": 3.087890625,
+      "learning_rate": 8.785857238158774e-06,
+      "loss": 1.3528,
+      "step": 13683
+    },
+    {
+      "epoch": 12.6323563351027,
+      "grad_norm": 3.267578125,
+      "learning_rate": 8.779186124082723e-06,
+      "loss": 1.3721,
+      "step": 13684
+    },
+    {
+      "epoch": 12.633279483037157,
+      "grad_norm": 3.1640625,
+      "learning_rate": 8.772515010006672e-06,
+      "loss": 1.3586,
+      "step": 13685
+    },
+    {
+      "epoch": 12.634202630971613,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.765843895930622e-06,
+      "loss": 1.1278,
+      "step": 13686
+    },
+    {
+      "epoch": 12.635125778906069,
+      "grad_norm": 3.369140625,
+      "learning_rate": 8.759172781854571e-06,
+      "loss": 1.0679,
+      "step": 13687
+    },
+    {
+      "epoch": 12.636048926840527,
+      "grad_norm": 3.01171875,
+      "learning_rate": 8.75250166777852e-06,
+      "loss": 1.0719,
+      "step": 13688
+    },
+    {
+      "epoch": 12.636972074774983,
+      "grad_norm": 3.361328125,
+      "learning_rate": 8.745830553702469e-06,
+      "loss": 1.4478,
+      "step": 13689
+    },
+    {
+      "epoch": 12.637895222709439,
+      "grad_norm": 3.06640625,
+      "learning_rate": 8.739159439626417e-06,
+      "loss": 1.3181,
+      "step": 13690
+    },
+    {
+      "epoch": 12.638818370643897,
+      "grad_norm": 2.833984375,
+      "learning_rate": 8.732488325550366e-06,
+      "loss": 1.3125,
+      "step": 13691
+    },
+    {
+      "epoch": 12.639741518578353,
+      "grad_norm": 3.154296875,
+      "learning_rate": 8.725817211474317e-06,
+      "loss": 1.2903,
+      "step": 13692
+    },
+    {
+      "epoch": 12.640664666512809,
+      "grad_norm": 2.931640625,
+      "learning_rate": 8.719146097398265e-06,
+      "loss": 1.3755,
+      "step": 13693
+    },
+    {
+      "epoch": 12.641587814447265,
+      "grad_norm": 3.193359375,
+      "learning_rate": 8.712474983322214e-06,
+      "loss": 0.94,
+      "step": 13694
+    },
+    {
+      "epoch": 12.642510962381722,
+      "grad_norm": 3.10546875,
+      "learning_rate": 8.705803869246165e-06,
+      "loss": 1.3118,
+      "step": 13695
+    },
+    {
+      "epoch": 12.643434110316178,
+      "grad_norm": 2.986328125,
+      "learning_rate": 8.699132755170114e-06,
+      "loss": 1.3127,
+      "step": 13696
+    },
+    {
+      "epoch": 12.644357258250635,
+      "grad_norm": 2.87890625,
+      "learning_rate": 8.692461641094062e-06,
+      "loss": 1.3711,
+      "step": 13697
+    },
+    {
+      "epoch": 12.64528040618509,
+      "grad_norm": 2.884765625,
+      "learning_rate": 8.685790527018011e-06,
+      "loss": 1.3289,
+      "step": 13698
+    },
+    {
+      "epoch": 12.646203554119548,
+      "grad_norm": 3.0078125,
+      "learning_rate": 8.679119412941962e-06,
+      "loss": 1.1637,
+      "step": 13699
+    },
+    {
+      "epoch": 12.647126702054004,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.67244829886591e-06,
+      "loss": 1.3545,
+      "step": 13700
+    },
+    {
+      "epoch": 12.64804984998846,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.66577718478986e-06,
+      "loss": 1.3188,
+      "step": 13701
+    },
+    {
+      "epoch": 12.648972997922916,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.65910607071381e-06,
+      "loss": 1.2222,
+      "step": 13702
+    },
+    {
+      "epoch": 12.649896145857374,
+      "grad_norm": 3.01171875,
+      "learning_rate": 8.652434956637759e-06,
+      "loss": 1.261,
+      "step": 13703
+    },
+    {
+      "epoch": 12.65081929379183,
+      "grad_norm": 3.462890625,
+      "learning_rate": 8.645763842561708e-06,
+      "loss": 1.1071,
+      "step": 13704
+    },
+    {
+      "epoch": 12.651742441726286,
+      "grad_norm": 3.005859375,
+      "learning_rate": 8.639092728485658e-06,
+      "loss": 1.1814,
+      "step": 13705
+    },
+    {
+      "epoch": 12.652665589660742,
+      "grad_norm": 3.212890625,
+      "learning_rate": 8.632421614409607e-06,
+      "loss": 1.1754,
+      "step": 13706
+    },
+    {
+      "epoch": 12.6535887375952,
+      "grad_norm": 3.486328125,
+      "learning_rate": 8.625750500333556e-06,
+      "loss": 1.1134,
+      "step": 13707
+    },
+    {
+      "epoch": 12.654511885529656,
+      "grad_norm": 3.3828125,
+      "learning_rate": 8.619079386257506e-06,
+      "loss": 1.4143,
+      "step": 13708
+    },
+    {
+      "epoch": 12.655435033464112,
+      "grad_norm": 2.986328125,
+      "learning_rate": 8.612408272181455e-06,
+      "loss": 0.8829,
+      "step": 13709
+    },
+    {
+      "epoch": 12.65635818139857,
+      "grad_norm": 3.4453125,
+      "learning_rate": 8.605737158105404e-06,
+      "loss": 1.2738,
+      "step": 13710
+    },
+    {
+      "epoch": 12.657281329333026,
+      "grad_norm": 2.6640625,
+      "learning_rate": 8.599066044029354e-06,
+      "loss": 1.1138,
+      "step": 13711
+    },
+    {
+      "epoch": 12.658204477267482,
+      "grad_norm": 3.07421875,
+      "learning_rate": 8.592394929953303e-06,
+      "loss": 1.3615,
+      "step": 13712
+    },
+    {
+      "epoch": 12.659127625201938,
+      "grad_norm": 2.8828125,
+      "learning_rate": 8.585723815877252e-06,
+      "loss": 1.1595,
+      "step": 13713
+    },
+    {
+      "epoch": 12.660050773136396,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.579052701801203e-06,
+      "loss": 1.0182,
+      "step": 13714
+    },
+    {
+      "epoch": 12.660973921070852,
+      "grad_norm": 3.572265625,
+      "learning_rate": 8.572381587725151e-06,
+      "loss": 1.3591,
+      "step": 13715
+    },
+    {
+      "epoch": 12.661897069005308,
+      "grad_norm": 3.677734375,
+      "learning_rate": 8.5657104736491e-06,
+      "loss": 1.2335,
+      "step": 13716
+    },
+    {
+      "epoch": 12.662820216939764,
+      "grad_norm": 2.95703125,
+      "learning_rate": 8.559039359573049e-06,
+      "loss": 1.3091,
+      "step": 13717
+    },
+    {
+      "epoch": 12.663743364874222,
+      "grad_norm": 3.107421875,
+      "learning_rate": 8.552368245496998e-06,
+      "loss": 1.3219,
+      "step": 13718
+    },
+    {
+      "epoch": 12.664666512808678,
+      "grad_norm": 2.83203125,
+      "learning_rate": 8.545697131420947e-06,
+      "loss": 1.1359,
+      "step": 13719
+    },
+    {
+      "epoch": 12.665589660743134,
+      "grad_norm": 3.314453125,
+      "learning_rate": 8.539026017344897e-06,
+      "loss": 1.3599,
+      "step": 13720
+    },
+    {
+      "epoch": 12.66651280867759,
+      "grad_norm": 3.732421875,
+      "learning_rate": 8.532354903268846e-06,
+      "loss": 0.9822,
+      "step": 13721
+    },
+    {
+      "epoch": 12.667435956612048,
+      "grad_norm": 3.544921875,
+      "learning_rate": 8.525683789192795e-06,
+      "loss": 1.295,
+      "step": 13722
+    },
+    {
+      "epoch": 12.668359104546504,
+      "grad_norm": 3.5390625,
+      "learning_rate": 8.519012675116744e-06,
+      "loss": 1.2533,
+      "step": 13723
+    },
+    {
+      "epoch": 12.66928225248096,
+      "grad_norm": 2.96484375,
+      "learning_rate": 8.512341561040694e-06,
+      "loss": 1.2338,
+      "step": 13724
+    },
+    {
+      "epoch": 12.670205400415416,
+      "grad_norm": 3.27734375,
+      "learning_rate": 8.505670446964643e-06,
+      "loss": 1.0275,
+      "step": 13725
+    },
+    {
+      "epoch": 12.671128548349873,
+      "grad_norm": 3.30859375,
+      "learning_rate": 8.498999332888592e-06,
+      "loss": 1.2566,
+      "step": 13726
+    },
+    {
+      "epoch": 12.67205169628433,
+      "grad_norm": 2.904296875,
+      "learning_rate": 8.492328218812542e-06,
+      "loss": 1.2114,
+      "step": 13727
+    },
+    {
+      "epoch": 12.672974844218786,
+      "grad_norm": 3.55078125,
+      "learning_rate": 8.485657104736491e-06,
+      "loss": 1.156,
+      "step": 13728
+    },
+    {
+      "epoch": 12.673897992153243,
+      "grad_norm": 3.130859375,
+      "learning_rate": 8.47898599066044e-06,
+      "loss": 1.2902,
+      "step": 13729
+    },
+    {
+      "epoch": 12.6748211400877,
+      "grad_norm": 3.009765625,
+      "learning_rate": 8.47231487658439e-06,
+      "loss": 1.2964,
+      "step": 13730
+    },
+    {
+      "epoch": 12.675744288022155,
+      "grad_norm": 3.064453125,
+      "learning_rate": 8.46564376250834e-06,
+      "loss": 1.4661,
+      "step": 13731
+    },
+    {
+      "epoch": 12.676667435956611,
+      "grad_norm": 3.095703125,
+      "learning_rate": 8.458972648432288e-06,
+      "loss": 1.4482,
+      "step": 13732
+    },
+    {
+      "epoch": 12.67759058389107,
+      "grad_norm": 3.123046875,
+      "learning_rate": 8.452301534356239e-06,
+      "loss": 1.1273,
+      "step": 13733
+    },
+    {
+      "epoch": 12.678513731825525,
+      "grad_norm": 3.052734375,
+      "learning_rate": 8.445630420280187e-06,
+      "loss": 1.1841,
+      "step": 13734
+    },
+    {
+      "epoch": 12.679436879759981,
+      "grad_norm": 2.95703125,
+      "learning_rate": 8.438959306204136e-06,
+      "loss": 1.3228,
+      "step": 13735
+    },
+    {
+      "epoch": 12.680360027694437,
+      "grad_norm": 3.404296875,
+      "learning_rate": 8.432288192128087e-06,
+      "loss": 1.1034,
+      "step": 13736
+    },
+    {
+      "epoch": 12.681283175628895,
+      "grad_norm": 3.28515625,
+      "learning_rate": 8.425617078052036e-06,
+      "loss": 1.594,
+      "step": 13737
+    },
+    {
+      "epoch": 12.682206323563351,
+      "grad_norm": 3.439453125,
+      "learning_rate": 8.418945963975984e-06,
+      "loss": 1.0859,
+      "step": 13738
+    },
+    {
+      "epoch": 12.683129471497807,
+      "grad_norm": 3.408203125,
+      "learning_rate": 8.412274849899935e-06,
+      "loss": 1.2966,
+      "step": 13739
+    },
+    {
+      "epoch": 12.684052619432265,
+      "grad_norm": 3.83984375,
+      "learning_rate": 8.405603735823884e-06,
+      "loss": 1.1085,
+      "step": 13740
+    },
+    {
+      "epoch": 12.684975767366721,
+      "grad_norm": 2.90234375,
+      "learning_rate": 8.398932621747832e-06,
+      "loss": 1.2478,
+      "step": 13741
+    },
+    {
+      "epoch": 12.685898915301177,
+      "grad_norm": 3.21484375,
+      "learning_rate": 8.392261507671783e-06,
+      "loss": 1.1486,
+      "step": 13742
+    },
+    {
+      "epoch": 12.686822063235633,
+      "grad_norm": 3.849609375,
+      "learning_rate": 8.385590393595732e-06,
+      "loss": 1.2317,
+      "step": 13743
+    },
+    {
+      "epoch": 12.68774521117009,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.37891927951968e-06,
+      "loss": 1.1018,
+      "step": 13744
+    },
+    {
+      "epoch": 12.688668359104547,
+      "grad_norm": 3.287109375,
+      "learning_rate": 8.37224816544363e-06,
+      "loss": 1.251,
+      "step": 13745
+    },
+    {
+      "epoch": 12.689591507039003,
+      "grad_norm": 3.3203125,
+      "learning_rate": 8.365577051367578e-06,
+      "loss": 1.1252,
+      "step": 13746
+    },
+    {
+      "epoch": 12.690514654973459,
+      "grad_norm": 3.3359375,
+      "learning_rate": 8.358905937291527e-06,
+      "loss": 1.2384,
+      "step": 13747
+    },
+    {
+      "epoch": 12.691437802907917,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.352234823215478e-06,
+      "loss": 1.3076,
+      "step": 13748
+    },
+    {
+      "epoch": 12.692360950842373,
+      "grad_norm": 3.052734375,
+      "learning_rate": 8.345563709139426e-06,
+      "loss": 1.449,
+      "step": 13749
+    },
+    {
+      "epoch": 12.693284098776829,
+      "grad_norm": 3.275390625,
+      "learning_rate": 8.338892595063375e-06,
+      "loss": 1.1479,
+      "step": 13750
+    },
+    {
+      "epoch": 12.694207246711285,
+      "grad_norm": 3.470703125,
+      "learning_rate": 8.332221480987324e-06,
+      "loss": 1.2262,
+      "step": 13751
+    },
+    {
+      "epoch": 12.695130394645743,
+      "grad_norm": 3.2421875,
+      "learning_rate": 8.325550366911275e-06,
+      "loss": 1.3031,
+      "step": 13752
+    },
+    {
+      "epoch": 12.696053542580199,
+      "grad_norm": 3.2890625,
+      "learning_rate": 8.318879252835223e-06,
+      "loss": 1.4995,
+      "step": 13753
+    },
+    {
+      "epoch": 12.696976690514655,
+      "grad_norm": 3.232421875,
+      "learning_rate": 8.312208138759172e-06,
+      "loss": 1.361,
+      "step": 13754
+    },
+    {
+      "epoch": 12.69789983844911,
+      "grad_norm": 3.341796875,
+      "learning_rate": 8.305537024683123e-06,
+      "loss": 1.5684,
+      "step": 13755
+    },
+    {
+      "epoch": 12.698822986383568,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.298865910607072e-06,
+      "loss": 1.4297,
+      "step": 13756
+    },
+    {
+      "epoch": 12.699746134318024,
+      "grad_norm": 3.005859375,
+      "learning_rate": 8.29219479653102e-06,
+      "loss": 1.2242,
+      "step": 13757
+    },
+    {
+      "epoch": 12.70066928225248,
+      "grad_norm": 3.029296875,
+      "learning_rate": 8.285523682454971e-06,
+      "loss": 1.2781,
+      "step": 13758
+    },
+    {
+      "epoch": 12.701592430186938,
+      "grad_norm": 3.337890625,
+      "learning_rate": 8.27885256837892e-06,
+      "loss": 1.2786,
+      "step": 13759
+    },
+    {
+      "epoch": 12.702515578121394,
+      "grad_norm": 3.18359375,
+      "learning_rate": 8.272181454302869e-06,
+      "loss": 1.0981,
+      "step": 13760
+    },
+    {
+      "epoch": 12.70343872605585,
+      "grad_norm": 2.958984375,
+      "learning_rate": 8.265510340226819e-06,
+      "loss": 1.4507,
+      "step": 13761
+    },
+    {
+      "epoch": 12.704361873990306,
+      "grad_norm": 3.72265625,
+      "learning_rate": 8.258839226150768e-06,
+      "loss": 1.4106,
+      "step": 13762
+    },
+    {
+      "epoch": 12.705285021924764,
+      "grad_norm": 3.080078125,
+      "learning_rate": 8.252168112074717e-06,
+      "loss": 1.1815,
+      "step": 13763
+    },
+    {
+      "epoch": 12.70620816985922,
+      "grad_norm": 3.1640625,
+      "learning_rate": 8.245496997998667e-06,
+      "loss": 1.1829,
+      "step": 13764
+    },
+    {
+      "epoch": 12.707131317793676,
+      "grad_norm": 3.095703125,
+      "learning_rate": 8.238825883922616e-06,
+      "loss": 1.1648,
+      "step": 13765
+    },
+    {
+      "epoch": 12.708054465728132,
+      "grad_norm": 3.212890625,
+      "learning_rate": 8.232154769846565e-06,
+      "loss": 1.121,
+      "step": 13766
+    },
+    {
+      "epoch": 12.70897761366259,
+      "grad_norm": 3.6015625,
+      "learning_rate": 8.225483655770514e-06,
+      "loss": 1.1947,
+      "step": 13767
+    },
+    {
+      "epoch": 12.709900761597046,
+      "grad_norm": 3.642578125,
+      "learning_rate": 8.218812541694464e-06,
+      "loss": 1.3389,
+      "step": 13768
+    },
+    {
+      "epoch": 12.710823909531502,
+      "grad_norm": 2.93359375,
+      "learning_rate": 8.212141427618413e-06,
+      "loss": 0.9518,
+      "step": 13769
+    },
+    {
+      "epoch": 12.711747057465958,
+      "grad_norm": 2.9453125,
+      "learning_rate": 8.205470313542362e-06,
+      "loss": 1.4202,
+      "step": 13770
+    },
+    {
+      "epoch": 12.712670205400416,
+      "grad_norm": 3.1015625,
+      "learning_rate": 8.198799199466312e-06,
+      "loss": 1.183,
+      "step": 13771
+    },
+    {
+      "epoch": 12.713593353334872,
+      "grad_norm": 3.396484375,
+      "learning_rate": 8.192128085390261e-06,
+      "loss": 1.4119,
+      "step": 13772
+    },
+    {
+      "epoch": 12.714516501269328,
+      "grad_norm": 3.263671875,
+      "learning_rate": 8.18545697131421e-06,
+      "loss": 1.1626,
+      "step": 13773
+    },
+    {
+      "epoch": 12.715439649203784,
+      "grad_norm": 3.0546875,
+      "learning_rate": 8.178785857238159e-06,
+      "loss": 1.3569,
+      "step": 13774
+    },
+    {
+      "epoch": 12.716362797138242,
+      "grad_norm": 3.310546875,
+      "learning_rate": 8.172114743162108e-06,
+      "loss": 1.2089,
+      "step": 13775
+    },
+    {
+      "epoch": 12.717285945072698,
+      "grad_norm": 3.322265625,
+      "learning_rate": 8.165443629086058e-06,
+      "loss": 1.3535,
+      "step": 13776
+    },
+    {
+      "epoch": 12.718209093007154,
+      "grad_norm": 3.291015625,
+      "learning_rate": 8.158772515010007e-06,
+      "loss": 1.2528,
+      "step": 13777
+    },
+    {
+      "epoch": 12.719132240941612,
+      "grad_norm": 3.6796875,
+      "learning_rate": 8.152101400933956e-06,
+      "loss": 1.2063,
+      "step": 13778
+    },
+    {
+      "epoch": 12.720055388876068,
+      "grad_norm": 2.908203125,
+      "learning_rate": 8.145430286857905e-06,
+      "loss": 1.2551,
+      "step": 13779
+    },
+    {
+      "epoch": 12.720978536810524,
+      "grad_norm": 2.861328125,
+      "learning_rate": 8.138759172781855e-06,
+      "loss": 1.2939,
+      "step": 13780
+    },
+    {
+      "epoch": 12.72190168474498,
+      "grad_norm": 3.501953125,
+      "learning_rate": 8.132088058705804e-06,
+      "loss": 1.396,
+      "step": 13781
+    },
+    {
+      "epoch": 12.722824832679438,
+      "grad_norm": 3.33203125,
+      "learning_rate": 8.125416944629753e-06,
+      "loss": 0.9516,
+      "step": 13782
+    },
+    {
+      "epoch": 12.723747980613894,
+      "grad_norm": 3.4296875,
+      "learning_rate": 8.118745830553703e-06,
+      "loss": 1.1384,
+      "step": 13783
+    },
+    {
+      "epoch": 12.72467112854835,
+      "grad_norm": 3.181640625,
+      "learning_rate": 8.112074716477652e-06,
+      "loss": 1.3328,
+      "step": 13784
+    },
+    {
+      "epoch": 12.725594276482806,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.105403602401601e-06,
+      "loss": 1.1584,
+      "step": 13785
+    },
+    {
+      "epoch": 12.726517424417263,
+      "grad_norm": 3.013671875,
+      "learning_rate": 8.098732488325551e-06,
+      "loss": 1.28,
+      "step": 13786
+    },
+    {
+      "epoch": 12.72744057235172,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.0920613742495e-06,
+      "loss": 1.1217,
+      "step": 13787
+    },
+    {
+      "epoch": 12.728363720286175,
+      "grad_norm": 3.158203125,
+      "learning_rate": 8.085390260173449e-06,
+      "loss": 1.3389,
+      "step": 13788
+    },
+    {
+      "epoch": 12.729286868220633,
+      "grad_norm": 3.28515625,
+      "learning_rate": 8.0787191460974e-06,
+      "loss": 1.2268,
+      "step": 13789
+    },
+    {
+      "epoch": 12.73021001615509,
+      "grad_norm": 2.87890625,
+      "learning_rate": 8.072048032021348e-06,
+      "loss": 1.4216,
+      "step": 13790
+    },
+    {
+      "epoch": 12.731133164089545,
+      "grad_norm": 3.00390625,
+      "learning_rate": 8.065376917945297e-06,
+      "loss": 1.3892,
+      "step": 13791
+    },
+    {
+      "epoch": 12.732056312024001,
+      "grad_norm": 3.12890625,
+      "learning_rate": 8.058705803869246e-06,
+      "loss": 1.3672,
+      "step": 13792
+    },
+    {
+      "epoch": 12.73297945995846,
+      "grad_norm": 3.412109375,
+      "learning_rate": 8.052034689793196e-06,
+      "loss": 0.8112,
+      "step": 13793
+    },
+    {
+      "epoch": 12.733902607892915,
+      "grad_norm": 3.923828125,
+      "learning_rate": 8.045363575717145e-06,
+      "loss": 1.1766,
+      "step": 13794
+    },
+    {
+      "epoch": 12.734825755827371,
+      "grad_norm": 3.3203125,
+      "learning_rate": 8.038692461641094e-06,
+      "loss": 1.2888,
+      "step": 13795
+    },
+    {
+      "epoch": 12.735748903761827,
+      "grad_norm": 3.009765625,
+      "learning_rate": 8.032021347565045e-06,
+      "loss": 1.1798,
+      "step": 13796
+    },
+    {
+      "epoch": 12.736672051696285,
+      "grad_norm": 2.89453125,
+      "learning_rate": 8.025350233488993e-06,
+      "loss": 1.5344,
+      "step": 13797
+    },
+    {
+      "epoch": 12.737595199630741,
+      "grad_norm": 3.515625,
+      "learning_rate": 8.018679119412942e-06,
+      "loss": 1.3225,
+      "step": 13798
+    },
+    {
+      "epoch": 12.738518347565197,
+      "grad_norm": 3.466796875,
+      "learning_rate": 8.012008005336893e-06,
+      "loss": 1.2129,
+      "step": 13799
+    },
+    {
+      "epoch": 12.739441495499653,
+      "grad_norm": 2.90234375,
+      "learning_rate": 8.005336891260842e-06,
+      "loss": 1.3254,
+      "step": 13800
+    },
+    {
+      "epoch": 12.740364643434111,
+      "grad_norm": 2.75390625,
+      "learning_rate": 7.99866577718479e-06,
+      "loss": 1.0883,
+      "step": 13801
+    },
+    {
+      "epoch": 12.741287791368567,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.99199466310874e-06,
+      "loss": 1.2616,
+      "step": 13802
+    },
+    {
+      "epoch": 12.742210939303023,
+      "grad_norm": 3.162109375,
+      "learning_rate": 7.985323549032688e-06,
+      "loss": 1.1073,
+      "step": 13803
+    },
+    {
+      "epoch": 12.743134087237479,
+      "grad_norm": 3.474609375,
+      "learning_rate": 7.978652434956639e-06,
+      "loss": 1.1924,
+      "step": 13804
+    },
+    {
+      "epoch": 12.744057235171937,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.971981320880587e-06,
+      "loss": 1.1644,
+      "step": 13805
+    },
+    {
+      "epoch": 12.744980383106393,
+      "grad_norm": 3.080078125,
+      "learning_rate": 7.965310206804536e-06,
+      "loss": 1.1407,
+      "step": 13806
+    },
+    {
+      "epoch": 12.745903531040849,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.958639092728485e-06,
+      "loss": 1.3104,
+      "step": 13807
+    },
+    {
+      "epoch": 12.746826678975307,
+      "grad_norm": 3.3984375,
+      "learning_rate": 7.951967978652436e-06,
+      "loss": 0.9166,
+      "step": 13808
+    },
+    {
+      "epoch": 12.747749826909763,
+      "grad_norm": 3.076171875,
+      "learning_rate": 7.945296864576384e-06,
+      "loss": 1.4062,
+      "step": 13809
+    },
+    {
+      "epoch": 12.748672974844219,
+      "grad_norm": 3.0,
+      "learning_rate": 7.938625750500333e-06,
+      "loss": 1.2314,
+      "step": 13810
+    },
+    {
+      "epoch": 12.749596122778675,
+      "grad_norm": 3.642578125,
+      "learning_rate": 7.931954636424284e-06,
+      "loss": 1.397,
+      "step": 13811
+    },
+    {
+      "epoch": 12.750519270713133,
+      "grad_norm": 3.166015625,
+      "learning_rate": 7.925283522348233e-06,
+      "loss": 1.3704,
+      "step": 13812
+    },
+    {
+      "epoch": 12.751442418647589,
+      "grad_norm": 3.1953125,
+      "learning_rate": 7.918612408272181e-06,
+      "loss": 1.15,
+      "step": 13813
+    },
+    {
+      "epoch": 12.752365566582045,
+      "grad_norm": 3.412109375,
+      "learning_rate": 7.91194129419613e-06,
+      "loss": 1.2395,
+      "step": 13814
+    },
+    {
+      "epoch": 12.7532887145165,
+      "grad_norm": 2.755859375,
+      "learning_rate": 7.90527018012008e-06,
+      "loss": 0.959,
+      "step": 13815
+    },
+    {
+      "epoch": 12.754211862450958,
+      "grad_norm": 3.1640625,
+      "learning_rate": 7.89859906604403e-06,
+      "loss": 1.4575,
+      "step": 13816
+    },
+    {
+      "epoch": 12.755135010385414,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.891927951967978e-06,
+      "loss": 1.2869,
+      "step": 13817
+    },
+    {
+      "epoch": 12.75605815831987,
+      "grad_norm": 3.34765625,
+      "learning_rate": 7.885256837891929e-06,
+      "loss": 1.2789,
+      "step": 13818
+    },
+    {
+      "epoch": 12.756981306254326,
+      "grad_norm": 2.974609375,
+      "learning_rate": 7.878585723815878e-06,
+      "loss": 1.182,
+      "step": 13819
+    },
+    {
+      "epoch": 12.757904454188784,
+      "grad_norm": 3.119140625,
+      "learning_rate": 7.871914609739826e-06,
+      "loss": 1.1888,
+      "step": 13820
+    },
+    {
+      "epoch": 12.75882760212324,
+      "grad_norm": 3.05078125,
+      "learning_rate": 7.865243495663777e-06,
+      "loss": 1.0544,
+      "step": 13821
+    },
+    {
+      "epoch": 12.759750750057696,
+      "grad_norm": 3.19140625,
+      "learning_rate": 7.858572381587726e-06,
+      "loss": 1.4397,
+      "step": 13822
+    },
+    {
+      "epoch": 12.760673897992152,
+      "grad_norm": 3.28515625,
+      "learning_rate": 7.851901267511675e-06,
+      "loss": 1.1014,
+      "step": 13823
+    },
+    {
+      "epoch": 12.76159704592661,
+      "grad_norm": 2.978515625,
+      "learning_rate": 7.845230153435625e-06,
+      "loss": 1.3074,
+      "step": 13824
+    },
+    {
+      "epoch": 12.762520193861066,
+      "grad_norm": 3.11328125,
+      "learning_rate": 7.838559039359574e-06,
+      "loss": 1.1138,
+      "step": 13825
+    },
+    {
+      "epoch": 12.763443341795522,
+      "grad_norm": 2.880859375,
+      "learning_rate": 7.831887925283523e-06,
+      "loss": 1.2764,
+      "step": 13826
+    },
+    {
+      "epoch": 12.76436648972998,
+      "grad_norm": 3.1484375,
+      "learning_rate": 7.825216811207473e-06,
+      "loss": 1.3887,
+      "step": 13827
+    },
+    {
+      "epoch": 12.765289637664436,
+      "grad_norm": 3.232421875,
+      "learning_rate": 7.818545697131422e-06,
+      "loss": 1.1495,
+      "step": 13828
+    },
+    {
+      "epoch": 12.766212785598892,
+      "grad_norm": 3.1328125,
+      "learning_rate": 7.811874583055371e-06,
+      "loss": 1.0717,
+      "step": 13829
+    },
+    {
+      "epoch": 12.767135933533348,
+      "grad_norm": 3.447265625,
+      "learning_rate": 7.80520346897932e-06,
+      "loss": 0.7673,
+      "step": 13830
+    },
+    {
+      "epoch": 12.768059081467806,
+      "grad_norm": 3.0234375,
+      "learning_rate": 7.798532354903269e-06,
+      "loss": 1.4788,
+      "step": 13831
+    },
+    {
+      "epoch": 12.768982229402262,
+      "grad_norm": 3.3359375,
+      "learning_rate": 7.791861240827217e-06,
+      "loss": 1.2178,
+      "step": 13832
+    },
+    {
+      "epoch": 12.769905377336718,
+      "grad_norm": 3.29296875,
+      "learning_rate": 7.785190126751168e-06,
+      "loss": 1.3037,
+      "step": 13833
+    },
+    {
+      "epoch": 12.770828525271174,
+      "grad_norm": 3.5390625,
+      "learning_rate": 7.778519012675117e-06,
+      "loss": 1.0355,
+      "step": 13834
+    },
+    {
+      "epoch": 12.771751673205632,
+      "grad_norm": 3.47265625,
+      "learning_rate": 7.771847898599065e-06,
+      "loss": 1.301,
+      "step": 13835
+    },
+    {
+      "epoch": 12.772674821140088,
+      "grad_norm": 2.740234375,
+      "learning_rate": 7.765176784523016e-06,
+      "loss": 1.2275,
+      "step": 13836
+    },
+    {
+      "epoch": 12.773597969074544,
+      "grad_norm": 2.998046875,
+      "learning_rate": 7.758505670446965e-06,
+      "loss": 1.3101,
+      "step": 13837
+    },
+    {
+      "epoch": 12.774521117009002,
+      "grad_norm": 3.240234375,
+      "learning_rate": 7.751834556370914e-06,
+      "loss": 1.0825,
+      "step": 13838
+    },
+    {
+      "epoch": 12.775444264943458,
+      "grad_norm": 3.138671875,
+      "learning_rate": 7.745163442294862e-06,
+      "loss": 1.0386,
+      "step": 13839
+    },
+    {
+      "epoch": 12.776367412877914,
+      "grad_norm": 3.2265625,
+      "learning_rate": 7.738492328218813e-06,
+      "loss": 1.0981,
+      "step": 13840
+    },
+    {
+      "epoch": 12.77729056081237,
+      "grad_norm": 3.591796875,
+      "learning_rate": 7.731821214142762e-06,
+      "loss": 0.8163,
+      "step": 13841
+    },
+    {
+      "epoch": 12.778213708746827,
+      "grad_norm": 3.0859375,
+      "learning_rate": 7.72515010006671e-06,
+      "loss": 1.163,
+      "step": 13842
+    },
+    {
+      "epoch": 12.779136856681284,
+      "grad_norm": 2.91796875,
+      "learning_rate": 7.718478985990661e-06,
+      "loss": 1.2585,
+      "step": 13843
+    },
+    {
+      "epoch": 12.78006000461574,
+      "grad_norm": 2.892578125,
+      "learning_rate": 7.71180787191461e-06,
+      "loss": 1.1703,
+      "step": 13844
+    },
+    {
+      "epoch": 12.780983152550196,
+      "grad_norm": 3.572265625,
+      "learning_rate": 7.705136757838559e-06,
+      "loss": 1.0764,
+      "step": 13845
+    },
+    {
+      "epoch": 12.781906300484653,
+      "grad_norm": 2.962890625,
+      "learning_rate": 7.69846564376251e-06,
+      "loss": 1.1036,
+      "step": 13846
+    },
+    {
+      "epoch": 12.78282944841911,
+      "grad_norm": 2.900390625,
+      "learning_rate": 7.691794529686458e-06,
+      "loss": 1.1449,
+      "step": 13847
+    },
+    {
+      "epoch": 12.783752596353565,
+      "grad_norm": 3.5078125,
+      "learning_rate": 7.685123415610407e-06,
+      "loss": 1.3035,
+      "step": 13848
+    },
+    {
+      "epoch": 12.784675744288021,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.678452301534357e-06,
+      "loss": 1.3452,
+      "step": 13849
+    },
+    {
+      "epoch": 12.78559889222248,
+      "grad_norm": 2.974609375,
+      "learning_rate": 7.671781187458306e-06,
+      "loss": 1.078,
+      "step": 13850
+    },
+    {
+      "epoch": 12.786522040156935,
+      "grad_norm": 3.076171875,
+      "learning_rate": 7.665110073382255e-06,
+      "loss": 1.4846,
+      "step": 13851
+    },
+    {
+      "epoch": 12.787445188091391,
+      "grad_norm": 3.306640625,
+      "learning_rate": 7.658438959306206e-06,
+      "loss": 1.1597,
+      "step": 13852
+    },
+    {
+      "epoch": 12.788368336025847,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.651767845230154e-06,
+      "loss": 1.4888,
+      "step": 13853
+    },
+    {
+      "epoch": 12.789291483960305,
+      "grad_norm": 3.248046875,
+      "learning_rate": 7.645096731154103e-06,
+      "loss": 1.2227,
+      "step": 13854
+    },
+    {
+      "epoch": 12.790214631894761,
+      "grad_norm": 3.34765625,
+      "learning_rate": 7.638425617078054e-06,
+      "loss": 1.2175,
+      "step": 13855
+    },
+    {
+      "epoch": 12.791137779829217,
+      "grad_norm": 3.33203125,
+      "learning_rate": 7.631754503002003e-06,
+      "loss": 1.05,
+      "step": 13856
+    },
+    {
+      "epoch": 12.792060927763675,
+      "grad_norm": 3.34765625,
+      "learning_rate": 7.6250833889259505e-06,
+      "loss": 1.3037,
+      "step": 13857
+    },
+    {
+      "epoch": 12.792984075698131,
+      "grad_norm": 3.251953125,
+      "learning_rate": 7.618412274849901e-06,
+      "loss": 1.3455,
+      "step": 13858
+    },
+    {
+      "epoch": 12.793907223632587,
+      "grad_norm": 3.365234375,
+      "learning_rate": 7.61174116077385e-06,
+      "loss": 1.359,
+      "step": 13859
+    },
+    {
+      "epoch": 12.794830371567043,
+      "grad_norm": 3.10546875,
+      "learning_rate": 7.605070046697799e-06,
+      "loss": 1.1305,
+      "step": 13860
+    },
+    {
+      "epoch": 12.7957535195015,
+      "grad_norm": 3.564453125,
+      "learning_rate": 7.5983989326217475e-06,
+      "loss": 1.1881,
+      "step": 13861
+    },
+    {
+      "epoch": 12.796676667435957,
+      "grad_norm": 2.826171875,
+      "learning_rate": 7.591727818545698e-06,
+      "loss": 0.7156,
+      "step": 13862
+    },
+    {
+      "epoch": 12.797599815370413,
+      "grad_norm": 2.962890625,
+      "learning_rate": 7.585056704469647e-06,
+      "loss": 1.21,
+      "step": 13863
+    },
+    {
+      "epoch": 12.798522963304869,
+      "grad_norm": 3.119140625,
+      "learning_rate": 7.578385590393596e-06,
+      "loss": 1.1085,
+      "step": 13864
+    },
+    {
+      "epoch": 12.799446111239327,
+      "grad_norm": 2.822265625,
+      "learning_rate": 7.571714476317545e-06,
+      "loss": 0.9885,
+      "step": 13865
+    },
+    {
+      "epoch": 12.800369259173783,
+      "grad_norm": 3.0078125,
+      "learning_rate": 7.565043362241495e-06,
+      "loss": 1.2063,
+      "step": 13866
+    },
+    {
+      "epoch": 12.801292407108239,
+      "grad_norm": 3.369140625,
+      "learning_rate": 7.558372248165444e-06,
+      "loss": 1.0957,
+      "step": 13867
+    },
+    {
+      "epoch": 12.802215555042695,
+      "grad_norm": 3.177734375,
+      "learning_rate": 7.5517011340893935e-06,
+      "loss": 1.0342,
+      "step": 13868
+    },
+    {
+      "epoch": 12.803138702977153,
+      "grad_norm": 3.431640625,
+      "learning_rate": 7.545030020013342e-06,
+      "loss": 1.0903,
+      "step": 13869
+    },
+    {
+      "epoch": 12.804061850911609,
+      "grad_norm": 2.873046875,
+      "learning_rate": 7.538358905937291e-06,
+      "loss": 1.4478,
+      "step": 13870
+    },
+    {
+      "epoch": 12.804984998846065,
+      "grad_norm": 3.30859375,
+      "learning_rate": 7.531687791861242e-06,
+      "loss": 1.3411,
+      "step": 13871
+    },
+    {
+      "epoch": 12.80590814678052,
+      "grad_norm": 3.1171875,
+      "learning_rate": 7.5250166777851904e-06,
+      "loss": 1.4282,
+      "step": 13872
+    },
+    {
+      "epoch": 12.806831294714978,
+      "grad_norm": 3.19921875,
+      "learning_rate": 7.518345563709139e-06,
+      "loss": 1.2054,
+      "step": 13873
+    },
+    {
+      "epoch": 12.807754442649435,
+      "grad_norm": 3.091796875,
+      "learning_rate": 7.51167444963309e-06,
+      "loss": 1.2729,
+      "step": 13874
+    },
+    {
+      "epoch": 12.80867759058389,
+      "grad_norm": 2.94921875,
+      "learning_rate": 7.5050033355570386e-06,
+      "loss": 1.1378,
+      "step": 13875
+    },
+    {
+      "epoch": 12.809600738518348,
+      "grad_norm": 3.244140625,
+      "learning_rate": 7.498332221480987e-06,
+      "loss": 1.4343,
+      "step": 13876
+    },
+    {
+      "epoch": 12.810523886452804,
+      "grad_norm": 3.1796875,
+      "learning_rate": 7.491661107404938e-06,
+      "loss": 1.46,
+      "step": 13877
+    },
+    {
+      "epoch": 12.81144703438726,
+      "grad_norm": 3.1953125,
+      "learning_rate": 7.484989993328887e-06,
+      "loss": 1.2489,
+      "step": 13878
+    },
+    {
+      "epoch": 12.812370182321716,
+      "grad_norm": 3.048828125,
+      "learning_rate": 7.4783188792528355e-06,
+      "loss": 1.1675,
+      "step": 13879
+    },
+    {
+      "epoch": 12.813293330256174,
+      "grad_norm": 3.111328125,
+      "learning_rate": 7.471647765176785e-06,
+      "loss": 1.0547,
+      "step": 13880
+    },
+    {
+      "epoch": 12.81421647819063,
+      "grad_norm": 3.591796875,
+      "learning_rate": 7.464976651100734e-06,
+      "loss": 1.197,
+      "step": 13881
+    },
+    {
+      "epoch": 12.815139626125086,
+      "grad_norm": 3.19921875,
+      "learning_rate": 7.458305537024683e-06,
+      "loss": 1.373,
+      "step": 13882
+    },
+    {
+      "epoch": 12.816062774059542,
+      "grad_norm": 3.0390625,
+      "learning_rate": 7.451634422948633e-06,
+      "loss": 1.4165,
+      "step": 13883
+    },
+    {
+      "epoch": 12.816985921994,
+      "grad_norm": 2.994140625,
+      "learning_rate": 7.444963308872582e-06,
+      "loss": 1.3109,
+      "step": 13884
+    },
+    {
+      "epoch": 12.817909069928456,
+      "grad_norm": 3.029296875,
+      "learning_rate": 7.438292194796531e-06,
+      "loss": 1.1272,
+      "step": 13885
+    },
+    {
+      "epoch": 12.818832217862912,
+      "grad_norm": 3.5859375,
+      "learning_rate": 7.43162108072048e-06,
+      "loss": 1.1821,
+      "step": 13886
+    },
+    {
+      "epoch": 12.81975536579737,
+      "grad_norm": 2.8828125,
+      "learning_rate": 7.42494996664443e-06,
+      "loss": 1.2676,
+      "step": 13887
+    },
+    {
+      "epoch": 12.820678513731826,
+      "grad_norm": 2.767578125,
+      "learning_rate": 7.418278852568379e-06,
+      "loss": 1.2854,
+      "step": 13888
+    },
+    {
+      "epoch": 12.821601661666282,
+      "grad_norm": 3.115234375,
+      "learning_rate": 7.411607738492328e-06,
+      "loss": 1.312,
+      "step": 13889
+    },
+    {
+      "epoch": 12.822524809600738,
+      "grad_norm": 3.0546875,
+      "learning_rate": 7.4049366244162785e-06,
+      "loss": 1.3584,
+      "step": 13890
+    },
+    {
+      "epoch": 12.823447957535196,
+      "grad_norm": 3.37109375,
+      "learning_rate": 7.398265510340227e-06,
+      "loss": 0.92,
+      "step": 13891
+    },
+    {
+      "epoch": 12.824371105469652,
+      "grad_norm": 3.2265625,
+      "learning_rate": 7.391594396264176e-06,
+      "loss": 1.1794,
+      "step": 13892
+    },
+    {
+      "epoch": 12.825294253404108,
+      "grad_norm": 3.5234375,
+      "learning_rate": 7.384923282188126e-06,
+      "loss": 1.2238,
+      "step": 13893
+    },
+    {
+      "epoch": 12.826217401338564,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.3782521681120754e-06,
+      "loss": 1.2791,
+      "step": 13894
+    },
+    {
+      "epoch": 12.827140549273022,
+      "grad_norm": 2.97265625,
+      "learning_rate": 7.371581054036024e-06,
+      "loss": 1.2449,
+      "step": 13895
+    },
+    {
+      "epoch": 12.828063697207478,
+      "grad_norm": 2.99609375,
+      "learning_rate": 7.364909939959974e-06,
+      "loss": 0.9579,
+      "step": 13896
+    },
+    {
+      "epoch": 12.828986845141934,
+      "grad_norm": 3.02734375,
+      "learning_rate": 7.358238825883923e-06,
+      "loss": 1.1497,
+      "step": 13897
+    },
+    {
+      "epoch": 12.82990999307639,
+      "grad_norm": 3.384765625,
+      "learning_rate": 7.3515677118078716e-06,
+      "loss": 1.2898,
+      "step": 13898
+    },
+    {
+      "epoch": 12.830833141010848,
+      "grad_norm": 3.44921875,
+      "learning_rate": 7.344896597731822e-06,
+      "loss": 1.2849,
+      "step": 13899
+    },
+    {
+      "epoch": 12.831756288945304,
+      "grad_norm": 3.271484375,
+      "learning_rate": 7.338225483655771e-06,
+      "loss": 1.5513,
+      "step": 13900
+    },
+    {
+      "epoch": 12.83267943687976,
+      "grad_norm": 3.177734375,
+      "learning_rate": 7.33155436957972e-06,
+      "loss": 1.1467,
+      "step": 13901
+    },
+    {
+      "epoch": 12.833602584814216,
+      "grad_norm": 3.2578125,
+      "learning_rate": 7.32488325550367e-06,
+      "loss": 1.3479,
+      "step": 13902
+    },
+    {
+      "epoch": 12.834525732748673,
+      "grad_norm": 3.05078125,
+      "learning_rate": 7.318212141427619e-06,
+      "loss": 1.1653,
+      "step": 13903
+    },
+    {
+      "epoch": 12.83544888068313,
+      "grad_norm": 3.16015625,
+      "learning_rate": 7.311541027351568e-06,
+      "loss": 1.353,
+      "step": 13904
+    },
+    {
+      "epoch": 12.836372028617586,
+      "grad_norm": 3.443359375,
+      "learning_rate": 7.304869913275518e-06,
+      "loss": 1.5957,
+      "step": 13905
+    },
+    {
+      "epoch": 12.837295176552043,
+      "grad_norm": 3.25390625,
+      "learning_rate": 7.298198799199467e-06,
+      "loss": 1.2189,
+      "step": 13906
+    },
+    {
+      "epoch": 12.8382183244865,
+      "grad_norm": 3.060546875,
+      "learning_rate": 7.291527685123416e-06,
+      "loss": 1.333,
+      "step": 13907
+    },
+    {
+      "epoch": 12.839141472420955,
+      "grad_norm": 3.70703125,
+      "learning_rate": 7.284856571047365e-06,
+      "loss": 1.1766,
+      "step": 13908
+    },
+    {
+      "epoch": 12.840064620355411,
+      "grad_norm": 3.189453125,
+      "learning_rate": 7.2781854569713145e-06,
+      "loss": 1.3093,
+      "step": 13909
+    },
+    {
+      "epoch": 12.84098776828987,
+      "grad_norm": 3.080078125,
+      "learning_rate": 7.271514342895263e-06,
+      "loss": 1.3511,
+      "step": 13910
+    },
+    {
+      "epoch": 12.841910916224325,
+      "grad_norm": 3.052734375,
+      "learning_rate": 7.264843228819212e-06,
+      "loss": 1.1829,
+      "step": 13911
+    },
+    {
+      "epoch": 12.842834064158781,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.258172114743163e-06,
+      "loss": 1.1973,
+      "step": 13912
+    },
+    {
+      "epoch": 12.843757212093237,
+      "grad_norm": 3.42578125,
+      "learning_rate": 7.2515010006671115e-06,
+      "loss": 1.4839,
+      "step": 13913
+    },
+    {
+      "epoch": 12.844680360027695,
+      "grad_norm": 3.396484375,
+      "learning_rate": 7.24482988659106e-06,
+      "loss": 1.2661,
+      "step": 13914
+    },
+    {
+      "epoch": 12.845603507962151,
+      "grad_norm": 3.060546875,
+      "learning_rate": 7.238158772515011e-06,
+      "loss": 1.129,
+      "step": 13915
+    },
+    {
+      "epoch": 12.846526655896607,
+      "grad_norm": 3.337890625,
+      "learning_rate": 7.23148765843896e-06,
+      "loss": 1.325,
+      "step": 13916
+    },
+    {
+      "epoch": 12.847449803831063,
+      "grad_norm": 2.951171875,
+      "learning_rate": 7.2248165443629084e-06,
+      "loss": 1.0145,
+      "step": 13917
+    },
+    {
+      "epoch": 12.848372951765521,
+      "grad_norm": 3.169921875,
+      "learning_rate": 7.218145430286859e-06,
+      "loss": 1.2734,
+      "step": 13918
+    },
+    {
+      "epoch": 12.849296099699977,
+      "grad_norm": 3.00390625,
+      "learning_rate": 7.211474316210808e-06,
+      "loss": 1.2545,
+      "step": 13919
+    },
+    {
+      "epoch": 12.850219247634433,
+      "grad_norm": 2.87890625,
+      "learning_rate": 7.204803202134757e-06,
+      "loss": 1.3413,
+      "step": 13920
+    },
+    {
+      "epoch": 12.851142395568889,
+      "grad_norm": 2.931640625,
+      "learning_rate": 7.198132088058706e-06,
+      "loss": 1.5203,
+      "step": 13921
+    },
+    {
+      "epoch": 12.852065543503347,
+      "grad_norm": 2.888671875,
+      "learning_rate": 7.191460973982656e-06,
+      "loss": 1.2815,
+      "step": 13922
+    },
+    {
+      "epoch": 12.852988691437803,
+      "grad_norm": 3.41796875,
+      "learning_rate": 7.184789859906605e-06,
+      "loss": 1.3713,
+      "step": 13923
+    },
+    {
+      "epoch": 12.853911839372259,
+      "grad_norm": 2.98046875,
+      "learning_rate": 7.178118745830554e-06,
+      "loss": 1.4011,
+      "step": 13924
+    },
+    {
+      "epoch": 12.854834987306717,
+      "grad_norm": 3.103515625,
+      "learning_rate": 7.171447631754503e-06,
+      "loss": 1.3772,
+      "step": 13925
+    },
+    {
+      "epoch": 12.855758135241173,
+      "grad_norm": 3.041015625,
+      "learning_rate": 7.164776517678452e-06,
+      "loss": 1.3059,
+      "step": 13926
+    },
+    {
+      "epoch": 12.856681283175629,
+      "grad_norm": 3.248046875,
+      "learning_rate": 7.1581054036024026e-06,
+      "loss": 1.3706,
+      "step": 13927
+    },
+    {
+      "epoch": 12.857604431110085,
+      "grad_norm": 2.951171875,
+      "learning_rate": 7.151434289526351e-06,
+      "loss": 1.3746,
+      "step": 13928
+    },
+    {
+      "epoch": 12.858527579044543,
+      "grad_norm": 3.005859375,
+      "learning_rate": 7.1447631754503e-06,
+      "loss": 1.1838,
+      "step": 13929
+    },
+    {
+      "epoch": 12.859450726978999,
+      "grad_norm": 3.177734375,
+      "learning_rate": 7.138092061374251e-06,
+      "loss": 1.2849,
+      "step": 13930
+    },
+    {
+      "epoch": 12.860373874913455,
+      "grad_norm": 3.298828125,
+      "learning_rate": 7.1314209472981995e-06,
+      "loss": 1.2539,
+      "step": 13931
+    },
+    {
+      "epoch": 12.86129702284791,
+      "grad_norm": 2.990234375,
+      "learning_rate": 7.124749833222148e-06,
+      "loss": 1.4072,
+      "step": 13932
+    },
+    {
+      "epoch": 12.862220170782368,
+      "grad_norm": 3.271484375,
+      "learning_rate": 7.118078719146097e-06,
+      "loss": 1.4172,
+      "step": 13933
+    },
+    {
+      "epoch": 12.863143318716824,
+      "grad_norm": 3.291015625,
+      "learning_rate": 7.111407605070048e-06,
+      "loss": 1.4229,
+      "step": 13934
+    },
+    {
+      "epoch": 12.86406646665128,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.1047364909939965e-06,
+      "loss": 1.416,
+      "step": 13935
+    },
+    {
+      "epoch": 12.864989614585737,
+      "grad_norm": 2.974609375,
+      "learning_rate": 7.098065376917945e-06,
+      "loss": 1.192,
+      "step": 13936
+    },
+    {
+      "epoch": 12.865912762520194,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.091394262841895e-06,
+      "loss": 1.2042,
+      "step": 13937
+    },
+    {
+      "epoch": 12.86683591045465,
+      "grad_norm": 3.0078125,
+      "learning_rate": 7.084723148765844e-06,
+      "loss": 1.3154,
+      "step": 13938
+    },
+    {
+      "epoch": 12.867759058389106,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.078052034689793e-06,
+      "loss": 1.373,
+      "step": 13939
+    },
+    {
+      "epoch": 12.868682206323562,
+      "grad_norm": 2.939453125,
+      "learning_rate": 7.071380920613743e-06,
+      "loss": 1.426,
+      "step": 13940
+    },
+    {
+      "epoch": 12.86960535425802,
+      "grad_norm": 2.736328125,
+      "learning_rate": 7.064709806537692e-06,
+      "loss": 1.1584,
+      "step": 13941
+    },
+    {
+      "epoch": 12.870528502192476,
+      "grad_norm": 3.255859375,
+      "learning_rate": 7.058038692461641e-06,
+      "loss": 0.8938,
+      "step": 13942
+    },
+    {
+      "epoch": 12.871451650126932,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.051367578385591e-06,
+      "loss": 1.4075,
+      "step": 13943
+    },
+    {
+      "epoch": 12.87237479806139,
+      "grad_norm": 2.826171875,
+      "learning_rate": 7.04469646430954e-06,
+      "loss": 1.2139,
+      "step": 13944
+    },
+    {
+      "epoch": 12.873297945995846,
+      "grad_norm": 3.2109375,
+      "learning_rate": 7.038025350233489e-06,
+      "loss": 1.0851,
+      "step": 13945
+    },
+    {
+      "epoch": 12.874221093930302,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.0313542361574394e-06,
+      "loss": 1.2449,
+      "step": 13946
+    },
+    {
+      "epoch": 12.875144241864758,
+      "grad_norm": 3.08203125,
+      "learning_rate": 7.024683122081388e-06,
+      "loss": 1.3787,
+      "step": 13947
+    },
+    {
+      "epoch": 12.876067389799216,
+      "grad_norm": 3.146484375,
+      "learning_rate": 7.018012008005337e-06,
+      "loss": 1.0999,
+      "step": 13948
+    },
+    {
+      "epoch": 12.876990537733672,
+      "grad_norm": 3.0078125,
+      "learning_rate": 7.011340893929287e-06,
+      "loss": 1.4277,
+      "step": 13949
+    },
+    {
+      "epoch": 12.877913685668128,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.0046697798532356e-06,
+      "loss": 1.1489,
+      "step": 13950
+    },
+    {
+      "epoch": 12.878836833602584,
+      "grad_norm": 3.24609375,
+      "learning_rate": 6.997998665777185e-06,
+      "loss": 1.4355,
+      "step": 13951
+    },
+    {
+      "epoch": 12.879759981537042,
+      "grad_norm": 3.3671875,
+      "learning_rate": 6.991327551701135e-06,
+      "loss": 1.4392,
+      "step": 13952
+    },
+    {
+      "epoch": 12.880683129471498,
+      "grad_norm": 3.185546875,
+      "learning_rate": 6.984656437625084e-06,
+      "loss": 1.4846,
+      "step": 13953
+    },
+    {
+      "epoch": 12.881606277405954,
+      "grad_norm": 3.025390625,
+      "learning_rate": 6.9779853235490325e-06,
+      "loss": 1.4111,
+      "step": 13954
+    },
+    {
+      "epoch": 12.882529425340412,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.971314209472981e-06,
+      "loss": 1.1724,
+      "step": 13955
+    },
+    {
+      "epoch": 12.883452573274868,
+      "grad_norm": 3.486328125,
+      "learning_rate": 6.964643095396932e-06,
+      "loss": 1.1184,
+      "step": 13956
+    },
+    {
+      "epoch": 12.884375721209324,
+      "grad_norm": 3.1015625,
+      "learning_rate": 6.957971981320881e-06,
+      "loss": 1.3264,
+      "step": 13957
+    },
+    {
+      "epoch": 12.88529886914378,
+      "grad_norm": 3.44140625,
+      "learning_rate": 6.9513008672448295e-06,
+      "loss": 0.9978,
+      "step": 13958
+    },
+    {
+      "epoch": 12.886222017078238,
+      "grad_norm": 3.408203125,
+      "learning_rate": 6.94462975316878e-06,
+      "loss": 1.1462,
+      "step": 13959
+    },
+    {
+      "epoch": 12.887145165012694,
+      "grad_norm": 3.31640625,
+      "learning_rate": 6.937958639092729e-06,
+      "loss": 1.3945,
+      "step": 13960
+    },
+    {
+      "epoch": 12.88806831294715,
+      "grad_norm": 3.134765625,
+      "learning_rate": 6.931287525016678e-06,
+      "loss": 1.416,
+      "step": 13961
+    },
+    {
+      "epoch": 12.888991460881606,
+      "grad_norm": 3.134765625,
+      "learning_rate": 6.924616410940628e-06,
+      "loss": 0.9609,
+      "step": 13962
+    },
+    {
+      "epoch": 12.889914608816063,
+      "grad_norm": 3.033203125,
+      "learning_rate": 6.917945296864577e-06,
+      "loss": 1.2644,
+      "step": 13963
+    },
+    {
+      "epoch": 12.89083775675052,
+      "grad_norm": 3.31640625,
+      "learning_rate": 6.911274182788526e-06,
+      "loss": 1.2679,
+      "step": 13964
+    },
+    {
+      "epoch": 12.891760904684975,
+      "grad_norm": 3.0234375,
+      "learning_rate": 6.9046030687124755e-06,
+      "loss": 1.2029,
+      "step": 13965
+    },
+    {
+      "epoch": 12.892684052619432,
+      "grad_norm": 3.05859375,
+      "learning_rate": 6.897931954636424e-06,
+      "loss": 1.0068,
+      "step": 13966
+    },
+    {
+      "epoch": 12.89360720055389,
+      "grad_norm": 3.25390625,
+      "learning_rate": 6.891260840560373e-06,
+      "loss": 1.2651,
+      "step": 13967
+    },
+    {
+      "epoch": 12.894530348488345,
+      "grad_norm": 2.95703125,
+      "learning_rate": 6.884589726484324e-06,
+      "loss": 1.2963,
+      "step": 13968
+    },
+    {
+      "epoch": 12.895453496422801,
+      "grad_norm": 3.263671875,
+      "learning_rate": 6.8779186124082724e-06,
+      "loss": 1.4387,
+      "step": 13969
+    },
+    {
+      "epoch": 12.896376644357257,
+      "grad_norm": 2.947265625,
+      "learning_rate": 6.871247498332221e-06,
+      "loss": 1.2487,
+      "step": 13970
+    },
+    {
+      "epoch": 12.897299792291715,
+      "grad_norm": 3.154296875,
+      "learning_rate": 6.864576384256172e-06,
+      "loss": 1.1147,
+      "step": 13971
+    },
+    {
+      "epoch": 12.898222940226171,
+      "grad_norm": 3.337890625,
+      "learning_rate": 6.8579052701801206e-06,
+      "loss": 1.1896,
+      "step": 13972
+    },
+    {
+      "epoch": 12.899146088160627,
+      "grad_norm": 3.041015625,
+      "learning_rate": 6.851234156104069e-06,
+      "loss": 1.4836,
+      "step": 13973
+    },
+    {
+      "epoch": 12.900069236095085,
+      "grad_norm": 3.05859375,
+      "learning_rate": 6.84456304202802e-06,
+      "loss": 0.8144,
+      "step": 13974
+    },
+    {
+      "epoch": 12.900992384029541,
+      "grad_norm": 3.240234375,
+      "learning_rate": 6.837891927951969e-06,
+      "loss": 1.4375,
+      "step": 13975
+    },
+    {
+      "epoch": 12.901915531963997,
+      "grad_norm": 3.30078125,
+      "learning_rate": 6.8312208138759175e-06,
+      "loss": 1.1467,
+      "step": 13976
+    },
+    {
+      "epoch": 12.902838679898453,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.824549699799867e-06,
+      "loss": 0.9631,
+      "step": 13977
+    },
+    {
+      "epoch": 12.903761827832911,
+      "grad_norm": 2.623046875,
+      "learning_rate": 6.817878585723816e-06,
+      "loss": 1.0717,
+      "step": 13978
+    },
+    {
+      "epoch": 12.904684975767367,
+      "grad_norm": 3.41796875,
+      "learning_rate": 6.811207471647766e-06,
+      "loss": 1.2966,
+      "step": 13979
+    },
+    {
+      "epoch": 12.905608123701823,
+      "grad_norm": 3.189453125,
+      "learning_rate": 6.8045363575717145e-06,
+      "loss": 1.3269,
+      "step": 13980
+    },
+    {
+      "epoch": 12.906531271636279,
+      "grad_norm": 3.173828125,
+      "learning_rate": 6.797865243495664e-06,
+      "loss": 1.1646,
+      "step": 13981
+    },
+    {
+      "epoch": 12.907454419570737,
+      "grad_norm": 3.619140625,
+      "learning_rate": 6.791194129419613e-06,
+      "loss": 1.4131,
+      "step": 13982
+    },
+    {
+      "epoch": 12.908377567505193,
+      "grad_norm": 3.134765625,
+      "learning_rate": 6.784523015343562e-06,
+      "loss": 1.3152,
+      "step": 13983
+    },
+    {
+      "epoch": 12.909300715439649,
+      "grad_norm": 3.25390625,
+      "learning_rate": 6.777851901267512e-06,
+      "loss": 1.1836,
+      "step": 13984
+    },
+    {
+      "epoch": 12.910223863374105,
+      "grad_norm": 3.236328125,
+      "learning_rate": 6.771180787191461e-06,
+      "loss": 1.0928,
+      "step": 13985
+    },
+    {
+      "epoch": 12.911147011308563,
+      "grad_norm": 3.3046875,
+      "learning_rate": 6.76450967311541e-06,
+      "loss": 1.0861,
+      "step": 13986
+    },
+    {
+      "epoch": 12.912070159243019,
+      "grad_norm": 2.966796875,
+      "learning_rate": 6.7578385590393605e-06,
+      "loss": 0.9614,
+      "step": 13987
+    },
+    {
+      "epoch": 12.912993307177475,
+      "grad_norm": 3.212890625,
+      "learning_rate": 6.751167444963309e-06,
+      "loss": 1.1083,
+      "step": 13988
+    },
+    {
+      "epoch": 12.91391645511193,
+      "grad_norm": 3.15234375,
+      "learning_rate": 6.744496330887258e-06,
+      "loss": 1.4761,
+      "step": 13989
+    },
+    {
+      "epoch": 12.914839603046389,
+      "grad_norm": 3.07421875,
+      "learning_rate": 6.737825216811209e-06,
+      "loss": 1.1073,
+      "step": 13990
+    },
+    {
+      "epoch": 12.915762750980845,
+      "grad_norm": 3.31640625,
+      "learning_rate": 6.7311541027351574e-06,
+      "loss": 1.363,
+      "step": 13991
+    },
+    {
+      "epoch": 12.9166858989153,
+      "grad_norm": 3.033203125,
+      "learning_rate": 6.724482988659106e-06,
+      "loss": 1.3875,
+      "step": 13992
+    },
+    {
+      "epoch": 12.917609046849758,
+      "grad_norm": 2.85546875,
+      "learning_rate": 6.717811874583056e-06,
+      "loss": 1.3577,
+      "step": 13993
+    },
+    {
+      "epoch": 12.918532194784214,
+      "grad_norm": 3.16015625,
+      "learning_rate": 6.711140760507005e-06,
+      "loss": 1.1888,
+      "step": 13994
+    },
+    {
+      "epoch": 12.91945534271867,
+      "grad_norm": 3.8203125,
+      "learning_rate": 6.7044696464309536e-06,
+      "loss": 1.501,
+      "step": 13995
+    },
+    {
+      "epoch": 12.920378490653126,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.697798532354904e-06,
+      "loss": 1.1542,
+      "step": 13996
+    },
+    {
+      "epoch": 12.921301638587584,
+      "grad_norm": 2.935546875,
+      "learning_rate": 6.691127418278853e-06,
+      "loss": 1.4343,
+      "step": 13997
+    },
+    {
+      "epoch": 12.92222478652204,
+      "grad_norm": 2.830078125,
+      "learning_rate": 6.684456304202802e-06,
+      "loss": 1.2476,
+      "step": 13998
+    },
+    {
+      "epoch": 12.923147934456496,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.677785190126752e-06,
+      "loss": 1.3735,
+      "step": 13999
+    },
+    {
+      "epoch": 12.924071082390952,
+      "grad_norm": 3.08203125,
+      "learning_rate": 6.671114076050701e-06,
+      "loss": 1.0608,
+      "step": 14000
+    },
+    {
+      "epoch": 12.924071082390952,
+      "eval_loss": 0.8701171875,
+      "eval_runtime": 23.3623,
+      "eval_samples_per_second": 975.974,
+      "eval_steps_per_second": 61.039,
+      "step": 14000
+    },
+    {
+      "epoch": 12.92499423032541,
+      "grad_norm": 3.087890625,
+      "learning_rate": 6.66444296197465e-06,
+      "loss": 1.439,
+      "step": 14001
+    },
+    {
+      "epoch": 12.925917378259866,
+      "grad_norm": 3.263671875,
+      "learning_rate": 6.657771847898599e-06,
+      "loss": 1.2646,
+      "step": 14002
+    },
+    {
+      "epoch": 12.926840526194322,
+      "grad_norm": 3.5859375,
+      "learning_rate": 6.651100733822549e-06,
+      "loss": 1.3738,
+      "step": 14003
+    },
+    {
+      "epoch": 12.92776367412878,
+      "grad_norm": 3.201171875,
+      "learning_rate": 6.644429619746498e-06,
+      "loss": 1.249,
+      "step": 14004
+    },
+    {
+      "epoch": 12.928686822063236,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.637758505670447e-06,
+      "loss": 1.5774,
+      "step": 14005
+    },
+    {
+      "epoch": 12.929609969997692,
+      "grad_norm": 3.255859375,
+      "learning_rate": 6.6310873915943965e-06,
+      "loss": 1.4473,
+      "step": 14006
+    },
+    {
+      "epoch": 12.930533117932148,
+      "grad_norm": 3.279296875,
+      "learning_rate": 6.624416277518346e-06,
+      "loss": 1.2136,
+      "step": 14007
+    },
+    {
+      "epoch": 12.931456265866606,
+      "grad_norm": 3.07421875,
+      "learning_rate": 6.617745163442295e-06,
+      "loss": 1.1882,
+      "step": 14008
+    },
+    {
+      "epoch": 12.932379413801062,
+      "grad_norm": 3.146484375,
+      "learning_rate": 6.611074049366245e-06,
+      "loss": 1.3142,
+      "step": 14009
+    },
+    {
+      "epoch": 12.933302561735518,
+      "grad_norm": 3.17578125,
+      "learning_rate": 6.6044029352901935e-06,
+      "loss": 1.3191,
+      "step": 14010
+    },
+    {
+      "epoch": 12.934225709669974,
+      "grad_norm": 3.158203125,
+      "learning_rate": 6.597731821214142e-06,
+      "loss": 1.1969,
+      "step": 14011
+    },
+    {
+      "epoch": 12.935148857604432,
+      "grad_norm": 2.845703125,
+      "learning_rate": 6.591060707138093e-06,
+      "loss": 1.2195,
+      "step": 14012
+    },
+    {
+      "epoch": 12.936072005538888,
+      "grad_norm": 3.431640625,
+      "learning_rate": 6.584389593062042e-06,
+      "loss": 1.2114,
+      "step": 14013
+    },
+    {
+      "epoch": 12.936995153473344,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.5777184789859904e-06,
+      "loss": 1.4673,
+      "step": 14014
+    },
+    {
+      "epoch": 12.9379183014078,
+      "grad_norm": 3.798828125,
+      "learning_rate": 6.571047364909941e-06,
+      "loss": 1.3527,
+      "step": 14015
+    },
+    {
+      "epoch": 12.938841449342258,
+      "grad_norm": 3.33203125,
+      "learning_rate": 6.56437625083389e-06,
+      "loss": 1.2244,
+      "step": 14016
+    },
+    {
+      "epoch": 12.939764597276714,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.557705136757839e-06,
+      "loss": 1.2422,
+      "step": 14017
+    },
+    {
+      "epoch": 12.94068774521117,
+      "grad_norm": 3.056640625,
+      "learning_rate": 6.551034022681789e-06,
+      "loss": 1.1012,
+      "step": 14018
+    },
+    {
+      "epoch": 12.941610893145626,
+      "grad_norm": 3.189453125,
+      "learning_rate": 6.544362908605738e-06,
+      "loss": 1.3601,
+      "step": 14019
+    },
+    {
+      "epoch": 12.942534041080084,
+      "grad_norm": 2.970703125,
+      "learning_rate": 6.537691794529687e-06,
+      "loss": 1.3599,
+      "step": 14020
+    },
+    {
+      "epoch": 12.94345718901454,
+      "grad_norm": 3.39453125,
+      "learning_rate": 6.531020680453636e-06,
+      "loss": 1.2042,
+      "step": 14021
+    },
+    {
+      "epoch": 12.944380336948996,
+      "grad_norm": 3.23046875,
+      "learning_rate": 6.524349566377585e-06,
+      "loss": 1.3828,
+      "step": 14022
+    },
+    {
+      "epoch": 12.945303484883453,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.517678452301534e-06,
+      "loss": 1.4043,
+      "step": 14023
+    },
+    {
+      "epoch": 12.94622663281791,
+      "grad_norm": 3.126953125,
+      "learning_rate": 6.5110073382254846e-06,
+      "loss": 1.3435,
+      "step": 14024
+    },
+    {
+      "epoch": 12.947149780752365,
+      "grad_norm": 2.95703125,
+      "learning_rate": 6.504336224149433e-06,
+      "loss": 1.418,
+      "step": 14025
+    },
+    {
+      "epoch": 12.948072928686821,
+      "grad_norm": 2.98046875,
+      "learning_rate": 6.497665110073382e-06,
+      "loss": 1.3195,
+      "step": 14026
+    },
+    {
+      "epoch": 12.94899607662128,
+      "grad_norm": 3.208984375,
+      "learning_rate": 6.490993995997331e-06,
+      "loss": 1.2775,
+      "step": 14027
+    },
+    {
+      "epoch": 12.949919224555735,
+      "grad_norm": 2.888671875,
+      "learning_rate": 6.4843228819212815e-06,
+      "loss": 1.5544,
+      "step": 14028
+    },
+    {
+      "epoch": 12.950842372490191,
+      "grad_norm": 3.189453125,
+      "learning_rate": 6.47765176784523e-06,
+      "loss": 1.3369,
+      "step": 14029
+    },
+    {
+      "epoch": 12.951765520424647,
+      "grad_norm": 3.1015625,
+      "learning_rate": 6.470980653769179e-06,
+      "loss": 1.3347,
+      "step": 14030
+    },
+    {
+      "epoch": 12.952688668359105,
+      "grad_norm": 2.93359375,
+      "learning_rate": 6.46430953969313e-06,
+      "loss": 1.1353,
+      "step": 14031
+    },
+    {
+      "epoch": 12.953611816293561,
+      "grad_norm": 3.025390625,
+      "learning_rate": 6.4576384256170785e-06,
+      "loss": 1.1499,
+      "step": 14032
+    },
+    {
+      "epoch": 12.954534964228017,
+      "grad_norm": 3.58203125,
+      "learning_rate": 6.450967311541027e-06,
+      "loss": 1.1951,
+      "step": 14033
+    },
+    {
+      "epoch": 12.955458112162473,
+      "grad_norm": 3.177734375,
+      "learning_rate": 6.444296197464977e-06,
+      "loss": 1.2444,
+      "step": 14034
+    },
+    {
+      "epoch": 12.956381260096931,
+      "grad_norm": 3.400390625,
+      "learning_rate": 6.437625083388927e-06,
+      "loss": 1.4641,
+      "step": 14035
+    },
+    {
+      "epoch": 12.957304408031387,
+      "grad_norm": 3.455078125,
+      "learning_rate": 6.4309539693128755e-06,
+      "loss": 1.2104,
+      "step": 14036
+    },
+    {
+      "epoch": 12.958227555965843,
+      "grad_norm": 3.220703125,
+      "learning_rate": 6.424282855236825e-06,
+      "loss": 0.8189,
+      "step": 14037
+    },
+    {
+      "epoch": 12.959150703900299,
+      "grad_norm": 3.65234375,
+      "learning_rate": 6.417611741160774e-06,
+      "loss": 1.061,
+      "step": 14038
+    },
+    {
+      "epoch": 12.960073851834757,
+      "grad_norm": 2.845703125,
+      "learning_rate": 6.410940627084723e-06,
+      "loss": 0.8507,
+      "step": 14039
+    },
+    {
+      "epoch": 12.960996999769213,
+      "grad_norm": 3.388671875,
+      "learning_rate": 6.404269513008673e-06,
+      "loss": 1.0255,
+      "step": 14040
+    },
+    {
+      "epoch": 12.961920147703669,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.397598398932622e-06,
+      "loss": 1.4792,
+      "step": 14041
+    },
+    {
+      "epoch": 12.962843295638127,
+      "grad_norm": 2.990234375,
+      "learning_rate": 6.390927284856571e-06,
+      "loss": 1.1676,
+      "step": 14042
+    },
+    {
+      "epoch": 12.963766443572583,
+      "grad_norm": 2.91796875,
+      "learning_rate": 6.3842561707805214e-06,
+      "loss": 1.2424,
+      "step": 14043
+    },
+    {
+      "epoch": 12.964689591507039,
+      "grad_norm": 3.15234375,
+      "learning_rate": 6.37758505670447e-06,
+      "loss": 1.1562,
+      "step": 14044
+    },
+    {
+      "epoch": 12.965612739441495,
+      "grad_norm": 3.65234375,
+      "learning_rate": 6.370913942628419e-06,
+      "loss": 1.3315,
+      "step": 14045
+    },
+    {
+      "epoch": 12.966535887375953,
+      "grad_norm": 3.15234375,
+      "learning_rate": 6.36424282855237e-06,
+      "loss": 1.1573,
+      "step": 14046
+    },
+    {
+      "epoch": 12.967459035310409,
+      "grad_norm": 3.419921875,
+      "learning_rate": 6.357571714476318e-06,
+      "loss": 1.2549,
+      "step": 14047
+    },
+    {
+      "epoch": 12.968382183244865,
+      "grad_norm": 3.49609375,
+      "learning_rate": 6.350900600400267e-06,
+      "loss": 1.3916,
+      "step": 14048
+    },
+    {
+      "epoch": 12.96930533117932,
+      "grad_norm": 3.19140625,
+      "learning_rate": 6.344229486324216e-06,
+      "loss": 1.3098,
+      "step": 14049
+    },
+    {
+      "epoch": 12.970228479113779,
+      "grad_norm": 3.076171875,
+      "learning_rate": 6.337558372248166e-06,
+      "loss": 1.3789,
+      "step": 14050
+    },
+    {
+      "epoch": 12.971151627048235,
+      "grad_norm": 3.08203125,
+      "learning_rate": 6.3308872581721145e-06,
+      "loss": 1.3899,
+      "step": 14051
+    },
+    {
+      "epoch": 12.97207477498269,
+      "grad_norm": 2.685546875,
+      "learning_rate": 6.324216144096063e-06,
+      "loss": 1.3296,
+      "step": 14052
+    },
+    {
+      "epoch": 12.972997922917148,
+      "grad_norm": 2.97265625,
+      "learning_rate": 6.317545030020014e-06,
+      "loss": 1.3948,
+      "step": 14053
+    },
+    {
+      "epoch": 12.973921070851604,
+      "grad_norm": 3.08203125,
+      "learning_rate": 6.310873915943963e-06,
+      "loss": 1.3379,
+      "step": 14054
+    },
+    {
+      "epoch": 12.97484421878606,
+      "grad_norm": 3.298828125,
+      "learning_rate": 6.3042028018679115e-06,
+      "loss": 1.4133,
+      "step": 14055
+    },
+    {
+      "epoch": 12.975767366720516,
+      "grad_norm": 3.478515625,
+      "learning_rate": 6.297531687791862e-06,
+      "loss": 1.2668,
+      "step": 14056
+    },
+    {
+      "epoch": 12.976690514654974,
+      "grad_norm": 3.3671875,
+      "learning_rate": 6.290860573715811e-06,
+      "loss": 1.4202,
+      "step": 14057
+    },
+    {
+      "epoch": 12.97761366258943,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.28418945963976e-06,
+      "loss": 1.3379,
+      "step": 14058
+    },
+    {
+      "epoch": 12.978536810523886,
+      "grad_norm": 2.89453125,
+      "learning_rate": 6.27751834556371e-06,
+      "loss": 1.5452,
+      "step": 14059
+    },
+    {
+      "epoch": 12.979459958458342,
+      "grad_norm": 3.017578125,
+      "learning_rate": 6.270847231487659e-06,
+      "loss": 1.0731,
+      "step": 14060
+    },
+    {
+      "epoch": 12.9803831063928,
+      "grad_norm": 3.3984375,
+      "learning_rate": 6.264176117411608e-06,
+      "loss": 1.2576,
+      "step": 14061
+    },
+    {
+      "epoch": 12.981306254327256,
+      "grad_norm": 2.9921875,
+      "learning_rate": 6.2575050033355575e-06,
+      "loss": 1.1988,
+      "step": 14062
+    },
+    {
+      "epoch": 12.982229402261712,
+      "grad_norm": 3.064453125,
+      "learning_rate": 6.250833889259506e-06,
+      "loss": 1.3428,
+      "step": 14063
+    },
+    {
+      "epoch": 12.983152550196168,
+      "grad_norm": 2.99609375,
+      "learning_rate": 6.244162775183456e-06,
+      "loss": 1.3301,
+      "step": 14064
+    },
+    {
+      "epoch": 12.984075698130626,
+      "grad_norm": 3.15234375,
+      "learning_rate": 6.237491661107405e-06,
+      "loss": 1.0721,
+      "step": 14065
+    },
+    {
+      "epoch": 12.984998846065082,
+      "grad_norm": 3.158203125,
+      "learning_rate": 6.2308205470313544e-06,
+      "loss": 1.3158,
+      "step": 14066
+    },
+    {
+      "epoch": 12.985921993999538,
+      "grad_norm": 3.38671875,
+      "learning_rate": 6.224149432955303e-06,
+      "loss": 1.2434,
+      "step": 14067
+    },
+    {
+      "epoch": 12.986845141933994,
+      "grad_norm": 3.134765625,
+      "learning_rate": 6.217478318879253e-06,
+      "loss": 1.3113,
+      "step": 14068
+    },
+    {
+      "epoch": 12.987768289868452,
+      "grad_norm": 3.685546875,
+      "learning_rate": 6.2108072048032026e-06,
+      "loss": 1.2991,
+      "step": 14069
+    },
+    {
+      "epoch": 12.988691437802908,
+      "grad_norm": 3.49609375,
+      "learning_rate": 6.204136090727151e-06,
+      "loss": 1.448,
+      "step": 14070
+    },
+    {
+      "epoch": 12.989614585737364,
+      "grad_norm": 3.322265625,
+      "learning_rate": 6.197464976651101e-06,
+      "loss": 1.3594,
+      "step": 14071
+    },
+    {
+      "epoch": 12.990537733671822,
+      "grad_norm": 3.19921875,
+      "learning_rate": 6.190793862575051e-06,
+      "loss": 1.3391,
+      "step": 14072
+    },
+    {
+      "epoch": 12.991460881606278,
+      "grad_norm": 3.35546875,
+      "learning_rate": 6.1841227484989995e-06,
+      "loss": 1.3584,
+      "step": 14073
+    },
+    {
+      "epoch": 12.992384029540734,
+      "grad_norm": 3.291015625,
+      "learning_rate": 6.177451634422949e-06,
+      "loss": 1.2708,
+      "step": 14074
+    },
+    {
+      "epoch": 12.99330717747519,
+      "grad_norm": 3.220703125,
+      "learning_rate": 6.170780520346899e-06,
+      "loss": 1.4319,
+      "step": 14075
+    },
+    {
+      "epoch": 12.994230325409648,
+      "grad_norm": 3.2734375,
+      "learning_rate": 6.164109406270848e-06,
+      "loss": 1.1116,
+      "step": 14076
+    },
+    {
+      "epoch": 12.995153473344104,
+      "grad_norm": 3.44921875,
+      "learning_rate": 6.1574382921947965e-06,
+      "loss": 1.3391,
+      "step": 14077
+    },
+    {
+      "epoch": 12.99607662127856,
+      "grad_norm": 3.28515625,
+      "learning_rate": 6.150767178118746e-06,
+      "loss": 1.3992,
+      "step": 14078
+    },
+    {
+      "epoch": 12.996999769213016,
+      "grad_norm": 3.107421875,
+      "learning_rate": 6.144096064042695e-06,
+      "loss": 1.2013,
+      "step": 14079
+    },
+    {
+      "epoch": 12.997922917147473,
+      "grad_norm": 3.396484375,
+      "learning_rate": 6.137424949966645e-06,
+      "loss": 1.2156,
+      "step": 14080
+    },
+    {
+      "epoch": 12.99884606508193,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.1307538358905935e-06,
+      "loss": 1.0871,
+      "step": 14081
+    },
+    {
+      "epoch": 12.999769213016386,
+      "grad_norm": 3.53515625,
+      "learning_rate": 6.124082721814543e-06,
+      "loss": 1.3403,
+      "step": 14082
+    },
+    {
+      "epoch": 13.000692360950842,
+      "grad_norm": 3.01953125,
+      "learning_rate": 6.117411607738493e-06,
+      "loss": 0.8462,
+      "step": 14083
+    },
+    {
+      "epoch": 13.0016155088853,
+      "grad_norm": 2.974609375,
+      "learning_rate": 6.110740493662442e-06,
+      "loss": 1.1694,
+      "step": 14084
+    },
+    {
+      "epoch": 13.002538656819755,
+      "grad_norm": 2.564453125,
+      "learning_rate": 6.104069379586391e-06,
+      "loss": 0.9765,
+      "step": 14085
+    },
+    {
+      "epoch": 13.003461804754211,
+      "grad_norm": 3.02734375,
+      "learning_rate": 6.097398265510341e-06,
+      "loss": 1.2551,
+      "step": 14086
+    },
+    {
+      "epoch": 13.00438495268867,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.09072715143429e-06,
+      "loss": 1.3048,
+      "step": 14087
+    },
+    {
+      "epoch": 13.005308100623125,
+      "grad_norm": 3.337890625,
+      "learning_rate": 6.0840560373582394e-06,
+      "loss": 1.2513,
+      "step": 14088
+    },
+    {
+      "epoch": 13.006231248557581,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.077384923282189e-06,
+      "loss": 1.0962,
+      "step": 14089
+    },
+    {
+      "epoch": 13.007154396492037,
+      "grad_norm": 3.048828125,
+      "learning_rate": 6.070713809206138e-06,
+      "loss": 1.2423,
+      "step": 14090
+    },
+    {
+      "epoch": 13.008077544426495,
+      "grad_norm": 3.125,
+      "learning_rate": 6.064042695130087e-06,
+      "loss": 1.2734,
+      "step": 14091
+    },
+    {
+      "epoch": 13.009000692360951,
+      "grad_norm": 2.98046875,
+      "learning_rate": 6.057371581054036e-06,
+      "loss": 1.2446,
+      "step": 14092
+    },
+    {
+      "epoch": 13.009923840295407,
+      "grad_norm": 2.771484375,
+      "learning_rate": 6.050700466977985e-06,
+      "loss": 1.3325,
+      "step": 14093
+    },
+    {
+      "epoch": 13.010846988229863,
+      "grad_norm": 2.974609375,
+      "learning_rate": 6.044029352901935e-06,
+      "loss": 1.2891,
+      "step": 14094
+    },
+    {
+      "epoch": 13.011770136164321,
+      "grad_norm": 3.6328125,
+      "learning_rate": 6.037358238825884e-06,
+      "loss": 1.4348,
+      "step": 14095
+    },
+    {
+      "epoch": 13.012693284098777,
+      "grad_norm": 3.083984375,
+      "learning_rate": 6.030687124749833e-06,
+      "loss": 1.1,
+      "step": 14096
+    },
+    {
+      "epoch": 13.013616432033233,
+      "grad_norm": 3.119140625,
+      "learning_rate": 6.024016010673783e-06,
+      "loss": 1.1294,
+      "step": 14097
+    },
+    {
+      "epoch": 13.014539579967689,
+      "grad_norm": 3.650390625,
+      "learning_rate": 6.017344896597732e-06,
+      "loss": 0.8315,
+      "step": 14098
+    },
+    {
+      "epoch": 13.015462727902147,
+      "grad_norm": 3.48828125,
+      "learning_rate": 6.0106737825216815e-06,
+      "loss": 1.3765,
+      "step": 14099
+    },
+    {
+      "epoch": 13.016385875836603,
+      "grad_norm": 2.9765625,
+      "learning_rate": 6.004002668445631e-06,
+      "loss": 1.3848,
+      "step": 14100
+    },
+    {
+      "epoch": 13.017309023771059,
+      "grad_norm": 3.1328125,
+      "learning_rate": 5.99733155436958e-06,
+      "loss": 1.2369,
+      "step": 14101
+    },
+    {
+      "epoch": 13.018232171705515,
+      "grad_norm": 3.333984375,
+      "learning_rate": 5.99066044029353e-06,
+      "loss": 1.2876,
+      "step": 14102
+    },
+    {
+      "epoch": 13.019155319639973,
+      "grad_norm": 3.564453125,
+      "learning_rate": 5.9839893262174785e-06,
+      "loss": 1.3086,
+      "step": 14103
+    },
+    {
+      "epoch": 13.020078467574429,
+      "grad_norm": 3.435546875,
+      "learning_rate": 5.977318212141428e-06,
+      "loss": 1.2686,
+      "step": 14104
+    },
+    {
+      "epoch": 13.021001615508885,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.970647098065377e-06,
+      "loss": 1.215,
+      "step": 14105
+    },
+    {
+      "epoch": 13.021924763443343,
+      "grad_norm": 2.869140625,
+      "learning_rate": 5.963975983989327e-06,
+      "loss": 1.1371,
+      "step": 14106
+    },
+    {
+      "epoch": 13.022847911377799,
+      "grad_norm": 3.244140625,
+      "learning_rate": 5.9573048699132755e-06,
+      "loss": 1.2869,
+      "step": 14107
+    },
+    {
+      "epoch": 13.023771059312255,
+      "grad_norm": 3.994140625,
+      "learning_rate": 5.950633755837225e-06,
+      "loss": 0.5029,
+      "step": 14108
+    },
+    {
+      "epoch": 13.02469420724671,
+      "grad_norm": 3.748046875,
+      "learning_rate": 5.943962641761174e-06,
+      "loss": 1.4148,
+      "step": 14109
+    },
+    {
+      "epoch": 13.025617355181168,
+      "grad_norm": 3.025390625,
+      "learning_rate": 5.937291527685124e-06,
+      "loss": 1.207,
+      "step": 14110
+    },
+    {
+      "epoch": 13.026540503115625,
+      "grad_norm": 3.056640625,
+      "learning_rate": 5.930620413609073e-06,
+      "loss": 1.3245,
+      "step": 14111
+    },
+    {
+      "epoch": 13.02746365105008,
+      "grad_norm": 3.271484375,
+      "learning_rate": 5.923949299533022e-06,
+      "loss": 1.2241,
+      "step": 14112
+    },
+    {
+      "epoch": 13.028386798984537,
+      "grad_norm": 3.048828125,
+      "learning_rate": 5.917278185456972e-06,
+      "loss": 0.9452,
+      "step": 14113
+    },
+    {
+      "epoch": 13.029309946918994,
+      "grad_norm": 3.44140625,
+      "learning_rate": 5.910607071380921e-06,
+      "loss": 1.3378,
+      "step": 14114
+    },
+    {
+      "epoch": 13.03023309485345,
+      "grad_norm": 3.677734375,
+      "learning_rate": 5.90393595730487e-06,
+      "loss": 1.1644,
+      "step": 14115
+    },
+    {
+      "epoch": 13.031156242787906,
+      "grad_norm": 3.009765625,
+      "learning_rate": 5.89726484322882e-06,
+      "loss": 1.1227,
+      "step": 14116
+    },
+    {
+      "epoch": 13.032079390722362,
+      "grad_norm": 3.130859375,
+      "learning_rate": 5.890593729152769e-06,
+      "loss": 1.4153,
+      "step": 14117
+    },
+    {
+      "epoch": 13.03300253865682,
+      "grad_norm": 3.599609375,
+      "learning_rate": 5.883922615076718e-06,
+      "loss": 1.4578,
+      "step": 14118
+    },
+    {
+      "epoch": 13.033925686591276,
+      "grad_norm": 2.888671875,
+      "learning_rate": 5.877251501000667e-06,
+      "loss": 1.2073,
+      "step": 14119
+    },
+    {
+      "epoch": 13.034848834525732,
+      "grad_norm": 3.458984375,
+      "learning_rate": 5.870580386924617e-06,
+      "loss": 1.0308,
+      "step": 14120
+    },
+    {
+      "epoch": 13.03577198246019,
+      "grad_norm": 3.279296875,
+      "learning_rate": 5.863909272848566e-06,
+      "loss": 1.3425,
+      "step": 14121
+    },
+    {
+      "epoch": 13.036695130394646,
+      "grad_norm": 3.0546875,
+      "learning_rate": 5.857238158772515e-06,
+      "loss": 1.343,
+      "step": 14122
+    },
+    {
+      "epoch": 13.037618278329102,
+      "grad_norm": 3.0,
+      "learning_rate": 5.850567044696464e-06,
+      "loss": 1.376,
+      "step": 14123
+    },
+    {
+      "epoch": 13.038541426263558,
+      "grad_norm": 3.4765625,
+      "learning_rate": 5.843895930620414e-06,
+      "loss": 0.9597,
+      "step": 14124
+    },
+    {
+      "epoch": 13.039464574198016,
+      "grad_norm": 2.810546875,
+      "learning_rate": 5.837224816544363e-06,
+      "loss": 1.25,
+      "step": 14125
+    },
+    {
+      "epoch": 13.040387722132472,
+      "grad_norm": 3.08984375,
+      "learning_rate": 5.830553702468312e-06,
+      "loss": 1.4443,
+      "step": 14126
+    },
+    {
+      "epoch": 13.041310870066928,
+      "grad_norm": 3.375,
+      "learning_rate": 5.823882588392262e-06,
+      "loss": 1.0783,
+      "step": 14127
+    },
+    {
+      "epoch": 13.042234018001384,
+      "grad_norm": 3.751953125,
+      "learning_rate": 5.817211474316211e-06,
+      "loss": 1.4487,
+      "step": 14128
+    },
+    {
+      "epoch": 13.043157165935842,
+      "grad_norm": 2.888671875,
+      "learning_rate": 5.8105403602401605e-06,
+      "loss": 1.3544,
+      "step": 14129
+    },
+    {
+      "epoch": 13.044080313870298,
+      "grad_norm": 2.994140625,
+      "learning_rate": 5.80386924616411e-06,
+      "loss": 1.3599,
+      "step": 14130
+    },
+    {
+      "epoch": 13.045003461804754,
+      "grad_norm": 2.810546875,
+      "learning_rate": 5.797198132088059e-06,
+      "loss": 1.4763,
+      "step": 14131
+    },
+    {
+      "epoch": 13.04592660973921,
+      "grad_norm": 3.341796875,
+      "learning_rate": 5.790527018012009e-06,
+      "loss": 1.23,
+      "step": 14132
+    },
+    {
+      "epoch": 13.046849757673668,
+      "grad_norm": 2.951171875,
+      "learning_rate": 5.7838559039359575e-06,
+      "loss": 0.7553,
+      "step": 14133
+    },
+    {
+      "epoch": 13.047772905608124,
+      "grad_norm": 3.40234375,
+      "learning_rate": 5.777184789859907e-06,
+      "loss": 1.3877,
+      "step": 14134
+    },
+    {
+      "epoch": 13.04869605354258,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.770513675783856e-06,
+      "loss": 1.3257,
+      "step": 14135
+    },
+    {
+      "epoch": 13.049619201477038,
+      "grad_norm": 2.9296875,
+      "learning_rate": 5.763842561707806e-06,
+      "loss": 1.4133,
+      "step": 14136
+    },
+    {
+      "epoch": 13.050542349411494,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.7571714476317544e-06,
+      "loss": 1.3225,
+      "step": 14137
+    },
+    {
+      "epoch": 13.05146549734595,
+      "grad_norm": 3.248046875,
+      "learning_rate": 5.750500333555704e-06,
+      "loss": 1.2632,
+      "step": 14138
+    },
+    {
+      "epoch": 13.052388645280406,
+      "grad_norm": 3.162109375,
+      "learning_rate": 5.743829219479653e-06,
+      "loss": 1.3391,
+      "step": 14139
+    },
+    {
+      "epoch": 13.053311793214863,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.737158105403603e-06,
+      "loss": 1.3909,
+      "step": 14140
+    },
+    {
+      "epoch": 13.05423494114932,
+      "grad_norm": 3.150390625,
+      "learning_rate": 5.730486991327552e-06,
+      "loss": 1.2734,
+      "step": 14141
+    },
+    {
+      "epoch": 13.055158089083776,
+      "grad_norm": 3.22265625,
+      "learning_rate": 5.723815877251501e-06,
+      "loss": 1.2181,
+      "step": 14142
+    },
+    {
+      "epoch": 13.056081237018232,
+      "grad_norm": 3.26953125,
+      "learning_rate": 5.717144763175451e-06,
+      "loss": 1.4858,
+      "step": 14143
+    },
+    {
+      "epoch": 13.05700438495269,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.7104736490994e-06,
+      "loss": 1.3369,
+      "step": 14144
+    },
+    {
+      "epoch": 13.057927532887145,
+      "grad_norm": 3.13671875,
+      "learning_rate": 5.703802535023349e-06,
+      "loss": 1.1807,
+      "step": 14145
+    },
+    {
+      "epoch": 13.058850680821601,
+      "grad_norm": 3.470703125,
+      "learning_rate": 5.697131420947299e-06,
+      "loss": 1.1907,
+      "step": 14146
+    },
+    {
+      "epoch": 13.059773828756057,
+      "grad_norm": 3.31640625,
+      "learning_rate": 5.690460306871248e-06,
+      "loss": 1.2844,
+      "step": 14147
+    },
+    {
+      "epoch": 13.060696976690515,
+      "grad_norm": 3.134765625,
+      "learning_rate": 5.683789192795197e-06,
+      "loss": 1.322,
+      "step": 14148
+    },
+    {
+      "epoch": 13.061620124624971,
+      "grad_norm": 3.15234375,
+      "learning_rate": 5.677118078719146e-06,
+      "loss": 1.3937,
+      "step": 14149
+    },
+    {
+      "epoch": 13.062543272559427,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.670446964643095e-06,
+      "loss": 1.262,
+      "step": 14150
+    },
+    {
+      "epoch": 13.063466420493883,
+      "grad_norm": 2.720703125,
+      "learning_rate": 5.663775850567045e-06,
+      "loss": 0.781,
+      "step": 14151
+    },
+    {
+      "epoch": 13.064389568428341,
+      "grad_norm": 3.09765625,
+      "learning_rate": 5.657104736490994e-06,
+      "loss": 1.201,
+      "step": 14152
+    },
+    {
+      "epoch": 13.065312716362797,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.650433622414943e-06,
+      "loss": 1.5503,
+      "step": 14153
+    },
+    {
+      "epoch": 13.066235864297253,
+      "grad_norm": 3.556640625,
+      "learning_rate": 5.643762508338893e-06,
+      "loss": 1.1019,
+      "step": 14154
+    },
+    {
+      "epoch": 13.067159012231711,
+      "grad_norm": 3.580078125,
+      "learning_rate": 5.6370913942628425e-06,
+      "loss": 1.139,
+      "step": 14155
+    },
+    {
+      "epoch": 13.068082160166167,
+      "grad_norm": 2.86328125,
+      "learning_rate": 5.630420280186791e-06,
+      "loss": 1.3433,
+      "step": 14156
+    },
+    {
+      "epoch": 13.069005308100623,
+      "grad_norm": 3.009765625,
+      "learning_rate": 5.623749166110741e-06,
+      "loss": 1.2061,
+      "step": 14157
+    },
+    {
+      "epoch": 13.069928456035079,
+      "grad_norm": 3.208984375,
+      "learning_rate": 5.617078052034691e-06,
+      "loss": 1.1156,
+      "step": 14158
+    },
+    {
+      "epoch": 13.070851603969537,
+      "grad_norm": 3.556640625,
+      "learning_rate": 5.6104069379586395e-06,
+      "loss": 1.1097,
+      "step": 14159
+    },
+    {
+      "epoch": 13.071774751903993,
+      "grad_norm": 3.11328125,
+      "learning_rate": 5.603735823882589e-06,
+      "loss": 1.3345,
+      "step": 14160
+    },
+    {
+      "epoch": 13.072697899838449,
+      "grad_norm": 3.388671875,
+      "learning_rate": 5.597064709806538e-06,
+      "loss": 1.168,
+      "step": 14161
+    },
+    {
+      "epoch": 13.073621047772905,
+      "grad_norm": 3.083984375,
+      "learning_rate": 5.590393595730488e-06,
+      "loss": 1.3298,
+      "step": 14162
+    },
+    {
+      "epoch": 13.074544195707363,
+      "grad_norm": 3.349609375,
+      "learning_rate": 5.583722481654436e-06,
+      "loss": 1.4197,
+      "step": 14163
+    },
+    {
+      "epoch": 13.075467343641819,
+      "grad_norm": 3.341796875,
+      "learning_rate": 5.577051367578385e-06,
+      "loss": 1.2817,
+      "step": 14164
+    },
+    {
+      "epoch": 13.076390491576275,
+      "grad_norm": 2.740234375,
+      "learning_rate": 5.570380253502335e-06,
+      "loss": 1.4814,
+      "step": 14165
+    },
+    {
+      "epoch": 13.07731363951073,
+      "grad_norm": 3.330078125,
+      "learning_rate": 5.5637091394262846e-06,
+      "loss": 1.2067,
+      "step": 14166
+    },
+    {
+      "epoch": 13.078236787445189,
+      "grad_norm": 3.7421875,
+      "learning_rate": 5.557038025350233e-06,
+      "loss": 1.1686,
+      "step": 14167
+    },
+    {
+      "epoch": 13.079159935379645,
+      "grad_norm": 3.53515625,
+      "learning_rate": 5.550366911274183e-06,
+      "loss": 1.2349,
+      "step": 14168
+    },
+    {
+      "epoch": 13.0800830833141,
+      "grad_norm": 2.98046875,
+      "learning_rate": 5.543695797198133e-06,
+      "loss": 1.4707,
+      "step": 14169
+    },
+    {
+      "epoch": 13.081006231248558,
+      "grad_norm": 3.416015625,
+      "learning_rate": 5.5370246831220815e-06,
+      "loss": 1.2905,
+      "step": 14170
+    },
+    {
+      "epoch": 13.081929379183014,
+      "grad_norm": 3.43359375,
+      "learning_rate": 5.530353569046031e-06,
+      "loss": 1.1464,
+      "step": 14171
+    },
+    {
+      "epoch": 13.08285252711747,
+      "grad_norm": 3.267578125,
+      "learning_rate": 5.52368245496998e-06,
+      "loss": 1.3958,
+      "step": 14172
+    },
+    {
+      "epoch": 13.083775675051927,
+      "grad_norm": 3.189453125,
+      "learning_rate": 5.51701134089393e-06,
+      "loss": 1.354,
+      "step": 14173
+    },
+    {
+      "epoch": 13.084698822986384,
+      "grad_norm": 3.19140625,
+      "learning_rate": 5.510340226817879e-06,
+      "loss": 1.5889,
+      "step": 14174
+    },
+    {
+      "epoch": 13.08562197092084,
+      "grad_norm": 3.17578125,
+      "learning_rate": 5.503669112741828e-06,
+      "loss": 1.2371,
+      "step": 14175
+    },
+    {
+      "epoch": 13.086545118855296,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.496997998665777e-06,
+      "loss": 1.4072,
+      "step": 14176
+    },
+    {
+      "epoch": 13.087468266789752,
+      "grad_norm": 3.67578125,
+      "learning_rate": 5.490326884589727e-06,
+      "loss": 1.2673,
+      "step": 14177
+    },
+    {
+      "epoch": 13.08839141472421,
+      "grad_norm": 3.30078125,
+      "learning_rate": 5.4836557705136755e-06,
+      "loss": 1.4177,
+      "step": 14178
+    },
+    {
+      "epoch": 13.089314562658666,
+      "grad_norm": 3.205078125,
+      "learning_rate": 5.476984656437625e-06,
+      "loss": 1.3953,
+      "step": 14179
+    },
+    {
+      "epoch": 13.090237710593122,
+      "grad_norm": 3.125,
+      "learning_rate": 5.470313542361575e-06,
+      "loss": 1.4834,
+      "step": 14180
+    },
+    {
+      "epoch": 13.091160858527578,
+      "grad_norm": 3.423828125,
+      "learning_rate": 5.463642428285524e-06,
+      "loss": 1.5676,
+      "step": 14181
+    },
+    {
+      "epoch": 13.092084006462036,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.456971314209473e-06,
+      "loss": 1.3525,
+      "step": 14182
+    },
+    {
+      "epoch": 13.093007154396492,
+      "grad_norm": 2.818359375,
+      "learning_rate": 5.450300200133423e-06,
+      "loss": 1.007,
+      "step": 14183
+    },
+    {
+      "epoch": 13.093930302330948,
+      "grad_norm": 3.4296875,
+      "learning_rate": 5.443629086057372e-06,
+      "loss": 1.3259,
+      "step": 14184
+    },
+    {
+      "epoch": 13.094853450265404,
+      "grad_norm": 3.138671875,
+      "learning_rate": 5.4369579719813214e-06,
+      "loss": 1.3767,
+      "step": 14185
+    },
+    {
+      "epoch": 13.095776598199862,
+      "grad_norm": 3.71484375,
+      "learning_rate": 5.43028685790527e-06,
+      "loss": 1.3213,
+      "step": 14186
+    },
+    {
+      "epoch": 13.096699746134318,
+      "grad_norm": 2.82421875,
+      "learning_rate": 5.42361574382922e-06,
+      "loss": 1.5361,
+      "step": 14187
+    },
+    {
+      "epoch": 13.097622894068774,
+      "grad_norm": 3.455078125,
+      "learning_rate": 5.41694462975317e-06,
+      "loss": 1.2118,
+      "step": 14188
+    },
+    {
+      "epoch": 13.098546042003232,
+      "grad_norm": 3.12109375,
+      "learning_rate": 5.410273515677118e-06,
+      "loss": 1.2128,
+      "step": 14189
+    },
+    {
+      "epoch": 13.099469189937688,
+      "grad_norm": 3.025390625,
+      "learning_rate": 5.403602401601067e-06,
+      "loss": 1.4187,
+      "step": 14190
+    },
+    {
+      "epoch": 13.100392337872144,
+      "grad_norm": 3.01171875,
+      "learning_rate": 5.396931287525017e-06,
+      "loss": 1.2854,
+      "step": 14191
+    },
+    {
+      "epoch": 13.1013154858066,
+      "grad_norm": 3.509765625,
+      "learning_rate": 5.390260173448966e-06,
+      "loss": 1.0776,
+      "step": 14192
+    },
+    {
+      "epoch": 13.102238633741058,
+      "grad_norm": 3.005859375,
+      "learning_rate": 5.383589059372915e-06,
+      "loss": 1.3154,
+      "step": 14193
+    },
+    {
+      "epoch": 13.103161781675514,
+      "grad_norm": 2.99609375,
+      "learning_rate": 5.376917945296865e-06,
+      "loss": 1.2446,
+      "step": 14194
+    },
+    {
+      "epoch": 13.10408492960997,
+      "grad_norm": 3.416015625,
+      "learning_rate": 5.370246831220814e-06,
+      "loss": 1.0931,
+      "step": 14195
+    },
+    {
+      "epoch": 13.105008077544426,
+      "grad_norm": 3.1328125,
+      "learning_rate": 5.3635757171447635e-06,
+      "loss": 1.2664,
+      "step": 14196
+    },
+    {
+      "epoch": 13.105931225478884,
+      "grad_norm": 3.888671875,
+      "learning_rate": 5.356904603068712e-06,
+      "loss": 1.2632,
+      "step": 14197
+    },
+    {
+      "epoch": 13.10685437341334,
+      "grad_norm": 3.15234375,
+      "learning_rate": 5.350233488992662e-06,
+      "loss": 0.9819,
+      "step": 14198
+    },
+    {
+      "epoch": 13.107777521347796,
+      "grad_norm": 2.978515625,
+      "learning_rate": 5.343562374916612e-06,
+      "loss": 1.1222,
+      "step": 14199
+    },
+    {
+      "epoch": 13.108700669282252,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.3368912608405605e-06,
+      "loss": 1.2324,
+      "step": 14200
+    },
+    {
+      "epoch": 13.10962381721671,
+      "grad_norm": 3.53515625,
+      "learning_rate": 5.33022014676451e-06,
+      "loss": 1.3546,
+      "step": 14201
+    },
+    {
+      "epoch": 13.110546965151165,
+      "grad_norm": 2.82421875,
+      "learning_rate": 5.32354903268846e-06,
+      "loss": 1.2854,
+      "step": 14202
+    },
+    {
+      "epoch": 13.111470113085621,
+      "grad_norm": 3.236328125,
+      "learning_rate": 5.316877918612409e-06,
+      "loss": 1.1295,
+      "step": 14203
+    },
+    {
+      "epoch": 13.11239326102008,
+      "grad_norm": 2.8671875,
+      "learning_rate": 5.3102068045363575e-06,
+      "loss": 1.3262,
+      "step": 14204
+    },
+    {
+      "epoch": 13.113316408954535,
+      "grad_norm": 3.490234375,
+      "learning_rate": 5.303535690460307e-06,
+      "loss": 1.3552,
+      "step": 14205
+    },
+    {
+      "epoch": 13.114239556888991,
+      "grad_norm": 3.37890625,
+      "learning_rate": 5.296864576384256e-06,
+      "loss": 1.3113,
+      "step": 14206
+    },
+    {
+      "epoch": 13.115162704823447,
+      "grad_norm": 3.1015625,
+      "learning_rate": 5.290193462308206e-06,
+      "loss": 1.1869,
+      "step": 14207
+    },
+    {
+      "epoch": 13.116085852757905,
+      "grad_norm": 3.044921875,
+      "learning_rate": 5.2835223482321544e-06,
+      "loss": 1.3125,
+      "step": 14208
+    },
+    {
+      "epoch": 13.117009000692361,
+      "grad_norm": 3.443359375,
+      "learning_rate": 5.276851234156104e-06,
+      "loss": 1.1948,
+      "step": 14209
+    },
+    {
+      "epoch": 13.117932148626817,
+      "grad_norm": 3.03515625,
+      "learning_rate": 5.270180120080054e-06,
+      "loss": 1.3174,
+      "step": 14210
+    },
+    {
+      "epoch": 13.118855296561273,
+      "grad_norm": 2.99609375,
+      "learning_rate": 5.263509006004003e-06,
+      "loss": 1.1042,
+      "step": 14211
+    },
+    {
+      "epoch": 13.119778444495731,
+      "grad_norm": 2.900390625,
+      "learning_rate": 5.256837891927952e-06,
+      "loss": 1.4297,
+      "step": 14212
+    },
+    {
+      "epoch": 13.120701592430187,
+      "grad_norm": 3.451171875,
+      "learning_rate": 5.250166777851902e-06,
+      "loss": 1.4902,
+      "step": 14213
+    },
+    {
+      "epoch": 13.121624740364643,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.243495663775851e-06,
+      "loss": 1.1559,
+      "step": 14214
+    },
+    {
+      "epoch": 13.1225478882991,
+      "grad_norm": 3.56640625,
+      "learning_rate": 5.2368245496998e-06,
+      "loss": 1.1624,
+      "step": 14215
+    },
+    {
+      "epoch": 13.123471036233557,
+      "grad_norm": 3.013671875,
+      "learning_rate": 5.23015343562375e-06,
+      "loss": 1.4382,
+      "step": 14216
+    },
+    {
+      "epoch": 13.124394184168013,
+      "grad_norm": 3.400390625,
+      "learning_rate": 5.223482321547699e-06,
+      "loss": 1.3075,
+      "step": 14217
+    },
+    {
+      "epoch": 13.125317332102469,
+      "grad_norm": 3.107421875,
+      "learning_rate": 5.216811207471648e-06,
+      "loss": 1.2097,
+      "step": 14218
+    },
+    {
+      "epoch": 13.126240480036927,
+      "grad_norm": 2.958984375,
+      "learning_rate": 5.210140093395597e-06,
+      "loss": 1.4258,
+      "step": 14219
+    },
+    {
+      "epoch": 13.127163627971383,
+      "grad_norm": 3.23828125,
+      "learning_rate": 5.203468979319546e-06,
+      "loss": 1.2034,
+      "step": 14220
+    },
+    {
+      "epoch": 13.128086775905839,
+      "grad_norm": 3.330078125,
+      "learning_rate": 5.196797865243496e-06,
+      "loss": 1.4368,
+      "step": 14221
+    },
+    {
+      "epoch": 13.129009923840295,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.190126751167445e-06,
+      "loss": 1.0023,
+      "step": 14222
+    },
+    {
+      "epoch": 13.129933071774753,
+      "grad_norm": 3.2109375,
+      "learning_rate": 5.183455637091394e-06,
+      "loss": 1.239,
+      "step": 14223
+    },
+    {
+      "epoch": 13.130856219709209,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.176784523015344e-06,
+      "loss": 1.3103,
+      "step": 14224
+    },
+    {
+      "epoch": 13.131779367643665,
+      "grad_norm": 2.873046875,
+      "learning_rate": 5.170113408939293e-06,
+      "loss": 1.2234,
+      "step": 14225
+    },
+    {
+      "epoch": 13.13270251557812,
+      "grad_norm": 2.736328125,
+      "learning_rate": 5.1634422948632425e-06,
+      "loss": 1.4087,
+      "step": 14226
+    },
+    {
+      "epoch": 13.133625663512579,
+      "grad_norm": 3.48046875,
+      "learning_rate": 5.156771180787192e-06,
+      "loss": 1.437,
+      "step": 14227
+    },
+    {
+      "epoch": 13.134548811447035,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.150100066711141e-06,
+      "loss": 1.4443,
+      "step": 14228
+    },
+    {
+      "epoch": 13.13547195938149,
+      "grad_norm": 3.68359375,
+      "learning_rate": 5.143428952635091e-06,
+      "loss": 1.1962,
+      "step": 14229
+    },
+    {
+      "epoch": 13.136395107315947,
+      "grad_norm": 3.005859375,
+      "learning_rate": 5.13675783855904e-06,
+      "loss": 1.4292,
+      "step": 14230
+    },
+    {
+      "epoch": 13.137318255250404,
+      "grad_norm": 2.955078125,
+      "learning_rate": 5.130086724482989e-06,
+      "loss": 1.3914,
+      "step": 14231
+    },
+    {
+      "epoch": 13.13824140318486,
+      "grad_norm": 3.05859375,
+      "learning_rate": 5.123415610406938e-06,
+      "loss": 1.115,
+      "step": 14232
+    },
+    {
+      "epoch": 13.139164551119316,
+      "grad_norm": 2.947265625,
+      "learning_rate": 5.116744496330888e-06,
+      "loss": 1.2705,
+      "step": 14233
+    },
+    {
+      "epoch": 13.140087699053772,
+      "grad_norm": 3.30078125,
+      "learning_rate": 5.1100733822548364e-06,
+      "loss": 0.9193,
+      "step": 14234
+    },
+    {
+      "epoch": 13.14101084698823,
+      "grad_norm": 3.3515625,
+      "learning_rate": 5.103402268178786e-06,
+      "loss": 1.1187,
+      "step": 14235
+    },
+    {
+      "epoch": 13.141933994922686,
+      "grad_norm": 2.74609375,
+      "learning_rate": 5.096731154102735e-06,
+      "loss": 1.1711,
+      "step": 14236
+    },
+    {
+      "epoch": 13.142857142857142,
+      "grad_norm": 3.36328125,
+      "learning_rate": 5.090060040026685e-06,
+      "loss": 1.408,
+      "step": 14237
+    },
+    {
+      "epoch": 13.1437802907916,
+      "grad_norm": 3.439453125,
+      "learning_rate": 5.083388925950634e-06,
+      "loss": 1.2775,
+      "step": 14238
+    },
+    {
+      "epoch": 13.144703438726056,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.076717811874583e-06,
+      "loss": 1.091,
+      "step": 14239
+    },
+    {
+      "epoch": 13.145626586660512,
+      "grad_norm": 2.900390625,
+      "learning_rate": 5.070046697798533e-06,
+      "loss": 1.3235,
+      "step": 14240
+    },
+    {
+      "epoch": 13.146549734594968,
+      "grad_norm": 3.197265625,
+      "learning_rate": 5.063375583722482e-06,
+      "loss": 1.2976,
+      "step": 14241
+    },
+    {
+      "epoch": 13.147472882529426,
+      "grad_norm": 3.1484375,
+      "learning_rate": 5.056704469646431e-06,
+      "loss": 1.3579,
+      "step": 14242
+    },
+    {
+      "epoch": 13.148396030463882,
+      "grad_norm": 2.99609375,
+      "learning_rate": 5.050033355570381e-06,
+      "loss": 1.3582,
+      "step": 14243
+    },
+    {
+      "epoch": 13.149319178398338,
+      "grad_norm": 3.09765625,
+      "learning_rate": 5.04336224149433e-06,
+      "loss": 1.4004,
+      "step": 14244
+    },
+    {
+      "epoch": 13.150242326332794,
+      "grad_norm": 3.251953125,
+      "learning_rate": 5.036691127418279e-06,
+      "loss": 1.0536,
+      "step": 14245
+    },
+    {
+      "epoch": 13.151165474267252,
+      "grad_norm": 3.017578125,
+      "learning_rate": 5.030020013342228e-06,
+      "loss": 1.0217,
+      "step": 14246
+    },
+    {
+      "epoch": 13.152088622201708,
+      "grad_norm": 3.4140625,
+      "learning_rate": 5.023348899266178e-06,
+      "loss": 1.3181,
+      "step": 14247
+    },
+    {
+      "epoch": 13.153011770136164,
+      "grad_norm": 3.3984375,
+      "learning_rate": 5.016677785190127e-06,
+      "loss": 0.9963,
+      "step": 14248
+    },
+    {
+      "epoch": 13.15393491807062,
+      "grad_norm": 2.55859375,
+      "learning_rate": 5.010006671114076e-06,
+      "loss": 1.0285,
+      "step": 14249
+    },
+    {
+      "epoch": 13.154858066005078,
+      "grad_norm": 3.04296875,
+      "learning_rate": 5.003335557038025e-06,
+      "loss": 1.2407,
+      "step": 14250
+    },
+    {
+      "epoch": 13.155781213939534,
+      "grad_norm": 3.95703125,
+      "learning_rate": 4.996664442961975e-06,
+      "loss": 1.2227,
+      "step": 14251
+    },
+    {
+      "epoch": 13.15670436187399,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.9899933288859245e-06,
+      "loss": 1.3292,
+      "step": 14252
+    },
+    {
+      "epoch": 13.157627509808448,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.983322214809873e-06,
+      "loss": 1.006,
+      "step": 14253
+    },
+    {
+      "epoch": 13.158550657742904,
+      "grad_norm": 3.6953125,
+      "learning_rate": 4.976651100733823e-06,
+      "loss": 1.2974,
+      "step": 14254
+    },
+    {
+      "epoch": 13.15947380567736,
+      "grad_norm": 2.603515625,
+      "learning_rate": 4.969979986657772e-06,
+      "loss": 1.0518,
+      "step": 14255
+    },
+    {
+      "epoch": 13.160396953611816,
+      "grad_norm": 3.208984375,
+      "learning_rate": 4.9633088725817215e-06,
+      "loss": 1.3584,
+      "step": 14256
+    },
+    {
+      "epoch": 13.161320101546274,
+      "grad_norm": 3.75390625,
+      "learning_rate": 4.956637758505671e-06,
+      "loss": 1.251,
+      "step": 14257
+    },
+    {
+      "epoch": 13.16224324948073,
+      "grad_norm": 3.44921875,
+      "learning_rate": 4.94996664442962e-06,
+      "loss": 1.2239,
+      "step": 14258
+    },
+    {
+      "epoch": 13.163166397415186,
+      "grad_norm": 3.005859375,
+      "learning_rate": 4.94329553035357e-06,
+      "loss": 1.4229,
+      "step": 14259
+    },
+    {
+      "epoch": 13.164089545349642,
+      "grad_norm": 2.88671875,
+      "learning_rate": 4.936624416277518e-06,
+      "loss": 1.1407,
+      "step": 14260
+    },
+    {
+      "epoch": 13.1650126932841,
+      "grad_norm": 2.9765625,
+      "learning_rate": 4.929953302201468e-06,
+      "loss": 1.408,
+      "step": 14261
+    },
+    {
+      "epoch": 13.165935841218555,
+      "grad_norm": 2.912109375,
+      "learning_rate": 4.923282188125417e-06,
+      "loss": 1.2842,
+      "step": 14262
+    },
+    {
+      "epoch": 13.166858989153011,
+      "grad_norm": 3.4765625,
+      "learning_rate": 4.9166110740493666e-06,
+      "loss": 1.5059,
+      "step": 14263
+    },
+    {
+      "epoch": 13.167782137087467,
+      "grad_norm": 2.982421875,
+      "learning_rate": 4.909939959973315e-06,
+      "loss": 1.5225,
+      "step": 14264
+    },
+    {
+      "epoch": 13.168705285021925,
+      "grad_norm": 2.98046875,
+      "learning_rate": 4.903268845897265e-06,
+      "loss": 1.0614,
+      "step": 14265
+    },
+    {
+      "epoch": 13.169628432956381,
+      "grad_norm": 3.248046875,
+      "learning_rate": 4.896597731821214e-06,
+      "loss": 1.1157,
+      "step": 14266
+    },
+    {
+      "epoch": 13.170551580890837,
+      "grad_norm": 2.86328125,
+      "learning_rate": 4.8899266177451635e-06,
+      "loss": 1.3821,
+      "step": 14267
+    },
+    {
+      "epoch": 13.171474728825295,
+      "grad_norm": 3.01171875,
+      "learning_rate": 4.883255503669113e-06,
+      "loss": 1.4458,
+      "step": 14268
+    },
+    {
+      "epoch": 13.172397876759751,
+      "grad_norm": 3.23046875,
+      "learning_rate": 4.876584389593062e-06,
+      "loss": 1.1949,
+      "step": 14269
+    },
+    {
+      "epoch": 13.173321024694207,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.869913275517012e-06,
+      "loss": 1.1156,
+      "step": 14270
+    },
+    {
+      "epoch": 13.174244172628663,
+      "grad_norm": 2.958984375,
+      "learning_rate": 4.863242161440961e-06,
+      "loss": 1.2321,
+      "step": 14271
+    },
+    {
+      "epoch": 13.175167320563121,
+      "grad_norm": 3.330078125,
+      "learning_rate": 4.85657104736491e-06,
+      "loss": 1.1741,
+      "step": 14272
+    },
+    {
+      "epoch": 13.176090468497577,
+      "grad_norm": 3.228515625,
+      "learning_rate": 4.84989993328886e-06,
+      "loss": 1.2756,
+      "step": 14273
+    },
+    {
+      "epoch": 13.177013616432033,
+      "grad_norm": 3.388671875,
+      "learning_rate": 4.843228819212809e-06,
+      "loss": 1.302,
+      "step": 14274
+    },
+    {
+      "epoch": 13.177936764366489,
+      "grad_norm": 2.60546875,
+      "learning_rate": 4.836557705136758e-06,
+      "loss": 0.8997,
+      "step": 14275
+    },
+    {
+      "epoch": 13.178859912300947,
+      "grad_norm": 3.248046875,
+      "learning_rate": 4.829886591060707e-06,
+      "loss": 1.3372,
+      "step": 14276
+    },
+    {
+      "epoch": 13.179783060235403,
+      "grad_norm": 3.3046875,
+      "learning_rate": 4.823215476984657e-06,
+      "loss": 1.1393,
+      "step": 14277
+    },
+    {
+      "epoch": 13.180706208169859,
+      "grad_norm": 3.083984375,
+      "learning_rate": 4.816544362908606e-06,
+      "loss": 1.1368,
+      "step": 14278
+    },
+    {
+      "epoch": 13.181629356104315,
+      "grad_norm": 2.568359375,
+      "learning_rate": 4.809873248832555e-06,
+      "loss": 0.9952,
+      "step": 14279
+    },
+    {
+      "epoch": 13.182552504038773,
+      "grad_norm": 3.23046875,
+      "learning_rate": 4.803202134756504e-06,
+      "loss": 1.5173,
+      "step": 14280
+    },
+    {
+      "epoch": 13.183475651973229,
+      "grad_norm": 3.111328125,
+      "learning_rate": 4.796531020680454e-06,
+      "loss": 1.3369,
+      "step": 14281
+    },
+    {
+      "epoch": 13.184398799907685,
+      "grad_norm": 2.884765625,
+      "learning_rate": 4.7898599066044034e-06,
+      "loss": 1.1152,
+      "step": 14282
+    },
+    {
+      "epoch": 13.18532194784214,
+      "grad_norm": 3.09765625,
+      "learning_rate": 4.783188792528352e-06,
+      "loss": 1.4216,
+      "step": 14283
+    },
+    {
+      "epoch": 13.186245095776599,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.776517678452302e-06,
+      "loss": 1.3962,
+      "step": 14284
+    },
+    {
+      "epoch": 13.187168243711055,
+      "grad_norm": 3.03515625,
+      "learning_rate": 4.769846564376252e-06,
+      "loss": 1.1045,
+      "step": 14285
+    },
+    {
+      "epoch": 13.18809139164551,
+      "grad_norm": 3.287109375,
+      "learning_rate": 4.7631754503002e-06,
+      "loss": 1.2858,
+      "step": 14286
+    },
+    {
+      "epoch": 13.189014539579968,
+      "grad_norm": 2.91015625,
+      "learning_rate": 4.75650433622415e-06,
+      "loss": 1.1907,
+      "step": 14287
+    },
+    {
+      "epoch": 13.189937687514425,
+      "grad_norm": 3.78515625,
+      "learning_rate": 4.749833222148099e-06,
+      "loss": 1.0179,
+      "step": 14288
+    },
+    {
+      "epoch": 13.19086083544888,
+      "grad_norm": 3.279296875,
+      "learning_rate": 4.743162108072048e-06,
+      "loss": 1.2645,
+      "step": 14289
+    },
+    {
+      "epoch": 13.191783983383337,
+      "grad_norm": 3.244140625,
+      "learning_rate": 4.736490993995997e-06,
+      "loss": 1.29,
+      "step": 14290
+    },
+    {
+      "epoch": 13.192707131317794,
+      "grad_norm": 3.216796875,
+      "learning_rate": 4.729819879919946e-06,
+      "loss": 1.0446,
+      "step": 14291
+    },
+    {
+      "epoch": 13.19363027925225,
+      "grad_norm": 3.322265625,
+      "learning_rate": 4.723148765843896e-06,
+      "loss": 1.1661,
+      "step": 14292
+    },
+    {
+      "epoch": 13.194553427186706,
+      "grad_norm": 2.947265625,
+      "learning_rate": 4.7164776517678455e-06,
+      "loss": 1.2627,
+      "step": 14293
+    },
+    {
+      "epoch": 13.195476575121162,
+      "grad_norm": 3.599609375,
+      "learning_rate": 4.709806537691794e-06,
+      "loss": 1.1455,
+      "step": 14294
+    },
+    {
+      "epoch": 13.19639972305562,
+      "grad_norm": 3.21484375,
+      "learning_rate": 4.703135423615744e-06,
+      "loss": 1.3041,
+      "step": 14295
+    },
+    {
+      "epoch": 13.197322870990076,
+      "grad_norm": 2.88671875,
+      "learning_rate": 4.696464309539694e-06,
+      "loss": 1.0045,
+      "step": 14296
+    },
+    {
+      "epoch": 13.198246018924532,
+      "grad_norm": 2.853515625,
+      "learning_rate": 4.6897931954636425e-06,
+      "loss": 1.3171,
+      "step": 14297
+    },
+    {
+      "epoch": 13.199169166858988,
+      "grad_norm": 3.384765625,
+      "learning_rate": 4.683122081387592e-06,
+      "loss": 1.3738,
+      "step": 14298
+    },
+    {
+      "epoch": 13.200092314793446,
+      "grad_norm": 3.173828125,
+      "learning_rate": 4.676450967311542e-06,
+      "loss": 0.787,
+      "step": 14299
+    },
+    {
+      "epoch": 13.201015462727902,
+      "grad_norm": 3.021484375,
+      "learning_rate": 4.669779853235491e-06,
+      "loss": 1.1396,
+      "step": 14300
+    },
+    {
+      "epoch": 13.201938610662358,
+      "grad_norm": 3.376953125,
+      "learning_rate": 4.66310873915944e-06,
+      "loss": 1.3019,
+      "step": 14301
+    },
+    {
+      "epoch": 13.202861758596816,
+      "grad_norm": 3.3828125,
+      "learning_rate": 4.656437625083389e-06,
+      "loss": 1.2834,
+      "step": 14302
+    },
+    {
+      "epoch": 13.203784906531272,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.649766511007338e-06,
+      "loss": 1.1804,
+      "step": 14303
+    },
+    {
+      "epoch": 13.204708054465728,
+      "grad_norm": 3.2265625,
+      "learning_rate": 4.643095396931288e-06,
+      "loss": 1.0133,
+      "step": 14304
+    },
+    {
+      "epoch": 13.205631202400184,
+      "grad_norm": 3.02734375,
+      "learning_rate": 4.6364242828552364e-06,
+      "loss": 1.3506,
+      "step": 14305
+    },
+    {
+      "epoch": 13.206554350334642,
+      "grad_norm": 3.193359375,
+      "learning_rate": 4.629753168779186e-06,
+      "loss": 1.1106,
+      "step": 14306
+    },
+    {
+      "epoch": 13.207477498269098,
+      "grad_norm": 4.015625,
+      "learning_rate": 4.623082054703136e-06,
+      "loss": 1.2812,
+      "step": 14307
+    },
+    {
+      "epoch": 13.208400646203554,
+      "grad_norm": 2.833984375,
+      "learning_rate": 4.616410940627085e-06,
+      "loss": 1.2832,
+      "step": 14308
+    },
+    {
+      "epoch": 13.20932379413801,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.609739826551034e-06,
+      "loss": 1.4878,
+      "step": 14309
+    },
+    {
+      "epoch": 13.210246942072468,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.603068712474984e-06,
+      "loss": 1.1016,
+      "step": 14310
+    },
+    {
+      "epoch": 13.211170090006924,
+      "grad_norm": 3.23828125,
+      "learning_rate": 4.596397598398933e-06,
+      "loss": 1.4692,
+      "step": 14311
+    },
+    {
+      "epoch": 13.21209323794138,
+      "grad_norm": 3.26953125,
+      "learning_rate": 4.589726484322882e-06,
+      "loss": 1.1245,
+      "step": 14312
+    },
+    {
+      "epoch": 13.213016385875836,
+      "grad_norm": 2.939453125,
+      "learning_rate": 4.583055370246831e-06,
+      "loss": 1.2582,
+      "step": 14313
+    },
+    {
+      "epoch": 13.213939533810294,
+      "grad_norm": 3.4140625,
+      "learning_rate": 4.576384256170781e-06,
+      "loss": 1.2246,
+      "step": 14314
+    },
+    {
+      "epoch": 13.21486268174475,
+      "grad_norm": 3.505859375,
+      "learning_rate": 4.5697131420947306e-06,
+      "loss": 1.1785,
+      "step": 14315
+    },
+    {
+      "epoch": 13.215785829679206,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.563042028018679e-06,
+      "loss": 1.511,
+      "step": 14316
+    },
+    {
+      "epoch": 13.216708977613663,
+      "grad_norm": 3.107421875,
+      "learning_rate": 4.556370913942628e-06,
+      "loss": 1.3629,
+      "step": 14317
+    },
+    {
+      "epoch": 13.21763212554812,
+      "grad_norm": 3.19921875,
+      "learning_rate": 4.549699799866578e-06,
+      "loss": 1.1666,
+      "step": 14318
+    },
+    {
+      "epoch": 13.218555273482576,
+      "grad_norm": 3.298828125,
+      "learning_rate": 4.543028685790527e-06,
+      "loss": 1.2622,
+      "step": 14319
+    },
+    {
+      "epoch": 13.219478421417032,
+      "grad_norm": 3.189453125,
+      "learning_rate": 4.536357571714476e-06,
+      "loss": 1.3208,
+      "step": 14320
+    },
+    {
+      "epoch": 13.22040156935149,
+      "grad_norm": 3.70703125,
+      "learning_rate": 4.529686457638426e-06,
+      "loss": 1.29,
+      "step": 14321
+    },
+    {
+      "epoch": 13.221324717285945,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.523015343562375e-06,
+      "loss": 1.1924,
+      "step": 14322
+    },
+    {
+      "epoch": 13.222247865220401,
+      "grad_norm": 3.06640625,
+      "learning_rate": 4.5163442294863245e-06,
+      "loss": 1.3916,
+      "step": 14323
+    },
+    {
+      "epoch": 13.223171013154857,
+      "grad_norm": 3.458984375,
+      "learning_rate": 4.509673115410274e-06,
+      "loss": 1.3608,
+      "step": 14324
+    },
+    {
+      "epoch": 13.224094161089315,
+      "grad_norm": 3.51171875,
+      "learning_rate": 4.503002001334223e-06,
+      "loss": 1.17,
+      "step": 14325
+    },
+    {
+      "epoch": 13.225017309023771,
+      "grad_norm": 3.056640625,
+      "learning_rate": 4.496330887258173e-06,
+      "loss": 0.9134,
+      "step": 14326
+    },
+    {
+      "epoch": 13.225940456958227,
+      "grad_norm": 2.892578125,
+      "learning_rate": 4.4896597731821215e-06,
+      "loss": 1.2302,
+      "step": 14327
+    },
+    {
+      "epoch": 13.226863604892683,
+      "grad_norm": 3.103515625,
+      "learning_rate": 4.482988659106071e-06,
+      "loss": 1.3462,
+      "step": 14328
+    },
+    {
+      "epoch": 13.227786752827141,
+      "grad_norm": 3.158203125,
+      "learning_rate": 4.476317545030021e-06,
+      "loss": 1.2303,
+      "step": 14329
+    },
+    {
+      "epoch": 13.228709900761597,
+      "grad_norm": 3.05859375,
+      "learning_rate": 4.46964643095397e-06,
+      "loss": 1.2892,
+      "step": 14330
+    },
+    {
+      "epoch": 13.229633048696053,
+      "grad_norm": 2.677734375,
+      "learning_rate": 4.4629753168779184e-06,
+      "loss": 0.9528,
+      "step": 14331
+    },
+    {
+      "epoch": 13.23055619663051,
+      "grad_norm": 3.318359375,
+      "learning_rate": 4.456304202801868e-06,
+      "loss": 1.2966,
+      "step": 14332
+    },
+    {
+      "epoch": 13.231479344564967,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.449633088725817e-06,
+      "loss": 1.4211,
+      "step": 14333
+    },
+    {
+      "epoch": 13.232402492499423,
+      "grad_norm": 2.9921875,
+      "learning_rate": 4.442961974649767e-06,
+      "loss": 1.3284,
+      "step": 14334
+    },
+    {
+      "epoch": 13.233325640433879,
+      "grad_norm": 3.15234375,
+      "learning_rate": 4.436290860573716e-06,
+      "loss": 1.2115,
+      "step": 14335
+    },
+    {
+      "epoch": 13.234248788368337,
+      "grad_norm": 3.052734375,
+      "learning_rate": 4.429619746497665e-06,
+      "loss": 1.2444,
+      "step": 14336
+    },
+    {
+      "epoch": 13.235171936302793,
+      "grad_norm": 3.107421875,
+      "learning_rate": 4.422948632421615e-06,
+      "loss": 1.2415,
+      "step": 14337
+    },
+    {
+      "epoch": 13.236095084237249,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.4162775183455635e-06,
+      "loss": 1.1445,
+      "step": 14338
+    },
+    {
+      "epoch": 13.237018232171705,
+      "grad_norm": 3.38671875,
+      "learning_rate": 4.409606404269513e-06,
+      "loss": 1.2345,
+      "step": 14339
+    },
+    {
+      "epoch": 13.237941380106163,
+      "grad_norm": 4.02734375,
+      "learning_rate": 4.402935290193463e-06,
+      "loss": 1.12,
+      "step": 14340
+    },
+    {
+      "epoch": 13.238864528040619,
+      "grad_norm": 3.853515625,
+      "learning_rate": 4.396264176117412e-06,
+      "loss": 1.0304,
+      "step": 14341
+    },
+    {
+      "epoch": 13.239787675975075,
+      "grad_norm": 3.185546875,
+      "learning_rate": 4.389593062041361e-06,
+      "loss": 1.2737,
+      "step": 14342
+    },
+    {
+      "epoch": 13.24071082390953,
+      "grad_norm": 3.525390625,
+      "learning_rate": 4.382921947965311e-06,
+      "loss": 1.0546,
+      "step": 14343
+    },
+    {
+      "epoch": 13.241633971843989,
+      "grad_norm": 3.56640625,
+      "learning_rate": 4.37625083388926e-06,
+      "loss": 1.6392,
+      "step": 14344
+    },
+    {
+      "epoch": 13.242557119778445,
+      "grad_norm": 3.11328125,
+      "learning_rate": 4.369579719813209e-06,
+      "loss": 1.2512,
+      "step": 14345
+    },
+    {
+      "epoch": 13.2434802677129,
+      "grad_norm": 2.8671875,
+      "learning_rate": 4.362908605737158e-06,
+      "loss": 1.3738,
+      "step": 14346
+    },
+    {
+      "epoch": 13.244403415647357,
+      "grad_norm": 3.337890625,
+      "learning_rate": 4.356237491661107e-06,
+      "loss": 1.3411,
+      "step": 14347
+    },
+    {
+      "epoch": 13.245326563581814,
+      "grad_norm": 3.111328125,
+      "learning_rate": 4.349566377585057e-06,
+      "loss": 1.241,
+      "step": 14348
+    },
+    {
+      "epoch": 13.24624971151627,
+      "grad_norm": 3.900390625,
+      "learning_rate": 4.342895263509006e-06,
+      "loss": 0.9414,
+      "step": 14349
+    },
+    {
+      "epoch": 13.247172859450727,
+      "grad_norm": 3.642578125,
+      "learning_rate": 4.336224149432955e-06,
+      "loss": 1.2976,
+      "step": 14350
+    },
+    {
+      "epoch": 13.248096007385184,
+      "grad_norm": 3.205078125,
+      "learning_rate": 4.329553035356905e-06,
+      "loss": 1.2529,
+      "step": 14351
+    },
+    {
+      "epoch": 13.24901915531964,
+      "grad_norm": 3.130859375,
+      "learning_rate": 4.322881921280854e-06,
+      "loss": 0.9111,
+      "step": 14352
+    },
+    {
+      "epoch": 13.249942303254096,
+      "grad_norm": 3.345703125,
+      "learning_rate": 4.3162108072048035e-06,
+      "loss": 1.2661,
+      "step": 14353
+    },
+    {
+      "epoch": 13.250865451188552,
+      "grad_norm": 3.076171875,
+      "learning_rate": 4.309539693128753e-06,
+      "loss": 1.3057,
+      "step": 14354
+    },
+    {
+      "epoch": 13.25178859912301,
+      "grad_norm": 3.447265625,
+      "learning_rate": 4.302868579052702e-06,
+      "loss": 1.0583,
+      "step": 14355
+    },
+    {
+      "epoch": 13.252711747057466,
+      "grad_norm": 3.384765625,
+      "learning_rate": 4.296197464976652e-06,
+      "loss": 1.2572,
+      "step": 14356
+    },
+    {
+      "epoch": 13.253634894991922,
+      "grad_norm": 2.970703125,
+      "learning_rate": 4.289526350900601e-06,
+      "loss": 1.3403,
+      "step": 14357
+    },
+    {
+      "epoch": 13.254558042926378,
+      "grad_norm": 3.494140625,
+      "learning_rate": 4.28285523682455e-06,
+      "loss": 1.2133,
+      "step": 14358
+    },
+    {
+      "epoch": 13.255481190860836,
+      "grad_norm": 3.4765625,
+      "learning_rate": 4.276184122748499e-06,
+      "loss": 1.2307,
+      "step": 14359
+    },
+    {
+      "epoch": 13.256404338795292,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.2695130086724486e-06,
+      "loss": 0.9991,
+      "step": 14360
+    },
+    {
+      "epoch": 13.257327486729748,
+      "grad_norm": 2.923828125,
+      "learning_rate": 4.262841894596397e-06,
+      "loss": 1.0094,
+      "step": 14361
+    },
+    {
+      "epoch": 13.258250634664204,
+      "grad_norm": 3.619140625,
+      "learning_rate": 4.256170780520347e-06,
+      "loss": 1.0415,
+      "step": 14362
+    },
+    {
+      "epoch": 13.259173782598662,
+      "grad_norm": 3.125,
+      "learning_rate": 4.249499666444296e-06,
+      "loss": 1.2449,
+      "step": 14363
+    },
+    {
+      "epoch": 13.260096930533118,
+      "grad_norm": 2.826171875,
+      "learning_rate": 4.2428285523682455e-06,
+      "loss": 1.2583,
+      "step": 14364
+    },
+    {
+      "epoch": 13.261020078467574,
+      "grad_norm": 2.91015625,
+      "learning_rate": 4.236157438292195e-06,
+      "loss": 1.2441,
+      "step": 14365
+    },
+    {
+      "epoch": 13.261943226402032,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.229486324216144e-06,
+      "loss": 1.1654,
+      "step": 14366
+    },
+    {
+      "epoch": 13.262866374336488,
+      "grad_norm": 3.04296875,
+      "learning_rate": 4.222815210140094e-06,
+      "loss": 1.1014,
+      "step": 14367
+    },
+    {
+      "epoch": 13.263789522270944,
+      "grad_norm": 2.97265625,
+      "learning_rate": 4.216144096064043e-06,
+      "loss": 1.3928,
+      "step": 14368
+    },
+    {
+      "epoch": 13.2647126702054,
+      "grad_norm": 3.125,
+      "learning_rate": 4.209472981987992e-06,
+      "loss": 1.3611,
+      "step": 14369
+    },
+    {
+      "epoch": 13.265635818139858,
+      "grad_norm": 2.9609375,
+      "learning_rate": 4.202801867911942e-06,
+      "loss": 1.1747,
+      "step": 14370
+    },
+    {
+      "epoch": 13.266558966074314,
+      "grad_norm": 2.95703125,
+      "learning_rate": 4.1961307538358915e-06,
+      "loss": 1.1935,
+      "step": 14371
+    },
+    {
+      "epoch": 13.26748211400877,
+      "grad_norm": 3.076171875,
+      "learning_rate": 4.18945963975984e-06,
+      "loss": 0.9768,
+      "step": 14372
+    },
+    {
+      "epoch": 13.268405261943226,
+      "grad_norm": 2.875,
+      "learning_rate": 4.182788525683789e-06,
+      "loss": 1.3408,
+      "step": 14373
+    },
+    {
+      "epoch": 13.269328409877684,
+      "grad_norm": 2.95703125,
+      "learning_rate": 4.176117411607739e-06,
+      "loss": 1.5742,
+      "step": 14374
+    },
+    {
+      "epoch": 13.27025155781214,
+      "grad_norm": 3.115234375,
+      "learning_rate": 4.169446297531688e-06,
+      "loss": 1.4651,
+      "step": 14375
+    },
+    {
+      "epoch": 13.271174705746596,
+      "grad_norm": 3.212890625,
+      "learning_rate": 4.162775183455637e-06,
+      "loss": 1.4089,
+      "step": 14376
+    },
+    {
+      "epoch": 13.272097853681052,
+      "grad_norm": 2.26171875,
+      "learning_rate": 4.156104069379586e-06,
+      "loss": 1.0674,
+      "step": 14377
+    },
+    {
+      "epoch": 13.27302100161551,
+      "grad_norm": 3.34765625,
+      "learning_rate": 4.149432955303536e-06,
+      "loss": 1.2241,
+      "step": 14378
+    },
+    {
+      "epoch": 13.273944149549965,
+      "grad_norm": 3.154296875,
+      "learning_rate": 4.1427618412274854e-06,
+      "loss": 1.4163,
+      "step": 14379
+    },
+    {
+      "epoch": 13.274867297484422,
+      "grad_norm": 3.197265625,
+      "learning_rate": 4.136090727151434e-06,
+      "loss": 1.4399,
+      "step": 14380
+    },
+    {
+      "epoch": 13.275790445418878,
+      "grad_norm": 3.169921875,
+      "learning_rate": 4.129419613075384e-06,
+      "loss": 1.3606,
+      "step": 14381
+    },
+    {
+      "epoch": 13.276713593353335,
+      "grad_norm": 3.26171875,
+      "learning_rate": 4.122748498999334e-06,
+      "loss": 1.1671,
+      "step": 14382
+    },
+    {
+      "epoch": 13.277636741287791,
+      "grad_norm": 3.263671875,
+      "learning_rate": 4.116077384923282e-06,
+      "loss": 0.8978,
+      "step": 14383
+    },
+    {
+      "epoch": 13.278559889222247,
+      "grad_norm": 3.19921875,
+      "learning_rate": 4.109406270847232e-06,
+      "loss": 1.1504,
+      "step": 14384
+    },
+    {
+      "epoch": 13.279483037156705,
+      "grad_norm": 3.23828125,
+      "learning_rate": 4.102735156771181e-06,
+      "loss": 1.2747,
+      "step": 14385
+    },
+    {
+      "epoch": 13.280406185091161,
+      "grad_norm": 2.857421875,
+      "learning_rate": 4.0960640426951306e-06,
+      "loss": 1.0569,
+      "step": 14386
+    },
+    {
+      "epoch": 13.281329333025617,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.089392928619079e-06,
+      "loss": 1.2454,
+      "step": 14387
+    },
+    {
+      "epoch": 13.282252480960073,
+      "grad_norm": 3.068359375,
+      "learning_rate": 4.082721814543029e-06,
+      "loss": 1.1615,
+      "step": 14388
+    },
+    {
+      "epoch": 13.283175628894531,
+      "grad_norm": 3.25,
+      "learning_rate": 4.076050700466978e-06,
+      "loss": 1.0946,
+      "step": 14389
+    },
+    {
+      "epoch": 13.284098776828987,
+      "grad_norm": 2.95703125,
+      "learning_rate": 4.0693795863909275e-06,
+      "loss": 1.4373,
+      "step": 14390
+    },
+    {
+      "epoch": 13.285021924763443,
+      "grad_norm": 3.10546875,
+      "learning_rate": 4.062708472314876e-06,
+      "loss": 1.3137,
+      "step": 14391
+    },
+    {
+      "epoch": 13.2859450726979,
+      "grad_norm": 3.001953125,
+      "learning_rate": 4.056037358238826e-06,
+      "loss": 1.2416,
+      "step": 14392
+    },
+    {
+      "epoch": 13.286868220632357,
+      "grad_norm": 3.154296875,
+      "learning_rate": 4.049366244162776e-06,
+      "loss": 1.3361,
+      "step": 14393
+    },
+    {
+      "epoch": 13.287791368566813,
+      "grad_norm": 3.208984375,
+      "learning_rate": 4.0426951300867245e-06,
+      "loss": 1.1635,
+      "step": 14394
+    },
+    {
+      "epoch": 13.288714516501269,
+      "grad_norm": 3.19140625,
+      "learning_rate": 4.036024016010674e-06,
+      "loss": 1.2329,
+      "step": 14395
+    },
+    {
+      "epoch": 13.289637664435725,
+      "grad_norm": 3.17578125,
+      "learning_rate": 4.029352901934623e-06,
+      "loss": 1.2939,
+      "step": 14396
+    },
+    {
+      "epoch": 13.290560812370183,
+      "grad_norm": 2.943359375,
+      "learning_rate": 4.022681787858573e-06,
+      "loss": 1.189,
+      "step": 14397
+    },
+    {
+      "epoch": 13.291483960304639,
+      "grad_norm": 3.5234375,
+      "learning_rate": 4.016010673782522e-06,
+      "loss": 1.4265,
+      "step": 14398
+    },
+    {
+      "epoch": 13.292407108239095,
+      "grad_norm": 3.04296875,
+      "learning_rate": 4.009339559706471e-06,
+      "loss": 1.0726,
+      "step": 14399
+    },
+    {
+      "epoch": 13.293330256173551,
+      "grad_norm": 3.291015625,
+      "learning_rate": 4.002668445630421e-06,
+      "loss": 1.302,
+      "step": 14400
+    },
+    {
+      "epoch": 13.294253404108009,
+      "grad_norm": 3.685546875,
+      "learning_rate": 3.99599733155437e-06,
+      "loss": 1.2986,
+      "step": 14401
+    },
+    {
+      "epoch": 13.295176552042465,
+      "grad_norm": 3.43359375,
+      "learning_rate": 3.989326217478319e-06,
+      "loss": 1.3499,
+      "step": 14402
+    },
+    {
+      "epoch": 13.29609969997692,
+      "grad_norm": 3.16015625,
+      "learning_rate": 3.982655103402268e-06,
+      "loss": 1.1831,
+      "step": 14403
+    },
+    {
+      "epoch": 13.297022847911379,
+      "grad_norm": 3.146484375,
+      "learning_rate": 3.975983989326218e-06,
+      "loss": 1.1611,
+      "step": 14404
+    },
+    {
+      "epoch": 13.297945995845835,
+      "grad_norm": 3.068359375,
+      "learning_rate": 3.969312875250167e-06,
+      "loss": 1.256,
+      "step": 14405
+    },
+    {
+      "epoch": 13.29886914378029,
+      "grad_norm": 3.2578125,
+      "learning_rate": 3.962641761174116e-06,
+      "loss": 1.4794,
+      "step": 14406
+    },
+    {
+      "epoch": 13.299792291714747,
+      "grad_norm": 2.8984375,
+      "learning_rate": 3.955970647098065e-06,
+      "loss": 1.4207,
+      "step": 14407
+    },
+    {
+      "epoch": 13.300715439649204,
+      "grad_norm": 3.455078125,
+      "learning_rate": 3.949299533022015e-06,
+      "loss": 1.4048,
+      "step": 14408
+    },
+    {
+      "epoch": 13.30163858758366,
+      "grad_norm": 3.224609375,
+      "learning_rate": 3.942628418945964e-06,
+      "loss": 1.2135,
+      "step": 14409
+    },
+    {
+      "epoch": 13.302561735518116,
+      "grad_norm": 3.076171875,
+      "learning_rate": 3.935957304869913e-06,
+      "loss": 1.229,
+      "step": 14410
+    },
+    {
+      "epoch": 13.303484883452573,
+      "grad_norm": 2.98828125,
+      "learning_rate": 3.929286190793863e-06,
+      "loss": 1.424,
+      "step": 14411
+    },
+    {
+      "epoch": 13.30440803138703,
+      "grad_norm": 3.18359375,
+      "learning_rate": 3.9226150767178126e-06,
+      "loss": 1.3535,
+      "step": 14412
+    },
+    {
+      "epoch": 13.305331179321486,
+      "grad_norm": 3.009765625,
+      "learning_rate": 3.915943962641761e-06,
+      "loss": 0.9666,
+      "step": 14413
+    },
+    {
+      "epoch": 13.306254327255942,
+      "grad_norm": 3.126953125,
+      "learning_rate": 3.909272848565711e-06,
+      "loss": 1.0307,
+      "step": 14414
+    },
+    {
+      "epoch": 13.3071774751904,
+      "grad_norm": 3.1171875,
+      "learning_rate": 3.90260173448966e-06,
+      "loss": 1.3333,
+      "step": 14415
+    },
+    {
+      "epoch": 13.308100623124856,
+      "grad_norm": 3.091796875,
+      "learning_rate": 3.895930620413609e-06,
+      "loss": 1.3882,
+      "step": 14416
+    },
+    {
+      "epoch": 13.309023771059312,
+      "grad_norm": 2.96484375,
+      "learning_rate": 3.889259506337558e-06,
+      "loss": 1.4666,
+      "step": 14417
+    },
+    {
+      "epoch": 13.309946918993768,
+      "grad_norm": 3.513671875,
+      "learning_rate": 3.882588392261508e-06,
+      "loss": 1.2317,
+      "step": 14418
+    },
+    {
+      "epoch": 13.310870066928226,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.875917278185457e-06,
+      "loss": 1.2559,
+      "step": 14419
+    },
+    {
+      "epoch": 13.311793214862682,
+      "grad_norm": 3.32421875,
+      "learning_rate": 3.8692461641094065e-06,
+      "loss": 1.4351,
+      "step": 14420
+    },
+    {
+      "epoch": 13.312716362797138,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.862575050033355e-06,
+      "loss": 1.1691,
+      "step": 14421
+    },
+    {
+      "epoch": 13.313639510731594,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.855903935957305e-06,
+      "loss": 1.323,
+      "step": 14422
+    },
+    {
+      "epoch": 13.314562658666052,
+      "grad_norm": 3.255859375,
+      "learning_rate": 3.849232821881255e-06,
+      "loss": 1.2523,
+      "step": 14423
+    },
+    {
+      "epoch": 13.315485806600508,
+      "grad_norm": 3.513671875,
+      "learning_rate": 3.8425617078052035e-06,
+      "loss": 1.0935,
+      "step": 14424
+    },
+    {
+      "epoch": 13.316408954534964,
+      "grad_norm": 3.24609375,
+      "learning_rate": 3.835890593729153e-06,
+      "loss": 1.2521,
+      "step": 14425
+    },
+    {
+      "epoch": 13.31733210246942,
+      "grad_norm": 2.892578125,
+      "learning_rate": 3.829219479653103e-06,
+      "loss": 1.3569,
+      "step": 14426
+    },
+    {
+      "epoch": 13.318255250403878,
+      "grad_norm": 3.00390625,
+      "learning_rate": 3.822548365577052e-06,
+      "loss": 1.3127,
+      "step": 14427
+    },
+    {
+      "epoch": 13.319178398338334,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.815877251501001e-06,
+      "loss": 1.4326,
+      "step": 14428
+    },
+    {
+      "epoch": 13.32010154627279,
+      "grad_norm": 2.990234375,
+      "learning_rate": 3.8092061374249505e-06,
+      "loss": 1.2991,
+      "step": 14429
+    },
+    {
+      "epoch": 13.321024694207246,
+      "grad_norm": 3.44140625,
+      "learning_rate": 3.8025350233488993e-06,
+      "loss": 0.9515,
+      "step": 14430
+    },
+    {
+      "epoch": 13.321947842141704,
+      "grad_norm": 3.15234375,
+      "learning_rate": 3.795863909272849e-06,
+      "loss": 1.3447,
+      "step": 14431
+    },
+    {
+      "epoch": 13.32287099007616,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.789192795196798e-06,
+      "loss": 1.1868,
+      "step": 14432
+    },
+    {
+      "epoch": 13.323794138010616,
+      "grad_norm": 3.435546875,
+      "learning_rate": 3.7825216811207475e-06,
+      "loss": 1.191,
+      "step": 14433
+    },
+    {
+      "epoch": 13.324717285945074,
+      "grad_norm": 3.22265625,
+      "learning_rate": 3.7758505670446967e-06,
+      "loss": 1.2773,
+      "step": 14434
+    },
+    {
+      "epoch": 13.32564043387953,
+      "grad_norm": 3.39453125,
+      "learning_rate": 3.7691794529686455e-06,
+      "loss": 1.3909,
+      "step": 14435
+    },
+    {
+      "epoch": 13.326563581813986,
+      "grad_norm": 3.02734375,
+      "learning_rate": 3.7625083388925952e-06,
+      "loss": 1.4468,
+      "step": 14436
+    },
+    {
+      "epoch": 13.327486729748442,
+      "grad_norm": 3.037109375,
+      "learning_rate": 3.755837224816545e-06,
+      "loss": 1.1692,
+      "step": 14437
+    },
+    {
+      "epoch": 13.3284098776829,
+      "grad_norm": 2.79296875,
+      "learning_rate": 3.7491661107404937e-06,
+      "loss": 0.9352,
+      "step": 14438
+    },
+    {
+      "epoch": 13.329333025617355,
+      "grad_norm": 3.099609375,
+      "learning_rate": 3.7424949966644434e-06,
+      "loss": 1.3945,
+      "step": 14439
+    },
+    {
+      "epoch": 13.330256173551811,
+      "grad_norm": 3.12890625,
+      "learning_rate": 3.7358238825883926e-06,
+      "loss": 1.2968,
+      "step": 14440
+    },
+    {
+      "epoch": 13.331179321486267,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.7291527685123414e-06,
+      "loss": 0.9736,
+      "step": 14441
+    },
+    {
+      "epoch": 13.332102469420725,
+      "grad_norm": 3.0234375,
+      "learning_rate": 3.722481654436291e-06,
+      "loss": 1.2905,
+      "step": 14442
+    },
+    {
+      "epoch": 13.333025617355181,
+      "grad_norm": 3.392578125,
+      "learning_rate": 3.71581054036024e-06,
+      "loss": 1.1927,
+      "step": 14443
+    },
+    {
+      "epoch": 13.333948765289637,
+      "grad_norm": 3.208984375,
+      "learning_rate": 3.7091394262841896e-06,
+      "loss": 1.3853,
+      "step": 14444
+    },
+    {
+      "epoch": 13.334871913224093,
+      "grad_norm": 3.322265625,
+      "learning_rate": 3.7024683122081392e-06,
+      "loss": 1.1698,
+      "step": 14445
+    },
+    {
+      "epoch": 13.335795061158551,
+      "grad_norm": 3.3515625,
+      "learning_rate": 3.695797198132088e-06,
+      "loss": 1.4763,
+      "step": 14446
+    },
+    {
+      "epoch": 13.336718209093007,
+      "grad_norm": 3.625,
+      "learning_rate": 3.6891260840560377e-06,
+      "loss": 1.1415,
+      "step": 14447
+    },
+    {
+      "epoch": 13.337641357027463,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.682454969979987e-06,
+      "loss": 0.9836,
+      "step": 14448
+    },
+    {
+      "epoch": 13.33856450496192,
+      "grad_norm": 3.173828125,
+      "learning_rate": 3.6757838559039358e-06,
+      "loss": 1.386,
+      "step": 14449
+    },
+    {
+      "epoch": 13.339487652896377,
+      "grad_norm": 2.81640625,
+      "learning_rate": 3.6691127418278855e-06,
+      "loss": 1.0376,
+      "step": 14450
+    },
+    {
+      "epoch": 13.340410800830833,
+      "grad_norm": 3.130859375,
+      "learning_rate": 3.662441627751835e-06,
+      "loss": 1.3403,
+      "step": 14451
+    },
+    {
+      "epoch": 13.341333948765289,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.655770513675784e-06,
+      "loss": 1.2954,
+      "step": 14452
+    },
+    {
+      "epoch": 13.342257096699747,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.6490993995997336e-06,
+      "loss": 1.1588,
+      "step": 14453
+    },
+    {
+      "epoch": 13.343180244634203,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.6424282855236824e-06,
+      "loss": 1.3357,
+      "step": 14454
+    },
+    {
+      "epoch": 13.344103392568659,
+      "grad_norm": 3.00390625,
+      "learning_rate": 3.6357571714476317e-06,
+      "loss": 1.4075,
+      "step": 14455
+    },
+    {
+      "epoch": 13.345026540503115,
+      "grad_norm": 3.29296875,
+      "learning_rate": 3.6290860573715813e-06,
+      "loss": 0.8976,
+      "step": 14456
+    },
+    {
+      "epoch": 13.345949688437573,
+      "grad_norm": 3.341796875,
+      "learning_rate": 3.62241494329553e-06,
+      "loss": 1.3594,
+      "step": 14457
+    },
+    {
+      "epoch": 13.346872836372029,
+      "grad_norm": 3.40234375,
+      "learning_rate": 3.61574382921948e-06,
+      "loss": 1.5212,
+      "step": 14458
+    },
+    {
+      "epoch": 13.347795984306485,
+      "grad_norm": 3.50390625,
+      "learning_rate": 3.6090727151434295e-06,
+      "loss": 1.1877,
+      "step": 14459
+    },
+    {
+      "epoch": 13.34871913224094,
+      "grad_norm": 2.7890625,
+      "learning_rate": 3.6024016010673783e-06,
+      "loss": 0.9126,
+      "step": 14460
+    },
+    {
+      "epoch": 13.349642280175399,
+      "grad_norm": 3.00390625,
+      "learning_rate": 3.595730486991328e-06,
+      "loss": 1.3782,
+      "step": 14461
+    },
+    {
+      "epoch": 13.350565428109855,
+      "grad_norm": 3.033203125,
+      "learning_rate": 3.589059372915277e-06,
+      "loss": 1.153,
+      "step": 14462
+    },
+    {
+      "epoch": 13.35148857604431,
+      "grad_norm": 3.013671875,
+      "learning_rate": 3.582388258839226e-06,
+      "loss": 1.407,
+      "step": 14463
+    },
+    {
+      "epoch": 13.352411723978767,
+      "grad_norm": 3.361328125,
+      "learning_rate": 3.5757171447631757e-06,
+      "loss": 1.1743,
+      "step": 14464
+    },
+    {
+      "epoch": 13.353334871913225,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.5690460306871254e-06,
+      "loss": 1.151,
+      "step": 14465
+    },
+    {
+      "epoch": 13.35425801984768,
+      "grad_norm": 3.197265625,
+      "learning_rate": 3.562374916611074e-06,
+      "loss": 1.1747,
+      "step": 14466
+    },
+    {
+      "epoch": 13.355181167782137,
+      "grad_norm": 3.205078125,
+      "learning_rate": 3.555703802535024e-06,
+      "loss": 1.007,
+      "step": 14467
+    },
+    {
+      "epoch": 13.356104315716594,
+      "grad_norm": 3.05859375,
+      "learning_rate": 3.5490326884589727e-06,
+      "loss": 1.4504,
+      "step": 14468
+    },
+    {
+      "epoch": 13.35702746365105,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.542361574382922e-06,
+      "loss": 1.4165,
+      "step": 14469
+    },
+    {
+      "epoch": 13.357950611585506,
+      "grad_norm": 2.814453125,
+      "learning_rate": 3.5356904603068716e-06,
+      "loss": 1.428,
+      "step": 14470
+    },
+    {
+      "epoch": 13.358873759519962,
+      "grad_norm": 2.947265625,
+      "learning_rate": 3.5290193462308204e-06,
+      "loss": 1.4011,
+      "step": 14471
+    },
+    {
+      "epoch": 13.35979690745442,
+      "grad_norm": 3.146484375,
+      "learning_rate": 3.52234823215477e-06,
+      "loss": 1.4863,
+      "step": 14472
+    },
+    {
+      "epoch": 13.360720055388876,
+      "grad_norm": 3.927734375,
+      "learning_rate": 3.5156771180787197e-06,
+      "loss": 1.5144,
+      "step": 14473
+    },
+    {
+      "epoch": 13.361643203323332,
+      "grad_norm": 3.142578125,
+      "learning_rate": 3.5090060040026685e-06,
+      "loss": 1.22,
+      "step": 14474
+    },
+    {
+      "epoch": 13.362566351257788,
+      "grad_norm": 3.029296875,
+      "learning_rate": 3.5023348899266178e-06,
+      "loss": 1.469,
+      "step": 14475
+    },
+    {
+      "epoch": 13.363489499192246,
+      "grad_norm": 2.861328125,
+      "learning_rate": 3.4956637758505674e-06,
+      "loss": 1.1322,
+      "step": 14476
+    },
+    {
+      "epoch": 13.364412647126702,
+      "grad_norm": 3.13671875,
+      "learning_rate": 3.4889926617745163e-06,
+      "loss": 1.4548,
+      "step": 14477
+    },
+    {
+      "epoch": 13.365335795061158,
+      "grad_norm": 3.779296875,
+      "learning_rate": 3.482321547698466e-06,
+      "loss": 1.3333,
+      "step": 14478
+    },
+    {
+      "epoch": 13.366258942995614,
+      "grad_norm": 3.1640625,
+      "learning_rate": 3.4756504336224147e-06,
+      "loss": 1.2944,
+      "step": 14479
+    },
+    {
+      "epoch": 13.367182090930072,
+      "grad_norm": 3.20703125,
+      "learning_rate": 3.4689793195463644e-06,
+      "loss": 1.3237,
+      "step": 14480
+    },
+    {
+      "epoch": 13.368105238864528,
+      "grad_norm": 3.16015625,
+      "learning_rate": 3.462308205470314e-06,
+      "loss": 1.0548,
+      "step": 14481
+    },
+    {
+      "epoch": 13.369028386798984,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.455637091394263e-06,
+      "loss": 1.2676,
+      "step": 14482
+    },
+    {
+      "epoch": 13.369951534733442,
+      "grad_norm": 3.625,
+      "learning_rate": 3.448965977318212e-06,
+      "loss": 1.1841,
+      "step": 14483
+    },
+    {
+      "epoch": 13.370874682667898,
+      "grad_norm": 3.408203125,
+      "learning_rate": 3.442294863242162e-06,
+      "loss": 1.0576,
+      "step": 14484
+    },
+    {
+      "epoch": 13.371797830602354,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.4356237491661106e-06,
+      "loss": 1.0518,
+      "step": 14485
+    },
+    {
+      "epoch": 13.37272097853681,
+      "grad_norm": 3.275390625,
+      "learning_rate": 3.4289526350900603e-06,
+      "loss": 1.4871,
+      "step": 14486
+    },
+    {
+      "epoch": 13.373644126471268,
+      "grad_norm": 3.3046875,
+      "learning_rate": 3.42228152101401e-06,
+      "loss": 1.4536,
+      "step": 14487
+    },
+    {
+      "epoch": 13.374567274405724,
+      "grad_norm": 3.51171875,
+      "learning_rate": 3.4156104069379588e-06,
+      "loss": 1.3167,
+      "step": 14488
+    },
+    {
+      "epoch": 13.37549042234018,
+      "grad_norm": 2.486328125,
+      "learning_rate": 3.408939292861908e-06,
+      "loss": 0.9017,
+      "step": 14489
+    },
+    {
+      "epoch": 13.376413570274636,
+      "grad_norm": 2.97265625,
+      "learning_rate": 3.4022681787858573e-06,
+      "loss": 1.2705,
+      "step": 14490
+    },
+    {
+      "epoch": 13.377336718209094,
+      "grad_norm": 2.982421875,
+      "learning_rate": 3.3955970647098065e-06,
+      "loss": 1.2878,
+      "step": 14491
+    },
+    {
+      "epoch": 13.37825986614355,
+      "grad_norm": 3.14453125,
+      "learning_rate": 3.388925950633756e-06,
+      "loss": 1.2766,
+      "step": 14492
+    },
+    {
+      "epoch": 13.379183014078006,
+      "grad_norm": 3.287109375,
+      "learning_rate": 3.382254836557705e-06,
+      "loss": 1.3594,
+      "step": 14493
+    },
+    {
+      "epoch": 13.380106162012462,
+      "grad_norm": 3.0546875,
+      "learning_rate": 3.3755837224816546e-06,
+      "loss": 1.2888,
+      "step": 14494
+    },
+    {
+      "epoch": 13.38102930994692,
+      "grad_norm": 2.939453125,
+      "learning_rate": 3.3689126084056043e-06,
+      "loss": 1.2234,
+      "step": 14495
+    },
+    {
+      "epoch": 13.381952457881376,
+      "grad_norm": 3.17578125,
+      "learning_rate": 3.362241494329553e-06,
+      "loss": 1.4746,
+      "step": 14496
+    },
+    {
+      "epoch": 13.382875605815832,
+      "grad_norm": 3.11328125,
+      "learning_rate": 3.3555703802535024e-06,
+      "loss": 1.5339,
+      "step": 14497
+    },
+    {
+      "epoch": 13.383798753750288,
+      "grad_norm": 3.2265625,
+      "learning_rate": 3.348899266177452e-06,
+      "loss": 1.0376,
+      "step": 14498
+    },
+    {
+      "epoch": 13.384721901684745,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.342228152101401e-06,
+      "loss": 1.0621,
+      "step": 14499
+    },
+    {
+      "epoch": 13.385645049619201,
+      "grad_norm": 3.021484375,
+      "learning_rate": 3.3355570380253505e-06,
+      "loss": 1.4004,
+      "step": 14500
+    },
+    {
+      "epoch": 13.386568197553657,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.3288859239492993e-06,
+      "loss": 1.3384,
+      "step": 14501
+    },
+    {
+      "epoch": 13.387491345488115,
+      "grad_norm": 2.98828125,
+      "learning_rate": 3.322214809873249e-06,
+      "loss": 1.2896,
+      "step": 14502
+    },
+    {
+      "epoch": 13.388414493422571,
+      "grad_norm": 3.001953125,
+      "learning_rate": 3.3155436957971983e-06,
+      "loss": 1.3,
+      "step": 14503
+    },
+    {
+      "epoch": 13.389337641357027,
+      "grad_norm": 3.18359375,
+      "learning_rate": 3.3088725817211475e-06,
+      "loss": 1.1799,
+      "step": 14504
+    },
+    {
+      "epoch": 13.390260789291483,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.3022014676450967e-06,
+      "loss": 1.3201,
+      "step": 14505
+    },
+    {
+      "epoch": 13.391183937225941,
+      "grad_norm": 3.0859375,
+      "learning_rate": 3.2955303535690464e-06,
+      "loss": 1.292,
+      "step": 14506
+    },
+    {
+      "epoch": 13.392107085160397,
+      "grad_norm": 3.1015625,
+      "learning_rate": 3.2888592394929952e-06,
+      "loss": 0.9803,
+      "step": 14507
+    },
+    {
+      "epoch": 13.393030233094853,
+      "grad_norm": 3.458984375,
+      "learning_rate": 3.282188125416945e-06,
+      "loss": 1.3816,
+      "step": 14508
+    },
+    {
+      "epoch": 13.39395338102931,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.2755170113408946e-06,
+      "loss": 1.1937,
+      "step": 14509
+    },
+    {
+      "epoch": 13.394876528963767,
+      "grad_norm": 2.8984375,
+      "learning_rate": 3.2688458972648434e-06,
+      "loss": 1.231,
+      "step": 14510
+    },
+    {
+      "epoch": 13.395799676898223,
+      "grad_norm": 3.2421875,
+      "learning_rate": 3.2621747831887926e-06,
+      "loss": 1.2969,
+      "step": 14511
+    },
+    {
+      "epoch": 13.396722824832679,
+      "grad_norm": 3.212890625,
+      "learning_rate": 3.2555036691127423e-06,
+      "loss": 1.0443,
+      "step": 14512
+    },
+    {
+      "epoch": 13.397645972767135,
+      "grad_norm": 3.7578125,
+      "learning_rate": 3.248832555036691e-06,
+      "loss": 1.2345,
+      "step": 14513
+    },
+    {
+      "epoch": 13.398569120701593,
+      "grad_norm": 3.119140625,
+      "learning_rate": 3.2421614409606408e-06,
+      "loss": 1.3821,
+      "step": 14514
+    },
+    {
+      "epoch": 13.399492268636049,
+      "grad_norm": 2.3203125,
+      "learning_rate": 3.2354903268845896e-06,
+      "loss": 1.0524,
+      "step": 14515
+    },
+    {
+      "epoch": 13.400415416570505,
+      "grad_norm": 3.509765625,
+      "learning_rate": 3.2288192128085392e-06,
+      "loss": 1.0989,
+      "step": 14516
+    },
+    {
+      "epoch": 13.401338564504963,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.2221480987324885e-06,
+      "loss": 1.353,
+      "step": 14517
+    },
+    {
+      "epoch": 13.402261712439419,
+      "grad_norm": 2.95703125,
+      "learning_rate": 3.2154769846564377e-06,
+      "loss": 1.2664,
+      "step": 14518
+    },
+    {
+      "epoch": 13.403184860373875,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.208805870580387e-06,
+      "loss": 1.3276,
+      "step": 14519
+    },
+    {
+      "epoch": 13.40410800830833,
+      "grad_norm": 3.146484375,
+      "learning_rate": 3.2021347565043366e-06,
+      "loss": 1.3064,
+      "step": 14520
+    },
+    {
+      "epoch": 13.405031156242789,
+      "grad_norm": 3.42578125,
+      "learning_rate": 3.1954636424282855e-06,
+      "loss": 1.4409,
+      "step": 14521
+    },
+    {
+      "epoch": 13.405954304177245,
+      "grad_norm": 3.27734375,
+      "learning_rate": 3.188792528352235e-06,
+      "loss": 1.2366,
+      "step": 14522
+    },
+    {
+      "epoch": 13.4068774521117,
+      "grad_norm": 3.216796875,
+      "learning_rate": 3.182121414276185e-06,
+      "loss": 1.4299,
+      "step": 14523
+    },
+    {
+      "epoch": 13.407800600046157,
+      "grad_norm": 3.44921875,
+      "learning_rate": 3.1754503002001336e-06,
+      "loss": 1.2559,
+      "step": 14524
+    },
+    {
+      "epoch": 13.408723747980615,
+      "grad_norm": 3.240234375,
+      "learning_rate": 3.168779186124083e-06,
+      "loss": 1.152,
+      "step": 14525
+    },
+    {
+      "epoch": 13.40964689591507,
+      "grad_norm": 3.04296875,
+      "learning_rate": 3.1621080720480317e-06,
+      "loss": 1.2162,
+      "step": 14526
+    },
+    {
+      "epoch": 13.410570043849527,
+      "grad_norm": 3.6953125,
+      "learning_rate": 3.1554369579719813e-06,
+      "loss": 1.4495,
+      "step": 14527
+    },
+    {
+      "epoch": 13.411493191783983,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.148765843895931e-06,
+      "loss": 1.3804,
+      "step": 14528
+    },
+    {
+      "epoch": 13.41241633971844,
+      "grad_norm": 3.11328125,
+      "learning_rate": 3.14209472981988e-06,
+      "loss": 1.0961,
+      "step": 14529
+    },
+    {
+      "epoch": 13.413339487652896,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.1354236157438295e-06,
+      "loss": 1.2974,
+      "step": 14530
+    },
+    {
+      "epoch": 13.414262635587352,
+      "grad_norm": 3.08984375,
+      "learning_rate": 3.1287525016677787e-06,
+      "loss": 1.4531,
+      "step": 14531
+    },
+    {
+      "epoch": 13.41518578352181,
+      "grad_norm": 3.509765625,
+      "learning_rate": 3.122081387591728e-06,
+      "loss": 1.3569,
+      "step": 14532
+    },
+    {
+      "epoch": 13.416108931456266,
+      "grad_norm": 3.208984375,
+      "learning_rate": 3.1154102735156772e-06,
+      "loss": 1.0168,
+      "step": 14533
+    },
+    {
+      "epoch": 13.417032079390722,
+      "grad_norm": 3.61328125,
+      "learning_rate": 3.1087391594396265e-06,
+      "loss": 1.2646,
+      "step": 14534
+    },
+    {
+      "epoch": 13.417955227325178,
+      "grad_norm": 3.001953125,
+      "learning_rate": 3.1020680453635757e-06,
+      "loss": 1.1804,
+      "step": 14535
+    },
+    {
+      "epoch": 13.418878375259636,
+      "grad_norm": 3.267578125,
+      "learning_rate": 3.0953969312875254e-06,
+      "loss": 1.3792,
+      "step": 14536
+    },
+    {
+      "epoch": 13.419801523194092,
+      "grad_norm": 3.302734375,
+      "learning_rate": 3.0887258172114746e-06,
+      "loss": 1.3931,
+      "step": 14537
+    },
+    {
+      "epoch": 13.420724671128548,
+      "grad_norm": 3.48828125,
+      "learning_rate": 3.082054703135424e-06,
+      "loss": 1.1968,
+      "step": 14538
+    },
+    {
+      "epoch": 13.421647819063004,
+      "grad_norm": 3.185546875,
+      "learning_rate": 3.075383589059373e-06,
+      "loss": 1.375,
+      "step": 14539
+    },
+    {
+      "epoch": 13.422570966997462,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.0687124749833223e-06,
+      "loss": 1.4219,
+      "step": 14540
+    },
+    {
+      "epoch": 13.423494114931918,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.0620413609072716e-06,
+      "loss": 0.8576,
+      "step": 14541
+    },
+    {
+      "epoch": 13.424417262866374,
+      "grad_norm": 3.087890625,
+      "learning_rate": 3.055370246831221e-06,
+      "loss": 1.3506,
+      "step": 14542
+    },
+    {
+      "epoch": 13.42534041080083,
+      "grad_norm": 3.125,
+      "learning_rate": 3.0486991327551705e-06,
+      "loss": 1.0433,
+      "step": 14543
+    },
+    {
+      "epoch": 13.426263558735288,
+      "grad_norm": 3.576171875,
+      "learning_rate": 3.0420280186791197e-06,
+      "loss": 1.3816,
+      "step": 14544
+    },
+    {
+      "epoch": 13.427186706669744,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.035356904603069e-06,
+      "loss": 0.9077,
+      "step": 14545
+    },
+    {
+      "epoch": 13.4281098546042,
+      "grad_norm": 3.033203125,
+      "learning_rate": 3.028685790527018e-06,
+      "loss": 1.3042,
+      "step": 14546
+    },
+    {
+      "epoch": 13.429033002538656,
+      "grad_norm": 3.1640625,
+      "learning_rate": 3.0220146764509674e-06,
+      "loss": 1.269,
+      "step": 14547
+    },
+    {
+      "epoch": 13.429956150473114,
+      "grad_norm": 3.654296875,
+      "learning_rate": 3.0153435623749167e-06,
+      "loss": 1.2388,
+      "step": 14548
+    },
+    {
+      "epoch": 13.43087929840757,
+      "grad_norm": 3.41796875,
+      "learning_rate": 3.008672448298866e-06,
+      "loss": 1.1545,
+      "step": 14549
+    },
+    {
+      "epoch": 13.431802446342026,
+      "grad_norm": 3.3671875,
+      "learning_rate": 3.0020013342228156e-06,
+      "loss": 0.9951,
+      "step": 14550
+    },
+    {
+      "epoch": 13.432725594276484,
+      "grad_norm": 3.20703125,
+      "learning_rate": 2.995330220146765e-06,
+      "loss": 1.2742,
+      "step": 14551
+    },
+    {
+      "epoch": 13.43364874221094,
+      "grad_norm": 3.02734375,
+      "learning_rate": 2.988659106070714e-06,
+      "loss": 1.2817,
+      "step": 14552
+    },
+    {
+      "epoch": 13.434571890145396,
+      "grad_norm": 3.537109375,
+      "learning_rate": 2.9819879919946633e-06,
+      "loss": 1.2802,
+      "step": 14553
+    },
+    {
+      "epoch": 13.435495038079852,
+      "grad_norm": 3.32421875,
+      "learning_rate": 2.9753168779186126e-06,
+      "loss": 1.1281,
+      "step": 14554
+    },
+    {
+      "epoch": 13.43641818601431,
+      "grad_norm": 3.525390625,
+      "learning_rate": 2.968645763842562e-06,
+      "loss": 1.0227,
+      "step": 14555
+    },
+    {
+      "epoch": 13.437341333948766,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.961974649766511e-06,
+      "loss": 1.3492,
+      "step": 14556
+    },
+    {
+      "epoch": 13.438264481883222,
+      "grad_norm": 3.279296875,
+      "learning_rate": 2.9553035356904603e-06,
+      "loss": 1.1685,
+      "step": 14557
+    },
+    {
+      "epoch": 13.439187629817678,
+      "grad_norm": 3.244140625,
+      "learning_rate": 2.94863242161441e-06,
+      "loss": 0.9657,
+      "step": 14558
+    },
+    {
+      "epoch": 13.440110777752135,
+      "grad_norm": 2.873046875,
+      "learning_rate": 2.941961307538359e-06,
+      "loss": 1.1511,
+      "step": 14559
+    },
+    {
+      "epoch": 13.441033925686591,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.9352901934623084e-06,
+      "loss": 1.215,
+      "step": 14560
+    },
+    {
+      "epoch": 13.441957073621047,
+      "grad_norm": 3.33984375,
+      "learning_rate": 2.9286190793862577e-06,
+      "loss": 1.4258,
+      "step": 14561
+    },
+    {
+      "epoch": 13.442880221555503,
+      "grad_norm": 3.57421875,
+      "learning_rate": 2.921947965310207e-06,
+      "loss": 1.1589,
+      "step": 14562
+    },
+    {
+      "epoch": 13.443803369489961,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.915276851234156e-06,
+      "loss": 1.1418,
+      "step": 14563
+    },
+    {
+      "epoch": 13.444726517424417,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9086057371581054e-06,
+      "loss": 1.2803,
+      "step": 14564
+    },
+    {
+      "epoch": 13.445649665358873,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.901934623082055e-06,
+      "loss": 0.8501,
+      "step": 14565
+    },
+    {
+      "epoch": 13.446572813293331,
+      "grad_norm": 3.05859375,
+      "learning_rate": 2.8952635090060043e-06,
+      "loss": 1.3196,
+      "step": 14566
+    },
+    {
+      "epoch": 13.447495961227787,
+      "grad_norm": 3.48828125,
+      "learning_rate": 2.8885923949299536e-06,
+      "loss": 1.1615,
+      "step": 14567
+    },
+    {
+      "epoch": 13.448419109162243,
+      "grad_norm": 3.4609375,
+      "learning_rate": 2.881921280853903e-06,
+      "loss": 1.2589,
+      "step": 14568
+    },
+    {
+      "epoch": 13.4493422570967,
+      "grad_norm": 3.193359375,
+      "learning_rate": 2.875250166777852e-06,
+      "loss": 1.2031,
+      "step": 14569
+    },
+    {
+      "epoch": 13.450265405031157,
+      "grad_norm": 3.443359375,
+      "learning_rate": 2.8685790527018013e-06,
+      "loss": 1.256,
+      "step": 14570
+    },
+    {
+      "epoch": 13.451188552965613,
+      "grad_norm": 3.380859375,
+      "learning_rate": 2.8619079386257505e-06,
+      "loss": 1.3777,
+      "step": 14571
+    },
+    {
+      "epoch": 13.452111700900069,
+      "grad_norm": 3.076171875,
+      "learning_rate": 2.8552368245497e-06,
+      "loss": 1.3875,
+      "step": 14572
+    },
+    {
+      "epoch": 13.453034848834525,
+      "grad_norm": 3.302734375,
+      "learning_rate": 2.8485657104736494e-06,
+      "loss": 1.0266,
+      "step": 14573
+    },
+    {
+      "epoch": 13.453957996768983,
+      "grad_norm": 2.982421875,
+      "learning_rate": 2.8418945963975987e-06,
+      "loss": 1.2379,
+      "step": 14574
+    },
+    {
+      "epoch": 13.454881144703439,
+      "grad_norm": 3.38671875,
+      "learning_rate": 2.8352234823215475e-06,
+      "loss": 1.3965,
+      "step": 14575
+    },
+    {
+      "epoch": 13.455804292637895,
+      "grad_norm": 3.099609375,
+      "learning_rate": 2.828552368245497e-06,
+      "loss": 0.8393,
+      "step": 14576
+    },
+    {
+      "epoch": 13.456727440572351,
+      "grad_norm": 3.24609375,
+      "learning_rate": 2.8218812541694464e-06,
+      "loss": 1.27,
+      "step": 14577
+    },
+    {
+      "epoch": 13.457650588506809,
+      "grad_norm": 2.787109375,
+      "learning_rate": 2.8152101400933957e-06,
+      "loss": 1.2012,
+      "step": 14578
+    },
+    {
+      "epoch": 13.458573736441265,
+      "grad_norm": 3.0234375,
+      "learning_rate": 2.8085390260173453e-06,
+      "loss": 1.2062,
+      "step": 14579
+    },
+    {
+      "epoch": 13.45949688437572,
+      "grad_norm": 2.92578125,
+      "learning_rate": 2.8018679119412946e-06,
+      "loss": 1.1663,
+      "step": 14580
+    },
+    {
+      "epoch": 13.460420032310179,
+      "grad_norm": 3.798828125,
+      "learning_rate": 2.795196797865244e-06,
+      "loss": 1.2532,
+      "step": 14581
+    },
+    {
+      "epoch": 13.461343180244635,
+      "grad_norm": 2.849609375,
+      "learning_rate": 2.7885256837891926e-06,
+      "loss": 1.1659,
+      "step": 14582
+    },
+    {
+      "epoch": 13.46226632817909,
+      "grad_norm": 3.1015625,
+      "learning_rate": 2.7818545697131423e-06,
+      "loss": 1.3425,
+      "step": 14583
+    },
+    {
+      "epoch": 13.463189476113547,
+      "grad_norm": 2.85546875,
+      "learning_rate": 2.7751834556370915e-06,
+      "loss": 1.0966,
+      "step": 14584
+    },
+    {
+      "epoch": 13.464112624048004,
+      "grad_norm": 2.919921875,
+      "learning_rate": 2.7685123415610408e-06,
+      "loss": 1.1238,
+      "step": 14585
+    },
+    {
+      "epoch": 13.46503577198246,
+      "grad_norm": 3.185546875,
+      "learning_rate": 2.76184122748499e-06,
+      "loss": 1.4702,
+      "step": 14586
+    },
+    {
+      "epoch": 13.465958919916917,
+      "grad_norm": 3.66796875,
+      "learning_rate": 2.7551701134089397e-06,
+      "loss": 1.1782,
+      "step": 14587
+    },
+    {
+      "epoch": 13.466882067851373,
+      "grad_norm": 3.00390625,
+      "learning_rate": 2.7484989993328885e-06,
+      "loss": 1.4031,
+      "step": 14588
+    },
+    {
+      "epoch": 13.46780521578583,
+      "grad_norm": 3.08203125,
+      "learning_rate": 2.7418278852568377e-06,
+      "loss": 1.1152,
+      "step": 14589
+    },
+    {
+      "epoch": 13.468728363720286,
+      "grad_norm": 3.23828125,
+      "learning_rate": 2.7351567711807874e-06,
+      "loss": 1.2588,
+      "step": 14590
+    },
+    {
+      "epoch": 13.469651511654742,
+      "grad_norm": 2.853515625,
+      "learning_rate": 2.7284856571047366e-06,
+      "loss": 1.51,
+      "step": 14591
+    },
+    {
+      "epoch": 13.470574659589198,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.721814543028686e-06,
+      "loss": 1.007,
+      "step": 14592
+    },
+    {
+      "epoch": 13.471497807523656,
+      "grad_norm": 2.87890625,
+      "learning_rate": 2.715143428952635e-06,
+      "loss": 1.106,
+      "step": 14593
+    },
+    {
+      "epoch": 13.472420955458112,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.708472314876585e-06,
+      "loss": 1.181,
+      "step": 14594
+    },
+    {
+      "epoch": 13.473344103392568,
+      "grad_norm": 3.630859375,
+      "learning_rate": 2.7018012008005336e-06,
+      "loss": 1.3352,
+      "step": 14595
+    },
+    {
+      "epoch": 13.474267251327024,
+      "grad_norm": 2.8828125,
+      "learning_rate": 2.695130086724483e-06,
+      "loss": 1.1316,
+      "step": 14596
+    },
+    {
+      "epoch": 13.475190399261482,
+      "grad_norm": 3.23828125,
+      "learning_rate": 2.6884589726484325e-06,
+      "loss": 1.1749,
+      "step": 14597
+    },
+    {
+      "epoch": 13.476113547195938,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.6817878585723818e-06,
+      "loss": 0.9271,
+      "step": 14598
+    },
+    {
+      "epoch": 13.477036695130394,
+      "grad_norm": 2.912109375,
+      "learning_rate": 2.675116744496331e-06,
+      "loss": 1.3743,
+      "step": 14599
+    },
+    {
+      "epoch": 13.477959843064852,
+      "grad_norm": 3.357421875,
+      "learning_rate": 2.6684456304202803e-06,
+      "loss": 1.248,
+      "step": 14600
+    },
+    {
+      "epoch": 13.478882990999308,
+      "grad_norm": 3.365234375,
+      "learning_rate": 2.66177451634423e-06,
+      "loss": 1.3486,
+      "step": 14601
+    },
+    {
+      "epoch": 13.479806138933764,
+      "grad_norm": 3.43359375,
+      "learning_rate": 2.6551034022681787e-06,
+      "loss": 1.2058,
+      "step": 14602
+    },
+    {
+      "epoch": 13.48072928686822,
+      "grad_norm": 2.490234375,
+      "learning_rate": 2.648432288192128e-06,
+      "loss": 1.0387,
+      "step": 14603
+    },
+    {
+      "epoch": 13.481652434802678,
+      "grad_norm": 3.9609375,
+      "learning_rate": 2.6417611741160772e-06,
+      "loss": 1.265,
+      "step": 14604
+    },
+    {
+      "epoch": 13.482575582737134,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.635090060040027e-06,
+      "loss": 1.2074,
+      "step": 14605
+    },
+    {
+      "epoch": 13.48349873067159,
+      "grad_norm": 2.830078125,
+      "learning_rate": 2.628418945963976e-06,
+      "loss": 1.2607,
+      "step": 14606
+    },
+    {
+      "epoch": 13.484421878606046,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.6217478318879254e-06,
+      "loss": 1.0713,
+      "step": 14607
+    },
+    {
+      "epoch": 13.485345026540504,
+      "grad_norm": 2.771484375,
+      "learning_rate": 2.615076717811875e-06,
+      "loss": 1.3418,
+      "step": 14608
+    },
+    {
+      "epoch": 13.48626817447496,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.608405603735824e-06,
+      "loss": 1.1917,
+      "step": 14609
+    },
+    {
+      "epoch": 13.487191322409416,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.601734489659773e-06,
+      "loss": 1.3354,
+      "step": 14610
+    },
+    {
+      "epoch": 13.488114470343872,
+      "grad_norm": 2.90234375,
+      "learning_rate": 2.5950633755837223e-06,
+      "loss": 1.2906,
+      "step": 14611
+    },
+    {
+      "epoch": 13.48903761827833,
+      "grad_norm": 3.33984375,
+      "learning_rate": 2.588392261507672e-06,
+      "loss": 1.3467,
+      "step": 14612
+    },
+    {
+      "epoch": 13.489960766212786,
+      "grad_norm": 3.099609375,
+      "learning_rate": 2.5817211474316212e-06,
+      "loss": 1.0498,
+      "step": 14613
+    },
+    {
+      "epoch": 13.490883914147242,
+      "grad_norm": 3.416015625,
+      "learning_rate": 2.5750500333555705e-06,
+      "loss": 1.1848,
+      "step": 14614
+    },
+    {
+      "epoch": 13.4918070620817,
+      "grad_norm": 2.80078125,
+      "learning_rate": 2.56837891927952e-06,
+      "loss": 1.3479,
+      "step": 14615
+    },
+    {
+      "epoch": 13.492730210016155,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.561707805203469e-06,
+      "loss": 1.0665,
+      "step": 14616
+    },
+    {
+      "epoch": 13.493653357950611,
+      "grad_norm": 3.36328125,
+      "learning_rate": 2.5550366911274182e-06,
+      "loss": 1.1364,
+      "step": 14617
+    },
+    {
+      "epoch": 13.494576505885068,
+      "grad_norm": 3.330078125,
+      "learning_rate": 2.5483655770513675e-06,
+      "loss": 1.3306,
+      "step": 14618
+    },
+    {
+      "epoch": 13.495499653819525,
+      "grad_norm": 3.490234375,
+      "learning_rate": 2.541694462975317e-06,
+      "loss": 1.5164,
+      "step": 14619
+    },
+    {
+      "epoch": 13.496422801753981,
+      "grad_norm": 2.865234375,
+      "learning_rate": 2.5350233488992664e-06,
+      "loss": 1.152,
+      "step": 14620
+    },
+    {
+      "epoch": 13.497345949688437,
+      "grad_norm": 2.708984375,
+      "learning_rate": 2.5283522348232156e-06,
+      "loss": 1.072,
+      "step": 14621
+    },
+    {
+      "epoch": 13.498269097622893,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.521681120747165e-06,
+      "loss": 1.293,
+      "step": 14622
+    },
+    {
+      "epoch": 13.499192245557351,
+      "grad_norm": 3.28515625,
+      "learning_rate": 2.515010006671114e-06,
+      "loss": 1.5093,
+      "step": 14623
+    },
+    {
+      "epoch": 13.500115393491807,
+      "grad_norm": 3.310546875,
+      "learning_rate": 2.5083388925950633e-06,
+      "loss": 1.3306,
+      "step": 14624
+    },
+    {
+      "epoch": 13.501038541426263,
+      "grad_norm": 3.1484375,
+      "learning_rate": 2.5016677785190126e-06,
+      "loss": 1.2672,
+      "step": 14625
+    },
+    {
+      "epoch": 13.50196168936072,
+      "grad_norm": 3.251953125,
+      "learning_rate": 2.4949966644429622e-06,
+      "loss": 1.29,
+      "step": 14626
+    },
+    {
+      "epoch": 13.502884837295177,
+      "grad_norm": 3.150390625,
+      "learning_rate": 2.4883255503669115e-06,
+      "loss": 1.3342,
+      "step": 14627
+    },
+    {
+      "epoch": 13.503807985229633,
+      "grad_norm": 2.83984375,
+      "learning_rate": 2.4816544362908607e-06,
+      "loss": 1.2326,
+      "step": 14628
+    },
+    {
+      "epoch": 13.50473113316409,
+      "grad_norm": 3.7265625,
+      "learning_rate": 2.47498332221481e-06,
+      "loss": 1.3206,
+      "step": 14629
+    },
+    {
+      "epoch": 13.505654281098547,
+      "grad_norm": 3.31640625,
+      "learning_rate": 2.468312208138759e-06,
+      "loss": 1.1218,
+      "step": 14630
+    },
+    {
+      "epoch": 13.506577429033003,
+      "grad_norm": 3.60546875,
+      "learning_rate": 2.4616410940627085e-06,
+      "loss": 1.4136,
+      "step": 14631
+    },
+    {
+      "epoch": 13.507500576967459,
+      "grad_norm": 3.220703125,
+      "learning_rate": 2.4549699799866577e-06,
+      "loss": 1.2224,
+      "step": 14632
+    },
+    {
+      "epoch": 13.508423724901915,
+      "grad_norm": 3.064453125,
+      "learning_rate": 2.448298865910607e-06,
+      "loss": 1.2803,
+      "step": 14633
+    },
+    {
+      "epoch": 13.509346872836373,
+      "grad_norm": 3.0703125,
+      "learning_rate": 2.4416277518345566e-06,
+      "loss": 1.1259,
+      "step": 14634
+    },
+    {
+      "epoch": 13.510270020770829,
+      "grad_norm": 3.396484375,
+      "learning_rate": 2.434956637758506e-06,
+      "loss": 1.2306,
+      "step": 14635
+    },
+    {
+      "epoch": 13.511193168705285,
+      "grad_norm": 2.998046875,
+      "learning_rate": 2.428285523682455e-06,
+      "loss": 1.2986,
+      "step": 14636
+    },
+    {
+      "epoch": 13.51211631663974,
+      "grad_norm": 3.38671875,
+      "learning_rate": 2.4216144096064043e-06,
+      "loss": 1.2607,
+      "step": 14637
+    },
+    {
+      "epoch": 13.513039464574199,
+      "grad_norm": 3.25390625,
+      "learning_rate": 2.4149432955303536e-06,
+      "loss": 1.2378,
+      "step": 14638
+    },
+    {
+      "epoch": 13.513962612508655,
+      "grad_norm": 3.0703125,
+      "learning_rate": 2.408272181454303e-06,
+      "loss": 1.1416,
+      "step": 14639
+    },
+    {
+      "epoch": 13.51488576044311,
+      "grad_norm": 3.197265625,
+      "learning_rate": 2.401601067378252e-06,
+      "loss": 1.249,
+      "step": 14640
+    },
+    {
+      "epoch": 13.515808908377567,
+      "grad_norm": 2.888671875,
+      "learning_rate": 2.3949299533022017e-06,
+      "loss": 1.149,
+      "step": 14641
+    },
+    {
+      "epoch": 13.516732056312025,
+      "grad_norm": 2.82421875,
+      "learning_rate": 2.388258839226151e-06,
+      "loss": 1.2395,
+      "step": 14642
+    },
+    {
+      "epoch": 13.51765520424648,
+      "grad_norm": 3.962890625,
+      "learning_rate": 2.3815877251501e-06,
+      "loss": 1.1107,
+      "step": 14643
+    },
+    {
+      "epoch": 13.518578352180937,
+      "grad_norm": 3.11328125,
+      "learning_rate": 2.3749166110740494e-06,
+      "loss": 1.3938,
+      "step": 14644
+    },
+    {
+      "epoch": 13.519501500115393,
+      "grad_norm": 2.728515625,
+      "learning_rate": 2.3682454969979987e-06,
+      "loss": 1.0738,
+      "step": 14645
+    },
+    {
+      "epoch": 13.52042464804985,
+      "grad_norm": 3.5390625,
+      "learning_rate": 2.361574382921948e-06,
+      "loss": 1.1025,
+      "step": 14646
+    },
+    {
+      "epoch": 13.521347795984306,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.354903268845897e-06,
+      "loss": 0.8692,
+      "step": 14647
+    },
+    {
+      "epoch": 13.522270943918762,
+      "grad_norm": 3.376953125,
+      "learning_rate": 2.348232154769847e-06,
+      "loss": 1.2058,
+      "step": 14648
+    },
+    {
+      "epoch": 13.52319409185322,
+      "grad_norm": 2.908203125,
+      "learning_rate": 2.341561040693796e-06,
+      "loss": 1.2311,
+      "step": 14649
+    },
+    {
+      "epoch": 13.524117239787676,
+      "grad_norm": 2.998046875,
+      "learning_rate": 2.3348899266177453e-06,
+      "loss": 1.1754,
+      "step": 14650
+    },
+    {
+      "epoch": 13.525040387722132,
+      "grad_norm": 3.35546875,
+      "learning_rate": 2.3282188125416946e-06,
+      "loss": 1.3286,
+      "step": 14651
+    },
+    {
+      "epoch": 13.525963535656588,
+      "grad_norm": 3.455078125,
+      "learning_rate": 2.321547698465644e-06,
+      "loss": 1.175,
+      "step": 14652
+    },
+    {
+      "epoch": 13.526886683591046,
+      "grad_norm": 3.158203125,
+      "learning_rate": 2.314876584389593e-06,
+      "loss": 1.0278,
+      "step": 14653
+    },
+    {
+      "epoch": 13.527809831525502,
+      "grad_norm": 3.275390625,
+      "learning_rate": 2.3082054703135423e-06,
+      "loss": 1.3311,
+      "step": 14654
+    },
+    {
+      "epoch": 13.528732979459958,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.301534356237492e-06,
+      "loss": 1.3608,
+      "step": 14655
+    },
+    {
+      "epoch": 13.529656127394414,
+      "grad_norm": 3.662109375,
+      "learning_rate": 2.294863242161441e-06,
+      "loss": 1.2783,
+      "step": 14656
+    },
+    {
+      "epoch": 13.530579275328872,
+      "grad_norm": 3.1640625,
+      "learning_rate": 2.2881921280853904e-06,
+      "loss": 1.2676,
+      "step": 14657
+    },
+    {
+      "epoch": 13.531502423263328,
+      "grad_norm": 2.86328125,
+      "learning_rate": 2.2815210140093397e-06,
+      "loss": 1.3729,
+      "step": 14658
+    },
+    {
+      "epoch": 13.532425571197784,
+      "grad_norm": 2.78515625,
+      "learning_rate": 2.274849899933289e-06,
+      "loss": 1.4041,
+      "step": 14659
+    },
+    {
+      "epoch": 13.53334871913224,
+      "grad_norm": 3.044921875,
+      "learning_rate": 2.268178785857238e-06,
+      "loss": 1.4106,
+      "step": 14660
+    },
+    {
+      "epoch": 13.534271867066698,
+      "grad_norm": 3.349609375,
+      "learning_rate": 2.2615076717811874e-06,
+      "loss": 1.3267,
+      "step": 14661
+    },
+    {
+      "epoch": 13.535195015001154,
+      "grad_norm": 3.142578125,
+      "learning_rate": 2.254836557705137e-06,
+      "loss": 1.1821,
+      "step": 14662
+    },
+    {
+      "epoch": 13.53611816293561,
+      "grad_norm": 3.306640625,
+      "learning_rate": 2.2481654436290863e-06,
+      "loss": 1.4009,
+      "step": 14663
+    },
+    {
+      "epoch": 13.537041310870066,
+      "grad_norm": 3.07421875,
+      "learning_rate": 2.2414943295530356e-06,
+      "loss": 1.2567,
+      "step": 14664
+    },
+    {
+      "epoch": 13.537964458804524,
+      "grad_norm": 3.28515625,
+      "learning_rate": 2.234823215476985e-06,
+      "loss": 1.4753,
+      "step": 14665
+    },
+    {
+      "epoch": 13.53888760673898,
+      "grad_norm": 2.935546875,
+      "learning_rate": 2.228152101400934e-06,
+      "loss": 1.2893,
+      "step": 14666
+    },
+    {
+      "epoch": 13.539810754673436,
+      "grad_norm": 3.4765625,
+      "learning_rate": 2.2214809873248833e-06,
+      "loss": 1.3126,
+      "step": 14667
+    },
+    {
+      "epoch": 13.540733902607894,
+      "grad_norm": 2.810546875,
+      "learning_rate": 2.2148098732488325e-06,
+      "loss": 1.269,
+      "step": 14668
+    },
+    {
+      "epoch": 13.54165705054235,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.2081387591727818e-06,
+      "loss": 1.5479,
+      "step": 14669
+    },
+    {
+      "epoch": 13.542580198476806,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.2014676450967314e-06,
+      "loss": 1.1763,
+      "step": 14670
+    },
+    {
+      "epoch": 13.543503346411262,
+      "grad_norm": 3.322265625,
+      "learning_rate": 2.1947965310206807e-06,
+      "loss": 1.2942,
+      "step": 14671
+    },
+    {
+      "epoch": 13.54442649434572,
+      "grad_norm": 3.23046875,
+      "learning_rate": 2.18812541694463e-06,
+      "loss": 1.0167,
+      "step": 14672
+    },
+    {
+      "epoch": 13.545349642280176,
+      "grad_norm": 3.083984375,
+      "learning_rate": 2.181454302868579e-06,
+      "loss": 1.1754,
+      "step": 14673
+    },
+    {
+      "epoch": 13.546272790214632,
+      "grad_norm": 3.1953125,
+      "learning_rate": 2.1747831887925284e-06,
+      "loss": 1.2606,
+      "step": 14674
+    },
+    {
+      "epoch": 13.547195938149088,
+      "grad_norm": 2.8671875,
+      "learning_rate": 2.1681120747164777e-06,
+      "loss": 1.1744,
+      "step": 14675
+    },
+    {
+      "epoch": 13.548119086083545,
+      "grad_norm": 3.2421875,
+      "learning_rate": 2.161440960640427e-06,
+      "loss": 1.152,
+      "step": 14676
+    },
+    {
+      "epoch": 13.549042234018001,
+      "grad_norm": 2.8203125,
+      "learning_rate": 2.1547698465643766e-06,
+      "loss": 1.3345,
+      "step": 14677
+    },
+    {
+      "epoch": 13.549965381952457,
+      "grad_norm": 3.66796875,
+      "learning_rate": 2.148098732488326e-06,
+      "loss": 1.4333,
+      "step": 14678
+    },
+    {
+      "epoch": 13.550888529886915,
+      "grad_norm": 3.10546875,
+      "learning_rate": 2.141427618412275e-06,
+      "loss": 1.1339,
+      "step": 14679
+    },
+    {
+      "epoch": 13.551811677821371,
+      "grad_norm": 3.41796875,
+      "learning_rate": 2.1347565043362243e-06,
+      "loss": 1.3438,
+      "step": 14680
+    },
+    {
+      "epoch": 13.552734825755827,
+      "grad_norm": 3.376953125,
+      "learning_rate": 2.1280853902601735e-06,
+      "loss": 1.2603,
+      "step": 14681
+    },
+    {
+      "epoch": 13.553657973690283,
+      "grad_norm": 3.57421875,
+      "learning_rate": 2.1214142761841228e-06,
+      "loss": 1.4019,
+      "step": 14682
+    },
+    {
+      "epoch": 13.554581121624741,
+      "grad_norm": 3.21484375,
+      "learning_rate": 2.114743162108072e-06,
+      "loss": 1.1222,
+      "step": 14683
+    },
+    {
+      "epoch": 13.555504269559197,
+      "grad_norm": 3.669921875,
+      "learning_rate": 2.1080720480320217e-06,
+      "loss": 1.2546,
+      "step": 14684
+    },
+    {
+      "epoch": 13.556427417493653,
+      "grad_norm": 3.119140625,
+      "learning_rate": 2.101400933955971e-06,
+      "loss": 1.2351,
+      "step": 14685
+    },
+    {
+      "epoch": 13.55735056542811,
+      "grad_norm": 2.943359375,
+      "learning_rate": 2.09472981987992e-06,
+      "loss": 1.2118,
+      "step": 14686
+    },
+    {
+      "epoch": 13.558273713362567,
+      "grad_norm": 3.125,
+      "learning_rate": 2.0880587058038694e-06,
+      "loss": 1.3125,
+      "step": 14687
+    },
+    {
+      "epoch": 13.559196861297023,
+      "grad_norm": 3.404296875,
+      "learning_rate": 2.0813875917278186e-06,
+      "loss": 1.1334,
+      "step": 14688
+    },
+    {
+      "epoch": 13.560120009231479,
+      "grad_norm": 3.49609375,
+      "learning_rate": 2.074716477651768e-06,
+      "loss": 1.4551,
+      "step": 14689
+    },
+    {
+      "epoch": 13.561043157165935,
+      "grad_norm": 3.173828125,
+      "learning_rate": 2.068045363575717e-06,
+      "loss": 1.2258,
+      "step": 14690
+    },
+    {
+      "epoch": 13.561966305100393,
+      "grad_norm": 2.498046875,
+      "learning_rate": 2.061374249499667e-06,
+      "loss": 1.2065,
+      "step": 14691
+    },
+    {
+      "epoch": 13.562889453034849,
+      "grad_norm": 2.939453125,
+      "learning_rate": 2.054703135423616e-06,
+      "loss": 1.2847,
+      "step": 14692
+    },
+    {
+      "epoch": 13.563812600969305,
+      "grad_norm": 2.9609375,
+      "learning_rate": 2.0480320213475653e-06,
+      "loss": 1.2034,
+      "step": 14693
+    },
+    {
+      "epoch": 13.564735748903761,
+      "grad_norm": 3.826171875,
+      "learning_rate": 2.0413609072715145e-06,
+      "loss": 1.0906,
+      "step": 14694
+    },
+    {
+      "epoch": 13.565658896838219,
+      "grad_norm": 3.47265625,
+      "learning_rate": 2.0346897931954638e-06,
+      "loss": 1.2573,
+      "step": 14695
+    },
+    {
+      "epoch": 13.566582044772675,
+      "grad_norm": 3.130859375,
+      "learning_rate": 2.028018679119413e-06,
+      "loss": 1.2643,
+      "step": 14696
+    },
+    {
+      "epoch": 13.56750519270713,
+      "grad_norm": 3.115234375,
+      "learning_rate": 2.0213475650433622e-06,
+      "loss": 1.1479,
+      "step": 14697
+    },
+    {
+      "epoch": 13.568428340641589,
+      "grad_norm": 3.26171875,
+      "learning_rate": 2.0146764509673115e-06,
+      "loss": 1.1909,
+      "step": 14698
+    },
+    {
+      "epoch": 13.569351488576045,
+      "grad_norm": 3.38671875,
+      "learning_rate": 2.008005336891261e-06,
+      "loss": 1.2156,
+      "step": 14699
+    },
+    {
+      "epoch": 13.5702746365105,
+      "grad_norm": 3.154296875,
+      "learning_rate": 2.0013342228152104e-06,
+      "loss": 1.5015,
+      "step": 14700
+    },
+    {
+      "epoch": 13.571197784444957,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.9946631087391596e-06,
+      "loss": 1.3445,
+      "step": 14701
+    },
+    {
+      "epoch": 13.572120932379415,
+      "grad_norm": 2.947265625,
+      "learning_rate": 1.987991994663109e-06,
+      "loss": 1.2834,
+      "step": 14702
+    },
+    {
+      "epoch": 13.57304408031387,
+      "grad_norm": 3.060546875,
+      "learning_rate": 1.981320880587058e-06,
+      "loss": 1.4126,
+      "step": 14703
+    },
+    {
+      "epoch": 13.573967228248327,
+      "grad_norm": 2.78515625,
+      "learning_rate": 1.9746497665110074e-06,
+      "loss": 1.0856,
+      "step": 14704
+    },
+    {
+      "epoch": 13.574890376182783,
+      "grad_norm": 2.82421875,
+      "learning_rate": 1.9679786524349566e-06,
+      "loss": 1.0906,
+      "step": 14705
+    },
+    {
+      "epoch": 13.57581352411724,
+      "grad_norm": 3.283203125,
+      "learning_rate": 1.9613075383589063e-06,
+      "loss": 1.4534,
+      "step": 14706
+    },
+    {
+      "epoch": 13.576736672051696,
+      "grad_norm": 3.12890625,
+      "learning_rate": 1.9546364242828555e-06,
+      "loss": 0.976,
+      "step": 14707
+    },
+    {
+      "epoch": 13.577659819986152,
+      "grad_norm": 3.05859375,
+      "learning_rate": 1.9479653102068043e-06,
+      "loss": 1.4033,
+      "step": 14708
+    },
+    {
+      "epoch": 13.578582967920608,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.941294196130754e-06,
+      "loss": 1.1497,
+      "step": 14709
+    },
+    {
+      "epoch": 13.579506115855066,
+      "grad_norm": 3.080078125,
+      "learning_rate": 1.9346230820547032e-06,
+      "loss": 1.3442,
+      "step": 14710
+    },
+    {
+      "epoch": 13.580429263789522,
+      "grad_norm": 2.830078125,
+      "learning_rate": 1.9279519679786525e-06,
+      "loss": 1.3687,
+      "step": 14711
+    },
+    {
+      "epoch": 13.581352411723978,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.9212808539026017e-06,
+      "loss": 1.1029,
+      "step": 14712
+    },
+    {
+      "epoch": 13.582275559658434,
+      "grad_norm": 3.0390625,
+      "learning_rate": 1.9146097398265514e-06,
+      "loss": 1.0668,
+      "step": 14713
+    },
+    {
+      "epoch": 13.583198707592892,
+      "grad_norm": 3.115234375,
+      "learning_rate": 1.9079386257505006e-06,
+      "loss": 1.2927,
+      "step": 14714
+    },
+    {
+      "epoch": 13.584121855527348,
+      "grad_norm": 2.830078125,
+      "learning_rate": 1.9012675116744497e-06,
+      "loss": 1.3203,
+      "step": 14715
+    },
+    {
+      "epoch": 13.585045003461804,
+      "grad_norm": 3.3203125,
+      "learning_rate": 1.894596397598399e-06,
+      "loss": 1.2791,
+      "step": 14716
+    },
+    {
+      "epoch": 13.585968151396262,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.8879252835223484e-06,
+      "loss": 1.095,
+      "step": 14717
+    },
+    {
+      "epoch": 13.586891299330718,
+      "grad_norm": 3.16796875,
+      "learning_rate": 1.8812541694462976e-06,
+      "loss": 1.3921,
+      "step": 14718
+    },
+    {
+      "epoch": 13.587814447265174,
+      "grad_norm": 3.3984375,
+      "learning_rate": 1.8745830553702468e-06,
+      "loss": 0.8198,
+      "step": 14719
+    },
+    {
+      "epoch": 13.58873759519963,
+      "grad_norm": 3.51953125,
+      "learning_rate": 1.8679119412941963e-06,
+      "loss": 1.2375,
+      "step": 14720
+    },
+    {
+      "epoch": 13.589660743134088,
+      "grad_norm": 3.07421875,
+      "learning_rate": 1.8612408272181455e-06,
+      "loss": 1.2441,
+      "step": 14721
+    },
+    {
+      "epoch": 13.590583891068544,
+      "grad_norm": 2.95703125,
+      "learning_rate": 1.8545697131420948e-06,
+      "loss": 1.2466,
+      "step": 14722
+    },
+    {
+      "epoch": 13.591507039003,
+      "grad_norm": 3.1328125,
+      "learning_rate": 1.847898599066044e-06,
+      "loss": 0.9722,
+      "step": 14723
+    },
+    {
+      "epoch": 13.592430186937456,
+      "grad_norm": 3.189453125,
+      "learning_rate": 1.8412274849899935e-06,
+      "loss": 1.2896,
+      "step": 14724
+    },
+    {
+      "epoch": 13.593353334871914,
+      "grad_norm": 3.146484375,
+      "learning_rate": 1.8345563709139427e-06,
+      "loss": 1.181,
+      "step": 14725
+    },
+    {
+      "epoch": 13.59427648280637,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.827885256837892e-06,
+      "loss": 1.3501,
+      "step": 14726
+    },
+    {
+      "epoch": 13.595199630740826,
+      "grad_norm": 3.53515625,
+      "learning_rate": 1.8212141427618412e-06,
+      "loss": 1.068,
+      "step": 14727
+    },
+    {
+      "epoch": 13.596122778675284,
+      "grad_norm": 3.4296875,
+      "learning_rate": 1.8145430286857907e-06,
+      "loss": 1.2677,
+      "step": 14728
+    },
+    {
+      "epoch": 13.59704592660974,
+      "grad_norm": 3.0,
+      "learning_rate": 1.80787191460974e-06,
+      "loss": 0.9198,
+      "step": 14729
+    },
+    {
+      "epoch": 13.597969074544196,
+      "grad_norm": 3.333984375,
+      "learning_rate": 1.8012008005336891e-06,
+      "loss": 1.3723,
+      "step": 14730
+    },
+    {
+      "epoch": 13.598892222478652,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.7945296864576386e-06,
+      "loss": 1.3823,
+      "step": 14731
+    },
+    {
+      "epoch": 13.59981537041311,
+      "grad_norm": 2.8828125,
+      "learning_rate": 1.7878585723815878e-06,
+      "loss": 1.2571,
+      "step": 14732
+    },
+    {
+      "epoch": 13.600738518347566,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.781187458305537e-06,
+      "loss": 1.2397,
+      "step": 14733
+    },
+    {
+      "epoch": 13.601661666282022,
+      "grad_norm": 3.267578125,
+      "learning_rate": 1.7745163442294863e-06,
+      "loss": 1.0687,
+      "step": 14734
+    },
+    {
+      "epoch": 13.602584814216478,
+      "grad_norm": 3.232421875,
+      "learning_rate": 1.7678452301534358e-06,
+      "loss": 1.0846,
+      "step": 14735
+    },
+    {
+      "epoch": 13.603507962150935,
+      "grad_norm": 2.3984375,
+      "learning_rate": 1.761174116077385e-06,
+      "loss": 1.0702,
+      "step": 14736
+    },
+    {
+      "epoch": 13.604431110085391,
+      "grad_norm": 2.9609375,
+      "learning_rate": 1.7545030020013343e-06,
+      "loss": 1.0295,
+      "step": 14737
+    },
+    {
+      "epoch": 13.605354258019847,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.7478318879252837e-06,
+      "loss": 1.2422,
+      "step": 14738
+    },
+    {
+      "epoch": 13.606277405954303,
+      "grad_norm": 3.13671875,
+      "learning_rate": 1.741160773849233e-06,
+      "loss": 1.167,
+      "step": 14739
+    },
+    {
+      "epoch": 13.607200553888761,
+      "grad_norm": 2.9453125,
+      "learning_rate": 1.7344896597731822e-06,
+      "loss": 1.3611,
+      "step": 14740
+    },
+    {
+      "epoch": 13.608123701823217,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.7278185456971314e-06,
+      "loss": 1.4055,
+      "step": 14741
+    },
+    {
+      "epoch": 13.609046849757673,
+      "grad_norm": 2.912109375,
+      "learning_rate": 1.721147431621081e-06,
+      "loss": 1.4211,
+      "step": 14742
+    },
+    {
+      "epoch": 13.60996999769213,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.7144763175450301e-06,
+      "loss": 1.1831,
+      "step": 14743
+    },
+    {
+      "epoch": 13.610893145626587,
+      "grad_norm": 3.3828125,
+      "learning_rate": 1.7078052034689794e-06,
+      "loss": 1.1683,
+      "step": 14744
+    },
+    {
+      "epoch": 13.611816293561043,
+      "grad_norm": 2.77734375,
+      "learning_rate": 1.7011340893929286e-06,
+      "loss": 1.186,
+      "step": 14745
+    },
+    {
+      "epoch": 13.6127394414955,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.694462975316878e-06,
+      "loss": 1.26,
+      "step": 14746
+    },
+    {
+      "epoch": 13.613662589429957,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.6877918612408273e-06,
+      "loss": 1.3386,
+      "step": 14747
+    },
+    {
+      "epoch": 13.614585737364413,
+      "grad_norm": 3.0546875,
+      "learning_rate": 1.6811207471647766e-06,
+      "loss": 1.2645,
+      "step": 14748
+    },
+    {
+      "epoch": 13.615508885298869,
+      "grad_norm": 3.48046875,
+      "learning_rate": 1.674449633088726e-06,
+      "loss": 1.0092,
+      "step": 14749
+    },
+    {
+      "epoch": 13.616432033233325,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.6677785190126753e-06,
+      "loss": 1.2666,
+      "step": 14750
+    },
+    {
+      "epoch": 13.617355181167783,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.6611074049366245e-06,
+      "loss": 1.3904,
+      "step": 14751
+    },
+    {
+      "epoch": 13.618278329102239,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.6544362908605737e-06,
+      "loss": 1.2002,
+      "step": 14752
+    },
+    {
+      "epoch": 13.619201477036695,
+      "grad_norm": 3.060546875,
+      "learning_rate": 1.6477651767845232e-06,
+      "loss": 1.5913,
+      "step": 14753
+    },
+    {
+      "epoch": 13.620124624971151,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.6410940627084724e-06,
+      "loss": 1.212,
+      "step": 14754
+    },
+    {
+      "epoch": 13.621047772905609,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6344229486324217e-06,
+      "loss": 1.4097,
+      "step": 14755
+    },
+    {
+      "epoch": 13.621970920840065,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6277518345563711e-06,
+      "loss": 1.0925,
+      "step": 14756
+    },
+    {
+      "epoch": 13.62289406877452,
+      "grad_norm": 3.45703125,
+      "learning_rate": 1.6210807204803204e-06,
+      "loss": 1.2871,
+      "step": 14757
+    },
+    {
+      "epoch": 13.623817216708977,
+      "grad_norm": 2.935546875,
+      "learning_rate": 1.6144096064042696e-06,
+      "loss": 1.005,
+      "step": 14758
+    },
+    {
+      "epoch": 13.624740364643435,
+      "grad_norm": 3.001953125,
+      "learning_rate": 1.6077384923282189e-06,
+      "loss": 1.3574,
+      "step": 14759
+    },
+    {
+      "epoch": 13.62566351257789,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.6010673782521683e-06,
+      "loss": 1.3154,
+      "step": 14760
+    },
+    {
+      "epoch": 13.626586660512347,
+      "grad_norm": 3.302734375,
+      "learning_rate": 1.5943962641761176e-06,
+      "loss": 1.1528,
+      "step": 14761
+    },
+    {
+      "epoch": 13.627509808446803,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.5877251501000668e-06,
+      "loss": 1.0625,
+      "step": 14762
+    },
+    {
+      "epoch": 13.62843295638126,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.5810540360240158e-06,
+      "loss": 1.2424,
+      "step": 14763
+    },
+    {
+      "epoch": 13.629356104315717,
+      "grad_norm": 3.32421875,
+      "learning_rate": 1.5743829219479655e-06,
+      "loss": 1.1627,
+      "step": 14764
+    },
+    {
+      "epoch": 13.630279252250173,
+      "grad_norm": 2.6015625,
+      "learning_rate": 1.5677118078719147e-06,
+      "loss": 1.3447,
+      "step": 14765
+    },
+    {
+      "epoch": 13.63120240018463,
+      "grad_norm": 3.416015625,
+      "learning_rate": 1.561040693795864e-06,
+      "loss": 1.3474,
+      "step": 14766
+    },
+    {
+      "epoch": 13.632125548119086,
+      "grad_norm": 3.080078125,
+      "learning_rate": 1.5543695797198132e-06,
+      "loss": 1.2133,
+      "step": 14767
+    },
+    {
+      "epoch": 13.633048696053542,
+      "grad_norm": 6.40234375,
+      "learning_rate": 1.5476984656437627e-06,
+      "loss": 1.1606,
+      "step": 14768
+    },
+    {
+      "epoch": 13.633971843987998,
+      "grad_norm": 3.32421875,
+      "learning_rate": 1.541027351567712e-06,
+      "loss": 1.114,
+      "step": 14769
+    },
+    {
+      "epoch": 13.634894991922456,
+      "grad_norm": 3.322265625,
+      "learning_rate": 1.5343562374916612e-06,
+      "loss": 1.3326,
+      "step": 14770
+    },
+    {
+      "epoch": 13.635818139856912,
+      "grad_norm": 2.837890625,
+      "learning_rate": 1.5276851234156104e-06,
+      "loss": 1.2714,
+      "step": 14771
+    },
+    {
+      "epoch": 13.636741287791368,
+      "grad_norm": 3.18359375,
+      "learning_rate": 1.5210140093395599e-06,
+      "loss": 1.0756,
+      "step": 14772
+    },
+    {
+      "epoch": 13.637664435725824,
+      "grad_norm": 3.66015625,
+      "learning_rate": 1.514342895263509e-06,
+      "loss": 1.0939,
+      "step": 14773
+    },
+    {
+      "epoch": 13.638587583660282,
+      "grad_norm": 2.931640625,
+      "learning_rate": 1.5076717811874583e-06,
+      "loss": 1.3022,
+      "step": 14774
+    },
+    {
+      "epoch": 13.639510731594738,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.5010006671114078e-06,
+      "loss": 1.3301,
+      "step": 14775
+    },
+    {
+      "epoch": 13.640433879529194,
+      "grad_norm": 3.17578125,
+      "learning_rate": 1.494329553035357e-06,
+      "loss": 1.1274,
+      "step": 14776
+    },
+    {
+      "epoch": 13.641357027463652,
+      "grad_norm": 3.181640625,
+      "learning_rate": 1.4876584389593063e-06,
+      "loss": 1.3862,
+      "step": 14777
+    },
+    {
+      "epoch": 13.642280175398108,
+      "grad_norm": 2.82421875,
+      "learning_rate": 1.4809873248832555e-06,
+      "loss": 1.2795,
+      "step": 14778
+    },
+    {
+      "epoch": 13.643203323332564,
+      "grad_norm": 2.6484375,
+      "learning_rate": 1.474316210807205e-06,
+      "loss": 1.1932,
+      "step": 14779
+    },
+    {
+      "epoch": 13.64412647126702,
+      "grad_norm": 3.42578125,
+      "learning_rate": 1.4676450967311542e-06,
+      "loss": 1.014,
+      "step": 14780
+    },
+    {
+      "epoch": 13.645049619201478,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.4609739826551035e-06,
+      "loss": 1.0336,
+      "step": 14781
+    },
+    {
+      "epoch": 13.645972767135934,
+      "grad_norm": 2.97265625,
+      "learning_rate": 1.4543028685790527e-06,
+      "loss": 1.2561,
+      "step": 14782
+    },
+    {
+      "epoch": 13.64689591507039,
+      "grad_norm": 3.26171875,
+      "learning_rate": 1.4476317545030022e-06,
+      "loss": 1.384,
+      "step": 14783
+    },
+    {
+      "epoch": 13.647819063004846,
+      "grad_norm": 3.130859375,
+      "learning_rate": 1.4409606404269514e-06,
+      "loss": 1.2426,
+      "step": 14784
+    },
+    {
+      "epoch": 13.648742210939304,
+      "grad_norm": 2.705078125,
+      "learning_rate": 1.4342895263509006e-06,
+      "loss": 1.1938,
+      "step": 14785
+    },
+    {
+      "epoch": 13.64966535887376,
+      "grad_norm": 2.751953125,
+      "learning_rate": 1.42761841227485e-06,
+      "loss": 1.4041,
+      "step": 14786
+    },
+    {
+      "epoch": 13.650588506808216,
+      "grad_norm": 3.26171875,
+      "learning_rate": 1.4209472981987993e-06,
+      "loss": 1.3442,
+      "step": 14787
+    },
+    {
+      "epoch": 13.651511654742672,
+      "grad_norm": 3.0546875,
+      "learning_rate": 1.4142761841227486e-06,
+      "loss": 1.3101,
+      "step": 14788
+    },
+    {
+      "epoch": 13.65243480267713,
+      "grad_norm": 2.974609375,
+      "learning_rate": 1.4076050700466978e-06,
+      "loss": 1.2295,
+      "step": 14789
+    },
+    {
+      "epoch": 13.653357950611586,
+      "grad_norm": 3.341796875,
+      "learning_rate": 1.4009339559706473e-06,
+      "loss": 1.4963,
+      "step": 14790
+    },
+    {
+      "epoch": 13.654281098546042,
+      "grad_norm": 3.033203125,
+      "learning_rate": 1.3942628418945963e-06,
+      "loss": 1.0674,
+      "step": 14791
+    },
+    {
+      "epoch": 13.655204246480498,
+      "grad_norm": 3.09765625,
+      "learning_rate": 1.3875917278185458e-06,
+      "loss": 1.1895,
+      "step": 14792
+    },
+    {
+      "epoch": 13.656127394414955,
+      "grad_norm": 3.1484375,
+      "learning_rate": 1.380920613742495e-06,
+      "loss": 1.3484,
+      "step": 14793
+    },
+    {
+      "epoch": 13.657050542349412,
+      "grad_norm": 3.240234375,
+      "learning_rate": 1.3742494996664442e-06,
+      "loss": 1.2285,
+      "step": 14794
+    },
+    {
+      "epoch": 13.657973690283868,
+      "grad_norm": 3.16015625,
+      "learning_rate": 1.3675783855903937e-06,
+      "loss": 1.1409,
+      "step": 14795
+    },
+    {
+      "epoch": 13.658896838218325,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.360907271514343e-06,
+      "loss": 1.4741,
+      "step": 14796
+    },
+    {
+      "epoch": 13.659819986152781,
+      "grad_norm": 3.083984375,
+      "learning_rate": 1.3542361574382924e-06,
+      "loss": 0.9375,
+      "step": 14797
+    },
+    {
+      "epoch": 13.660743134087237,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.3475650433622414e-06,
+      "loss": 1.2781,
+      "step": 14798
+    },
+    {
+      "epoch": 13.661666282021693,
+      "grad_norm": 2.841796875,
+      "learning_rate": 1.3408939292861909e-06,
+      "loss": 1.6201,
+      "step": 14799
+    },
+    {
+      "epoch": 13.662589429956151,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.3342228152101401e-06,
+      "loss": 1.0594,
+      "step": 14800
+    },
+    {
+      "epoch": 13.663512577890607,
+      "grad_norm": 3.005859375,
+      "learning_rate": 1.3275517011340894e-06,
+      "loss": 1.2363,
+      "step": 14801
+    },
+    {
+      "epoch": 13.664435725825063,
+      "grad_norm": 3.1484375,
+      "learning_rate": 1.3208805870580386e-06,
+      "loss": 1.467,
+      "step": 14802
+    },
+    {
+      "epoch": 13.66535887375952,
+      "grad_norm": 2.89453125,
+      "learning_rate": 1.314209472981988e-06,
+      "loss": 1.3845,
+      "step": 14803
+    },
+    {
+      "epoch": 13.666282021693977,
+      "grad_norm": 2.923828125,
+      "learning_rate": 1.3075383589059375e-06,
+      "loss": 1.5444,
+      "step": 14804
+    },
+    {
+      "epoch": 13.667205169628433,
+      "grad_norm": 3.314453125,
+      "learning_rate": 1.3008672448298865e-06,
+      "loss": 1.1593,
+      "step": 14805
+    },
+    {
+      "epoch": 13.66812831756289,
+      "grad_norm": 3.380859375,
+      "learning_rate": 1.294196130753836e-06,
+      "loss": 1.4045,
+      "step": 14806
+    },
+    {
+      "epoch": 13.669051465497345,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2875250166777852e-06,
+      "loss": 1.1428,
+      "step": 14807
+    },
+    {
+      "epoch": 13.669974613431803,
+      "grad_norm": 3.033203125,
+      "learning_rate": 1.2808539026017345e-06,
+      "loss": 1.1549,
+      "step": 14808
+    },
+    {
+      "epoch": 13.670897761366259,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.2741827885256837e-06,
+      "loss": 1.3105,
+      "step": 14809
+    },
+    {
+      "epoch": 13.671820909300715,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.2675116744496332e-06,
+      "loss": 1.2417,
+      "step": 14810
+    },
+    {
+      "epoch": 13.672744057235171,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.2608405603735824e-06,
+      "loss": 1.3081,
+      "step": 14811
+    },
+    {
+      "epoch": 13.673667205169629,
+      "grad_norm": 3.01171875,
+      "learning_rate": 1.2541694462975317e-06,
+      "loss": 1.1375,
+      "step": 14812
+    },
+    {
+      "epoch": 13.674590353104085,
+      "grad_norm": 3.052734375,
+      "learning_rate": 1.2474983322214811e-06,
+      "loss": 1.3218,
+      "step": 14813
+    },
+    {
+      "epoch": 13.675513501038541,
+      "grad_norm": 3.54296875,
+      "learning_rate": 1.2408272181454304e-06,
+      "loss": 1.0469,
+      "step": 14814
+    },
+    {
+      "epoch": 13.676436648972999,
+      "grad_norm": 3.205078125,
+      "learning_rate": 1.2341561040693796e-06,
+      "loss": 1.1309,
+      "step": 14815
+    },
+    {
+      "epoch": 13.677359796907455,
+      "grad_norm": 3.052734375,
+      "learning_rate": 1.2274849899933288e-06,
+      "loss": 1.3308,
+      "step": 14816
+    },
+    {
+      "epoch": 13.67828294484191,
+      "grad_norm": 3.2109375,
+      "learning_rate": 1.2208138759172783e-06,
+      "loss": 1.3013,
+      "step": 14817
+    },
+    {
+      "epoch": 13.679206092776367,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2141427618412275e-06,
+      "loss": 1.2449,
+      "step": 14818
+    },
+    {
+      "epoch": 13.680129240710825,
+      "grad_norm": 2.962890625,
+      "learning_rate": 1.2074716477651768e-06,
+      "loss": 1.2888,
+      "step": 14819
+    },
+    {
+      "epoch": 13.68105238864528,
+      "grad_norm": 3.02734375,
+      "learning_rate": 1.200800533689126e-06,
+      "loss": 1.1973,
+      "step": 14820
+    },
+    {
+      "epoch": 13.681975536579737,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.1941294196130755e-06,
+      "loss": 1.2688,
+      "step": 14821
+    },
+    {
+      "epoch": 13.682898684514193,
+      "grad_norm": 3.17578125,
+      "learning_rate": 1.1874583055370247e-06,
+      "loss": 1.1636,
+      "step": 14822
+    },
+    {
+      "epoch": 13.68382183244865,
+      "grad_norm": 3.1953125,
+      "learning_rate": 1.180787191460974e-06,
+      "loss": 1.1921,
+      "step": 14823
+    },
+    {
+      "epoch": 13.684744980383106,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.1741160773849234e-06,
+      "loss": 1.2595,
+      "step": 14824
+    },
+    {
+      "epoch": 13.685668128317563,
+      "grad_norm": 3.517578125,
+      "learning_rate": 1.1674449633088727e-06,
+      "loss": 1.4551,
+      "step": 14825
+    },
+    {
+      "epoch": 13.68659127625202,
+      "grad_norm": 3.072265625,
+      "learning_rate": 1.160773849232822e-06,
+      "loss": 1.4058,
+      "step": 14826
+    },
+    {
+      "epoch": 13.687514424186476,
+      "grad_norm": 3.103515625,
+      "learning_rate": 1.1541027351567711e-06,
+      "loss": 1.4939,
+      "step": 14827
+    },
+    {
+      "epoch": 13.688437572120932,
+      "grad_norm": 3.130859375,
+      "learning_rate": 1.1474316210807206e-06,
+      "loss": 1.3218,
+      "step": 14828
+    },
+    {
+      "epoch": 13.689360720055388,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.1407605070046698e-06,
+      "loss": 1.1514,
+      "step": 14829
+    },
+    {
+      "epoch": 13.690283867989846,
+      "grad_norm": 2.962890625,
+      "learning_rate": 1.134089392928619e-06,
+      "loss": 1.426,
+      "step": 14830
+    },
+    {
+      "epoch": 13.691207015924302,
+      "grad_norm": 3.279296875,
+      "learning_rate": 1.1274182788525685e-06,
+      "loss": 0.8996,
+      "step": 14831
+    },
+    {
+      "epoch": 13.692130163858758,
+      "grad_norm": 2.91015625,
+      "learning_rate": 1.1207471647765178e-06,
+      "loss": 1.3596,
+      "step": 14832
+    },
+    {
+      "epoch": 13.693053311793214,
+      "grad_norm": 3.982421875,
+      "learning_rate": 1.114076050700467e-06,
+      "loss": 1.2722,
+      "step": 14833
+    },
+    {
+      "epoch": 13.693976459727672,
+      "grad_norm": 3.236328125,
+      "learning_rate": 1.1074049366244163e-06,
+      "loss": 1.2529,
+      "step": 14834
+    },
+    {
+      "epoch": 13.694899607662128,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.1007338225483657e-06,
+      "loss": 1.342,
+      "step": 14835
+    },
+    {
+      "epoch": 13.695822755596584,
+      "grad_norm": 3.322265625,
+      "learning_rate": 1.094062708472315e-06,
+      "loss": 1.2244,
+      "step": 14836
+    },
+    {
+      "epoch": 13.69674590353104,
+      "grad_norm": 3.083984375,
+      "learning_rate": 1.0873915943962642e-06,
+      "loss": 1.2676,
+      "step": 14837
+    },
+    {
+      "epoch": 13.697669051465498,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0807204803202134e-06,
+      "loss": 1.3423,
+      "step": 14838
+    },
+    {
+      "epoch": 13.698592199399954,
+      "grad_norm": 3.154296875,
+      "learning_rate": 1.074049366244163e-06,
+      "loss": 1.3752,
+      "step": 14839
+    },
+    {
+      "epoch": 13.69951534733441,
+      "grad_norm": 3.630859375,
+      "learning_rate": 1.0673782521681121e-06,
+      "loss": 1.3381,
+      "step": 14840
+    },
+    {
+      "epoch": 13.700438495268866,
+      "grad_norm": 3.529296875,
+      "learning_rate": 1.0607071380920614e-06,
+      "loss": 1.4875,
+      "step": 14841
+    },
+    {
+      "epoch": 13.701361643203324,
+      "grad_norm": 3.4296875,
+      "learning_rate": 1.0540360240160108e-06,
+      "loss": 1.2795,
+      "step": 14842
+    },
+    {
+      "epoch": 13.70228479113778,
+      "grad_norm": 2.962890625,
+      "learning_rate": 1.04736490993996e-06,
+      "loss": 1.2456,
+      "step": 14843
+    },
+    {
+      "epoch": 13.703207939072236,
+      "grad_norm": 3.3828125,
+      "learning_rate": 1.0406937958639093e-06,
+      "loss": 1.2649,
+      "step": 14844
+    },
+    {
+      "epoch": 13.704131087006694,
+      "grad_norm": 3.02734375,
+      "learning_rate": 1.0340226817878586e-06,
+      "loss": 1.3076,
+      "step": 14845
+    },
+    {
+      "epoch": 13.70505423494115,
+      "grad_norm": 3.8359375,
+      "learning_rate": 1.027351567711808e-06,
+      "loss": 0.9589,
+      "step": 14846
+    },
+    {
+      "epoch": 13.705977382875606,
+      "grad_norm": 3.349609375,
+      "learning_rate": 1.0206804536357573e-06,
+      "loss": 1.0261,
+      "step": 14847
+    },
+    {
+      "epoch": 13.706900530810062,
+      "grad_norm": 2.900390625,
+      "learning_rate": 1.0140093395597065e-06,
+      "loss": 1.4229,
+      "step": 14848
+    },
+    {
+      "epoch": 13.70782367874452,
+      "grad_norm": 2.83203125,
+      "learning_rate": 1.0073382254836557e-06,
+      "loss": 1.1736,
+      "step": 14849
+    },
+    {
+      "epoch": 13.708746826678976,
+      "grad_norm": 2.837890625,
+      "learning_rate": 1.0006671114076052e-06,
+      "loss": 1.2478,
+      "step": 14850
+    },
+    {
+      "epoch": 13.709669974613432,
+      "grad_norm": 3.10546875,
+      "learning_rate": 9.939959973315544e-07,
+      "loss": 1.063,
+      "step": 14851
+    },
+    {
+      "epoch": 13.710593122547888,
+      "grad_norm": 3.271484375,
+      "learning_rate": 9.873248832555037e-07,
+      "loss": 1.4272,
+      "step": 14852
+    },
+    {
+      "epoch": 13.711516270482345,
+      "grad_norm": 2.947265625,
+      "learning_rate": 9.806537691794531e-07,
+      "loss": 1.214,
+      "step": 14853
+    },
+    {
+      "epoch": 13.712439418416801,
+      "grad_norm": 3.080078125,
+      "learning_rate": 9.739826551034022e-07,
+      "loss": 1.325,
+      "step": 14854
+    },
+    {
+      "epoch": 13.713362566351257,
+      "grad_norm": 3.080078125,
+      "learning_rate": 9.673115410273516e-07,
+      "loss": 1.3032,
+      "step": 14855
+    },
+    {
+      "epoch": 13.714285714285714,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.606404269513009e-07,
+      "loss": 1.272,
+      "step": 14856
+    },
+    {
+      "epoch": 13.715208862220171,
+      "grad_norm": 2.822265625,
+      "learning_rate": 9.539693128752503e-07,
+      "loss": 1.217,
+      "step": 14857
+    },
+    {
+      "epoch": 13.716132010154627,
+      "grad_norm": 3.05859375,
+      "learning_rate": 9.472981987991995e-07,
+      "loss": 1.1985,
+      "step": 14858
+    },
+    {
+      "epoch": 13.717055158089083,
+      "grad_norm": 3.44140625,
+      "learning_rate": 9.406270847231488e-07,
+      "loss": 1.2712,
+      "step": 14859
+    },
+    {
+      "epoch": 13.71797830602354,
+      "grad_norm": 3.033203125,
+      "learning_rate": 9.339559706470982e-07,
+      "loss": 1.3303,
+      "step": 14860
+    },
+    {
+      "epoch": 13.718901453957997,
+      "grad_norm": 3.32421875,
+      "learning_rate": 9.272848565710474e-07,
+      "loss": 1.2583,
+      "step": 14861
+    },
+    {
+      "epoch": 13.719824601892453,
+      "grad_norm": 3.212890625,
+      "learning_rate": 9.206137424949967e-07,
+      "loss": 1.3771,
+      "step": 14862
+    },
+    {
+      "epoch": 13.72074774982691,
+      "grad_norm": 3.017578125,
+      "learning_rate": 9.13942628418946e-07,
+      "loss": 1.1378,
+      "step": 14863
+    },
+    {
+      "epoch": 13.721670897761367,
+      "grad_norm": 2.92578125,
+      "learning_rate": 9.072715143428953e-07,
+      "loss": 1.2656,
+      "step": 14864
+    },
+    {
+      "epoch": 13.722594045695823,
+      "grad_norm": 2.705078125,
+      "learning_rate": 9.006004002668446e-07,
+      "loss": 1.3317,
+      "step": 14865
+    },
+    {
+      "epoch": 13.723517193630279,
+      "grad_norm": 2.966796875,
+      "learning_rate": 8.939292861907939e-07,
+      "loss": 1.1152,
+      "step": 14866
+    },
+    {
+      "epoch": 13.724440341564735,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.872581721147432e-07,
+      "loss": 1.1775,
+      "step": 14867
+    },
+    {
+      "epoch": 13.725363489499193,
+      "grad_norm": 3.26953125,
+      "learning_rate": 8.805870580386925e-07,
+      "loss": 1.2639,
+      "step": 14868
+    },
+    {
+      "epoch": 13.726286637433649,
+      "grad_norm": 3.025390625,
+      "learning_rate": 8.739159439626419e-07,
+      "loss": 1.229,
+      "step": 14869
+    },
+    {
+      "epoch": 13.727209785368105,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.672448298865911e-07,
+      "loss": 1.2421,
+      "step": 14870
+    },
+    {
+      "epoch": 13.728132933302561,
+      "grad_norm": 3.12890625,
+      "learning_rate": 8.605737158105405e-07,
+      "loss": 1.2866,
+      "step": 14871
+    },
+    {
+      "epoch": 13.729056081237019,
+      "grad_norm": 3.193359375,
+      "learning_rate": 8.539026017344897e-07,
+      "loss": 1.141,
+      "step": 14872
+    },
+    {
+      "epoch": 13.729979229171475,
+      "grad_norm": 2.826171875,
+      "learning_rate": 8.47231487658439e-07,
+      "loss": 1.1655,
+      "step": 14873
+    },
+    {
+      "epoch": 13.73090237710593,
+      "grad_norm": 3.044921875,
+      "learning_rate": 8.405603735823883e-07,
+      "loss": 1.2539,
+      "step": 14874
+    },
+    {
+      "epoch": 13.731825525040387,
+      "grad_norm": 3.404296875,
+      "learning_rate": 8.338892595063376e-07,
+      "loss": 1.2098,
+      "step": 14875
+    },
+    {
+      "epoch": 13.732748672974845,
+      "grad_norm": 3.5234375,
+      "learning_rate": 8.272181454302869e-07,
+      "loss": 1.1982,
+      "step": 14876
+    },
+    {
+      "epoch": 13.7336718209093,
+      "grad_norm": 3.623046875,
+      "learning_rate": 8.205470313542362e-07,
+      "loss": 1.2422,
+      "step": 14877
+    },
+    {
+      "epoch": 13.734594968843757,
+      "grad_norm": 3.064453125,
+      "learning_rate": 8.138759172781856e-07,
+      "loss": 1.2271,
+      "step": 14878
+    },
+    {
+      "epoch": 13.735518116778213,
+      "grad_norm": 3.650390625,
+      "learning_rate": 8.072048032021348e-07,
+      "loss": 1.179,
+      "step": 14879
+    },
+    {
+      "epoch": 13.73644126471267,
+      "grad_norm": 3.638671875,
+      "learning_rate": 8.005336891260842e-07,
+      "loss": 1.3159,
+      "step": 14880
+    },
+    {
+      "epoch": 13.737364412647127,
+      "grad_norm": 2.87109375,
+      "learning_rate": 7.938625750500334e-07,
+      "loss": 0.9206,
+      "step": 14881
+    },
+    {
+      "epoch": 13.738287560581583,
+      "grad_norm": 3.27734375,
+      "learning_rate": 7.871914609739828e-07,
+      "loss": 1.1455,
+      "step": 14882
+    },
+    {
+      "epoch": 13.73921070851604,
+      "grad_norm": 3.375,
+      "learning_rate": 7.80520346897932e-07,
+      "loss": 1.323,
+      "step": 14883
+    },
+    {
+      "epoch": 13.740133856450496,
+      "grad_norm": 3.044921875,
+      "learning_rate": 7.738492328218813e-07,
+      "loss": 0.9618,
+      "step": 14884
+    },
+    {
+      "epoch": 13.741057004384952,
+      "grad_norm": 2.875,
+      "learning_rate": 7.671781187458306e-07,
+      "loss": 1.3501,
+      "step": 14885
+    },
+    {
+      "epoch": 13.741980152319409,
+      "grad_norm": 3.01953125,
+      "learning_rate": 7.605070046697799e-07,
+      "loss": 1.21,
+      "step": 14886
+    },
+    {
+      "epoch": 13.742903300253866,
+      "grad_norm": 3.486328125,
+      "learning_rate": 7.538358905937292e-07,
+      "loss": 1.1102,
+      "step": 14887
+    },
+    {
+      "epoch": 13.743826448188322,
+      "grad_norm": 2.85546875,
+      "learning_rate": 7.471647765176785e-07,
+      "loss": 1.3022,
+      "step": 14888
+    },
+    {
+      "epoch": 13.744749596122778,
+      "grad_norm": 3.044921875,
+      "learning_rate": 7.404936624416278e-07,
+      "loss": 1.1069,
+      "step": 14889
+    },
+    {
+      "epoch": 13.745672744057234,
+      "grad_norm": 3.25,
+      "learning_rate": 7.338225483655771e-07,
+      "loss": 1.4141,
+      "step": 14890
+    },
+    {
+      "epoch": 13.746595891991692,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.271514342895264e-07,
+      "loss": 1.1322,
+      "step": 14891
+    },
+    {
+      "epoch": 13.747519039926148,
+      "grad_norm": 3.119140625,
+      "learning_rate": 7.204803202134757e-07,
+      "loss": 1.1721,
+      "step": 14892
+    },
+    {
+      "epoch": 13.748442187860604,
+      "grad_norm": 3.123046875,
+      "learning_rate": 7.13809206137425e-07,
+      "loss": 1.2539,
+      "step": 14893
+    },
+    {
+      "epoch": 13.749365335795062,
+      "grad_norm": 2.98828125,
+      "learning_rate": 7.071380920613743e-07,
+      "loss": 1.3378,
+      "step": 14894
+    },
+    {
+      "epoch": 13.750288483729518,
+      "grad_norm": 2.919921875,
+      "learning_rate": 7.004669779853236e-07,
+      "loss": 1.3693,
+      "step": 14895
+    },
+    {
+      "epoch": 13.751211631663974,
+      "grad_norm": 2.822265625,
+      "learning_rate": 6.937958639092729e-07,
+      "loss": 1.2532,
+      "step": 14896
+    },
+    {
+      "epoch": 13.75213477959843,
+      "grad_norm": 3.07421875,
+      "learning_rate": 6.871247498332221e-07,
+      "loss": 1.1102,
+      "step": 14897
+    },
+    {
+      "epoch": 13.753057927532888,
+      "grad_norm": 3.693359375,
+      "learning_rate": 6.804536357571715e-07,
+      "loss": 1.2516,
+      "step": 14898
+    },
+    {
+      "epoch": 13.753981075467344,
+      "grad_norm": 3.095703125,
+      "learning_rate": 6.737825216811207e-07,
+      "loss": 1.2646,
+      "step": 14899
+    },
+    {
+      "epoch": 13.7549042234018,
+      "grad_norm": 3.197265625,
+      "learning_rate": 6.671114076050701e-07,
+      "loss": 1.0741,
+      "step": 14900
+    },
+    {
+      "epoch": 13.755827371336256,
+      "grad_norm": 3.41796875,
+      "learning_rate": 6.604402935290193e-07,
+      "loss": 1.209,
+      "step": 14901
+    },
+    {
+      "epoch": 13.756750519270714,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.537691794529688e-07,
+      "loss": 1.3767,
+      "step": 14902
+    },
+    {
+      "epoch": 13.75767366720517,
+      "grad_norm": 3.158203125,
+      "learning_rate": 6.47098065376918e-07,
+      "loss": 1.4548,
+      "step": 14903
+    },
+    {
+      "epoch": 13.758596815139626,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.404269513008672e-07,
+      "loss": 1.4694,
+      "step": 14904
+    },
+    {
+      "epoch": 13.759519963074082,
+      "grad_norm": 3.173828125,
+      "learning_rate": 6.337558372248166e-07,
+      "loss": 1.2649,
+      "step": 14905
+    },
+    {
+      "epoch": 13.76044311100854,
+      "grad_norm": 3.525390625,
+      "learning_rate": 6.270847231487658e-07,
+      "loss": 0.9916,
+      "step": 14906
+    },
+    {
+      "epoch": 13.761366258942996,
+      "grad_norm": 3.189453125,
+      "learning_rate": 6.204136090727152e-07,
+      "loss": 0.9954,
+      "step": 14907
+    },
+    {
+      "epoch": 13.762289406877452,
+      "grad_norm": 2.89453125,
+      "learning_rate": 6.137424949966644e-07,
+      "loss": 1.3357,
+      "step": 14908
+    },
+    {
+      "epoch": 13.763212554811908,
+      "grad_norm": 2.814453125,
+      "learning_rate": 6.070713809206138e-07,
+      "loss": 1.3716,
+      "step": 14909
+    },
+    {
+      "epoch": 13.764135702746366,
+      "grad_norm": 3.26953125,
+      "learning_rate": 6.00400266844563e-07,
+      "loss": 1.1565,
+      "step": 14910
+    },
+    {
+      "epoch": 13.765058850680822,
+      "grad_norm": 3.201171875,
+      "learning_rate": 5.937291527685124e-07,
+      "loss": 1.3284,
+      "step": 14911
+    },
+    {
+      "epoch": 13.765981998615278,
+      "grad_norm": 3.193359375,
+      "learning_rate": 5.870580386924617e-07,
+      "loss": 1.1664,
+      "step": 14912
+    },
+    {
+      "epoch": 13.766905146549735,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.80386924616411e-07,
+      "loss": 1.1906,
+      "step": 14913
+    },
+    {
+      "epoch": 13.767828294484191,
+      "grad_norm": 2.759765625,
+      "learning_rate": 5.737158105403603e-07,
+      "loss": 0.8841,
+      "step": 14914
+    },
+    {
+      "epoch": 13.768751442418647,
+      "grad_norm": 3.025390625,
+      "learning_rate": 5.670446964643095e-07,
+      "loss": 1.1385,
+      "step": 14915
+    },
+    {
+      "epoch": 13.769674590353103,
+      "grad_norm": 3.1953125,
+      "learning_rate": 5.603735823882589e-07,
+      "loss": 1.446,
+      "step": 14916
+    },
+    {
+      "epoch": 13.770597738287561,
+      "grad_norm": 2.96484375,
+      "learning_rate": 5.537024683122081e-07,
+      "loss": 1.1753,
+      "step": 14917
+    },
+    {
+      "epoch": 13.771520886222017,
+      "grad_norm": 3.0234375,
+      "learning_rate": 5.470313542361575e-07,
+      "loss": 1.2754,
+      "step": 14918
+    },
+    {
+      "epoch": 13.772444034156473,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.403602401601067e-07,
+      "loss": 1.0724,
+      "step": 14919
+    },
+    {
+      "epoch": 13.77336718209093,
+      "grad_norm": 3.076171875,
+      "learning_rate": 5.336891260840561e-07,
+      "loss": 1.5435,
+      "step": 14920
+    },
+    {
+      "epoch": 13.774290330025387,
+      "grad_norm": 3.09765625,
+      "learning_rate": 5.270180120080054e-07,
+      "loss": 1.2507,
+      "step": 14921
+    },
+    {
+      "epoch": 13.775213477959843,
+      "grad_norm": 3.76171875,
+      "learning_rate": 5.203468979319547e-07,
+      "loss": 1.2668,
+      "step": 14922
+    },
+    {
+      "epoch": 13.7761366258943,
+      "grad_norm": 3.0703125,
+      "learning_rate": 5.13675783855904e-07,
+      "loss": 1.0994,
+      "step": 14923
+    },
+    {
+      "epoch": 13.777059773828755,
+      "grad_norm": 3.0859375,
+      "learning_rate": 5.070046697798533e-07,
+      "loss": 1.3381,
+      "step": 14924
+    },
+    {
+      "epoch": 13.777982921763213,
+      "grad_norm": 2.974609375,
+      "learning_rate": 5.003335557038026e-07,
+      "loss": 1.2693,
+      "step": 14925
+    },
+    {
+      "epoch": 13.778906069697669,
+      "grad_norm": 2.8984375,
+      "learning_rate": 4.936624416277518e-07,
+      "loss": 1.3035,
+      "step": 14926
+    },
+    {
+      "epoch": 13.779829217632125,
+      "grad_norm": 3.0390625,
+      "learning_rate": 4.869913275517011e-07,
+      "loss": 1.1537,
+      "step": 14927
+    },
+    {
+      "epoch": 13.780752365566581,
+      "grad_norm": 2.8828125,
+      "learning_rate": 4.803202134756504e-07,
+      "loss": 1.0222,
+      "step": 14928
+    },
+    {
+      "epoch": 13.781675513501039,
+      "grad_norm": 3.009765625,
+      "learning_rate": 4.7364909939959973e-07,
+      "loss": 1.3151,
+      "step": 14929
+    },
+    {
+      "epoch": 13.782598661435495,
+      "grad_norm": 2.943359375,
+      "learning_rate": 4.669779853235491e-07,
+      "loss": 1.4163,
+      "step": 14930
+    },
+    {
+      "epoch": 13.783521809369951,
+      "grad_norm": 3.115234375,
+      "learning_rate": 4.6030687124749837e-07,
+      "loss": 0.9012,
+      "step": 14931
+    },
+    {
+      "epoch": 13.784444957304409,
+      "grad_norm": 3.349609375,
+      "learning_rate": 4.5363575717144767e-07,
+      "loss": 1.2515,
+      "step": 14932
+    },
+    {
+      "epoch": 13.785368105238865,
+      "grad_norm": 3.3203125,
+      "learning_rate": 4.4696464309539696e-07,
+      "loss": 1.1743,
+      "step": 14933
+    },
+    {
+      "epoch": 13.78629125317332,
+      "grad_norm": 3.029296875,
+      "learning_rate": 4.4029352901934626e-07,
+      "loss": 1.1936,
+      "step": 14934
+    },
+    {
+      "epoch": 13.787214401107777,
+      "grad_norm": 3.181640625,
+      "learning_rate": 4.3362241494329555e-07,
+      "loss": 1.324,
+      "step": 14935
+    },
+    {
+      "epoch": 13.788137549042235,
+      "grad_norm": 3.267578125,
+      "learning_rate": 4.2695130086724485e-07,
+      "loss": 1.343,
+      "step": 14936
+    },
+    {
+      "epoch": 13.78906069697669,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.2028018679119414e-07,
+      "loss": 1.3428,
+      "step": 14937
+    },
+    {
+      "epoch": 13.789983844911147,
+      "grad_norm": 3.3046875,
+      "learning_rate": 4.1360907271514344e-07,
+      "loss": 1.1895,
+      "step": 14938
+    },
+    {
+      "epoch": 13.790906992845603,
+      "grad_norm": 3.06640625,
+      "learning_rate": 4.069379586390928e-07,
+      "loss": 1.1924,
+      "step": 14939
+    },
+    {
+      "epoch": 13.79183014078006,
+      "grad_norm": 3.212890625,
+      "learning_rate": 4.002668445630421e-07,
+      "loss": 1.1759,
+      "step": 14940
+    },
+    {
+      "epoch": 13.792753288714517,
+      "grad_norm": 2.958984375,
+      "learning_rate": 3.935957304869914e-07,
+      "loss": 1.3219,
+      "step": 14941
+    },
+    {
+      "epoch": 13.793676436648973,
+      "grad_norm": 3.68359375,
+      "learning_rate": 3.8692461641094067e-07,
+      "loss": 0.9993,
+      "step": 14942
+    },
+    {
+      "epoch": 13.79459958458343,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.8025350233488997e-07,
+      "loss": 1.252,
+      "step": 14943
+    },
+    {
+      "epoch": 13.795522732517886,
+      "grad_norm": 3.134765625,
+      "learning_rate": 3.7358238825883926e-07,
+      "loss": 1.2814,
+      "step": 14944
+    },
+    {
+      "epoch": 13.796445880452342,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.6691127418278856e-07,
+      "loss": 1.4014,
+      "step": 14945
+    },
+    {
+      "epoch": 13.797369028386798,
+      "grad_norm": 3.224609375,
+      "learning_rate": 3.6024016010673785e-07,
+      "loss": 1.2013,
+      "step": 14946
+    },
+    {
+      "epoch": 13.798292176321256,
+      "grad_norm": 3.580078125,
+      "learning_rate": 3.5356904603068715e-07,
+      "loss": 1.2263,
+      "step": 14947
+    },
+    {
+      "epoch": 13.799215324255712,
+      "grad_norm": 2.8203125,
+      "learning_rate": 3.4689793195463644e-07,
+      "loss": 1.3689,
+      "step": 14948
+    },
+    {
+      "epoch": 13.800138472190168,
+      "grad_norm": 3.048828125,
+      "learning_rate": 3.4022681787858574e-07,
+      "loss": 1.238,
+      "step": 14949
+    },
+    {
+      "epoch": 13.801061620124624,
+      "grad_norm": 2.998046875,
+      "learning_rate": 3.3355570380253503e-07,
+      "loss": 1.1469,
+      "step": 14950
+    },
+    {
+      "epoch": 13.801984768059082,
+      "grad_norm": 2.76953125,
+      "learning_rate": 3.268845897264844e-07,
+      "loss": 1.2979,
+      "step": 14951
+    },
+    {
+      "epoch": 13.802907915993538,
+      "grad_norm": 3.87109375,
+      "learning_rate": 3.202134756504336e-07,
+      "loss": 1.3074,
+      "step": 14952
+    },
+    {
+      "epoch": 13.803831063927994,
+      "grad_norm": 3.25390625,
+      "learning_rate": 3.135423615743829e-07,
+      "loss": 1.2373,
+      "step": 14953
+    },
+    {
+      "epoch": 13.80475421186245,
+      "grad_norm": 3.12109375,
+      "learning_rate": 3.068712474983322e-07,
+      "loss": 0.9799,
+      "step": 14954
+    },
+    {
+      "epoch": 13.805677359796908,
+      "grad_norm": 3.36328125,
+      "learning_rate": 3.002001334222815e-07,
+      "loss": 1.4048,
+      "step": 14955
+    },
+    {
+      "epoch": 13.806600507731364,
+      "grad_norm": 3.353515625,
+      "learning_rate": 2.9352901934623086e-07,
+      "loss": 1.2922,
+      "step": 14956
+    },
+    {
+      "epoch": 13.80752365566582,
+      "grad_norm": 2.53515625,
+      "learning_rate": 2.8685790527018015e-07,
+      "loss": 1.1176,
+      "step": 14957
+    },
+    {
+      "epoch": 13.808446803600276,
+      "grad_norm": 3.099609375,
+      "learning_rate": 2.8018679119412945e-07,
+      "loss": 1.3108,
+      "step": 14958
+    },
+    {
+      "epoch": 13.809369951534734,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7351567711807874e-07,
+      "loss": 1.1622,
+      "step": 14959
+    },
+    {
+      "epoch": 13.81029309946919,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.6684456304202804e-07,
+      "loss": 1.3574,
+      "step": 14960
+    },
+    {
+      "epoch": 13.811216247403646,
+      "grad_norm": 2.998046875,
+      "learning_rate": 2.6017344896597733e-07,
+      "loss": 1.2676,
+      "step": 14961
+    },
+    {
+      "epoch": 13.812139395338104,
+      "grad_norm": 2.626953125,
+      "learning_rate": 2.535023348899266e-07,
+      "loss": 1.2048,
+      "step": 14962
+    },
+    {
+      "epoch": 13.81306254327256,
+      "grad_norm": 3.072265625,
+      "learning_rate": 2.468312208138759e-07,
+      "loss": 1.2671,
+      "step": 14963
+    },
+    {
+      "epoch": 13.813985691207016,
+      "grad_norm": 3.228515625,
+      "learning_rate": 2.401601067378252e-07,
+      "loss": 1.2739,
+      "step": 14964
+    },
+    {
+      "epoch": 13.814908839141472,
+      "grad_norm": 3.017578125,
+      "learning_rate": 2.3348899266177454e-07,
+      "loss": 1.078,
+      "step": 14965
+    },
+    {
+      "epoch": 13.81583198707593,
+      "grad_norm": 3.51953125,
+      "learning_rate": 2.2681787858572383e-07,
+      "loss": 1.3542,
+      "step": 14966
+    },
+    {
+      "epoch": 13.816755135010386,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.2014676450967313e-07,
+      "loss": 1.4678,
+      "step": 14967
+    },
+    {
+      "epoch": 13.817678282944842,
+      "grad_norm": 3.634765625,
+      "learning_rate": 2.1347565043362242e-07,
+      "loss": 1.1473,
+      "step": 14968
+    },
+    {
+      "epoch": 13.818601430879298,
+      "grad_norm": 3.3515625,
+      "learning_rate": 2.0680453635757172e-07,
+      "loss": 1.3757,
+      "step": 14969
+    },
+    {
+      "epoch": 13.819524578813756,
+      "grad_norm": 3.447265625,
+      "learning_rate": 2.0013342228152104e-07,
+      "loss": 1.0769,
+      "step": 14970
+    },
+    {
+      "epoch": 13.820447726748212,
+      "grad_norm": 2.869140625,
+      "learning_rate": 1.9346230820547034e-07,
+      "loss": 1.3009,
+      "step": 14971
+    },
+    {
+      "epoch": 13.821370874682668,
+      "grad_norm": 3.16015625,
+      "learning_rate": 1.8679119412941963e-07,
+      "loss": 1.2571,
+      "step": 14972
+    },
+    {
+      "epoch": 13.822294022617124,
+      "grad_norm": 3.05859375,
+      "learning_rate": 1.8012008005336893e-07,
+      "loss": 1.4604,
+      "step": 14973
+    },
+    {
+      "epoch": 13.823217170551581,
+      "grad_norm": 2.95703125,
+      "learning_rate": 1.7344896597731822e-07,
+      "loss": 1.3553,
+      "step": 14974
+    },
+    {
+      "epoch": 13.824140318486037,
+      "grad_norm": 3.4453125,
+      "learning_rate": 1.6677785190126752e-07,
+      "loss": 1.292,
+      "step": 14975
+    },
+    {
+      "epoch": 13.825063466420493,
+      "grad_norm": 2.818359375,
+      "learning_rate": 1.601067378252168e-07,
+      "loss": 1.2405,
+      "step": 14976
+    },
+    {
+      "epoch": 13.82598661435495,
+      "grad_norm": 3.076171875,
+      "learning_rate": 1.534356237491661e-07,
+      "loss": 1.2522,
+      "step": 14977
+    },
+    {
+      "epoch": 13.826909762289407,
+      "grad_norm": 3.12109375,
+      "learning_rate": 1.4676450967311543e-07,
+      "loss": 0.905,
+      "step": 14978
+    },
+    {
+      "epoch": 13.827832910223863,
+      "grad_norm": 3.123046875,
+      "learning_rate": 1.4009339559706472e-07,
+      "loss": 1.2693,
+      "step": 14979
+    },
+    {
+      "epoch": 13.82875605815832,
+      "grad_norm": 3.00390625,
+      "learning_rate": 1.3342228152101402e-07,
+      "loss": 1.2103,
+      "step": 14980
+    },
+    {
+      "epoch": 13.829679206092777,
+      "grad_norm": 3.384765625,
+      "learning_rate": 1.267511674449633e-07,
+      "loss": 1.3911,
+      "step": 14981
+    },
+    {
+      "epoch": 13.830602354027233,
+      "grad_norm": 3.111328125,
+      "learning_rate": 1.200800533689126e-07,
+      "loss": 1.2632,
+      "step": 14982
+    },
+    {
+      "epoch": 13.83152550196169,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.1340893929286192e-07,
+      "loss": 1.4126,
+      "step": 14983
+    },
+    {
+      "epoch": 13.832448649896145,
+      "grad_norm": 3.353515625,
+      "learning_rate": 1.0673782521681121e-07,
+      "loss": 1.2424,
+      "step": 14984
+    },
+    {
+      "epoch": 13.833371797830603,
+      "grad_norm": 2.947265625,
+      "learning_rate": 1.0006671114076052e-07,
+      "loss": 1.1925,
+      "step": 14985
+    },
+    {
+      "epoch": 13.834294945765059,
+      "grad_norm": 3.091796875,
+      "learning_rate": 9.339559706470982e-08,
+      "loss": 1.2665,
+      "step": 14986
+    },
+    {
+      "epoch": 13.835218093699515,
+      "grad_norm": 2.91015625,
+      "learning_rate": 8.672448298865911e-08,
+      "loss": 1.3105,
+      "step": 14987
+    },
+    {
+      "epoch": 13.836141241633971,
+      "grad_norm": 2.962890625,
+      "learning_rate": 8.00533689126084e-08,
+      "loss": 1.2395,
+      "step": 14988
+    },
+    {
+      "epoch": 13.837064389568429,
+      "grad_norm": 3.3515625,
+      "learning_rate": 7.338225483655771e-08,
+      "loss": 1.2917,
+      "step": 14989
+    },
+    {
+      "epoch": 13.837987537502885,
+      "grad_norm": 3.4140625,
+      "learning_rate": 6.671114076050701e-08,
+      "loss": 1.2729,
+      "step": 14990
+    },
+    {
+      "epoch": 13.838910685437341,
+      "grad_norm": 3.056640625,
+      "learning_rate": 6.00400266844563e-08,
+      "loss": 1.1812,
+      "step": 14991
+    },
+    {
+      "epoch": 13.839833833371799,
+      "grad_norm": 2.99609375,
+      "learning_rate": 5.3368912608405606e-08,
+      "loss": 0.9284,
+      "step": 14992
+    },
+    {
+      "epoch": 13.840756981306255,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.669779853235491e-08,
+      "loss": 1.3263,
+      "step": 14993
+    },
+    {
+      "epoch": 13.84168012924071,
+      "grad_norm": 3.375,
+      "learning_rate": 4.00266844563042e-08,
+      "loss": 1.4631,
+      "step": 14994
+    },
+    {
+      "epoch": 13.842603277175167,
+      "grad_norm": 3.228515625,
+      "learning_rate": 3.3355570380253504e-08,
+      "loss": 1.3418,
+      "step": 14995
+    },
+    {
+      "epoch": 13.843526425109625,
+      "grad_norm": 3.005859375,
+      "learning_rate": 2.6684456304202803e-08,
+      "loss": 1.2433,
+      "step": 14996
+    },
+    {
+      "epoch": 13.84444957304408,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.00133422281521e-08,
+      "loss": 1.2687,
+      "step": 14997
+    },
+    {
+      "epoch": 13.845372720978537,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.3342228152101401e-08,
+      "loss": 1.4851,
+      "step": 14998
+    },
+    {
+      "epoch": 13.846295868912993,
+      "grad_norm": 3.017578125,
+      "learning_rate": 6.671114076050701e-09,
+      "loss": 1.1838,
+      "step": 14999
+    },
+    {
+      "epoch": 13.84721901684745,
+      "grad_norm": 3.244140625,
+      "learning_rate": 0.0,
+      "loss": 1.3088,
+      "step": 15000
+    },
+    {
+      "epoch": 13.84721901684745,
+      "eval_loss": 0.86865234375,
+      "eval_runtime": 23.3402,
+      "eval_samples_per_second": 976.897,
+      "eval_steps_per_second": 61.096,
+      "step": 15000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 15000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 14,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.8371474817024e+17,
+  "train_batch_size": 100,
+  "trial_name": null,
+  "trial_params": null
+}