diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,45624 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.24282982791587,
+  "eval_steps": 500,
+  "global_step": 6500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019120458891013384,
+      "grad_norm": 9.746596336364746,
+      "learning_rate": 0.0,
+      "loss": 0.8644,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003824091778202677,
+      "grad_norm": 7.742455005645752,
+      "learning_rate": 4.404279284341596e-07,
+      "loss": 0.6036,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005736137667304016,
+      "grad_norm": 4.959498405456543,
+      "learning_rate": 6.980617508384441e-07,
+      "loss": 0.5138,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007648183556405354,
+      "grad_norm": 12.7059326171875,
+      "learning_rate": 8.808558568683192e-07,
+      "loss": 0.6537,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009560229445506692,
+      "grad_norm": 9.178160667419434,
+      "learning_rate": 1.0226419808043158e-06,
+      "loss": 0.5527,
+      "step": 5
+    },
+    {
+      "epoch": 0.001147227533460803,
+      "grad_norm": 9.766277313232422,
+      "learning_rate": 1.1384896792726035e-06,
+      "loss": 0.4063,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013384321223709368,
+      "grad_norm": 5.554647922515869,
+      "learning_rate": 1.236437512701272e-06,
+      "loss": 0.7499,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015296367112810707,
+      "grad_norm": 8.689834594726562,
+      "learning_rate": 1.3212837853024787e-06,
+      "loss": 1.4177,
+      "step": 8
+    },
+    {
+      "epoch": 0.0017208413001912047,
+      "grad_norm": 3.352787494659424,
+      "learning_rate": 1.3961235016768883e-06,
+      "loss": 0.5945,
+      "step": 9
+    },
+    {
+      "epoch": 0.0019120458891013384,
+      "grad_norm": 8.300889015197754,
+      "learning_rate": 1.4630699092384754e-06,
+      "loss": 0.7862,
+      "step": 10
+    },
+    {
+      "epoch": 0.002103250478011472,
+      "grad_norm": 4.62298583984375,
+      "learning_rate": 1.5236303013560567e-06,
+      "loss": 0.4052,
+      "step": 11
+    },
+    {
+      "epoch": 0.002294455066921606,
+      "grad_norm": 4.900755405426025,
+      "learning_rate": 1.578917607706763e-06,
+      "loss": 0.2796,
+      "step": 12
+    },
+    {
+      "epoch": 0.00248565965583174,
+      "grad_norm": 4.150928497314453,
+      "learning_rate": 1.6297769993563666e-06,
+      "loss": 0.3957,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026768642447418736,
+      "grad_norm": 3.4511361122131348,
+      "learning_rate": 1.6768654411354316e-06,
+      "loss": 0.5152,
+      "step": 14
+    },
+    {
+      "epoch": 0.0028680688336520078,
+      "grad_norm": 3.906311511993408,
+      "learning_rate": 1.7207037316427597e-06,
+      "loss": 0.5973,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030592734225621415,
+      "grad_norm": 5.145625591278076,
+      "learning_rate": 1.7617117137366383e-06,
+      "loss": 0.5577,
+      "step": 16
+    },
+    {
+      "epoch": 0.003250478011472275,
+      "grad_norm": 4.597131252288818,
+      "learning_rate": 1.800232791723491e-06,
+      "loss": 0.4065,
+      "step": 17
+    },
+    {
+      "epoch": 0.0034416826003824093,
+      "grad_norm": 5.256167411804199,
+      "learning_rate": 1.8365514301110474e-06,
+      "loss": 0.4937,
+      "step": 18
+    },
+    {
+      "epoch": 0.003632887189292543,
+      "grad_norm": 6.020605087280273,
+      "learning_rate": 1.8709059148844288e-06,
+      "loss": 0.2741,
+      "step": 19
+    },
+    {
+      "epoch": 0.0038240917782026767,
+      "grad_norm": 6.1135969161987305,
+      "learning_rate": 1.9034978376726347e-06,
+      "loss": 0.6598,
+      "step": 20
+    },
+    {
+      "epoch": 0.00401529636711281,
+      "grad_norm": 2.16821026802063,
+      "learning_rate": 1.934499263539716e-06,
+      "loss": 0.4142,
+      "step": 21
+    },
+    {
+      "epoch": 0.004206500956022944,
+      "grad_norm": 4.964519023895264,
+      "learning_rate": 1.964058229790216e-06,
+      "loss": 0.5294,
+      "step": 22
+    },
+    {
+      "epoch": 0.004397705544933079,
+      "grad_norm": 2.558614492416382,
+      "learning_rate": 1.992303021449765e-06,
+      "loss": 0.451,
+      "step": 23
+    },
+    {
+      "epoch": 0.004588910133843212,
+      "grad_norm": 5.189733028411865,
+      "learning_rate": 2.019345536140923e-06,
+      "loss": 0.5024,
+      "step": 24
+    },
+    {
+      "epoch": 0.004780114722753346,
+      "grad_norm": 5.467967510223389,
+      "learning_rate": 2.0452839616086316e-06,
+      "loss": 0.6787,
+      "step": 25
+    },
+    {
+      "epoch": 0.00497131931166348,
+      "grad_norm": 4.121450424194336,
+      "learning_rate": 2.070204927790526e-06,
+      "loss": 0.6174,
+      "step": 26
+    },
+    {
+      "epoch": 0.0051625239005736135,
+      "grad_norm": 2.465310573577881,
+      "learning_rate": 2.094185252515332e-06,
+      "loss": 0.3246,
+      "step": 27
+    },
+    {
+      "epoch": 0.005353728489483747,
+      "grad_norm": 3.9040215015411377,
+      "learning_rate": 2.117293369569591e-06,
+      "loss": 0.3824,
+      "step": 28
+    },
+    {
+      "epoch": 0.005544933078393882,
+      "grad_norm": 4.060995101928711,
+      "learning_rate": 2.1395905060565537e-06,
+      "loss": 0.6113,
+      "step": 29
+    },
+    {
+      "epoch": 0.0057361376673040155,
+      "grad_norm": 3.5145015716552734,
+      "learning_rate": 2.1611316600769195e-06,
+      "loss": 0.25,
+      "step": 30
+    },
+    {
+      "epoch": 0.005927342256214149,
+      "grad_norm": 4.692299842834473,
+      "learning_rate": 2.1819664180398482e-06,
+      "loss": 0.3016,
+      "step": 31
+    },
+    {
+      "epoch": 0.006118546845124283,
+      "grad_norm": 4.2027788162231445,
+      "learning_rate": 2.202139642170798e-06,
+      "loss": 0.873,
+      "step": 32
+    },
+    {
+      "epoch": 0.006309751434034417,
+      "grad_norm": 2.9636874198913574,
+      "learning_rate": 2.2216920521945006e-06,
+      "loss": 0.8008,
+      "step": 33
+    },
+    {
+      "epoch": 0.00650095602294455,
+      "grad_norm": 1.739428162574768,
+      "learning_rate": 2.240660720157651e-06,
+      "loss": 0.1883,
+      "step": 34
+    },
+    {
+      "epoch": 0.006692160611854685,
+      "grad_norm": 4.254518032073975,
+      "learning_rate": 2.2590794935055878e-06,
+      "loss": 0.5216,
+      "step": 35
+    },
+    {
+      "epoch": 0.006883365200764819,
+      "grad_norm": 3.1227238178253174,
+      "learning_rate": 2.276979358545207e-06,
+      "loss": 0.4041,
+      "step": 36
+    },
+    {
+      "epoch": 0.007074569789674952,
+      "grad_norm": 3.3138492107391357,
+      "learning_rate": 2.2943887540983186e-06,
+      "loss": 0.2946,
+      "step": 37
+    },
+    {
+      "epoch": 0.007265774378585086,
+      "grad_norm": 4.404989719390869,
+      "learning_rate": 2.3113338433185884e-06,
+      "loss": 0.476,
+      "step": 38
+    },
+    {
+      "epoch": 0.00745697896749522,
+      "grad_norm": 3.9010329246520996,
+      "learning_rate": 2.3278387501948105e-06,
+      "loss": 0.7358,
+      "step": 39
+    },
+    {
+      "epoch": 0.0076481835564053535,
+      "grad_norm": 4.1798624992370605,
+      "learning_rate": 2.3439257661067945e-06,
+      "loss": 0.51,
+      "step": 40
+    },
+    {
+      "epoch": 0.007839388145315488,
+      "grad_norm": 3.3263652324676514,
+      "learning_rate": 2.3596155308722216e-06,
+      "loss": 0.4974,
+      "step": 41
+    },
+    {
+      "epoch": 0.00803059273422562,
+      "grad_norm": 5.814546585083008,
+      "learning_rate": 2.3749271919738757e-06,
+      "loss": 0.4005,
+      "step": 42
+    },
+    {
+      "epoch": 0.008221797323135755,
+      "grad_norm": 2.7462821006774902,
+      "learning_rate": 2.389878545048738e-06,
+      "loss": 0.391,
+      "step": 43
+    },
+    {
+      "epoch": 0.008413001912045888,
+      "grad_norm": 4.502755165100098,
+      "learning_rate": 2.4044861582243756e-06,
+      "loss": 0.2376,
+      "step": 44
+    },
+    {
+      "epoch": 0.008604206500956023,
+      "grad_norm": 4.496071815490723,
+      "learning_rate": 2.4187654824812036e-06,
+      "loss": 0.8074,
+      "step": 45
+    },
+    {
+      "epoch": 0.008795411089866157,
+      "grad_norm": 4.023655414581299,
+      "learning_rate": 2.4327309498839246e-06,
+      "loss": 0.5484,
+      "step": 46
+    },
+    {
+      "epoch": 0.00898661567877629,
+      "grad_norm": 2.9175775051116943,
+      "learning_rate": 2.446396061247859e-06,
+      "loss": 0.3416,
+      "step": 47
+    },
+    {
+      "epoch": 0.009177820267686425,
+      "grad_norm": 2.873981237411499,
+      "learning_rate": 2.4597734645750825e-06,
+      "loss": 0.5039,
+      "step": 48
+    },
+    {
+      "epoch": 0.009369024856596558,
+      "grad_norm": 2.632161855697632,
+      "learning_rate": 2.472875025402544e-06,
+      "loss": 0.4038,
+      "step": 49
+    },
+    {
+      "epoch": 0.009560229445506692,
+      "grad_norm": 3.235424041748047,
+      "learning_rate": 2.4857118900427907e-06,
+      "loss": 0.3458,
+      "step": 50
+    },
+    {
+      "epoch": 0.009751434034416827,
+      "grad_norm": 2.7174906730651855,
+      "learning_rate": 2.498294542561935e-06,
+      "loss": 0.4972,
+      "step": 51
+    },
+    {
+      "epoch": 0.00994263862332696,
+      "grad_norm": 3.14343523979187,
+      "learning_rate": 2.5106328562246856e-06,
+      "loss": 0.4773,
+      "step": 52
+    },
+    {
+      "epoch": 0.010133843212237094,
+      "grad_norm": 3.828115224838257,
+      "learning_rate": 2.5227361400389193e-06,
+      "loss": 0.4097,
+      "step": 53
+    },
+    {
+      "epoch": 0.010325047801147227,
+      "grad_norm": 3.4634366035461426,
+      "learning_rate": 2.5346131809494915e-06,
+      "loss": 0.582,
+      "step": 54
+    },
+    {
+      "epoch": 0.010516252390057362,
+      "grad_norm": 3.1296660900115967,
+      "learning_rate": 2.5462722821603727e-06,
+      "loss": 0.3681,
+      "step": 55
+    },
+    {
+      "epoch": 0.010707456978967494,
+      "grad_norm": 2.358661651611328,
+      "learning_rate": 2.557721298003751e-06,
+      "loss": 0.1711,
+      "step": 56
+    },
+    {
+      "epoch": 0.010898661567877629,
+      "grad_norm": 4.519606113433838,
+      "learning_rate": 2.568967665722873e-06,
+      "loss": 0.7541,
+      "step": 57
+    },
+    {
+      "epoch": 0.011089866156787764,
+      "grad_norm": 2.1992197036743164,
+      "learning_rate": 2.580018434490713e-06,
+      "loss": 0.2106,
+      "step": 58
+    },
+    {
+      "epoch": 0.011281070745697896,
+      "grad_norm": 2.597233295440674,
+      "learning_rate": 2.5908802919480436e-06,
+      "loss": 0.4043,
+      "step": 59
+    },
+    {
+      "epoch": 0.011472275334608031,
+      "grad_norm": 3.097158670425415,
+      "learning_rate": 2.601559588511079e-06,
+      "loss": 0.3287,
+      "step": 60
+    },
+    {
+      "epoch": 0.011663479923518164,
+      "grad_norm": 2.8475654125213623,
+      "learning_rate": 2.6120623596699453e-06,
+      "loss": 0.3242,
+      "step": 61
+    },
+    {
+      "epoch": 0.011854684512428298,
+      "grad_norm": 2.1070339679718018,
+      "learning_rate": 2.622394346474008e-06,
+      "loss": 0.1107,
+      "step": 62
+    },
+    {
+      "epoch": 0.012045889101338431,
+      "grad_norm": 3.0335943698883057,
+      "learning_rate": 2.63256101437816e-06,
+      "loss": 0.5753,
+      "step": 63
+    },
+    {
+      "epoch": 0.012237093690248566,
+      "grad_norm": 5.001894950866699,
+      "learning_rate": 2.6425675706049575e-06,
+      "loss": 0.6856,
+      "step": 64
+    },
+    {
+      "epoch": 0.0124282982791587,
+      "grad_norm": 2.677110433578491,
+      "learning_rate": 2.652418980160682e-06,
+      "loss": 0.1709,
+      "step": 65
+    },
+    {
+      "epoch": 0.012619502868068833,
+      "grad_norm": 2.887890100479126,
+      "learning_rate": 2.6621199806286598e-06,
+      "loss": 0.5874,
+      "step": 66
+    },
+    {
+      "epoch": 0.012810707456978968,
+      "grad_norm": 2.3578383922576904,
+      "learning_rate": 2.6716750958501646e-06,
+      "loss": 0.1294,
+      "step": 67
+    },
+    {
+      "epoch": 0.0130019120458891,
+      "grad_norm": 3.1397762298583984,
+      "learning_rate": 2.6810886485918104e-06,
+      "loss": 0.3335,
+      "step": 68
+    },
+    {
+      "epoch": 0.013193116634799235,
+      "grad_norm": 2.663269281387329,
+      "learning_rate": 2.690364772288209e-06,
+      "loss": 0.2245,
+      "step": 69
+    },
+    {
+      "epoch": 0.01338432122370937,
+      "grad_norm": 5.426806449890137,
+      "learning_rate": 2.6995074219397478e-06,
+      "loss": 0.5171,
+      "step": 70
+    },
+    {
+      "epoch": 0.013575525812619503,
+      "grad_norm": 5.457904815673828,
+      "learning_rate": 2.708520384237387e-06,
+      "loss": 0.3069,
+      "step": 71
+    },
+    {
+      "epoch": 0.013766730401529637,
+      "grad_norm": 2.8190979957580566,
+      "learning_rate": 2.7174072869793666e-06,
+      "loss": 0.4272,
+      "step": 72
+    },
+    {
+      "epoch": 0.01395793499043977,
+      "grad_norm": 2.0407586097717285,
+      "learning_rate": 2.7261716078384117e-06,
+      "loss": 0.2898,
+      "step": 73
+    },
+    {
+      "epoch": 0.014149139579349905,
+      "grad_norm": 2.2423548698425293,
+      "learning_rate": 2.734816682532479e-06,
+      "loss": 0.2225,
+      "step": 74
+    },
+    {
+      "epoch": 0.014340344168260038,
+      "grad_norm": 2.801767587661743,
+      "learning_rate": 2.7433457124470753e-06,
+      "loss": 0.1297,
+      "step": 75
+    },
+    {
+      "epoch": 0.014531548757170172,
+      "grad_norm": 3.7313790321350098,
+      "learning_rate": 2.751761771752748e-06,
+      "loss": 0.6037,
+      "step": 76
+    },
+    {
+      "epoch": 0.014722753346080307,
+      "grad_norm": 2.962400436401367,
+      "learning_rate": 2.760067814057329e-06,
+      "loss": 0.4091,
+      "step": 77
+    },
+    {
+      "epoch": 0.01491395793499044,
+      "grad_norm": 1.2434055805206299,
+      "learning_rate": 2.76826667862897e-06,
+      "loss": 0.3433,
+      "step": 78
+    },
+    {
+      "epoch": 0.015105162523900574,
+      "grad_norm": 4.778526782989502,
+      "learning_rate": 2.7763610962227783e-06,
+      "loss": 0.4217,
+      "step": 79
+    },
+    {
+      "epoch": 0.015296367112810707,
+      "grad_norm": 1.6460481882095337,
+      "learning_rate": 2.7843536945409537e-06,
+      "loss": 0.0981,
+      "step": 80
+    },
+    {
+      "epoch": 0.015487571701720841,
+      "grad_norm": 3.009162187576294,
+      "learning_rate": 2.7922470033537765e-06,
+      "loss": 0.1744,
+      "step": 81
+    },
+    {
+      "epoch": 0.015678776290630976,
+      "grad_norm": 4.354190826416016,
+      "learning_rate": 2.8000434593063812e-06,
+      "loss": 0.4956,
+      "step": 82
+    },
+    {
+      "epoch": 0.01586998087954111,
+      "grad_norm": 3.888239860534668,
+      "learning_rate": 2.807745410434209e-06,
+      "loss": 0.4807,
+      "step": 83
+    },
+    {
+      "epoch": 0.01606118546845124,
+      "grad_norm": 2.944679021835327,
+      "learning_rate": 2.8153551204080353e-06,
+      "loss": 0.4177,
+      "step": 84
+    },
+    {
+      "epoch": 0.016252390057361378,
+      "grad_norm": 3.1436245441436768,
+      "learning_rate": 2.822874772527807e-06,
+      "loss": 0.2304,
+      "step": 85
+    },
+    {
+      "epoch": 0.01644359464627151,
+      "grad_norm": 4.823106288909912,
+      "learning_rate": 2.830306473482897e-06,
+      "loss": 0.5006,
+      "step": 86
+    },
+    {
+      "epoch": 0.016634799235181644,
+      "grad_norm": 2.8725435733795166,
+      "learning_rate": 2.837652256894998e-06,
+      "loss": 0.128,
+      "step": 87
+    },
+    {
+      "epoch": 0.016826003824091777,
+      "grad_norm": 4.338479995727539,
+      "learning_rate": 2.8449140866585356e-06,
+      "loss": 0.2579,
+      "step": 88
+    },
+    {
+      "epoch": 0.017017208413001913,
+      "grad_norm": 4.045529365539551,
+      "learning_rate": 2.852093860092363e-06,
+      "loss": 0.9362,
+      "step": 89
+    },
+    {
+      "epoch": 0.017208413001912046,
+      "grad_norm": 3.180530309677124,
+      "learning_rate": 2.8591934109153636e-06,
+      "loss": 0.3104,
+      "step": 90
+    },
+    {
+      "epoch": 0.01739961759082218,
+      "grad_norm": 3.4100022315979004,
+      "learning_rate": 2.8662145120576384e-06,
+      "loss": 0.3072,
+      "step": 91
+    },
+    {
+      "epoch": 0.017590822179732315,
+      "grad_norm": 2.9597487449645996,
+      "learning_rate": 2.873158878318084e-06,
+      "loss": 0.2439,
+      "step": 92
+    },
+    {
+      "epoch": 0.017782026768642448,
+      "grad_norm": 3.1329643726348877,
+      "learning_rate": 2.8800281688782923e-06,
+      "loss": 0.1225,
+      "step": 93
+    },
+    {
+      "epoch": 0.01797323135755258,
+      "grad_norm": 3.316901445388794,
+      "learning_rate": 2.8868239896820188e-06,
+      "loss": 0.2914,
+      "step": 94
+    },
+    {
+      "epoch": 0.018164435946462717,
+      "grad_norm": 2.207570791244507,
+      "learning_rate": 2.8935478956887446e-06,
+      "loss": 0.294,
+      "step": 95
+    },
+    {
+      "epoch": 0.01835564053537285,
+      "grad_norm": 2.481140375137329,
+      "learning_rate": 2.900201393009242e-06,
+      "loss": 0.5148,
+      "step": 96
+    },
+    {
+      "epoch": 0.018546845124282983,
+      "grad_norm": 2.634587526321411,
+      "learning_rate": 2.906785940930483e-06,
+      "loss": 0.5636,
+      "step": 97
+    },
+    {
+      "epoch": 0.018738049713193115,
+      "grad_norm": 2.602517604827881,
+      "learning_rate": 2.913302953836704e-06,
+      "loss": 0.0733,
+      "step": 98
+    },
+    {
+      "epoch": 0.01892925430210325,
+      "grad_norm": 2.664775848388672,
+      "learning_rate": 2.9197538030329443e-06,
+      "loss": 0.2034,
+      "step": 99
+    },
+    {
+      "epoch": 0.019120458891013385,
+      "grad_norm": 4.784229278564453,
+      "learning_rate": 2.9261398184769508e-06,
+      "loss": 0.4161,
+      "step": 100
+    },
+    {
+      "epoch": 0.019311663479923517,
+      "grad_norm": 20.801212310791016,
+      "learning_rate": 2.9324622904249074e-06,
+      "loss": 0.312,
+      "step": 101
+    },
+    {
+      "epoch": 0.019502868068833654,
+      "grad_norm": 2.7974696159362793,
+      "learning_rate": 2.938722470996094e-06,
+      "loss": 0.436,
+      "step": 102
+    },
+    {
+      "epoch": 0.019694072657743786,
+      "grad_norm": 2.719644784927368,
+      "learning_rate": 2.944921575661221e-06,
+      "loss": 0.2729,
+      "step": 103
+    },
+    {
+      "epoch": 0.01988527724665392,
+      "grad_norm": 2.62326717376709,
+      "learning_rate": 2.951060784658845e-06,
+      "loss": 0.1857,
+      "step": 104
+    },
+    {
+      "epoch": 0.020076481835564052,
+      "grad_norm": 3.2945799827575684,
+      "learning_rate": 2.957141244344032e-06,
+      "loss": 0.2618,
+      "step": 105
+    },
+    {
+      "epoch": 0.02026768642447419,
+      "grad_norm": 3.1361422538757324,
+      "learning_rate": 2.963164068473079e-06,
+      "loss": 0.2186,
+      "step": 106
+    },
+    {
+      "epoch": 0.02045889101338432,
+      "grad_norm": 3.901604175567627,
+      "learning_rate": 2.9691303394279335e-06,
+      "loss": 0.5334,
+      "step": 107
+    },
+    {
+      "epoch": 0.020650095602294454,
+      "grad_norm": 4.767889499664307,
+      "learning_rate": 2.975041109383651e-06,
+      "loss": 0.4494,
+      "step": 108
+    },
+    {
+      "epoch": 0.02084130019120459,
+      "grad_norm": 2.728724241256714,
+      "learning_rate": 2.9808974014220527e-06,
+      "loss": 0.3868,
+      "step": 109
+    },
+    {
+      "epoch": 0.021032504780114723,
+      "grad_norm": 3.014492988586426,
+      "learning_rate": 2.986700210594532e-06,
+      "loss": 0.3775,
+      "step": 110
+    },
+    {
+      "epoch": 0.021223709369024856,
+      "grad_norm": 2.4681789875030518,
+      "learning_rate": 2.9924505049367623e-06,
+      "loss": 0.3131,
+      "step": 111
+    },
+    {
+      "epoch": 0.02141491395793499,
+      "grad_norm": 3.5321526527404785,
+      "learning_rate": 2.9981492264379103e-06,
+      "loss": 0.3168,
+      "step": 112
+    },
+    {
+      "epoch": 0.021606118546845125,
+      "grad_norm": 2.0447781085968018,
+      "learning_rate": 3.003797291966757e-06,
+      "loss": 0.3725,
+      "step": 113
+    },
+    {
+      "epoch": 0.021797323135755258,
+      "grad_norm": 3.804521322250366,
+      "learning_rate": 3.0093955941570325e-06,
+      "loss": 0.571,
+      "step": 114
+    },
+    {
+      "epoch": 0.02198852772466539,
+      "grad_norm": 2.5446314811706543,
+      "learning_rate": 3.014945002254081e-06,
+      "loss": 0.3944,
+      "step": 115
+    },
+    {
+      "epoch": 0.022179732313575527,
+      "grad_norm": 2.264204502105713,
+      "learning_rate": 3.0204463629248725e-06,
+      "loss": 0.2348,
+      "step": 116
+    },
+    {
+      "epoch": 0.02237093690248566,
+      "grad_norm": 2.108393907546997,
+      "learning_rate": 3.0259005010332546e-06,
+      "loss": 0.259,
+      "step": 117
+    },
+    {
+      "epoch": 0.022562141491395793,
+      "grad_norm": 4.435889720916748,
+      "learning_rate": 3.031308220382203e-06,
+      "loss": 0.2287,
+      "step": 118
+    },
+    {
+      "epoch": 0.022753346080305926,
+      "grad_norm": 2.867352247238159,
+      "learning_rate": 3.0366703044247632e-06,
+      "loss": 0.1899,
+      "step": 119
+    },
+    {
+      "epoch": 0.022944550669216062,
+      "grad_norm": 4.31091833114624,
+      "learning_rate": 3.0419875169452382e-06,
+      "loss": 0.771,
+      "step": 120
+    },
+    {
+      "epoch": 0.023135755258126195,
+      "grad_norm": 2.2673258781433105,
+      "learning_rate": 3.0472606027121134e-06,
+      "loss": 0.3879,
+      "step": 121
+    },
+    {
+      "epoch": 0.023326959847036328,
+      "grad_norm": 2.5156142711639404,
+      "learning_rate": 3.0524902881041045e-06,
+      "loss": 0.2798,
+      "step": 122
+    },
+    {
+      "epoch": 0.023518164435946464,
+      "grad_norm": 1.8811085224151611,
+      "learning_rate": 3.0576772817106653e-06,
+      "loss": 0.1592,
+      "step": 123
+    },
+    {
+      "epoch": 0.023709369024856597,
+      "grad_norm": 2.619985818862915,
+      "learning_rate": 3.062822274908168e-06,
+      "loss": 0.2492,
+      "step": 124
+    },
+    {
+      "epoch": 0.02390057361376673,
+      "grad_norm": 2.3442344665527344,
+      "learning_rate": 3.067925942412948e-06,
+      "loss": 0.1004,
+      "step": 125
+    },
+    {
+      "epoch": 0.024091778202676863,
+      "grad_norm": 3.19574236869812,
+      "learning_rate": 3.0729889428123194e-06,
+      "loss": 0.6887,
+      "step": 126
+    },
+    {
+      "epoch": 0.024282982791587,
+      "grad_norm": 1.725019097328186,
+      "learning_rate": 3.0780119190745983e-06,
+      "loss": 0.2958,
+      "step": 127
+    },
+    {
+      "epoch": 0.024474187380497132,
+      "grad_norm": 2.455963373184204,
+      "learning_rate": 3.082995499039117e-06,
+      "loss": 0.2027,
+      "step": 128
+    },
+    {
+      "epoch": 0.024665391969407265,
+      "grad_norm": 6.7689619064331055,
+      "learning_rate": 3.0879402958871817e-06,
+      "loss": 0.0587,
+      "step": 129
+    },
+    {
+      "epoch": 0.0248565965583174,
+      "grad_norm": 1.6544665098190308,
+      "learning_rate": 3.0928469085948413e-06,
+      "loss": 0.0546,
+      "step": 130
+    },
+    {
+      "epoch": 0.025047801147227534,
+      "grad_norm": 3.8221983909606934,
+      "learning_rate": 3.0977159223683077e-06,
+      "loss": 0.1562,
+      "step": 131
+    },
+    {
+      "epoch": 0.025239005736137667,
+      "grad_norm": 2.911043167114258,
+      "learning_rate": 3.102547909062819e-06,
+      "loss": 0.3307,
+      "step": 132
+    },
+    {
+      "epoch": 0.025430210325047803,
+      "grad_norm": 3.8134329319000244,
+      "learning_rate": 3.1073434275857012e-06,
+      "loss": 0.6647,
+      "step": 133
+    },
+    {
+      "epoch": 0.025621414913957936,
+      "grad_norm": 3.2995803356170654,
+      "learning_rate": 3.112103024284324e-06,
+      "loss": 0.1994,
+      "step": 134
+    },
+    {
+      "epoch": 0.02581261950286807,
+      "grad_norm": 1.8985707759857178,
+      "learning_rate": 3.1168272333196477e-06,
+      "loss": 0.2707,
+      "step": 135
+    },
+    {
+      "epoch": 0.0260038240917782,
+      "grad_norm": 2.5111820697784424,
+      "learning_rate": 3.1215165770259696e-06,
+      "loss": 0.2573,
+      "step": 136
+    },
+    {
+      "epoch": 0.026195028680688338,
+      "grad_norm": 1.4703351259231567,
+      "learning_rate": 3.1261715662575076e-06,
+      "loss": 0.032,
+      "step": 137
+    },
+    {
+      "epoch": 0.02638623326959847,
+      "grad_norm": 3.487067937850952,
+      "learning_rate": 3.1307927007223687e-06,
+      "loss": 0.3179,
+      "step": 138
+    },
+    {
+      "epoch": 0.026577437858508603,
+      "grad_norm": 3.4684128761291504,
+      "learning_rate": 3.135380469304468e-06,
+      "loss": 0.4447,
+      "step": 139
+    },
+    {
+      "epoch": 0.02676864244741874,
+      "grad_norm": 3.0979361534118652,
+      "learning_rate": 3.1399353503739065e-06,
+      "loss": 0.2441,
+      "step": 140
+    },
+    {
+      "epoch": 0.026959847036328873,
+      "grad_norm": 3.7381913661956787,
+      "learning_rate": 3.1444578120863033e-06,
+      "loss": 0.3272,
+      "step": 141
+    },
+    {
+      "epoch": 0.027151051625239005,
+      "grad_norm": 2.558560609817505,
+      "learning_rate": 3.148948312671547e-06,
+      "loss": 0.3229,
+      "step": 142
+    },
+    {
+      "epoch": 0.027342256214149138,
+      "grad_norm": 2.115128755569458,
+      "learning_rate": 3.153407300712423e-06,
+      "loss": 0.1838,
+      "step": 143
+    },
+    {
+      "epoch": 0.027533460803059275,
+      "grad_norm": 3.6304619312286377,
+      "learning_rate": 3.157835215413526e-06,
+      "loss": 0.3281,
+      "step": 144
+    },
+    {
+      "epoch": 0.027724665391969407,
+      "grad_norm": 7.865360260009766,
+      "learning_rate": 3.1622324868608695e-06,
+      "loss": 0.8346,
+      "step": 145
+    },
+    {
+      "epoch": 0.02791586998087954,
+      "grad_norm": 3.5878102779388428,
+      "learning_rate": 3.166599536272571e-06,
+      "loss": 0.5395,
+      "step": 146
+    },
+    {
+      "epoch": 0.028107074569789676,
+      "grad_norm": 3.4643099308013916,
+      "learning_rate": 3.170936776240988e-06,
+      "loss": 0.3664,
+      "step": 147
+    },
+    {
+      "epoch": 0.02829827915869981,
+      "grad_norm": 3.340297222137451,
+      "learning_rate": 3.1752446109666377e-06,
+      "loss": 0.3959,
+      "step": 148
+    },
+    {
+      "epoch": 0.028489483747609942,
+      "grad_norm": 3.389150619506836,
+      "learning_rate": 3.1795234364842463e-06,
+      "loss": 0.0883,
+      "step": 149
+    },
+    {
+      "epoch": 0.028680688336520075,
+      "grad_norm": 8.659844398498535,
+      "learning_rate": 3.183773640881235e-06,
+      "loss": 0.2891,
+      "step": 150
+    },
+    {
+      "epoch": 0.02887189292543021,
+      "grad_norm": 2.904853582382202,
+      "learning_rate": 3.1879956045089473e-06,
+      "loss": 0.4851,
+      "step": 151
+    },
+    {
+      "epoch": 0.029063097514340344,
+      "grad_norm": 2.432720184326172,
+      "learning_rate": 3.192189700186908e-06,
+      "loss": 0.4643,
+      "step": 152
+    },
+    {
+      "epoch": 0.029254302103250477,
+      "grad_norm": 2.321363925933838,
+      "learning_rate": 3.196356293400379e-06,
+      "loss": 0.1767,
+      "step": 153
+    },
+    {
+      "epoch": 0.029445506692160613,
+      "grad_norm": 1.9591343402862549,
+      "learning_rate": 3.2004957424914884e-06,
+      "loss": 0.1611,
+      "step": 154
+    },
+    {
+      "epoch": 0.029636711281070746,
+      "grad_norm": 2.822035551071167,
+      "learning_rate": 3.204608398844164e-06,
+      "loss": 0.2053,
+      "step": 155
+    },
+    {
+      "epoch": 0.02982791586998088,
+      "grad_norm": 2.73078989982605,
+      "learning_rate": 3.2086946070631297e-06,
+      "loss": 0.095,
+      "step": 156
+    },
+    {
+      "epoch": 0.030019120458891012,
+      "grad_norm": 4.012321949005127,
+      "learning_rate": 3.212754705147177e-06,
+      "loss": 0.4317,
+      "step": 157
+    },
+    {
+      "epoch": 0.030210325047801148,
+      "grad_norm": 2.332084894180298,
+      "learning_rate": 3.216789024656938e-06,
+      "loss": 0.4723,
+      "step": 158
+    },
+    {
+      "epoch": 0.03040152963671128,
+      "grad_norm": 3.0491888523101807,
+      "learning_rate": 3.2207978908773634e-06,
+      "loss": 0.4482,
+      "step": 159
+    },
+    {
+      "epoch": 0.030592734225621414,
+      "grad_norm": 1.8330838680267334,
+      "learning_rate": 3.2247816229751133e-06,
+      "loss": 0.1647,
+      "step": 160
+    },
+    {
+      "epoch": 0.03078393881453155,
+      "grad_norm": 2.9033660888671875,
+      "learning_rate": 3.228740534151037e-06,
+      "loss": 0.3954,
+      "step": 161
+    },
+    {
+      "epoch": 0.030975143403441683,
+      "grad_norm": 2.846273899078369,
+      "learning_rate": 3.2326749317879352e-06,
+      "loss": 0.1128,
+      "step": 162
+    },
+    {
+      "epoch": 0.031166347992351816,
+      "grad_norm": 3.386489152908325,
+      "learning_rate": 3.2365851175937783e-06,
+      "loss": 0.4637,
+      "step": 163
+    },
+    {
+      "epoch": 0.03135755258126195,
+      "grad_norm": 2.893277883529663,
+      "learning_rate": 3.240471387740541e-06,
+      "loss": 0.5229,
+      "step": 164
+    },
+    {
+      "epoch": 0.03154875717017208,
+      "grad_norm": 2.0880420207977295,
+      "learning_rate": 3.244334032998816e-06,
+      "loss": 0.4664,
+      "step": 165
+    },
+    {
+      "epoch": 0.03173996175908222,
+      "grad_norm": 4.319767475128174,
+      "learning_rate": 3.2481733388683686e-06,
+      "loss": 0.5711,
+      "step": 166
+    },
+    {
+      "epoch": 0.031931166347992354,
+      "grad_norm": 3.115651845932007,
+      "learning_rate": 3.251989585704759e-06,
+      "loss": 0.1742,
+      "step": 167
+    },
+    {
+      "epoch": 0.03212237093690248,
+      "grad_norm": 3.5364396572113037,
+      "learning_rate": 3.255783048842195e-06,
+      "loss": 0.1186,
+      "step": 168
+    },
+    {
+      "epoch": 0.03231357552581262,
+      "grad_norm": 2.261878252029419,
+      "learning_rate": 3.2595539987127332e-06,
+      "loss": 0.1794,
+      "step": 169
+    },
+    {
+      "epoch": 0.032504780114722756,
+      "grad_norm": 3.346421003341675,
+      "learning_rate": 3.2633027009619666e-06,
+      "loss": 0.5637,
+      "step": 170
+    },
+    {
+      "epoch": 0.032695984703632885,
+      "grad_norm": 3.0582921504974365,
+      "learning_rate": 3.267029416561317e-06,
+      "loss": 0.3209,
+      "step": 171
+    },
+    {
+      "epoch": 0.03288718929254302,
+      "grad_norm": 2.956470251083374,
+      "learning_rate": 3.270734401917057e-06,
+      "loss": 0.3218,
+      "step": 172
+    },
+    {
+      "epoch": 0.03307839388145316,
+      "grad_norm": 1.4681670665740967,
+      "learning_rate": 3.27441790897617e-06,
+      "loss": 0.2129,
+      "step": 173
+    },
+    {
+      "epoch": 0.03326959847036329,
+      "grad_norm": 4.20260763168335,
+      "learning_rate": 3.278080185329157e-06,
+      "loss": 0.3151,
+      "step": 174
+    },
+    {
+      "epoch": 0.033460803059273424,
+      "grad_norm": 4.02779483795166,
+      "learning_rate": 3.281721474309904e-06,
+      "loss": 0.3233,
+      "step": 175
+    },
+    {
+      "epoch": 0.03365200764818355,
+      "grad_norm": 3.5959632396698,
+      "learning_rate": 3.2853420150926944e-06,
+      "loss": 0.7,
+      "step": 176
+    },
+    {
+      "epoch": 0.03384321223709369,
+      "grad_norm": 2.106358528137207,
+      "learning_rate": 3.2889420427864873e-06,
+      "loss": 0.6374,
+      "step": 177
+    },
+    {
+      "epoch": 0.034034416826003826,
+      "grad_norm": 2.9504098892211914,
+      "learning_rate": 3.2925217885265225e-06,
+      "loss": 0.408,
+      "step": 178
+    },
+    {
+      "epoch": 0.034225621414913955,
+      "grad_norm": 2.7030961513519287,
+      "learning_rate": 3.296081479563376e-06,
+      "loss": 0.3322,
+      "step": 179
+    },
+    {
+      "epoch": 0.03441682600382409,
+      "grad_norm": 2.271064281463623,
+      "learning_rate": 3.299621339349523e-06,
+      "loss": 0.1597,
+      "step": 180
+    },
+    {
+      "epoch": 0.03460803059273423,
+      "grad_norm": 3.3034112453460693,
+      "learning_rate": 3.3031415876235085e-06,
+      "loss": 0.1215,
+      "step": 181
+    },
+    {
+      "epoch": 0.03479923518164436,
+      "grad_norm": 3.9516632556915283,
+      "learning_rate": 3.306642440491798e-06,
+      "loss": 0.6046,
+      "step": 182
+    },
+    {
+      "epoch": 0.03499043977055449,
+      "grad_norm": 5.616006374359131,
+      "learning_rate": 3.310124110508389e-06,
+      "loss": 0.8098,
+      "step": 183
+    },
+    {
+      "epoch": 0.03518164435946463,
+      "grad_norm": 1.1683584451675415,
+      "learning_rate": 3.3135868067522437e-06,
+      "loss": 0.2074,
+      "step": 184
+    },
+    {
+      "epoch": 0.03537284894837476,
+      "grad_norm": 4.46233606338501,
+      "learning_rate": 3.3170307349026344e-06,
+      "loss": 0.2912,
+      "step": 185
+    },
+    {
+      "epoch": 0.035564053537284895,
+      "grad_norm": 2.703484535217285,
+      "learning_rate": 3.3204560973124523e-06,
+      "loss": 0.1338,
+      "step": 186
+    },
+    {
+      "epoch": 0.03575525812619503,
+      "grad_norm": 2.688272476196289,
+      "learning_rate": 3.3238630930795473e-06,
+      "loss": 0.1497,
+      "step": 187
+    },
+    {
+      "epoch": 0.03594646271510516,
+      "grad_norm": 3.20407772064209,
+      "learning_rate": 3.3272519181161784e-06,
+      "loss": 0.3101,
+      "step": 188
+    },
+    {
+      "epoch": 0.0361376673040153,
+      "grad_norm": 3.1082072257995605,
+      "learning_rate": 3.330622765216604e-06,
+      "loss": 0.2251,
+      "step": 189
+    },
+    {
+      "epoch": 0.036328871892925434,
+      "grad_norm": 2.8750967979431152,
+      "learning_rate": 3.333975824122904e-06,
+      "loss": 0.475,
+      "step": 190
+    },
+    {
+      "epoch": 0.03652007648183556,
+      "grad_norm": 3.8573219776153564,
+      "learning_rate": 3.337311281589066e-06,
+      "loss": 0.093,
+      "step": 191
+    },
+    {
+      "epoch": 0.0367112810707457,
+      "grad_norm": 1.341035008430481,
+      "learning_rate": 3.3406293214434016e-06,
+      "loss": 0.053,
+      "step": 192
+    },
+    {
+      "epoch": 0.03690248565965583,
+      "grad_norm": 2.519542694091797,
+      "learning_rate": 3.343930124649337e-06,
+      "loss": 0.1018,
+      "step": 193
+    },
+    {
+      "epoch": 0.037093690248565965,
+      "grad_norm": 3.034273386001587,
+      "learning_rate": 3.3472138693646427e-06,
+      "loss": 0.3481,
+      "step": 194
+    },
+    {
+      "epoch": 0.0372848948374761,
+      "grad_norm": 1.9574315547943115,
+      "learning_rate": 3.350480730999126e-06,
+      "loss": 0.2867,
+      "step": 195
+    },
+    {
+      "epoch": 0.03747609942638623,
+      "grad_norm": 2.153280735015869,
+      "learning_rate": 3.353730882270863e-06,
+      "loss": 0.3375,
+      "step": 196
+    },
+    {
+      "epoch": 0.03766730401529637,
+      "grad_norm": 3.426328420639038,
+      "learning_rate": 3.3569644932609887e-06,
+      "loss": 0.4611,
+      "step": 197
+    },
+    {
+      "epoch": 0.0378585086042065,
+      "grad_norm": 2.072275161743164,
+      "learning_rate": 3.360181731467104e-06,
+      "loss": 0.2144,
+      "step": 198
+    },
+    {
+      "epoch": 0.03804971319311663,
+      "grad_norm": 2.887531280517578,
+      "learning_rate": 3.3633827618553393e-06,
+      "loss": 0.2432,
+      "step": 199
+    },
+    {
+      "epoch": 0.03824091778202677,
+      "grad_norm": 5.023746013641357,
+      "learning_rate": 3.3665677469111103e-06,
+      "loss": 0.2016,
+      "step": 200
+    },
+    {
+      "epoch": 0.038432122370936905,
+      "grad_norm": 3.695518970489502,
+      "learning_rate": 3.3697368466886087e-06,
+      "loss": 0.5206,
+      "step": 201
+    },
+    {
+      "epoch": 0.038623326959847035,
+      "grad_norm": 3.6168696880340576,
+      "learning_rate": 3.3728902188590666e-06,
+      "loss": 0.5451,
+      "step": 202
+    },
+    {
+      "epoch": 0.03881453154875717,
+      "grad_norm": 4.44756555557251,
+      "learning_rate": 3.376028018757826e-06,
+      "loss": 0.4795,
+      "step": 203
+    },
+    {
+      "epoch": 0.03900573613766731,
+      "grad_norm": 3.0557193756103516,
+      "learning_rate": 3.3791503994302537e-06,
+      "loss": 0.2987,
+      "step": 204
+    },
+    {
+      "epoch": 0.03919694072657744,
+      "grad_norm": 3.2316536903381348,
+      "learning_rate": 3.382257511676538e-06,
+      "loss": 0.3986,
+      "step": 205
+    },
+    {
+      "epoch": 0.03938814531548757,
+      "grad_norm": 3.7219014167785645,
+      "learning_rate": 3.3853495040953806e-06,
+      "loss": 0.3432,
+      "step": 206
+    },
+    {
+      "epoch": 0.0395793499043977,
+      "grad_norm": 3.6749324798583984,
+      "learning_rate": 3.388426523126653e-06,
+      "loss": 0.5342,
+      "step": 207
+    },
+    {
+      "epoch": 0.03977055449330784,
+      "grad_norm": 3.951026439666748,
+      "learning_rate": 3.3914887130930047e-06,
+      "loss": 0.6743,
+      "step": 208
+    },
+    {
+      "epoch": 0.039961759082217975,
+      "grad_norm": 2.203550338745117,
+      "learning_rate": 3.3945362162404853e-06,
+      "loss": 0.272,
+      "step": 209
+    },
+    {
+      "epoch": 0.040152963671128104,
+      "grad_norm": 3.5759997367858887,
+      "learning_rate": 3.397569172778191e-06,
+      "loss": 0.4403,
+      "step": 210
+    },
+    {
+      "epoch": 0.04034416826003824,
+      "grad_norm": 2.468674421310425,
+      "learning_rate": 3.400587720916976e-06,
+      "loss": 0.2374,
+      "step": 211
+    },
+    {
+      "epoch": 0.04053537284894838,
+      "grad_norm": 1.7186170816421509,
+      "learning_rate": 3.4035919969072384e-06,
+      "loss": 0.1339,
+      "step": 212
+    },
+    {
+      "epoch": 0.040726577437858506,
+      "grad_norm": 3.243154525756836,
+      "learning_rate": 3.4065821350758317e-06,
+      "loss": 0.2984,
+      "step": 213
+    },
+    {
+      "epoch": 0.04091778202676864,
+      "grad_norm": 3.2775168418884277,
+      "learning_rate": 3.409558267862093e-06,
+      "loss": 0.5302,
+      "step": 214
+    },
+    {
+      "epoch": 0.04110898661567878,
+      "grad_norm": 1.8264061212539673,
+      "learning_rate": 3.4125205258530534e-06,
+      "loss": 0.5952,
+      "step": 215
+    },
+    {
+      "epoch": 0.04130019120458891,
+      "grad_norm": 5.033473491668701,
+      "learning_rate": 3.4154690378178107e-06,
+      "loss": 0.2228,
+      "step": 216
+    },
+    {
+      "epoch": 0.041491395793499045,
+      "grad_norm": 1.7373385429382324,
+      "learning_rate": 3.4184039307411206e-06,
+      "loss": 0.1134,
+      "step": 217
+    },
+    {
+      "epoch": 0.04168260038240918,
+      "grad_norm": 2.2630789279937744,
+      "learning_rate": 3.4213253298562123e-06,
+      "loss": 0.1515,
+      "step": 218
+    },
+    {
+      "epoch": 0.04187380497131931,
+      "grad_norm": 3.2471377849578857,
+      "learning_rate": 3.4242333586768554e-06,
+      "loss": 0.1851,
+      "step": 219
+    },
+    {
+      "epoch": 0.04206500956022945,
+      "grad_norm": 4.196602821350098,
+      "learning_rate": 3.4271281390286914e-06,
+      "loss": 0.9757,
+      "step": 220
+    },
+    {
+      "epoch": 0.042256214149139576,
+      "grad_norm": 3.0383758544921875,
+      "learning_rate": 3.4300097910798572e-06,
+      "loss": 0.426,
+      "step": 221
+    },
+    {
+      "epoch": 0.04244741873804971,
+      "grad_norm": 2.5604920387268066,
+      "learning_rate": 3.4328784333709227e-06,
+      "loss": 0.2252,
+      "step": 222
+    },
+    {
+      "epoch": 0.04263862332695985,
+      "grad_norm": 3.8304648399353027,
+      "learning_rate": 3.4357341828441426e-06,
+      "loss": 0.3957,
+      "step": 223
+    },
+    {
+      "epoch": 0.04282982791586998,
+      "grad_norm": 2.3922934532165527,
+      "learning_rate": 3.43857715487207e-06,
+      "loss": 0.1992,
+      "step": 224
+    },
+    {
+      "epoch": 0.043021032504780114,
+      "grad_norm": 3.5151939392089844,
+      "learning_rate": 3.4414074632855194e-06,
+      "loss": 0.2571,
+      "step": 225
+    },
+    {
+      "epoch": 0.04321223709369025,
+      "grad_norm": 3.84830904006958,
+      "learning_rate": 3.4442252204009168e-06,
+      "loss": 0.7981,
+      "step": 226
+    },
+    {
+      "epoch": 0.04340344168260038,
+      "grad_norm": 3.0907559394836426,
+      "learning_rate": 3.447030537047043e-06,
+      "loss": 0.4319,
+      "step": 227
+    },
+    {
+      "epoch": 0.043594646271510516,
+      "grad_norm": 3.6448781490325928,
+      "learning_rate": 3.4498235225911925e-06,
+      "loss": 0.4267,
+      "step": 228
+    },
+    {
+      "epoch": 0.04378585086042065,
+      "grad_norm": 1.7876298427581787,
+      "learning_rate": 3.452604284964754e-06,
+      "loss": 0.2185,
+      "step": 229
+    },
+    {
+      "epoch": 0.04397705544933078,
+      "grad_norm": 1.8112496137619019,
+      "learning_rate": 3.4553729306882404e-06,
+      "loss": 0.1537,
+      "step": 230
+    },
+    {
+      "epoch": 0.04416826003824092,
+      "grad_norm": 4.762514591217041,
+      "learning_rate": 3.4581295648957726e-06,
+      "loss": 0.4439,
+      "step": 231
+    },
+    {
+      "epoch": 0.044359464627151055,
+      "grad_norm": 2.5081305503845215,
+      "learning_rate": 3.4608742913590325e-06,
+      "loss": 0.2384,
+      "step": 232
+    },
+    {
+      "epoch": 0.044550669216061184,
+      "grad_norm": 4.172029495239258,
+      "learning_rate": 3.4636072125107046e-06,
+      "loss": 0.6397,
+      "step": 233
+    },
+    {
+      "epoch": 0.04474187380497132,
+      "grad_norm": 2.4700138568878174,
+      "learning_rate": 3.4663284294674142e-06,
+      "loss": 0.5316,
+      "step": 234
+    },
+    {
+      "epoch": 0.044933078393881457,
+      "grad_norm": 2.9252090454101562,
+      "learning_rate": 3.4690380420521746e-06,
+      "loss": 0.2955,
+      "step": 235
+    },
+    {
+      "epoch": 0.045124282982791586,
+      "grad_norm": 2.3378536701202393,
+      "learning_rate": 3.4717361488163627e-06,
+      "loss": 0.3204,
+      "step": 236
+    },
+    {
+      "epoch": 0.04531548757170172,
+      "grad_norm": 5.816604137420654,
+      "learning_rate": 3.4744228470612224e-06,
+      "loss": 0.1649,
+      "step": 237
+    },
+    {
+      "epoch": 0.04550669216061185,
+      "grad_norm": 2.326578140258789,
+      "learning_rate": 3.4770982328589232e-06,
+      "loss": 0.201,
+      "step": 238
+    },
+    {
+      "epoch": 0.04569789674952199,
+      "grad_norm": 4.211781024932861,
+      "learning_rate": 3.4797624010731722e-06,
+      "loss": 0.7306,
+      "step": 239
+    },
+    {
+      "epoch": 0.045889101338432124,
+      "grad_norm": 1.4428257942199707,
+      "learning_rate": 3.482415445379398e-06,
+      "loss": 0.1414,
+      "step": 240
+    },
+    {
+      "epoch": 0.046080305927342254,
+      "grad_norm": 3.0482757091522217,
+      "learning_rate": 3.4850574582845136e-06,
+      "loss": 0.5237,
+      "step": 241
+    },
+    {
+      "epoch": 0.04627151051625239,
+      "grad_norm": 2.439641237258911,
+      "learning_rate": 3.4876885311462725e-06,
+      "loss": 0.2411,
+      "step": 242
+    },
+    {
+      "epoch": 0.046462715105162526,
+      "grad_norm": 3.4746406078338623,
+      "learning_rate": 3.4903087541922198e-06,
+      "loss": 0.2901,
+      "step": 243
+    },
+    {
+      "epoch": 0.046653919694072656,
+      "grad_norm": 2.937816619873047,
+      "learning_rate": 3.492918216538264e-06,
+      "loss": 0.1849,
+      "step": 244
+    },
+    {
+      "epoch": 0.04684512428298279,
+      "grad_norm": 4.481417655944824,
+      "learning_rate": 3.49551700620686e-06,
+      "loss": 0.6734,
+      "step": 245
+    },
+    {
+      "epoch": 0.04703632887189293,
+      "grad_norm": 3.3815813064575195,
+      "learning_rate": 3.498105210144825e-06,
+      "loss": 0.7956,
+      "step": 246
+    },
+    {
+      "epoch": 0.04722753346080306,
+      "grad_norm": 1.8896292448043823,
+      "learning_rate": 3.5006829142407957e-06,
+      "loss": 0.1625,
+      "step": 247
+    },
+    {
+      "epoch": 0.047418738049713194,
+      "grad_norm": 2.2965781688690186,
+      "learning_rate": 3.503250203342327e-06,
+      "loss": 0.1043,
+      "step": 248
+    },
+    {
+      "epoch": 0.04760994263862333,
+      "grad_norm": 3.055568218231201,
+      "learning_rate": 3.5058071612726523e-06,
+      "loss": 0.2573,
+      "step": 249
+    },
+    {
+      "epoch": 0.04780114722753346,
+      "grad_norm": 2.876594305038452,
+      "learning_rate": 3.5083538708471065e-06,
+      "loss": 0.1837,
+      "step": 250
+    },
+    {
+      "epoch": 0.047992351816443596,
+      "grad_norm": 2.6785998344421387,
+      "learning_rate": 3.5108904138892164e-06,
+      "loss": 0.3405,
+      "step": 251
+    },
+    {
+      "epoch": 0.048183556405353725,
+      "grad_norm": 3.1598598957061768,
+      "learning_rate": 3.5134168712464794e-06,
+      "loss": 0.3016,
+      "step": 252
+    },
+    {
+      "epoch": 0.04837476099426386,
+      "grad_norm": 2.128307342529297,
+      "learning_rate": 3.515933322805821e-06,
+      "loss": 0.3255,
+      "step": 253
+    },
+    {
+      "epoch": 0.048565965583174,
+      "grad_norm": 2.671571969985962,
+      "learning_rate": 3.518439847508758e-06,
+      "loss": 0.2336,
+      "step": 254
+    },
+    {
+      "epoch": 0.04875717017208413,
+      "grad_norm": 5.808933734893799,
+      "learning_rate": 3.5209365233662508e-06,
+      "loss": 0.2801,
+      "step": 255
+    },
+    {
+      "epoch": 0.048948374760994263,
+      "grad_norm": 2.248617649078369,
+      "learning_rate": 3.5234234274732767e-06,
+      "loss": 0.1844,
+      "step": 256
+    },
+    {
+      "epoch": 0.0491395793499044,
+      "grad_norm": 2.858762264251709,
+      "learning_rate": 3.5259006360231123e-06,
+      "loss": 0.4732,
+      "step": 257
+    },
+    {
+      "epoch": 0.04933078393881453,
+      "grad_norm": 2.7468910217285156,
+      "learning_rate": 3.5283682243213417e-06,
+      "loss": 0.2897,
+      "step": 258
+    },
+    {
+      "epoch": 0.049521988527724665,
+      "grad_norm": 1.2300924062728882,
+      "learning_rate": 3.5308262667995906e-06,
+      "loss": 0.2216,
+      "step": 259
+    },
+    {
+      "epoch": 0.0497131931166348,
+      "grad_norm": 3.6707677841186523,
+      "learning_rate": 3.5332748370290014e-06,
+      "loss": 0.5293,
+      "step": 260
+    },
+    {
+      "epoch": 0.04990439770554493,
+      "grad_norm": 2.7596583366394043,
+      "learning_rate": 3.5357140077334416e-06,
+      "loss": 0.2129,
+      "step": 261
+    },
+    {
+      "epoch": 0.05009560229445507,
+      "grad_norm": 5.763503074645996,
+      "learning_rate": 3.5381438508024672e-06,
+      "loss": 0.1661,
+      "step": 262
+    },
+    {
+      "epoch": 0.050286806883365204,
+      "grad_norm": 3.31182599067688,
+      "learning_rate": 3.5405644373040366e-06,
+      "loss": 0.3727,
+      "step": 263
+    },
+    {
+      "epoch": 0.05047801147227533,
+      "grad_norm": 2.9972293376922607,
+      "learning_rate": 3.5429758374969793e-06,
+      "loss": 0.3564,
+      "step": 264
+    },
+    {
+      "epoch": 0.05066921606118547,
+      "grad_norm": 2.814368963241577,
+      "learning_rate": 3.5453781208432355e-06,
+      "loss": 0.4932,
+      "step": 265
+    },
+    {
+      "epoch": 0.050860420650095606,
+      "grad_norm": 4.843621730804443,
+      "learning_rate": 3.547771356019861e-06,
+      "loss": 0.3832,
+      "step": 266
+    },
+    {
+      "epoch": 0.051051625239005735,
+      "grad_norm": 3.9224517345428467,
+      "learning_rate": 3.550155610930807e-06,
+      "loss": 0.1663,
+      "step": 267
+    },
+    {
+      "epoch": 0.05124282982791587,
+      "grad_norm": 2.4663190841674805,
+      "learning_rate": 3.5525309527184838e-06,
+      "loss": 0.0759,
+      "step": 268
+    },
+    {
+      "epoch": 0.051434034416826,
+      "grad_norm": 2.4381537437438965,
+      "learning_rate": 3.5548974477751014e-06,
+      "loss": 0.1891,
+      "step": 269
+    },
+    {
+      "epoch": 0.05162523900573614,
+      "grad_norm": 3.2417197227478027,
+      "learning_rate": 3.5572551617538078e-06,
+      "loss": 0.5146,
+      "step": 270
+    },
+    {
+      "epoch": 0.05181644359464627,
+      "grad_norm": 2.6413793563842773,
+      "learning_rate": 3.559604159579614e-06,
+      "loss": 0.2155,
+      "step": 271
+    },
+    {
+      "epoch": 0.0520076481835564,
+      "grad_norm": 2.974642753601074,
+      "learning_rate": 3.56194450546013e-06,
+      "loss": 0.3031,
+      "step": 272
+    },
+    {
+      "epoch": 0.05219885277246654,
+      "grad_norm": 2.5996057987213135,
+      "learning_rate": 3.5642762628960825e-06,
+      "loss": 0.236,
+      "step": 273
+    },
+    {
+      "epoch": 0.052390057361376675,
+      "grad_norm": 2.6430821418762207,
+      "learning_rate": 3.566599494691667e-06,
+      "loss": 0.1634,
+      "step": 274
+    },
+    {
+      "epoch": 0.052581261950286805,
+      "grad_norm": 4.012979984283447,
+      "learning_rate": 3.5689142629646885e-06,
+      "loss": 0.1486,
+      "step": 275
+    },
+    {
+      "epoch": 0.05277246653919694,
+      "grad_norm": 2.250370740890503,
+      "learning_rate": 3.571220629156528e-06,
+      "loss": 0.3647,
+      "step": 276
+    },
+    {
+      "epoch": 0.05296367112810708,
+      "grad_norm": 2.271556854248047,
+      "learning_rate": 3.5735186540419348e-06,
+      "loss": 0.21,
+      "step": 277
+    },
+    {
+      "epoch": 0.05315487571701721,
+      "grad_norm": 1.3703874349594116,
+      "learning_rate": 3.5758083977386276e-06,
+      "loss": 0.2091,
+      "step": 278
+    },
+    {
+      "epoch": 0.05334608030592734,
+      "grad_norm": 2.649837017059326,
+      "learning_rate": 3.5780899197167356e-06,
+      "loss": 0.1355,
+      "step": 279
+    },
+    {
+      "epoch": 0.05353728489483748,
+      "grad_norm": 3.166032552719116,
+      "learning_rate": 3.580363278808066e-06,
+      "loss": 0.4169,
+      "step": 280
+    },
+    {
+      "epoch": 0.05372848948374761,
+      "grad_norm": 4.7268290519714355,
+      "learning_rate": 3.582628533215206e-06,
+      "loss": 0.2532,
+      "step": 281
+    },
+    {
+      "epoch": 0.053919694072657745,
+      "grad_norm": 2.934070587158203,
+      "learning_rate": 3.584885740520463e-06,
+      "loss": 0.4672,
+      "step": 282
+    },
+    {
+      "epoch": 0.054110898661567874,
+      "grad_norm": 2.0285329818725586,
+      "learning_rate": 3.5871349576946483e-06,
+      "loss": 0.3499,
+      "step": 283
+    },
+    {
+      "epoch": 0.05430210325047801,
+      "grad_norm": 2.122689962387085,
+      "learning_rate": 3.5893762411057063e-06,
+      "loss": 0.2839,
+      "step": 284
+    },
+    {
+      "epoch": 0.05449330783938815,
+      "grad_norm": 3.105699300765991,
+      "learning_rate": 3.5916096465271888e-06,
+      "loss": 0.4167,
+      "step": 285
+    },
+    {
+      "epoch": 0.054684512428298276,
+      "grad_norm": 2.428645372390747,
+      "learning_rate": 3.593835229146582e-06,
+      "loss": 0.2457,
+      "step": 286
+    },
+    {
+      "epoch": 0.05487571701720841,
+      "grad_norm": 2.716440439224243,
+      "learning_rate": 3.5960530435734936e-06,
+      "loss": 0.252,
+      "step": 287
+    },
+    {
+      "epoch": 0.05506692160611855,
+      "grad_norm": 2.986215829849243,
+      "learning_rate": 3.598263143847686e-06,
+      "loss": 0.1737,
+      "step": 288
+    },
+    {
+      "epoch": 0.05525812619502868,
+      "grad_norm": 1.8082088232040405,
+      "learning_rate": 3.600465583446982e-06,
+      "loss": 0.1929,
+      "step": 289
+    },
+    {
+      "epoch": 0.055449330783938815,
+      "grad_norm": 2.961946964263916,
+      "learning_rate": 3.602660415295029e-06,
+      "loss": 0.2653,
+      "step": 290
+    },
+    {
+      "epoch": 0.05564053537284895,
+      "grad_norm": 3.455840587615967,
+      "learning_rate": 3.6048476917689273e-06,
+      "loss": 0.3367,
+      "step": 291
+    },
+    {
+      "epoch": 0.05583173996175908,
+      "grad_norm": 1.9973338842391968,
+      "learning_rate": 3.607027464706731e-06,
+      "loss": 0.1757,
+      "step": 292
+    },
+    {
+      "epoch": 0.05602294455066922,
+      "grad_norm": 3.1876704692840576,
+      "learning_rate": 3.609199785414821e-06,
+      "loss": 0.3237,
+      "step": 293
+    },
+    {
+      "epoch": 0.05621414913957935,
+      "grad_norm": 2.5493903160095215,
+      "learning_rate": 3.6113647046751477e-06,
+      "loss": 0.1988,
+      "step": 294
+    },
+    {
+      "epoch": 0.05640535372848948,
+      "grad_norm": 3.4992916584014893,
+      "learning_rate": 3.6135222727523593e-06,
+      "loss": 0.3859,
+      "step": 295
+    },
+    {
+      "epoch": 0.05659655831739962,
+      "grad_norm": 1.860656499862671,
+      "learning_rate": 3.6156725394007973e-06,
+      "loss": 0.2747,
+      "step": 296
+    },
+    {
+      "epoch": 0.05678776290630975,
+      "grad_norm": 2.110995054244995,
+      "learning_rate": 3.6178155538713884e-06,
+      "loss": 0.2186,
+      "step": 297
+    },
+    {
+      "epoch": 0.056978967495219884,
+      "grad_norm": 2.3457248210906982,
+      "learning_rate": 3.6199513649184063e-06,
+      "loss": 0.2764,
+      "step": 298
+    },
+    {
+      "epoch": 0.05717017208413002,
+      "grad_norm": 2.250549554824829,
+      "learning_rate": 3.622080020806132e-06,
+      "loss": 0.1175,
+      "step": 299
+    },
+    {
+      "epoch": 0.05736137667304015,
+      "grad_norm": 3.1696152687072754,
+      "learning_rate": 3.6242015693153945e-06,
+      "loss": 0.1506,
+      "step": 300
+    },
+    {
+      "epoch": 0.057552581261950286,
+      "grad_norm": 2.9207370281219482,
+      "learning_rate": 3.62631605775001e-06,
+      "loss": 0.3109,
+      "step": 301
+    },
+    {
+      "epoch": 0.05774378585086042,
+      "grad_norm": 2.2986581325531006,
+      "learning_rate": 3.6284235329431073e-06,
+      "loss": 0.2265,
+      "step": 302
+    },
+    {
+      "epoch": 0.05793499043977055,
+      "grad_norm": 2.269641876220703,
+      "learning_rate": 3.6305240412633507e-06,
+      "loss": 0.1851,
+      "step": 303
+    },
+    {
+      "epoch": 0.05812619502868069,
+      "grad_norm": 2.343125104904175,
+      "learning_rate": 3.632617628621067e-06,
+      "loss": 0.1895,
+      "step": 304
+    },
+    {
+      "epoch": 0.058317399617590825,
+      "grad_norm": 3.1341187953948975,
+      "learning_rate": 3.634704340474261e-06,
+      "loss": 0.3916,
+      "step": 305
+    },
+    {
+      "epoch": 0.058508604206500954,
+      "grad_norm": 1.92935311794281,
+      "learning_rate": 3.6367842218345383e-06,
+      "loss": 0.1707,
+      "step": 306
+    },
+    {
+      "epoch": 0.05869980879541109,
+      "grad_norm": 2.106807231903076,
+      "learning_rate": 3.6388573172729357e-06,
+      "loss": 0.3863,
+      "step": 307
+    },
+    {
+      "epoch": 0.05889101338432123,
+      "grad_norm": 0.9721649885177612,
+      "learning_rate": 3.640923670925647e-06,
+      "loss": 0.3013,
+      "step": 308
+    },
+    {
+      "epoch": 0.059082217973231356,
+      "grad_norm": 2.205176830291748,
+      "learning_rate": 3.642983326499665e-06,
+      "loss": 0.2873,
+      "step": 309
+    },
+    {
+      "epoch": 0.05927342256214149,
+      "grad_norm": 2.2065727710723877,
+      "learning_rate": 3.6450363272783236e-06,
+      "loss": 0.2406,
+      "step": 310
+    },
+    {
+      "epoch": 0.05946462715105163,
+      "grad_norm": 5.1225361824035645,
+      "learning_rate": 3.647082716126761e-06,
+      "loss": 0.143,
+      "step": 311
+    },
+    {
+      "epoch": 0.05965583173996176,
+      "grad_norm": 3.16314435005188,
+      "learning_rate": 3.6491225354972893e-06,
+      "loss": 0.1193,
+      "step": 312
+    },
+    {
+      "epoch": 0.059847036328871894,
+      "grad_norm": 3.749077558517456,
+      "learning_rate": 3.651155827434673e-06,
+      "loss": 0.6098,
+      "step": 313
+    },
+    {
+      "epoch": 0.060038240917782024,
+      "grad_norm": 3.1285479068756104,
+      "learning_rate": 3.6531826335813365e-06,
+      "loss": 0.6629,
+      "step": 314
+    },
+    {
+      "epoch": 0.06022944550669216,
+      "grad_norm": 3.448866128921509,
+      "learning_rate": 3.6552029951824756e-06,
+      "loss": 0.3374,
+      "step": 315
+    },
+    {
+      "epoch": 0.060420650095602296,
+      "grad_norm": 1.7784894704818726,
+      "learning_rate": 3.6572169530910974e-06,
+      "loss": 0.1474,
+      "step": 316
+    },
+    {
+      "epoch": 0.060611854684512426,
+      "grad_norm": 2.4160869121551514,
+      "learning_rate": 3.6592245477729737e-06,
+      "loss": 0.3054,
+      "step": 317
+    },
+    {
+      "epoch": 0.06080305927342256,
+      "grad_norm": 3.9507014751434326,
+      "learning_rate": 3.661225819311523e-06,
+      "loss": 0.2749,
+      "step": 318
+    },
+    {
+      "epoch": 0.0609942638623327,
+      "grad_norm": 1.2071470022201538,
+      "learning_rate": 3.66322080741261e-06,
+      "loss": 0.1813,
+      "step": 319
+    },
+    {
+      "epoch": 0.06118546845124283,
+      "grad_norm": 4.6793646812438965,
+      "learning_rate": 3.665209551409273e-06,
+      "loss": 0.3954,
+      "step": 320
+    },
+    {
+      "epoch": 0.061376673040152964,
+      "grad_norm": 3.9796018600463867,
+      "learning_rate": 3.6671920902663776e-06,
+      "loss": 0.5827,
+      "step": 321
+    },
+    {
+      "epoch": 0.0615678776290631,
+      "grad_norm": 3.5386040210723877,
+      "learning_rate": 3.669168462585197e-06,
+      "loss": 0.5768,
+      "step": 322
+    },
+    {
+      "epoch": 0.06175908221797323,
+      "grad_norm": 1.5878996849060059,
+      "learning_rate": 3.67113870660792e-06,
+      "loss": 0.203,
+      "step": 323
+    },
+    {
+      "epoch": 0.061950286806883366,
+      "grad_norm": 1.8842241764068604,
+      "learning_rate": 3.673102860222095e-06,
+      "loss": 0.1778,
+      "step": 324
+    },
+    {
+      "epoch": 0.0621414913957935,
+      "grad_norm": 2.419157028198242,
+      "learning_rate": 3.675060960964998e-06,
+      "loss": 0.1822,
+      "step": 325
+    },
+    {
+      "epoch": 0.06233269598470363,
+      "grad_norm": 3.7156035900115967,
+      "learning_rate": 3.6770130460279383e-06,
+      "loss": 0.6939,
+      "step": 326
+    },
+    {
+      "epoch": 0.06252390057361376,
+      "grad_norm": 2.3214938640594482,
+      "learning_rate": 3.678959152260497e-06,
+      "loss": 0.3233,
+      "step": 327
+    },
+    {
+      "epoch": 0.0627151051625239,
+      "grad_norm": 6.248021602630615,
+      "learning_rate": 3.6808993161747004e-06,
+      "loss": 0.291,
+      "step": 328
+    },
+    {
+      "epoch": 0.06290630975143403,
+      "grad_norm": 3.342606782913208,
+      "learning_rate": 3.682833573949131e-06,
+      "loss": 0.1396,
+      "step": 329
+    },
+    {
+      "epoch": 0.06309751434034416,
+      "grad_norm": 2.0580525398254395,
+      "learning_rate": 3.6847619614329755e-06,
+      "loss": 0.0865,
+      "step": 330
+    },
+    {
+      "epoch": 0.0632887189292543,
+      "grad_norm": 2.6703569889068604,
+      "learning_rate": 3.68668451415001e-06,
+      "loss": 0.1062,
+      "step": 331
+    },
+    {
+      "epoch": 0.06347992351816444,
+      "grad_norm": 3.680206775665283,
+      "learning_rate": 3.6886012673025277e-06,
+      "loss": 0.7988,
+      "step": 332
+    },
+    {
+      "epoch": 0.06367112810707456,
+      "grad_norm": 2.1449387073516846,
+      "learning_rate": 3.6905122557752073e-06,
+      "loss": 0.2702,
+      "step": 333
+    },
+    {
+      "epoch": 0.06386233269598471,
+      "grad_norm": 3.1063783168792725,
+      "learning_rate": 3.6924175141389183e-06,
+      "loss": 0.2693,
+      "step": 334
+    },
+    {
+      "epoch": 0.06405353728489484,
+      "grad_norm": 2.218688488006592,
+      "learning_rate": 3.6943170766544804e-06,
+      "loss": 0.3532,
+      "step": 335
+    },
+    {
+      "epoch": 0.06424474187380497,
+      "grad_norm": 2.191000461578369,
+      "learning_rate": 3.6962109772763544e-06,
+      "loss": 0.2497,
+      "step": 336
+    },
+    {
+      "epoch": 0.06443594646271511,
+      "grad_norm": 2.405402898788452,
+      "learning_rate": 3.6980992496562857e-06,
+      "loss": 0.1078,
+      "step": 337
+    },
+    {
+      "epoch": 0.06462715105162524,
+      "grad_norm": 2.5214316844940186,
+      "learning_rate": 3.699981927146893e-06,
+      "loss": 0.1989,
+      "step": 338
+    },
+    {
+      "epoch": 0.06481835564053537,
+      "grad_norm": 3.761211633682251,
+      "learning_rate": 3.701859042805201e-06,
+      "loss": 0.4005,
+      "step": 339
+    },
+    {
+      "epoch": 0.06500956022944551,
+      "grad_norm": 1.934031367301941,
+      "learning_rate": 3.7037306293961262e-06,
+      "loss": 0.4611,
+      "step": 340
+    },
+    {
+      "epoch": 0.06520076481835564,
+      "grad_norm": 2.1387293338775635,
+      "learning_rate": 3.7055967193959047e-06,
+      "loss": 0.2271,
+      "step": 341
+    },
+    {
+      "epoch": 0.06539196940726577,
+      "grad_norm": 2.621551990509033,
+      "learning_rate": 3.7074573449954767e-06,
+      "loss": 0.1125,
+      "step": 342
+    },
+    {
+      "epoch": 0.06558317399617591,
+      "grad_norm": 1.7968058586120605,
+      "learning_rate": 3.709312538103816e-06,
+      "loss": 0.113,
+      "step": 343
+    },
+    {
+      "epoch": 0.06577437858508604,
+      "grad_norm": 2.509631395339966,
+      "learning_rate": 3.7111623303512168e-06,
+      "loss": 0.2285,
+      "step": 344
+    },
+    {
+      "epoch": 0.06596558317399617,
+      "grad_norm": 2.8840248584747314,
+      "learning_rate": 3.7130067530925253e-06,
+      "loss": 0.2937,
+      "step": 345
+    },
+    {
+      "epoch": 0.06615678776290632,
+      "grad_norm": 3.3847973346710205,
+      "learning_rate": 3.7148458374103296e-06,
+      "loss": 0.1772,
+      "step": 346
+    },
+    {
+      "epoch": 0.06634799235181645,
+      "grad_norm": 2.8074636459350586,
+      "learning_rate": 3.716679614118107e-06,
+      "loss": 0.3393,
+      "step": 347
+    },
+    {
+      "epoch": 0.06653919694072657,
+      "grad_norm": 9.130067825317383,
+      "learning_rate": 3.7185081137633166e-06,
+      "loss": 1.0307,
+      "step": 348
+    },
+    {
+      "epoch": 0.06673040152963672,
+      "grad_norm": 1.991129994392395,
+      "learning_rate": 3.7203313666304545e-06,
+      "loss": 0.1372,
+      "step": 349
+    },
+    {
+      "epoch": 0.06692160611854685,
+      "grad_norm": 3.1855194568634033,
+      "learning_rate": 3.722149402744063e-06,
+      "loss": 0.1485,
+      "step": 350
+    },
+    {
+      "epoch": 0.06711281070745698,
+      "grad_norm": 3.3848493099212646,
+      "learning_rate": 3.7239622518716983e-06,
+      "loss": 0.5177,
+      "step": 351
+    },
+    {
+      "epoch": 0.0673040152963671,
+      "grad_norm": 3.134887933731079,
+      "learning_rate": 3.725769943526854e-06,
+      "loss": 0.6924,
+      "step": 352
+    },
+    {
+      "epoch": 0.06749521988527725,
+      "grad_norm": 3.605255126953125,
+      "learning_rate": 3.727572506971844e-06,
+      "loss": 0.4911,
+      "step": 353
+    },
+    {
+      "epoch": 0.06768642447418738,
+      "grad_norm": 2.0924346446990967,
+      "learning_rate": 3.7293699712206464e-06,
+      "loss": 0.2743,
+      "step": 354
+    },
+    {
+      "epoch": 0.06787762906309751,
+      "grad_norm": 3.687945604324341,
+      "learning_rate": 3.731162365041703e-06,
+      "loss": 0.1613,
+      "step": 355
+    },
+    {
+      "epoch": 0.06806883365200765,
+      "grad_norm": 5.654581546783447,
+      "learning_rate": 3.7329497169606825e-06,
+      "loss": 0.2055,
+      "step": 356
+    },
+    {
+      "epoch": 0.06826003824091778,
+      "grad_norm": 1.5821623802185059,
+      "learning_rate": 3.734732055263207e-06,
+      "loss": 0.1343,
+      "step": 357
+    },
+    {
+      "epoch": 0.06845124282982791,
+      "grad_norm": 3.3485209941864014,
+      "learning_rate": 3.736509407997536e-06,
+      "loss": 0.6085,
+      "step": 358
+    },
+    {
+      "epoch": 0.06864244741873805,
+      "grad_norm": 3.3173418045043945,
+      "learning_rate": 3.738281802977213e-06,
+      "loss": 0.3427,
+      "step": 359
+    },
+    {
+      "epoch": 0.06883365200764818,
+      "grad_norm": 2.4630320072174072,
+      "learning_rate": 3.7400492677836824e-06,
+      "loss": 0.1692,
+      "step": 360
+    },
+    {
+      "epoch": 0.06902485659655831,
+      "grad_norm": 2.6465280055999756,
+      "learning_rate": 3.7418118297688577e-06,
+      "loss": 0.1822,
+      "step": 361
+    },
+    {
+      "epoch": 0.06921606118546846,
+      "grad_norm": 2.0654947757720947,
+      "learning_rate": 3.743569516057668e-06,
+      "loss": 0.091,
+      "step": 362
+    },
+    {
+      "epoch": 0.06940726577437858,
+      "grad_norm": 3.139086961746216,
+      "learning_rate": 3.7453223535505566e-06,
+      "loss": 0.2461,
+      "step": 363
+    },
+    {
+      "epoch": 0.06959847036328871,
+      "grad_norm": 3.0427610874176025,
+      "learning_rate": 3.747070368925958e-06,
+      "loss": 0.3526,
+      "step": 364
+    },
+    {
+      "epoch": 0.06978967495219886,
+      "grad_norm": 3.79207181930542,
+      "learning_rate": 3.7488135886427275e-06,
+      "loss": 0.4793,
+      "step": 365
+    },
+    {
+      "epoch": 0.06998087954110899,
+      "grad_norm": 1.8372082710266113,
+      "learning_rate": 3.750552038942548e-06,
+      "loss": 0.1645,
+      "step": 366
+    },
+    {
+      "epoch": 0.07017208413001912,
+      "grad_norm": 2.4285526275634766,
+      "learning_rate": 3.7522857458523022e-06,
+      "loss": 0.1195,
+      "step": 367
+    },
+    {
+      "epoch": 0.07036328871892926,
+      "grad_norm": 2.658700942993164,
+      "learning_rate": 3.7540147351864037e-06,
+      "loss": 0.2826,
+      "step": 368
+    },
+    {
+      "epoch": 0.07055449330783939,
+      "grad_norm": 3.1620702743530273,
+      "learning_rate": 3.7557390325491095e-06,
+      "loss": 0.2795,
+      "step": 369
+    },
+    {
+      "epoch": 0.07074569789674952,
+      "grad_norm": 2.690842628479004,
+      "learning_rate": 3.757458663336794e-06,
+      "loss": 0.4693,
+      "step": 370
+    },
+    {
+      "epoch": 0.07093690248565966,
+      "grad_norm": 2.008033275604248,
+      "learning_rate": 3.7591736527401917e-06,
+      "loss": 0.2736,
+      "step": 371
+    },
+    {
+      "epoch": 0.07112810707456979,
+      "grad_norm": 3.240271806716919,
+      "learning_rate": 3.760884025746611e-06,
+      "loss": 0.3499,
+      "step": 372
+    },
+    {
+      "epoch": 0.07131931166347992,
+      "grad_norm": 1.3944287300109863,
+      "learning_rate": 3.7625898071421226e-06,
+      "loss": 0.0905,
+      "step": 373
+    },
+    {
+      "epoch": 0.07151051625239006,
+      "grad_norm": 2.2824788093566895,
+      "learning_rate": 3.7642910215137073e-06,
+      "loss": 0.2282,
+      "step": 374
+    },
+    {
+      "epoch": 0.07170172084130019,
+      "grad_norm": 4.351092338562012,
+      "learning_rate": 3.765987693251391e-06,
+      "loss": 0.3069,
+      "step": 375
+    },
+    {
+      "epoch": 0.07189292543021032,
+      "grad_norm": 4.460353851318359,
+      "learning_rate": 3.767679846550338e-06,
+      "loss": 0.6941,
+      "step": 376
+    },
+    {
+      "epoch": 0.07208413001912047,
+      "grad_norm": 2.765394926071167,
+      "learning_rate": 3.76936750541292e-06,
+      "loss": 0.4267,
+      "step": 377
+    },
+    {
+      "epoch": 0.0722753346080306,
+      "grad_norm": 2.731600284576416,
+      "learning_rate": 3.771050693650764e-06,
+      "loss": 0.3801,
+      "step": 378
+    },
+    {
+      "epoch": 0.07246653919694072,
+      "grad_norm": 1.8360090255737305,
+      "learning_rate": 3.772729434886761e-06,
+      "loss": 0.2339,
+      "step": 379
+    },
+    {
+      "epoch": 0.07265774378585087,
+      "grad_norm": 1.9759913682937622,
+      "learning_rate": 3.774403752557064e-06,
+      "loss": 0.3144,
+      "step": 380
+    },
+    {
+      "epoch": 0.072848948374761,
+      "grad_norm": 1.7601401805877686,
+      "learning_rate": 3.7760736699130425e-06,
+      "loss": 0.1128,
+      "step": 381
+    },
+    {
+      "epoch": 0.07304015296367113,
+      "grad_norm": 4.08651065826416,
+      "learning_rate": 3.7777392100232265e-06,
+      "loss": 0.4442,
+      "step": 382
+    },
+    {
+      "epoch": 0.07323135755258126,
+      "grad_norm": 3.4187510013580322,
+      "learning_rate": 3.7794003957752135e-06,
+      "loss": 0.4448,
+      "step": 383
+    },
+    {
+      "epoch": 0.0734225621414914,
+      "grad_norm": 1.9716267585754395,
+      "learning_rate": 3.781057249877561e-06,
+      "loss": 0.3026,
+      "step": 384
+    },
+    {
+      "epoch": 0.07361376673040153,
+      "grad_norm": 1.3850023746490479,
+      "learning_rate": 3.7827097948616442e-06,
+      "loss": 0.2962,
+      "step": 385
+    },
+    {
+      "epoch": 0.07380497131931166,
+      "grad_norm": 3.1978061199188232,
+      "learning_rate": 3.784358053083497e-06,
+      "loss": 0.4785,
+      "step": 386
+    },
+    {
+      "epoch": 0.0739961759082218,
+      "grad_norm": 2.065589666366577,
+      "learning_rate": 3.786002046725626e-06,
+      "loss": 0.1068,
+      "step": 387
+    },
+    {
+      "epoch": 0.07418738049713193,
+      "grad_norm": 2.613898277282715,
+      "learning_rate": 3.7876417977988023e-06,
+      "loss": 0.2658,
+      "step": 388
+    },
+    {
+      "epoch": 0.07437858508604206,
+      "grad_norm": 2.723231077194214,
+      "learning_rate": 3.7892773281438285e-06,
+      "loss": 0.451,
+      "step": 389
+    },
+    {
+      "epoch": 0.0745697896749522,
+      "grad_norm": 11.916027069091797,
+      "learning_rate": 3.7909086594332863e-06,
+      "loss": 0.4356,
+      "step": 390
+    },
+    {
+      "epoch": 0.07476099426386233,
+      "grad_norm": 1.5050503015518188,
+      "learning_rate": 3.792535813173256e-06,
+      "loss": 0.1137,
+      "step": 391
+    },
+    {
+      "epoch": 0.07495219885277246,
+      "grad_norm": 2.0423741340637207,
+      "learning_rate": 3.7941588107050227e-06,
+      "loss": 0.228,
+      "step": 392
+    },
+    {
+      "epoch": 0.0751434034416826,
+      "grad_norm": 5.4225969314575195,
+      "learning_rate": 3.795777673206752e-06,
+      "loss": 0.3271,
+      "step": 393
+    },
+    {
+      "epoch": 0.07533460803059273,
+      "grad_norm": 2.6373708248138428,
+      "learning_rate": 3.7973924216951487e-06,
+      "loss": 0.2295,
+      "step": 394
+    },
+    {
+      "epoch": 0.07552581261950286,
+      "grad_norm": 7.188790798187256,
+      "learning_rate": 3.7990030770270936e-06,
+      "loss": 0.2779,
+      "step": 395
+    },
+    {
+      "epoch": 0.075717017208413,
+      "grad_norm": 2.465902805328369,
+      "learning_rate": 3.800609659901264e-06,
+      "loss": 0.2997,
+      "step": 396
+    },
+    {
+      "epoch": 0.07590822179732314,
+      "grad_norm": 2.4251866340637207,
+      "learning_rate": 3.802212190859722e-06,
+      "loss": 0.3933,
+      "step": 397
+    },
+    {
+      "epoch": 0.07609942638623327,
+      "grad_norm": 2.4671387672424316,
+      "learning_rate": 3.803810690289499e-06,
+      "loss": 0.2564,
+      "step": 398
+    },
+    {
+      "epoch": 0.07629063097514341,
+      "grad_norm": 2.4797089099884033,
+      "learning_rate": 3.8054051784241454e-06,
+      "loss": 0.1865,
+      "step": 399
+    },
+    {
+      "epoch": 0.07648183556405354,
+      "grad_norm": 2.997403860092163,
+      "learning_rate": 3.8069956753452695e-06,
+      "loss": 0.3207,
+      "step": 400
+    },
+    {
+      "epoch": 0.07667304015296367,
+      "grad_norm": 1.8929557800292969,
+      "learning_rate": 3.808582200984058e-06,
+      "loss": 0.1249,
+      "step": 401
+    },
+    {
+      "epoch": 0.07686424474187381,
+      "grad_norm": 3.276320695877075,
+      "learning_rate": 3.8101647751227683e-06,
+      "loss": 0.4781,
+      "step": 402
+    },
+    {
+      "epoch": 0.07705544933078394,
+      "grad_norm": 1.9976557493209839,
+      "learning_rate": 3.811743417396214e-06,
+      "loss": 0.184,
+      "step": 403
+    },
+    {
+      "epoch": 0.07724665391969407,
+      "grad_norm": 3.980438232421875,
+      "learning_rate": 3.813318147293226e-06,
+      "loss": 0.4318,
+      "step": 404
+    },
+    {
+      "epoch": 0.07743785850860421,
+      "grad_norm": 2.1951019763946533,
+      "learning_rate": 3.8148889841580915e-06,
+      "loss": 0.2133,
+      "step": 405
+    },
+    {
+      "epoch": 0.07762906309751434,
+      "grad_norm": 4.515348434448242,
+      "learning_rate": 3.816455947191985e-06,
+      "loss": 0.5805,
+      "step": 406
+    },
+    {
+      "epoch": 0.07782026768642447,
+      "grad_norm": 2.6555490493774414,
+      "learning_rate": 3.818019055454375e-06,
+      "loss": 0.3141,
+      "step": 407
+    },
+    {
+      "epoch": 0.07801147227533461,
+      "grad_norm": 1.6512001752853394,
+      "learning_rate": 3.819578327864414e-06,
+      "loss": 0.2455,
+      "step": 408
+    },
+    {
+      "epoch": 0.07820267686424474,
+      "grad_norm": 5.637848854064941,
+      "learning_rate": 3.821133783202312e-06,
+      "loss": 0.335,
+      "step": 409
+    },
+    {
+      "epoch": 0.07839388145315487,
+      "grad_norm": 1.5848608016967773,
+      "learning_rate": 3.8226854401106974e-06,
+      "loss": 0.2007,
+      "step": 410
+    },
+    {
+      "epoch": 0.078585086042065,
+      "grad_norm": 2.4538025856018066,
+      "learning_rate": 3.824233317095951e-06,
+      "loss": 0.309,
+      "step": 411
+    },
+    {
+      "epoch": 0.07877629063097515,
+      "grad_norm": 3.901400089263916,
+      "learning_rate": 3.82577743252954e-06,
+      "loss": 0.1759,
+      "step": 412
+    },
+    {
+      "epoch": 0.07896749521988528,
+      "grad_norm": 3.340503454208374,
+      "learning_rate": 3.8273178046493155e-06,
+      "loss": 0.4415,
+      "step": 413
+    },
+    {
+      "epoch": 0.0791586998087954,
+      "grad_norm": 3.1490180492401123,
+      "learning_rate": 3.828854451560812e-06,
+      "loss": 0.3878,
+      "step": 414
+    },
+    {
+      "epoch": 0.07934990439770555,
+      "grad_norm": 1.568618893623352,
+      "learning_rate": 3.830387391238524e-06,
+      "loss": 0.2918,
+      "step": 415
+    },
+    {
+      "epoch": 0.07954110898661568,
+      "grad_norm": 2.177776336669922,
+      "learning_rate": 3.831916641527165e-06,
+      "loss": 0.1442,
+      "step": 416
+    },
+    {
+      "epoch": 0.0797323135755258,
+      "grad_norm": 2.933709144592285,
+      "learning_rate": 3.833442220142912e-06,
+      "loss": 0.4187,
+      "step": 417
+    },
+    {
+      "epoch": 0.07992351816443595,
+      "grad_norm": 1.9458264112472534,
+      "learning_rate": 3.834964144674645e-06,
+      "loss": 0.0881,
+      "step": 418
+    },
+    {
+      "epoch": 0.08011472275334608,
+      "grad_norm": 2.3854639530181885,
+      "learning_rate": 3.836482432585154e-06,
+      "loss": 0.1411,
+      "step": 419
+    },
+    {
+      "epoch": 0.08030592734225621,
+      "grad_norm": 2.503591299057007,
+      "learning_rate": 3.837997101212351e-06,
+      "loss": 0.2308,
+      "step": 420
+    },
+    {
+      "epoch": 0.08049713193116635,
+      "grad_norm": 2.7284929752349854,
+      "learning_rate": 3.839508167770447e-06,
+      "loss": 0.5046,
+      "step": 421
+    },
+    {
+      "epoch": 0.08068833652007648,
+      "grad_norm": 2.8158528804779053,
+      "learning_rate": 3.8410156493511355e-06,
+      "loss": 0.4343,
+      "step": 422
+    },
+    {
+      "epoch": 0.08087954110898661,
+      "grad_norm": 2.605060577392578,
+      "learning_rate": 3.842519562924747e-06,
+      "loss": 0.242,
+      "step": 423
+    },
+    {
+      "epoch": 0.08107074569789675,
+      "grad_norm": 2.482698440551758,
+      "learning_rate": 3.8440199253413985e-06,
+      "loss": 0.2665,
+      "step": 424
+    },
+    {
+      "epoch": 0.08126195028680688,
+      "grad_norm": 2.105672836303711,
+      "learning_rate": 3.8455167533321224e-06,
+      "loss": 0.1729,
+      "step": 425
+    },
+    {
+      "epoch": 0.08145315487571701,
+      "grad_norm": 4.377143383026123,
+      "learning_rate": 3.847010063509991e-06,
+      "loss": 0.5099,
+      "step": 426
+    },
+    {
+      "epoch": 0.08164435946462716,
+      "grad_norm": 1.9579622745513916,
+      "learning_rate": 3.848499872371217e-06,
+      "loss": 0.1786,
+      "step": 427
+    },
+    {
+      "epoch": 0.08183556405353729,
+      "grad_norm": 2.5516083240509033,
+      "learning_rate": 3.849986196296252e-06,
+      "loss": 0.1474,
+      "step": 428
+    },
+    {
+      "epoch": 0.08202676864244741,
+      "grad_norm": 2.7220704555511475,
+      "learning_rate": 3.851469051550867e-06,
+      "loss": 0.3341,
+      "step": 429
+    },
+    {
+      "epoch": 0.08221797323135756,
+      "grad_norm": 1.9412697553634644,
+      "learning_rate": 3.852948454287213e-06,
+      "loss": 0.163,
+      "step": 430
+    },
+    {
+      "epoch": 0.08240917782026769,
+      "grad_norm": 2.0283761024475098,
+      "learning_rate": 3.854424420544888e-06,
+      "loss": 0.0943,
+      "step": 431
+    },
+    {
+      "epoch": 0.08260038240917782,
+      "grad_norm": 2.609199285507202,
+      "learning_rate": 3.855896966251971e-06,
+      "loss": 0.246,
+      "step": 432
+    },
+    {
+      "epoch": 0.08279158699808796,
+      "grad_norm": 1.5453547239303589,
+      "learning_rate": 3.857366107226057e-06,
+      "loss": 0.224,
+      "step": 433
+    },
+    {
+      "epoch": 0.08298279158699809,
+      "grad_norm": 1.983446717262268,
+      "learning_rate": 3.85883185917528e-06,
+      "loss": 0.2313,
+      "step": 434
+    },
+    {
+      "epoch": 0.08317399617590822,
+      "grad_norm": 2.101289749145508,
+      "learning_rate": 3.860294237699313e-06,
+      "loss": 0.1372,
+      "step": 435
+    },
+    {
+      "epoch": 0.08336520076481836,
+      "grad_norm": 1.78105628490448,
+      "learning_rate": 3.861753258290372e-06,
+      "loss": 0.1718,
+      "step": 436
+    },
+    {
+      "epoch": 0.08355640535372849,
+      "grad_norm": 2.88455867767334,
+      "learning_rate": 3.863208936334194e-06,
+      "loss": 0.1571,
+      "step": 437
+    },
+    {
+      "epoch": 0.08374760994263862,
+      "grad_norm": 2.6141340732574463,
+      "learning_rate": 3.864661287111015e-06,
+      "loss": 0.2811,
+      "step": 438
+    },
+    {
+      "epoch": 0.08393881453154876,
+      "grad_norm": 2.1665003299713135,
+      "learning_rate": 3.866110325796531e-06,
+      "loss": 0.1615,
+      "step": 439
+    },
+    {
+      "epoch": 0.0841300191204589,
+      "grad_norm": 2.2355403900146484,
+      "learning_rate": 3.86755606746285e-06,
+      "loss": 0.2134,
+      "step": 440
+    },
+    {
+      "epoch": 0.08432122370936902,
+      "grad_norm": 25.508962631225586,
+      "learning_rate": 3.868998527079432e-06,
+      "loss": 0.5128,
+      "step": 441
+    },
+    {
+      "epoch": 0.08451242829827915,
+      "grad_norm": 3.2912681102752686,
+      "learning_rate": 3.870437719514017e-06,
+      "loss": 0.2624,
+      "step": 442
+    },
+    {
+      "epoch": 0.0847036328871893,
+      "grad_norm": 2.363339900970459,
+      "learning_rate": 3.871873659533549e-06,
+      "loss": 0.1605,
+      "step": 443
+    },
+    {
+      "epoch": 0.08489483747609942,
+      "grad_norm": 4.588367938995361,
+      "learning_rate": 3.873306361805082e-06,
+      "loss": 0.3243,
+      "step": 444
+    },
+    {
+      "epoch": 0.08508604206500955,
+      "grad_norm": 4.233880519866943,
+      "learning_rate": 3.874735840896679e-06,
+      "loss": 0.6475,
+      "step": 445
+    },
+    {
+      "epoch": 0.0852772466539197,
+      "grad_norm": 2.803955316543579,
+      "learning_rate": 3.876162111278302e-06,
+      "loss": 0.2044,
+      "step": 446
+    },
+    {
+      "epoch": 0.08546845124282983,
+      "grad_norm": 2.126622438430786,
+      "learning_rate": 3.877585187322691e-06,
+      "loss": 0.1412,
+      "step": 447
+    },
+    {
+      "epoch": 0.08565965583173996,
+      "grad_norm": 2.711604356765747,
+      "learning_rate": 3.879005083306229e-06,
+      "loss": 0.4514,
+      "step": 448
+    },
+    {
+      "epoch": 0.0858508604206501,
+      "grad_norm": 1.7695578336715698,
+      "learning_rate": 3.88042181340981e-06,
+      "loss": 0.0854,
+      "step": 449
+    },
+    {
+      "epoch": 0.08604206500956023,
+      "grad_norm": 3.0042665004730225,
+      "learning_rate": 3.881835391719679e-06,
+      "loss": 0.1316,
+      "step": 450
+    },
+    {
+      "epoch": 0.08623326959847036,
+      "grad_norm": 4.021963119506836,
+      "learning_rate": 3.883245832228278e-06,
+      "loss": 0.5587,
+      "step": 451
+    },
+    {
+      "epoch": 0.0864244741873805,
+      "grad_norm": 2.566257953643799,
+      "learning_rate": 3.884653148835076e-06,
+      "loss": 0.5083,
+      "step": 452
+    },
+    {
+      "epoch": 0.08661567877629063,
+      "grad_norm": 2.0462639331817627,
+      "learning_rate": 3.8860573553473914e-06,
+      "loss": 0.161,
+      "step": 453
+    },
+    {
+      "epoch": 0.08680688336520076,
+      "grad_norm": 2.4460551738739014,
+      "learning_rate": 3.887458465481203e-06,
+      "loss": 0.1692,
+      "step": 454
+    },
+    {
+      "epoch": 0.0869980879541109,
+      "grad_norm": 2.9268269538879395,
+      "learning_rate": 3.888856492861955e-06,
+      "loss": 0.3161,
+      "step": 455
+    },
+    {
+      "epoch": 0.08718929254302103,
+      "grad_norm": 2.3640105724334717,
+      "learning_rate": 3.890251451025352e-06,
+      "loss": 0.1184,
+      "step": 456
+    },
+    {
+      "epoch": 0.08738049713193116,
+      "grad_norm": 3.2469921112060547,
+      "learning_rate": 3.8916433534181465e-06,
+      "loss": 0.3819,
+      "step": 457
+    },
+    {
+      "epoch": 0.0875717017208413,
+      "grad_norm": 3.044752359390259,
+      "learning_rate": 3.893032213398913e-06,
+      "loss": 0.5221,
+      "step": 458
+    },
+    {
+      "epoch": 0.08776290630975143,
+      "grad_norm": 1.492080807685852,
+      "learning_rate": 3.894418044238823e-06,
+      "loss": 0.1747,
+      "step": 459
+    },
+    {
+      "epoch": 0.08795411089866156,
+      "grad_norm": 1.967621922492981,
+      "learning_rate": 3.8958008591224e-06,
+      "loss": 0.4279,
+      "step": 460
+    },
+    {
+      "epoch": 0.08814531548757171,
+      "grad_norm": 4.419346809387207,
+      "learning_rate": 3.897180671148275e-06,
+      "loss": 0.222,
+      "step": 461
+    },
+    {
+      "epoch": 0.08833652007648184,
+      "grad_norm": 3.055210590362549,
+      "learning_rate": 3.898557493329932e-06,
+      "loss": 0.203,
+      "step": 462
+    },
+    {
+      "epoch": 0.08852772466539197,
+      "grad_norm": 2.82248854637146,
+      "learning_rate": 3.899931338596441e-06,
+      "loss": 0.2677,
+      "step": 463
+    },
+    {
+      "epoch": 0.08871892925430211,
+      "grad_norm": 4.254226207733154,
+      "learning_rate": 3.901302219793192e-06,
+      "loss": 0.6093,
+      "step": 464
+    },
+    {
+      "epoch": 0.08891013384321224,
+      "grad_norm": 3.696042537689209,
+      "learning_rate": 3.902670149682608e-06,
+      "loss": 0.6688,
+      "step": 465
+    },
+    {
+      "epoch": 0.08910133843212237,
+      "grad_norm": 2.048828601837158,
+      "learning_rate": 3.904035140944864e-06,
+      "loss": 0.0932,
+      "step": 466
+    },
+    {
+      "epoch": 0.08929254302103251,
+      "grad_norm": 1.4884488582611084,
+      "learning_rate": 3.905397206178592e-06,
+      "loss": 0.1926,
+      "step": 467
+    },
+    {
+      "epoch": 0.08948374760994264,
+      "grad_norm": 4.860929012298584,
+      "learning_rate": 3.906756357901574e-06,
+      "loss": 0.2957,
+      "step": 468
+    },
+    {
+      "epoch": 0.08967495219885277,
+      "grad_norm": 3.0578386783599854,
+      "learning_rate": 3.908112608551437e-06,
+      "loss": 0.2792,
+      "step": 469
+    },
+    {
+      "epoch": 0.08986615678776291,
+      "grad_norm": 2.4391660690307617,
+      "learning_rate": 3.9094659704863346e-06,
+      "loss": 0.3934,
+      "step": 470
+    },
+    {
+      "epoch": 0.09005736137667304,
+      "grad_norm": 3.1707816123962402,
+      "learning_rate": 3.910816455985621e-06,
+      "loss": 0.4196,
+      "step": 471
+    },
+    {
+      "epoch": 0.09024856596558317,
+      "grad_norm": 3.7489700317382812,
+      "learning_rate": 3.912164077250522e-06,
+      "loss": 0.5584,
+      "step": 472
+    },
+    {
+      "epoch": 0.0904397705544933,
+      "grad_norm": 3.4774014949798584,
+      "learning_rate": 3.9135088464047945e-06,
+      "loss": 0.3015,
+      "step": 473
+    },
+    {
+      "epoch": 0.09063097514340344,
+      "grad_norm": 3.203939914703369,
+      "learning_rate": 3.9148507754953815e-06,
+      "loss": 0.1764,
+      "step": 474
+    },
+    {
+      "epoch": 0.09082217973231357,
+      "grad_norm": 5.508978366851807,
+      "learning_rate": 3.91618987649306e-06,
+      "loss": 0.1797,
+      "step": 475
+    },
+    {
+      "epoch": 0.0910133843212237,
+      "grad_norm": 3.0496325492858887,
+      "learning_rate": 3.917526161293082e-06,
+      "loss": 0.2904,
+      "step": 476
+    },
+    {
+      "epoch": 0.09120458891013385,
+      "grad_norm": 2.569993257522583,
+      "learning_rate": 3.9188596417158075e-06,
+      "loss": 0.2979,
+      "step": 477
+    },
+    {
+      "epoch": 0.09139579349904398,
+      "grad_norm": 2.830799102783203,
+      "learning_rate": 3.920190329507332e-06,
+      "loss": 0.1373,
+      "step": 478
+    },
+    {
+      "epoch": 0.0915869980879541,
+      "grad_norm": 1.5657826662063599,
+      "learning_rate": 3.921518236340108e-06,
+      "loss": 0.1339,
+      "step": 479
+    },
+    {
+      "epoch": 0.09177820267686425,
+      "grad_norm": 1.5757731199264526,
+      "learning_rate": 3.922843373813557e-06,
+      "loss": 0.1992,
+      "step": 480
+    },
+    {
+      "epoch": 0.09196940726577438,
+      "grad_norm": 3.687089443206787,
+      "learning_rate": 3.924165753454686e-06,
+      "loss": 0.1859,
+      "step": 481
+    },
+    {
+      "epoch": 0.09216061185468451,
+      "grad_norm": 4.091752052307129,
+      "learning_rate": 3.925485386718674e-06,
+      "loss": 0.2528,
+      "step": 482
+    },
+    {
+      "epoch": 0.09235181644359465,
+      "grad_norm": 1.9640034437179565,
+      "learning_rate": 3.926802284989481e-06,
+      "loss": 0.3403,
+      "step": 483
+    },
+    {
+      "epoch": 0.09254302103250478,
+      "grad_norm": 2.387704849243164,
+      "learning_rate": 3.928116459580432e-06,
+      "loss": 0.4675,
+      "step": 484
+    },
+    {
+      "epoch": 0.09273422562141491,
+      "grad_norm": 1.9029995203018188,
+      "learning_rate": 3.9294279217347985e-06,
+      "loss": 0.2266,
+      "step": 485
+    },
+    {
+      "epoch": 0.09292543021032505,
+      "grad_norm": 3.2065725326538086,
+      "learning_rate": 3.930736682626379e-06,
+      "loss": 0.25,
+      "step": 486
+    },
+    {
+      "epoch": 0.09311663479923518,
+      "grad_norm": 2.709934711456299,
+      "learning_rate": 3.932042753360069e-06,
+      "loss": 0.2074,
+      "step": 487
+    },
+    {
+      "epoch": 0.09330783938814531,
+      "grad_norm": 3.0203776359558105,
+      "learning_rate": 3.933346144972424e-06,
+      "loss": 0.3794,
+      "step": 488
+    },
+    {
+      "epoch": 0.09349904397705545,
+      "grad_norm": 3.989088296890259,
+      "learning_rate": 3.934646868432223e-06,
+      "loss": 0.3628,
+      "step": 489
+    },
+    {
+      "epoch": 0.09369024856596558,
+      "grad_norm": 1.333524465560913,
+      "learning_rate": 3.93594493464102e-06,
+      "loss": 0.2219,
+      "step": 490
+    },
+    {
+      "epoch": 0.09388145315487571,
+      "grad_norm": 2.184206247329712,
+      "learning_rate": 3.937240354433692e-06,
+      "loss": 0.4398,
+      "step": 491
+    },
+    {
+      "epoch": 0.09407265774378586,
+      "grad_norm": 2.465756893157959,
+      "learning_rate": 3.9385331385789845e-06,
+      "loss": 0.1526,
+      "step": 492
+    },
+    {
+      "epoch": 0.09426386233269599,
+      "grad_norm": 2.6147758960723877,
+      "learning_rate": 3.939823297780045e-06,
+      "loss": 0.1571,
+      "step": 493
+    },
+    {
+      "epoch": 0.09445506692160612,
+      "grad_norm": 2.757384777069092,
+      "learning_rate": 3.941110842674955e-06,
+      "loss": 0.178,
+      "step": 494
+    },
+    {
+      "epoch": 0.09464627151051626,
+      "grad_norm": 3.065559148788452,
+      "learning_rate": 3.94239578383726e-06,
+      "loss": 0.5259,
+      "step": 495
+    },
+    {
+      "epoch": 0.09483747609942639,
+      "grad_norm": 3.7982611656188965,
+      "learning_rate": 3.9436781317764865e-06,
+      "loss": 0.6654,
+      "step": 496
+    },
+    {
+      "epoch": 0.09502868068833652,
+      "grad_norm": 1.8163230419158936,
+      "learning_rate": 3.944957896938659e-06,
+      "loss": 0.2204,
+      "step": 497
+    },
+    {
+      "epoch": 0.09521988527724666,
+      "grad_norm": 2.6650335788726807,
+      "learning_rate": 3.946235089706813e-06,
+      "loss": 0.2265,
+      "step": 498
+    },
+    {
+      "epoch": 0.09541108986615679,
+      "grad_norm": 1.7298262119293213,
+      "learning_rate": 3.947509720401494e-06,
+      "loss": 0.0825,
+      "step": 499
+    },
+    {
+      "epoch": 0.09560229445506692,
+      "grad_norm": 2.199256181716919,
+      "learning_rate": 3.948781799281266e-06,
+      "loss": 0.0587,
+      "step": 500
+    },
+    {
+      "epoch": 0.09560229445506692,
+      "eval_runtime": 758.2728,
+      "eval_samples_per_second": 2.023,
+      "eval_steps_per_second": 0.253,
+      "step": 500
+    },
+    {
+      "epoch": 0.09579349904397706,
+      "grad_norm": 2.560211658477783,
+      "learning_rate": 3.950051336543202e-06,
+      "loss": 0.3781,
+      "step": 501
+    },
+    {
+      "epoch": 0.09598470363288719,
+      "grad_norm": 2.047006845474243,
+      "learning_rate": 3.951318342323376e-06,
+      "loss": 0.3203,
+      "step": 502
+    },
+    {
+      "epoch": 0.09617590822179732,
+      "grad_norm": 2.891329050064087,
+      "learning_rate": 3.952582826697346e-06,
+      "loss": 0.3811,
+      "step": 503
+    },
+    {
+      "epoch": 0.09636711281070745,
+      "grad_norm": 2.8886239528656006,
+      "learning_rate": 3.953844799680639e-06,
+      "loss": 0.269,
+      "step": 504
+    },
+    {
+      "epoch": 0.0965583173996176,
+      "grad_norm": 1.302022099494934,
+      "learning_rate": 3.955104271229223e-06,
+      "loss": 0.0563,
+      "step": 505
+    },
+    {
+      "epoch": 0.09674952198852772,
+      "grad_norm": 2.516767740249634,
+      "learning_rate": 3.9563612512399815e-06,
+      "loss": 0.1752,
+      "step": 506
+    },
+    {
+      "epoch": 0.09694072657743785,
+      "grad_norm": 5.090005874633789,
+      "learning_rate": 3.957615749551177e-06,
+      "loss": 0.3137,
+      "step": 507
+    },
+    {
+      "epoch": 0.097131931166348,
+      "grad_norm": 2.3271141052246094,
+      "learning_rate": 3.958867775942918e-06,
+      "loss": 0.2331,
+      "step": 508
+    },
+    {
+      "epoch": 0.09732313575525812,
+      "grad_norm": 5.051674842834473,
+      "learning_rate": 3.960117340137608e-06,
+      "loss": 0.2622,
+      "step": 509
+    },
+    {
+      "epoch": 0.09751434034416825,
+      "grad_norm": 2.13322114944458,
+      "learning_rate": 3.96136445180041e-06,
+      "loss": 0.1548,
+      "step": 510
+    },
+    {
+      "epoch": 0.0977055449330784,
+      "grad_norm": 2.8355138301849365,
+      "learning_rate": 3.962609120539684e-06,
+      "loss": 0.1487,
+      "step": 511
+    },
+    {
+      "epoch": 0.09789674952198853,
+      "grad_norm": 1.6841158866882324,
+      "learning_rate": 3.963851355907436e-06,
+      "loss": 0.1633,
+      "step": 512
+    },
+    {
+      "epoch": 0.09808795411089866,
+      "grad_norm": 3.454113483428955,
+      "learning_rate": 3.965091167399761e-06,
+      "loss": 0.2414,
+      "step": 513
+    },
+    {
+      "epoch": 0.0982791586998088,
+      "grad_norm": 3.103166103363037,
+      "learning_rate": 3.966328564457273e-06,
+      "loss": 0.2673,
+      "step": 514
+    },
+    {
+      "epoch": 0.09847036328871893,
+      "grad_norm": 3.4055354595184326,
+      "learning_rate": 3.967563556465537e-06,
+      "loss": 0.434,
+      "step": 515
+    },
+    {
+      "epoch": 0.09866156787762906,
+      "grad_norm": 2.7894411087036133,
+      "learning_rate": 3.968796152755501e-06,
+      "loss": 0.2902,
+      "step": 516
+    },
+    {
+      "epoch": 0.0988527724665392,
+      "grad_norm": 2.1280155181884766,
+      "learning_rate": 3.970026362603916e-06,
+      "loss": 0.2733,
+      "step": 517
+    },
+    {
+      "epoch": 0.09904397705544933,
+      "grad_norm": 3.4343643188476562,
+      "learning_rate": 3.97125419523375e-06,
+      "loss": 0.5206,
+      "step": 518
+    },
+    {
+      "epoch": 0.09923518164435946,
+      "grad_norm": 2.7457900047302246,
+      "learning_rate": 3.972479659814613e-06,
+      "loss": 0.2337,
+      "step": 519
+    },
+    {
+      "epoch": 0.0994263862332696,
+      "grad_norm": 3.7752184867858887,
+      "learning_rate": 3.973702765463161e-06,
+      "loss": 0.5257,
+      "step": 520
+    },
+    {
+      "epoch": 0.09961759082217973,
+      "grad_norm": 2.813133478164673,
+      "learning_rate": 3.974923521243501e-06,
+      "loss": 0.2109,
+      "step": 521
+    },
+    {
+      "epoch": 0.09980879541108986,
+      "grad_norm": 2.6638405323028564,
+      "learning_rate": 3.976141936167601e-06,
+      "loss": 0.2409,
+      "step": 522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6184728145599365,
+      "learning_rate": 3.9773580191956855e-06,
+      "loss": 0.4386,
+      "step": 523
+    },
+    {
+      "epoch": 0.10019120458891013,
+      "grad_norm": 2.8022634983062744,
+      "learning_rate": 3.978571779236627e-06,
+      "loss": 0.2676,
+      "step": 524
+    },
+    {
+      "epoch": 0.10038240917782026,
+      "grad_norm": 3.0874242782592773,
+      "learning_rate": 3.979783225148348e-06,
+      "loss": 0.1031,
+      "step": 525
+    },
+    {
+      "epoch": 0.10057361376673041,
+      "grad_norm": 2.6263234615325928,
+      "learning_rate": 3.980992365738197e-06,
+      "loss": 0.3226,
+      "step": 526
+    },
+    {
+      "epoch": 0.10076481835564054,
+      "grad_norm": 3.469144821166992,
+      "learning_rate": 3.982199209763339e-06,
+      "loss": 0.1735,
+      "step": 527
+    },
+    {
+      "epoch": 0.10095602294455067,
+      "grad_norm": 3.713228940963745,
+      "learning_rate": 3.983403765931139e-06,
+      "loss": 0.2317,
+      "step": 528
+    },
+    {
+      "epoch": 0.10114722753346081,
+      "grad_norm": 4.205535411834717,
+      "learning_rate": 3.98460604289953e-06,
+      "loss": 0.1452,
+      "step": 529
+    },
+    {
+      "epoch": 0.10133843212237094,
+      "grad_norm": 2.1234936714172363,
+      "learning_rate": 3.985806049277395e-06,
+      "loss": 0.1145,
+      "step": 530
+    },
+    {
+      "epoch": 0.10152963671128107,
+      "grad_norm": 2.30922794342041,
+      "learning_rate": 3.987003793624931e-06,
+      "loss": 0.144,
+      "step": 531
+    },
+    {
+      "epoch": 0.10172084130019121,
+      "grad_norm": 1.6173641681671143,
+      "learning_rate": 3.98819928445402e-06,
+      "loss": 0.3568,
+      "step": 532
+    },
+    {
+      "epoch": 0.10191204588910134,
+      "grad_norm": 3.590470552444458,
+      "learning_rate": 3.989392530228588e-06,
+      "loss": 0.4316,
+      "step": 533
+    },
+    {
+      "epoch": 0.10210325047801147,
+      "grad_norm": 3.209644079208374,
+      "learning_rate": 3.990583539364967e-06,
+      "loss": 0.4416,
+      "step": 534
+    },
+    {
+      "epoch": 0.1022944550669216,
+      "grad_norm": 3.220745801925659,
+      "learning_rate": 3.99177232023225e-06,
+      "loss": 0.1935,
+      "step": 535
+    },
+    {
+      "epoch": 0.10248565965583174,
+      "grad_norm": 1.88006591796875,
+      "learning_rate": 3.992958881152644e-06,
+      "loss": 0.2317,
+      "step": 536
+    },
+    {
+      "epoch": 0.10267686424474187,
+      "grad_norm": 1.5669440031051636,
+      "learning_rate": 3.9941432304018205e-06,
+      "loss": 0.0725,
+      "step": 537
+    },
+    {
+      "epoch": 0.102868068833652,
+      "grad_norm": 2.0758368968963623,
+      "learning_rate": 3.995325376209261e-06,
+      "loss": 0.2837,
+      "step": 538
+    },
+    {
+      "epoch": 0.10305927342256214,
+      "grad_norm": 3.3589015007019043,
+      "learning_rate": 3.996505326758601e-06,
+      "loss": 0.5328,
+      "step": 539
+    },
+    {
+      "epoch": 0.10325047801147227,
+      "grad_norm": 2.2015810012817383,
+      "learning_rate": 3.997683090187967e-06,
+      "loss": 0.3724,
+      "step": 540
+    },
+    {
+      "epoch": 0.1034416826003824,
+      "grad_norm": 3.714031934738159,
+      "learning_rate": 3.998858674590317e-06,
+      "loss": 0.2856,
+      "step": 541
+    },
+    {
+      "epoch": 0.10363288718929255,
+      "grad_norm": 2.511425018310547,
+      "learning_rate": 4.000032088013775e-06,
+      "loss": 0.2015,
+      "step": 542
+    },
+    {
+      "epoch": 0.10382409177820268,
+      "grad_norm": 2.7374191284179688,
+      "learning_rate": 4.001203338461952e-06,
+      "loss": 0.1779,
+      "step": 543
+    },
+    {
+      "epoch": 0.1040152963671128,
+      "grad_norm": 3.1923158168792725,
+      "learning_rate": 4.002372433894288e-06,
+      "loss": 0.2595,
+      "step": 544
+    },
+    {
+      "epoch": 0.10420650095602295,
+      "grad_norm": 2.5342373847961426,
+      "learning_rate": 4.0035393822263685e-06,
+      "loss": 0.3994,
+      "step": 545
+    },
+    {
+      "epoch": 0.10439770554493308,
+      "grad_norm": 2.794058084487915,
+      "learning_rate": 4.0047041913302425e-06,
+      "loss": 0.5341,
+      "step": 546
+    },
+    {
+      "epoch": 0.10458891013384321,
+      "grad_norm": 2.975745439529419,
+      "learning_rate": 4.005866869034748e-06,
+      "loss": 0.3757,
+      "step": 547
+    },
+    {
+      "epoch": 0.10478011472275335,
+      "grad_norm": 1.5141457319259644,
+      "learning_rate": 4.0070274231258276e-06,
+      "loss": 0.2261,
+      "step": 548
+    },
+    {
+      "epoch": 0.10497131931166348,
+      "grad_norm": 1.4179555177688599,
+      "learning_rate": 4.008185861346832e-06,
+      "loss": 0.1121,
+      "step": 549
+    },
+    {
+      "epoch": 0.10516252390057361,
+      "grad_norm": 1.3781744241714478,
+      "learning_rate": 4.009342191398848e-06,
+      "loss": 0.0929,
+      "step": 550
+    },
+    {
+      "epoch": 0.10535372848948375,
+      "grad_norm": 3.2945396900177,
+      "learning_rate": 4.010496420940983e-06,
+      "loss": 0.5864,
+      "step": 551
+    },
+    {
+      "epoch": 0.10554493307839388,
+      "grad_norm": 2.5260989665985107,
+      "learning_rate": 4.0116485575906874e-06,
+      "loss": 0.3214,
+      "step": 552
+    },
+    {
+      "epoch": 0.10573613766730401,
+      "grad_norm": 2.2214319705963135,
+      "learning_rate": 4.01279860892405e-06,
+      "loss": 0.3037,
+      "step": 553
+    },
+    {
+      "epoch": 0.10592734225621415,
+      "grad_norm": 2.315720796585083,
+      "learning_rate": 4.013946582476095e-06,
+      "loss": 0.3353,
+      "step": 554
+    },
+    {
+      "epoch": 0.10611854684512428,
+      "grad_norm": 2.7136638164520264,
+      "learning_rate": 4.015092485741078e-06,
+      "loss": 0.2516,
+      "step": 555
+    },
+    {
+      "epoch": 0.10630975143403441,
+      "grad_norm": 2.8678059577941895,
+      "learning_rate": 4.0162363261727876e-06,
+      "loss": 0.1223,
+      "step": 556
+    },
+    {
+      "epoch": 0.10650095602294456,
+      "grad_norm": 3.997982978820801,
+      "learning_rate": 4.017378111184824e-06,
+      "loss": 0.2095,
+      "step": 557
+    },
+    {
+      "epoch": 0.10669216061185469,
+      "grad_norm": 2.3306539058685303,
+      "learning_rate": 4.018517848150896e-06,
+      "loss": 0.2294,
+      "step": 558
+    },
+    {
+      "epoch": 0.10688336520076482,
+      "grad_norm": 1.9543094635009766,
+      "learning_rate": 4.019655544405104e-06,
+      "loss": 0.1078,
+      "step": 559
+    },
+    {
+      "epoch": 0.10707456978967496,
+      "grad_norm": 2.487516403198242,
+      "learning_rate": 4.0207912072422265e-06,
+      "loss": 0.2011,
+      "step": 560
+    },
+    {
+      "epoch": 0.10726577437858509,
+      "grad_norm": 1.3109228610992432,
+      "learning_rate": 4.0219248439179914e-06,
+      "loss": 0.1198,
+      "step": 561
+    },
+    {
+      "epoch": 0.10745697896749522,
+      "grad_norm": 1.7776157855987549,
+      "learning_rate": 4.023056461649366e-06,
+      "loss": 0.0965,
+      "step": 562
+    },
+    {
+      "epoch": 0.10764818355640535,
+      "grad_norm": 2.4455111026763916,
+      "learning_rate": 4.024186067614824e-06,
+      "loss": 0.1997,
+      "step": 563
+    },
+    {
+      "epoch": 0.10783938814531549,
+      "grad_norm": 2.7624261379241943,
+      "learning_rate": 4.0253136689546225e-06,
+      "loss": 0.6754,
+      "step": 564
+    },
+    {
+      "epoch": 0.10803059273422562,
+      "grad_norm": 1.7886794805526733,
+      "learning_rate": 4.026439272771073e-06,
+      "loss": 0.292,
+      "step": 565
+    },
+    {
+      "epoch": 0.10822179732313575,
+      "grad_norm": 1.396551489830017,
+      "learning_rate": 4.027562886128808e-06,
+      "loss": 0.095,
+      "step": 566
+    },
+    {
+      "epoch": 0.10841300191204589,
+      "grad_norm": 2.8856143951416016,
+      "learning_rate": 4.028684516055048e-06,
+      "loss": 0.2296,
+      "step": 567
+    },
+    {
+      "epoch": 0.10860420650095602,
+      "grad_norm": 2.7726597785949707,
+      "learning_rate": 4.029804169539866e-06,
+      "loss": 0.1276,
+      "step": 568
+    },
+    {
+      "epoch": 0.10879541108986615,
+      "grad_norm": 2.4296844005584717,
+      "learning_rate": 4.030921853536447e-06,
+      "loss": 0.2504,
+      "step": 569
+    },
+    {
+      "epoch": 0.1089866156787763,
+      "grad_norm": 2.3982651233673096,
+      "learning_rate": 4.032037574961348e-06,
+      "loss": 0.487,
+      "step": 570
+    },
+    {
+      "epoch": 0.10917782026768642,
+      "grad_norm": 2.2958669662475586,
+      "learning_rate": 4.033151340694757e-06,
+      "loss": 0.352,
+      "step": 571
+    },
+    {
+      "epoch": 0.10936902485659655,
+      "grad_norm": 2.792046546936035,
+      "learning_rate": 4.0342631575807424e-06,
+      "loss": 0.3026,
+      "step": 572
+    },
+    {
+      "epoch": 0.1095602294455067,
+      "grad_norm": 2.0497055053710938,
+      "learning_rate": 4.035373032427511e-06,
+      "loss": 0.207,
+      "step": 573
+    },
+    {
+      "epoch": 0.10975143403441683,
+      "grad_norm": 1.5417126417160034,
+      "learning_rate": 4.036480972007654e-06,
+      "loss": 0.0919,
+      "step": 574
+    },
+    {
+      "epoch": 0.10994263862332695,
+      "grad_norm": 2.187443494796753,
+      "learning_rate": 4.037586983058396e-06,
+      "loss": 0.134,
+      "step": 575
+    },
+    {
+      "epoch": 0.1101338432122371,
+      "grad_norm": 3.1193389892578125,
+      "learning_rate": 4.038691072281846e-06,
+      "loss": 0.5783,
+      "step": 576
+    },
+    {
+      "epoch": 0.11032504780114723,
+      "grad_norm": 2.0621426105499268,
+      "learning_rate": 4.0397932463452296e-06,
+      "loss": 0.3393,
+      "step": 577
+    },
+    {
+      "epoch": 0.11051625239005736,
+      "grad_norm": 3.7791013717651367,
+      "learning_rate": 4.040893511881142e-06,
+      "loss": 0.469,
+      "step": 578
+    },
+    {
+      "epoch": 0.1107074569789675,
+      "grad_norm": 1.6285041570663452,
+      "learning_rate": 4.0419918754877816e-06,
+      "loss": 0.0855,
+      "step": 579
+    },
+    {
+      "epoch": 0.11089866156787763,
+      "grad_norm": 2.735530138015747,
+      "learning_rate": 4.043088343729189e-06,
+      "loss": 0.2253,
+      "step": 580
+    },
+    {
+      "epoch": 0.11108986615678776,
+      "grad_norm": 3.3225202560424805,
+      "learning_rate": 4.0441829231354805e-06,
+      "loss": 0.2688,
+      "step": 581
+    },
+    {
+      "epoch": 0.1112810707456979,
+      "grad_norm": 2.049799680709839,
+      "learning_rate": 4.0452756202030864e-06,
+      "loss": 0.2561,
+      "step": 582
+    },
+    {
+      "epoch": 0.11147227533460803,
+      "grad_norm": 2.2837252616882324,
+      "learning_rate": 4.046366441394976e-06,
+      "loss": 0.4603,
+      "step": 583
+    },
+    {
+      "epoch": 0.11166347992351816,
+      "grad_norm": 1.8498018980026245,
+      "learning_rate": 4.0474553931408905e-06,
+      "loss": 0.3059,
+      "step": 584
+    },
+    {
+      "epoch": 0.1118546845124283,
+      "grad_norm": 2.7567930221557617,
+      "learning_rate": 4.0485424818375704e-06,
+      "loss": 0.3465,
+      "step": 585
+    },
+    {
+      "epoch": 0.11204588910133843,
+      "grad_norm": 2.339862585067749,
+      "learning_rate": 4.04962771384898e-06,
+      "loss": 0.1157,
+      "step": 586
+    },
+    {
+      "epoch": 0.11223709369024856,
+      "grad_norm": 2.2230653762817383,
+      "learning_rate": 4.050711095506532e-06,
+      "loss": 0.1013,
+      "step": 587
+    },
+    {
+      "epoch": 0.1124282982791587,
+      "grad_norm": 4.637607574462891,
+      "learning_rate": 4.051792633109307e-06,
+      "loss": 0.3786,
+      "step": 588
+    },
+    {
+      "epoch": 0.11261950286806884,
+      "grad_norm": 2.5454115867614746,
+      "learning_rate": 4.052872332924277e-06,
+      "loss": 0.223,
+      "step": 589
+    },
+    {
+      "epoch": 0.11281070745697896,
+      "grad_norm": 2.093069076538086,
+      "learning_rate": 4.053950201186518e-06,
+      "loss": 0.1641,
+      "step": 590
+    },
+    {
+      "epoch": 0.11300191204588911,
+      "grad_norm": 2.5368142127990723,
+      "learning_rate": 4.055026244099433e-06,
+      "loss": 0.3767,
+      "step": 591
+    },
+    {
+      "epoch": 0.11319311663479924,
+      "grad_norm": 1.601399302482605,
+      "learning_rate": 4.056100467834957e-06,
+      "loss": 0.1006,
+      "step": 592
+    },
+    {
+      "epoch": 0.11338432122370937,
+      "grad_norm": 2.5372474193573,
+      "learning_rate": 4.05717287853378e-06,
+      "loss": 0.1315,
+      "step": 593
+    },
+    {
+      "epoch": 0.1135755258126195,
+      "grad_norm": 2.6447410583496094,
+      "learning_rate": 4.058243482305548e-06,
+      "loss": 0.1529,
+      "step": 594
+    },
+    {
+      "epoch": 0.11376673040152964,
+      "grad_norm": 3.427429676055908,
+      "learning_rate": 4.059312285229079e-06,
+      "loss": 0.6611,
+      "step": 595
+    },
+    {
+      "epoch": 0.11395793499043977,
+      "grad_norm": 2.3087265491485596,
+      "learning_rate": 4.060379293352566e-06,
+      "loss": 0.226,
+      "step": 596
+    },
+    {
+      "epoch": 0.1141491395793499,
+      "grad_norm": 2.3593530654907227,
+      "learning_rate": 4.061444512693784e-06,
+      "loss": 0.3704,
+      "step": 597
+    },
+    {
+      "epoch": 0.11434034416826004,
+      "grad_norm": 2.03262996673584,
+      "learning_rate": 4.062507949240291e-06,
+      "loss": 0.1638,
+      "step": 598
+    },
+    {
+      "epoch": 0.11453154875717017,
+      "grad_norm": 2.4928765296936035,
+      "learning_rate": 4.063569608949637e-06,
+      "loss": 0.1744,
+      "step": 599
+    },
+    {
+      "epoch": 0.1147227533460803,
+      "grad_norm": 6.897516250610352,
+      "learning_rate": 4.0646294977495545e-06,
+      "loss": 0.2404,
+      "step": 600
+    },
+    {
+      "epoch": 0.11491395793499044,
+      "grad_norm": 1.9926093816757202,
+      "learning_rate": 4.065687621538164e-06,
+      "loss": 0.2995,
+      "step": 601
+    },
+    {
+      "epoch": 0.11510516252390057,
+      "grad_norm": 3.1814188957214355,
+      "learning_rate": 4.066743986184169e-06,
+      "loss": 0.4707,
+      "step": 602
+    },
+    {
+      "epoch": 0.1152963671128107,
+      "grad_norm": 1.9630004167556763,
+      "learning_rate": 4.067798597527053e-06,
+      "loss": 0.2011,
+      "step": 603
+    },
+    {
+      "epoch": 0.11548757170172085,
+      "grad_norm": 3.1169941425323486,
+      "learning_rate": 4.068851461377267e-06,
+      "loss": 0.1188,
+      "step": 604
+    },
+    {
+      "epoch": 0.11567877629063097,
+      "grad_norm": 1.6685168743133545,
+      "learning_rate": 4.069902583516429e-06,
+      "loss": 0.0651,
+      "step": 605
+    },
+    {
+      "epoch": 0.1158699808795411,
+      "grad_norm": 3.466280937194824,
+      "learning_rate": 4.070951969697511e-06,
+      "loss": 0.1449,
+      "step": 606
+    },
+    {
+      "epoch": 0.11606118546845125,
+      "grad_norm": 2.8022851943969727,
+      "learning_rate": 4.071999625645027e-06,
+      "loss": 0.2323,
+      "step": 607
+    },
+    {
+      "epoch": 0.11625239005736138,
+      "grad_norm": 2.7856459617614746,
+      "learning_rate": 4.073045557055227e-06,
+      "loss": 0.4914,
+      "step": 608
+    },
+    {
+      "epoch": 0.1164435946462715,
+      "grad_norm": 2.7790656089782715,
+      "learning_rate": 4.07408976959627e-06,
+      "loss": 0.5342,
+      "step": 609
+    },
+    {
+      "epoch": 0.11663479923518165,
+      "grad_norm": 2.095447063446045,
+      "learning_rate": 4.075132268908421e-06,
+      "loss": 0.2162,
+      "step": 610
+    },
+    {
+      "epoch": 0.11682600382409178,
+      "grad_norm": 1.7271842956542969,
+      "learning_rate": 4.076173060604226e-06,
+      "loss": 0.1058,
+      "step": 611
+    },
+    {
+      "epoch": 0.11701720841300191,
+      "grad_norm": 3.3794445991516113,
+      "learning_rate": 4.077212150268698e-06,
+      "loss": 0.1215,
+      "step": 612
+    },
+    {
+      "epoch": 0.11720841300191205,
+      "grad_norm": 5.20923376083374,
+      "learning_rate": 4.078249543459495e-06,
+      "loss": 0.2695,
+      "step": 613
+    },
+    {
+      "epoch": 0.11739961759082218,
+      "grad_norm": 3.670212984085083,
+      "learning_rate": 4.079285245707096e-06,
+      "loss": 0.9118,
+      "step": 614
+    },
+    {
+      "epoch": 0.11759082217973231,
+      "grad_norm": 4.551514148712158,
+      "learning_rate": 4.080319262514981e-06,
+      "loss": 0.407,
+      "step": 615
+    },
+    {
+      "epoch": 0.11778202676864245,
+      "grad_norm": 2.7722604274749756,
+      "learning_rate": 4.081351599359807e-06,
+      "loss": 0.3355,
+      "step": 616
+    },
+    {
+      "epoch": 0.11797323135755258,
+      "grad_norm": 2.6713385581970215,
+      "learning_rate": 4.0823822616915795e-06,
+      "loss": 0.1726,
+      "step": 617
+    },
+    {
+      "epoch": 0.11816443594646271,
+      "grad_norm": 2.3525755405426025,
+      "learning_rate": 4.083411254933824e-06,
+      "loss": 0.1748,
+      "step": 618
+    },
+    {
+      "epoch": 0.11835564053537286,
+      "grad_norm": 2.2042713165283203,
+      "learning_rate": 4.084438584483764e-06,
+      "loss": 0.2662,
+      "step": 619
+    },
+    {
+      "epoch": 0.11854684512428298,
+      "grad_norm": 2.0608386993408203,
+      "learning_rate": 4.085464255712483e-06,
+      "loss": 0.2041,
+      "step": 620
+    },
+    {
+      "epoch": 0.11873804971319311,
+      "grad_norm": 2.0503997802734375,
+      "learning_rate": 4.0864882739650965e-06,
+      "loss": 0.2242,
+      "step": 621
+    },
+    {
+      "epoch": 0.11892925430210326,
+      "grad_norm": 1.8372803926467896,
+      "learning_rate": 4.087510644560921e-06,
+      "loss": 0.1795,
+      "step": 622
+    },
+    {
+      "epoch": 0.11912045889101339,
+      "grad_norm": 1.664029598236084,
+      "learning_rate": 4.088531372793636e-06,
+      "loss": 0.1379,
+      "step": 623
+    },
+    {
+      "epoch": 0.11931166347992352,
+      "grad_norm": 1.8293601274490356,
+      "learning_rate": 4.089550463931449e-06,
+      "loss": 0.1597,
+      "step": 624
+    },
+    {
+      "epoch": 0.11950286806883365,
+      "grad_norm": 3.36576771736145,
+      "learning_rate": 4.090567923217263e-06,
+      "loss": 0.1534,
+      "step": 625
+    },
+    {
+      "epoch": 0.11969407265774379,
+      "grad_norm": 2.3628504276275635,
+      "learning_rate": 4.091583755868833e-06,
+      "loss": 0.348,
+      "step": 626
+    },
+    {
+      "epoch": 0.11988527724665392,
+      "grad_norm": 2.525311231613159,
+      "learning_rate": 4.0925979670789294e-06,
+      "loss": 0.4548,
+      "step": 627
+    },
+    {
+      "epoch": 0.12007648183556405,
+      "grad_norm": 2.786006450653076,
+      "learning_rate": 4.093610562015496e-06,
+      "loss": 0.3034,
+      "step": 628
+    },
+    {
+      "epoch": 0.12026768642447419,
+      "grad_norm": 2.4381093978881836,
+      "learning_rate": 4.09462154582181e-06,
+      "loss": 0.3277,
+      "step": 629
+    },
+    {
+      "epoch": 0.12045889101338432,
+      "grad_norm": 1.9374213218688965,
+      "learning_rate": 4.095630923616636e-06,
+      "loss": 0.1188,
+      "step": 630
+    },
+    {
+      "epoch": 0.12065009560229445,
+      "grad_norm": 3.029554843902588,
+      "learning_rate": 4.096638700494381e-06,
+      "loss": 0.1719,
+      "step": 631
+    },
+    {
+      "epoch": 0.12084130019120459,
+      "grad_norm": 2.0337328910827637,
+      "learning_rate": 4.097644881525257e-06,
+      "loss": 0.3093,
+      "step": 632
+    },
+    {
+      "epoch": 0.12103250478011472,
+      "grad_norm": 2.4461922645568848,
+      "learning_rate": 4.098649471755419e-06,
+      "loss": 0.3777,
+      "step": 633
+    },
+    {
+      "epoch": 0.12122370936902485,
+      "grad_norm": 2.149134635925293,
+      "learning_rate": 4.099652476207133e-06,
+      "loss": 0.1888,
+      "step": 634
+    },
+    {
+      "epoch": 0.121414913957935,
+      "grad_norm": 3.174314022064209,
+      "learning_rate": 4.1006538998789146e-06,
+      "loss": 0.4976,
+      "step": 635
+    },
+    {
+      "epoch": 0.12160611854684512,
+      "grad_norm": 2.157924175262451,
+      "learning_rate": 4.101653747745683e-06,
+      "loss": 0.2177,
+      "step": 636
+    },
+    {
+      "epoch": 0.12179732313575525,
+      "grad_norm": 9.099021911621094,
+      "learning_rate": 4.102652024758911e-06,
+      "loss": 0.3181,
+      "step": 637
+    },
+    {
+      "epoch": 0.1219885277246654,
+      "grad_norm": 2.279358386993408,
+      "learning_rate": 4.10364873584677e-06,
+      "loss": 0.1611,
+      "step": 638
+    },
+    {
+      "epoch": 0.12217973231357553,
+      "grad_norm": 6.471635818481445,
+      "learning_rate": 4.104643885914275e-06,
+      "loss": 0.2495,
+      "step": 639
+    },
+    {
+      "epoch": 0.12237093690248566,
+      "grad_norm": 1.9951330423355103,
+      "learning_rate": 4.1056374798434325e-06,
+      "loss": 0.1462,
+      "step": 640
+    },
+    {
+      "epoch": 0.1225621414913958,
+      "grad_norm": 2.0457870960235596,
+      "learning_rate": 4.106629522493381e-06,
+      "loss": 0.2578,
+      "step": 641
+    },
+    {
+      "epoch": 0.12275334608030593,
+      "grad_norm": 2.4925954341888428,
+      "learning_rate": 4.107620018700538e-06,
+      "loss": 0.2238,
+      "step": 642
+    },
+    {
+      "epoch": 0.12294455066921606,
+      "grad_norm": 3.2228965759277344,
+      "learning_rate": 4.1086089732787325e-06,
+      "loss": 0.1979,
+      "step": 643
+    },
+    {
+      "epoch": 0.1231357552581262,
+      "grad_norm": 1.5814366340637207,
+      "learning_rate": 4.109596391019356e-06,
+      "loss": 0.097,
+      "step": 644
+    },
+    {
+      "epoch": 0.12332695984703633,
+      "grad_norm": 3.6417267322540283,
+      "learning_rate": 4.110582276691498e-06,
+      "loss": 0.354,
+      "step": 645
+    },
+    {
+      "epoch": 0.12351816443594646,
+      "grad_norm": 2.278165102005005,
+      "learning_rate": 4.11156663504208e-06,
+      "loss": 0.1463,
+      "step": 646
+    },
+    {
+      "epoch": 0.1237093690248566,
+      "grad_norm": 1.5671286582946777,
+      "learning_rate": 4.112549470795998e-06,
+      "loss": 0.1445,
+      "step": 647
+    },
+    {
+      "epoch": 0.12390057361376673,
+      "grad_norm": 3.1425745487213135,
+      "learning_rate": 4.113530788656255e-06,
+      "loss": 0.2828,
+      "step": 648
+    },
+    {
+      "epoch": 0.12409177820267686,
+      "grad_norm": 3.6428639888763428,
+      "learning_rate": 4.1145105933041e-06,
+      "loss": 0.1176,
+      "step": 649
+    },
+    {
+      "epoch": 0.124282982791587,
+      "grad_norm": 1.9820659160614014,
+      "learning_rate": 4.115488889399157e-06,
+      "loss": 0.1025,
+      "step": 650
+    },
+    {
+      "epoch": 0.12447418738049713,
+      "grad_norm": 2.4328808784484863,
+      "learning_rate": 4.116465681579564e-06,
+      "loss": 0.3629,
+      "step": 651
+    },
+    {
+      "epoch": 0.12466539196940726,
+      "grad_norm": 2.9326987266540527,
+      "learning_rate": 4.117440974462098e-06,
+      "loss": 0.1771,
+      "step": 652
+    },
+    {
+      "epoch": 0.1248565965583174,
+      "grad_norm": 2.000542640686035,
+      "learning_rate": 4.11841477264231e-06,
+      "loss": 0.1175,
+      "step": 653
+    },
+    {
+      "epoch": 0.12504780114722752,
+      "grad_norm": 2.4583964347839355,
+      "learning_rate": 4.119387080694656e-06,
+      "loss": 0.2045,
+      "step": 654
+    },
+    {
+      "epoch": 0.12523900573613767,
+      "grad_norm": 2.519341230392456,
+      "learning_rate": 4.1203579031726235e-06,
+      "loss": 0.2256,
+      "step": 655
+    },
+    {
+      "epoch": 0.1254302103250478,
+      "grad_norm": 1.732568621635437,
+      "learning_rate": 4.12132724460886e-06,
+      "loss": 0.0891,
+      "step": 656
+    },
+    {
+      "epoch": 0.12562141491395792,
+      "grad_norm": 3.676959276199341,
+      "learning_rate": 4.122295109515299e-06,
+      "loss": 0.5975,
+      "step": 657
+    },
+    {
+      "epoch": 0.12581261950286807,
+      "grad_norm": 3.2852377891540527,
+      "learning_rate": 4.123261502383291e-06,
+      "loss": 0.6327,
+      "step": 658
+    },
+    {
+      "epoch": 0.1260038240917782,
+      "grad_norm": 2.5524721145629883,
+      "learning_rate": 4.12422642768372e-06,
+      "loss": 0.2364,
+      "step": 659
+    },
+    {
+      "epoch": 0.12619502868068833,
+      "grad_norm": 2.0922858715057373,
+      "learning_rate": 4.125189889867135e-06,
+      "loss": 0.147,
+      "step": 660
+    },
+    {
+      "epoch": 0.12638623326959847,
+      "grad_norm": 2.2585864067077637,
+      "learning_rate": 4.126151893363871e-06,
+      "loss": 0.1199,
+      "step": 661
+    },
+    {
+      "epoch": 0.1265774378585086,
+      "grad_norm": 1.344876766204834,
+      "learning_rate": 4.12711244258417e-06,
+      "loss": 0.0836,
+      "step": 662
+    },
+    {
+      "epoch": 0.12676864244741873,
+      "grad_norm": 4.379039764404297,
+      "learning_rate": 4.128071541918302e-06,
+      "loss": 0.2658,
+      "step": 663
+    },
+    {
+      "epoch": 0.12695984703632887,
+      "grad_norm": 2.2609031200408936,
+      "learning_rate": 4.129029195736687e-06,
+      "loss": 0.3937,
+      "step": 664
+    },
+    {
+      "epoch": 0.12715105162523901,
+      "grad_norm": 2.849423408508301,
+      "learning_rate": 4.129985408390017e-06,
+      "loss": 0.1937,
+      "step": 665
+    },
+    {
+      "epoch": 0.12734225621414913,
+      "grad_norm": 3.229069471359253,
+      "learning_rate": 4.130940184209367e-06,
+      "loss": 0.4889,
+      "step": 666
+    },
+    {
+      "epoch": 0.12753346080305927,
+      "grad_norm": 3.168386459350586,
+      "learning_rate": 4.131893527506318e-06,
+      "loss": 0.3612,
+      "step": 667
+    },
+    {
+      "epoch": 0.12772466539196942,
+      "grad_norm": 1.0114667415618896,
+      "learning_rate": 4.132845442573078e-06,
+      "loss": 0.0365,
+      "step": 668
+    },
+    {
+      "epoch": 0.12791586998087953,
+      "grad_norm": 1.751471757888794,
+      "learning_rate": 4.133795933682587e-06,
+      "loss": 0.097,
+      "step": 669
+    },
+    {
+      "epoch": 0.12810707456978968,
+      "grad_norm": 7.675336837768555,
+      "learning_rate": 4.13474500508864e-06,
+      "loss": 0.6252,
+      "step": 670
+    },
+    {
+      "epoch": 0.12829827915869982,
+      "grad_norm": 2.195767402648926,
+      "learning_rate": 4.135692661026002e-06,
+      "loss": 0.2651,
+      "step": 671
+    },
+    {
+      "epoch": 0.12848948374760993,
+      "grad_norm": 2.2194595336914062,
+      "learning_rate": 4.136638905710514e-06,
+      "loss": 0.1452,
+      "step": 672
+    },
+    {
+      "epoch": 0.12868068833652008,
+      "grad_norm": 2.8059849739074707,
+      "learning_rate": 4.137583743339215e-06,
+      "loss": 0.1959,
+      "step": 673
+    },
+    {
+      "epoch": 0.12887189292543022,
+      "grad_norm": 2.4581644535064697,
+      "learning_rate": 4.138527178090445e-06,
+      "loss": 0.1656,
+      "step": 674
+    },
+    {
+      "epoch": 0.12906309751434034,
+      "grad_norm": 2.713122606277466,
+      "learning_rate": 4.1394692141239635e-06,
+      "loss": 0.172,
+      "step": 675
+    },
+    {
+      "epoch": 0.12925430210325048,
+      "grad_norm": 3.6402595043182373,
+      "learning_rate": 4.140409855581052e-06,
+      "loss": 0.5809,
+      "step": 676
+    },
+    {
+      "epoch": 0.12944550669216062,
+      "grad_norm": 3.1227595806121826,
+      "learning_rate": 4.14134910658463e-06,
+      "loss": 0.3983,
+      "step": 677
+    },
+    {
+      "epoch": 0.12963671128107074,
+      "grad_norm": 2.864481210708618,
+      "learning_rate": 4.14228697123936e-06,
+      "loss": 0.3984,
+      "step": 678
+    },
+    {
+      "epoch": 0.12982791586998088,
+      "grad_norm": 1.6978756189346313,
+      "learning_rate": 4.143223453631755e-06,
+      "loss": 0.074,
+      "step": 679
+    },
+    {
+      "epoch": 0.13001912045889102,
+      "grad_norm": 1.8725084066390991,
+      "learning_rate": 4.144158557830285e-06,
+      "loss": 0.1087,
+      "step": 680
+    },
+    {
+      "epoch": 0.13021032504780114,
+      "grad_norm": 3.0994725227355957,
+      "learning_rate": 4.1450922878854865e-06,
+      "loss": 0.3181,
+      "step": 681
+    },
+    {
+      "epoch": 0.13040152963671128,
+      "grad_norm": 2.7692768573760986,
+      "learning_rate": 4.146024647830064e-06,
+      "loss": 0.3154,
+      "step": 682
+    },
+    {
+      "epoch": 0.13059273422562143,
+      "grad_norm": 2.107736825942993,
+      "learning_rate": 4.146955641678995e-06,
+      "loss": 0.3758,
+      "step": 683
+    },
+    {
+      "epoch": 0.13078393881453154,
+      "grad_norm": 1.6708070039749146,
+      "learning_rate": 4.147885273429636e-06,
+      "loss": 0.1446,
+      "step": 684
+    },
+    {
+      "epoch": 0.13097514340344169,
+      "grad_norm": 2.343808174133301,
+      "learning_rate": 4.148813547061823e-06,
+      "loss": 0.2898,
+      "step": 685
+    },
+    {
+      "epoch": 0.13116634799235183,
+      "grad_norm": 3.282604932785034,
+      "learning_rate": 4.1497404665379755e-06,
+      "loss": 0.3803,
+      "step": 686
+    },
+    {
+      "epoch": 0.13135755258126194,
+      "grad_norm": 2.8634464740753174,
+      "learning_rate": 4.150666035803198e-06,
+      "loss": 0.1605,
+      "step": 687
+    },
+    {
+      "epoch": 0.1315487571701721,
+      "grad_norm": 2.119781255722046,
+      "learning_rate": 4.151590258785376e-06,
+      "loss": 0.1974,
+      "step": 688
+    },
+    {
+      "epoch": 0.13173996175908223,
+      "grad_norm": 2.3915913105010986,
+      "learning_rate": 4.1525131393952865e-06,
+      "loss": 0.2118,
+      "step": 689
+    },
+    {
+      "epoch": 0.13193116634799235,
+      "grad_norm": 2.3511242866516113,
+      "learning_rate": 4.153434681526684e-06,
+      "loss": 0.3729,
+      "step": 690
+    },
+    {
+      "epoch": 0.1321223709369025,
+      "grad_norm": 2.7544589042663574,
+      "learning_rate": 4.154354889056412e-06,
+      "loss": 0.3998,
+      "step": 691
+    },
+    {
+      "epoch": 0.13231357552581263,
+      "grad_norm": 2.3922064304351807,
+      "learning_rate": 4.155273765844489e-06,
+      "loss": 0.1746,
+      "step": 692
+    },
+    {
+      "epoch": 0.13250478011472275,
+      "grad_norm": 2.2704615592956543,
+      "learning_rate": 4.156191315734217e-06,
+      "loss": 0.1901,
+      "step": 693
+    },
+    {
+      "epoch": 0.1326959847036329,
+      "grad_norm": 1.6022690534591675,
+      "learning_rate": 4.157107542552267e-06,
+      "loss": 0.1309,
+      "step": 694
+    },
+    {
+      "epoch": 0.13288718929254303,
+      "grad_norm": 2.8731253147125244,
+      "learning_rate": 4.158022450108784e-06,
+      "loss": 0.3855,
+      "step": 695
+    },
+    {
+      "epoch": 0.13307839388145315,
+      "grad_norm": 2.325465679168701,
+      "learning_rate": 4.158936042197477e-06,
+      "loss": 0.1592,
+      "step": 696
+    },
+    {
+      "epoch": 0.1332695984703633,
+      "grad_norm": 3.571281909942627,
+      "learning_rate": 4.1598483225957125e-06,
+      "loss": 0.5375,
+      "step": 697
+    },
+    {
+      "epoch": 0.13346080305927344,
+      "grad_norm": 2.267007350921631,
+      "learning_rate": 4.1607592950646145e-06,
+      "loss": 0.1196,
+      "step": 698
+    },
+    {
+      "epoch": 0.13365200764818355,
+      "grad_norm": 2.3478941917419434,
+      "learning_rate": 4.161668963349148e-06,
+      "loss": 0.1641,
+      "step": 699
+    },
+    {
+      "epoch": 0.1338432122370937,
+      "grad_norm": 2.05322265625,
+      "learning_rate": 4.162577331178222e-06,
+      "loss": 0.131,
+      "step": 700
+    },
+    {
+      "epoch": 0.1340344168260038,
+      "grad_norm": 4.005086898803711,
+      "learning_rate": 4.163484402264773e-06,
+      "loss": 0.4016,
+      "step": 701
+    },
+    {
+      "epoch": 0.13422562141491395,
+      "grad_norm": 3.990110397338867,
+      "learning_rate": 4.1643901803058575e-06,
+      "loss": 0.4845,
+      "step": 702
+    },
+    {
+      "epoch": 0.1344168260038241,
+      "grad_norm": 2.7093546390533447,
+      "learning_rate": 4.165294668982747e-06,
+      "loss": 0.3717,
+      "step": 703
+    },
+    {
+      "epoch": 0.1346080305927342,
+      "grad_norm": 2.439100980758667,
+      "learning_rate": 4.1661978719610135e-06,
+      "loss": 0.3189,
+      "step": 704
+    },
+    {
+      "epoch": 0.13479923518164436,
+      "grad_norm": 2.0016961097717285,
+      "learning_rate": 4.167099792890619e-06,
+      "loss": 0.1995,
+      "step": 705
+    },
+    {
+      "epoch": 0.1349904397705545,
+      "grad_norm": 2.2571163177490234,
+      "learning_rate": 4.168000435406005e-06,
+      "loss": 0.1032,
+      "step": 706
+    },
+    {
+      "epoch": 0.13518164435946461,
+      "grad_norm": 2.6907873153686523,
+      "learning_rate": 4.168899803126179e-06,
+      "loss": 0.3268,
+      "step": 707
+    },
+    {
+      "epoch": 0.13537284894837476,
+      "grad_norm": 2.1795923709869385,
+      "learning_rate": 4.169797899654807e-06,
+      "loss": 0.2979,
+      "step": 708
+    },
+    {
+      "epoch": 0.1355640535372849,
+      "grad_norm": 1.801849126815796,
+      "learning_rate": 4.1706947285802905e-06,
+      "loss": 0.2202,
+      "step": 709
+    },
+    {
+      "epoch": 0.13575525812619502,
+      "grad_norm": 3.593956232070923,
+      "learning_rate": 4.1715902934758625e-06,
+      "loss": 0.4659,
+      "step": 710
+    },
+    {
+      "epoch": 0.13594646271510516,
+      "grad_norm": 2.4480674266815186,
+      "learning_rate": 4.172484597899666e-06,
+      "loss": 0.1247,
+      "step": 711
+    },
+    {
+      "epoch": 0.1361376673040153,
+      "grad_norm": 2.3875763416290283,
+      "learning_rate": 4.1733776453948425e-06,
+      "loss": 0.064,
+      "step": 712
+    },
+    {
+      "epoch": 0.13632887189292542,
+      "grad_norm": 2.754842758178711,
+      "learning_rate": 4.174269439489614e-06,
+      "loss": 0.1734,
+      "step": 713
+    },
+    {
+      "epoch": 0.13652007648183556,
+      "grad_norm": 3.3177061080932617,
+      "learning_rate": 4.175159983697367e-06,
+      "loss": 0.452,
+      "step": 714
+    },
+    {
+      "epoch": 0.1367112810707457,
+      "grad_norm": 1.6432844400405884,
+      "learning_rate": 4.176049281516738e-06,
+      "loss": 0.1835,
+      "step": 715
+    },
+    {
+      "epoch": 0.13690248565965582,
+      "grad_norm": 3.1820502281188965,
+      "learning_rate": 4.176937336431696e-06,
+      "loss": 0.3835,
+      "step": 716
+    },
+    {
+      "epoch": 0.13709369024856596,
+      "grad_norm": 2.664802312850952,
+      "learning_rate": 4.177824151911616e-06,
+      "loss": 0.1186,
+      "step": 717
+    },
+    {
+      "epoch": 0.1372848948374761,
+      "grad_norm": 2.4787399768829346,
+      "learning_rate": 4.178709731411373e-06,
+      "loss": 0.3117,
+      "step": 718
+    },
+    {
+      "epoch": 0.13747609942638622,
+      "grad_norm": 2.4012043476104736,
+      "learning_rate": 4.179594078371414e-06,
+      "loss": 0.0682,
+      "step": 719
+    },
+    {
+      "epoch": 0.13766730401529637,
+      "grad_norm": 2.465858221054077,
+      "learning_rate": 4.180477196217842e-06,
+      "loss": 0.335,
+      "step": 720
+    },
+    {
+      "epoch": 0.1378585086042065,
+      "grad_norm": 2.301172971725464,
+      "learning_rate": 4.181359088362493e-06,
+      "loss": 0.3335,
+      "step": 721
+    },
+    {
+      "epoch": 0.13804971319311662,
+      "grad_norm": 1.929276466369629,
+      "learning_rate": 4.182239758203017e-06,
+      "loss": 0.0915,
+      "step": 722
+    },
+    {
+      "epoch": 0.13824091778202677,
+      "grad_norm": 2.7580771446228027,
+      "learning_rate": 4.183119209122958e-06,
+      "loss": 0.3561,
+      "step": 723
+    },
+    {
+      "epoch": 0.1384321223709369,
+      "grad_norm": 1.6526697874069214,
+      "learning_rate": 4.183997444491827e-06,
+      "loss": 0.2137,
+      "step": 724
+    },
+    {
+      "epoch": 0.13862332695984703,
+      "grad_norm": 2.3815810680389404,
+      "learning_rate": 4.184874467665185e-06,
+      "loss": 0.1152,
+      "step": 725
+    },
+    {
+      "epoch": 0.13881453154875717,
+      "grad_norm": 2.2507095336914062,
+      "learning_rate": 4.185750281984717e-06,
+      "loss": 0.3442,
+      "step": 726
+    },
+    {
+      "epoch": 0.1390057361376673,
+      "grad_norm": 3.3361330032348633,
+      "learning_rate": 4.1866248907783065e-06,
+      "loss": 0.2196,
+      "step": 727
+    },
+    {
+      "epoch": 0.13919694072657743,
+      "grad_norm": 2.5841755867004395,
+      "learning_rate": 4.187498297360117e-06,
+      "loss": 0.2489,
+      "step": 728
+    },
+    {
+      "epoch": 0.13938814531548757,
+      "grad_norm": 2.5357816219329834,
+      "learning_rate": 4.188370505030664e-06,
+      "loss": 0.1614,
+      "step": 729
+    },
+    {
+      "epoch": 0.13957934990439771,
+      "grad_norm": 2.3405823707580566,
+      "learning_rate": 4.189241517076887e-06,
+      "loss": 0.1311,
+      "step": 730
+    },
+    {
+      "epoch": 0.13977055449330783,
+      "grad_norm": 2.4516336917877197,
+      "learning_rate": 4.190111336772229e-06,
+      "loss": 0.1509,
+      "step": 731
+    },
+    {
+      "epoch": 0.13996175908221797,
+      "grad_norm": 3.084892511367798,
+      "learning_rate": 4.190979967376708e-06,
+      "loss": 0.2953,
+      "step": 732
+    },
+    {
+      "epoch": 0.14015296367112812,
+      "grad_norm": 2.2274246215820312,
+      "learning_rate": 4.19184741213699e-06,
+      "loss": 0.4075,
+      "step": 733
+    },
+    {
+      "epoch": 0.14034416826003823,
+      "grad_norm": 3.0207102298736572,
+      "learning_rate": 4.192713674286461e-06,
+      "loss": 0.2865,
+      "step": 734
+    },
+    {
+      "epoch": 0.14053537284894838,
+      "grad_norm": 1.836266040802002,
+      "learning_rate": 4.193578757045304e-06,
+      "loss": 0.1552,
+      "step": 735
+    },
+    {
+      "epoch": 0.14072657743785852,
+      "grad_norm": 3.116115093231201,
+      "learning_rate": 4.194442663620563e-06,
+      "loss": 0.2001,
+      "step": 736
+    },
+    {
+      "epoch": 0.14091778202676863,
+      "grad_norm": 3.3889942169189453,
+      "learning_rate": 4.1953053972062215e-06,
+      "loss": 0.1891,
+      "step": 737
+    },
+    {
+      "epoch": 0.14110898661567878,
+      "grad_norm": 2.072631359100342,
+      "learning_rate": 4.196166960983269e-06,
+      "loss": 0.1645,
+      "step": 738
+    },
+    {
+      "epoch": 0.14130019120458892,
+      "grad_norm": 3.678781032562256,
+      "learning_rate": 4.197027358119775e-06,
+      "loss": 0.5332,
+      "step": 739
+    },
+    {
+      "epoch": 0.14149139579349904,
+      "grad_norm": 2.225128650665283,
+      "learning_rate": 4.1978865917709535e-06,
+      "loss": 0.1687,
+      "step": 740
+    },
+    {
+      "epoch": 0.14168260038240918,
+      "grad_norm": 5.0827131271362305,
+      "learning_rate": 4.198744665079239e-06,
+      "loss": 0.4454,
+      "step": 741
+    },
+    {
+      "epoch": 0.14187380497131932,
+      "grad_norm": 3.1659598350524902,
+      "learning_rate": 4.199601581174351e-06,
+      "loss": 0.4498,
+      "step": 742
+    },
+    {
+      "epoch": 0.14206500956022944,
+      "grad_norm": 2.8837954998016357,
+      "learning_rate": 4.200457343173363e-06,
+      "loss": 0.1854,
+      "step": 743
+    },
+    {
+      "epoch": 0.14225621414913958,
+      "grad_norm": 2.6731128692626953,
+      "learning_rate": 4.20131195418077e-06,
+      "loss": 0.2493,
+      "step": 744
+    },
+    {
+      "epoch": 0.14244741873804972,
+      "grad_norm": 2.6178650856018066,
+      "learning_rate": 4.2021654172885625e-06,
+      "loss": 0.4237,
+      "step": 745
+    },
+    {
+      "epoch": 0.14263862332695984,
+      "grad_norm": 2.731067657470703,
+      "learning_rate": 4.203017735576281e-06,
+      "loss": 0.3222,
+      "step": 746
+    },
+    {
+      "epoch": 0.14282982791586998,
+      "grad_norm": 2.4263575077056885,
+      "learning_rate": 4.203868912111097e-06,
+      "loss": 0.3628,
+      "step": 747
+    },
+    {
+      "epoch": 0.14302103250478013,
+      "grad_norm": 1.9994317293167114,
+      "learning_rate": 4.204718949947867e-06,
+      "loss": 0.1756,
+      "step": 748
+    },
+    {
+      "epoch": 0.14321223709369024,
+      "grad_norm": 3.0634548664093018,
+      "learning_rate": 4.205567852129206e-06,
+      "loss": 0.2029,
+      "step": 749
+    },
+    {
+      "epoch": 0.14340344168260039,
+      "grad_norm": 2.2715470790863037,
+      "learning_rate": 4.20641562168555e-06,
+      "loss": 0.2621,
+      "step": 750
+    },
+    {
+      "epoch": 0.14359464627151053,
+      "grad_norm": 4.5768842697143555,
+      "learning_rate": 4.207262261635224e-06,
+      "loss": 0.3184,
+      "step": 751
+    },
+    {
+      "epoch": 0.14378585086042064,
+      "grad_norm": 2.8422701358795166,
+      "learning_rate": 4.208107774984498e-06,
+      "loss": 0.3711,
+      "step": 752
+    },
+    {
+      "epoch": 0.1439770554493308,
+      "grad_norm": 1.645248293876648,
+      "learning_rate": 4.20895216472766e-06,
+      "loss": 0.1714,
+      "step": 753
+    },
+    {
+      "epoch": 0.14416826003824093,
+      "grad_norm": 2.0097696781158447,
+      "learning_rate": 4.20979543384708e-06,
+      "loss": 0.2713,
+      "step": 754
+    },
+    {
+      "epoch": 0.14435946462715105,
+      "grad_norm": 1.5780304670333862,
+      "learning_rate": 4.210637585313263e-06,
+      "loss": 0.1144,
+      "step": 755
+    },
+    {
+      "epoch": 0.1445506692160612,
+      "grad_norm": 2.3878743648529053,
+      "learning_rate": 4.2114786220849235e-06,
+      "loss": 0.1314,
+      "step": 756
+    },
+    {
+      "epoch": 0.14474187380497133,
+      "grad_norm": 2.5952939987182617,
+      "learning_rate": 4.212318547109039e-06,
+      "loss": 0.3168,
+      "step": 757
+    },
+    {
+      "epoch": 0.14493307839388145,
+      "grad_norm": 1.6484967470169067,
+      "learning_rate": 4.213157363320921e-06,
+      "loss": 0.195,
+      "step": 758
+    },
+    {
+      "epoch": 0.1451242829827916,
+      "grad_norm": 1.6973950862884521,
+      "learning_rate": 4.213995073644266e-06,
+      "loss": 0.0901,
+      "step": 759
+    },
+    {
+      "epoch": 0.14531548757170173,
+      "grad_norm": 2.196223258972168,
+      "learning_rate": 4.214831680991224e-06,
+      "loss": 0.1255,
+      "step": 760
+    },
+    {
+      "epoch": 0.14550669216061185,
+      "grad_norm": 1.6075400114059448,
+      "learning_rate": 4.2156671882624575e-06,
+      "loss": 0.2015,
+      "step": 761
+    },
+    {
+      "epoch": 0.145697896749522,
+      "grad_norm": 1.5651001930236816,
+      "learning_rate": 4.2165015983472025e-06,
+      "loss": 0.0658,
+      "step": 762
+    },
+    {
+      "epoch": 0.1458891013384321,
+      "grad_norm": 2.242047071456909,
+      "learning_rate": 4.217334914123325e-06,
+      "loss": 0.2173,
+      "step": 763
+    },
+    {
+      "epoch": 0.14608030592734225,
+      "grad_norm": 4.271725654602051,
+      "learning_rate": 4.218167138457386e-06,
+      "loss": 0.3126,
+      "step": 764
+    },
+    {
+      "epoch": 0.1462715105162524,
+      "grad_norm": 1.832981824874878,
+      "learning_rate": 4.218998274204695e-06,
+      "loss": 0.1925,
+      "step": 765
+    },
+    {
+      "epoch": 0.1464627151051625,
+      "grad_norm": 2.5712363719940186,
+      "learning_rate": 4.219828324209373e-06,
+      "loss": 0.2876,
+      "step": 766
+    },
+    {
+      "epoch": 0.14665391969407265,
+      "grad_norm": 1.7425990104675293,
+      "learning_rate": 4.2206572913044095e-06,
+      "loss": 0.1229,
+      "step": 767
+    },
+    {
+      "epoch": 0.1468451242829828,
+      "grad_norm": 2.2643516063690186,
+      "learning_rate": 4.22148517831172e-06,
+      "loss": 0.1774,
+      "step": 768
+    },
+    {
+      "epoch": 0.1470363288718929,
+      "grad_norm": 2.998650312423706,
+      "learning_rate": 4.222311988042205e-06,
+      "loss": 0.1941,
+      "step": 769
+    },
+    {
+      "epoch": 0.14722753346080306,
+      "grad_norm": 3.894683361053467,
+      "learning_rate": 4.223137723295803e-06,
+      "loss": 0.6313,
+      "step": 770
+    },
+    {
+      "epoch": 0.1474187380497132,
+      "grad_norm": 1.2565118074417114,
+      "learning_rate": 4.2239623868615564e-06,
+      "loss": 0.2454,
+      "step": 771
+    },
+    {
+      "epoch": 0.14760994263862331,
+      "grad_norm": 3.5236380100250244,
+      "learning_rate": 4.224785981517657e-06,
+      "loss": 0.3753,
+      "step": 772
+    },
+    {
+      "epoch": 0.14780114722753346,
+      "grad_norm": 2.124140501022339,
+      "learning_rate": 4.225608510031509e-06,
+      "loss": 0.241,
+      "step": 773
+    },
+    {
+      "epoch": 0.1479923518164436,
+      "grad_norm": 1.7390496730804443,
+      "learning_rate": 4.226429975159786e-06,
+      "loss": 0.0968,
+      "step": 774
+    },
+    {
+      "epoch": 0.14818355640535372,
+      "grad_norm": 3.544078826904297,
+      "learning_rate": 4.22725037964848e-06,
+      "loss": 0.1911,
+      "step": 775
+    },
+    {
+      "epoch": 0.14837476099426386,
+      "grad_norm": 4.519067287445068,
+      "learning_rate": 4.228069726232962e-06,
+      "loss": 0.8982,
+      "step": 776
+    },
+    {
+      "epoch": 0.148565965583174,
+      "grad_norm": 2.7855448722839355,
+      "learning_rate": 4.228888017638035e-06,
+      "loss": 0.3423,
+      "step": 777
+    },
+    {
+      "epoch": 0.14875717017208412,
+      "grad_norm": 3.33547043800354,
+      "learning_rate": 4.229705256577988e-06,
+      "loss": 0.2157,
+      "step": 778
+    },
+    {
+      "epoch": 0.14894837476099426,
+      "grad_norm": 2.3527631759643555,
+      "learning_rate": 4.2305214457566505e-06,
+      "loss": 0.1733,
+      "step": 779
+    },
+    {
+      "epoch": 0.1491395793499044,
+      "grad_norm": 3.6699483394622803,
+      "learning_rate": 4.231336587867446e-06,
+      "loss": 0.2432,
+      "step": 780
+    },
+    {
+      "epoch": 0.14933078393881452,
+      "grad_norm": 1.638396143913269,
+      "learning_rate": 4.232150685593444e-06,
+      "loss": 0.0568,
+      "step": 781
+    },
+    {
+      "epoch": 0.14952198852772466,
+      "grad_norm": 3.646986961364746,
+      "learning_rate": 4.232963741607416e-06,
+      "loss": 0.4378,
+      "step": 782
+    },
+    {
+      "epoch": 0.1497131931166348,
+      "grad_norm": 2.220862865447998,
+      "learning_rate": 4.233775758571886e-06,
+      "loss": 0.2993,
+      "step": 783
+    },
+    {
+      "epoch": 0.14990439770554492,
+      "grad_norm": 3.7839043140411377,
+      "learning_rate": 4.234586739139182e-06,
+      "loss": 0.47,
+      "step": 784
+    },
+    {
+      "epoch": 0.15009560229445507,
+      "grad_norm": 1.9454379081726074,
+      "learning_rate": 4.235396685951493e-06,
+      "loss": 0.1672,
+      "step": 785
+    },
+    {
+      "epoch": 0.1502868068833652,
+      "grad_norm": 2.693408727645874,
+      "learning_rate": 4.236205601640911e-06,
+      "loss": 0.3557,
+      "step": 786
+    },
+    {
+      "epoch": 0.15047801147227532,
+      "grad_norm": 2.1505634784698486,
+      "learning_rate": 4.237013488829494e-06,
+      "loss": 0.0909,
+      "step": 787
+    },
+    {
+      "epoch": 0.15066921606118547,
+      "grad_norm": 1.8870456218719482,
+      "learning_rate": 4.237820350129308e-06,
+      "loss": 0.1915,
+      "step": 788
+    },
+    {
+      "epoch": 0.1508604206500956,
+      "grad_norm": 3.362215518951416,
+      "learning_rate": 4.23862618814248e-06,
+      "loss": 0.549,
+      "step": 789
+    },
+    {
+      "epoch": 0.15105162523900573,
+      "grad_norm": 3.406558036804199,
+      "learning_rate": 4.239431005461254e-06,
+      "loss": 0.3686,
+      "step": 790
+    },
+    {
+      "epoch": 0.15124282982791587,
+      "grad_norm": 2.572495460510254,
+      "learning_rate": 4.240234804668029e-06,
+      "loss": 0.2132,
+      "step": 791
+    },
+    {
+      "epoch": 0.151434034416826,
+      "grad_norm": 3.3531334400177,
+      "learning_rate": 4.2410375883354235e-06,
+      "loss": 0.2027,
+      "step": 792
+    },
+    {
+      "epoch": 0.15162523900573613,
+      "grad_norm": 1.9929064512252808,
+      "learning_rate": 4.241839359026311e-06,
+      "loss": 0.0925,
+      "step": 793
+    },
+    {
+      "epoch": 0.15181644359464627,
+      "grad_norm": 4.2206010818481445,
+      "learning_rate": 4.242640119293882e-06,
+      "loss": 0.3048,
+      "step": 794
+    },
+    {
+      "epoch": 0.15200764818355642,
+      "grad_norm": 3.0227558612823486,
+      "learning_rate": 4.24343987168168e-06,
+      "loss": 0.372,
+      "step": 795
+    },
+    {
+      "epoch": 0.15219885277246653,
+      "grad_norm": 2.1193037033081055,
+      "learning_rate": 4.2442386187236584e-06,
+      "loss": 0.2249,
+      "step": 796
+    },
+    {
+      "epoch": 0.15239005736137667,
+      "grad_norm": 2.863941192626953,
+      "learning_rate": 4.2450363629442295e-06,
+      "loss": 0.3385,
+      "step": 797
+    },
+    {
+      "epoch": 0.15258126195028682,
+      "grad_norm": 1.4282262325286865,
+      "learning_rate": 4.245833106858305e-06,
+      "loss": 0.1862,
+      "step": 798
+    },
+    {
+      "epoch": 0.15277246653919693,
+      "grad_norm": 3.935479164123535,
+      "learning_rate": 4.2466288529713505e-06,
+      "loss": 0.2419,
+      "step": 799
+    },
+    {
+      "epoch": 0.15296367112810708,
+      "grad_norm": 1.3507992029190063,
+      "learning_rate": 4.247423603779429e-06,
+      "loss": 0.1504,
+      "step": 800
+    },
+    {
+      "epoch": 0.15315487571701722,
+      "grad_norm": 3.4311065673828125,
+      "learning_rate": 4.248217361769252e-06,
+      "loss": 0.4921,
+      "step": 801
+    },
+    {
+      "epoch": 0.15334608030592733,
+      "grad_norm": 1.9401636123657227,
+      "learning_rate": 4.2490101294182175e-06,
+      "loss": 0.3157,
+      "step": 802
+    },
+    {
+      "epoch": 0.15353728489483748,
+      "grad_norm": 2.6393685340881348,
+      "learning_rate": 4.249801909194468e-06,
+      "loss": 0.3513,
+      "step": 803
+    },
+    {
+      "epoch": 0.15372848948374762,
+      "grad_norm": 2.4486782550811768,
+      "learning_rate": 4.250592703556928e-06,
+      "loss": 0.1468,
+      "step": 804
+    },
+    {
+      "epoch": 0.15391969407265774,
+      "grad_norm": 1.5397764444351196,
+      "learning_rate": 4.251382514955353e-06,
+      "loss": 0.1181,
+      "step": 805
+    },
+    {
+      "epoch": 0.15411089866156788,
+      "grad_norm": 2.7118899822235107,
+      "learning_rate": 4.252171345830375e-06,
+      "loss": 0.1423,
+      "step": 806
+    },
+    {
+      "epoch": 0.15430210325047802,
+      "grad_norm": 2.3002309799194336,
+      "learning_rate": 4.252959198613545e-06,
+      "loss": 0.2643,
+      "step": 807
+    },
+    {
+      "epoch": 0.15449330783938814,
+      "grad_norm": 4.120339393615723,
+      "learning_rate": 4.253746075727386e-06,
+      "loss": 0.3035,
+      "step": 808
+    },
+    {
+      "epoch": 0.15468451242829828,
+      "grad_norm": 1.6706206798553467,
+      "learning_rate": 4.254531979585426e-06,
+      "loss": 0.1403,
+      "step": 809
+    },
+    {
+      "epoch": 0.15487571701720843,
+      "grad_norm": 1.846777319908142,
+      "learning_rate": 4.2553169125922515e-06,
+      "loss": 0.1069,
+      "step": 810
+    },
+    {
+      "epoch": 0.15506692160611854,
+      "grad_norm": 2.311530590057373,
+      "learning_rate": 4.256100877143548e-06,
+      "loss": 0.1826,
+      "step": 811
+    },
+    {
+      "epoch": 0.15525812619502868,
+      "grad_norm": 1.510845422744751,
+      "learning_rate": 4.256883875626145e-06,
+      "loss": 0.0622,
+      "step": 812
+    },
+    {
+      "epoch": 0.15544933078393883,
+      "grad_norm": 1.923916220664978,
+      "learning_rate": 4.2576659104180575e-06,
+      "loss": 0.2099,
+      "step": 813
+    },
+    {
+      "epoch": 0.15564053537284894,
+      "grad_norm": 3.773207902908325,
+      "learning_rate": 4.258446983888535e-06,
+      "loss": 0.5412,
+      "step": 814
+    },
+    {
+      "epoch": 0.15583173996175909,
+      "grad_norm": 2.4419755935668945,
+      "learning_rate": 4.259227098398094e-06,
+      "loss": 0.4255,
+      "step": 815
+    },
+    {
+      "epoch": 0.15602294455066923,
+      "grad_norm": 4.022161960601807,
+      "learning_rate": 4.260006256298574e-06,
+      "loss": 0.4799,
+      "step": 816
+    },
+    {
+      "epoch": 0.15621414913957934,
+      "grad_norm": 1.609773874282837,
+      "learning_rate": 4.260784459933167e-06,
+      "loss": 0.1932,
+      "step": 817
+    },
+    {
+      "epoch": 0.1564053537284895,
+      "grad_norm": 1.0987634658813477,
+      "learning_rate": 4.261561711636471e-06,
+      "loss": 0.051,
+      "step": 818
+    },
+    {
+      "epoch": 0.15659655831739963,
+      "grad_norm": 2.9759936332702637,
+      "learning_rate": 4.262338013734527e-06,
+      "loss": 0.178,
+      "step": 819
+    },
+    {
+      "epoch": 0.15678776290630975,
+      "grad_norm": 1.6191765069961548,
+      "learning_rate": 4.263113368544856e-06,
+      "loss": 0.1116,
+      "step": 820
+    },
+    {
+      "epoch": 0.1569789674952199,
+      "grad_norm": 2.261420965194702,
+      "learning_rate": 4.2638877783765115e-06,
+      "loss": 0.2843,
+      "step": 821
+    },
+    {
+      "epoch": 0.15717017208413,
+      "grad_norm": 1.752107858657837,
+      "learning_rate": 4.26466124553011e-06,
+      "loss": 0.2237,
+      "step": 822
+    },
+    {
+      "epoch": 0.15736137667304015,
+      "grad_norm": 2.615859270095825,
+      "learning_rate": 4.265433772297882e-06,
+      "loss": 0.1974,
+      "step": 823
+    },
+    {
+      "epoch": 0.1575525812619503,
+      "grad_norm": 1.7172234058380127,
+      "learning_rate": 4.2662053609637e-06,
+      "loss": 0.1186,
+      "step": 824
+    },
+    {
+      "epoch": 0.1577437858508604,
+      "grad_norm": 1.4844310283660889,
+      "learning_rate": 4.266976013803132e-06,
+      "loss": 0.0642,
+      "step": 825
+    },
+    {
+      "epoch": 0.15793499043977055,
+      "grad_norm": 2.956435441970825,
+      "learning_rate": 4.267745733083475e-06,
+      "loss": 0.2152,
+      "step": 826
+    },
+    {
+      "epoch": 0.1581261950286807,
+      "grad_norm": 3.139362335205078,
+      "learning_rate": 4.268514521063796e-06,
+      "loss": 0.3454,
+      "step": 827
+    },
+    {
+      "epoch": 0.1583173996175908,
+      "grad_norm": 3.0984015464782715,
+      "learning_rate": 4.269282379994972e-06,
+      "loss": 0.3052,
+      "step": 828
+    },
+    {
+      "epoch": 0.15850860420650095,
+      "grad_norm": 2.265077829360962,
+      "learning_rate": 4.27004931211973e-06,
+      "loss": 0.1137,
+      "step": 829
+    },
+    {
+      "epoch": 0.1586998087954111,
+      "grad_norm": 3.9762401580810547,
+      "learning_rate": 4.270815319672684e-06,
+      "loss": 0.3162,
+      "step": 830
+    },
+    {
+      "epoch": 0.1588910133843212,
+      "grad_norm": 2.093393325805664,
+      "learning_rate": 4.2715804048803785e-06,
+      "loss": 0.0985,
+      "step": 831
+    },
+    {
+      "epoch": 0.15908221797323135,
+      "grad_norm": 1.6958585977554321,
+      "learning_rate": 4.272344569961324e-06,
+      "loss": 0.1414,
+      "step": 832
+    },
+    {
+      "epoch": 0.1592734225621415,
+      "grad_norm": 1.754672884941101,
+      "learning_rate": 4.273107817126036e-06,
+      "loss": 0.1115,
+      "step": 833
+    },
+    {
+      "epoch": 0.1594646271510516,
+      "grad_norm": 2.3991124629974365,
+      "learning_rate": 4.273870148577072e-06,
+      "loss": 0.3149,
+      "step": 834
+    },
+    {
+      "epoch": 0.15965583173996176,
+      "grad_norm": 1.6448285579681396,
+      "learning_rate": 4.2746315665090745e-06,
+      "loss": 0.0904,
+      "step": 835
+    },
+    {
+      "epoch": 0.1598470363288719,
+      "grad_norm": 2.705641984939575,
+      "learning_rate": 4.275392073108804e-06,
+      "loss": 0.2193,
+      "step": 836
+    },
+    {
+      "epoch": 0.16003824091778202,
+      "grad_norm": 3.0307109355926514,
+      "learning_rate": 4.27615167055518e-06,
+      "loss": 0.1033,
+      "step": 837
+    },
+    {
+      "epoch": 0.16022944550669216,
+      "grad_norm": 3.0683274269104004,
+      "learning_rate": 4.276910361019314e-06,
+      "loss": 0.3433,
+      "step": 838
+    },
+    {
+      "epoch": 0.1604206500956023,
+      "grad_norm": 2.956444501876831,
+      "learning_rate": 4.277668146664553e-06,
+      "loss": 0.3323,
+      "step": 839
+    },
+    {
+      "epoch": 0.16061185468451242,
+      "grad_norm": 1.917574167251587,
+      "learning_rate": 4.278425029646511e-06,
+      "loss": 0.1704,
+      "step": 840
+    },
+    {
+      "epoch": 0.16080305927342256,
+      "grad_norm": 2.8657712936401367,
+      "learning_rate": 4.2791810121131075e-06,
+      "loss": 0.217,
+      "step": 841
+    },
+    {
+      "epoch": 0.1609942638623327,
+      "grad_norm": 2.1672983169555664,
+      "learning_rate": 4.279936096204607e-06,
+      "loss": 0.1102,
+      "step": 842
+    },
+    {
+      "epoch": 0.16118546845124282,
+      "grad_norm": 2.7191829681396484,
+      "learning_rate": 4.2806902840536505e-06,
+      "loss": 0.2781,
+      "step": 843
+    },
+    {
+      "epoch": 0.16137667304015296,
+      "grad_norm": 1.849647045135498,
+      "learning_rate": 4.2814435777852955e-06,
+      "loss": 0.2105,
+      "step": 844
+    },
+    {
+      "epoch": 0.1615678776290631,
+      "grad_norm": 2.904238700866699,
+      "learning_rate": 4.2821959795170494e-06,
+      "loss": 0.5104,
+      "step": 845
+    },
+    {
+      "epoch": 0.16175908221797322,
+      "grad_norm": 1.7526648044586182,
+      "learning_rate": 4.282947491358906e-06,
+      "loss": 0.1464,
+      "step": 846
+    },
+    {
+      "epoch": 0.16195028680688336,
+      "grad_norm": 3.317267417907715,
+      "learning_rate": 4.283698115413385e-06,
+      "loss": 0.4144,
+      "step": 847
+    },
+    {
+      "epoch": 0.1621414913957935,
+      "grad_norm": 1.7838001251220703,
+      "learning_rate": 4.284447853775558e-06,
+      "loss": 0.0672,
+      "step": 848
+    },
+    {
+      "epoch": 0.16233269598470362,
+      "grad_norm": 2.0435261726379395,
+      "learning_rate": 4.2851967085330925e-06,
+      "loss": 0.1293,
+      "step": 849
+    },
+    {
+      "epoch": 0.16252390057361377,
+      "grad_norm": 2.0593795776367188,
+      "learning_rate": 4.2859446817662824e-06,
+      "loss": 0.0951,
+      "step": 850
+    },
+    {
+      "epoch": 0.1627151051625239,
+      "grad_norm": 2.386403799057007,
+      "learning_rate": 4.286691775548084e-06,
+      "loss": 0.3164,
+      "step": 851
+    },
+    {
+      "epoch": 0.16290630975143403,
+      "grad_norm": 1.7959471940994263,
+      "learning_rate": 4.28743799194415e-06,
+      "loss": 0.1443,
+      "step": 852
+    },
+    {
+      "epoch": 0.16309751434034417,
+      "grad_norm": 4.4852681159973145,
+      "learning_rate": 4.288183333012865e-06,
+      "loss": 0.6776,
+      "step": 853
+    },
+    {
+      "epoch": 0.1632887189292543,
+      "grad_norm": 1.6934727430343628,
+      "learning_rate": 4.288927800805377e-06,
+      "loss": 0.1368,
+      "step": 854
+    },
+    {
+      "epoch": 0.16347992351816443,
+      "grad_norm": 3.038947582244873,
+      "learning_rate": 4.289671397365632e-06,
+      "loss": 0.3026,
+      "step": 855
+    },
+    {
+      "epoch": 0.16367112810707457,
+      "grad_norm": 2.0727596282958984,
+      "learning_rate": 4.290414124730413e-06,
+      "loss": 0.1034,
+      "step": 856
+    },
+    {
+      "epoch": 0.1638623326959847,
+      "grad_norm": 3.337085008621216,
+      "learning_rate": 4.291155984929362e-06,
+      "loss": 0.4055,
+      "step": 857
+    },
+    {
+      "epoch": 0.16405353728489483,
+      "grad_norm": 2.652127504348755,
+      "learning_rate": 4.291896979985027e-06,
+      "loss": 0.3585,
+      "step": 858
+    },
+    {
+      "epoch": 0.16424474187380497,
+      "grad_norm": 3.1562952995300293,
+      "learning_rate": 4.292637111912882e-06,
+      "loss": 0.446,
+      "step": 859
+    },
+    {
+      "epoch": 0.16443594646271512,
+      "grad_norm": 2.3302462100982666,
+      "learning_rate": 4.293376382721373e-06,
+      "loss": 0.2846,
+      "step": 860
+    },
+    {
+      "epoch": 0.16462715105162523,
+      "grad_norm": 1.904043197631836,
+      "learning_rate": 4.294114794411938e-06,
+      "loss": 0.1925,
+      "step": 861
+    },
+    {
+      "epoch": 0.16481835564053537,
+      "grad_norm": 2.2849550247192383,
+      "learning_rate": 4.294852348979047e-06,
+      "loss": 0.1391,
+      "step": 862
+    },
+    {
+      "epoch": 0.16500956022944552,
+      "grad_norm": 2.6952767372131348,
+      "learning_rate": 4.295589048410235e-06,
+      "loss": 0.302,
+      "step": 863
+    },
+    {
+      "epoch": 0.16520076481835563,
+      "grad_norm": 3.0628738403320312,
+      "learning_rate": 4.2963248946861294e-06,
+      "loss": 0.4776,
+      "step": 864
+    },
+    {
+      "epoch": 0.16539196940726578,
+      "grad_norm": 2.940101385116577,
+      "learning_rate": 4.297059889780485e-06,
+      "loss": 0.2961,
+      "step": 865
+    },
+    {
+      "epoch": 0.16558317399617592,
+      "grad_norm": 2.9635932445526123,
+      "learning_rate": 4.297794035660217e-06,
+      "loss": 0.4084,
+      "step": 866
+    },
+    {
+      "epoch": 0.16577437858508604,
+      "grad_norm": 1.4240761995315552,
+      "learning_rate": 4.298527334285426e-06,
+      "loss": 0.0877,
+      "step": 867
+    },
+    {
+      "epoch": 0.16596558317399618,
+      "grad_norm": 1.576123595237732,
+      "learning_rate": 4.29925978760944e-06,
+      "loss": 0.1101,
+      "step": 868
+    },
+    {
+      "epoch": 0.16615678776290632,
+      "grad_norm": 2.8661000728607178,
+      "learning_rate": 4.299991397578835e-06,
+      "loss": 0.1854,
+      "step": 869
+    },
+    {
+      "epoch": 0.16634799235181644,
+      "grad_norm": 2.202101469039917,
+      "learning_rate": 4.300722166133473e-06,
+      "loss": 0.1717,
+      "step": 870
+    },
+    {
+      "epoch": 0.16653919694072658,
+      "grad_norm": 1.9804801940917969,
+      "learning_rate": 4.301452095206531e-06,
+      "loss": 0.419,
+      "step": 871
+    },
+    {
+      "epoch": 0.16673040152963672,
+      "grad_norm": 1.402471899986267,
+      "learning_rate": 4.302181186724532e-06,
+      "loss": 0.1489,
+      "step": 872
+    },
+    {
+      "epoch": 0.16692160611854684,
+      "grad_norm": 2.836812973022461,
+      "learning_rate": 4.302909442607371e-06,
+      "loss": 0.2623,
+      "step": 873
+    },
+    {
+      "epoch": 0.16711281070745698,
+      "grad_norm": 1.8831483125686646,
+      "learning_rate": 4.303636864768353e-06,
+      "loss": 0.1072,
+      "step": 874
+    },
+    {
+      "epoch": 0.16730401529636713,
+      "grad_norm": 1.8472790718078613,
+      "learning_rate": 4.304363455114219e-06,
+      "loss": 0.1246,
+      "step": 875
+    },
+    {
+      "epoch": 0.16749521988527724,
+      "grad_norm": 2.792137622833252,
+      "learning_rate": 4.305089215545175e-06,
+      "loss": 0.3111,
+      "step": 876
+    },
+    {
+      "epoch": 0.16768642447418738,
+      "grad_norm": 2.943768262863159,
+      "learning_rate": 4.305814147954922e-06,
+      "loss": 0.6068,
+      "step": 877
+    },
+    {
+      "epoch": 0.16787762906309753,
+      "grad_norm": 2.1078765392303467,
+      "learning_rate": 4.306538254230691e-06,
+      "loss": 0.1484,
+      "step": 878
+    },
+    {
+      "epoch": 0.16806883365200764,
+      "grad_norm": 3.882357597351074,
+      "learning_rate": 4.307261536253264e-06,
+      "loss": 0.2091,
+      "step": 879
+    },
+    {
+      "epoch": 0.1682600382409178,
+      "grad_norm": 1.381016492843628,
+      "learning_rate": 4.30798399589701e-06,
+      "loss": 0.0775,
+      "step": 880
+    },
+    {
+      "epoch": 0.16845124282982793,
+      "grad_norm": 2.2446751594543457,
+      "learning_rate": 4.308705635029911e-06,
+      "loss": 0.0931,
+      "step": 881
+    },
+    {
+      "epoch": 0.16864244741873805,
+      "grad_norm": 3.000154733657837,
+      "learning_rate": 4.309426455513591e-06,
+      "loss": 0.3437,
+      "step": 882
+    },
+    {
+      "epoch": 0.1688336520076482,
+      "grad_norm": 2.4180312156677246,
+      "learning_rate": 4.310146459203349e-06,
+      "loss": 0.1738,
+      "step": 883
+    },
+    {
+      "epoch": 0.1690248565965583,
+      "grad_norm": 1.9331579208374023,
+      "learning_rate": 4.310865647948177e-06,
+      "loss": 0.2825,
+      "step": 884
+    },
+    {
+      "epoch": 0.16921606118546845,
+      "grad_norm": 3.176043748855591,
+      "learning_rate": 4.311584023590803e-06,
+      "loss": 0.435,
+      "step": 885
+    },
+    {
+      "epoch": 0.1694072657743786,
+      "grad_norm": 2.9748902320861816,
+      "learning_rate": 4.312301587967709e-06,
+      "loss": 0.442,
+      "step": 886
+    },
+    {
+      "epoch": 0.1695984703632887,
+      "grad_norm": 1.8928931951522827,
+      "learning_rate": 4.313018342909162e-06,
+      "loss": 0.0825,
+      "step": 887
+    },
+    {
+      "epoch": 0.16978967495219885,
+      "grad_norm": 3.570660352706909,
+      "learning_rate": 4.3137342902392415e-06,
+      "loss": 0.4021,
+      "step": 888
+    },
+    {
+      "epoch": 0.169980879541109,
+      "grad_norm": 1.958633542060852,
+      "learning_rate": 4.31444943177587e-06,
+      "loss": 0.2978,
+      "step": 889
+    },
+    {
+      "epoch": 0.1701720841300191,
+      "grad_norm": 2.055731773376465,
+      "learning_rate": 4.315163769330838e-06,
+      "loss": 0.2606,
+      "step": 890
+    },
+    {
+      "epoch": 0.17036328871892925,
+      "grad_norm": 2.733567714691162,
+      "learning_rate": 4.3158773047098325e-06,
+      "loss": 0.3208,
+      "step": 891
+    },
+    {
+      "epoch": 0.1705544933078394,
+      "grad_norm": 2.0781280994415283,
+      "learning_rate": 4.3165900397124626e-06,
+      "loss": 0.2838,
+      "step": 892
+    },
+    {
+      "epoch": 0.1707456978967495,
+      "grad_norm": 4.172160625457764,
+      "learning_rate": 4.317301976132288e-06,
+      "loss": 0.1485,
+      "step": 893
+    },
+    {
+      "epoch": 0.17093690248565965,
+      "grad_norm": 2.199040174484253,
+      "learning_rate": 4.31801311575685e-06,
+      "loss": 0.1481,
+      "step": 894
+    },
+    {
+      "epoch": 0.1711281070745698,
+      "grad_norm": 3.143498420715332,
+      "learning_rate": 4.318723460367692e-06,
+      "loss": 0.5856,
+      "step": 895
+    },
+    {
+      "epoch": 0.1713193116634799,
+      "grad_norm": 2.4403462409973145,
+      "learning_rate": 4.319433011740389e-06,
+      "loss": 0.4743,
+      "step": 896
+    },
+    {
+      "epoch": 0.17151051625239006,
+      "grad_norm": 1.9097824096679688,
+      "learning_rate": 4.3201417716445755e-06,
+      "loss": 0.1086,
+      "step": 897
+    },
+    {
+      "epoch": 0.1717017208413002,
+      "grad_norm": 2.7453420162200928,
+      "learning_rate": 4.32084974184397e-06,
+      "loss": 0.1981,
+      "step": 898
+    },
+    {
+      "epoch": 0.1718929254302103,
+      "grad_norm": 1.3835936784744263,
+      "learning_rate": 4.321556924096402e-06,
+      "loss": 0.1109,
+      "step": 899
+    },
+    {
+      "epoch": 0.17208413001912046,
+      "grad_norm": 1.855629563331604,
+      "learning_rate": 4.322263320153839e-06,
+      "loss": 0.0757,
+      "step": 900
+    },
+    {
+      "epoch": 0.1722753346080306,
+      "grad_norm": 2.9253125190734863,
+      "learning_rate": 4.322968931762411e-06,
+      "loss": 0.2818,
+      "step": 901
+    },
+    {
+      "epoch": 0.17246653919694072,
+      "grad_norm": 2.60461163520813,
+      "learning_rate": 4.323673760662438e-06,
+      "loss": 0.2558,
+      "step": 902
+    },
+    {
+      "epoch": 0.17265774378585086,
+      "grad_norm": 1.764817476272583,
+      "learning_rate": 4.324377808588454e-06,
+      "loss": 0.1228,
+      "step": 903
+    },
+    {
+      "epoch": 0.172848948374761,
+      "grad_norm": 2.949141502380371,
+      "learning_rate": 4.3250810772692355e-06,
+      "loss": 0.3251,
+      "step": 904
+    },
+    {
+      "epoch": 0.17304015296367112,
+      "grad_norm": 2.435056686401367,
+      "learning_rate": 4.3257835684278235e-06,
+      "loss": 0.3511,
+      "step": 905
+    },
+    {
+      "epoch": 0.17323135755258126,
+      "grad_norm": 2.2405662536621094,
+      "learning_rate": 4.3264852837815515e-06,
+      "loss": 0.0803,
+      "step": 906
+    },
+    {
+      "epoch": 0.1734225621414914,
+      "grad_norm": 2.5514614582061768,
+      "learning_rate": 4.327186225042066e-06,
+      "loss": 0.3172,
+      "step": 907
+    },
+    {
+      "epoch": 0.17361376673040152,
+      "grad_norm": 2.14652419090271,
+      "learning_rate": 4.327886393915363e-06,
+      "loss": 0.1575,
+      "step": 908
+    },
+    {
+      "epoch": 0.17380497131931166,
+      "grad_norm": 1.3867746591567993,
+      "learning_rate": 4.328585792101795e-06,
+      "loss": 0.1291,
+      "step": 909
+    },
+    {
+      "epoch": 0.1739961759082218,
+      "grad_norm": 3.6421027183532715,
+      "learning_rate": 4.329284421296114e-06,
+      "loss": 0.5589,
+      "step": 910
+    },
+    {
+      "epoch": 0.17418738049713192,
+      "grad_norm": 2.516690731048584,
+      "learning_rate": 4.329982283187484e-06,
+      "loss": 0.1707,
+      "step": 911
+    },
+    {
+      "epoch": 0.17437858508604206,
+      "grad_norm": 3.2310264110565186,
+      "learning_rate": 4.330679379459511e-06,
+      "loss": 0.2891,
+      "step": 912
+    },
+    {
+      "epoch": 0.1745697896749522,
+      "grad_norm": 1.8623366355895996,
+      "learning_rate": 4.331375711790265e-06,
+      "loss": 0.1513,
+      "step": 913
+    },
+    {
+      "epoch": 0.17476099426386232,
+      "grad_norm": 3.0351459980010986,
+      "learning_rate": 4.332071281852306e-06,
+      "loss": 0.3003,
+      "step": 914
+    },
+    {
+      "epoch": 0.17495219885277247,
+      "grad_norm": 3.0663366317749023,
+      "learning_rate": 4.332766091312704e-06,
+      "loss": 0.3235,
+      "step": 915
+    },
+    {
+      "epoch": 0.1751434034416826,
+      "grad_norm": 1.8554519414901733,
+      "learning_rate": 4.333460141833072e-06,
+      "loss": 0.1387,
+      "step": 916
+    },
+    {
+      "epoch": 0.17533460803059273,
+      "grad_norm": 3.4480252265930176,
+      "learning_rate": 4.33415343506958e-06,
+      "loss": 0.2172,
+      "step": 917
+    },
+    {
+      "epoch": 0.17552581261950287,
+      "grad_norm": 2.396263837814331,
+      "learning_rate": 4.334845972672983e-06,
+      "loss": 0.2369,
+      "step": 918
+    },
+    {
+      "epoch": 0.175717017208413,
+      "grad_norm": 2.6219048500061035,
+      "learning_rate": 4.335537756288644e-06,
+      "loss": 0.2771,
+      "step": 919
+    },
+    {
+      "epoch": 0.17590822179732313,
+      "grad_norm": 2.162111759185791,
+      "learning_rate": 4.3362287875565595e-06,
+      "loss": 0.3263,
+      "step": 920
+    },
+    {
+      "epoch": 0.17609942638623327,
+      "grad_norm": 2.3528635501861572,
+      "learning_rate": 4.33691906811138e-06,
+      "loss": 0.1726,
+      "step": 921
+    },
+    {
+      "epoch": 0.17629063097514341,
+      "grad_norm": 3.782550096511841,
+      "learning_rate": 4.337608599582434e-06,
+      "loss": 0.3164,
+      "step": 922
+    },
+    {
+      "epoch": 0.17648183556405353,
+      "grad_norm": 2.536262035369873,
+      "learning_rate": 4.338297383593754e-06,
+      "loss": 0.3199,
+      "step": 923
+    },
+    {
+      "epoch": 0.17667304015296367,
+      "grad_norm": 4.701323986053467,
+      "learning_rate": 4.338985421764091e-06,
+      "loss": 0.258,
+      "step": 924
+    },
+    {
+      "epoch": 0.17686424474187382,
+      "grad_norm": 2.707467555999756,
+      "learning_rate": 4.33967271570695e-06,
+      "loss": 0.105,
+      "step": 925
+    },
+    {
+      "epoch": 0.17705544933078393,
+      "grad_norm": 3.847241163253784,
+      "learning_rate": 4.340359267030601e-06,
+      "loss": 0.6552,
+      "step": 926
+    },
+    {
+      "epoch": 0.17724665391969407,
+      "grad_norm": 1.8708347082138062,
+      "learning_rate": 4.341045077338109e-06,
+      "loss": 0.2187,
+      "step": 927
+    },
+    {
+      "epoch": 0.17743785850860422,
+      "grad_norm": 2.500690460205078,
+      "learning_rate": 4.341730148227351e-06,
+      "loss": 0.1601,
+      "step": 928
+    },
+    {
+      "epoch": 0.17762906309751433,
+      "grad_norm": 1.6813768148422241,
+      "learning_rate": 4.342414481291046e-06,
+      "loss": 0.0899,
+      "step": 929
+    },
+    {
+      "epoch": 0.17782026768642448,
+      "grad_norm": 1.971083641052246,
+      "learning_rate": 4.343098078116767e-06,
+      "loss": 0.1689,
+      "step": 930
+    },
+    {
+      "epoch": 0.17801147227533462,
+      "grad_norm": 1.656010389328003,
+      "learning_rate": 4.343780940286974e-06,
+      "loss": 0.0773,
+      "step": 931
+    },
+    {
+      "epoch": 0.17820267686424474,
+      "grad_norm": 3.547978639602661,
+      "learning_rate": 4.344463069379024e-06,
+      "loss": 0.4317,
+      "step": 932
+    },
+    {
+      "epoch": 0.17839388145315488,
+      "grad_norm": 1.7893317937850952,
+      "learning_rate": 4.345144466965205e-06,
+      "loss": 0.1723,
+      "step": 933
+    },
+    {
+      "epoch": 0.17858508604206502,
+      "grad_norm": 1.7867648601531982,
+      "learning_rate": 4.345825134612752e-06,
+      "loss": 0.1345,
+      "step": 934
+    },
+    {
+      "epoch": 0.17877629063097514,
+      "grad_norm": 3.299664258956909,
+      "learning_rate": 4.3465050738838635e-06,
+      "loss": 0.4693,
+      "step": 935
+    },
+    {
+      "epoch": 0.17896749521988528,
+      "grad_norm": 2.6684377193450928,
+      "learning_rate": 4.347184286335733e-06,
+      "loss": 0.3825,
+      "step": 936
+    },
+    {
+      "epoch": 0.17915869980879542,
+      "grad_norm": 1.9024335145950317,
+      "learning_rate": 4.347862773520565e-06,
+      "loss": 0.1348,
+      "step": 937
+    },
+    {
+      "epoch": 0.17934990439770554,
+      "grad_norm": 2.407212972640991,
+      "learning_rate": 4.348540536985596e-06,
+      "loss": 0.3481,
+      "step": 938
+    },
+    {
+      "epoch": 0.17954110898661568,
+      "grad_norm": 2.431903600692749,
+      "learning_rate": 4.349217578273117e-06,
+      "loss": 0.326,
+      "step": 939
+    },
+    {
+      "epoch": 0.17973231357552583,
+      "grad_norm": 2.8086395263671875,
+      "learning_rate": 4.349893898920495e-06,
+      "loss": 0.3743,
+      "step": 940
+    },
+    {
+      "epoch": 0.17992351816443594,
+      "grad_norm": 1.9264482259750366,
+      "learning_rate": 4.350569500460188e-06,
+      "loss": 0.2072,
+      "step": 941
+    },
+    {
+      "epoch": 0.18011472275334608,
+      "grad_norm": 1.8798692226409912,
+      "learning_rate": 4.35124438441978e-06,
+      "loss": 0.0674,
+      "step": 942
+    },
+    {
+      "epoch": 0.1803059273422562,
+      "grad_norm": 1.4653009176254272,
+      "learning_rate": 4.351918552321986e-06,
+      "loss": 0.0763,
+      "step": 943
+    },
+    {
+      "epoch": 0.18049713193116634,
+      "grad_norm": 2.255692958831787,
+      "learning_rate": 4.3525920056846815e-06,
+      "loss": 0.124,
+      "step": 944
+    },
+    {
+      "epoch": 0.1806883365200765,
+      "grad_norm": 2.478980779647827,
+      "learning_rate": 4.35326474602092e-06,
+      "loss": 0.3314,
+      "step": 945
+    },
+    {
+      "epoch": 0.1808795411089866,
+      "grad_norm": 1.5225646495819092,
+      "learning_rate": 4.353936774838954e-06,
+      "loss": 0.353,
+      "step": 946
+    },
+    {
+      "epoch": 0.18107074569789675,
+      "grad_norm": 1.214327096939087,
+      "learning_rate": 4.354608093642257e-06,
+      "loss": 0.0809,
+      "step": 947
+    },
+    {
+      "epoch": 0.1812619502868069,
+      "grad_norm": 2.2034029960632324,
+      "learning_rate": 4.355278703929541e-06,
+      "loss": 0.2837,
+      "step": 948
+    },
+    {
+      "epoch": 0.181453154875717,
+      "grad_norm": 3.38645076751709,
+      "learning_rate": 4.3559486071947785e-06,
+      "loss": 0.378,
+      "step": 949
+    },
+    {
+      "epoch": 0.18164435946462715,
+      "grad_norm": 1.7553558349609375,
+      "learning_rate": 4.3566178049272204e-06,
+      "loss": 0.1021,
+      "step": 950
+    },
+    {
+      "epoch": 0.1818355640535373,
+      "grad_norm": 2.1390013694763184,
+      "learning_rate": 4.357286298611418e-06,
+      "loss": 0.2978,
+      "step": 951
+    },
+    {
+      "epoch": 0.1820267686424474,
+      "grad_norm": 2.9334323406219482,
+      "learning_rate": 4.3579540897272416e-06,
+      "loss": 0.5175,
+      "step": 952
+    },
+    {
+      "epoch": 0.18221797323135755,
+      "grad_norm": 1.2445647716522217,
+      "learning_rate": 4.358621179749902e-06,
+      "loss": 0.108,
+      "step": 953
+    },
+    {
+      "epoch": 0.1824091778202677,
+      "grad_norm": 1.3846619129180908,
+      "learning_rate": 4.359287570149967e-06,
+      "loss": 0.1046,
+      "step": 954
+    },
+    {
+      "epoch": 0.1826003824091778,
+      "grad_norm": 2.8698277473449707,
+      "learning_rate": 4.359953262393382e-06,
+      "loss": 0.1461,
+      "step": 955
+    },
+    {
+      "epoch": 0.18279158699808795,
+      "grad_norm": 1.6638492345809937,
+      "learning_rate": 4.3606182579414905e-06,
+      "loss": 0.1026,
+      "step": 956
+    },
+    {
+      "epoch": 0.1829827915869981,
+      "grad_norm": 2.5066792964935303,
+      "learning_rate": 4.361282558251054e-06,
+      "loss": 0.2583,
+      "step": 957
+    },
+    {
+      "epoch": 0.1831739961759082,
+      "grad_norm": 3.5391042232513428,
+      "learning_rate": 4.361946164774267e-06,
+      "loss": 0.5604,
+      "step": 958
+    },
+    {
+      "epoch": 0.18336520076481835,
+      "grad_norm": 2.347865104675293,
+      "learning_rate": 4.36260907895878e-06,
+      "loss": 0.2277,
+      "step": 959
+    },
+    {
+      "epoch": 0.1835564053537285,
+      "grad_norm": 2.174129009246826,
+      "learning_rate": 4.363271302247718e-06,
+      "loss": 0.1745,
+      "step": 960
+    },
+    {
+      "epoch": 0.1837476099426386,
+      "grad_norm": 1.4797195196151733,
+      "learning_rate": 4.3639328360796964e-06,
+      "loss": 0.1074,
+      "step": 961
+    },
+    {
+      "epoch": 0.18393881453154876,
+      "grad_norm": 2.326012134552002,
+      "learning_rate": 4.364593681888845e-06,
+      "loss": 0.1232,
+      "step": 962
+    },
+    {
+      "epoch": 0.1841300191204589,
+      "grad_norm": 2.7982571125030518,
+      "learning_rate": 4.365253841104822e-06,
+      "loss": 0.5077,
+      "step": 963
+    },
+    {
+      "epoch": 0.18432122370936901,
+      "grad_norm": 3.0020434856414795,
+      "learning_rate": 4.365913315152833e-06,
+      "loss": 0.5043,
+      "step": 964
+    },
+    {
+      "epoch": 0.18451242829827916,
+      "grad_norm": 4.000023365020752,
+      "learning_rate": 4.366572105453653e-06,
+      "loss": 0.3364,
+      "step": 965
+    },
+    {
+      "epoch": 0.1847036328871893,
+      "grad_norm": 1.9444904327392578,
+      "learning_rate": 4.367230213423641e-06,
+      "loss": 0.1445,
+      "step": 966
+    },
+    {
+      "epoch": 0.18489483747609942,
+      "grad_norm": 1.875847578048706,
+      "learning_rate": 4.367887640474758e-06,
+      "loss": 0.2279,
+      "step": 967
+    },
+    {
+      "epoch": 0.18508604206500956,
+      "grad_norm": 1.9915648698806763,
+      "learning_rate": 4.368544388014591e-06,
+      "loss": 0.1222,
+      "step": 968
+    },
+    {
+      "epoch": 0.1852772466539197,
+      "grad_norm": 2.044407844543457,
+      "learning_rate": 4.369200457446364e-06,
+      "loss": 0.1119,
+      "step": 969
+    },
+    {
+      "epoch": 0.18546845124282982,
+      "grad_norm": 1.7450714111328125,
+      "learning_rate": 4.3698558501689585e-06,
+      "loss": 0.1761,
+      "step": 970
+    },
+    {
+      "epoch": 0.18565965583173996,
+      "grad_norm": 1.9846551418304443,
+      "learning_rate": 4.370510567576933e-06,
+      "loss": 0.2268,
+      "step": 971
+    },
+    {
+      "epoch": 0.1858508604206501,
+      "grad_norm": 2.735543727874756,
+      "learning_rate": 4.3711646110605385e-06,
+      "loss": 0.2611,
+      "step": 972
+    },
+    {
+      "epoch": 0.18604206500956022,
+      "grad_norm": 2.1292803287506104,
+      "learning_rate": 4.371817982005741e-06,
+      "loss": 0.1536,
+      "step": 973
+    },
+    {
+      "epoch": 0.18623326959847036,
+      "grad_norm": 2.738206386566162,
+      "learning_rate": 4.372470681794229e-06,
+      "loss": 0.1698,
+      "step": 974
+    },
+    {
+      "epoch": 0.1864244741873805,
+      "grad_norm": 3.367114543914795,
+      "learning_rate": 4.373122711803442e-06,
+      "loss": 0.1996,
+      "step": 975
+    },
+    {
+      "epoch": 0.18661567877629062,
+      "grad_norm": 3.1926040649414062,
+      "learning_rate": 4.373774073406583e-06,
+      "loss": 0.3569,
+      "step": 976
+    },
+    {
+      "epoch": 0.18680688336520077,
+      "grad_norm": 3.1637473106384277,
+      "learning_rate": 4.374424767972636e-06,
+      "loss": 0.228,
+      "step": 977
+    },
+    {
+      "epoch": 0.1869980879541109,
+      "grad_norm": 3.3024322986602783,
+      "learning_rate": 4.375074796866383e-06,
+      "loss": 0.5118,
+      "step": 978
+    },
+    {
+      "epoch": 0.18718929254302102,
+      "grad_norm": 3.079582929611206,
+      "learning_rate": 4.375724161448419e-06,
+      "loss": 0.1612,
+      "step": 979
+    },
+    {
+      "epoch": 0.18738049713193117,
+      "grad_norm": 1.2228554487228394,
+      "learning_rate": 4.376372863075179e-06,
+      "loss": 0.0437,
+      "step": 980
+    },
+    {
+      "epoch": 0.1875717017208413,
+      "grad_norm": 3.335177183151245,
+      "learning_rate": 4.377020903098941e-06,
+      "loss": 0.169,
+      "step": 981
+    },
+    {
+      "epoch": 0.18776290630975143,
+      "grad_norm": 2.6680805683135986,
+      "learning_rate": 4.377668282867852e-06,
+      "loss": 0.3572,
+      "step": 982
+    },
+    {
+      "epoch": 0.18795411089866157,
+      "grad_norm": 2.7457680702209473,
+      "learning_rate": 4.378315003725942e-06,
+      "loss": 0.6047,
+      "step": 983
+    },
+    {
+      "epoch": 0.1881453154875717,
+      "grad_norm": 2.021188974380493,
+      "learning_rate": 4.3789610670131445e-06,
+      "loss": 0.2667,
+      "step": 984
+    },
+    {
+      "epoch": 0.18833652007648183,
+      "grad_norm": 2.8314571380615234,
+      "learning_rate": 4.3796064740653045e-06,
+      "loss": 0.4146,
+      "step": 985
+    },
+    {
+      "epoch": 0.18852772466539197,
+      "grad_norm": 1.4372998476028442,
+      "learning_rate": 4.380251226214205e-06,
+      "loss": 0.1061,
+      "step": 986
+    },
+    {
+      "epoch": 0.18871892925430211,
+      "grad_norm": 3.8322417736053467,
+      "learning_rate": 4.380895324787575e-06,
+      "loss": 0.2334,
+      "step": 987
+    },
+    {
+      "epoch": 0.18891013384321223,
+      "grad_norm": 1.867491602897644,
+      "learning_rate": 4.381538771109115e-06,
+      "loss": 0.1069,
+      "step": 988
+    },
+    {
+      "epoch": 0.18910133843212237,
+      "grad_norm": 2.0068299770355225,
+      "learning_rate": 4.382181566498504e-06,
+      "loss": 0.1687,
+      "step": 989
+    },
+    {
+      "epoch": 0.18929254302103252,
+      "grad_norm": 2.4695441722869873,
+      "learning_rate": 4.382823712271419e-06,
+      "loss": 0.2867,
+      "step": 990
+    },
+    {
+      "epoch": 0.18948374760994263,
+      "grad_norm": 3.3172388076782227,
+      "learning_rate": 4.383465209739559e-06,
+      "loss": 0.2976,
+      "step": 991
+    },
+    {
+      "epoch": 0.18967495219885278,
+      "grad_norm": 1.749539852142334,
+      "learning_rate": 4.384106060210646e-06,
+      "loss": 0.1326,
+      "step": 992
+    },
+    {
+      "epoch": 0.18986615678776292,
+      "grad_norm": 2.1479973793029785,
+      "learning_rate": 4.384746264988454e-06,
+      "loss": 0.1488,
+      "step": 993
+    },
+    {
+      "epoch": 0.19005736137667303,
+      "grad_norm": 2.0028390884399414,
+      "learning_rate": 4.385385825372819e-06,
+      "loss": 0.1353,
+      "step": 994
+    },
+    {
+      "epoch": 0.19024856596558318,
+      "grad_norm": 2.371243953704834,
+      "learning_rate": 4.386024742659655e-06,
+      "loss": 0.2165,
+      "step": 995
+    },
+    {
+      "epoch": 0.19043977055449332,
+      "grad_norm": 1.8113404512405396,
+      "learning_rate": 4.386663018140972e-06,
+      "loss": 0.2138,
+      "step": 996
+    },
+    {
+      "epoch": 0.19063097514340344,
+      "grad_norm": 2.1039295196533203,
+      "learning_rate": 4.38730065310489e-06,
+      "loss": 0.1077,
+      "step": 997
+    },
+    {
+      "epoch": 0.19082217973231358,
+      "grad_norm": 1.7272989749908447,
+      "learning_rate": 4.387937648835653e-06,
+      "loss": 0.1083,
+      "step": 998
+    },
+    {
+      "epoch": 0.19101338432122372,
+      "grad_norm": 3.76637601852417,
+      "learning_rate": 4.388574006613651e-06,
+      "loss": 0.2184,
+      "step": 999
+    },
+    {
+      "epoch": 0.19120458891013384,
+      "grad_norm": 2.589611053466797,
+      "learning_rate": 4.389209727715426e-06,
+      "loss": 0.111,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19120458891013384,
+      "eval_runtime": 801.9588,
+      "eval_samples_per_second": 1.913,
+      "eval_steps_per_second": 0.239,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19139579349904398,
+      "grad_norm": 2.5449726581573486,
+      "learning_rate": 4.389844813413695e-06,
+      "loss": 0.2094,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19158699808795412,
+      "grad_norm": 2.380741834640503,
+      "learning_rate": 4.390479264977362e-06,
+      "loss": 0.2957,
+      "step": 1002
+    },
+    {
+      "epoch": 0.19177820267686424,
+      "grad_norm": 2.056061029434204,
+      "learning_rate": 4.391113083671535e-06,
+      "loss": 0.3615,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19196940726577438,
+      "grad_norm": 2.0398001670837402,
+      "learning_rate": 4.391746270757536e-06,
+      "loss": 0.1292,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1921606118546845,
+      "grad_norm": 4.531376838684082,
+      "learning_rate": 4.392378827492925e-06,
+      "loss": 0.1914,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19235181644359464,
+      "grad_norm": 2.163278102874756,
+      "learning_rate": 4.393010755131506e-06,
+      "loss": 0.0862,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19254302103250479,
+      "grad_norm": 2.507943868637085,
+      "learning_rate": 4.393642054923349e-06,
+      "loss": 0.4044,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1927342256214149,
+      "grad_norm": 3.3545026779174805,
+      "learning_rate": 4.394272728114798e-06,
+      "loss": 0.4105,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19292543021032504,
+      "grad_norm": 3.325700044631958,
+      "learning_rate": 4.3949027759484944e-06,
+      "loss": 0.3183,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1931166347992352,
+      "grad_norm": 1.7217702865600586,
+      "learning_rate": 4.395532199663382e-06,
+      "loss": 0.1855,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1933078393881453,
+      "grad_norm": 1.5247466564178467,
+      "learning_rate": 4.396161000494729e-06,
+      "loss": 0.0998,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19349904397705545,
+      "grad_norm": 2.8211965560913086,
+      "learning_rate": 4.396789179674141e-06,
+      "loss": 0.0913,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1936902485659656,
+      "grad_norm": 3.427171468734741,
+      "learning_rate": 4.397416738429569e-06,
+      "loss": 0.1638,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1938814531548757,
+      "grad_norm": 2.2865030765533447,
+      "learning_rate": 4.398043677985337e-06,
+      "loss": 0.1951,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19407265774378585,
+      "grad_norm": 1.854646921157837,
+      "learning_rate": 4.398669999562141e-06,
+      "loss": 0.1315,
+      "step": 1015
+    },
+    {
+      "epoch": 0.194263862332696,
+      "grad_norm": 2.7197680473327637,
+      "learning_rate": 4.399295704377077e-06,
+      "loss": 0.3981,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1944550669216061,
+      "grad_norm": 1.8270578384399414,
+      "learning_rate": 4.399920793643645e-06,
+      "loss": 0.1379,
+      "step": 1017
+    },
+    {
+      "epoch": 0.19464627151051625,
+      "grad_norm": 3.02530574798584,
+      "learning_rate": 4.400545268571768e-06,
+      "loss": 0.2511,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1948374760994264,
+      "grad_norm": 2.5796334743499756,
+      "learning_rate": 4.401169130367807e-06,
+      "loss": 0.1854,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1950286806883365,
+      "grad_norm": 3.3476648330688477,
+      "learning_rate": 4.4017923802345695e-06,
+      "loss": 0.3143,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19521988527724665,
+      "grad_norm": 2.5378341674804688,
+      "learning_rate": 4.4024150193713316e-06,
+      "loss": 0.4362,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1954110898661568,
+      "grad_norm": 1.8505767583847046,
+      "learning_rate": 4.403037048973844e-06,
+      "loss": 0.2285,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1956022944550669,
+      "grad_norm": 2.6891872882843018,
+      "learning_rate": 4.403658470234348e-06,
+      "loss": 0.2091,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19579349904397705,
+      "grad_norm": 2.9312164783477783,
+      "learning_rate": 4.404279284341596e-06,
+      "loss": 0.2019,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1959847036328872,
+      "grad_norm": 2.8806815147399902,
+      "learning_rate": 4.404899492480853e-06,
+      "loss": 0.1688,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1961759082217973,
+      "grad_norm": 3.487785577774048,
+      "learning_rate": 4.405519095833921e-06,
+      "loss": 0.502,
+      "step": 1026
+    },
+    {
+      "epoch": 0.19636711281070746,
+      "grad_norm": 2.5800065994262695,
+      "learning_rate": 4.406138095579144e-06,
+      "loss": 0.3615,
+      "step": 1027
+    },
+    {
+      "epoch": 0.1965583173996176,
+      "grad_norm": 2.3235068321228027,
+      "learning_rate": 4.406756492891431e-06,
+      "loss": 0.2984,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19674952198852771,
+      "grad_norm": 4.1836628913879395,
+      "learning_rate": 4.407374288942261e-06,
+      "loss": 0.3625,
+      "step": 1029
+    },
+    {
+      "epoch": 0.19694072657743786,
+      "grad_norm": 3.1203243732452393,
+      "learning_rate": 4.407991484899696e-06,
+      "loss": 0.1182,
+      "step": 1030
+    },
+    {
+      "epoch": 0.197131931166348,
+      "grad_norm": 1.4133257865905762,
+      "learning_rate": 4.408608081928404e-06,
+      "loss": 0.0475,
+      "step": 1031
+    },
+    {
+      "epoch": 0.19732313575525812,
+      "grad_norm": 2.494713068008423,
+      "learning_rate": 4.4092240811896605e-06,
+      "loss": 0.2211,
+      "step": 1032
+    },
+    {
+      "epoch": 0.19751434034416826,
+      "grad_norm": 1.6299430131912231,
+      "learning_rate": 4.40983948384137e-06,
+      "loss": 0.1825,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1977055449330784,
+      "grad_norm": 3.060887575149536,
+      "learning_rate": 4.410454291038076e-06,
+      "loss": 0.307,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19789674952198852,
+      "grad_norm": 2.6344666481018066,
+      "learning_rate": 4.411068503930969e-06,
+      "loss": 0.2975,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19808795411089866,
+      "grad_norm": 1.8494610786437988,
+      "learning_rate": 4.41168212366791e-06,
+      "loss": 0.0681,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1982791586998088,
+      "grad_norm": 2.327357530593872,
+      "learning_rate": 4.412295151393437e-06,
+      "loss": 0.1178,
+      "step": 1037
+    },
+    {
+      "epoch": 0.19847036328871892,
+      "grad_norm": 1.8669428825378418,
+      "learning_rate": 4.412907588248774e-06,
+      "loss": 0.2461,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19866156787762906,
+      "grad_norm": 3.062044143676758,
+      "learning_rate": 4.413519435371853e-06,
+      "loss": 0.468,
+      "step": 1039
+    },
+    {
+      "epoch": 0.1988527724665392,
+      "grad_norm": 3.0303285121917725,
+      "learning_rate": 4.4141306938973205e-06,
+      "loss": 0.3054,
+      "step": 1040
+    },
+    {
+      "epoch": 0.19904397705544932,
+      "grad_norm": 2.5444071292877197,
+      "learning_rate": 4.414741364956551e-06,
+      "loss": 0.1862,
+      "step": 1041
+    },
+    {
+      "epoch": 0.19923518164435947,
+      "grad_norm": 2.047647476196289,
+      "learning_rate": 4.415351449677661e-06,
+      "loss": 0.2525,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1994263862332696,
+      "grad_norm": 3.2048020362854004,
+      "learning_rate": 4.415960949185519e-06,
+      "loss": 0.1087,
+      "step": 1043
+    },
+    {
+      "epoch": 0.19961759082217972,
+      "grad_norm": 1.5147563219070435,
+      "learning_rate": 4.416569864601761e-06,
+      "loss": 0.0919,
+      "step": 1044
+    },
+    {
+      "epoch": 0.19980879541108987,
+      "grad_norm": 2.4613304138183594,
+      "learning_rate": 4.4171781970448015e-06,
+      "loss": 0.4064,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.334059953689575,
+      "learning_rate": 4.417785947629845e-06,
+      "loss": 0.2635,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20019120458891013,
+      "grad_norm": 1.446562647819519,
+      "learning_rate": 4.418393117468899e-06,
+      "loss": 0.1417,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20038240917782027,
+      "grad_norm": 3.4641342163085938,
+      "learning_rate": 4.418999707670787e-06,
+      "loss": 0.3766,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2005736137667304,
+      "grad_norm": 1.2733123302459717,
+      "learning_rate": 4.4196057193411596e-06,
+      "loss": 0.0523,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20076481835564053,
+      "grad_norm": 1.9554495811462402,
+      "learning_rate": 4.420211153582507e-06,
+      "loss": 0.1459,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20095602294455067,
+      "grad_norm": 3.5012214183807373,
+      "learning_rate": 4.4208160114941716e-06,
+      "loss": 0.4968,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20114722753346082,
+      "grad_norm": 2.0732827186584473,
+      "learning_rate": 4.421420294172356e-06,
+      "loss": 0.318,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20133843212237093,
+      "grad_norm": 2.6593754291534424,
+      "learning_rate": 4.422024002710142e-06,
+      "loss": 0.2944,
+      "step": 1053
+    },
+    {
+      "epoch": 0.20152963671128107,
+      "grad_norm": 2.268272876739502,
+      "learning_rate": 4.422627138197499e-06,
+      "loss": 0.2813,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20172084130019122,
+      "grad_norm": 2.346163034439087,
+      "learning_rate": 4.423229701721291e-06,
+      "loss": 0.1823,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20191204588910133,
+      "grad_norm": 1.8060640096664429,
+      "learning_rate": 4.423831694365298e-06,
+      "loss": 0.1057,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20210325047801148,
+      "grad_norm": 2.9678351879119873,
+      "learning_rate": 4.424433117210219e-06,
+      "loss": 0.4058,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20229445506692162,
+      "grad_norm": 2.440826416015625,
+      "learning_rate": 4.42503397133369e-06,
+      "loss": 0.314,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20248565965583173,
+      "grad_norm": 2.8080196380615234,
+      "learning_rate": 4.425634257810289e-06,
+      "loss": 0.6267,
+      "step": 1059
+    },
+    {
+      "epoch": 0.20267686424474188,
+      "grad_norm": 3.3547277450561523,
+      "learning_rate": 4.426233977711554e-06,
+      "loss": 0.2995,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20286806883365202,
+      "grad_norm": 1.7534793615341187,
+      "learning_rate": 4.426833132105993e-06,
+      "loss": 0.2257,
+      "step": 1061
+    },
+    {
+      "epoch": 0.20305927342256214,
+      "grad_norm": 1.749898910522461,
+      "learning_rate": 4.4274317220590905e-06,
+      "loss": 0.1218,
+      "step": 1062
+    },
+    {
+      "epoch": 0.20325047801147228,
+      "grad_norm": 1.7936410903930664,
+      "learning_rate": 4.428029748633326e-06,
+      "loss": 0.1347,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20344168260038242,
+      "grad_norm": 3.712153196334839,
+      "learning_rate": 4.42862721288818e-06,
+      "loss": 0.4845,
+      "step": 1064
+    },
+    {
+      "epoch": 0.20363288718929254,
+      "grad_norm": 3.04533052444458,
+      "learning_rate": 4.429224115880146e-06,
+      "loss": 0.3209,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20382409177820268,
+      "grad_norm": 1.8066027164459229,
+      "learning_rate": 4.429820458662747e-06,
+      "loss": 0.1314,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2040152963671128,
+      "grad_norm": 1.466841697692871,
+      "learning_rate": 4.43041624228654e-06,
+      "loss": 0.0976,
+      "step": 1067
+    },
+    {
+      "epoch": 0.20420650095602294,
+      "grad_norm": 2.905015468597412,
+      "learning_rate": 4.431011467799127e-06,
+      "loss": 0.273,
+      "step": 1068
+    },
+    {
+      "epoch": 0.20439770554493308,
+      "grad_norm": 2.144706964492798,
+      "learning_rate": 4.4316061362451714e-06,
+      "loss": 0.1751,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2045889101338432,
+      "grad_norm": 2.6781883239746094,
+      "learning_rate": 4.432200248666409e-06,
+      "loss": 0.535,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20478011472275334,
+      "grad_norm": 2.1494102478027344,
+      "learning_rate": 4.4327938061016515e-06,
+      "loss": 0.1759,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20497131931166349,
+      "grad_norm": 2.51594877243042,
+      "learning_rate": 4.433386809586803e-06,
+      "loss": 0.1144,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2051625239005736,
+      "grad_norm": 1.9181592464447021,
+      "learning_rate": 4.433979260154872e-06,
+      "loss": 0.1752,
+      "step": 1073
+    },
+    {
+      "epoch": 0.20535372848948374,
+      "grad_norm": 2.254838466644287,
+      "learning_rate": 4.4345711588359805e-06,
+      "loss": 0.0968,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2055449330783939,
+      "grad_norm": 2.161107063293457,
+      "learning_rate": 4.435162506657369e-06,
+      "loss": 0.0813,
+      "step": 1075
+    },
+    {
+      "epoch": 0.205736137667304,
+      "grad_norm": 2.8448431491851807,
+      "learning_rate": 4.435753304643421e-06,
+      "loss": 0.6405,
+      "step": 1076
+    },
+    {
+      "epoch": 0.20592734225621415,
+      "grad_norm": 1.856491208076477,
+      "learning_rate": 4.436343553815657e-06,
+      "loss": 0.1512,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2061185468451243,
+      "grad_norm": 2.017235517501831,
+      "learning_rate": 4.43693325519276e-06,
+      "loss": 0.1791,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2063097514340344,
+      "grad_norm": 2.606391429901123,
+      "learning_rate": 4.437522409790575e-06,
+      "loss": 0.2601,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20650095602294455,
+      "grad_norm": 2.5404584407806396,
+      "learning_rate": 4.438111018622127e-06,
+      "loss": 0.2392,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2066921606118547,
+      "grad_norm": 2.630312919616699,
+      "learning_rate": 4.438699082697624e-06,
+      "loss": 0.1322,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2068833652007648,
+      "grad_norm": 2.1289303302764893,
+      "learning_rate": 4.4392866030244765e-06,
+      "loss": 0.2648,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20707456978967495,
+      "grad_norm": 3.3554677963256836,
+      "learning_rate": 4.439873580607301e-06,
+      "loss": 0.5565,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2072657743785851,
+      "grad_norm": 2.1717491149902344,
+      "learning_rate": 4.440460016447934e-06,
+      "loss": 0.1442,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2074569789674952,
+      "grad_norm": 2.714172601699829,
+      "learning_rate": 4.441045911545436e-06,
+      "loss": 0.3317,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20764818355640535,
+      "grad_norm": 2.02815318107605,
+      "learning_rate": 4.441631266896111e-06,
+      "loss": 0.1441,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2078393881453155,
+      "grad_norm": 1.8019261360168457,
+      "learning_rate": 4.442216083493512e-06,
+      "loss": 0.0926,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2080305927342256,
+      "grad_norm": 1.846985101699829,
+      "learning_rate": 4.442800362328448e-06,
+      "loss": 0.1832,
+      "step": 1088
+    },
+    {
+      "epoch": 0.20822179732313575,
+      "grad_norm": 2.089106321334839,
+      "learning_rate": 4.443384104389001e-06,
+      "loss": 0.1436,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2084130019120459,
+      "grad_norm": 2.037452459335327,
+      "learning_rate": 4.443967310660528e-06,
+      "loss": 0.1578,
+      "step": 1090
+    },
+    {
+      "epoch": 0.208604206500956,
+      "grad_norm": 2.5428967475891113,
+      "learning_rate": 4.444549982125679e-06,
+      "loss": 0.5048,
+      "step": 1091
+    },
+    {
+      "epoch": 0.20879541108986616,
+      "grad_norm": 1.7422195672988892,
+      "learning_rate": 4.445132119764402e-06,
+      "loss": 0.1375,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2089866156787763,
+      "grad_norm": 1.4205963611602783,
+      "learning_rate": 4.445713724553953e-06,
+      "loss": 0.0716,
+      "step": 1093
+    },
+    {
+      "epoch": 0.20917782026768642,
+      "grad_norm": 2.83103346824646,
+      "learning_rate": 4.446294797468908e-06,
+      "loss": 0.1791,
+      "step": 1094
+    },
+    {
+      "epoch": 0.20936902485659656,
+      "grad_norm": 2.2832252979278564,
+      "learning_rate": 4.446875339481172e-06,
+      "loss": 0.2271,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2095602294455067,
+      "grad_norm": 2.6716253757476807,
+      "learning_rate": 4.447455351559987e-06,
+      "loss": 0.4096,
+      "step": 1096
+    },
+    {
+      "epoch": 0.20975143403441682,
+      "grad_norm": 1.877040147781372,
+      "learning_rate": 4.448034834671944e-06,
+      "loss": 0.27,
+      "step": 1097
+    },
+    {
+      "epoch": 0.20994263862332696,
+      "grad_norm": 1.4391602277755737,
+      "learning_rate": 4.448613789780993e-06,
+      "loss": 0.076,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2101338432122371,
+      "grad_norm": 1.844596266746521,
+      "learning_rate": 4.449192217848449e-06,
+      "loss": 0.1119,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21032504780114722,
+      "grad_norm": 4.228497505187988,
+      "learning_rate": 4.449770119833007e-06,
+      "loss": 0.1945,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21051625239005736,
+      "grad_norm": 2.4177355766296387,
+      "learning_rate": 4.450347496690746e-06,
+      "loss": 0.3176,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2107074569789675,
+      "grad_norm": 3.1131842136383057,
+      "learning_rate": 4.450924349375142e-06,
+      "loss": 0.3662,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21089866156787762,
+      "grad_norm": 1.61544668674469,
+      "learning_rate": 4.4515006788370775e-06,
+      "loss": 0.2676,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21108986615678776,
+      "grad_norm": 4.033684253692627,
+      "learning_rate": 4.4520764860248474e-06,
+      "loss": 0.4563,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2112810707456979,
+      "grad_norm": 2.592637062072754,
+      "learning_rate": 4.452651771884173e-06,
+      "loss": 0.3416,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21147227533460802,
+      "grad_norm": 1.7672158479690552,
+      "learning_rate": 4.45322653735821e-06,
+      "loss": 0.1088,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21166347992351817,
+      "grad_norm": 1.9779208898544312,
+      "learning_rate": 4.453800783387554e-06,
+      "loss": 0.2377,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2118546845124283,
+      "grad_norm": 2.543431043624878,
+      "learning_rate": 4.454374510910254e-06,
+      "loss": 0.2436,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21204588910133843,
+      "grad_norm": 4.803256988525391,
+      "learning_rate": 4.454947720861822e-06,
+      "loss": 0.3721,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21223709369024857,
+      "grad_norm": 2.2504935264587402,
+      "learning_rate": 4.455520414175238e-06,
+      "loss": 0.2385,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2124282982791587,
+      "grad_norm": 2.656966209411621,
+      "learning_rate": 4.456092591780964e-06,
+      "loss": 0.1749,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21261950286806883,
+      "grad_norm": 2.746769666671753,
+      "learning_rate": 4.4566642546069476e-06,
+      "loss": 0.109,
+      "step": 1112
+    },
+    {
+      "epoch": 0.21281070745697897,
+      "grad_norm": 3.2161478996276855,
+      "learning_rate": 4.457235403578636e-06,
+      "loss": 0.2271,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2130019120458891,
+      "grad_norm": 4.615819931030273,
+      "learning_rate": 4.457806039618983e-06,
+      "loss": 0.5054,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21319311663479923,
+      "grad_norm": 2.4721786975860596,
+      "learning_rate": 4.458376163648458e-06,
+      "loss": 0.2637,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21338432122370937,
+      "grad_norm": 1.3068219423294067,
+      "learning_rate": 4.458945776585056e-06,
+      "loss": 0.1306,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21357552581261952,
+      "grad_norm": 1.7614368200302124,
+      "learning_rate": 4.459514879344301e-06,
+      "loss": 0.0854,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21376673040152963,
+      "grad_norm": 2.025942325592041,
+      "learning_rate": 4.460083472839265e-06,
+      "loss": 0.1001,
+      "step": 1118
+    },
+    {
+      "epoch": 0.21395793499043977,
+      "grad_norm": 1.3237824440002441,
+      "learning_rate": 4.460651557980566e-06,
+      "loss": 0.091,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21414913957934992,
+      "grad_norm": 3.109684705734253,
+      "learning_rate": 4.461219135676386e-06,
+      "loss": 0.6804,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21434034416826003,
+      "grad_norm": 3.659759044647217,
+      "learning_rate": 4.461786206832473e-06,
+      "loss": 0.4264,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21453154875717018,
+      "grad_norm": 3.0191221237182617,
+      "learning_rate": 4.462352772352151e-06,
+      "loss": 0.2061,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21472275334608032,
+      "grad_norm": 0.91303551197052,
+      "learning_rate": 4.4629188331363334e-06,
+      "loss": 0.1592,
+      "step": 1123
+    },
+    {
+      "epoch": 0.21491395793499043,
+      "grad_norm": 2.4983134269714355,
+      "learning_rate": 4.463484390083525e-06,
+      "loss": 0.1401,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21510516252390058,
+      "grad_norm": 2.4315645694732666,
+      "learning_rate": 4.464049444089835e-06,
+      "loss": 0.1035,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2152963671128107,
+      "grad_norm": 2.098078489303589,
+      "learning_rate": 4.464613996048983e-06,
+      "loss": 0.2432,
+      "step": 1126
+    },
+    {
+      "epoch": 0.21548757170172084,
+      "grad_norm": 2.7613978385925293,
+      "learning_rate": 4.465178046852309e-06,
+      "loss": 0.401,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21567877629063098,
+      "grad_norm": 2.0903496742248535,
+      "learning_rate": 4.465741597388782e-06,
+      "loss": 0.1743,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2158699808795411,
+      "grad_norm": 2.099884271621704,
+      "learning_rate": 4.466304648545006e-06,
+      "loss": 0.1594,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21606118546845124,
+      "grad_norm": 1.7775846719741821,
+      "learning_rate": 4.466867201205232e-06,
+      "loss": 0.0927,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21625239005736138,
+      "grad_norm": 1.9474855661392212,
+      "learning_rate": 4.467429256251365e-06,
+      "loss": 0.1128,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2164435946462715,
+      "grad_norm": 2.1535003185272217,
+      "learning_rate": 4.467990814562967e-06,
+      "loss": 0.2644,
+      "step": 1132
+    },
+    {
+      "epoch": 0.21663479923518164,
+      "grad_norm": 3.28275203704834,
+      "learning_rate": 4.4685518770172775e-06,
+      "loss": 0.4298,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21682600382409178,
+      "grad_norm": 2.098271608352661,
+      "learning_rate": 4.469112444489207e-06,
+      "loss": 0.1012,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2170172084130019,
+      "grad_norm": 2.1434457302093506,
+      "learning_rate": 4.469672517851359e-06,
+      "loss": 0.2505,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21720841300191204,
+      "grad_norm": 2.248441219329834,
+      "learning_rate": 4.470232097974025e-06,
+      "loss": 0.2486,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2173996175908222,
+      "grad_norm": 1.4229638576507568,
+      "learning_rate": 4.470791185725206e-06,
+      "loss": 0.0635,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2175908221797323,
+      "grad_norm": 2.0013930797576904,
+      "learning_rate": 4.4713497819706065e-06,
+      "loss": 0.1122,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21778202676864244,
+      "grad_norm": 2.527571678161621,
+      "learning_rate": 4.471907887573656e-06,
+      "loss": 0.3839,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2179732313575526,
+      "grad_norm": 3.140901565551758,
+      "learning_rate": 4.472465503395508e-06,
+      "loss": 0.5159,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2181644359464627,
+      "grad_norm": 3.4777331352233887,
+      "learning_rate": 4.473022630295051e-06,
+      "loss": 0.2159,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21835564053537285,
+      "grad_norm": 1.6637810468673706,
+      "learning_rate": 4.473579269128917e-06,
+      "loss": 0.2454,
+      "step": 1142
+    },
+    {
+      "epoch": 0.218546845124283,
+      "grad_norm": 3.4433083534240723,
+      "learning_rate": 4.474135420751486e-06,
+      "loss": 0.2348,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2187380497131931,
+      "grad_norm": 2.1617848873138428,
+      "learning_rate": 4.474691086014902e-06,
+      "loss": 0.2981,
+      "step": 1144
+    },
+    {
+      "epoch": 0.21892925430210325,
+      "grad_norm": 2.460859775543213,
+      "learning_rate": 4.475246265769069e-06,
+      "loss": 0.4682,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2191204588910134,
+      "grad_norm": 1.9391944408416748,
+      "learning_rate": 4.47580096086167e-06,
+      "loss": 0.2622,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2193116634799235,
+      "grad_norm": 2.875788927078247,
+      "learning_rate": 4.476355172138167e-06,
+      "loss": 0.4374,
+      "step": 1147
+    },
+    {
+      "epoch": 0.21950286806883365,
+      "grad_norm": 2.8496835231781006,
+      "learning_rate": 4.476908900441812e-06,
+      "loss": 0.297,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2196940726577438,
+      "grad_norm": 2.6283702850341797,
+      "learning_rate": 4.477462146613657e-06,
+      "loss": 0.1253,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2198852772466539,
+      "grad_norm": 1.8835588693618774,
+      "learning_rate": 4.478014911492556e-06,
+      "loss": 0.1859,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22007648183556405,
+      "grad_norm": 13.009642601013184,
+      "learning_rate": 4.478567195915176e-06,
+      "loss": 0.3384,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2202676864244742,
+      "grad_norm": 0.9838836193084717,
+      "learning_rate": 4.479119000716005e-06,
+      "loss": 0.0698,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2204588910133843,
+      "grad_norm": 2.835535764694214,
+      "learning_rate": 4.479670326727359e-06,
+      "loss": 0.394,
+      "step": 1153
+    },
+    {
+      "epoch": 0.22065009560229445,
+      "grad_norm": 2.8879711627960205,
+      "learning_rate": 4.480221174779389e-06,
+      "loss": 0.2163,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2208413001912046,
+      "grad_norm": 2.8582956790924072,
+      "learning_rate": 4.480771545700088e-06,
+      "loss": 0.0787,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2210325047801147,
+      "grad_norm": 2.854708194732666,
+      "learning_rate": 4.481321440315302e-06,
+      "loss": 0.2055,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22122370936902486,
+      "grad_norm": 1.8422484397888184,
+      "learning_rate": 4.481870859448731e-06,
+      "loss": 0.1446,
+      "step": 1157
+    },
+    {
+      "epoch": 0.221414913957935,
+      "grad_norm": 2.160968542098999,
+      "learning_rate": 4.482419803921941e-06,
+      "loss": 0.2785,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22160611854684512,
+      "grad_norm": 2.0334866046905518,
+      "learning_rate": 4.482968274554374e-06,
+      "loss": 0.1943,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22179732313575526,
+      "grad_norm": 1.5216292142868042,
+      "learning_rate": 4.483516272163348e-06,
+      "loss": 0.0838,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2219885277246654,
+      "grad_norm": 3.173959970474243,
+      "learning_rate": 4.4840637975640696e-06,
+      "loss": 0.272,
+      "step": 1161
+    },
+    {
+      "epoch": 0.22217973231357552,
+      "grad_norm": 1.9395912885665894,
+      "learning_rate": 4.4846108515696406e-06,
+      "loss": 0.2258,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22237093690248566,
+      "grad_norm": 1.6804591417312622,
+      "learning_rate": 4.485157434991062e-06,
+      "loss": 0.0927,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2225621414913958,
+      "grad_norm": 2.5178074836730957,
+      "learning_rate": 4.485703548637246e-06,
+      "loss": 0.2079,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22275334608030592,
+      "grad_norm": 2.004671096801758,
+      "learning_rate": 4.48624919331502e-06,
+      "loss": 0.232,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22294455066921606,
+      "grad_norm": 1.9772891998291016,
+      "learning_rate": 4.486794369829135e-06,
+      "loss": 0.1463,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2231357552581262,
+      "grad_norm": 2.044581174850464,
+      "learning_rate": 4.487339078982273e-06,
+      "loss": 0.1012,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22332695984703632,
+      "grad_norm": 2.186039924621582,
+      "learning_rate": 4.48788332157505e-06,
+      "loss": 0.1192,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22351816443594646,
+      "grad_norm": 1.783949851989746,
+      "learning_rate": 4.488427098406031e-06,
+      "loss": 0.0964,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2237093690248566,
+      "grad_norm": 2.8822665214538574,
+      "learning_rate": 4.4889704102717304e-06,
+      "loss": 0.402,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22390057361376672,
+      "grad_norm": 2.142519235610962,
+      "learning_rate": 4.48951325796662e-06,
+      "loss": 0.302,
+      "step": 1171
+    },
+    {
+      "epoch": 0.22409177820267687,
+      "grad_norm": 1.857020616531372,
+      "learning_rate": 4.49005564228314e-06,
+      "loss": 0.3136,
+      "step": 1172
+    },
+    {
+      "epoch": 0.224282982791587,
+      "grad_norm": 2.0371313095092773,
+      "learning_rate": 4.490597564011701e-06,
+      "loss": 0.2399,
+      "step": 1173
+    },
+    {
+      "epoch": 0.22447418738049713,
+      "grad_norm": 2.0099854469299316,
+      "learning_rate": 4.491139023940692e-06,
+      "loss": 0.0889,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22466539196940727,
+      "grad_norm": 2.0622718334198,
+      "learning_rate": 4.49168002285649e-06,
+      "loss": 0.1362,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2248565965583174,
+      "grad_norm": 3.3179798126220703,
+      "learning_rate": 4.492220561543467e-06,
+      "loss": 0.4877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22504780114722753,
+      "grad_norm": 2.5217320919036865,
+      "learning_rate": 4.49276064078399e-06,
+      "loss": 0.5047,
+      "step": 1177
+    },
+    {
+      "epoch": 0.22523900573613767,
+      "grad_norm": 3.4361255168914795,
+      "learning_rate": 4.493300261358436e-06,
+      "loss": 0.5802,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22543021032504781,
+      "grad_norm": 2.50624942779541,
+      "learning_rate": 4.493839424045196e-06,
+      "loss": 0.2938,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22562141491395793,
+      "grad_norm": 2.448352575302124,
+      "learning_rate": 4.494378129620678e-06,
+      "loss": 0.1441,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22581261950286807,
+      "grad_norm": 1.622539758682251,
+      "learning_rate": 4.494916378859321e-06,
+      "loss": 0.092,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22600382409177822,
+      "grad_norm": 2.642896890640259,
+      "learning_rate": 4.495454172533592e-06,
+      "loss": 0.3514,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22619502868068833,
+      "grad_norm": 2.202483892440796,
+      "learning_rate": 4.495991511414005e-06,
+      "loss": 0.2871,
+      "step": 1183
+    },
+    {
+      "epoch": 0.22638623326959847,
+      "grad_norm": 1.2011011838912964,
+      "learning_rate": 4.496528396269117e-06,
+      "loss": 0.0971,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22657743785850862,
+      "grad_norm": 2.026979923248291,
+      "learning_rate": 4.497064827865537e-06,
+      "loss": 0.1101,
+      "step": 1185
+    },
+    {
+      "epoch": 0.22676864244741873,
+      "grad_norm": 1.4967538118362427,
+      "learning_rate": 4.49760080696794e-06,
+      "loss": 0.2007,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22695984703632888,
+      "grad_norm": 2.381471872329712,
+      "learning_rate": 4.498136334339059e-06,
+      "loss": 0.0943,
+      "step": 1187
+    },
+    {
+      "epoch": 0.227151051625239,
+      "grad_norm": 2.089556932449341,
+      "learning_rate": 4.498671410739708e-06,
+      "loss": 0.1357,
+      "step": 1188
+    },
+    {
+      "epoch": 0.22734225621414914,
+      "grad_norm": 3.158884048461914,
+      "learning_rate": 4.4992060369287745e-06,
+      "loss": 0.4473,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22753346080305928,
+      "grad_norm": 1.9269723892211914,
+      "learning_rate": 4.499740213663238e-06,
+      "loss": 0.2718,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2277246653919694,
+      "grad_norm": 3.3276264667510986,
+      "learning_rate": 4.500273941698166e-06,
+      "loss": 0.4174,
+      "step": 1191
+    },
+    {
+      "epoch": 0.22791586998087954,
+      "grad_norm": 1.854248285293579,
+      "learning_rate": 4.500807221786725e-06,
+      "loss": 0.1561,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22810707456978968,
+      "grad_norm": 5.148835182189941,
+      "learning_rate": 4.50134005468019e-06,
+      "loss": 0.3131,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2282982791586998,
+      "grad_norm": 2.1514906883239746,
+      "learning_rate": 4.501872441127943e-06,
+      "loss": 0.1113,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22848948374760994,
+      "grad_norm": 2.2153143882751465,
+      "learning_rate": 4.502404381877488e-06,
+      "loss": 0.2921,
+      "step": 1195
+    },
+    {
+      "epoch": 0.22868068833652008,
+      "grad_norm": 1.5667790174484253,
+      "learning_rate": 4.502935877674451e-06,
+      "loss": 0.1318,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2288718929254302,
+      "grad_norm": 3.2153244018554688,
+      "learning_rate": 4.503466929262589e-06,
+      "loss": 0.5028,
+      "step": 1197
+    },
+    {
+      "epoch": 0.22906309751434034,
+      "grad_norm": 1.3133012056350708,
+      "learning_rate": 4.503997537383796e-06,
+      "loss": 0.1566,
+      "step": 1198
+    },
+    {
+      "epoch": 0.22925430210325048,
+      "grad_norm": 1.9280524253845215,
+      "learning_rate": 4.50452770277811e-06,
+      "loss": 0.2353,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2294455066921606,
+      "grad_norm": 2.680346965789795,
+      "learning_rate": 4.505057426183714e-06,
+      "loss": 0.1869,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22963671128107074,
+      "grad_norm": 3.4344375133514404,
+      "learning_rate": 4.505586708336951e-06,
+      "loss": 0.4891,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2298279158699809,
+      "grad_norm": 1.4899710416793823,
+      "learning_rate": 4.506115549972324e-06,
+      "loss": 0.1304,
+      "step": 1202
+    },
+    {
+      "epoch": 0.230019120458891,
+      "grad_norm": 2.121846914291382,
+      "learning_rate": 4.506643951822501e-06,
+      "loss": 0.1658,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23021032504780115,
+      "grad_norm": 2.3687522411346436,
+      "learning_rate": 4.507171914618329e-06,
+      "loss": 0.3938,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2304015296367113,
+      "grad_norm": 1.6581428050994873,
+      "learning_rate": 4.5076994390888294e-06,
+      "loss": 0.1058,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2305927342256214,
+      "grad_norm": 1.87697434425354,
+      "learning_rate": 4.508226525961212e-06,
+      "loss": 0.1238,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23078393881453155,
+      "grad_norm": 2.034977912902832,
+      "learning_rate": 4.508753175960878e-06,
+      "loss": 0.183,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2309751434034417,
+      "grad_norm": 2.1185624599456787,
+      "learning_rate": 4.509279389811426e-06,
+      "loss": 0.2888,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2311663479923518,
+      "grad_norm": 2.6012215614318848,
+      "learning_rate": 4.509805168234659e-06,
+      "loss": 0.3785,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23135755258126195,
+      "grad_norm": 1.3720933198928833,
+      "learning_rate": 4.510330511950588e-06,
+      "loss": 0.1433,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2315487571701721,
+      "grad_norm": 1.6616960763931274,
+      "learning_rate": 4.510855421677442e-06,
+      "loss": 0.1277,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2317399617590822,
+      "grad_norm": 1.7231650352478027,
+      "learning_rate": 4.511379898131671e-06,
+      "loss": 0.1076,
+      "step": 1212
+    },
+    {
+      "epoch": 0.23193116634799235,
+      "grad_norm": 2.2084357738494873,
+      "learning_rate": 4.511903942027949e-06,
+      "loss": 0.2608,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2321223709369025,
+      "grad_norm": 2.0652458667755127,
+      "learning_rate": 4.512427554079188e-06,
+      "loss": 0.2228,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2323135755258126,
+      "grad_norm": 1.695842981338501,
+      "learning_rate": 4.512950734996536e-06,
+      "loss": 0.1546,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23250478011472275,
+      "grad_norm": 3.044466972351074,
+      "learning_rate": 4.513473485489387e-06,
+      "loss": 0.3004,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2326959847036329,
+      "grad_norm": 2.2261133193969727,
+      "learning_rate": 4.513995806265384e-06,
+      "loss": 0.2103,
+      "step": 1217
+    },
+    {
+      "epoch": 0.232887189292543,
+      "grad_norm": 2.5913712978363037,
+      "learning_rate": 4.51451769803043e-06,
+      "loss": 0.3076,
+      "step": 1218
+    },
+    {
+      "epoch": 0.23307839388145316,
+      "grad_norm": 2.4403324127197266,
+      "learning_rate": 4.515039161488684e-06,
+      "loss": 0.1843,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2332695984703633,
+      "grad_norm": 3.694639205932617,
+      "learning_rate": 4.51556019734258e-06,
+      "loss": 0.3922,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23346080305927341,
+      "grad_norm": 3.163925886154175,
+      "learning_rate": 4.516080806292819e-06,
+      "loss": 0.3171,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23365200764818356,
+      "grad_norm": 2.774327039718628,
+      "learning_rate": 4.516600989038385e-06,
+      "loss": 0.3307,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2338432122370937,
+      "grad_norm": 1.897996425628662,
+      "learning_rate": 4.517120746276545e-06,
+      "loss": 0.1318,
+      "step": 1223
+    },
+    {
+      "epoch": 0.23403441682600382,
+      "grad_norm": 1.8080636262893677,
+      "learning_rate": 4.517640078702858e-06,
+      "loss": 0.089,
+      "step": 1224
+    },
+    {
+      "epoch": 0.23422562141491396,
+      "grad_norm": 1.8752037286758423,
+      "learning_rate": 4.5181589870111755e-06,
+      "loss": 0.1162,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2344168260038241,
+      "grad_norm": 1.7227205038070679,
+      "learning_rate": 4.518677471893654e-06,
+      "loss": 0.316,
+      "step": 1226
+    },
+    {
+      "epoch": 0.23460803059273422,
+      "grad_norm": 2.1087758541107178,
+      "learning_rate": 4.519195534040756e-06,
+      "loss": 0.1585,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23479923518164436,
+      "grad_norm": 2.5245354175567627,
+      "learning_rate": 4.519713174141255e-06,
+      "loss": 0.2401,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2349904397705545,
+      "grad_norm": 2.966695547103882,
+      "learning_rate": 4.520230392882245e-06,
+      "loss": 0.0818,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23518164435946462,
+      "grad_norm": 1.3945024013519287,
+      "learning_rate": 4.520747190949141e-06,
+      "loss": 0.1178,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23537284894837476,
+      "grad_norm": 2.4835667610168457,
+      "learning_rate": 4.5212635690256885e-06,
+      "loss": 0.0745,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2355640535372849,
+      "grad_norm": 3.491046667098999,
+      "learning_rate": 4.521779527793967e-06,
+      "loss": 0.2478,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23575525812619502,
+      "grad_norm": 1.6101934909820557,
+      "learning_rate": 4.522295067934395e-06,
+      "loss": 0.1821,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23594646271510517,
+      "grad_norm": 1.2053178548812866,
+      "learning_rate": 4.522810190125739e-06,
+      "loss": 0.1249,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2361376673040153,
+      "grad_norm": 2.9630486965179443,
+      "learning_rate": 4.523324895045111e-06,
+      "loss": 0.2589,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23632887189292542,
+      "grad_norm": 1.9277501106262207,
+      "learning_rate": 4.523839183367984e-06,
+      "loss": 0.2001,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23652007648183557,
+      "grad_norm": 3.066760540008545,
+      "learning_rate": 4.5243530557681885e-06,
+      "loss": 0.4899,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2367112810707457,
+      "grad_norm": 1.5749143362045288,
+      "learning_rate": 4.524866512917924e-06,
+      "loss": 0.0899,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23690248565965583,
+      "grad_norm": 2.5981080532073975,
+      "learning_rate": 4.525379555487759e-06,
+      "loss": 0.4254,
+      "step": 1239
+    },
+    {
+      "epoch": 0.23709369024856597,
+      "grad_norm": 2.3466320037841797,
+      "learning_rate": 4.525892184146643e-06,
+      "loss": 0.3603,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2372848948374761,
+      "grad_norm": 1.4082063436508179,
+      "learning_rate": 4.526404399561903e-06,
+      "loss": 0.1948,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23747609942638623,
+      "grad_norm": 2.5637047290802,
+      "learning_rate": 4.5269162023992565e-06,
+      "loss": 0.2129,
+      "step": 1242
+    },
+    {
+      "epoch": 0.23766730401529637,
+      "grad_norm": 6.297000885009766,
+      "learning_rate": 4.527427593322813e-06,
+      "loss": 0.1063,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23785850860420651,
+      "grad_norm": 1.5783936977386475,
+      "learning_rate": 4.527938572995081e-06,
+      "loss": 0.0618,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23804971319311663,
+      "grad_norm": 2.314439535140991,
+      "learning_rate": 4.5284491420769685e-06,
+      "loss": 0.2352,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23824091778202677,
+      "grad_norm": 2.707080841064453,
+      "learning_rate": 4.528959301227795e-06,
+      "loss": 0.5317,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2384321223709369,
+      "grad_norm": 3.3706955909729004,
+      "learning_rate": 4.529469051105292e-06,
+      "loss": 0.2381,
+      "step": 1247
+    },
+    {
+      "epoch": 0.23862332695984703,
+      "grad_norm": 2.027712345123291,
+      "learning_rate": 4.5299783923656084e-06,
+      "loss": 0.2689,
+      "step": 1248
+    },
+    {
+      "epoch": 0.23881453154875718,
+      "grad_norm": 2.501119375228882,
+      "learning_rate": 4.530487325663318e-06,
+      "loss": 0.1832,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2390057361376673,
+      "grad_norm": 1.212647795677185,
+      "learning_rate": 4.530995851651422e-06,
+      "loss": 0.0478,
+      "step": 1250
+    },
+    {
+      "epoch": 0.23919694072657743,
+      "grad_norm": 3.2780611515045166,
+      "learning_rate": 4.531503970981357e-06,
+      "loss": 0.5961,
+      "step": 1251
+    },
+    {
+      "epoch": 0.23938814531548758,
+      "grad_norm": 2.3319296836853027,
+      "learning_rate": 4.532011684302993e-06,
+      "loss": 0.1756,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2395793499043977,
+      "grad_norm": 1.8689370155334473,
+      "learning_rate": 4.532518992264648e-06,
+      "loss": 0.0948,
+      "step": 1253
+    },
+    {
+      "epoch": 0.23977055449330784,
+      "grad_norm": 2.4608871936798096,
+      "learning_rate": 4.5330258955130894e-06,
+      "loss": 0.1658,
+      "step": 1254
+    },
+    {
+      "epoch": 0.23996175908221798,
+      "grad_norm": 3.9478836059570312,
+      "learning_rate": 4.533532394693532e-06,
+      "loss": 0.3201,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2401529636711281,
+      "grad_norm": 1.9749811887741089,
+      "learning_rate": 4.534038490449656e-06,
+      "loss": 0.0873,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24034416826003824,
+      "grad_norm": 1.7174595594406128,
+      "learning_rate": 4.534544183423599e-06,
+      "loss": 0.1389,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24053537284894838,
+      "grad_norm": 1.479738712310791,
+      "learning_rate": 4.5350494742559694e-06,
+      "loss": 0.2478,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2407265774378585,
+      "grad_norm": 1.9513295888900757,
+      "learning_rate": 4.535554363585849e-06,
+      "loss": 0.2366,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24091778202676864,
+      "grad_norm": 1.9584410190582275,
+      "learning_rate": 4.536058852050796e-06,
+      "loss": 0.1392,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24110898661567878,
+      "grad_norm": 1.891797423362732,
+      "learning_rate": 4.53656294028685e-06,
+      "loss": 0.0977,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2413001912045889,
+      "grad_norm": 1.9594978094100952,
+      "learning_rate": 4.537066628928541e-06,
+      "loss": 0.1411,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24149139579349904,
+      "grad_norm": 2.9643545150756836,
+      "learning_rate": 4.537569918608891e-06,
+      "loss": 0.2389,
+      "step": 1263
+    },
+    {
+      "epoch": 0.24168260038240919,
+      "grad_norm": 1.6669223308563232,
+      "learning_rate": 4.538072809959417e-06,
+      "loss": 0.1562,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2418738049713193,
+      "grad_norm": 2.446153402328491,
+      "learning_rate": 4.538575303610137e-06,
+      "loss": 0.1492,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24206500956022944,
+      "grad_norm": 1.535170555114746,
+      "learning_rate": 4.539077400189579e-06,
+      "loss": 0.0918,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2422562141491396,
+      "grad_norm": 1.8623381853103638,
+      "learning_rate": 4.53957910032478e-06,
+      "loss": 0.107,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2424474187380497,
+      "grad_norm": 2.7198643684387207,
+      "learning_rate": 4.5400804046412925e-06,
+      "loss": 0.256,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24263862332695985,
+      "grad_norm": 2.15179443359375,
+      "learning_rate": 4.540581313763191e-06,
+      "loss": 0.1535,
+      "step": 1269
+    },
+    {
+      "epoch": 0.24282982791587,
+      "grad_norm": 2.4457833766937256,
+      "learning_rate": 4.541081828313074e-06,
+      "loss": 0.4226,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2430210325047801,
+      "grad_norm": 2.3401358127593994,
+      "learning_rate": 4.54158194891207e-06,
+      "loss": 0.1621,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24321223709369025,
+      "grad_norm": 2.9228503704071045,
+      "learning_rate": 4.542081676179842e-06,
+      "loss": 0.1591,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2434034416826004,
+      "grad_norm": 1.8721615076065063,
+      "learning_rate": 4.542581010734594e-06,
+      "loss": 0.2154,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2435946462715105,
+      "grad_norm": 1.5572518110275269,
+      "learning_rate": 4.54307995319307e-06,
+      "loss": 0.084,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24378585086042065,
+      "grad_norm": 2.8869805335998535,
+      "learning_rate": 4.543578504170567e-06,
+      "loss": 0.1584,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2439770554493308,
+      "grad_norm": 2.540313243865967,
+      "learning_rate": 4.544076664280929e-06,
+      "loss": 0.2337,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2441682600382409,
+      "grad_norm": 2.9700610637664795,
+      "learning_rate": 4.544574434136564e-06,
+      "loss": 0.2049,
+      "step": 1277
+    },
+    {
+      "epoch": 0.24435946462715105,
+      "grad_norm": 2.179842710494995,
+      "learning_rate": 4.545071814348435e-06,
+      "loss": 0.3845,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2445506692160612,
+      "grad_norm": 2.47133731842041,
+      "learning_rate": 4.5455688055260765e-06,
+      "loss": 0.2824,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2447418738049713,
+      "grad_norm": 1.5787922143936157,
+      "learning_rate": 4.5460654082775925e-06,
+      "loss": 0.079,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24493307839388145,
+      "grad_norm": 2.6020150184631348,
+      "learning_rate": 4.546561623209661e-06,
+      "loss": 0.1166,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2451242829827916,
+      "grad_norm": 5.912609100341797,
+      "learning_rate": 4.547057450927541e-06,
+      "loss": 0.6601,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2453154875717017,
+      "grad_norm": 3.048389434814453,
+      "learning_rate": 4.547552892035077e-06,
+      "loss": 0.5557,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24550669216061186,
+      "grad_norm": 2.1050589084625244,
+      "learning_rate": 4.548047947134698e-06,
+      "loss": 0.1083,
+      "step": 1284
+    },
+    {
+      "epoch": 0.245697896749522,
+      "grad_norm": 2.0243983268737793,
+      "learning_rate": 4.5485426168274285e-06,
+      "loss": 0.1569,
+      "step": 1285
+    },
+    {
+      "epoch": 0.24588910133843211,
+      "grad_norm": 1.9739599227905273,
+      "learning_rate": 4.549036901712892e-06,
+      "loss": 0.1624,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24608030592734226,
+      "grad_norm": 1.8565188646316528,
+      "learning_rate": 4.549530802389311e-06,
+      "loss": 0.0998,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2462715105162524,
+      "grad_norm": 2.806781053543091,
+      "learning_rate": 4.550024319453516e-06,
+      "loss": 0.3977,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24646271510516252,
+      "grad_norm": 2.98370623588562,
+      "learning_rate": 4.550517453500946e-06,
+      "loss": 0.4309,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24665391969407266,
+      "grad_norm": 2.2158684730529785,
+      "learning_rate": 4.551010205125657e-06,
+      "loss": 0.3051,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2468451242829828,
+      "grad_norm": 2.2170090675354004,
+      "learning_rate": 4.551502574920322e-06,
+      "loss": 0.2447,
+      "step": 1291
+    },
+    {
+      "epoch": 0.24703632887189292,
+      "grad_norm": 2.093726634979248,
+      "learning_rate": 4.551994563476239e-06,
+      "loss": 0.1615,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24722753346080306,
+      "grad_norm": 0.953401505947113,
+      "learning_rate": 4.5524861713833315e-06,
+      "loss": 0.0517,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2474187380497132,
+      "grad_norm": 2.464068651199341,
+      "learning_rate": 4.552977399230156e-06,
+      "loss": 0.1319,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24760994263862332,
+      "grad_norm": 1.8242707252502441,
+      "learning_rate": 4.553468247603907e-06,
+      "loss": 0.1472,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24780114722753346,
+      "grad_norm": 1.8389755487442017,
+      "learning_rate": 4.553958717090414e-06,
+      "loss": 0.2597,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2479923518164436,
+      "grad_norm": 2.3550262451171875,
+      "learning_rate": 4.554448808274157e-06,
+      "loss": 0.1745,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24818355640535372,
+      "grad_norm": 1.627742886543274,
+      "learning_rate": 4.554938521738259e-06,
+      "loss": 0.0905,
+      "step": 1298
+    },
+    {
+      "epoch": 0.24837476099426387,
+      "grad_norm": 3.0560617446899414,
+      "learning_rate": 4.555427858064501e-06,
+      "loss": 0.5314,
+      "step": 1299
+    },
+    {
+      "epoch": 0.248565965583174,
+      "grad_norm": 2.0782976150512695,
+      "learning_rate": 4.555916817833317e-06,
+      "loss": 0.0902,
+      "step": 1300
+    },
+    {
+      "epoch": 0.24875717017208412,
+      "grad_norm": 2.113381862640381,
+      "learning_rate": 4.556405401623804e-06,
+      "loss": 0.4115,
+      "step": 1301
+    },
+    {
+      "epoch": 0.24894837476099427,
+      "grad_norm": 1.6574503183364868,
+      "learning_rate": 4.5568936100137235e-06,
+      "loss": 0.1394,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2491395793499044,
+      "grad_norm": 2.0017340183258057,
+      "learning_rate": 4.557381443579506e-06,
+      "loss": 0.1937,
+      "step": 1303
+    },
+    {
+      "epoch": 0.24933078393881453,
+      "grad_norm": 2.2529678344726562,
+      "learning_rate": 4.5578689028962575e-06,
+      "loss": 0.2572,
+      "step": 1304
+    },
+    {
+      "epoch": 0.24952198852772467,
+      "grad_norm": 1.7279530763626099,
+      "learning_rate": 4.558355988537758e-06,
+      "loss": 0.0845,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2497131931166348,
+      "grad_norm": 3.168016195297241,
+      "learning_rate": 4.558842701076469e-06,
+      "loss": 0.174,
+      "step": 1306
+    },
+    {
+      "epoch": 0.24990439770554493,
+      "grad_norm": 2.5194199085235596,
+      "learning_rate": 4.559329041083543e-06,
+      "loss": 0.3312,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25009560229445504,
+      "grad_norm": 3.844399929046631,
+      "learning_rate": 4.5598150091288164e-06,
+      "loss": 0.3948,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2502868068833652,
+      "grad_norm": 1.3260141611099243,
+      "learning_rate": 4.560300605780819e-06,
+      "loss": 0.0967,
+      "step": 1309
+    },
+    {
+      "epoch": 0.25047801147227533,
+      "grad_norm": 1.3077999353408813,
+      "learning_rate": 4.5607858316067835e-06,
+      "loss": 0.1281,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2506692160611855,
+      "grad_norm": 2.026874542236328,
+      "learning_rate": 4.561270687172638e-06,
+      "loss": 0.2637,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2508604206500956,
+      "grad_norm": 1.997843861579895,
+      "learning_rate": 4.561755173043019e-06,
+      "loss": 0.1711,
+      "step": 1312
+    },
+    {
+      "epoch": 0.25105162523900576,
+      "grad_norm": 1.434065341949463,
+      "learning_rate": 4.562239289781273e-06,
+      "loss": 0.1526,
+      "step": 1313
+    },
+    {
+      "epoch": 0.25124282982791585,
+      "grad_norm": 2.5661134719848633,
+      "learning_rate": 4.5627230379494595e-06,
+      "loss": 0.4677,
+      "step": 1314
+    },
+    {
+      "epoch": 0.251434034416826,
+      "grad_norm": 2.9416589736938477,
+      "learning_rate": 4.5632064181083524e-06,
+      "loss": 0.337,
+      "step": 1315
+    },
+    {
+      "epoch": 0.25162523900573613,
+      "grad_norm": 2.0680315494537354,
+      "learning_rate": 4.56368943081745e-06,
+      "loss": 0.2415,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2518164435946463,
+      "grad_norm": 1.4894253015518188,
+      "learning_rate": 4.564172076634976e-06,
+      "loss": 0.2191,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2520076481835564,
+      "grad_norm": 2.523341655731201,
+      "learning_rate": 4.56465435611788e-06,
+      "loss": 0.1231,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25219885277246656,
+      "grad_norm": 2.880688190460205,
+      "learning_rate": 4.5651362698218455e-06,
+      "loss": 0.2017,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25239005736137665,
+      "grad_norm": 2.6411757469177246,
+      "learning_rate": 4.565617818301295e-06,
+      "loss": 0.5086,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2525812619502868,
+      "grad_norm": 2.0425150394439697,
+      "learning_rate": 4.566099002109388e-06,
+      "loss": 0.2197,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25277246653919694,
+      "grad_norm": 2.203280448913574,
+      "learning_rate": 4.56657982179803e-06,
+      "loss": 0.2555,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2529636711281071,
+      "grad_norm": 1.5817688703536987,
+      "learning_rate": 4.567060277917876e-06,
+      "loss": 0.107,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2531548757170172,
+      "grad_norm": 2.743864059448242,
+      "learning_rate": 4.567540371018329e-06,
+      "loss": 0.1049,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25334608030592737,
+      "grad_norm": 1.1663135290145874,
+      "learning_rate": 4.568020101647551e-06,
+      "loss": 0.0449,
+      "step": 1325
+    },
+    {
+      "epoch": 0.25353728489483746,
+      "grad_norm": 2.348593235015869,
+      "learning_rate": 4.568499470352461e-06,
+      "loss": 0.4336,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2537284894837476,
+      "grad_norm": 2.174994468688965,
+      "learning_rate": 4.568978477678743e-06,
+      "loss": 0.4255,
+      "step": 1327
+    },
+    {
+      "epoch": 0.25391969407265774,
+      "grad_norm": 2.499854564666748,
+      "learning_rate": 4.5694571241708465e-06,
+      "loss": 0.1284,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2541108986615679,
+      "grad_norm": 1.8215583562850952,
+      "learning_rate": 4.5699354103719936e-06,
+      "loss": 0.1327,
+      "step": 1329
+    },
+    {
+      "epoch": 0.25430210325047803,
+      "grad_norm": 2.302138328552246,
+      "learning_rate": 4.570413336824176e-06,
+      "loss": 0.0985,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25449330783938817,
+      "grad_norm": 2.055473804473877,
+      "learning_rate": 4.570890904068169e-06,
+      "loss": 0.0946,
+      "step": 1331
+    },
+    {
+      "epoch": 0.25468451242829826,
+      "grad_norm": 1.9759926795959473,
+      "learning_rate": 4.571368112643526e-06,
+      "loss": 0.1447,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2548757170172084,
+      "grad_norm": 2.6056742668151855,
+      "learning_rate": 4.571844963088587e-06,
+      "loss": 0.4277,
+      "step": 1333
+    },
+    {
+      "epoch": 0.25506692160611855,
+      "grad_norm": 1.6580945253372192,
+      "learning_rate": 4.572321455940478e-06,
+      "loss": 0.3823,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2552581261950287,
+      "grad_norm": 3.8294076919555664,
+      "learning_rate": 4.572797591735123e-06,
+      "loss": 0.0972,
+      "step": 1335
+    },
+    {
+      "epoch": 0.25544933078393883,
+      "grad_norm": 1.2979289293289185,
+      "learning_rate": 4.573273371007238e-06,
+      "loss": 0.0785,
+      "step": 1336
+    },
+    {
+      "epoch": 0.255640535372849,
+      "grad_norm": 1.569770097732544,
+      "learning_rate": 4.573748794290339e-06,
+      "loss": 0.1083,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25583173996175906,
+      "grad_norm": 1.5974154472351074,
+      "learning_rate": 4.574223862116746e-06,
+      "loss": 0.1379,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2560229445506692,
+      "grad_norm": 2.4823997020721436,
+      "learning_rate": 4.574698575017587e-06,
+      "loss": 0.3444,
+      "step": 1339
+    },
+    {
+      "epoch": 0.25621414913957935,
+      "grad_norm": 1.7080856561660767,
+      "learning_rate": 4.5751729335227995e-06,
+      "loss": 0.1214,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2564053537284895,
+      "grad_norm": 2.0777196884155273,
+      "learning_rate": 4.575646938161135e-06,
+      "loss": 0.1628,
+      "step": 1341
+    },
+    {
+      "epoch": 0.25659655831739964,
+      "grad_norm": 1.514445424079895,
+      "learning_rate": 4.576120589460161e-06,
+      "loss": 0.1017,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2567877629063097,
+      "grad_norm": 3.241065502166748,
+      "learning_rate": 4.576593887946269e-06,
+      "loss": 0.1865,
+      "step": 1343
+    },
+    {
+      "epoch": 0.25697896749521987,
+      "grad_norm": 1.9930219650268555,
+      "learning_rate": 4.577066834144674e-06,
+      "loss": 0.2197,
+      "step": 1344
+    },
+    {
+      "epoch": 0.25717017208413,
+      "grad_norm": 2.6061015129089355,
+      "learning_rate": 4.577539428579417e-06,
+      "loss": 0.4761,
+      "step": 1345
+    },
+    {
+      "epoch": 0.25736137667304015,
+      "grad_norm": 1.4190788269042969,
+      "learning_rate": 4.578011671773374e-06,
+      "loss": 0.1455,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2575525812619503,
+      "grad_norm": 1.4043774604797363,
+      "learning_rate": 4.578483564248254e-06,
+      "loss": 0.0912,
+      "step": 1347
+    },
+    {
+      "epoch": 0.25774378585086044,
+      "grad_norm": 2.920255184173584,
+      "learning_rate": 4.578955106524605e-06,
+      "loss": 0.1808,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25793499043977053,
+      "grad_norm": 2.587714672088623,
+      "learning_rate": 4.5794262991218156e-06,
+      "loss": 0.1429,
+      "step": 1349
+    },
+    {
+      "epoch": 0.25812619502868067,
+      "grad_norm": 2.5600852966308594,
+      "learning_rate": 4.5798971425581235e-06,
+      "loss": 0.2295,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2583173996175908,
+      "grad_norm": 1.5082076787948608,
+      "learning_rate": 4.580367637350609e-06,
+      "loss": 0.174,
+      "step": 1351
+    },
+    {
+      "epoch": 0.25850860420650096,
+      "grad_norm": 1.4872022867202759,
+      "learning_rate": 4.580837784015212e-06,
+      "loss": 0.1193,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2586998087954111,
+      "grad_norm": 3.185533285140991,
+      "learning_rate": 4.581307583066722e-06,
+      "loss": 0.1592,
+      "step": 1353
+    },
+    {
+      "epoch": 0.25889101338432124,
+      "grad_norm": 2.905579090118408,
+      "learning_rate": 4.58177703501879e-06,
+      "loss": 0.1793,
+      "step": 1354
+    },
+    {
+      "epoch": 0.25908221797323133,
+      "grad_norm": 1.6768579483032227,
+      "learning_rate": 4.58224614038393e-06,
+      "loss": 0.0814,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2592734225621415,
+      "grad_norm": 4.425497055053711,
+      "learning_rate": 4.58271489967352e-06,
+      "loss": 0.2171,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2594646271510516,
+      "grad_norm": 2.09071946144104,
+      "learning_rate": 4.5831833133978085e-06,
+      "loss": 0.1503,
+      "step": 1357
+    },
+    {
+      "epoch": 0.25965583173996176,
+      "grad_norm": 2.006558895111084,
+      "learning_rate": 4.583651382065915e-06,
+      "loss": 0.4281,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2598470363288719,
+      "grad_norm": 1.6992747783660889,
+      "learning_rate": 4.584119106185835e-06,
+      "loss": 0.1838,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26003824091778205,
+      "grad_norm": 2.895442485809326,
+      "learning_rate": 4.584586486264445e-06,
+      "loss": 0.5134,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26022944550669214,
+      "grad_norm": 3.2242867946624756,
+      "learning_rate": 4.585053522807501e-06,
+      "loss": 0.1576,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2604206500956023,
+      "grad_norm": 3.0295143127441406,
+      "learning_rate": 4.5855202163196466e-06,
+      "loss": 0.3062,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2606118546845124,
+      "grad_norm": 1.9487686157226562,
+      "learning_rate": 4.585986567304413e-06,
+      "loss": 0.2205,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26080305927342257,
+      "grad_norm": 2.4985392093658447,
+      "learning_rate": 4.586452576264223e-06,
+      "loss": 0.3626,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2609942638623327,
+      "grad_norm": 3.514467716217041,
+      "learning_rate": 4.586918243700398e-06,
+      "loss": 0.365,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26118546845124285,
+      "grad_norm": 2.703725576400757,
+      "learning_rate": 4.587383570113155e-06,
+      "loss": 0.4229,
+      "step": 1366
+    },
+    {
+      "epoch": 0.26137667304015294,
+      "grad_norm": 2.9062862396240234,
+      "learning_rate": 4.587848556001613e-06,
+      "loss": 0.2676,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2615678776290631,
+      "grad_norm": 1.557067632675171,
+      "learning_rate": 4.588313201863795e-06,
+      "loss": 0.092,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2617590822179732,
+      "grad_norm": 2.868450880050659,
+      "learning_rate": 4.588777508196637e-06,
+      "loss": 0.2177,
+      "step": 1369
+    },
+    {
+      "epoch": 0.26195028680688337,
+      "grad_norm": 2.503908157348633,
+      "learning_rate": 4.589241475495983e-06,
+      "loss": 0.3252,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2621414913957935,
+      "grad_norm": 3.2766878604888916,
+      "learning_rate": 4.58970510425659e-06,
+      "loss": 0.5061,
+      "step": 1371
+    },
+    {
+      "epoch": 0.26233269598470366,
+      "grad_norm": 1.436415672302246,
+      "learning_rate": 4.5901683949721355e-06,
+      "loss": 0.113,
+      "step": 1372
+    },
+    {
+      "epoch": 0.26252390057361374,
+      "grad_norm": 1.4718507528305054,
+      "learning_rate": 4.590631348135217e-06,
+      "loss": 0.2098,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2627151051625239,
+      "grad_norm": 1.3220497369766235,
+      "learning_rate": 4.591093964237357e-06,
+      "loss": 0.0534,
+      "step": 1374
+    },
+    {
+      "epoch": 0.26290630975143403,
+      "grad_norm": 2.3449716567993164,
+      "learning_rate": 4.591556243769003e-06,
+      "loss": 0.1064,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2630975143403442,
+      "grad_norm": 2.357490301132202,
+      "learning_rate": 4.592018187219536e-06,
+      "loss": 0.5121,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2632887189292543,
+      "grad_norm": 2.1978540420532227,
+      "learning_rate": 4.5924797950772665e-06,
+      "loss": 0.2471,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26347992351816446,
+      "grad_norm": 2.617004871368408,
+      "learning_rate": 4.592941067829446e-06,
+      "loss": 0.3932,
+      "step": 1378
+    },
+    {
+      "epoch": 0.26367112810707455,
+      "grad_norm": 2.3759875297546387,
+      "learning_rate": 4.593402005962261e-06,
+      "loss": 0.3458,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2638623326959847,
+      "grad_norm": 1.166843056678772,
+      "learning_rate": 4.593862609960843e-06,
+      "loss": 0.0701,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26405353728489483,
+      "grad_norm": 2.4391300678253174,
+      "learning_rate": 4.594322880309272e-06,
+      "loss": 0.1061,
+      "step": 1381
+    },
+    {
+      "epoch": 0.264244741873805,
+      "grad_norm": 1.7300337553024292,
+      "learning_rate": 4.594782817490571e-06,
+      "loss": 0.1513,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2644359464627151,
+      "grad_norm": 2.3044793605804443,
+      "learning_rate": 4.595242421986719e-06,
+      "loss": 0.2708,
+      "step": 1383
+    },
+    {
+      "epoch": 0.26462715105162526,
+      "grad_norm": 1.8326455354690552,
+      "learning_rate": 4.595701694278649e-06,
+      "loss": 0.253,
+      "step": 1384
+    },
+    {
+      "epoch": 0.26481835564053535,
+      "grad_norm": 2.248283863067627,
+      "learning_rate": 4.5961606348462506e-06,
+      "loss": 0.2861,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2650095602294455,
+      "grad_norm": 3.2349724769592285,
+      "learning_rate": 4.596619244168376e-06,
+      "loss": 0.2644,
+      "step": 1386
+    },
+    {
+      "epoch": 0.26520076481835564,
+      "grad_norm": 3.4575836658477783,
+      "learning_rate": 4.59707752272284e-06,
+      "loss": 0.1486,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2653919694072658,
+      "grad_norm": 1.597733497619629,
+      "learning_rate": 4.597535470986426e-06,
+      "loss": 0.1201,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2655831739961759,
+      "grad_norm": 2.6269469261169434,
+      "learning_rate": 4.597993089434886e-06,
+      "loss": 0.18,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26577437858508607,
+      "grad_norm": 2.98217511177063,
+      "learning_rate": 4.598450378542943e-06,
+      "loss": 0.3596,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26596558317399616,
+      "grad_norm": 1.9663608074188232,
+      "learning_rate": 4.5989073387843e-06,
+      "loss": 0.2026,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2661567877629063,
+      "grad_norm": 1.8245049715042114,
+      "learning_rate": 4.599363970631637e-06,
+      "loss": 0.1057,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26634799235181644,
+      "grad_norm": 2.081806182861328,
+      "learning_rate": 4.599820274556611e-06,
+      "loss": 0.1292,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2665391969407266,
+      "grad_norm": 2.341907262802124,
+      "learning_rate": 4.6002762510298725e-06,
+      "loss": 0.159,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26673040152963673,
+      "grad_norm": 4.16287899017334,
+      "learning_rate": 4.600731900521051e-06,
+      "loss": 0.5912,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2669216061185469,
+      "grad_norm": 2.3653371334075928,
+      "learning_rate": 4.601187223498774e-06,
+      "loss": 0.1569,
+      "step": 1396
+    },
+    {
+      "epoch": 0.26711281070745696,
+      "grad_norm": 1.7293826341629028,
+      "learning_rate": 4.601642220430655e-06,
+      "loss": 0.1551,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2673040152963671,
+      "grad_norm": 2.207956314086914,
+      "learning_rate": 4.602096891783308e-06,
+      "loss": 0.1525,
+      "step": 1398
+    },
+    {
+      "epoch": 0.26749521988527725,
+      "grad_norm": 2.079894542694092,
+      "learning_rate": 4.6025512380223466e-06,
+      "loss": 0.2216,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2676864244741874,
+      "grad_norm": 1.731223702430725,
+      "learning_rate": 4.603005259612382e-06,
+      "loss": 0.1097,
+      "step": 1400
+    },
+    {
+      "epoch": 0.26787762906309753,
+      "grad_norm": 2.4742074012756348,
+      "learning_rate": 4.603458957017036e-06,
+      "loss": 0.3192,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2680688336520076,
+      "grad_norm": 2.1221306324005127,
+      "learning_rate": 4.603912330698932e-06,
+      "loss": 0.2922,
+      "step": 1402
+    },
+    {
+      "epoch": 0.26826003824091776,
+      "grad_norm": 3.109562873840332,
+      "learning_rate": 4.60436538111971e-06,
+      "loss": 0.3177,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2684512428298279,
+      "grad_norm": 2.970853567123413,
+      "learning_rate": 4.6048181087400175e-06,
+      "loss": 0.3191,
+      "step": 1404
+    },
+    {
+      "epoch": 0.26864244741873805,
+      "grad_norm": 1.673386812210083,
+      "learning_rate": 4.605270514019522e-06,
+      "loss": 0.0833,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2688336520076482,
+      "grad_norm": 2.224393606185913,
+      "learning_rate": 4.605722597416907e-06,
+      "loss": 0.2605,
+      "step": 1406
+    },
+    {
+      "epoch": 0.26902485659655834,
+      "grad_norm": 1.8234128952026367,
+      "learning_rate": 4.6061743593898815e-06,
+      "loss": 0.1346,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2692160611854684,
+      "grad_norm": 1.949435830116272,
+      "learning_rate": 4.6066258003951735e-06,
+      "loss": 0.157,
+      "step": 1408
+    },
+    {
+      "epoch": 0.26940726577437857,
+      "grad_norm": 1.833545446395874,
+      "learning_rate": 4.607076920888543e-06,
+      "loss": 0.1708,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2695984703632887,
+      "grad_norm": 2.0920236110687256,
+      "learning_rate": 4.607527721324779e-06,
+      "loss": 0.32,
+      "step": 1410
+    },
+    {
+      "epoch": 0.26978967495219885,
+      "grad_norm": 1.772287368774414,
+      "learning_rate": 4.6079782021577e-06,
+      "loss": 0.1181,
+      "step": 1411
+    },
+    {
+      "epoch": 0.269980879541109,
+      "grad_norm": 3.1658973693847656,
+      "learning_rate": 4.608428363840164e-06,
+      "loss": 0.1821,
+      "step": 1412
+    },
+    {
+      "epoch": 0.27017208413001914,
+      "grad_norm": 1.7124018669128418,
+      "learning_rate": 4.608878206824065e-06,
+      "loss": 0.0898,
+      "step": 1413
+    },
+    {
+      "epoch": 0.27036328871892923,
+      "grad_norm": 4.624269008636475,
+      "learning_rate": 4.6093277315603385e-06,
+      "loss": 0.461,
+      "step": 1414
+    },
+    {
+      "epoch": 0.27055449330783937,
+      "grad_norm": 3.7334680557250977,
+      "learning_rate": 4.609776938498964e-06,
+      "loss": 0.2932,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2707456978967495,
+      "grad_norm": 2.483091354370117,
+      "learning_rate": 4.610225828088966e-06,
+      "loss": 0.2618,
+      "step": 1416
+    },
+    {
+      "epoch": 0.27093690248565966,
+      "grad_norm": 2.0719807147979736,
+      "learning_rate": 4.610674400778419e-06,
+      "loss": 0.256,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2711281070745698,
+      "grad_norm": 2.7514090538024902,
+      "learning_rate": 4.6111226570144505e-06,
+      "loss": 0.2337,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27131931166347995,
+      "grad_norm": 4.372493267059326,
+      "learning_rate": 4.611570597243238e-06,
+      "loss": 0.348,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27151051625239003,
+      "grad_norm": 3.5149691104888916,
+      "learning_rate": 4.6120182219100225e-06,
+      "loss": 0.6993,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2717017208413002,
+      "grad_norm": 2.984778881072998,
+      "learning_rate": 4.612465531459098e-06,
+      "loss": 0.2772,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2718929254302103,
+      "grad_norm": 2.6740102767944336,
+      "learning_rate": 4.612912526333825e-06,
+      "loss": 0.3069,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27208413001912046,
+      "grad_norm": 1.7484631538391113,
+      "learning_rate": 4.613359206976629e-06,
+      "loss": 0.0702,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2722753346080306,
+      "grad_norm": 2.2864737510681152,
+      "learning_rate": 4.613805573829002e-06,
+      "loss": 0.1678,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27246653919694075,
+      "grad_norm": 1.9719682931900024,
+      "learning_rate": 4.614251627331505e-06,
+      "loss": 0.1079,
+      "step": 1425
+    },
+    {
+      "epoch": 0.27265774378585084,
+      "grad_norm": 3.1089677810668945,
+      "learning_rate": 4.614697367923773e-06,
+      "loss": 0.4868,
+      "step": 1426
+    },
+    {
+      "epoch": 0.272848948374761,
+      "grad_norm": 3.177788734436035,
+      "learning_rate": 4.615142796044517e-06,
+      "loss": 0.1517,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2730401529636711,
+      "grad_norm": 1.7205730676651,
+      "learning_rate": 4.615587912131526e-06,
+      "loss": 0.1392,
+      "step": 1428
+    },
+    {
+      "epoch": 0.27323135755258127,
+      "grad_norm": 2.4936370849609375,
+      "learning_rate": 4.61603271662167e-06,
+      "loss": 0.2818,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2734225621414914,
+      "grad_norm": 2.6876871585845947,
+      "learning_rate": 4.616477209950898e-06,
+      "loss": 0.2261,
+      "step": 1430
+    },
+    {
+      "epoch": 0.27361376673040155,
+      "grad_norm": 2.5818753242492676,
+      "learning_rate": 4.616921392554251e-06,
+      "loss": 0.1231,
+      "step": 1431
+    },
+    {
+      "epoch": 0.27380497131931164,
+      "grad_norm": 2.133584499359131,
+      "learning_rate": 4.617365264865855e-06,
+      "loss": 0.2894,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2739961759082218,
+      "grad_norm": 3.4310872554779053,
+      "learning_rate": 4.6178088273189265e-06,
+      "loss": 0.7312,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2741873804971319,
+      "grad_norm": 1.8406034708023071,
+      "learning_rate": 4.618252080345775e-06,
+      "loss": 0.1771,
+      "step": 1434
+    },
+    {
+      "epoch": 0.27437858508604207,
+      "grad_norm": 1.9556293487548828,
+      "learning_rate": 4.61869502437781e-06,
+      "loss": 0.1997,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2745697896749522,
+      "grad_norm": 2.400635242462158,
+      "learning_rate": 4.619137659845533e-06,
+      "loss": 0.1104,
+      "step": 1436
+    },
+    {
+      "epoch": 0.27476099426386236,
+      "grad_norm": 3.1298184394836426,
+      "learning_rate": 4.619579987178551e-06,
+      "loss": 0.2818,
+      "step": 1437
+    },
+    {
+      "epoch": 0.27495219885277244,
+      "grad_norm": 3.0977323055267334,
+      "learning_rate": 4.620022006805574e-06,
+      "loss": 0.5284,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2751434034416826,
+      "grad_norm": 1.5940463542938232,
+      "learning_rate": 4.620463719154416e-06,
+      "loss": 0.2201,
+      "step": 1439
+    },
+    {
+      "epoch": 0.27533460803059273,
+      "grad_norm": 2.438342809677124,
+      "learning_rate": 4.620905124652002e-06,
+      "loss": 0.2274,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2755258126195029,
+      "grad_norm": 1.9232128858566284,
+      "learning_rate": 4.6213462237243646e-06,
+      "loss": 0.1143,
+      "step": 1441
+    },
+    {
+      "epoch": 0.275717017208413,
+      "grad_norm": 2.3613779544830322,
+      "learning_rate": 4.621787016796653e-06,
+      "loss": 0.1331,
+      "step": 1442
+    },
+    {
+      "epoch": 0.27590822179732316,
+      "grad_norm": 1.824241280555725,
+      "learning_rate": 4.6222275042931295e-06,
+      "loss": 0.086,
+      "step": 1443
+    },
+    {
+      "epoch": 0.27609942638623325,
+      "grad_norm": 2.2183074951171875,
+      "learning_rate": 4.622667686637177e-06,
+      "loss": 0.1208,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2762906309751434,
+      "grad_norm": 2.0111348628997803,
+      "learning_rate": 4.623107564251298e-06,
+      "loss": 0.1653,
+      "step": 1445
+    },
+    {
+      "epoch": 0.27648183556405354,
+      "grad_norm": 2.0708916187286377,
+      "learning_rate": 4.623547137557118e-06,
+      "loss": 0.1958,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2766730401529637,
+      "grad_norm": 3.2641139030456543,
+      "learning_rate": 4.623986406975387e-06,
+      "loss": 0.2679,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2768642447418738,
+      "grad_norm": 1.6562097072601318,
+      "learning_rate": 4.624425372925986e-06,
+      "loss": 0.1593,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27705544933078396,
+      "grad_norm": 3.0824685096740723,
+      "learning_rate": 4.624864035827925e-06,
+      "loss": 0.1947,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27724665391969405,
+      "grad_norm": 2.1108744144439697,
+      "learning_rate": 4.6253023960993445e-06,
+      "loss": 0.0826,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2774378585086042,
+      "grad_norm": 1.9569447040557861,
+      "learning_rate": 4.625740454157524e-06,
+      "loss": 0.2948,
+      "step": 1451
+    },
+    {
+      "epoch": 0.27762906309751434,
+      "grad_norm": 2.475095272064209,
+      "learning_rate": 4.626178210418876e-06,
+      "loss": 0.2801,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2778202676864245,
+      "grad_norm": 1.949391484260559,
+      "learning_rate": 4.626615665298957e-06,
+      "loss": 0.2627,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2780114722753346,
+      "grad_norm": 3.6278679370880127,
+      "learning_rate": 4.627052819212466e-06,
+      "loss": 0.3923,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27820267686424477,
+      "grad_norm": 1.5531980991363525,
+      "learning_rate": 4.627489672573243e-06,
+      "loss": 0.0759,
+      "step": 1455
+    },
+    {
+      "epoch": 0.27839388145315486,
+      "grad_norm": 2.221863269805908,
+      "learning_rate": 4.627926225794277e-06,
+      "loss": 0.1375,
+      "step": 1456
+    },
+    {
+      "epoch": 0.278585086042065,
+      "grad_norm": 3.4804158210754395,
+      "learning_rate": 4.628362479287708e-06,
+      "loss": 0.31,
+      "step": 1457
+    },
+    {
+      "epoch": 0.27877629063097514,
+      "grad_norm": 2.319563627243042,
+      "learning_rate": 4.628798433464823e-06,
+      "loss": 0.3307,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2789674952198853,
+      "grad_norm": 2.1878955364227295,
+      "learning_rate": 4.62923408873607e-06,
+      "loss": 0.1792,
+      "step": 1459
+    },
+    {
+      "epoch": 0.27915869980879543,
+      "grad_norm": 1.8384076356887817,
+      "learning_rate": 4.629669445511046e-06,
+      "loss": 0.1425,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2793499043977055,
+      "grad_norm": 3.533768892288208,
+      "learning_rate": 4.630104504198513e-06,
+      "loss": 0.2667,
+      "step": 1461
+    },
+    {
+      "epoch": 0.27954110898661566,
+      "grad_norm": 2.315150499343872,
+      "learning_rate": 4.6305392652063885e-06,
+      "loss": 0.1398,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2797323135755258,
+      "grad_norm": 3.2896182537078857,
+      "learning_rate": 4.630973728941758e-06,
+      "loss": 0.3448,
+      "step": 1463
+    },
+    {
+      "epoch": 0.27992351816443595,
+      "grad_norm": 2.074253797531128,
+      "learning_rate": 4.631407895810868e-06,
+      "loss": 0.2655,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2801147227533461,
+      "grad_norm": 2.220587730407715,
+      "learning_rate": 4.631841766219136e-06,
+      "loss": 0.4133,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28030592734225623,
+      "grad_norm": 2.7555809020996094,
+      "learning_rate": 4.63227534057115e-06,
+      "loss": 0.3825,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2804971319311663,
+      "grad_norm": 1.95787513256073,
+      "learning_rate": 4.6327086192706666e-06,
+      "loss": 0.1425,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28068833652007646,
+      "grad_norm": 2.1310410499572754,
+      "learning_rate": 4.633141602720621e-06,
+      "loss": 0.1743,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2808795411089866,
+      "grad_norm": 1.3947631120681763,
+      "learning_rate": 4.633574291323124e-06,
+      "loss": 0.0875,
+      "step": 1469
+    },
+    {
+      "epoch": 0.28107074569789675,
+      "grad_norm": 3.2543869018554688,
+      "learning_rate": 4.6340066854794634e-06,
+      "loss": 0.5823,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2812619502868069,
+      "grad_norm": 2.7726821899414062,
+      "learning_rate": 4.634438785590112e-06,
+      "loss": 0.4261,
+      "step": 1471
+    },
+    {
+      "epoch": 0.28145315487571704,
+      "grad_norm": 1.8597933053970337,
+      "learning_rate": 4.634870592054722e-06,
+      "loss": 0.3469,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2816443594646271,
+      "grad_norm": 1.49411141872406,
+      "learning_rate": 4.635302105272136e-06,
+      "loss": 0.0695,
+      "step": 1473
+    },
+    {
+      "epoch": 0.28183556405353727,
+      "grad_norm": 2.444284677505493,
+      "learning_rate": 4.635733325640381e-06,
+      "loss": 0.2761,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2820267686424474,
+      "grad_norm": 2.646402359008789,
+      "learning_rate": 4.636164253556675e-06,
+      "loss": 0.1316,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28221797323135756,
+      "grad_norm": 2.195561408996582,
+      "learning_rate": 4.636594889417429e-06,
+      "loss": 0.4064,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2824091778202677,
+      "grad_norm": 1.763237476348877,
+      "learning_rate": 4.637025233618248e-06,
+      "loss": 0.1922,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28260038240917784,
+      "grad_norm": 2.47198748588562,
+      "learning_rate": 4.637455286553934e-06,
+      "loss": 0.4326,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28279158699808793,
+      "grad_norm": 1.6601879596710205,
+      "learning_rate": 4.637885048618489e-06,
+      "loss": 0.124,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2829827915869981,
+      "grad_norm": 1.9923487901687622,
+      "learning_rate": 4.6383145202051135e-06,
+      "loss": 0.1602,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2831739961759082,
+      "grad_norm": 1.6631501913070679,
+      "learning_rate": 4.638743701706214e-06,
+      "loss": 0.1,
+      "step": 1481
+    },
+    {
+      "epoch": 0.28336520076481836,
+      "grad_norm": 2.8882148265838623,
+      "learning_rate": 4.639172593513399e-06,
+      "loss": 0.3787,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2835564053537285,
+      "grad_norm": 2.454380750656128,
+      "learning_rate": 4.639601196017489e-06,
+      "loss": 0.3458,
+      "step": 1483
+    },
+    {
+      "epoch": 0.28374760994263865,
+      "grad_norm": 2.356696367263794,
+      "learning_rate": 4.640029509608511e-06,
+      "loss": 0.2473,
+      "step": 1484
+    },
+    {
+      "epoch": 0.28393881453154873,
+      "grad_norm": 2.13679838180542,
+      "learning_rate": 4.640457534675704e-06,
+      "loss": 0.318,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2841300191204589,
+      "grad_norm": 1.2381728887557983,
+      "learning_rate": 4.640885271607523e-06,
+      "loss": 0.0621,
+      "step": 1486
+    },
+    {
+      "epoch": 0.284321223709369,
+      "grad_norm": 2.054654121398926,
+      "learning_rate": 4.641312720791636e-06,
+      "loss": 0.33,
+      "step": 1487
+    },
+    {
+      "epoch": 0.28451242829827916,
+      "grad_norm": 3.917362928390503,
+      "learning_rate": 4.64173988261493e-06,
+      "loss": 0.1747,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2847036328871893,
+      "grad_norm": 2.648787260055542,
+      "learning_rate": 4.642166757463516e-06,
+      "loss": 0.5647,
+      "step": 1489
+    },
+    {
+      "epoch": 0.28489483747609945,
+      "grad_norm": 2.680501937866211,
+      "learning_rate": 4.6425933457227225e-06,
+      "loss": 0.3256,
+      "step": 1490
+    },
+    {
+      "epoch": 0.28508604206500954,
+      "grad_norm": 2.612128734588623,
+      "learning_rate": 4.643019647777103e-06,
+      "loss": 0.2337,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2852772466539197,
+      "grad_norm": 2.540062189102173,
+      "learning_rate": 4.6434456640104405e-06,
+      "loss": 0.1389,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2854684512428298,
+      "grad_norm": 3.286160469055176,
+      "learning_rate": 4.643871394805745e-06,
+      "loss": 0.1752,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28565965583173997,
+      "grad_norm": 1.7292871475219727,
+      "learning_rate": 4.644296840545256e-06,
+      "loss": 0.0987,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2858508604206501,
+      "grad_norm": 2.197441816329956,
+      "learning_rate": 4.644722001610448e-06,
+      "loss": 0.1405,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28604206500956025,
+      "grad_norm": 1.9219545125961304,
+      "learning_rate": 4.645146878382026e-06,
+      "loss": 0.2266,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28623326959847034,
+      "grad_norm": 2.036558151245117,
+      "learning_rate": 4.645571471239938e-06,
+      "loss": 0.1706,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2864244741873805,
+      "grad_norm": 1.5000076293945312,
+      "learning_rate": 4.6459957805633654e-06,
+      "loss": 0.1039,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2866156787762906,
+      "grad_norm": 2.3707828521728516,
+      "learning_rate": 4.646419806730734e-06,
+      "loss": 0.1623,
+      "step": 1499
+    },
+    {
+      "epoch": 0.28680688336520077,
+      "grad_norm": 1.9618526697158813,
+      "learning_rate": 4.64684355011971e-06,
+      "loss": 0.1254,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28680688336520077,
+      "eval_runtime": 761.9792,
+      "eval_samples_per_second": 2.013,
+      "eval_steps_per_second": 0.252,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2869980879541109,
+      "grad_norm": 2.2166504859924316,
+      "learning_rate": 4.6472670111072075e-06,
+      "loss": 0.37,
+      "step": 1501
+    },
+    {
+      "epoch": 0.28718929254302106,
+      "grad_norm": 2.407583236694336,
+      "learning_rate": 4.647690190069383e-06,
+      "loss": 0.3667,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28738049713193115,
+      "grad_norm": 2.0655288696289062,
+      "learning_rate": 4.648113087381647e-06,
+      "loss": 0.1062,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2875717017208413,
+      "grad_norm": 1.6839008331298828,
+      "learning_rate": 4.648535703418657e-06,
+      "loss": 0.122,
+      "step": 1504
+    },
+    {
+      "epoch": 0.28776290630975143,
+      "grad_norm": 2.2787880897521973,
+      "learning_rate": 4.648958038554326e-06,
+      "loss": 0.3146,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2879541108986616,
+      "grad_norm": 1.7219648361206055,
+      "learning_rate": 4.64938009316182e-06,
+      "loss": 0.0914,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2881453154875717,
+      "grad_norm": 2.2042160034179688,
+      "learning_rate": 4.6498018676135644e-06,
+      "loss": 0.291,
+      "step": 1507
+    },
+    {
+      "epoch": 0.28833652007648186,
+      "grad_norm": 2.321742057800293,
+      "learning_rate": 4.65022336228124e-06,
+      "loss": 0.2794,
+      "step": 1508
+    },
+    {
+      "epoch": 0.28852772466539195,
+      "grad_norm": 2.547917127609253,
+      "learning_rate": 4.650644577535791e-06,
+      "loss": 0.5054,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2887189292543021,
+      "grad_norm": 1.275026798248291,
+      "learning_rate": 4.651065513747423e-06,
+      "loss": 0.0787,
+      "step": 1510
+    },
+    {
+      "epoch": 0.28891013384321224,
+      "grad_norm": 3.1266400814056396,
+      "learning_rate": 4.651486171285608e-06,
+      "loss": 0.3595,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2891013384321224,
+      "grad_norm": 1.9492202997207642,
+      "learning_rate": 4.651906550519083e-06,
+      "loss": 0.123,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2892925430210325,
+      "grad_norm": 1.362431526184082,
+      "learning_rate": 4.652326651815854e-06,
+      "loss": 0.1088,
+      "step": 1513
+    },
+    {
+      "epoch": 0.28948374760994267,
+      "grad_norm": 1.5959398746490479,
+      "learning_rate": 4.652746475543199e-06,
+      "loss": 0.1612,
+      "step": 1514
+    },
+    {
+      "epoch": 0.28967495219885275,
+      "grad_norm": 1.9055774211883545,
+      "learning_rate": 4.6531660220676665e-06,
+      "loss": 0.2062,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2898661567877629,
+      "grad_norm": 1.3554742336273193,
+      "learning_rate": 4.653585291755081e-06,
+      "loss": 0.1628,
+      "step": 1516
+    },
+    {
+      "epoch": 0.29005736137667304,
+      "grad_norm": 1.883872628211975,
+      "learning_rate": 4.65400428497054e-06,
+      "loss": 0.1861,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2902485659655832,
+      "grad_norm": 1.0759433507919312,
+      "learning_rate": 4.654423002078425e-06,
+      "loss": 0.0531,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2904397705544933,
+      "grad_norm": 2.0732216835021973,
+      "learning_rate": 4.654841443442393e-06,
+      "loss": 0.1208,
+      "step": 1519
+    },
+    {
+      "epoch": 0.29063097514340347,
+      "grad_norm": 2.8344123363494873,
+      "learning_rate": 4.655259609425383e-06,
+      "loss": 0.7317,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29082217973231356,
+      "grad_norm": 1.3993836641311646,
+      "learning_rate": 4.655677500389621e-06,
+      "loss": 0.1063,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2910133843212237,
+      "grad_norm": 1.489634394645691,
+      "learning_rate": 4.6560951166966175e-06,
+      "loss": 0.1145,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29120458891013384,
+      "grad_norm": 1.8000231981277466,
+      "learning_rate": 4.656512458707168e-06,
+      "loss": 0.1213,
+      "step": 1523
+    },
+    {
+      "epoch": 0.291395793499044,
+      "grad_norm": 2.340092658996582,
+      "learning_rate": 4.656929526781362e-06,
+      "loss": 0.2737,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29158699808795413,
+      "grad_norm": 2.1177775859832764,
+      "learning_rate": 4.6573463212785765e-06,
+      "loss": 0.0867,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2917782026768642,
+      "grad_norm": 2.30710768699646,
+      "learning_rate": 4.657762842557484e-06,
+      "loss": 0.3876,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29196940726577436,
+      "grad_norm": 2.207430362701416,
+      "learning_rate": 4.658179090976053e-06,
+      "loss": 0.2844,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2921606118546845,
+      "grad_norm": 2.2533726692199707,
+      "learning_rate": 4.658595066891546e-06,
+      "loss": 0.1704,
+      "step": 1528
+    },
+    {
+      "epoch": 0.29235181644359465,
+      "grad_norm": 1.5354182720184326,
+      "learning_rate": 4.6590107706605244e-06,
+      "loss": 0.1379,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2925430210325048,
+      "grad_norm": 1.293443202972412,
+      "learning_rate": 4.659426202638854e-06,
+      "loss": 0.0735,
+      "step": 1530
+    },
+    {
+      "epoch": 0.29273422562141493,
+      "grad_norm": 2.447760820388794,
+      "learning_rate": 4.6598413631817005e-06,
+      "loss": 0.0711,
+      "step": 1531
+    },
+    {
+      "epoch": 0.292925430210325,
+      "grad_norm": 3.0195038318634033,
+      "learning_rate": 4.660256252643533e-06,
+      "loss": 0.3391,
+      "step": 1532
+    },
+    {
+      "epoch": 0.29311663479923517,
+      "grad_norm": 2.5726795196533203,
+      "learning_rate": 4.660670871378128e-06,
+      "loss": 0.456,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2933078393881453,
+      "grad_norm": 2.395726442337036,
+      "learning_rate": 4.6610852197385695e-06,
+      "loss": 0.4479,
+      "step": 1534
+    },
+    {
+      "epoch": 0.29349904397705545,
+      "grad_norm": 4.810997486114502,
+      "learning_rate": 4.661499298077252e-06,
+      "loss": 0.2757,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2936902485659656,
+      "grad_norm": 1.4835280179977417,
+      "learning_rate": 4.66191310674588e-06,
+      "loss": 0.1364,
+      "step": 1536
+    },
+    {
+      "epoch": 0.29388145315487574,
+      "grad_norm": 1.9309308528900146,
+      "learning_rate": 4.662326646095474e-06,
+      "loss": 0.081,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2940726577437858,
+      "grad_norm": 3.1075916290283203,
+      "learning_rate": 4.662739916476364e-06,
+      "loss": 0.3218,
+      "step": 1538
+    },
+    {
+      "epoch": 0.29426386233269597,
+      "grad_norm": 3.229728937149048,
+      "learning_rate": 4.663152918238205e-06,
+      "loss": 0.3299,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2944550669216061,
+      "grad_norm": 2.6888089179992676,
+      "learning_rate": 4.663565651729963e-06,
+      "loss": 0.2274,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29464627151051626,
+      "grad_norm": 1.4698622226715088,
+      "learning_rate": 4.66397811729993e-06,
+      "loss": 0.0906,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2948374760994264,
+      "grad_norm": 2.7869653701782227,
+      "learning_rate": 4.664390315295716e-06,
+      "loss": 0.17,
+      "step": 1542
+    },
+    {
+      "epoch": 0.29502868068833654,
+      "grad_norm": 3.202296257019043,
+      "learning_rate": 4.664802246064258e-06,
+      "loss": 0.1493,
+      "step": 1543
+    },
+    {
+      "epoch": 0.29521988527724663,
+      "grad_norm": 1.7716546058654785,
+      "learning_rate": 4.665213909951816e-06,
+      "loss": 0.0703,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2954110898661568,
+      "grad_norm": 4.290736675262451,
+      "learning_rate": 4.66562530730398e-06,
+      "loss": 0.3951,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2956022944550669,
+      "grad_norm": 1.5324119329452515,
+      "learning_rate": 4.666036438465668e-06,
+      "loss": 0.167,
+      "step": 1546
+    },
+    {
+      "epoch": 0.29579349904397706,
+      "grad_norm": 1.529327154159546,
+      "learning_rate": 4.66644730378113e-06,
+      "loss": 0.158,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2959847036328872,
+      "grad_norm": 1.6390931606292725,
+      "learning_rate": 4.666857903593945e-06,
+      "loss": 0.136,
+      "step": 1548
+    },
+    {
+      "epoch": 0.29617590822179735,
+      "grad_norm": 1.1221572160720825,
+      "learning_rate": 4.667268238247031e-06,
+      "loss": 0.0375,
+      "step": 1549
+    },
+    {
+      "epoch": 0.29636711281070743,
+      "grad_norm": 3.3055922985076904,
+      "learning_rate": 4.667678308082639e-06,
+      "loss": 0.1424,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2965583173996176,
+      "grad_norm": 1.8237640857696533,
+      "learning_rate": 4.668088113442359e-06,
+      "loss": 0.2064,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2967495219885277,
+      "grad_norm": 2.7643344402313232,
+      "learning_rate": 4.668497654667122e-06,
+      "loss": 0.415,
+      "step": 1552
+    },
+    {
+      "epoch": 0.29694072657743786,
+      "grad_norm": 3.103980302810669,
+      "learning_rate": 4.668906932097195e-06,
+      "loss": 0.5029,
+      "step": 1553
+    },
+    {
+      "epoch": 0.297131931166348,
+      "grad_norm": 2.982830286026001,
+      "learning_rate": 4.669315946072195e-06,
+      "loss": 0.4939,
+      "step": 1554
+    },
+    {
+      "epoch": 0.29732313575525815,
+      "grad_norm": 2.102238416671753,
+      "learning_rate": 4.669724696931077e-06,
+      "loss": 0.2851,
+      "step": 1555
+    },
+    {
+      "epoch": 0.29751434034416824,
+      "grad_norm": 2.107391357421875,
+      "learning_rate": 4.670133185012147e-06,
+      "loss": 0.1171,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2977055449330784,
+      "grad_norm": 7.794981002807617,
+      "learning_rate": 4.670541410653059e-06,
+      "loss": 0.3182,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2978967495219885,
+      "grad_norm": 3.63592529296875,
+      "learning_rate": 4.6709493741908105e-06,
+      "loss": 0.5255,
+      "step": 1558
+    },
+    {
+      "epoch": 0.29808795411089867,
+      "grad_norm": 2.3166344165802,
+      "learning_rate": 4.671357075961757e-06,
+      "loss": 0.2296,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2982791586998088,
+      "grad_norm": 1.7584208250045776,
+      "learning_rate": 4.671764516301605e-06,
+      "loss": 0.2356,
+      "step": 1560
+    },
+    {
+      "epoch": 0.29847036328871895,
+      "grad_norm": 1.8798422813415527,
+      "learning_rate": 4.672171695545415e-06,
+      "loss": 0.1173,
+      "step": 1561
+    },
+    {
+      "epoch": 0.29866156787762904,
+      "grad_norm": 2.359449863433838,
+      "learning_rate": 4.672578614027603e-06,
+      "loss": 0.1093,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2988527724665392,
+      "grad_norm": 2.927279233932495,
+      "learning_rate": 4.672985272081945e-06,
+      "loss": 0.4062,
+      "step": 1563
+    },
+    {
+      "epoch": 0.29904397705544933,
+      "grad_norm": 3.0216593742370605,
+      "learning_rate": 4.673391670041575e-06,
+      "loss": 0.5828,
+      "step": 1564
+    },
+    {
+      "epoch": 0.29923518164435947,
+      "grad_norm": 2.489259719848633,
+      "learning_rate": 4.673797808238989e-06,
+      "loss": 0.3301,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2994263862332696,
+      "grad_norm": 2.547424793243408,
+      "learning_rate": 4.674203687006046e-06,
+      "loss": 0.2759,
+      "step": 1566
+    },
+    {
+      "epoch": 0.29961759082217976,
+      "grad_norm": 1.899907112121582,
+      "learning_rate": 4.674609306673967e-06,
+      "loss": 0.1262,
+      "step": 1567
+    },
+    {
+      "epoch": 0.29980879541108985,
+      "grad_norm": 1.9912757873535156,
+      "learning_rate": 4.675014667573342e-06,
+      "loss": 0.1042,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6315016746520996,
+      "learning_rate": 4.675419770034128e-06,
+      "loss": 0.0955,
+      "step": 1569
+    },
+    {
+      "epoch": 0.30019120458891013,
+      "grad_norm": 2.126676321029663,
+      "learning_rate": 4.675824614385652e-06,
+      "loss": 0.4287,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3003824091778203,
+      "grad_norm": 1.8122913837432861,
+      "learning_rate": 4.6762292009566105e-06,
+      "loss": 0.1287,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3005736137667304,
+      "grad_norm": 16.377099990844727,
+      "learning_rate": 4.676633530075071e-06,
+      "loss": 0.2717,
+      "step": 1572
+    },
+    {
+      "epoch": 0.30076481835564056,
+      "grad_norm": 2.3182194232940674,
+      "learning_rate": 4.677037602068479e-06,
+      "loss": 0.2243,
+      "step": 1573
+    },
+    {
+      "epoch": 0.30095602294455065,
+      "grad_norm": 1.5239766836166382,
+      "learning_rate": 4.677441417263654e-06,
+      "loss": 0.124,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3011472275334608,
+      "grad_norm": 2.429558753967285,
+      "learning_rate": 4.677844975986791e-06,
+      "loss": 0.156,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30133843212237094,
+      "grad_norm": 2.8561208248138428,
+      "learning_rate": 4.678248278563468e-06,
+      "loss": 0.316,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3015296367112811,
+      "grad_norm": 2.1259777545928955,
+      "learning_rate": 4.678651325318638e-06,
+      "loss": 0.1972,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3017208413001912,
+      "grad_norm": 1.0598630905151367,
+      "learning_rate": 4.67905411657664e-06,
+      "loss": 0.0997,
+      "step": 1578
+    },
+    {
+      "epoch": 0.30191204588910137,
+      "grad_norm": 2.3135368824005127,
+      "learning_rate": 4.679456652661196e-06,
+      "loss": 0.1955,
+      "step": 1579
+    },
+    {
+      "epoch": 0.30210325047801145,
+      "grad_norm": 1.8581372499465942,
+      "learning_rate": 4.679858933895413e-06,
+      "loss": 0.1274,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3022944550669216,
+      "grad_norm": 1.0900535583496094,
+      "learning_rate": 4.680260960601784e-06,
+      "loss": 0.0515,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30248565965583174,
+      "grad_norm": 2.1805429458618164,
+      "learning_rate": 4.680662733102189e-06,
+      "loss": 0.4408,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3026768642447419,
+      "grad_norm": 3.509096145629883,
+      "learning_rate": 4.681064251717901e-06,
+      "loss": 0.4217,
+      "step": 1583
+    },
+    {
+      "epoch": 0.302868068833652,
+      "grad_norm": 2.643540620803833,
+      "learning_rate": 4.681465516769583e-06,
+      "loss": 0.1418,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3030592734225621,
+      "grad_norm": 1.8429009914398193,
+      "learning_rate": 4.681866528577289e-06,
+      "loss": 0.1108,
+      "step": 1585
+    },
+    {
+      "epoch": 0.30325047801147226,
+      "grad_norm": 1.879594087600708,
+      "learning_rate": 4.6822672874604705e-06,
+      "loss": 0.2356,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3034416826003824,
+      "grad_norm": 2.853865623474121,
+      "learning_rate": 4.6826677937379745e-06,
+      "loss": 0.1654,
+      "step": 1587
+    },
+    {
+      "epoch": 0.30363288718929254,
+      "grad_norm": 1.8738490343093872,
+      "learning_rate": 4.683068047728041e-06,
+      "loss": 0.2092,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3038240917782027,
+      "grad_norm": 3.2390451431274414,
+      "learning_rate": 4.683468049748315e-06,
+      "loss": 0.528,
+      "step": 1589
+    },
+    {
+      "epoch": 0.30401529636711283,
+      "grad_norm": 2.1751222610473633,
+      "learning_rate": 4.683867800115839e-06,
+      "loss": 0.3197,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3042065009560229,
+      "grad_norm": 3.2450201511383057,
+      "learning_rate": 4.684267299147057e-06,
+      "loss": 0.4242,
+      "step": 1591
+    },
+    {
+      "epoch": 0.30439770554493306,
+      "grad_norm": 1.056480050086975,
+      "learning_rate": 4.684666547157818e-06,
+      "loss": 0.1298,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3045889101338432,
+      "grad_norm": 1.7769147157669067,
+      "learning_rate": 4.685065544463375e-06,
+      "loss": 0.1012,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30478011472275335,
+      "grad_norm": 1.4946969747543335,
+      "learning_rate": 4.685464291378389e-06,
+      "loss": 0.0821,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3049713193116635,
+      "grad_norm": 2.63144850730896,
+      "learning_rate": 4.6858627882169256e-06,
+      "loss": 0.3787,
+      "step": 1595
+    },
+    {
+      "epoch": 0.30516252390057363,
+      "grad_norm": 2.6494715213775635,
+      "learning_rate": 4.686261035292464e-06,
+      "loss": 0.3208,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3053537284894837,
+      "grad_norm": 4.6903791427612305,
+      "learning_rate": 4.6866590329178915e-06,
+      "loss": 0.363,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30554493307839387,
+      "grad_norm": 2.102142572402954,
+      "learning_rate": 4.68705678140551e-06,
+      "loss": 0.1513,
+      "step": 1598
+    },
+    {
+      "epoch": 0.305736137667304,
+      "grad_norm": 1.5582478046417236,
+      "learning_rate": 4.687454281067032e-06,
+      "loss": 0.0899,
+      "step": 1599
+    },
+    {
+      "epoch": 0.30592734225621415,
+      "grad_norm": 1.9125760793685913,
+      "learning_rate": 4.687851532213589e-06,
+      "loss": 0.0984,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3061185468451243,
+      "grad_norm": 2.396864414215088,
+      "learning_rate": 4.688248535155727e-06,
+      "loss": 0.3593,
+      "step": 1601
+    },
+    {
+      "epoch": 0.30630975143403444,
+      "grad_norm": 1.5265965461730957,
+      "learning_rate": 4.688645290203411e-06,
+      "loss": 0.0922,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3065009560229445,
+      "grad_norm": 2.4949896335601807,
+      "learning_rate": 4.689041797666025e-06,
+      "loss": 0.1977,
+      "step": 1603
+    },
+    {
+      "epoch": 0.30669216061185467,
+      "grad_norm": 2.181450605392456,
+      "learning_rate": 4.6894380578523775e-06,
+      "loss": 0.1048,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3068833652007648,
+      "grad_norm": 1.2355436086654663,
+      "learning_rate": 4.689834071070693e-06,
+      "loss": 0.0392,
+      "step": 1605
+    },
+    {
+      "epoch": 0.30707456978967496,
+      "grad_norm": 3.254880666732788,
+      "learning_rate": 4.690229837628627e-06,
+      "loss": 0.1794,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3072657743785851,
+      "grad_norm": 2.0266425609588623,
+      "learning_rate": 4.690625357833257e-06,
+      "loss": 0.1934,
+      "step": 1607
+    },
+    {
+      "epoch": 0.30745697896749524,
+      "grad_norm": 1.5074000358581543,
+      "learning_rate": 4.6910206319910875e-06,
+      "loss": 0.206,
+      "step": 1608
+    },
+    {
+      "epoch": 0.30764818355640533,
+      "grad_norm": 1.8750319480895996,
+      "learning_rate": 4.6914156604080515e-06,
+      "loss": 0.1769,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3078393881453155,
+      "grad_norm": 1.9686181545257568,
+      "learning_rate": 4.691810443389513e-06,
+      "loss": 0.1552,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3080305927342256,
+      "grad_norm": 2.02163028717041,
+      "learning_rate": 4.692204981240264e-06,
+      "loss": 0.1021,
+      "step": 1611
+    },
+    {
+      "epoch": 0.30822179732313576,
+      "grad_norm": 2.6845357418060303,
+      "learning_rate": 4.692599274264534e-06,
+      "loss": 0.1429,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3084130019120459,
+      "grad_norm": 2.7308313846588135,
+      "learning_rate": 4.692993322765983e-06,
+      "loss": 0.3626,
+      "step": 1613
+    },
+    {
+      "epoch": 0.30860420650095605,
+      "grad_norm": 1.9989246129989624,
+      "learning_rate": 4.693387127047705e-06,
+      "loss": 0.3389,
+      "step": 1614
+    },
+    {
+      "epoch": 0.30879541108986613,
+      "grad_norm": 2.3005568981170654,
+      "learning_rate": 4.693780687412236e-06,
+      "loss": 0.3273,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3089866156787763,
+      "grad_norm": 2.354796886444092,
+      "learning_rate": 4.694174004161545e-06,
+      "loss": 0.1299,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3091778202676864,
+      "grad_norm": 1.5658437013626099,
+      "learning_rate": 4.6945670775970445e-06,
+      "loss": 0.0881,
+      "step": 1617
+    },
+    {
+      "epoch": 0.30936902485659656,
+      "grad_norm": 2.1076741218566895,
+      "learning_rate": 4.694959908019585e-06,
+      "loss": 0.2679,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3095602294455067,
+      "grad_norm": 2.51861572265625,
+      "learning_rate": 4.6953524957294615e-06,
+      "loss": 0.1314,
+      "step": 1619
+    },
+    {
+      "epoch": 0.30975143403441685,
+      "grad_norm": 2.3451156616210938,
+      "learning_rate": 4.695744841026411e-06,
+      "loss": 0.2426,
+      "step": 1620
+    },
+    {
+      "epoch": 0.30994263862332694,
+      "grad_norm": 2.2577333450317383,
+      "learning_rate": 4.696136944209617e-06,
+      "loss": 0.3822,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3101338432122371,
+      "grad_norm": 2.172588348388672,
+      "learning_rate": 4.696528805577708e-06,
+      "loss": 0.1405,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3103250478011472,
+      "grad_norm": 3.17950177192688,
+      "learning_rate": 4.696920425428762e-06,
+      "loss": 0.3541,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31051625239005737,
+      "grad_norm": 4.15213680267334,
+      "learning_rate": 4.6973118040603045e-06,
+      "loss": 0.1327,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3107074569789675,
+      "grad_norm": 1.5607051849365234,
+      "learning_rate": 4.697702941769314e-06,
+      "loss": 0.0799,
+      "step": 1625
+    },
+    {
+      "epoch": 0.31089866156787765,
+      "grad_norm": 1.972217082977295,
+      "learning_rate": 4.698093838852218e-06,
+      "loss": 0.3492,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31108986615678774,
+      "grad_norm": 2.502187728881836,
+      "learning_rate": 4.6984844956048994e-06,
+      "loss": 0.3884,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3112810707456979,
+      "grad_norm": 3.3736958503723145,
+      "learning_rate": 4.698874912322695e-06,
+      "loss": 0.2595,
+      "step": 1628
+    },
+    {
+      "epoch": 0.31147227533460803,
+      "grad_norm": 1.729385495185852,
+      "learning_rate": 4.699265089300396e-06,
+      "loss": 0.0982,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31166347992351817,
+      "grad_norm": 9.559845924377441,
+      "learning_rate": 4.699655026832254e-06,
+      "loss": 0.5083,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3118546845124283,
+      "grad_norm": 2.132229804992676,
+      "learning_rate": 4.700044725211977e-06,
+      "loss": 0.1439,
+      "step": 1631
+    },
+    {
+      "epoch": 0.31204588910133846,
+      "grad_norm": 2.046917676925659,
+      "learning_rate": 4.700434184732733e-06,
+      "loss": 0.345,
+      "step": 1632
+    },
+    {
+      "epoch": 0.31223709369024855,
+      "grad_norm": 2.525454044342041,
+      "learning_rate": 4.700823405687152e-06,
+      "loss": 0.4536,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3124282982791587,
+      "grad_norm": 2.227076768875122,
+      "learning_rate": 4.7012123883673265e-06,
+      "loss": 0.2261,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31261950286806883,
+      "grad_norm": 1.19818115234375,
+      "learning_rate": 4.701601133064812e-06,
+      "loss": 0.0679,
+      "step": 1635
+    },
+    {
+      "epoch": 0.312810707456979,
+      "grad_norm": 2.9948644638061523,
+      "learning_rate": 4.701989640070631e-06,
+      "loss": 0.1852,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3130019120458891,
+      "grad_norm": 2.5695416927337646,
+      "learning_rate": 4.70237790967527e-06,
+      "loss": 0.3184,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31319311663479926,
+      "grad_norm": 2.033766508102417,
+      "learning_rate": 4.702765942168687e-06,
+      "loss": 0.1388,
+      "step": 1638
+    },
+    {
+      "epoch": 0.31338432122370935,
+      "grad_norm": 2.1387429237365723,
+      "learning_rate": 4.703153737840303e-06,
+      "loss": 0.2016,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3135755258126195,
+      "grad_norm": 1.2150760889053345,
+      "learning_rate": 4.703541296979016e-06,
+      "loss": 0.1351,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31376673040152964,
+      "grad_norm": 1.4617748260498047,
+      "learning_rate": 4.703928619873192e-06,
+      "loss": 0.0998,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3139579349904398,
+      "grad_norm": 2.583080291748047,
+      "learning_rate": 4.704315706810671e-06,
+      "loss": 0.1376,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3141491395793499,
+      "grad_norm": 1.4851268529891968,
+      "learning_rate": 4.7047025580787675e-06,
+      "loss": 0.0733,
+      "step": 1643
+    },
+    {
+      "epoch": 0.31434034416826,
+      "grad_norm": 2.7585723400115967,
+      "learning_rate": 4.7050891739642704e-06,
+      "loss": 0.1577,
+      "step": 1644
+    },
+    {
+      "epoch": 0.31453154875717015,
+      "grad_norm": 2.7079718112945557,
+      "learning_rate": 4.705475554753447e-06,
+      "loss": 0.2468,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3147227533460803,
+      "grad_norm": 2.084747791290283,
+      "learning_rate": 4.705861700732041e-06,
+      "loss": 0.1464,
+      "step": 1646
+    },
+    {
+      "epoch": 0.31491395793499044,
+      "grad_norm": 3.7070984840393066,
+      "learning_rate": 4.706247612185277e-06,
+      "loss": 0.1427,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3151051625239006,
+      "grad_norm": 1.74429190158844,
+      "learning_rate": 4.70663328939786e-06,
+      "loss": 0.1169,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3152963671128107,
+      "grad_norm": 2.154435157775879,
+      "learning_rate": 4.707018732653974e-06,
+      "loss": 0.2084,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3154875717017208,
+      "grad_norm": 2.1565818786621094,
+      "learning_rate": 4.707403942237291e-06,
+      "loss": 0.1462,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31567877629063096,
+      "grad_norm": 3.50270676612854,
+      "learning_rate": 4.707788918430965e-06,
+      "loss": 0.6104,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3158699808795411,
+      "grad_norm": 1.7462185621261597,
+      "learning_rate": 4.708173661517635e-06,
+      "loss": 0.1656,
+      "step": 1652
+    },
+    {
+      "epoch": 0.31606118546845124,
+      "grad_norm": 1.6207178831100464,
+      "learning_rate": 4.708558171779426e-06,
+      "loss": 0.1279,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3162523900573614,
+      "grad_norm": 2.2919764518737793,
+      "learning_rate": 4.7089424494979555e-06,
+      "loss": 0.2565,
+      "step": 1654
+    },
+    {
+      "epoch": 0.31644359464627153,
+      "grad_norm": 2.5331718921661377,
+      "learning_rate": 4.709326494954326e-06,
+      "loss": 0.1646,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3166347992351816,
+      "grad_norm": 2.5295658111572266,
+      "learning_rate": 4.709710308429132e-06,
+      "loss": 0.1328,
+      "step": 1656
+    },
+    {
+      "epoch": 0.31682600382409176,
+      "grad_norm": 2.147109031677246,
+      "learning_rate": 4.710093890202459e-06,
+      "loss": 0.2947,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3170172084130019,
+      "grad_norm": 2.6603047847747803,
+      "learning_rate": 4.71047724055389e-06,
+      "loss": 0.4228,
+      "step": 1658
+    },
+    {
+      "epoch": 0.31720841300191205,
+      "grad_norm": 3.075491189956665,
+      "learning_rate": 4.710860359762494e-06,
+      "loss": 0.4414,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3173996175908222,
+      "grad_norm": 2.007510185241699,
+      "learning_rate": 4.711243248106844e-06,
+      "loss": 0.1255,
+      "step": 1660
+    },
+    {
+      "epoch": 0.31759082217973233,
+      "grad_norm": 1.8735504150390625,
+      "learning_rate": 4.711625905865004e-06,
+      "loss": 0.1407,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3177820267686424,
+      "grad_norm": 1.5158590078353882,
+      "learning_rate": 4.7120083333145385e-06,
+      "loss": 0.0653,
+      "step": 1662
+    },
+    {
+      "epoch": 0.31797323135755257,
+      "grad_norm": 2.0995800495147705,
+      "learning_rate": 4.712390530732511e-06,
+      "loss": 0.3412,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3181644359464627,
+      "grad_norm": 2.8367748260498047,
+      "learning_rate": 4.712772498395484e-06,
+      "loss": 0.3504,
+      "step": 1664
+    },
+    {
+      "epoch": 0.31835564053537285,
+      "grad_norm": 2.228076696395874,
+      "learning_rate": 4.713154236579523e-06,
+      "loss": 0.2741,
+      "step": 1665
+    },
+    {
+      "epoch": 0.318546845124283,
+      "grad_norm": 2.8032071590423584,
+      "learning_rate": 4.713535745560195e-06,
+      "loss": 0.4121,
+      "step": 1666
+    },
+    {
+      "epoch": 0.31873804971319314,
+      "grad_norm": 2.5112833976745605,
+      "learning_rate": 4.713917025612572e-06,
+      "loss": 0.4035,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3189292543021032,
+      "grad_norm": 4.7532758712768555,
+      "learning_rate": 4.714298077011231e-06,
+      "loss": 0.1276,
+      "step": 1668
+    },
+    {
+      "epoch": 0.31912045889101337,
+      "grad_norm": 2.9184582233428955,
+      "learning_rate": 4.714678900030255e-06,
+      "loss": 0.1993,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3193116634799235,
+      "grad_norm": 3.2624402046203613,
+      "learning_rate": 4.715059494943234e-06,
+      "loss": 0.5957,
+      "step": 1670
+    },
+    {
+      "epoch": 0.31950286806883366,
+      "grad_norm": 2.0953452587127686,
+      "learning_rate": 4.715439862023267e-06,
+      "loss": 0.2429,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3196940726577438,
+      "grad_norm": 2.3828635215759277,
+      "learning_rate": 4.715820001542965e-06,
+      "loss": 0.2674,
+      "step": 1672
+    },
+    {
+      "epoch": 0.31988527724665394,
+      "grad_norm": 1.713576078414917,
+      "learning_rate": 4.716199913774444e-06,
+      "loss": 0.1027,
+      "step": 1673
+    },
+    {
+      "epoch": 0.32007648183556403,
+      "grad_norm": 2.204185724258423,
+      "learning_rate": 4.71657959898934e-06,
+      "loss": 0.0693,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3202676864244742,
+      "grad_norm": 2.4515421390533447,
+      "learning_rate": 4.716959057458796e-06,
+      "loss": 0.2329,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3204588910133843,
+      "grad_norm": 2.277780055999756,
+      "learning_rate": 4.717338289453474e-06,
+      "loss": 0.2594,
+      "step": 1676
+    },
+    {
+      "epoch": 0.32065009560229446,
+      "grad_norm": 1.873704433441162,
+      "learning_rate": 4.717717295243549e-06,
+      "loss": 0.3301,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3208413001912046,
+      "grad_norm": 1.6470894813537598,
+      "learning_rate": 4.718096075098712e-06,
+      "loss": 0.1803,
+      "step": 1678
+    },
+    {
+      "epoch": 0.32103250478011475,
+      "grad_norm": 2.8109920024871826,
+      "learning_rate": 4.718474629288177e-06,
+      "loss": 0.1192,
+      "step": 1679
+    },
+    {
+      "epoch": 0.32122370936902483,
+      "grad_norm": 2.7735977172851562,
+      "learning_rate": 4.71885295808067e-06,
+      "loss": 0.3198,
+      "step": 1680
+    },
+    {
+      "epoch": 0.321414913957935,
+      "grad_norm": 3.0144803524017334,
+      "learning_rate": 4.719231061744443e-06,
+      "loss": 0.1269,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3216061185468451,
+      "grad_norm": 2.3814454078674316,
+      "learning_rate": 4.7196089405472675e-06,
+      "loss": 0.1253,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32179732313575526,
+      "grad_norm": 2.3879141807556152,
+      "learning_rate": 4.719986594756435e-06,
+      "loss": 0.4084,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3219885277246654,
+      "grad_norm": 1.6740145683288574,
+      "learning_rate": 4.720364024638766e-06,
+      "loss": 0.0784,
+      "step": 1684
+    },
+    {
+      "epoch": 0.32217973231357555,
+      "grad_norm": 2.241508722305298,
+      "learning_rate": 4.7207412304606015e-06,
+      "loss": 0.298,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32237093690248564,
+      "grad_norm": 1.0748635530471802,
+      "learning_rate": 4.72111821248781e-06,
+      "loss": 0.048,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3225621414913958,
+      "grad_norm": 2.42177677154541,
+      "learning_rate": 4.721494970985786e-06,
+      "loss": 0.1027,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3227533460803059,
+      "grad_norm": 2.5916969776153564,
+      "learning_rate": 4.721871506219455e-06,
+      "loss": 0.1818,
+      "step": 1688
+    },
+    {
+      "epoch": 0.32294455066921607,
+      "grad_norm": 1.406157374382019,
+      "learning_rate": 4.7222478184532676e-06,
+      "loss": 0.1544,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3231357552581262,
+      "grad_norm": 1.2543416023254395,
+      "learning_rate": 4.722623907951209e-06,
+      "loss": 0.1498,
+      "step": 1690
+    },
+    {
+      "epoch": 0.32332695984703635,
+      "grad_norm": 1.768796443939209,
+      "learning_rate": 4.722999774976792e-06,
+      "loss": 0.2372,
+      "step": 1691
+    },
+    {
+      "epoch": 0.32351816443594644,
+      "grad_norm": 2.0685126781463623,
+      "learning_rate": 4.723375419793066e-06,
+      "loss": 0.1395,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3237093690248566,
+      "grad_norm": 3.7179152965545654,
+      "learning_rate": 4.723750842662612e-06,
+      "loss": 0.1303,
+      "step": 1693
+    },
+    {
+      "epoch": 0.32390057361376673,
+      "grad_norm": 2.238060712814331,
+      "learning_rate": 4.7241260438475445e-06,
+      "loss": 0.2068,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3240917782026769,
+      "grad_norm": 2.760972023010254,
+      "learning_rate": 4.724501023609517e-06,
+      "loss": 0.4147,
+      "step": 1695
+    },
+    {
+      "epoch": 0.324282982791587,
+      "grad_norm": 3.159468412399292,
+      "learning_rate": 4.724875782209718e-06,
+      "loss": 0.6906,
+      "step": 1696
+    },
+    {
+      "epoch": 0.32447418738049716,
+      "grad_norm": 2.3904199600219727,
+      "learning_rate": 4.725250319908874e-06,
+      "loss": 0.4181,
+      "step": 1697
+    },
+    {
+      "epoch": 0.32466539196940725,
+      "grad_norm": 2.083958625793457,
+      "learning_rate": 4.725624636967252e-06,
+      "loss": 0.1321,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3248565965583174,
+      "grad_norm": 2.2649729251861572,
+      "learning_rate": 4.725998733644659e-06,
+      "loss": 0.1378,
+      "step": 1699
+    },
+    {
+      "epoch": 0.32504780114722753,
+      "grad_norm": 2.6033074855804443,
+      "learning_rate": 4.726372610200442e-06,
+      "loss": 0.1434,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3252390057361377,
+      "grad_norm": 2.7515151500701904,
+      "learning_rate": 4.726746266893492e-06,
+      "loss": 0.4188,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3254302103250478,
+      "grad_norm": 2.8080029487609863,
+      "learning_rate": 4.727119703982244e-06,
+      "loss": 0.3093,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3256214149139579,
+      "grad_norm": 3.603713274002075,
+      "learning_rate": 4.727492921724675e-06,
+      "loss": 0.4832,
+      "step": 1703
+    },
+    {
+      "epoch": 0.32581261950286805,
+      "grad_norm": 2.3233444690704346,
+      "learning_rate": 4.72786592037831e-06,
+      "loss": 0.3555,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3260038240917782,
+      "grad_norm": 3.610562801361084,
+      "learning_rate": 4.728238700200221e-06,
+      "loss": 0.2113,
+      "step": 1705
+    },
+    {
+      "epoch": 0.32619502868068834,
+      "grad_norm": 1.6778076887130737,
+      "learning_rate": 4.728611261447025e-06,
+      "loss": 0.0959,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3263862332695985,
+      "grad_norm": 2.616176128387451,
+      "learning_rate": 4.728983604374891e-06,
+      "loss": 0.3212,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3265774378585086,
+      "grad_norm": 2.2144649028778076,
+      "learning_rate": 4.7293557292395365e-06,
+      "loss": 0.3198,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3267686424474187,
+      "grad_norm": 1.6588743925094604,
+      "learning_rate": 4.72972763629623e-06,
+      "loss": 0.1418,
+      "step": 1709
+    },
+    {
+      "epoch": 0.32695984703632885,
+      "grad_norm": 3.0176148414611816,
+      "learning_rate": 4.730099325799792e-06,
+      "loss": 0.2423,
+      "step": 1710
+    },
+    {
+      "epoch": 0.327151051625239,
+      "grad_norm": 1.894723653793335,
+      "learning_rate": 4.730470798004597e-06,
+      "loss": 0.1414,
+      "step": 1711
+    },
+    {
+      "epoch": 0.32734225621414914,
+      "grad_norm": 2.1323254108428955,
+      "learning_rate": 4.730842053164572e-06,
+      "loss": 0.0995,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3275334608030593,
+      "grad_norm": 3.2055537700653076,
+      "learning_rate": 4.7312130915332e-06,
+      "loss": 0.3597,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3277246653919694,
+      "grad_norm": 2.552307605743408,
+      "learning_rate": 4.731583913363522e-06,
+      "loss": 0.1933,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3279158699808795,
+      "grad_norm": 1.2854914665222168,
+      "learning_rate": 4.731954518908132e-06,
+      "loss": 0.2917,
+      "step": 1715
+    },
+    {
+      "epoch": 0.32810707456978966,
+      "grad_norm": 2.1804897785186768,
+      "learning_rate": 4.732324908419186e-06,
+      "loss": 0.2287,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3282982791586998,
+      "grad_norm": 1.7139678001403809,
+      "learning_rate": 4.732695082148399e-06,
+      "loss": 0.1392,
+      "step": 1717
+    },
+    {
+      "epoch": 0.32848948374760994,
+      "grad_norm": 2.053713321685791,
+      "learning_rate": 4.733065040347042e-06,
+      "loss": 0.1129,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3286806883365201,
+      "grad_norm": 1.8551483154296875,
+      "learning_rate": 4.733434783265955e-06,
+      "loss": 0.096,
+      "step": 1719
+    },
+    {
+      "epoch": 0.32887189292543023,
+      "grad_norm": 2.319152593612671,
+      "learning_rate": 4.733804311155533e-06,
+      "loss": 0.5004,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3290630975143403,
+      "grad_norm": 2.038278102874756,
+      "learning_rate": 4.734173624265738e-06,
+      "loss": 0.1583,
+      "step": 1721
+    },
+    {
+      "epoch": 0.32925430210325046,
+      "grad_norm": 1.7227160930633545,
+      "learning_rate": 4.734542722846097e-06,
+      "loss": 0.2504,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3294455066921606,
+      "grad_norm": 2.386646270751953,
+      "learning_rate": 4.734911607145701e-06,
+      "loss": 0.3351,
+      "step": 1723
+    },
+    {
+      "epoch": 0.32963671128107075,
+      "grad_norm": 2.143439769744873,
+      "learning_rate": 4.735280277413207e-06,
+      "loss": 0.1811,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3298279158699809,
+      "grad_norm": 2.0049850940704346,
+      "learning_rate": 4.735648733896841e-06,
+      "loss": 0.1041,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33001912045889104,
+      "grad_norm": 2.2885913848876953,
+      "learning_rate": 4.736016976844395e-06,
+      "loss": 0.2313,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3302103250478011,
+      "grad_norm": 3.8230414390563965,
+      "learning_rate": 4.736385006503233e-06,
+      "loss": 0.6751,
+      "step": 1727
+    },
+    {
+      "epoch": 0.33040152963671127,
+      "grad_norm": 2.540400266647339,
+      "learning_rate": 4.7367528231202895e-06,
+      "loss": 0.2008,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3305927342256214,
+      "grad_norm": 2.929203987121582,
+      "learning_rate": 4.737120426942068e-06,
+      "loss": 0.3909,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33078393881453155,
+      "grad_norm": 1.263774037361145,
+      "learning_rate": 4.737487818214645e-06,
+      "loss": 0.1669,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3309751434034417,
+      "grad_norm": 1.2126981019973755,
+      "learning_rate": 4.737854997183673e-06,
+      "loss": 0.1578,
+      "step": 1731
+    },
+    {
+      "epoch": 0.33116634799235184,
+      "grad_norm": 3.105952739715576,
+      "learning_rate": 4.738221964094376e-06,
+      "loss": 0.4275,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3313575525812619,
+      "grad_norm": 1.326631784439087,
+      "learning_rate": 4.738588719191555e-06,
+      "loss": 0.1018,
+      "step": 1733
+    },
+    {
+      "epoch": 0.33154875717017207,
+      "grad_norm": 1.790029764175415,
+      "learning_rate": 4.738955262719585e-06,
+      "loss": 0.1735,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3317399617590822,
+      "grad_norm": 2.4458529949188232,
+      "learning_rate": 4.739321594922423e-06,
+      "loss": 0.1493,
+      "step": 1735
+    },
+    {
+      "epoch": 0.33193116634799236,
+      "grad_norm": 1.3313581943511963,
+      "learning_rate": 4.7396877160436e-06,
+      "loss": 0.048,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3321223709369025,
+      "grad_norm": 1.1828702688217163,
+      "learning_rate": 4.740053626326225e-06,
+      "loss": 0.0575,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33231357552581264,
+      "grad_norm": 3.091031551361084,
+      "learning_rate": 4.740419326012995e-06,
+      "loss": 0.4912,
+      "step": 1738
+    },
+    {
+      "epoch": 0.33250478011472273,
+      "grad_norm": 2.2998857498168945,
+      "learning_rate": 4.740784815346178e-06,
+      "loss": 0.2701,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3326959847036329,
+      "grad_norm": 1.3561292886734009,
+      "learning_rate": 4.741150094567632e-06,
+      "loss": 0.2125,
+      "step": 1740
+    },
+    {
+      "epoch": 0.332887189292543,
+      "grad_norm": 0.8512531518936157,
+      "learning_rate": 4.7415151639187964e-06,
+      "loss": 0.077,
+      "step": 1741
+    },
+    {
+      "epoch": 0.33307839388145316,
+      "grad_norm": 1.6298433542251587,
+      "learning_rate": 4.741880023640691e-06,
+      "loss": 0.1782,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3332695984703633,
+      "grad_norm": 3.3909101486206055,
+      "learning_rate": 4.742244673973925e-06,
+      "loss": 0.2529,
+      "step": 1743
+    },
+    {
+      "epoch": 0.33346080305927345,
+      "grad_norm": 2.4966087341308594,
+      "learning_rate": 4.742609115158691e-06,
+      "loss": 0.2019,
+      "step": 1744
+    },
+    {
+      "epoch": 0.33365200764818354,
+      "grad_norm": 4.023033142089844,
+      "learning_rate": 4.74297334743477e-06,
+      "loss": 0.4731,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3338432122370937,
+      "grad_norm": 3.0190768241882324,
+      "learning_rate": 4.74333737104153e-06,
+      "loss": 0.4793,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3340344168260038,
+      "grad_norm": 3.289560556411743,
+      "learning_rate": 4.743701186217929e-06,
+      "loss": 0.1652,
+      "step": 1747
+    },
+    {
+      "epoch": 0.33422562141491396,
+      "grad_norm": 2.139427423477173,
+      "learning_rate": 4.744064793202513e-06,
+      "loss": 0.137,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3344168260038241,
+      "grad_norm": 3.575605869293213,
+      "learning_rate": 4.74442819223342e-06,
+      "loss": 0.3851,
+      "step": 1749
+    },
+    {
+      "epoch": 0.33460803059273425,
+      "grad_norm": 3.1366186141967773,
+      "learning_rate": 4.744791383548379e-06,
+      "loss": 0.1949,
+      "step": 1750
+    },
+    {
+      "epoch": 0.33479923518164434,
+      "grad_norm": 1.9750595092773438,
+      "learning_rate": 4.745154367384712e-06,
+      "loss": 0.1648,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3349904397705545,
+      "grad_norm": 1.4047598838806152,
+      "learning_rate": 4.745517143979335e-06,
+      "loss": 0.1517,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3351816443594646,
+      "grad_norm": 1.289894700050354,
+      "learning_rate": 4.7458797135687565e-06,
+      "loss": 0.1399,
+      "step": 1753
+    },
+    {
+      "epoch": 0.33537284894837477,
+      "grad_norm": 0.9425712823867798,
+      "learning_rate": 4.746242076389082e-06,
+      "loss": 0.1055,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3355640535372849,
+      "grad_norm": 2.4297313690185547,
+      "learning_rate": 4.746604232676014e-06,
+      "loss": 0.1552,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33575525812619506,
+      "grad_norm": 2.6048977375030518,
+      "learning_rate": 4.746966182664851e-06,
+      "loss": 0.1051,
+      "step": 1756
+    },
+    {
+      "epoch": 0.33594646271510514,
+      "grad_norm": 3.050530433654785,
+      "learning_rate": 4.747327926590489e-06,
+      "loss": 0.3244,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3361376673040153,
+      "grad_norm": 2.425839900970459,
+      "learning_rate": 4.747689464687424e-06,
+      "loss": 0.4203,
+      "step": 1758
+    },
+    {
+      "epoch": 0.33632887189292543,
+      "grad_norm": 1.1890416145324707,
+      "learning_rate": 4.748050797189752e-06,
+      "loss": 0.1145,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3365200764818356,
+      "grad_norm": 2.5795624256134033,
+      "learning_rate": 4.74841192433117e-06,
+      "loss": 0.2012,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3367112810707457,
+      "grad_norm": 1.8465324640274048,
+      "learning_rate": 4.7487728463449755e-06,
+      "loss": 0.1086,
+      "step": 1761
+    },
+    {
+      "epoch": 0.33690248565965586,
+      "grad_norm": 1.6357309818267822,
+      "learning_rate": 4.749133563464071e-06,
+      "loss": 0.1407,
+      "step": 1762
+    },
+    {
+      "epoch": 0.33709369024856595,
+      "grad_norm": 1.5176138877868652,
+      "learning_rate": 4.749494075920959e-06,
+      "loss": 0.0951,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3372848948374761,
+      "grad_norm": 3.06494140625,
+      "learning_rate": 4.749854383947751e-06,
+      "loss": 0.453,
+      "step": 1764
+    },
+    {
+      "epoch": 0.33747609942638623,
+      "grad_norm": 3.943192481994629,
+      "learning_rate": 4.7502144877761604e-06,
+      "loss": 0.5386,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3376673040152964,
+      "grad_norm": 1.6208007335662842,
+      "learning_rate": 4.750574387637508e-06,
+      "loss": 0.1078,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3378585086042065,
+      "grad_norm": 1.6230381727218628,
+      "learning_rate": 4.750934083762721e-06,
+      "loss": 0.0768,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3380497131931166,
+      "grad_norm": 1.910230040550232,
+      "learning_rate": 4.751293576382336e-06,
+      "loss": 0.1267,
+      "step": 1768
+    },
+    {
+      "epoch": 0.33824091778202675,
+      "grad_norm": 2.327993869781494,
+      "learning_rate": 4.751652865726499e-06,
+      "loss": 0.1338,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3384321223709369,
+      "grad_norm": 2.1939034461975098,
+      "learning_rate": 4.752011952024963e-06,
+      "loss": 0.2042,
+      "step": 1770
+    },
+    {
+      "epoch": 0.33862332695984704,
+      "grad_norm": 1.9725449085235596,
+      "learning_rate": 4.752370835507094e-06,
+      "loss": 0.2497,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3388145315487572,
+      "grad_norm": 3.0306098461151123,
+      "learning_rate": 4.752729516401868e-06,
+      "loss": 0.3951,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3390057361376673,
+      "grad_norm": 2.3561394214630127,
+      "learning_rate": 4.753087994937877e-06,
+      "loss": 0.2254,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3391969407265774,
+      "grad_norm": 2.5845770835876465,
+      "learning_rate": 4.753446271343321e-06,
+      "loss": 0.2025,
+      "step": 1774
+    },
+    {
+      "epoch": 0.33938814531548755,
+      "grad_norm": 2.2755463123321533,
+      "learning_rate": 4.753804345846018e-06,
+      "loss": 0.1169,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3395793499043977,
+      "grad_norm": 2.320425271987915,
+      "learning_rate": 4.754162218673401e-06,
+      "loss": 0.3507,
+      "step": 1776
+    },
+    {
+      "epoch": 0.33977055449330784,
+      "grad_norm": 1.7574282884597778,
+      "learning_rate": 4.754519890052516e-06,
+      "loss": 0.3012,
+      "step": 1777
+    },
+    {
+      "epoch": 0.339961759082218,
+      "grad_norm": 1.9989606142044067,
+      "learning_rate": 4.75487736021003e-06,
+      "loss": 0.2198,
+      "step": 1778
+    },
+    {
+      "epoch": 0.34015296367112813,
+      "grad_norm": 1.8190861940383911,
+      "learning_rate": 4.7552346293722235e-06,
+      "loss": 0.1237,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3403441682600382,
+      "grad_norm": 1.9230531454086304,
+      "learning_rate": 4.755591697764998e-06,
+      "loss": 0.1621,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34053537284894836,
+      "grad_norm": 2.197153091430664,
+      "learning_rate": 4.755948565613874e-06,
+      "loss": 0.1253,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3407265774378585,
+      "grad_norm": 2.13446044921875,
+      "learning_rate": 4.756305233143992e-06,
+      "loss": 0.2251,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34091778202676865,
+      "grad_norm": 1.634322166442871,
+      "learning_rate": 4.756661700580113e-06,
+      "loss": 0.1433,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3411089866156788,
+      "grad_norm": 1.0584380626678467,
+      "learning_rate": 4.757017968146622e-06,
+      "loss": 0.0601,
+      "step": 1784
+    },
+    {
+      "epoch": 0.34130019120458893,
+      "grad_norm": 2.613311529159546,
+      "learning_rate": 4.757374036067523e-06,
+      "loss": 0.3052,
+      "step": 1785
+    },
+    {
+      "epoch": 0.341491395793499,
+      "grad_norm": 2.105159044265747,
+      "learning_rate": 4.757729904566448e-06,
+      "loss": 0.0758,
+      "step": 1786
+    },
+    {
+      "epoch": 0.34168260038240916,
+      "grad_norm": 2.2200515270233154,
+      "learning_rate": 4.75808557386665e-06,
+      "loss": 0.2331,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3418738049713193,
+      "grad_norm": 1.8687654733657837,
+      "learning_rate": 4.75844104419101e-06,
+      "loss": 0.1472,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34206500956022945,
+      "grad_norm": 3.0916874408721924,
+      "learning_rate": 4.758796315762033e-06,
+      "loss": 0.3935,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3422562141491396,
+      "grad_norm": 2.0571913719177246,
+      "learning_rate": 4.759151388801852e-06,
+      "loss": 0.1408,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34244741873804974,
+      "grad_norm": 2.078958511352539,
+      "learning_rate": 4.759506263532227e-06,
+      "loss": 0.2162,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3426386233269598,
+      "grad_norm": 1.2904046773910522,
+      "learning_rate": 4.759860940174549e-06,
+      "loss": 0.1126,
+      "step": 1792
+    },
+    {
+      "epoch": 0.34282982791586997,
+      "grad_norm": 1.762441635131836,
+      "learning_rate": 4.760215418949835e-06,
+      "loss": 0.1004,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3430210325047801,
+      "grad_norm": 1.8758217096328735,
+      "learning_rate": 4.760569700078735e-06,
+      "loss": 0.0913,
+      "step": 1794
+    },
+    {
+      "epoch": 0.34321223709369025,
+      "grad_norm": 2.6033754348754883,
+      "learning_rate": 4.760923783781529e-06,
+      "loss": 0.2553,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3434034416826004,
+      "grad_norm": 3.3962483406066895,
+      "learning_rate": 4.76127767027813e-06,
+      "loss": 0.5496,
+      "step": 1796
+    },
+    {
+      "epoch": 0.34359464627151054,
+      "grad_norm": 1.7760767936706543,
+      "learning_rate": 4.7616313597880805e-06,
+      "loss": 0.223,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3437858508604206,
+      "grad_norm": 2.5091898441314697,
+      "learning_rate": 4.761984852530561e-06,
+      "loss": 0.0785,
+      "step": 1798
+    },
+    {
+      "epoch": 0.34397705544933077,
+      "grad_norm": 2.1788854598999023,
+      "learning_rate": 4.762338148724385e-06,
+      "loss": 0.2495,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3441682600382409,
+      "grad_norm": 2.8695642948150635,
+      "learning_rate": 4.762691248587998e-06,
+      "loss": 0.1116,
+      "step": 1800
+    },
+    {
+      "epoch": 0.34435946462715106,
+      "grad_norm": 2.591621160507202,
+      "learning_rate": 4.763044152339487e-06,
+      "loss": 0.5715,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3445506692160612,
+      "grad_norm": 2.8537497520446777,
+      "learning_rate": 4.76339686019657e-06,
+      "loss": 0.4617,
+      "step": 1802
+    },
+    {
+      "epoch": 0.34474187380497134,
+      "grad_norm": 2.406893014907837,
+      "learning_rate": 4.763749372376608e-06,
+      "loss": 0.0832,
+      "step": 1803
+    },
+    {
+      "epoch": 0.34493307839388143,
+      "grad_norm": 1.8675955533981323,
+      "learning_rate": 4.764101689096597e-06,
+      "loss": 0.1626,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3451242829827916,
+      "grad_norm": 3.17629337310791,
+      "learning_rate": 4.7644538105731735e-06,
+      "loss": 0.3307,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3453154875717017,
+      "grad_norm": 2.8467295169830322,
+      "learning_rate": 4.764805737022614e-06,
+      "loss": 0.2995,
+      "step": 1806
+    },
+    {
+      "epoch": 0.34550669216061186,
+      "grad_norm": 2.333732843399048,
+      "learning_rate": 4.765157468660835e-06,
+      "loss": 0.3135,
+      "step": 1807
+    },
+    {
+      "epoch": 0.345697896749522,
+      "grad_norm": 3.3674793243408203,
+      "learning_rate": 4.7655090057033955e-06,
+      "loss": 0.651,
+      "step": 1808
+    },
+    {
+      "epoch": 0.34588910133843215,
+      "grad_norm": 2.2502448558807373,
+      "learning_rate": 4.7658603483654965e-06,
+      "loss": 0.2983,
+      "step": 1809
+    },
+    {
+      "epoch": 0.34608030592734224,
+      "grad_norm": 2.531153917312622,
+      "learning_rate": 4.7662114968619835e-06,
+      "loss": 0.3743,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3462715105162524,
+      "grad_norm": 2.929173231124878,
+      "learning_rate": 4.766562451407343e-06,
+      "loss": 0.2767,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3464627151051625,
+      "grad_norm": 2.2841217517852783,
+      "learning_rate": 4.766913212215711e-06,
+      "loss": 0.1143,
+      "step": 1812
+    },
+    {
+      "epoch": 0.34665391969407267,
+      "grad_norm": 1.7222204208374023,
+      "learning_rate": 4.767263779500863e-06,
+      "loss": 0.118,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3468451242829828,
+      "grad_norm": 2.1262247562408447,
+      "learning_rate": 4.767614153476226e-06,
+      "loss": 0.1722,
+      "step": 1814
+    },
+    {
+      "epoch": 0.34703632887189295,
+      "grad_norm": 1.8248090744018555,
+      "learning_rate": 4.7679643343548724e-06,
+      "loss": 0.1271,
+      "step": 1815
+    },
+    {
+      "epoch": 0.34722753346080304,
+      "grad_norm": 1.4642949104309082,
+      "learning_rate": 4.768314322349521e-06,
+      "loss": 0.0987,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3474187380497132,
+      "grad_norm": 2.0707128047943115,
+      "learning_rate": 4.768664117672543e-06,
+      "loss": 0.0991,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3476099426386233,
+      "grad_norm": 1.105255126953125,
+      "learning_rate": 4.769013720535954e-06,
+      "loss": 0.0739,
+      "step": 1818
+    },
+    {
+      "epoch": 0.34780114722753347,
+      "grad_norm": 1.9014976024627686,
+      "learning_rate": 4.769363131151425e-06,
+      "loss": 0.1378,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3479923518164436,
+      "grad_norm": 2.5240933895111084,
+      "learning_rate": 4.769712349730274e-06,
+      "loss": 0.4471,
+      "step": 1820
+    },
+    {
+      "epoch": 0.34818355640535376,
+      "grad_norm": 1.6601219177246094,
+      "learning_rate": 4.770061376483473e-06,
+      "loss": 0.1254,
+      "step": 1821
+    },
+    {
+      "epoch": 0.34837476099426384,
+      "grad_norm": 2.0070979595184326,
+      "learning_rate": 4.770410211621644e-06,
+      "loss": 0.1297,
+      "step": 1822
+    },
+    {
+      "epoch": 0.348565965583174,
+      "grad_norm": 2.647048234939575,
+      "learning_rate": 4.7707588553550665e-06,
+      "loss": 0.2616,
+      "step": 1823
+    },
+    {
+      "epoch": 0.34875717017208413,
+      "grad_norm": 1.199033260345459,
+      "learning_rate": 4.77110730789367e-06,
+      "loss": 0.0857,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3489483747609943,
+      "grad_norm": 2.3734302520751953,
+      "learning_rate": 4.771455569447043e-06,
+      "loss": 0.0676,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3491395793499044,
+      "grad_norm": 2.570103883743286,
+      "learning_rate": 4.771803640224424e-06,
+      "loss": 0.3698,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3493307839388145,
+      "grad_norm": 1.8896373510360718,
+      "learning_rate": 4.7721515204347135e-06,
+      "loss": 0.1165,
+      "step": 1827
+    },
+    {
+      "epoch": 0.34952198852772465,
+      "grad_norm": 2.3569037914276123,
+      "learning_rate": 4.772499210286465e-06,
+      "loss": 0.2643,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3497131931166348,
+      "grad_norm": 3.0266520977020264,
+      "learning_rate": 4.772846709987891e-06,
+      "loss": 0.3738,
+      "step": 1829
+    },
+    {
+      "epoch": 0.34990439770554493,
+      "grad_norm": 3.1515417098999023,
+      "learning_rate": 4.773194019746864e-06,
+      "loss": 0.1745,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3500956022944551,
+      "grad_norm": 1.1789801120758057,
+      "learning_rate": 4.773541139770914e-06,
+      "loss": 0.0523,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3502868068833652,
+      "grad_norm": 2.357591152191162,
+      "learning_rate": 4.7738880702672316e-06,
+      "loss": 0.3621,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3504780114722753,
+      "grad_norm": 1.8474746942520142,
+      "learning_rate": 4.77423481144267e-06,
+      "loss": 0.3042,
+      "step": 1833
+    },
+    {
+      "epoch": 0.35066921606118545,
+      "grad_norm": 1.6401458978652954,
+      "learning_rate": 4.77458136350374e-06,
+      "loss": 0.1098,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3508604206500956,
+      "grad_norm": 2.561168670654297,
+      "learning_rate": 4.774927726656617e-06,
+      "loss": 0.3783,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35105162523900574,
+      "grad_norm": 2.586186170578003,
+      "learning_rate": 4.775273901107143e-06,
+      "loss": 0.1671,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3512428298279159,
+      "grad_norm": 0.993121325969696,
+      "learning_rate": 4.775619887060815e-06,
+      "loss": 0.0358,
+      "step": 1837
+    },
+    {
+      "epoch": 0.351434034416826,
+      "grad_norm": 3.0413339138031006,
+      "learning_rate": 4.775965684722804e-06,
+      "loss": 0.1799,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3516252390057361,
+      "grad_norm": 2.575594902038574,
+      "learning_rate": 4.776311294297939e-06,
+      "loss": 0.3987,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35181644359464626,
+      "grad_norm": 2.484976053237915,
+      "learning_rate": 4.776656715990719e-06,
+      "loss": 0.201,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3520076481835564,
+      "grad_norm": 2.7293996810913086,
+      "learning_rate": 4.777001950005309e-06,
+      "loss": 0.5425,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35219885277246654,
+      "grad_norm": 1.7221540212631226,
+      "learning_rate": 4.77734699654554e-06,
+      "loss": 0.1215,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3523900573613767,
+      "grad_norm": 2.226285696029663,
+      "learning_rate": 4.777691855814912e-06,
+      "loss": 0.2002,
+      "step": 1843
+    },
+    {
+      "epoch": 0.35258126195028683,
+      "grad_norm": 3.086306571960449,
+      "learning_rate": 4.778036528016594e-06,
+      "loss": 0.2252,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3527724665391969,
+      "grad_norm": 2.252471685409546,
+      "learning_rate": 4.778381013353426e-06,
+      "loss": 0.2922,
+      "step": 1845
+    },
+    {
+      "epoch": 0.35296367112810706,
+      "grad_norm": 3.0337913036346436,
+      "learning_rate": 4.778725312027913e-06,
+      "loss": 0.3342,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3531548757170172,
+      "grad_norm": 2.6601107120513916,
+      "learning_rate": 4.7790694242422385e-06,
+      "loss": 0.2105,
+      "step": 1847
+    },
+    {
+      "epoch": 0.35334608030592735,
+      "grad_norm": 1.9365030527114868,
+      "learning_rate": 4.779413350198251e-06,
+      "loss": 0.192,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3535372848948375,
+      "grad_norm": 2.894613027572632,
+      "learning_rate": 4.779757090097476e-06,
+      "loss": 0.175,
+      "step": 1849
+    },
+    {
+      "epoch": 0.35372848948374763,
+      "grad_norm": 2.1319048404693604,
+      "learning_rate": 4.780100644141109e-06,
+      "loss": 0.1175,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3539196940726577,
+      "grad_norm": 2.9696707725524902,
+      "learning_rate": 4.780444012530022e-06,
+      "loss": 0.2393,
+      "step": 1851
+    },
+    {
+      "epoch": 0.35411089866156786,
+      "grad_norm": 2.417073965072632,
+      "learning_rate": 4.780787195464761e-06,
+      "loss": 0.335,
+      "step": 1852
+    },
+    {
+      "epoch": 0.354302103250478,
+      "grad_norm": 2.466055154800415,
+      "learning_rate": 4.7811301931455436e-06,
+      "loss": 0.3433,
+      "step": 1853
+    },
+    {
+      "epoch": 0.35449330783938815,
+      "grad_norm": 2.6776459217071533,
+      "learning_rate": 4.781473005772269e-06,
+      "loss": 0.3252,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3546845124282983,
+      "grad_norm": 1.4896794557571411,
+      "learning_rate": 4.7818156335445075e-06,
+      "loss": 0.1096,
+      "step": 1855
+    },
+    {
+      "epoch": 0.35487571701720844,
+      "grad_norm": 2.7048983573913574,
+      "learning_rate": 4.782158076661511e-06,
+      "loss": 0.2342,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3550669216061185,
+      "grad_norm": 2.1952528953552246,
+      "learning_rate": 4.782500335322208e-06,
+      "loss": 0.2551,
+      "step": 1857
+    },
+    {
+      "epoch": 0.35525812619502867,
+      "grad_norm": 2.7355797290802,
+      "learning_rate": 4.782842409725205e-06,
+      "loss": 0.4026,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3554493307839388,
+      "grad_norm": 1.9009243249893188,
+      "learning_rate": 4.78318430006879e-06,
+      "loss": 0.169,
+      "step": 1859
+    },
+    {
+      "epoch": 0.35564053537284895,
+      "grad_norm": 1.3028790950775146,
+      "learning_rate": 4.783526006550927e-06,
+      "loss": 0.1357,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3558317399617591,
+      "grad_norm": 2.2286794185638428,
+      "learning_rate": 4.783867529369265e-06,
+      "loss": 0.1787,
+      "step": 1861
+    },
+    {
+      "epoch": 0.35602294455066924,
+      "grad_norm": 1.97344970703125,
+      "learning_rate": 4.784208868721133e-06,
+      "loss": 0.0833,
+      "step": 1862
+    },
+    {
+      "epoch": 0.35621414913957933,
+      "grad_norm": 2.151817560195923,
+      "learning_rate": 4.784550024803541e-06,
+      "loss": 0.154,
+      "step": 1863
+    },
+    {
+      "epoch": 0.35640535372848947,
+      "grad_norm": 1.631117582321167,
+      "learning_rate": 4.784890997813184e-06,
+      "loss": 0.1474,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3565965583173996,
+      "grad_norm": 2.0637216567993164,
+      "learning_rate": 4.785231787946437e-06,
+      "loss": 0.1154,
+      "step": 1865
+    },
+    {
+      "epoch": 0.35678776290630976,
+      "grad_norm": 1.8624078035354614,
+      "learning_rate": 4.785572395399365e-06,
+      "loss": 0.1261,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3569789674952199,
+      "grad_norm": 2.4838550090789795,
+      "learning_rate": 4.785912820367712e-06,
+      "loss": 0.2282,
+      "step": 1867
+    },
+    {
+      "epoch": 0.35717017208413004,
+      "grad_norm": 2.778296709060669,
+      "learning_rate": 4.786253063046911e-06,
+      "loss": 0.3457,
+      "step": 1868
+    },
+    {
+      "epoch": 0.35736137667304013,
+      "grad_norm": 3.7155518531799316,
+      "learning_rate": 4.7865931236320795e-06,
+      "loss": 0.4199,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3575525812619503,
+      "grad_norm": 2.4353044033050537,
+      "learning_rate": 4.7869330023180235e-06,
+      "loss": 0.1885,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3577437858508604,
+      "grad_norm": 2.149007797241211,
+      "learning_rate": 4.787272699299234e-06,
+      "loss": 0.1746,
+      "step": 1871
+    },
+    {
+      "epoch": 0.35793499043977056,
+      "grad_norm": 3.386817216873169,
+      "learning_rate": 4.787612214769893e-06,
+      "loss": 0.2888,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3581261950286807,
+      "grad_norm": 1.4428246021270752,
+      "learning_rate": 4.78795154892387e-06,
+      "loss": 0.1703,
+      "step": 1873
+    },
+    {
+      "epoch": 0.35831739961759085,
+      "grad_norm": 1.719489574432373,
+      "learning_rate": 4.788290701954725e-06,
+      "loss": 0.113,
+      "step": 1874
+    },
+    {
+      "epoch": 0.35850860420650094,
+      "grad_norm": 2.9463346004486084,
+      "learning_rate": 4.788629674055707e-06,
+      "loss": 0.1265,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3586998087954111,
+      "grad_norm": 1.8955507278442383,
+      "learning_rate": 4.788968465419756e-06,
+      "loss": 0.1493,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3588910133843212,
+      "grad_norm": 1.885257601737976,
+      "learning_rate": 4.789307076239504e-06,
+      "loss": 0.3264,
+      "step": 1877
+    },
+    {
+      "epoch": 0.35908221797323137,
+      "grad_norm": 2.1236987113952637,
+      "learning_rate": 4.789645506707277e-06,
+      "loss": 0.1504,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3592734225621415,
+      "grad_norm": 1.592720627784729,
+      "learning_rate": 4.789983757015089e-06,
+      "loss": 0.2391,
+      "step": 1879
+    },
+    {
+      "epoch": 0.35946462715105165,
+      "grad_norm": 1.6620198488235474,
+      "learning_rate": 4.790321827354654e-06,
+      "loss": 0.2443,
+      "step": 1880
+    },
+    {
+      "epoch": 0.35965583173996174,
+      "grad_norm": 1.3143545389175415,
+      "learning_rate": 4.790659717917373e-06,
+      "loss": 0.1046,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3598470363288719,
+      "grad_norm": 1.7060904502868652,
+      "learning_rate": 4.790997428894348e-06,
+      "loss": 0.2334,
+      "step": 1882
+    },
+    {
+      "epoch": 0.360038240917782,
+      "grad_norm": 2.8742563724517822,
+      "learning_rate": 4.791334960476374e-06,
+      "loss": 0.4188,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36022944550669217,
+      "grad_norm": 3.0253193378448486,
+      "learning_rate": 4.79167231285394e-06,
+      "loss": 0.267,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3604206500956023,
+      "grad_norm": 1.5449588298797607,
+      "learning_rate": 4.792009486217236e-06,
+      "loss": 0.1025,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3606118546845124,
+      "grad_norm": 2.831562042236328,
+      "learning_rate": 4.792346480756146e-06,
+      "loss": 0.2385,
+      "step": 1886
+    },
+    {
+      "epoch": 0.36080305927342254,
+      "grad_norm": 2.17584490776062,
+      "learning_rate": 4.792683296660254e-06,
+      "loss": 0.1348,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3609942638623327,
+      "grad_norm": 2.886467456817627,
+      "learning_rate": 4.793019934118841e-06,
+      "loss": 0.3341,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36118546845124283,
+      "grad_norm": 2.703188896179199,
+      "learning_rate": 4.793356393320889e-06,
+      "loss": 0.5411,
+      "step": 1889
+    },
+    {
+      "epoch": 0.361376673040153,
+      "grad_norm": 2.883558988571167,
+      "learning_rate": 4.79369267445508e-06,
+      "loss": 0.4724,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3615678776290631,
+      "grad_norm": 2.0322422981262207,
+      "learning_rate": 4.794028777709793e-06,
+      "loss": 0.1397,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3617590822179732,
+      "grad_norm": 2.538921594619751,
+      "learning_rate": 4.794364703273114e-06,
+      "loss": 0.1836,
+      "step": 1892
+    },
+    {
+      "epoch": 0.36195028680688335,
+      "grad_norm": 1.7409203052520752,
+      "learning_rate": 4.7947004513328256e-06,
+      "loss": 0.1028,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3621414913957935,
+      "grad_norm": 2.6174840927124023,
+      "learning_rate": 4.795036022076417e-06,
+      "loss": 0.228,
+      "step": 1894
+    },
+    {
+      "epoch": 0.36233269598470363,
+      "grad_norm": 1.768669605255127,
+      "learning_rate": 4.795371415691077e-06,
+      "loss": 0.2068,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3625239005736138,
+      "grad_norm": 1.426334261894226,
+      "learning_rate": 4.795706632363701e-06,
+      "loss": 0.1385,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3627151051625239,
+      "grad_norm": 1.599705457687378,
+      "learning_rate": 4.796041672280887e-06,
+      "loss": 0.1014,
+      "step": 1897
+    },
+    {
+      "epoch": 0.362906309751434,
+      "grad_norm": 3.4072515964508057,
+      "learning_rate": 4.7963765356289385e-06,
+      "loss": 0.1648,
+      "step": 1898
+    },
+    {
+      "epoch": 0.36309751434034415,
+      "grad_norm": 2.493661642074585,
+      "learning_rate": 4.796711222593864e-06,
+      "loss": 0.106,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3632887189292543,
+      "grad_norm": 1.6522670984268188,
+      "learning_rate": 4.79704573336138e-06,
+      "loss": 0.0677,
+      "step": 1900
+    },
+    {
+      "epoch": 0.36347992351816444,
+      "grad_norm": 2.6305341720581055,
+      "learning_rate": 4.797380068116908e-06,
+      "loss": 0.3414,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3636711281070746,
+      "grad_norm": 2.2357025146484375,
+      "learning_rate": 4.797714227045577e-06,
+      "loss": 0.3488,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3638623326959847,
+      "grad_norm": 2.4729673862457275,
+      "learning_rate": 4.7980482103322265e-06,
+      "loss": 0.3622,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3640535372848948,
+      "grad_norm": 2.1345574855804443,
+      "learning_rate": 4.7983820181614016e-06,
+      "loss": 0.2182,
+      "step": 1904
+    },
+    {
+      "epoch": 0.36424474187380496,
+      "grad_norm": 3.113025188446045,
+      "learning_rate": 4.798715650717358e-06,
+      "loss": 0.1702,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3644359464627151,
+      "grad_norm": 1.9702924489974976,
+      "learning_rate": 4.799049108184062e-06,
+      "loss": 0.1392,
+      "step": 1906
+    },
+    {
+      "epoch": 0.36462715105162524,
+      "grad_norm": 3.3608129024505615,
+      "learning_rate": 4.799382390745188e-06,
+      "loss": 0.6729,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3648183556405354,
+      "grad_norm": 1.8882901668548584,
+      "learning_rate": 4.799715498584127e-06,
+      "loss": 0.1397,
+      "step": 1908
+    },
+    {
+      "epoch": 0.36500956022944553,
+      "grad_norm": 2.139448642730713,
+      "learning_rate": 4.800048431883974e-06,
+      "loss": 0.2139,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3652007648183556,
+      "grad_norm": 1.5788694620132446,
+      "learning_rate": 4.800381190827542e-06,
+      "loss": 0.1232,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36539196940726576,
+      "grad_norm": 1.1630561351776123,
+      "learning_rate": 4.800713775597355e-06,
+      "loss": 0.0644,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3655831739961759,
+      "grad_norm": 1.5363394021987915,
+      "learning_rate": 4.8010461863756505e-06,
+      "loss": 0.0722,
+      "step": 1912
+    },
+    {
+      "epoch": 0.36577437858508605,
+      "grad_norm": 2.2994883060455322,
+      "learning_rate": 4.801378423344381e-06,
+      "loss": 0.234,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3659655831739962,
+      "grad_norm": 2.314574956893921,
+      "learning_rate": 4.8017104866852135e-06,
+      "loss": 0.3868,
+      "step": 1914
+    },
+    {
+      "epoch": 0.36615678776290633,
+      "grad_norm": 2.861659288406372,
+      "learning_rate": 4.802042376579529e-06,
+      "loss": 0.3618,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3663479923518164,
+      "grad_norm": 2.1803858280181885,
+      "learning_rate": 4.802374093208426e-06,
+      "loss": 0.2072,
+      "step": 1916
+    },
+    {
+      "epoch": 0.36653919694072656,
+      "grad_norm": 2.282540798187256,
+      "learning_rate": 4.8027056367527195e-06,
+      "loss": 0.1491,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3667304015296367,
+      "grad_norm": 2.8533167839050293,
+      "learning_rate": 4.803037007392939e-06,
+      "loss": 0.2092,
+      "step": 1918
+    },
+    {
+      "epoch": 0.36692160611854685,
+      "grad_norm": 1.306018590927124,
+      "learning_rate": 4.803368205309336e-06,
+      "loss": 0.0812,
+      "step": 1919
+    },
+    {
+      "epoch": 0.367112810707457,
+      "grad_norm": 2.2597286701202393,
+      "learning_rate": 4.803699230681877e-06,
+      "loss": 0.3965,
+      "step": 1920
+    },
+    {
+      "epoch": 0.36730401529636714,
+      "grad_norm": 1.759872555732727,
+      "learning_rate": 4.804030083690248e-06,
+      "loss": 0.1801,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3674952198852772,
+      "grad_norm": 2.418565273284912,
+      "learning_rate": 4.804360764513856e-06,
+      "loss": 0.3234,
+      "step": 1922
+    },
+    {
+      "epoch": 0.36768642447418737,
+      "grad_norm": 2.85204815864563,
+      "learning_rate": 4.804691273331826e-06,
+      "loss": 0.1746,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3678776290630975,
+      "grad_norm": 3.226734161376953,
+      "learning_rate": 4.805021610323005e-06,
+      "loss": 0.115,
+      "step": 1924
+    },
+    {
+      "epoch": 0.36806883365200765,
+      "grad_norm": 2.01751446723938,
+      "learning_rate": 4.805351775665959e-06,
+      "loss": 0.1831,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3682600382409178,
+      "grad_norm": 6.024718761444092,
+      "learning_rate": 4.805681769538982e-06,
+      "loss": 0.3726,
+      "step": 1926
+    },
+    {
+      "epoch": 0.36845124282982794,
+      "grad_norm": 1.2591511011123657,
+      "learning_rate": 4.80601159212008e-06,
+      "loss": 0.1463,
+      "step": 1927
+    },
+    {
+      "epoch": 0.36864244741873803,
+      "grad_norm": 4.118191719055176,
+      "learning_rate": 4.806341243586993e-06,
+      "loss": 0.4118,
+      "step": 1928
+    },
+    {
+      "epoch": 0.36883365200764817,
+      "grad_norm": 3.087899923324585,
+      "learning_rate": 4.806670724117176e-06,
+      "loss": 0.4401,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3690248565965583,
+      "grad_norm": 1.9501676559448242,
+      "learning_rate": 4.807000033887813e-06,
+      "loss": 0.1647,
+      "step": 1930
+    },
+    {
+      "epoch": 0.36921606118546846,
+      "grad_norm": 1.5643742084503174,
+      "learning_rate": 4.807329173075811e-06,
+      "loss": 0.1077,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3694072657743786,
+      "grad_norm": 1.5807640552520752,
+      "learning_rate": 4.8076581418578e-06,
+      "loss": 0.2282,
+      "step": 1932
+    },
+    {
+      "epoch": 0.36959847036328874,
+      "grad_norm": 2.416539430618286,
+      "learning_rate": 4.807986940410142e-06,
+      "loss": 0.2228,
+      "step": 1933
+    },
+    {
+      "epoch": 0.36978967495219883,
+      "grad_norm": 1.8993134498596191,
+      "learning_rate": 4.808315568908919e-06,
+      "loss": 0.1577,
+      "step": 1934
+    },
+    {
+      "epoch": 0.369980879541109,
+      "grad_norm": 2.963465690612793,
+      "learning_rate": 4.808644027529942e-06,
+      "loss": 0.3729,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3701720841300191,
+      "grad_norm": 2.0274300575256348,
+      "learning_rate": 4.808972316448751e-06,
+      "loss": 0.1123,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37036328871892926,
+      "grad_norm": 1.7501300573349,
+      "learning_rate": 4.809300435840613e-06,
+      "loss": 0.0785,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3705544933078394,
+      "grad_norm": 1.8195841312408447,
+      "learning_rate": 4.809628385880523e-06,
+      "loss": 0.138,
+      "step": 1938
+    },
+    {
+      "epoch": 0.37074569789674955,
+      "grad_norm": 1.7135430574417114,
+      "learning_rate": 4.809956166743207e-06,
+      "loss": 0.1738,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37093690248565964,
+      "grad_norm": 2.0836780071258545,
+      "learning_rate": 4.8102837786031185e-06,
+      "loss": 0.1386,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3711281070745698,
+      "grad_norm": 1.6961652040481567,
+      "learning_rate": 4.8106112216344405e-06,
+      "loss": 0.1152,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3713193116634799,
+      "grad_norm": 1.7368990182876587,
+      "learning_rate": 4.810938496011093e-06,
+      "loss": 0.1961,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37151051625239007,
+      "grad_norm": 1.4816842079162598,
+      "learning_rate": 4.811265601906719e-06,
+      "loss": 0.0777,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3717017208413002,
+      "grad_norm": 2.1479954719543457,
+      "learning_rate": 4.8115925394946985e-06,
+      "loss": 0.2723,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37189292543021035,
+      "grad_norm": 2.2252044677734375,
+      "learning_rate": 4.811919308948144e-06,
+      "loss": 0.2639,
+      "step": 1945
+    },
+    {
+      "epoch": 0.37208413001912044,
+      "grad_norm": 1.9705275297164917,
+      "learning_rate": 4.8122459104399e-06,
+      "loss": 0.1681,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3722753346080306,
+      "grad_norm": 3.6255764961242676,
+      "learning_rate": 4.812572344142544e-06,
+      "loss": 0.1158,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3724665391969407,
+      "grad_norm": 1.2052736282348633,
+      "learning_rate": 4.812898610228388e-06,
+      "loss": 0.1344,
+      "step": 1948
+    },
+    {
+      "epoch": 0.37265774378585087,
+      "grad_norm": 2.316469669342041,
+      "learning_rate": 4.81322470886948e-06,
+      "loss": 0.2374,
+      "step": 1949
+    },
+    {
+      "epoch": 0.372848948374761,
+      "grad_norm": 2.0329482555389404,
+      "learning_rate": 4.813550640237602e-06,
+      "loss": 0.0955,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3730401529636711,
+      "grad_norm": 2.659797191619873,
+      "learning_rate": 4.813876404504271e-06,
+      "loss": 0.4339,
+      "step": 1951
+    },
+    {
+      "epoch": 0.37323135755258124,
+      "grad_norm": 1.6562273502349854,
+      "learning_rate": 4.814202001840742e-06,
+      "loss": 0.1375,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3734225621414914,
+      "grad_norm": 3.6155762672424316,
+      "learning_rate": 4.814527432418008e-06,
+      "loss": 0.1702,
+      "step": 1953
+    },
+    {
+      "epoch": 0.37361376673040153,
+      "grad_norm": 1.579047679901123,
+      "learning_rate": 4.814852696406796e-06,
+      "loss": 0.1116,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3738049713193117,
+      "grad_norm": 2.239649772644043,
+      "learning_rate": 4.815177793977572e-06,
+      "loss": 0.1462,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3739961759082218,
+      "grad_norm": 0.9502508640289307,
+      "learning_rate": 4.815502725300541e-06,
+      "loss": 0.0592,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3741873804971319,
+      "grad_norm": 3.136199712753296,
+      "learning_rate": 4.815827490545649e-06,
+      "loss": 0.5801,
+      "step": 1957
+    },
+    {
+      "epoch": 0.37437858508604205,
+      "grad_norm": 4.284992694854736,
+      "learning_rate": 4.8161520898825794e-06,
+      "loss": 0.1291,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3745697896749522,
+      "grad_norm": 2.919841766357422,
+      "learning_rate": 4.816476523480754e-06,
+      "loss": 0.4008,
+      "step": 1959
+    },
+    {
+      "epoch": 0.37476099426386233,
+      "grad_norm": 2.5471062660217285,
+      "learning_rate": 4.816800791509338e-06,
+      "loss": 0.0981,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3749521988527725,
+      "grad_norm": 2.21247935295105,
+      "learning_rate": 4.817124894137239e-06,
+      "loss": 0.153,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3751434034416826,
+      "grad_norm": 2.344128131866455,
+      "learning_rate": 4.8174488315331e-06,
+      "loss": 0.1729,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3753346080305927,
+      "grad_norm": 1.6539994478225708,
+      "learning_rate": 4.817772603865314e-06,
+      "loss": 0.1111,
+      "step": 1963
+    },
+    {
+      "epoch": 0.37552581261950285,
+      "grad_norm": 2.1132073402404785,
+      "learning_rate": 4.818096211302012e-06,
+      "loss": 0.3434,
+      "step": 1964
+    },
+    {
+      "epoch": 0.375717017208413,
+      "grad_norm": 2.918445110321045,
+      "learning_rate": 4.818419654011068e-06,
+      "loss": 0.4585,
+      "step": 1965
+    },
+    {
+      "epoch": 0.37590822179732314,
+      "grad_norm": 2.2636373043060303,
+      "learning_rate": 4.818742932160102e-06,
+      "loss": 0.3373,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3760994263862333,
+      "grad_norm": 6.420876502990723,
+      "learning_rate": 4.8190660459164775e-06,
+      "loss": 0.1555,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3762906309751434,
+      "grad_norm": 3.735738754272461,
+      "learning_rate": 4.819388995447304e-06,
+      "loss": 0.2304,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3764818355640535,
+      "grad_norm": 0.7488231062889099,
+      "learning_rate": 4.819711780919433e-06,
+      "loss": 0.0496,
+      "step": 1969
+    },
+    {
+      "epoch": 0.37667304015296366,
+      "grad_norm": 3.412890672683716,
+      "learning_rate": 4.8200344024994645e-06,
+      "loss": 0.6286,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3768642447418738,
+      "grad_norm": 1.569946050643921,
+      "learning_rate": 4.820356860353744e-06,
+      "loss": 0.1469,
+      "step": 1971
+    },
+    {
+      "epoch": 0.37705544933078394,
+      "grad_norm": 1.486785650253296,
+      "learning_rate": 4.820679154648364e-06,
+      "loss": 0.1043,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3772466539196941,
+      "grad_norm": 1.4714351892471313,
+      "learning_rate": 4.821001285549165e-06,
+      "loss": 0.0955,
+      "step": 1973
+    },
+    {
+      "epoch": 0.37743785850860423,
+      "grad_norm": 3.8283934593200684,
+      "learning_rate": 4.821323253221735e-06,
+      "loss": 0.24,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3776290630975143,
+      "grad_norm": 2.3622210025787354,
+      "learning_rate": 4.821645057831409e-06,
+      "loss": 0.2372,
+      "step": 1975
+    },
+    {
+      "epoch": 0.37782026768642446,
+      "grad_norm": 2.4711544513702393,
+      "learning_rate": 4.821966699543274e-06,
+      "loss": 0.2733,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3780114722753346,
+      "grad_norm": 3.0288548469543457,
+      "learning_rate": 4.822288178522164e-06,
+      "loss": 0.3468,
+      "step": 1977
+    },
+    {
+      "epoch": 0.37820267686424475,
+      "grad_norm": 2.11275577545166,
+      "learning_rate": 4.822609494932662e-06,
+      "loss": 0.4129,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3783938814531549,
+      "grad_norm": 2.2081563472747803,
+      "learning_rate": 4.822930648939106e-06,
+      "loss": 0.2547,
+      "step": 1979
+    },
+    {
+      "epoch": 0.37858508604206503,
+      "grad_norm": 1.3358185291290283,
+      "learning_rate": 4.823251640705579e-06,
+      "loss": 0.0734,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3787762906309751,
+      "grad_norm": 1.5681957006454468,
+      "learning_rate": 4.82357247039592e-06,
+      "loss": 0.0682,
+      "step": 1981
+    },
+    {
+      "epoch": 0.37896749521988526,
+      "grad_norm": 3.349141836166382,
+      "learning_rate": 4.823893138173719e-06,
+      "loss": 0.2218,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3791586998087954,
+      "grad_norm": 1.5513498783111572,
+      "learning_rate": 4.824213644202315e-06,
+      "loss": 0.1234,
+      "step": 1983
+    },
+    {
+      "epoch": 0.37934990439770555,
+      "grad_norm": 1.504607081413269,
+      "learning_rate": 4.824533988644806e-06,
+      "loss": 0.114,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3795411089866157,
+      "grad_norm": 1.5265016555786133,
+      "learning_rate": 4.8248541716640375e-06,
+      "loss": 0.1458,
+      "step": 1985
+    },
+    {
+      "epoch": 0.37973231357552584,
+      "grad_norm": 1.8571374416351318,
+      "learning_rate": 4.825174193422613e-06,
+      "loss": 0.0707,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3799235181644359,
+      "grad_norm": 1.7218410968780518,
+      "learning_rate": 4.82549405408289e-06,
+      "loss": 0.0912,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38011472275334607,
+      "grad_norm": 2.492093324661255,
+      "learning_rate": 4.825813753806979e-06,
+      "loss": 0.2172,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3803059273422562,
+      "grad_norm": 1.584241271018982,
+      "learning_rate": 4.8261332927567455e-06,
+      "loss": 0.1455,
+      "step": 1989
+    },
+    {
+      "epoch": 0.38049713193116635,
+      "grad_norm": 1.470940351486206,
+      "learning_rate": 4.826452671093815e-06,
+      "loss": 0.1538,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3806883365200765,
+      "grad_norm": 0.8330782055854797,
+      "learning_rate": 4.826771888979564e-06,
+      "loss": 0.0504,
+      "step": 1991
+    },
+    {
+      "epoch": 0.38087954110898664,
+      "grad_norm": 2.128312110900879,
+      "learning_rate": 4.827090946575131e-06,
+      "loss": 0.1854,
+      "step": 1992
+    },
+    {
+      "epoch": 0.38107074569789673,
+      "grad_norm": 2.1762027740478516,
+      "learning_rate": 4.827409844041409e-06,
+      "loss": 0.1338,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38126195028680687,
+      "grad_norm": 6.44410514831543,
+      "learning_rate": 4.827728581539049e-06,
+      "loss": 0.4493,
+      "step": 1994
+    },
+    {
+      "epoch": 0.381453154875717,
+      "grad_norm": 2.3031179904937744,
+      "learning_rate": 4.828047159228461e-06,
+      "loss": 0.1516,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38164435946462716,
+      "grad_norm": 1.6921497583389282,
+      "learning_rate": 4.828365577269813e-06,
+      "loss": 0.2124,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3818355640535373,
+      "grad_norm": 2.609638214111328,
+      "learning_rate": 4.828683835823034e-06,
+      "loss": 0.3682,
+      "step": 1997
+    },
+    {
+      "epoch": 0.38202676864244745,
+      "grad_norm": 1.6939135789871216,
+      "learning_rate": 4.82900193504781e-06,
+      "loss": 0.1457,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38221797323135753,
+      "grad_norm": 1.4097681045532227,
+      "learning_rate": 4.829319875103591e-06,
+      "loss": 0.0805,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3824091778202677,
+      "grad_norm": 1.39816153049469,
+      "learning_rate": 4.829637656149586e-06,
+      "loss": 0.1119,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3824091778202677,
+      "eval_runtime": 849.7838,
+      "eval_samples_per_second": 1.805,
+      "eval_steps_per_second": 0.226,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3826003824091778,
+      "grad_norm": 2.89485502243042,
+      "learning_rate": 4.829955278344763e-06,
+      "loss": 0.2111,
+      "step": 2001
+    },
+    {
+      "epoch": 0.38279158699808796,
+      "grad_norm": 2.1980834007263184,
+      "learning_rate": 4.830272741847855e-06,
+      "loss": 0.2678,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3829827915869981,
+      "grad_norm": 6.06950569152832,
+      "learning_rate": 4.830590046817356e-06,
+      "loss": 0.5378,
+      "step": 2003
+    },
+    {
+      "epoch": 0.38317399617590825,
+      "grad_norm": 1.7232050895690918,
+      "learning_rate": 4.830907193411522e-06,
+      "loss": 0.1375,
+      "step": 2004
+    },
+    {
+      "epoch": 0.38336520076481834,
+      "grad_norm": 2.8828940391540527,
+      "learning_rate": 4.831224181788373e-06,
+      "loss": 0.2377,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3835564053537285,
+      "grad_norm": 2.6020548343658447,
+      "learning_rate": 4.831541012105694e-06,
+      "loss": 0.1445,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3837476099426386,
+      "grad_norm": 4.2495341300964355,
+      "learning_rate": 4.83185768452103e-06,
+      "loss": 0.3976,
+      "step": 2007
+    },
+    {
+      "epoch": 0.38393881453154877,
+      "grad_norm": 1.559543251991272,
+      "learning_rate": 4.832174199191696e-06,
+      "loss": 0.2264,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3841300191204589,
+      "grad_norm": 1.8970059156417847,
+      "learning_rate": 4.8324905562747656e-06,
+      "loss": 0.1736,
+      "step": 2009
+    },
+    {
+      "epoch": 0.384321223709369,
+      "grad_norm": 1.7815033197402954,
+      "learning_rate": 4.832806755927084e-06,
+      "loss": 0.1395,
+      "step": 2010
+    },
+    {
+      "epoch": 0.38451242829827914,
+      "grad_norm": 1.9595319032669067,
+      "learning_rate": 4.833122798305259e-06,
+      "loss": 0.0992,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3847036328871893,
+      "grad_norm": 1.122615098953247,
+      "learning_rate": 4.8334386835656655e-06,
+      "loss": 0.0475,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3848948374760994,
+      "grad_norm": 1.7938095331192017,
+      "learning_rate": 4.8337544118644455e-06,
+      "loss": 0.0843,
+      "step": 2013
+    },
+    {
+      "epoch": 0.38508604206500957,
+      "grad_norm": 2.116589307785034,
+      "learning_rate": 4.834069983357508e-06,
+      "loss": 0.2812,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3852772466539197,
+      "grad_norm": 1.9789377450942993,
+      "learning_rate": 4.83438539820053e-06,
+      "loss": 0.1502,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3854684512428298,
+      "grad_norm": 1.4772371053695679,
+      "learning_rate": 4.834700656548958e-06,
+      "loss": 0.119,
+      "step": 2016
+    },
+    {
+      "epoch": 0.38565965583173994,
+      "grad_norm": 1.6071560382843018,
+      "learning_rate": 4.835015758558004e-06,
+      "loss": 0.1266,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3858508604206501,
+      "grad_norm": 1.8250123262405396,
+      "learning_rate": 4.835330704382654e-06,
+      "loss": 0.1408,
+      "step": 2018
+    },
+    {
+      "epoch": 0.38604206500956023,
+      "grad_norm": 1.7217341661453247,
+      "learning_rate": 4.835645494177658e-06,
+      "loss": 0.0784,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3862332695984704,
+      "grad_norm": 2.3684465885162354,
+      "learning_rate": 4.835960128097542e-06,
+      "loss": 0.336,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3864244741873805,
+      "grad_norm": 2.564565420150757,
+      "learning_rate": 4.836274606296597e-06,
+      "loss": 0.4045,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3866156787762906,
+      "grad_norm": 1.4181162118911743,
+      "learning_rate": 4.8365889289288894e-06,
+      "loss": 0.0994,
+      "step": 2022
+    },
+    {
+      "epoch": 0.38680688336520075,
+      "grad_norm": 2.774085760116577,
+      "learning_rate": 4.836903096148254e-06,
+      "loss": 0.2666,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3869980879541109,
+      "grad_norm": 2.653083086013794,
+      "learning_rate": 4.837217108108301e-06,
+      "loss": 0.3805,
+      "step": 2024
+    },
+    {
+      "epoch": 0.38718929254302104,
+      "grad_norm": 2.3239336013793945,
+      "learning_rate": 4.837530964962407e-06,
+      "loss": 0.086,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3873804971319312,
+      "grad_norm": 1.9220002889633179,
+      "learning_rate": 4.837844666863729e-06,
+      "loss": 0.2595,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3875717017208413,
+      "grad_norm": 2.506913900375366,
+      "learning_rate": 4.838158213965192e-06,
+      "loss": 0.3344,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3877629063097514,
+      "grad_norm": 2.293449640274048,
+      "learning_rate": 4.838471606419496e-06,
+      "loss": 0.1295,
+      "step": 2028
+    },
+    {
+      "epoch": 0.38795411089866155,
+      "grad_norm": 2.00219464302063,
+      "learning_rate": 4.8387848443791165e-06,
+      "loss": 0.1283,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3881453154875717,
+      "grad_norm": 2.1243393421173096,
+      "learning_rate": 4.839097927996301e-06,
+      "loss": 0.1196,
+      "step": 2030
+    },
+    {
+      "epoch": 0.38833652007648184,
+      "grad_norm": 1.4236724376678467,
+      "learning_rate": 4.839410857423075e-06,
+      "loss": 0.1003,
+      "step": 2031
+    },
+    {
+      "epoch": 0.388527724665392,
+      "grad_norm": 1.9056960344314575,
+      "learning_rate": 4.839723632811237e-06,
+      "loss": 0.1268,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3887189292543021,
+      "grad_norm": 2.2061145305633545,
+      "learning_rate": 4.840036254312363e-06,
+      "loss": 0.3048,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3889101338432122,
+      "grad_norm": 1.490218997001648,
+      "learning_rate": 4.840348722077805e-06,
+      "loss": 0.1018,
+      "step": 2034
+    },
+    {
+      "epoch": 0.38910133843212236,
+      "grad_norm": 1.3906128406524658,
+      "learning_rate": 4.840661036258691e-06,
+      "loss": 0.079,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3892925430210325,
+      "grad_norm": 2.2867212295532227,
+      "learning_rate": 4.840973197005928e-06,
+      "loss": 0.2561,
+      "step": 2036
+    },
+    {
+      "epoch": 0.38948374760994264,
+      "grad_norm": 2.6874730587005615,
+      "learning_rate": 4.841285204470199e-06,
+      "loss": 0.1251,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3896749521988528,
+      "grad_norm": 2.0944433212280273,
+      "learning_rate": 4.841597058801967e-06,
+      "loss": 0.2646,
+      "step": 2038
+    },
+    {
+      "epoch": 0.38986615678776293,
+      "grad_norm": 1.4850279092788696,
+      "learning_rate": 4.84190876015147e-06,
+      "loss": 0.1377,
+      "step": 2039
+    },
+    {
+      "epoch": 0.390057361376673,
+      "grad_norm": 1.729572057723999,
+      "learning_rate": 4.8422203086687295e-06,
+      "loss": 0.1603,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39024856596558316,
+      "grad_norm": 2.221207618713379,
+      "learning_rate": 4.842531704503544e-06,
+      "loss": 0.1859,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3904397705544933,
+      "grad_norm": 1.757230281829834,
+      "learning_rate": 4.842842947805491e-06,
+      "loss": 0.1199,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39063097514340345,
+      "grad_norm": 1.1581106185913086,
+      "learning_rate": 4.843154038723931e-06,
+      "loss": 0.0563,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3908221797323136,
+      "grad_norm": 1.65236234664917,
+      "learning_rate": 4.843464977408003e-06,
+      "loss": 0.1155,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39101338432122373,
+      "grad_norm": 3.325693130493164,
+      "learning_rate": 4.843775764006627e-06,
+      "loss": 0.4827,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3912045889101338,
+      "grad_norm": 2.740530014038086,
+      "learning_rate": 4.844086398668508e-06,
+      "loss": 0.3053,
+      "step": 2046
+    },
+    {
+      "epoch": 0.39139579349904396,
+      "grad_norm": 4.023008346557617,
+      "learning_rate": 4.844396881542128e-06,
+      "loss": 0.3783,
+      "step": 2047
+    },
+    {
+      "epoch": 0.3915869980879541,
+      "grad_norm": 1.8242052793502808,
+      "learning_rate": 4.844707212775756e-06,
+      "loss": 0.1804,
+      "step": 2048
+    },
+    {
+      "epoch": 0.39177820267686425,
+      "grad_norm": 1.4388545751571655,
+      "learning_rate": 4.8450173925174395e-06,
+      "loss": 0.0858,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3919694072657744,
+      "grad_norm": 2.0084853172302246,
+      "learning_rate": 4.845327420915012e-06,
+      "loss": 0.1291,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39216061185468454,
+      "grad_norm": 2.2301721572875977,
+      "learning_rate": 4.845637298116093e-06,
+      "loss": 0.3989,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3923518164435946,
+      "grad_norm": 1.3058888912200928,
+      "learning_rate": 4.84594702426808e-06,
+      "loss": 0.1435,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39254302103250477,
+      "grad_norm": 2.7791600227355957,
+      "learning_rate": 4.84625659951816e-06,
+      "loss": 0.4347,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3927342256214149,
+      "grad_norm": 2.30441951751709,
+      "learning_rate": 4.846566024013304e-06,
+      "loss": 0.1264,
+      "step": 2054
+    },
+    {
+      "epoch": 0.39292543021032506,
+      "grad_norm": 1.0920432806015015,
+      "learning_rate": 4.846875297900267e-06,
+      "loss": 0.0552,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3931166347992352,
+      "grad_norm": 1.924726128578186,
+      "learning_rate": 4.847184421325591e-06,
+      "loss": 0.1238,
+      "step": 2056
+    },
+    {
+      "epoch": 0.39330783938814534,
+      "grad_norm": 2.9873266220092773,
+      "learning_rate": 4.847493394435604e-06,
+      "loss": 0.3692,
+      "step": 2057
+    },
+    {
+      "epoch": 0.39349904397705543,
+      "grad_norm": 3.197446584701538,
+      "learning_rate": 4.84780221737642e-06,
+      "loss": 0.3972,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3936902485659656,
+      "grad_norm": 3.1228456497192383,
+      "learning_rate": 4.8481108902939405e-06,
+      "loss": 0.4105,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3938814531548757,
+      "grad_norm": 5.650298595428467,
+      "learning_rate": 4.8484194133338555e-06,
+      "loss": 0.3185,
+      "step": 2060
+    },
+    {
+      "epoch": 0.39407265774378586,
+      "grad_norm": 1.6081198453903198,
+      "learning_rate": 4.8487277866416415e-06,
+      "loss": 0.1289,
+      "step": 2061
+    },
+    {
+      "epoch": 0.394263862332696,
+      "grad_norm": 2.226999521255493,
+      "learning_rate": 4.849036010362564e-06,
+      "loss": 0.0994,
+      "step": 2062
+    },
+    {
+      "epoch": 0.39445506692160615,
+      "grad_norm": 2.3722176551818848,
+      "learning_rate": 4.8493440846416755e-06,
+      "loss": 0.1002,
+      "step": 2063
+    },
+    {
+      "epoch": 0.39464627151051623,
+      "grad_norm": 3.996410369873047,
+      "learning_rate": 4.8496520096238205e-06,
+      "loss": 0.5538,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3948374760994264,
+      "grad_norm": 2.3079299926757812,
+      "learning_rate": 4.8499597854536305e-06,
+      "loss": 0.2703,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3950286806883365,
+      "grad_norm": 2.9113502502441406,
+      "learning_rate": 4.85026741227553e-06,
+      "loss": 0.3456,
+      "step": 2066
+    },
+    {
+      "epoch": 0.39521988527724666,
+      "grad_norm": 2.183955192565918,
+      "learning_rate": 4.850574890233729e-06,
+      "loss": 0.11,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3954110898661568,
+      "grad_norm": 2.493313789367676,
+      "learning_rate": 4.850882219472235e-06,
+      "loss": 0.2962,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3956022944550669,
+      "grad_norm": 1.643670678138733,
+      "learning_rate": 4.851189400134838e-06,
+      "loss": 0.1328,
+      "step": 2069
+    },
+    {
+      "epoch": 0.39579349904397704,
+      "grad_norm": 2.830077886581421,
+      "learning_rate": 4.851496432365128e-06,
+      "loss": 0.3916,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3959847036328872,
+      "grad_norm": 1.6414769887924194,
+      "learning_rate": 4.851803316306482e-06,
+      "loss": 0.1513,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3961759082217973,
+      "grad_norm": 1.641139268875122,
+      "learning_rate": 4.85211005210207e-06,
+      "loss": 0.1359,
+      "step": 2072
+    },
+    {
+      "epoch": 0.39636711281070747,
+      "grad_norm": 2.166752576828003,
+      "learning_rate": 4.852416639894855e-06,
+      "loss": 0.1225,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3965583173996176,
+      "grad_norm": 2.580700397491455,
+      "learning_rate": 4.852723079827596e-06,
+      "loss": 0.2727,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3967495219885277,
+      "grad_norm": 1.7222399711608887,
+      "learning_rate": 4.85302937204284e-06,
+      "loss": 0.1269,
+      "step": 2075
+    },
+    {
+      "epoch": 0.39694072657743784,
+      "grad_norm": 1.9561702013015747,
+      "learning_rate": 4.853335516682933e-06,
+      "loss": 0.2164,
+      "step": 2076
+    },
+    {
+      "epoch": 0.397131931166348,
+      "grad_norm": 2.9243381023406982,
+      "learning_rate": 4.853641513890013e-06,
+      "loss": 0.3344,
+      "step": 2077
+    },
+    {
+      "epoch": 0.39732313575525813,
+      "grad_norm": 2.7178955078125,
+      "learning_rate": 4.853947363806012e-06,
+      "loss": 0.2405,
+      "step": 2078
+    },
+    {
+      "epoch": 0.39751434034416827,
+      "grad_norm": 1.9798709154129028,
+      "learning_rate": 4.85425306657266e-06,
+      "loss": 0.2382,
+      "step": 2079
+    },
+    {
+      "epoch": 0.3977055449330784,
+      "grad_norm": 3.2675580978393555,
+      "learning_rate": 4.8545586223314805e-06,
+      "loss": 0.1743,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3978967495219885,
+      "grad_norm": 2.0814285278320312,
+      "learning_rate": 4.8548640312237916e-06,
+      "loss": 0.0812,
+      "step": 2081
+    },
+    {
+      "epoch": 0.39808795411089865,
+      "grad_norm": 3.074506998062134,
+      "learning_rate": 4.855169293390711e-06,
+      "loss": 0.4034,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3982791586998088,
+      "grad_norm": 2.637707233428955,
+      "learning_rate": 4.85547440897315e-06,
+      "loss": 0.4595,
+      "step": 2083
+    },
+    {
+      "epoch": 0.39847036328871893,
+      "grad_norm": 1.7692375183105469,
+      "learning_rate": 4.855779378111821e-06,
+      "loss": 0.1297,
+      "step": 2084
+    },
+    {
+      "epoch": 0.3986615678776291,
+      "grad_norm": 1.5920054912567139,
+      "learning_rate": 4.856084200947228e-06,
+      "loss": 0.1393,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3988527724665392,
+      "grad_norm": 2.117548942565918,
+      "learning_rate": 4.856388877619678e-06,
+      "loss": 0.1037,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3990439770554493,
+      "grad_norm": 1.231958270072937,
+      "learning_rate": 4.856693408269275e-06,
+      "loss": 0.0662,
+      "step": 2087
+    },
+    {
+      "epoch": 0.39923518164435945,
+      "grad_norm": 1.5708861351013184,
+      "learning_rate": 4.85699779303592e-06,
+      "loss": 0.1032,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3994263862332696,
+      "grad_norm": 2.8182146549224854,
+      "learning_rate": 4.857302032059316e-06,
+      "loss": 0.5743,
+      "step": 2089
+    },
+    {
+      "epoch": 0.39961759082217974,
+      "grad_norm": 2.6597273349761963,
+      "learning_rate": 4.857606125478961e-06,
+      "loss": 0.3972,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3998087954110899,
+      "grad_norm": 2.8121769428253174,
+      "learning_rate": 4.857910073434157e-06,
+      "loss": 0.3327,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.188518762588501,
+      "learning_rate": 4.858213876064004e-06,
+      "loss": 0.2586,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4001912045889101,
+      "grad_norm": 1.4912960529327393,
+      "learning_rate": 4.858517533507403e-06,
+      "loss": 0.0882,
+      "step": 2093
+    },
+    {
+      "epoch": 0.40038240917782025,
+      "grad_norm": 3.883857011795044,
+      "learning_rate": 4.858821045903058e-06,
+      "loss": 0.1254,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4005736137667304,
+      "grad_norm": 3.3286659717559814,
+      "learning_rate": 4.85912441338947e-06,
+      "loss": 0.741,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40076481835564054,
+      "grad_norm": 2.54927134513855,
+      "learning_rate": 4.859427636104946e-06,
+      "loss": 0.1917,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4009560229445507,
+      "grad_norm": 3.058338165283203,
+      "learning_rate": 4.859730714187593e-06,
+      "loss": 0.2758,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4011472275334608,
+      "grad_norm": 2.1191134452819824,
+      "learning_rate": 4.860033647775319e-06,
+      "loss": 0.0992,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4013384321223709,
+      "grad_norm": 1.2811925411224365,
+      "learning_rate": 4.860336437005838e-06,
+      "loss": 0.066,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40152963671128106,
+      "grad_norm": 1.372006893157959,
+      "learning_rate": 4.860639082016667e-06,
+      "loss": 0.0831,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4017208413001912,
+      "grad_norm": 2.105086088180542,
+      "learning_rate": 4.8609415829451225e-06,
+      "loss": 0.1654,
+      "step": 2101
+    },
+    {
+      "epoch": 0.40191204588910134,
+      "grad_norm": 2.2075912952423096,
+      "learning_rate": 4.861243939928331e-06,
+      "loss": 0.2912,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4021032504780115,
+      "grad_norm": 1.9700030088424683,
+      "learning_rate": 4.861546153103217e-06,
+      "loss": 0.2135,
+      "step": 2103
+    },
+    {
+      "epoch": 0.40229445506692163,
+      "grad_norm": 1.7116718292236328,
+      "learning_rate": 4.861848222606516e-06,
+      "loss": 0.1778,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4024856596558317,
+      "grad_norm": 3.5548574924468994,
+      "learning_rate": 4.8621501485747625e-06,
+      "loss": 0.078,
+      "step": 2105
+    },
+    {
+      "epoch": 0.40267686424474186,
+      "grad_norm": 1.4873651266098022,
+      "learning_rate": 4.862451931144302e-06,
+      "loss": 0.0588,
+      "step": 2106
+    },
+    {
+      "epoch": 0.402868068833652,
+      "grad_norm": 2.477473497390747,
+      "learning_rate": 4.862753570451282e-06,
+      "loss": 0.2443,
+      "step": 2107
+    },
+    {
+      "epoch": 0.40305927342256215,
+      "grad_norm": 2.1252503395080566,
+      "learning_rate": 4.863055066631658e-06,
+      "loss": 0.4126,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4032504780114723,
+      "grad_norm": 1.0193214416503906,
+      "learning_rate": 4.863356419821191e-06,
+      "loss": 0.0968,
+      "step": 2109
+    },
+    {
+      "epoch": 0.40344168260038243,
+      "grad_norm": 3.261441946029663,
+      "learning_rate": 4.863657630155451e-06,
+      "loss": 0.5482,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4036328871892925,
+      "grad_norm": 1.2448210716247559,
+      "learning_rate": 4.863958697769811e-06,
+      "loss": 0.1252,
+      "step": 2111
+    },
+    {
+      "epoch": 0.40382409177820267,
+      "grad_norm": 2.254920482635498,
+      "learning_rate": 4.864259622799458e-06,
+      "loss": 0.0864,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4040152963671128,
+      "grad_norm": 2.846625804901123,
+      "learning_rate": 4.8645604053793795e-06,
+      "loss": 0.2891,
+      "step": 2113
+    },
+    {
+      "epoch": 0.40420650095602295,
+      "grad_norm": 2.426459789276123,
+      "learning_rate": 4.864861045644379e-06,
+      "loss": 0.2404,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4043977055449331,
+      "grad_norm": 2.1458327770233154,
+      "learning_rate": 4.865161543729063e-06,
+      "loss": 0.1713,
+      "step": 2115
+    },
+    {
+      "epoch": 0.40458891013384324,
+      "grad_norm": 3.3918213844299316,
+      "learning_rate": 4.865461899767849e-06,
+      "loss": 0.2598,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4047801147227533,
+      "grad_norm": 2.337883710861206,
+      "learning_rate": 4.865762113894966e-06,
+      "loss": 0.2028,
+      "step": 2117
+    },
+    {
+      "epoch": 0.40497131931166347,
+      "grad_norm": 1.5774587392807007,
+      "learning_rate": 4.866062186244448e-06,
+      "loss": 0.1215,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4051625239005736,
+      "grad_norm": 1.7613880634307861,
+      "learning_rate": 4.8663621169501456e-06,
+      "loss": 0.1357,
+      "step": 2119
+    },
+    {
+      "epoch": 0.40535372848948376,
+      "grad_norm": 3.2717082500457764,
+      "learning_rate": 4.866661906145713e-06,
+      "loss": 0.433,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4055449330783939,
+      "grad_norm": 2.3337671756744385,
+      "learning_rate": 4.866961553964623e-06,
+      "loss": 0.2526,
+      "step": 2121
+    },
+    {
+      "epoch": 0.40573613766730404,
+      "grad_norm": 1.9872267246246338,
+      "learning_rate": 4.867261060540153e-06,
+      "loss": 0.1941,
+      "step": 2122
+    },
+    {
+      "epoch": 0.40592734225621413,
+      "grad_norm": 1.2843959331512451,
+      "learning_rate": 4.867560426005394e-06,
+      "loss": 0.0682,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4061185468451243,
+      "grad_norm": 1.543263554573059,
+      "learning_rate": 4.8678596504932505e-06,
+      "loss": 0.1491,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4063097514340344,
+      "grad_norm": 2.1265337467193604,
+      "learning_rate": 4.868158734136438e-06,
+      "loss": 0.0946,
+      "step": 2125
+    },
+    {
+      "epoch": 0.40650095602294456,
+      "grad_norm": 1.6050384044647217,
+      "learning_rate": 4.868457677067485e-06,
+      "loss": 0.2419,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4066921606118547,
+      "grad_norm": 2.0895438194274902,
+      "learning_rate": 4.868756479418735e-06,
+      "loss": 0.1516,
+      "step": 2127
+    },
+    {
+      "epoch": 0.40688336520076485,
+      "grad_norm": 1.8586974143981934,
+      "learning_rate": 4.8690551413223396e-06,
+      "loss": 0.2194,
+      "step": 2128
+    },
+    {
+      "epoch": 0.40707456978967493,
+      "grad_norm": 2.5971615314483643,
+      "learning_rate": 4.869353662910269e-06,
+      "loss": 0.3191,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4072657743785851,
+      "grad_norm": 1.6561384201049805,
+      "learning_rate": 4.869652044314306e-06,
+      "loss": 0.1768,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4074569789674952,
+      "grad_norm": 2.3525571823120117,
+      "learning_rate": 4.869950285666048e-06,
+      "loss": 0.154,
+      "step": 2131
+    },
+    {
+      "epoch": 0.40764818355640536,
+      "grad_norm": 2.6129164695739746,
+      "learning_rate": 4.870248387096907e-06,
+      "loss": 0.1612,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4078393881453155,
+      "grad_norm": 1.5268852710723877,
+      "learning_rate": 4.87054634873811e-06,
+      "loss": 0.1151,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4080305927342256,
+      "grad_norm": 2.9393773078918457,
+      "learning_rate": 4.8708441707207e-06,
+      "loss": 0.1944,
+      "step": 2134
+    },
+    {
+      "epoch": 0.40822179732313574,
+      "grad_norm": 0.9411885738372803,
+      "learning_rate": 4.871141853175533e-06,
+      "loss": 0.0954,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4084130019120459,
+      "grad_norm": 1.751671552658081,
+      "learning_rate": 4.871439396233286e-06,
+      "loss": 0.2188,
+      "step": 2136
+    },
+    {
+      "epoch": 0.408604206500956,
+      "grad_norm": 2.3655753135681152,
+      "learning_rate": 4.8717368000244496e-06,
+      "loss": 0.3265,
+      "step": 2137
+    },
+    {
+      "epoch": 0.40879541108986617,
+      "grad_norm": 1.446745753288269,
+      "learning_rate": 4.8720340646793314e-06,
+      "loss": 0.1341,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4089866156787763,
+      "grad_norm": 1.964414358139038,
+      "learning_rate": 4.872331190328057e-06,
+      "loss": 0.3604,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4091778202676864,
+      "grad_norm": 1.6994702816009521,
+      "learning_rate": 4.872628177100569e-06,
+      "loss": 0.1487,
+      "step": 2140
+    },
+    {
+      "epoch": 0.40936902485659654,
+      "grad_norm": 2.1562836170196533,
+      "learning_rate": 4.872925025126627e-06,
+      "loss": 0.1985,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4095602294455067,
+      "grad_norm": 2.705679416656494,
+      "learning_rate": 4.873221734535811e-06,
+      "loss": 0.3949,
+      "step": 2142
+    },
+    {
+      "epoch": 0.40975143403441683,
+      "grad_norm": 2.84946870803833,
+      "learning_rate": 4.873518305457518e-06,
+      "loss": 0.1461,
+      "step": 2143
+    },
+    {
+      "epoch": 0.40994263862332697,
+      "grad_norm": 5.145754814147949,
+      "learning_rate": 4.873814738020963e-06,
+      "loss": 0.1118,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4101338432122371,
+      "grad_norm": 2.6604089736938477,
+      "learning_rate": 4.874111032355183e-06,
+      "loss": 0.3319,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4103250478011472,
+      "grad_norm": 2.2231945991516113,
+      "learning_rate": 4.874407188589032e-06,
+      "loss": 0.2436,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41051625239005735,
+      "grad_norm": 1.9829926490783691,
+      "learning_rate": 4.874703206851186e-06,
+      "loss": 0.1137,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4107074569789675,
+      "grad_norm": 1.793606162071228,
+      "learning_rate": 4.87499908727014e-06,
+      "loss": 0.1267,
+      "step": 2148
+    },
+    {
+      "epoch": 0.41089866156787763,
+      "grad_norm": 2.4400620460510254,
+      "learning_rate": 4.8752948299742085e-06,
+      "loss": 0.087,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4110898661567878,
+      "grad_norm": 1.6873563528060913,
+      "learning_rate": 4.875590435091529e-06,
+      "loss": 0.0621,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4112810707456979,
+      "grad_norm": 1.6609550714492798,
+      "learning_rate": 4.87588590275006e-06,
+      "loss": 0.1239,
+      "step": 2151
+    },
+    {
+      "epoch": 0.411472275334608,
+      "grad_norm": 1.9967857599258423,
+      "learning_rate": 4.876181233077581e-06,
+      "loss": 0.3056,
+      "step": 2152
+    },
+    {
+      "epoch": 0.41166347992351815,
+      "grad_norm": 2.5311481952667236,
+      "learning_rate": 4.876476426201691e-06,
+      "loss": 0.3258,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4118546845124283,
+      "grad_norm": 2.0180413722991943,
+      "learning_rate": 4.876771482249817e-06,
+      "loss": 0.2722,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41204588910133844,
+      "grad_norm": 1.2949248552322388,
+      "learning_rate": 4.877066401349204e-06,
+      "loss": 0.091,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4122370936902486,
+      "grad_norm": 1.2472654581069946,
+      "learning_rate": 4.87736118362692e-06,
+      "loss": 0.0818,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4124282982791587,
+      "grad_norm": 1.8975611925125122,
+      "learning_rate": 4.877655829209858e-06,
+      "loss": 0.2136,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4126195028680688,
+      "grad_norm": 2.191755771636963,
+      "learning_rate": 4.877950338224735e-06,
+      "loss": 0.3722,
+      "step": 2158
+    },
+    {
+      "epoch": 0.41281070745697895,
+      "grad_norm": 2.048837661743164,
+      "learning_rate": 4.87824471079809e-06,
+      "loss": 0.2638,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4130019120458891,
+      "grad_norm": 2.1126198768615723,
+      "learning_rate": 4.878538947056285e-06,
+      "loss": 0.2743,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41319311663479924,
+      "grad_norm": 2.0426793098449707,
+      "learning_rate": 4.878833047125512e-06,
+      "loss": 0.3363,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4133843212237094,
+      "grad_norm": 1.8592612743377686,
+      "learning_rate": 4.879127011131783e-06,
+      "loss": 0.1042,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4135755258126195,
+      "grad_norm": 7.511478424072266,
+      "learning_rate": 4.879420839200937e-06,
+      "loss": 0.2228,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4137667304015296,
+      "grad_norm": 1.5293527841567993,
+      "learning_rate": 4.8797145314586365e-06,
+      "loss": 0.1024,
+      "step": 2164
+    },
+    {
+      "epoch": 0.41395793499043976,
+      "grad_norm": 4.5838141441345215,
+      "learning_rate": 4.880008088030373e-06,
+      "loss": 0.5163,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4141491395793499,
+      "grad_norm": 2.471611738204956,
+      "learning_rate": 4.880301509041461e-06,
+      "loss": 0.2742,
+      "step": 2166
+    },
+    {
+      "epoch": 0.41434034416826004,
+      "grad_norm": 2.4510347843170166,
+      "learning_rate": 4.880594794617045e-06,
+      "loss": 0.1341,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4145315487571702,
+      "grad_norm": 2.3072543144226074,
+      "learning_rate": 4.880887944882094e-06,
+      "loss": 0.1352,
+      "step": 2168
+    },
+    {
+      "epoch": 0.41472275334608033,
+      "grad_norm": 0.9170379638671875,
+      "learning_rate": 4.881180959961403e-06,
+      "loss": 0.0542,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4149139579349904,
+      "grad_norm": 2.4316747188568115,
+      "learning_rate": 4.8814738399795956e-06,
+      "loss": 0.4565,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41510516252390056,
+      "grad_norm": 2.150325298309326,
+      "learning_rate": 4.881766585061125e-06,
+      "loss": 0.2609,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4152963671128107,
+      "grad_norm": 2.830181360244751,
+      "learning_rate": 4.8820591953302706e-06,
+      "loss": 0.3472,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41548757170172085,
+      "grad_norm": 1.9228092432022095,
+      "learning_rate": 4.882351670911141e-06,
+      "loss": 0.1312,
+      "step": 2173
+    },
+    {
+      "epoch": 0.415678776290631,
+      "grad_norm": 2.210085868835449,
+      "learning_rate": 4.8826440119276715e-06,
+      "loss": 0.1784,
+      "step": 2174
+    },
+    {
+      "epoch": 0.41586998087954113,
+      "grad_norm": 5.4337849617004395,
+      "learning_rate": 4.882936218503629e-06,
+      "loss": 0.1746,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4160611854684512,
+      "grad_norm": 1.8646759986877441,
+      "learning_rate": 4.883228290762608e-06,
+      "loss": 0.2276,
+      "step": 2176
+    },
+    {
+      "epoch": 0.41625239005736137,
+      "grad_norm": 2.343479871749878,
+      "learning_rate": 4.883520228828034e-06,
+      "loss": 0.2489,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4164435946462715,
+      "grad_norm": 2.6077628135681152,
+      "learning_rate": 4.88381203282316e-06,
+      "loss": 0.4214,
+      "step": 2178
+    },
+    {
+      "epoch": 0.41663479923518165,
+      "grad_norm": 2.1604247093200684,
+      "learning_rate": 4.884103702871074e-06,
+      "loss": 0.1684,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4168260038240918,
+      "grad_norm": 2.03054141998291,
+      "learning_rate": 4.884395239094688e-06,
+      "loss": 0.2568,
+      "step": 2180
+    },
+    {
+      "epoch": 0.41701720841300194,
+      "grad_norm": 1.8029322624206543,
+      "learning_rate": 4.88468664161675e-06,
+      "loss": 0.3231,
+      "step": 2181
+    },
+    {
+      "epoch": 0.417208413001912,
+      "grad_norm": 2.567142963409424,
+      "learning_rate": 4.884977910559839e-06,
+      "loss": 0.4366,
+      "step": 2182
+    },
+    {
+      "epoch": 0.41739961759082217,
+      "grad_norm": 1.7099403142929077,
+      "learning_rate": 4.885269046046362e-06,
+      "loss": 0.1528,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4175908221797323,
+      "grad_norm": 2.3315675258636475,
+      "learning_rate": 4.885560048198562e-06,
+      "loss": 0.5276,
+      "step": 2184
+    },
+    {
+      "epoch": 0.41778202676864246,
+      "grad_norm": 1.4180094003677368,
+      "learning_rate": 4.885850917138509e-06,
+      "loss": 0.0885,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4179732313575526,
+      "grad_norm": 1.8287487030029297,
+      "learning_rate": 4.886141652988113e-06,
+      "loss": 0.1236,
+      "step": 2186
+    },
+    {
+      "epoch": 0.41816443594646274,
+      "grad_norm": 2.6701676845550537,
+      "learning_rate": 4.886432255869108e-06,
+      "loss": 0.169,
+      "step": 2187
+    },
+    {
+      "epoch": 0.41835564053537283,
+      "grad_norm": 2.2629594802856445,
+      "learning_rate": 4.886722725903068e-06,
+      "loss": 0.2824,
+      "step": 2188
+    },
+    {
+      "epoch": 0.418546845124283,
+      "grad_norm": 2.599104881286621,
+      "learning_rate": 4.887013063211395e-06,
+      "loss": 0.2329,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4187380497131931,
+      "grad_norm": 2.1738619804382324,
+      "learning_rate": 4.887303267915331e-06,
+      "loss": 0.307,
+      "step": 2190
+    },
+    {
+      "epoch": 0.41892925430210326,
+      "grad_norm": 2.778266668319702,
+      "learning_rate": 4.887593340135946e-06,
+      "loss": 0.1392,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4191204588910134,
+      "grad_norm": 2.5189621448516846,
+      "learning_rate": 4.887883279994146e-06,
+      "loss": 0.3267,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4193116634799235,
+      "grad_norm": 2.6800177097320557,
+      "learning_rate": 4.888173087610673e-06,
+      "loss": 0.4112,
+      "step": 2193
+    },
+    {
+      "epoch": 0.41950286806883363,
+      "grad_norm": 4.370938777923584,
+      "learning_rate": 4.888462763106103e-06,
+      "loss": 0.2855,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4196940726577438,
+      "grad_norm": 2.941232442855835,
+      "learning_rate": 4.888752306600847e-06,
+      "loss": 0.4048,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4198852772466539,
+      "grad_norm": 2.167144775390625,
+      "learning_rate": 4.889041718215152e-06,
+      "loss": 0.3469,
+      "step": 2196
+    },
+    {
+      "epoch": 0.42007648183556406,
+      "grad_norm": 3.1527018547058105,
+      "learning_rate": 4.889330998069099e-06,
+      "loss": 0.4365,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4202676864244742,
+      "grad_norm": 2.6182942390441895,
+      "learning_rate": 4.889620146282609e-06,
+      "loss": 0.2454,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4204588910133843,
+      "grad_norm": 1.247412919998169,
+      "learning_rate": 4.889909162975434e-06,
+      "loss": 0.0454,
+      "step": 2199
+    },
+    {
+      "epoch": 0.42065009560229444,
+      "grad_norm": 2.049236536026001,
+      "learning_rate": 4.890198048267166e-06,
+      "loss": 0.0835,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4208413001912046,
+      "grad_norm": 2.416959285736084,
+      "learning_rate": 4.890486802277235e-06,
+      "loss": 0.2328,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4210325047801147,
+      "grad_norm": 2.8074474334716797,
+      "learning_rate": 4.890775425124906e-06,
+      "loss": 0.3223,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42122370936902487,
+      "grad_norm": 2.717810869216919,
+      "learning_rate": 4.89106391692928e-06,
+      "loss": 0.3238,
+      "step": 2203
+    },
+    {
+      "epoch": 0.421414913957935,
+      "grad_norm": 1.5587702989578247,
+      "learning_rate": 4.891352277809302e-06,
+      "loss": 0.1786,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4216061185468451,
+      "grad_norm": 1.952847957611084,
+      "learning_rate": 4.891640507883748e-06,
+      "loss": 0.174,
+      "step": 2205
+    },
+    {
+      "epoch": 0.42179732313575524,
+      "grad_norm": 3.144420623779297,
+      "learning_rate": 4.891928607271237e-06,
+      "loss": 0.1949,
+      "step": 2206
+    },
+    {
+      "epoch": 0.4219885277246654,
+      "grad_norm": 2.8562870025634766,
+      "learning_rate": 4.892216576090225e-06,
+      "loss": 0.3854,
+      "step": 2207
+    },
+    {
+      "epoch": 0.42217973231357553,
+      "grad_norm": 2.9668025970458984,
+      "learning_rate": 4.892504414459007e-06,
+      "loss": 0.4332,
+      "step": 2208
+    },
+    {
+      "epoch": 0.42237093690248567,
+      "grad_norm": 2.1173808574676514,
+      "learning_rate": 4.892792122495718e-06,
+      "loss": 0.1733,
+      "step": 2209
+    },
+    {
+      "epoch": 0.4225621414913958,
+      "grad_norm": 2.9007081985473633,
+      "learning_rate": 4.893079700318333e-06,
+      "loss": 0.2412,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4227533460803059,
+      "grad_norm": 2.2025012969970703,
+      "learning_rate": 4.893367148044665e-06,
+      "loss": 0.2325,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42294455066921605,
+      "grad_norm": 2.527050733566284,
+      "learning_rate": 4.893654465792369e-06,
+      "loss": 0.128,
+      "step": 2212
+    },
+    {
+      "epoch": 0.4231357552581262,
+      "grad_norm": 1.7862459421157837,
+      "learning_rate": 4.89394165367894e-06,
+      "loss": 0.2455,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42332695984703633,
+      "grad_norm": 1.6860567331314087,
+      "learning_rate": 4.894228711821714e-06,
+      "loss": 0.1739,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4235181644359465,
+      "grad_norm": 1.391953706741333,
+      "learning_rate": 4.894515640337865e-06,
+      "loss": 0.1023,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4237093690248566,
+      "grad_norm": 2.542024850845337,
+      "learning_rate": 4.894802439344414e-06,
+      "loss": 0.352,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4239005736137667,
+      "grad_norm": 1.902836799621582,
+      "learning_rate": 4.8950891089582185e-06,
+      "loss": 0.1247,
+      "step": 2217
+    },
+    {
+      "epoch": 0.42409177820267685,
+      "grad_norm": 1.8815006017684937,
+      "learning_rate": 4.895375649295982e-06,
+      "loss": 0.1437,
+      "step": 2218
+    },
+    {
+      "epoch": 0.424282982791587,
+      "grad_norm": 2.594043731689453,
+      "learning_rate": 4.895662060474246e-06,
+      "loss": 0.1224,
+      "step": 2219
+    },
+    {
+      "epoch": 0.42447418738049714,
+      "grad_norm": 2.592609405517578,
+      "learning_rate": 4.895948342609398e-06,
+      "loss": 0.5407,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4246653919694073,
+      "grad_norm": 1.1108200550079346,
+      "learning_rate": 4.8962344958176664e-06,
+      "loss": 0.1188,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4248565965583174,
+      "grad_norm": 1.9550951719284058,
+      "learning_rate": 4.896520520215123e-06,
+      "loss": 0.1843,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4250478011472275,
+      "grad_norm": 2.217928886413574,
+      "learning_rate": 4.8968064159176835e-06,
+      "loss": 0.1519,
+      "step": 2223
+    },
+    {
+      "epoch": 0.42523900573613765,
+      "grad_norm": 2.6040468215942383,
+      "learning_rate": 4.897092183041107e-06,
+      "loss": 0.2007,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4254302103250478,
+      "grad_norm": 1.7855995893478394,
+      "learning_rate": 4.897377821700995e-06,
+      "loss": 0.1069,
+      "step": 2225
+    },
+    {
+      "epoch": 0.42562141491395794,
+      "grad_norm": 2.076179027557373,
+      "learning_rate": 4.897663332012795e-06,
+      "loss": 0.4742,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4258126195028681,
+      "grad_norm": 2.4149672985076904,
+      "learning_rate": 4.897948714091799e-06,
+      "loss": 0.2418,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4260038240917782,
+      "grad_norm": 2.047053575515747,
+      "learning_rate": 4.898233968053142e-06,
+      "loss": 0.2112,
+      "step": 2228
+    },
+    {
+      "epoch": 0.4261950286806883,
+      "grad_norm": 1.8152729272842407,
+      "learning_rate": 4.8985190940118074e-06,
+      "loss": 0.0942,
+      "step": 2229
+    },
+    {
+      "epoch": 0.42638623326959846,
+      "grad_norm": 1.7877713441848755,
+      "learning_rate": 4.898804092082618e-06,
+      "loss": 0.1056,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4265774378585086,
+      "grad_norm": 1.7800778150558472,
+      "learning_rate": 4.899088962380248e-06,
+      "loss": 0.0929,
+      "step": 2231
+    },
+    {
+      "epoch": 0.42676864244741874,
+      "grad_norm": 3.8843071460723877,
+      "learning_rate": 4.899373705019215e-06,
+      "loss": 0.628,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4269598470363289,
+      "grad_norm": 3.350219964981079,
+      "learning_rate": 4.899658320113882e-06,
+      "loss": 0.591,
+      "step": 2233
+    },
+    {
+      "epoch": 0.42715105162523903,
+      "grad_norm": 1.9053051471710205,
+      "learning_rate": 4.899942807778461e-06,
+      "loss": 0.1806,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4273422562141491,
+      "grad_norm": 1.257057547569275,
+      "learning_rate": 4.900227168127006e-06,
+      "loss": 0.0753,
+      "step": 2235
+    },
+    {
+      "epoch": 0.42753346080305926,
+      "grad_norm": 1.812961220741272,
+      "learning_rate": 4.900511401273424e-06,
+      "loss": 0.1633,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4277246653919694,
+      "grad_norm": 2.06535267829895,
+      "learning_rate": 4.900795507331465e-06,
+      "loss": 0.1393,
+      "step": 2237
+    },
+    {
+      "epoch": 0.42791586998087955,
+      "grad_norm": 1.0033485889434814,
+      "learning_rate": 4.901079486414725e-06,
+      "loss": 0.0774,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4281070745697897,
+      "grad_norm": 2.0232021808624268,
+      "learning_rate": 4.9013633386366545e-06,
+      "loss": 0.2666,
+      "step": 2239
+    },
+    {
+      "epoch": 0.42829827915869984,
+      "grad_norm": 2.4945590496063232,
+      "learning_rate": 4.901647064110545e-06,
+      "loss": 0.4225,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4284894837476099,
+      "grad_norm": 2.9069554805755615,
+      "learning_rate": 4.901930662949541e-06,
+      "loss": 0.4286,
+      "step": 2241
+    },
+    {
+      "epoch": 0.42868068833652007,
+      "grad_norm": 1.4231261014938354,
+      "learning_rate": 4.902214135266632e-06,
+      "loss": 0.072,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4288718929254302,
+      "grad_norm": 1.2749229669570923,
+      "learning_rate": 4.902497481174659e-06,
+      "loss": 0.1584,
+      "step": 2243
+    },
+    {
+      "epoch": 0.42906309751434035,
+      "grad_norm": 3.1102404594421387,
+      "learning_rate": 4.902780700786311e-06,
+      "loss": 0.163,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4292543021032505,
+      "grad_norm": 2.0278851985931396,
+      "learning_rate": 4.903063794214126e-06,
+      "loss": 0.4301,
+      "step": 2245
+    },
+    {
+      "epoch": 0.42944550669216064,
+      "grad_norm": 2.2186572551727295,
+      "learning_rate": 4.903346761570493e-06,
+      "loss": 0.0785,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4296367112810707,
+      "grad_norm": 2.2448575496673584,
+      "learning_rate": 4.90362960296765e-06,
+      "loss": 0.2338,
+      "step": 2247
+    },
+    {
+      "epoch": 0.42982791586998087,
+      "grad_norm": 2.3453407287597656,
+      "learning_rate": 4.903912318517684e-06,
+      "loss": 0.269,
+      "step": 2248
+    },
+    {
+      "epoch": 0.430019120458891,
+      "grad_norm": 1.2500512599945068,
+      "learning_rate": 4.904194908332537e-06,
+      "loss": 0.0725,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43021032504780116,
+      "grad_norm": 2.4721336364746094,
+      "learning_rate": 4.904477372523995e-06,
+      "loss": 0.07,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4304015296367113,
+      "grad_norm": 2.766767740249634,
+      "learning_rate": 4.9047597112037e-06,
+      "loss": 0.5043,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4305927342256214,
+      "grad_norm": 1.717307686805725,
+      "learning_rate": 4.905041924483143e-06,
+      "loss": 0.1655,
+      "step": 2252
+    },
+    {
+      "epoch": 0.43078393881453153,
+      "grad_norm": 3.6042754650115967,
+      "learning_rate": 4.905324012473667e-06,
+      "loss": 0.4827,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4309751434034417,
+      "grad_norm": 1.890688419342041,
+      "learning_rate": 4.905605975286469e-06,
+      "loss": 0.1853,
+      "step": 2254
+    },
+    {
+      "epoch": 0.4311663479923518,
+      "grad_norm": 1.5629879236221313,
+      "learning_rate": 4.9058878130325935e-06,
+      "loss": 0.1271,
+      "step": 2255
+    },
+    {
+      "epoch": 0.43135755258126196,
+      "grad_norm": 2.144808053970337,
+      "learning_rate": 4.906169525822942e-06,
+      "loss": 0.1081,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4315487571701721,
+      "grad_norm": 1.9382816553115845,
+      "learning_rate": 4.9064511137682635e-06,
+      "loss": 0.2521,
+      "step": 2257
+    },
+    {
+      "epoch": 0.4317399617590822,
+      "grad_norm": 1.2302849292755127,
+      "learning_rate": 4.906732576979165e-06,
+      "loss": 0.1157,
+      "step": 2258
+    },
+    {
+      "epoch": 0.43193116634799233,
+      "grad_norm": 1.9480595588684082,
+      "learning_rate": 4.907013915566105e-06,
+      "loss": 0.22,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4321223709369025,
+      "grad_norm": 1.205723762512207,
+      "learning_rate": 4.907295129639391e-06,
+      "loss": 0.0779,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4323135755258126,
+      "grad_norm": 2.0031495094299316,
+      "learning_rate": 4.907576219309192e-06,
+      "loss": 0.2669,
+      "step": 2261
+    },
+    {
+      "epoch": 0.43250478011472276,
+      "grad_norm": 1.9002723693847656,
+      "learning_rate": 4.907857184685524e-06,
+      "loss": 0.0908,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4326959847036329,
+      "grad_norm": 2.721759796142578,
+      "learning_rate": 4.9081380258782595e-06,
+      "loss": 0.2764,
+      "step": 2263
+    },
+    {
+      "epoch": 0.432887189292543,
+      "grad_norm": 3.628314733505249,
+      "learning_rate": 4.908418742997127e-06,
+      "loss": 0.768,
+      "step": 2264
+    },
+    {
+      "epoch": 0.43307839388145314,
+      "grad_norm": 1.3902168273925781,
+      "learning_rate": 4.908699336151707e-06,
+      "loss": 0.1548,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4332695984703633,
+      "grad_norm": 2.48300838470459,
+      "learning_rate": 4.908979805451437e-06,
+      "loss": 0.4893,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4334608030592734,
+      "grad_norm": 2.2376482486724854,
+      "learning_rate": 4.909260151005608e-06,
+      "loss": 0.3067,
+      "step": 2267
+    },
+    {
+      "epoch": 0.43365200764818357,
+      "grad_norm": 1.2431788444519043,
+      "learning_rate": 4.909540372923367e-06,
+      "loss": 0.0815,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4338432122370937,
+      "grad_norm": 2.4078571796417236,
+      "learning_rate": 4.9098204713137175e-06,
+      "loss": 0.2497,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4340344168260038,
+      "grad_norm": 2.028752088546753,
+      "learning_rate": 4.910100446285518e-06,
+      "loss": 0.3107,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43422562141491394,
+      "grad_norm": 2.2157468795776367,
+      "learning_rate": 4.910380297947484e-06,
+      "loss": 0.2083,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4344168260038241,
+      "grad_norm": 2.5187478065490723,
+      "learning_rate": 4.910660026408185e-06,
+      "loss": 0.2862,
+      "step": 2272
+    },
+    {
+      "epoch": 0.43460803059273423,
+      "grad_norm": 2.122492790222168,
+      "learning_rate": 4.910939631776051e-06,
+      "loss": 0.1934,
+      "step": 2273
+    },
+    {
+      "epoch": 0.4347992351816444,
+      "grad_norm": 1.5224970579147339,
+      "learning_rate": 4.911219114159365e-06,
+      "loss": 0.0709,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4349904397705545,
+      "grad_norm": 1.999548077583313,
+      "learning_rate": 4.91149847366627e-06,
+      "loss": 0.1255,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4351816443594646,
+      "grad_norm": 1.97611665725708,
+      "learning_rate": 4.911777710404766e-06,
+      "loss": 0.1653,
+      "step": 2276
+    },
+    {
+      "epoch": 0.43537284894837475,
+      "grad_norm": 1.5954374074935913,
+      "learning_rate": 4.91205682448271e-06,
+      "loss": 0.365,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4355640535372849,
+      "grad_norm": 1.9923690557479858,
+      "learning_rate": 4.9123358160078154e-06,
+      "loss": 0.4458,
+      "step": 2278
+    },
+    {
+      "epoch": 0.43575525812619503,
+      "grad_norm": 2.6444168090820312,
+      "learning_rate": 4.912614685087658e-06,
+      "loss": 0.3102,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4359464627151052,
+      "grad_norm": 3.0847392082214355,
+      "learning_rate": 4.9128934318296675e-06,
+      "loss": 0.0966,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4361376673040153,
+      "grad_norm": 1.3775651454925537,
+      "learning_rate": 4.913172056341135e-06,
+      "loss": 0.0513,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4363288718929254,
+      "grad_norm": 2.156324625015259,
+      "learning_rate": 4.91345055872921e-06,
+      "loss": 0.3312,
+      "step": 2282
+    },
+    {
+      "epoch": 0.43652007648183555,
+      "grad_norm": 2.418484926223755,
+      "learning_rate": 4.913728939100901e-06,
+      "loss": 0.1921,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4367112810707457,
+      "grad_norm": 0.8661771416664124,
+      "learning_rate": 4.914007197563076e-06,
+      "loss": 0.0958,
+      "step": 2284
+    },
+    {
+      "epoch": 0.43690248565965584,
+      "grad_norm": 2.313511371612549,
+      "learning_rate": 4.914285334222461e-06,
+      "loss": 0.3075,
+      "step": 2285
+    },
+    {
+      "epoch": 0.437093690248566,
+      "grad_norm": 0.9290204048156738,
+      "learning_rate": 4.914563349185646e-06,
+      "loss": 0.091,
+      "step": 2286
+    },
+    {
+      "epoch": 0.4372848948374761,
+      "grad_norm": 1.7797731161117554,
+      "learning_rate": 4.914841242559077e-06,
+      "loss": 0.117,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4374760994263862,
+      "grad_norm": 3.0950605869293213,
+      "learning_rate": 4.915119014449062e-06,
+      "loss": 0.4731,
+      "step": 2288
+    },
+    {
+      "epoch": 0.43766730401529635,
+      "grad_norm": 2.342931032180786,
+      "learning_rate": 4.915396664961769e-06,
+      "loss": 0.1476,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4378585086042065,
+      "grad_norm": 2.6388490200042725,
+      "learning_rate": 4.915674194203229e-06,
+      "loss": 0.5004,
+      "step": 2290
+    },
+    {
+      "epoch": 0.43804971319311664,
+      "grad_norm": 1.8596402406692505,
+      "learning_rate": 4.9159516022793316e-06,
+      "loss": 0.1407,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4382409177820268,
+      "grad_norm": 2.9163615703582764,
+      "learning_rate": 4.916228889295829e-06,
+      "loss": 0.272,
+      "step": 2292
+    },
+    {
+      "epoch": 0.4384321223709369,
+      "grad_norm": 4.297892093658447,
+      "learning_rate": 4.916506055358336e-06,
+      "loss": 0.1782,
+      "step": 2293
+    },
+    {
+      "epoch": 0.438623326959847,
+      "grad_norm": 1.6136482954025269,
+      "learning_rate": 4.916783100572327e-06,
+      "loss": 0.1081,
+      "step": 2294
+    },
+    {
+      "epoch": 0.43881453154875716,
+      "grad_norm": 2.648573398590088,
+      "learning_rate": 4.917060025043139e-06,
+      "loss": 0.369,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4390057361376673,
+      "grad_norm": 2.309370756149292,
+      "learning_rate": 4.917336828875972e-06,
+      "loss": 0.3044,
+      "step": 2296
+    },
+    {
+      "epoch": 0.43919694072657744,
+      "grad_norm": 2.0419301986694336,
+      "learning_rate": 4.91761351217589e-06,
+      "loss": 0.1293,
+      "step": 2297
+    },
+    {
+      "epoch": 0.4393881453154876,
+      "grad_norm": 1.3703843355178833,
+      "learning_rate": 4.917890075047817e-06,
+      "loss": 0.0763,
+      "step": 2298
+    },
+    {
+      "epoch": 0.43957934990439773,
+      "grad_norm": 2.651688814163208,
+      "learning_rate": 4.918166517596542e-06,
+      "loss": 0.0794,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4397705544933078,
+      "grad_norm": 2.5015828609466553,
+      "learning_rate": 4.918442839926716e-06,
+      "loss": 0.1203,
+      "step": 2300
+    },
+    {
+      "epoch": 0.43996175908221796,
+      "grad_norm": 3.4868128299713135,
+      "learning_rate": 4.918719042142854e-06,
+      "loss": 0.3761,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4401529636711281,
+      "grad_norm": 3.072807550430298,
+      "learning_rate": 4.918995124349335e-06,
+      "loss": 0.4009,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44034416826003825,
+      "grad_norm": 1.4773715734481812,
+      "learning_rate": 4.9192710866504036e-06,
+      "loss": 0.1333,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4405353728489484,
+      "grad_norm": 1.7132066488265991,
+      "learning_rate": 4.919546929150165e-06,
+      "loss": 0.2852,
+      "step": 2304
+    },
+    {
+      "epoch": 0.44072657743785854,
+      "grad_norm": 1.9559893608093262,
+      "learning_rate": 4.919822651952591e-06,
+      "loss": 0.0896,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4409177820267686,
+      "grad_norm": 3.859649658203125,
+      "learning_rate": 4.920098255161518e-06,
+      "loss": 0.1275,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44110898661567877,
+      "grad_norm": 2.1365041732788086,
+      "learning_rate": 4.920373738880649e-06,
+      "loss": 0.2009,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4413001912045889,
+      "grad_norm": 2.118523120880127,
+      "learning_rate": 4.920649103213549e-06,
+      "loss": 0.2082,
+      "step": 2308
+    },
+    {
+      "epoch": 0.44149139579349905,
+      "grad_norm": 3.3567910194396973,
+      "learning_rate": 4.92092434826365e-06,
+      "loss": 0.2772,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4416826003824092,
+      "grad_norm": 2.266333818435669,
+      "learning_rate": 4.921199474134248e-06,
+      "loss": 0.2863,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4418738049713193,
+      "grad_norm": 2.4705424308776855,
+      "learning_rate": 4.921474480928509e-06,
+      "loss": 0.1557,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4420650095602294,
+      "grad_norm": 0.8634752631187439,
+      "learning_rate": 4.921749368749461e-06,
+      "loss": 0.0341,
+      "step": 2312
+    },
+    {
+      "epoch": 0.44225621414913957,
+      "grad_norm": 3.7431929111480713,
+      "learning_rate": 4.9220241377e-06,
+      "loss": 0.337,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4424474187380497,
+      "grad_norm": 3.0681843757629395,
+      "learning_rate": 4.92229878788289e-06,
+      "loss": 0.3988,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44263862332695986,
+      "grad_norm": 1.9662657976150513,
+      "learning_rate": 4.922573319400757e-06,
+      "loss": 0.2526,
+      "step": 2315
+    },
+    {
+      "epoch": 0.44282982791587,
+      "grad_norm": 2.1128218173980713,
+      "learning_rate": 4.9228477323561e-06,
+      "loss": 0.1578,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4430210325047801,
+      "grad_norm": 2.3029232025146484,
+      "learning_rate": 4.923122026851282e-06,
+      "loss": 0.188,
+      "step": 2317
+    },
+    {
+      "epoch": 0.44321223709369023,
+      "grad_norm": 1.4844509363174438,
+      "learning_rate": 4.923396202988534e-06,
+      "loss": 0.0395,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4434034416826004,
+      "grad_norm": 1.5008915662765503,
+      "learning_rate": 4.923670260869953e-06,
+      "loss": 0.0757,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4435946462715105,
+      "grad_norm": 3.172825813293457,
+      "learning_rate": 4.923944200597508e-06,
+      "loss": 0.5372,
+      "step": 2320
+    },
+    {
+      "epoch": 0.44378585086042066,
+      "grad_norm": 3.6113059520721436,
+      "learning_rate": 4.924218022273032e-06,
+      "loss": 0.4811,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4439770554493308,
+      "grad_norm": 1.0926949977874756,
+      "learning_rate": 4.9244917259982296e-06,
+      "loss": 0.1378,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4441682600382409,
+      "grad_norm": 2.9779396057128906,
+      "learning_rate": 4.9247653118746715e-06,
+      "loss": 0.1651,
+      "step": 2323
+    },
+    {
+      "epoch": 0.44435946462715104,
+      "grad_norm": 1.3684827089309692,
+      "learning_rate": 4.9250387800038e-06,
+      "loss": 0.1091,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4445506692160612,
+      "grad_norm": 1.5437180995941162,
+      "learning_rate": 4.9253121304869235e-06,
+      "loss": 0.0794,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4447418738049713,
+      "grad_norm": 1.8073052167892456,
+      "learning_rate": 4.925585363425222e-06,
+      "loss": 0.1806,
+      "step": 2326
+    },
+    {
+      "epoch": 0.44493307839388146,
+      "grad_norm": 2.79695463180542,
+      "learning_rate": 4.925858478919743e-06,
+      "loss": 0.4065,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4451242829827916,
+      "grad_norm": 2.834144115447998,
+      "learning_rate": 4.926131477071406e-06,
+      "loss": 0.4724,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4453154875717017,
+      "grad_norm": 1.8955655097961426,
+      "learning_rate": 4.926404357980999e-06,
+      "loss": 0.1503,
+      "step": 2329
+    },
+    {
+      "epoch": 0.44550669216061184,
+      "grad_norm": 2.6821250915527344,
+      "learning_rate": 4.9266771217491796e-06,
+      "loss": 0.172,
+      "step": 2330
+    },
+    {
+      "epoch": 0.445697896749522,
+      "grad_norm": 1.803351640701294,
+      "learning_rate": 4.926949768476479e-06,
+      "loss": 0.1085,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4458891013384321,
+      "grad_norm": 1.929173469543457,
+      "learning_rate": 4.927222298263295e-06,
+      "loss": 0.1473,
+      "step": 2332
+    },
+    {
+      "epoch": 0.44608030592734227,
+      "grad_norm": 1.8458374738693237,
+      "learning_rate": 4.927494711209899e-06,
+      "loss": 0.2217,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4462715105162524,
+      "grad_norm": 2.2545127868652344,
+      "learning_rate": 4.927767007416432e-06,
+      "loss": 0.2251,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4464627151051625,
+      "grad_norm": 2.347503185272217,
+      "learning_rate": 4.928039186982908e-06,
+      "loss": 0.2747,
+      "step": 2335
+    },
+    {
+      "epoch": 0.44665391969407264,
+      "grad_norm": 2.8446438312530518,
+      "learning_rate": 4.92831125000921e-06,
+      "loss": 0.2503,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4468451242829828,
+      "grad_norm": 1.7588080167770386,
+      "learning_rate": 4.928583196595095e-06,
+      "loss": 0.0731,
+      "step": 2337
+    },
+    {
+      "epoch": 0.44703632887189293,
+      "grad_norm": 2.6320197582244873,
+      "learning_rate": 4.92885502684019e-06,
+      "loss": 0.1976,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4472275334608031,
+      "grad_norm": 2.0502002239227295,
+      "learning_rate": 4.929126740843998e-06,
+      "loss": 0.2047,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4474187380497132,
+      "grad_norm": 1.7559682130813599,
+      "learning_rate": 4.929398338705889e-06,
+      "loss": 0.1507,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4476099426386233,
+      "grad_norm": 1.752091407775879,
+      "learning_rate": 4.92966982052511e-06,
+      "loss": 0.1281,
+      "step": 2341
+    },
+    {
+      "epoch": 0.44780114722753345,
+      "grad_norm": 3.454530954360962,
+      "learning_rate": 4.92994118640078e-06,
+      "loss": 0.2777,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4479923518164436,
+      "grad_norm": 3.0871379375457764,
+      "learning_rate": 4.930212436431887e-06,
+      "loss": 0.2577,
+      "step": 2343
+    },
+    {
+      "epoch": 0.44818355640535373,
+      "grad_norm": 2.506565809249878,
+      "learning_rate": 4.9304835707173e-06,
+      "loss": 0.1185,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4483747609942639,
+      "grad_norm": 2.834341287612915,
+      "learning_rate": 4.930754589355753e-06,
+      "loss": 0.4537,
+      "step": 2345
+    },
+    {
+      "epoch": 0.448565965583174,
+      "grad_norm": 3.0927183628082275,
+      "learning_rate": 4.931025492445859e-06,
+      "loss": 0.6336,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4487571701720841,
+      "grad_norm": 2.212230920791626,
+      "learning_rate": 4.931296280086106e-06,
+      "loss": 0.2144,
+      "step": 2347
+    },
+    {
+      "epoch": 0.44894837476099425,
+      "grad_norm": 1.8910106420516968,
+      "learning_rate": 4.931566952374851e-06,
+      "loss": 0.1857,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4491395793499044,
+      "grad_norm": 1.7347790002822876,
+      "learning_rate": 4.93183750941033e-06,
+      "loss": 0.094,
+      "step": 2349
+    },
+    {
+      "epoch": 0.44933078393881454,
+      "grad_norm": 2.411543846130371,
+      "learning_rate": 4.93210795129065e-06,
+      "loss": 0.1578,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4495219885277247,
+      "grad_norm": 1.223103642463684,
+      "learning_rate": 4.932378278113796e-06,
+      "loss": 0.1855,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4497131931166348,
+      "grad_norm": 2.896897554397583,
+      "learning_rate": 4.932648489977627e-06,
+      "loss": 0.562,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4499043977055449,
+      "grad_norm": 2.2476577758789062,
+      "learning_rate": 4.932918586979875e-06,
+      "loss": 0.2595,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45009560229445505,
+      "grad_norm": 1.3098033666610718,
+      "learning_rate": 4.93318856921815e-06,
+      "loss": 0.0972,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4502868068833652,
+      "grad_norm": 1.3734967708587646,
+      "learning_rate": 4.933458436789937e-06,
+      "loss": 0.1431,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45047801147227534,
+      "grad_norm": 1.1524994373321533,
+      "learning_rate": 4.933728189792596e-06,
+      "loss": 0.0973,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4506692160611855,
+      "grad_norm": 3.060199022293091,
+      "learning_rate": 4.933997828323365e-06,
+      "loss": 0.4424,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45086042065009563,
+      "grad_norm": 3.3533732891082764,
+      "learning_rate": 4.934267352479356e-06,
+      "loss": 0.3997,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4510516252390057,
+      "grad_norm": 2.261230945587158,
+      "learning_rate": 4.934536762357558e-06,
+      "loss": 0.2667,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45124282982791586,
+      "grad_norm": 1.9926918745040894,
+      "learning_rate": 4.934806058054837e-06,
+      "loss": 0.1408,
+      "step": 2360
+    },
+    {
+      "epoch": 0.451434034416826,
+      "grad_norm": 1.3215872049331665,
+      "learning_rate": 4.935075239667939e-06,
+      "loss": 0.1143,
+      "step": 2361
+    },
+    {
+      "epoch": 0.45162523900573615,
+      "grad_norm": 2.423924684524536,
+      "learning_rate": 4.93534430729348e-06,
+      "loss": 0.1111,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4518164435946463,
+      "grad_norm": 2.364931583404541,
+      "learning_rate": 4.935613261027959e-06,
+      "loss": 0.3041,
+      "step": 2363
+    },
+    {
+      "epoch": 0.45200764818355643,
+      "grad_norm": 3.6321725845336914,
+      "learning_rate": 4.9358821009677516e-06,
+      "loss": 0.7906,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4521988527724665,
+      "grad_norm": 1.2970473766326904,
+      "learning_rate": 4.93615082720911e-06,
+      "loss": 0.1492,
+      "step": 2365
+    },
+    {
+      "epoch": 0.45239005736137666,
+      "grad_norm": 1.3972055912017822,
+      "learning_rate": 4.936419439848165e-06,
+      "loss": 0.2284,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4525812619502868,
+      "grad_norm": 2.145207166671753,
+      "learning_rate": 4.936687938980925e-06,
+      "loss": 0.1145,
+      "step": 2367
+    },
+    {
+      "epoch": 0.45277246653919695,
+      "grad_norm": 1.5136973857879639,
+      "learning_rate": 4.936956324703276e-06,
+      "loss": 0.1319,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4529636711281071,
+      "grad_norm": 2.3273026943206787,
+      "learning_rate": 4.937224597110986e-06,
+      "loss": 0.2864,
+      "step": 2369
+    },
+    {
+      "epoch": 0.45315487571701724,
+      "grad_norm": 2.093433141708374,
+      "learning_rate": 4.937492756299697e-06,
+      "loss": 0.1958,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4533460803059273,
+      "grad_norm": 1.7759649753570557,
+      "learning_rate": 4.937760802364934e-06,
+      "loss": 0.3099,
+      "step": 2371
+    },
+    {
+      "epoch": 0.45353728489483747,
+      "grad_norm": 2.3750076293945312,
+      "learning_rate": 4.938028735402098e-06,
+      "loss": 0.2607,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4537284894837476,
+      "grad_norm": 1.7167375087738037,
+      "learning_rate": 4.938296555506473e-06,
+      "loss": 0.1179,
+      "step": 2373
+    },
+    {
+      "epoch": 0.45391969407265775,
+      "grad_norm": 2.261927604675293,
+      "learning_rate": 4.938564262773219e-06,
+      "loss": 0.1355,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4541108986615679,
+      "grad_norm": 1.3687342405319214,
+      "learning_rate": 4.938831857297376e-06,
+      "loss": 0.0843,
+      "step": 2375
+    },
+    {
+      "epoch": 0.454302103250478,
+      "grad_norm": 2.428593158721924,
+      "learning_rate": 4.939099339173867e-06,
+      "loss": 0.1995,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4544933078393881,
+      "grad_norm": 1.6332013607025146,
+      "learning_rate": 4.9393667084974925e-06,
+      "loss": 0.1263,
+      "step": 2377
+    },
+    {
+      "epoch": 0.45468451242829827,
+      "grad_norm": 1.8396623134613037,
+      "learning_rate": 4.9396339653629345e-06,
+      "loss": 0.153,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4548757170172084,
+      "grad_norm": 1.5843632221221924,
+      "learning_rate": 4.939901109864755e-06,
+      "loss": 0.1409,
+      "step": 2379
+    },
+    {
+      "epoch": 0.45506692160611856,
+      "grad_norm": 2.6990528106689453,
+      "learning_rate": 4.940168142097398e-06,
+      "loss": 0.2067,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4552581261950287,
+      "grad_norm": 1.8671232461929321,
+      "learning_rate": 4.940435062155186e-06,
+      "loss": 0.105,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4554493307839388,
+      "grad_norm": 2.4851431846618652,
+      "learning_rate": 4.940701870132325e-06,
+      "loss": 0.277,
+      "step": 2382
+    },
+    {
+      "epoch": 0.45564053537284893,
+      "grad_norm": 3.0492827892303467,
+      "learning_rate": 4.940968566122902e-06,
+      "loss": 0.4704,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4558317399617591,
+      "grad_norm": 3.329960823059082,
+      "learning_rate": 4.941235150220885e-06,
+      "loss": 0.1627,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4560229445506692,
+      "grad_norm": 2.0267832279205322,
+      "learning_rate": 4.941501622520123e-06,
+      "loss": 0.1374,
+      "step": 2385
+    },
+    {
+      "epoch": 0.45621414913957936,
+      "grad_norm": 1.8187408447265625,
+      "learning_rate": 4.941767983114349e-06,
+      "loss": 0.2506,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4564053537284895,
+      "grad_norm": 1.1165252923965454,
+      "learning_rate": 4.942034232097177e-06,
+      "loss": 0.1032,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4565965583173996,
+      "grad_norm": 2.010648250579834,
+      "learning_rate": 4.942300369562102e-06,
+      "loss": 0.1881,
+      "step": 2388
+    },
+    {
+      "epoch": 0.45678776290630974,
+      "grad_norm": 4.388535499572754,
+      "learning_rate": 4.942566395602506e-06,
+      "loss": 0.338,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4569789674952199,
+      "grad_norm": 2.1628904342651367,
+      "learning_rate": 4.942832310311647e-06,
+      "loss": 0.2078,
+      "step": 2390
+    },
+    {
+      "epoch": 0.45717017208413,
+      "grad_norm": 1.473975658416748,
+      "learning_rate": 4.943098113782672e-06,
+      "loss": 0.1039,
+      "step": 2391
+    },
+    {
+      "epoch": 0.45736137667304017,
+      "grad_norm": 2.243346691131592,
+      "learning_rate": 4.94336380610861e-06,
+      "loss": 0.0887,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4575525812619503,
+      "grad_norm": 1.2243520021438599,
+      "learning_rate": 4.9436293873823705e-06,
+      "loss": 0.1352,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4577437858508604,
+      "grad_norm": 1.9870951175689697,
+      "learning_rate": 4.943894857696749e-06,
+      "loss": 0.1092,
+      "step": 2394
+    },
+    {
+      "epoch": 0.45793499043977054,
+      "grad_norm": 1.85684335231781,
+      "learning_rate": 4.9441602171444225e-06,
+      "loss": 0.1827,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4581261950286807,
+      "grad_norm": 2.156700849533081,
+      "learning_rate": 4.944425465817956e-06,
+      "loss": 0.1789,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4583173996175908,
+      "grad_norm": 1.204390287399292,
+      "learning_rate": 4.944690603809794e-06,
+      "loss": 0.0855,
+      "step": 2397
+    },
+    {
+      "epoch": 0.45850860420650097,
+      "grad_norm": 1.6291753053665161,
+      "learning_rate": 4.944955631212269e-06,
+      "loss": 0.1035,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4586998087954111,
+      "grad_norm": 2.399693012237549,
+      "learning_rate": 4.945220548117595e-06,
+      "loss": 0.2533,
+      "step": 2399
+    },
+    {
+      "epoch": 0.4588910133843212,
+      "grad_norm": 1.6935648918151855,
+      "learning_rate": 4.945485354617874e-06,
+      "loss": 0.089,
+      "step": 2400
+    },
+    {
+      "epoch": 0.45908221797323134,
+      "grad_norm": 1.8837876319885254,
+      "learning_rate": 4.945750050805088e-06,
+      "loss": 0.2955,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4592734225621415,
+      "grad_norm": 1.520376205444336,
+      "learning_rate": 4.946014636771111e-06,
+      "loss": 0.0884,
+      "step": 2402
+    },
+    {
+      "epoch": 0.45946462715105163,
+      "grad_norm": 6.492906093597412,
+      "learning_rate": 4.946279112607695e-06,
+      "loss": 0.8193,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4596558317399618,
+      "grad_norm": 2.74912691116333,
+      "learning_rate": 4.946543478406484e-06,
+      "loss": 0.2926,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4598470363288719,
+      "grad_norm": 1.9996367692947388,
+      "learning_rate": 4.946807734259001e-06,
+      "loss": 0.1018,
+      "step": 2405
+    },
+    {
+      "epoch": 0.460038240917782,
+      "grad_norm": 1.903185486793518,
+      "learning_rate": 4.947071880256661e-06,
+      "loss": 0.243,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46022944550669215,
+      "grad_norm": 2.6955156326293945,
+      "learning_rate": 4.947335916490763e-06,
+      "loss": 0.3418,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4604206500956023,
+      "grad_norm": 4.031216144561768,
+      "learning_rate": 4.947599843052489e-06,
+      "loss": 0.4078,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46061185468451243,
+      "grad_norm": 2.110466241836548,
+      "learning_rate": 4.947863660032912e-06,
+      "loss": 0.2372,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4608030592734226,
+      "grad_norm": 1.633364200592041,
+      "learning_rate": 4.9481273675229894e-06,
+      "loss": 0.1019,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4609942638623327,
+      "grad_norm": 1.452785611152649,
+      "learning_rate": 4.948390965613565e-06,
+      "loss": 0.0943,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4611854684512428,
+      "grad_norm": 2.504720449447632,
+      "learning_rate": 4.948654454395372e-06,
+      "loss": 0.1355,
+      "step": 2412
+    },
+    {
+      "epoch": 0.46137667304015295,
+      "grad_norm": 2.703850269317627,
+      "learning_rate": 4.948917833959027e-06,
+      "loss": 0.3414,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4615678776290631,
+      "grad_norm": 2.818969964981079,
+      "learning_rate": 4.949181104395038e-06,
+      "loss": 0.528,
+      "step": 2414
+    },
+    {
+      "epoch": 0.46175908221797324,
+      "grad_norm": 2.11025071144104,
+      "learning_rate": 4.949444265793797e-06,
+      "loss": 0.1207,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4619502868068834,
+      "grad_norm": 2.019381523132324,
+      "learning_rate": 4.949707318245586e-06,
+      "loss": 0.1208,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4621414913957935,
+      "grad_norm": 2.8823390007019043,
+      "learning_rate": 4.949970261840574e-06,
+      "loss": 0.3685,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4623326959847036,
+      "grad_norm": 1.2599728107452393,
+      "learning_rate": 4.950233096668818e-06,
+      "loss": 0.0653,
+      "step": 2418
+    },
+    {
+      "epoch": 0.46252390057361376,
+      "grad_norm": 2.618422508239746,
+      "learning_rate": 4.950495822820266e-06,
+      "loss": 0.1693,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4627151051625239,
+      "grad_norm": 2.234100341796875,
+      "learning_rate": 4.950758440384748e-06,
+      "loss": 0.1476,
+      "step": 2420
+    },
+    {
+      "epoch": 0.46290630975143404,
+      "grad_norm": 1.7235227823257446,
+      "learning_rate": 4.95102094945199e-06,
+      "loss": 0.1403,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4630975143403442,
+      "grad_norm": 3.084684133529663,
+      "learning_rate": 4.951283350111601e-06,
+      "loss": 0.4766,
+      "step": 2422
+    },
+    {
+      "epoch": 0.46328871892925433,
+      "grad_norm": 1.7967596054077148,
+      "learning_rate": 4.951545642453086e-06,
+      "loss": 0.1491,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4634799235181644,
+      "grad_norm": 2.608936071395874,
+      "learning_rate": 4.95180782656583e-06,
+      "loss": 0.2224,
+      "step": 2424
+    },
+    {
+      "epoch": 0.46367112810707456,
+      "grad_norm": 1.7240525484085083,
+      "learning_rate": 4.952069902539114e-06,
+      "loss": 0.1001,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4638623326959847,
+      "grad_norm": 2.553189754486084,
+      "learning_rate": 4.952331870462108e-06,
+      "loss": 0.3537,
+      "step": 2426
+    },
+    {
+      "epoch": 0.46405353728489485,
+      "grad_norm": 1.4191901683807373,
+      "learning_rate": 4.952593730423869e-06,
+      "loss": 0.0765,
+      "step": 2427
+    },
+    {
+      "epoch": 0.464244741873805,
+      "grad_norm": 2.872032880783081,
+      "learning_rate": 4.952855482513347e-06,
+      "loss": 0.5042,
+      "step": 2428
+    },
+    {
+      "epoch": 0.46443594646271513,
+      "grad_norm": 1.699519157409668,
+      "learning_rate": 4.95311712681938e-06,
+      "loss": 0.2303,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4646271510516252,
+      "grad_norm": 3.224975109100342,
+      "learning_rate": 4.953378663430695e-06,
+      "loss": 0.1607,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46481835564053536,
+      "grad_norm": 2.4731459617614746,
+      "learning_rate": 4.953640092435914e-06,
+      "loss": 0.1178,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4650095602294455,
+      "grad_norm": 2.1776509284973145,
+      "learning_rate": 4.953901413923546e-06,
+      "loss": 0.2036,
+      "step": 2432
+    },
+    {
+      "epoch": 0.46520076481835565,
+      "grad_norm": 2.7886409759521484,
+      "learning_rate": 4.9541626279819915e-06,
+      "loss": 0.334,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4653919694072658,
+      "grad_norm": 1.9042888879776,
+      "learning_rate": 4.954423734699544e-06,
+      "loss": 0.2839,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4655831739961759,
+      "grad_norm": 3.304387092590332,
+      "learning_rate": 4.954684734164385e-06,
+      "loss": 0.5778,
+      "step": 2435
+    },
+    {
+      "epoch": 0.465774378585086,
+      "grad_norm": 2.7068235874176025,
+      "learning_rate": 4.954945626464589e-06,
+      "loss": 0.2185,
+      "step": 2436
+    },
+    {
+      "epoch": 0.46596558317399617,
+      "grad_norm": 2.0537827014923096,
+      "learning_rate": 4.955206411688123e-06,
+      "loss": 0.1307,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4661567877629063,
+      "grad_norm": 1.799789309501648,
+      "learning_rate": 4.955467089922844e-06,
+      "loss": 0.1601,
+      "step": 2438
+    },
+    {
+      "epoch": 0.46634799235181645,
+      "grad_norm": 2.1717841625213623,
+      "learning_rate": 4.955727661256503e-06,
+      "loss": 0.2017,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4665391969407266,
+      "grad_norm": 2.0859739780426025,
+      "learning_rate": 4.95598812577674e-06,
+      "loss": 0.1489,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4667304015296367,
+      "grad_norm": 1.309865117073059,
+      "learning_rate": 4.95624848357109e-06,
+      "loss": 0.0694,
+      "step": 2441
+    },
+    {
+      "epoch": 0.46692160611854683,
+      "grad_norm": 1.9916973114013672,
+      "learning_rate": 4.956508734726978e-06,
+      "loss": 0.1463,
+      "step": 2442
+    },
+    {
+      "epoch": 0.46711281070745697,
+      "grad_norm": 2.035287618637085,
+      "learning_rate": 4.956768879331726e-06,
+      "loss": 0.1042,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4673040152963671,
+      "grad_norm": 1.9258490800857544,
+      "learning_rate": 4.957028917472544e-06,
+      "loss": 0.1315,
+      "step": 2444
+    },
+    {
+      "epoch": 0.46749521988527726,
+      "grad_norm": 2.847909688949585,
+      "learning_rate": 4.957288849236539e-06,
+      "loss": 0.4624,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4676864244741874,
+      "grad_norm": 3.1480817794799805,
+      "learning_rate": 4.957548674710705e-06,
+      "loss": 0.3211,
+      "step": 2446
+    },
+    {
+      "epoch": 0.4678776290630975,
+      "grad_norm": 2.3119571208953857,
+      "learning_rate": 4.957808393981937e-06,
+      "loss": 0.2032,
+      "step": 2447
+    },
+    {
+      "epoch": 0.46806883365200763,
+      "grad_norm": 2.356323480606079,
+      "learning_rate": 4.9580680071370174e-06,
+      "loss": 0.2908,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4682600382409178,
+      "grad_norm": 0.683710515499115,
+      "learning_rate": 4.958327514262626e-06,
+      "loss": 0.0256,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4684512428298279,
+      "grad_norm": 2.147235155105591,
+      "learning_rate": 4.9585869154453355e-06,
+      "loss": 0.186,
+      "step": 2450
+    },
+    {
+      "epoch": 0.46864244741873806,
+      "grad_norm": 2.6111466884613037,
+      "learning_rate": 4.958846210771611e-06,
+      "loss": 0.3993,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4688336520076482,
+      "grad_norm": 1.6357756853103638,
+      "learning_rate": 4.959105400327814e-06,
+      "loss": 0.1385,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4690248565965583,
+      "grad_norm": 2.0699405670166016,
+      "learning_rate": 4.9593644842001994e-06,
+      "loss": 0.1961,
+      "step": 2453
+    },
+    {
+      "epoch": 0.46921606118546844,
+      "grad_norm": 2.4611475467681885,
+      "learning_rate": 4.959623462474916e-06,
+      "loss": 0.194,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4694072657743786,
+      "grad_norm": 2.70579195022583,
+      "learning_rate": 4.9598823352380075e-06,
+      "loss": 0.3494,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4695984703632887,
+      "grad_norm": 1.362758994102478,
+      "learning_rate": 4.9601411025754144e-06,
+      "loss": 0.0551,
+      "step": 2456
+    },
+    {
+      "epoch": 0.46978967495219887,
+      "grad_norm": 2.263684034347534,
+      "learning_rate": 4.960399764572971e-06,
+      "loss": 0.4414,
+      "step": 2457
+    },
+    {
+      "epoch": 0.469980879541109,
+      "grad_norm": 2.3129522800445557,
+      "learning_rate": 4.960658321316405e-06,
+      "loss": 0.337,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4701720841300191,
+      "grad_norm": 3.598054885864258,
+      "learning_rate": 4.960916772891341e-06,
+      "loss": 0.6028,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47036328871892924,
+      "grad_norm": 2.352046489715576,
+      "learning_rate": 4.9611751193833e-06,
+      "loss": 0.1063,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4705544933078394,
+      "grad_norm": 2.858907461166382,
+      "learning_rate": 4.9614333608776984e-06,
+      "loss": 0.3046,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4707456978967495,
+      "grad_norm": 1.6364027261734009,
+      "learning_rate": 4.9616914974598485e-06,
+      "loss": 0.0841,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47093690248565967,
+      "grad_norm": 3.1622374057769775,
+      "learning_rate": 4.961949529214955e-06,
+      "loss": 0.3993,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4711281070745698,
+      "grad_norm": 2.4431493282318115,
+      "learning_rate": 4.962207456228127e-06,
+      "loss": 0.3394,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4713193116634799,
+      "grad_norm": 1.4499131441116333,
+      "learning_rate": 4.96246527858436e-06,
+      "loss": 0.1154,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47151051625239004,
+      "grad_norm": 1.915971279144287,
+      "learning_rate": 4.962722996368555e-06,
+      "loss": 0.3057,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4717017208413002,
+      "grad_norm": 2.2309677600860596,
+      "learning_rate": 4.9629806096655045e-06,
+      "loss": 0.3469,
+      "step": 2467
+    },
+    {
+      "epoch": 0.47189292543021033,
+      "grad_norm": 1.4278203248977661,
+      "learning_rate": 4.963238118559899e-06,
+      "loss": 0.0782,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4720841300191205,
+      "grad_norm": 1.626359462738037,
+      "learning_rate": 4.963495523136326e-06,
+      "loss": 0.0824,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4722753346080306,
+      "grad_norm": 6.234444618225098,
+      "learning_rate": 4.96375282347927e-06,
+      "loss": 0.3704,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4724665391969407,
+      "grad_norm": 1.2472459077835083,
+      "learning_rate": 4.964010019673117e-06,
+      "loss": 0.1072,
+      "step": 2471
+    },
+    {
+      "epoch": 0.47265774378585085,
+      "grad_norm": 0.9773194193840027,
+      "learning_rate": 4.9642671118021435e-06,
+      "loss": 0.1166,
+      "step": 2472
+    },
+    {
+      "epoch": 0.472848948374761,
+      "grad_norm": 1.2697404623031616,
+      "learning_rate": 4.9645240999505284e-06,
+      "loss": 0.0847,
+      "step": 2473
+    },
+    {
+      "epoch": 0.47304015296367113,
+      "grad_norm": 1.7373942136764526,
+      "learning_rate": 4.9647809842023485e-06,
+      "loss": 0.1414,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4732313575525813,
+      "grad_norm": 1.2510534524917603,
+      "learning_rate": 4.965037764641576e-06,
+      "loss": 0.0646,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4734225621414914,
+      "grad_norm": 3.523763418197632,
+      "learning_rate": 4.965294441352084e-06,
+      "loss": 0.6183,
+      "step": 2476
+    },
+    {
+      "epoch": 0.4736137667304015,
+      "grad_norm": 2.536613941192627,
+      "learning_rate": 4.965551014417641e-06,
+      "loss": 0.307,
+      "step": 2477
+    },
+    {
+      "epoch": 0.47380497131931165,
+      "grad_norm": 1.2308690547943115,
+      "learning_rate": 4.965807483921919e-06,
+      "loss": 0.121,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4739961759082218,
+      "grad_norm": 1.2739108800888062,
+      "learning_rate": 4.966063849948484e-06,
+      "loss": 0.0985,
+      "step": 2479
+    },
+    {
+      "epoch": 0.47418738049713194,
+      "grad_norm": 4.273107528686523,
+      "learning_rate": 4.966320112580802e-06,
+      "loss": 0.1709,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4743785850860421,
+      "grad_norm": 2.6034741401672363,
+      "learning_rate": 4.96657627190224e-06,
+      "loss": 0.1497,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4745697896749522,
+      "grad_norm": 2.563988208770752,
+      "learning_rate": 4.966832327996062e-06,
+      "loss": 0.3513,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4747609942638623,
+      "grad_norm": 1.5361602306365967,
+      "learning_rate": 4.967088280945433e-06,
+      "loss": 0.1868,
+      "step": 2483
+    },
+    {
+      "epoch": 0.47495219885277246,
+      "grad_norm": 1.8172998428344727,
+      "learning_rate": 4.9673441308334165e-06,
+      "loss": 0.2232,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4751434034416826,
+      "grad_norm": 1.6137768030166626,
+      "learning_rate": 4.967599877742975e-06,
+      "loss": 0.1634,
+      "step": 2485
+    },
+    {
+      "epoch": 0.47533460803059274,
+      "grad_norm": 1.552061915397644,
+      "learning_rate": 4.967855521756973e-06,
+      "loss": 0.094,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4755258126195029,
+      "grad_norm": 2.0277292728424072,
+      "learning_rate": 4.9681110629581734e-06,
+      "loss": 0.1203,
+      "step": 2487
+    },
+    {
+      "epoch": 0.47571701720841303,
+      "grad_norm": 2.038464307785034,
+      "learning_rate": 4.96836650142924e-06,
+      "loss": 0.164,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4759082217973231,
+      "grad_norm": 2.4230268001556396,
+      "learning_rate": 4.968621837252737e-06,
+      "loss": 0.3603,
+      "step": 2489
+    },
+    {
+      "epoch": 0.47609942638623326,
+      "grad_norm": 1.4726464748382568,
+      "learning_rate": 4.9688770705111285e-06,
+      "loss": 0.1528,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4762906309751434,
+      "grad_norm": 1.6719284057617188,
+      "learning_rate": 4.969132201286779e-06,
+      "loss": 0.2506,
+      "step": 2491
+    },
+    {
+      "epoch": 0.47648183556405355,
+      "grad_norm": 1.5102533102035522,
+      "learning_rate": 4.969387229661954e-06,
+      "loss": 0.0726,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4766730401529637,
+      "grad_norm": 2.305950880050659,
+      "learning_rate": 4.969642155718823e-06,
+      "loss": 0.2434,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4768642447418738,
+      "grad_norm": 2.0720887184143066,
+      "learning_rate": 4.969896979539451e-06,
+      "loss": 0.249,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4770554493307839,
+      "grad_norm": 2.4147567749023438,
+      "learning_rate": 4.97015170120581e-06,
+      "loss": 0.3391,
+      "step": 2495
+    },
+    {
+      "epoch": 0.47724665391969406,
+      "grad_norm": 2.9447648525238037,
+      "learning_rate": 4.9704063207997684e-06,
+      "loss": 0.4681,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4774378585086042,
+      "grad_norm": 2.2584710121154785,
+      "learning_rate": 4.9706608384031e-06,
+      "loss": 0.1599,
+      "step": 2497
+    },
+    {
+      "epoch": 0.47762906309751435,
+      "grad_norm": 1.4370557069778442,
+      "learning_rate": 4.970915254097478e-06,
+      "loss": 0.1377,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4778202676864245,
+      "grad_norm": 1.5512670278549194,
+      "learning_rate": 4.971169567964479e-06,
+      "loss": 0.111,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4780114722753346,
+      "grad_norm": 1.5349520444869995,
+      "learning_rate": 4.9714237800855815e-06,
+      "loss": 0.0762,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4780114722753346,
+      "eval_runtime": 838.1854,
+      "eval_samples_per_second": 1.83,
+      "eval_steps_per_second": 0.229,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4782026768642447,
+      "grad_norm": 1.6777958869934082,
+      "learning_rate": 4.971677890542167e-06,
+      "loss": 0.1465,
+      "step": 2501
+    },
+    {
+      "epoch": 0.47839388145315487,
+      "grad_norm": 1.7863197326660156,
+      "learning_rate": 4.971931899415515e-06,
+      "loss": 0.1716,
+      "step": 2502
+    },
+    {
+      "epoch": 0.478585086042065,
+      "grad_norm": 1.1726549863815308,
+      "learning_rate": 4.972185806786815e-06,
+      "loss": 0.1211,
+      "step": 2503
+    },
+    {
+      "epoch": 0.47877629063097515,
+      "grad_norm": 3.095215320587158,
+      "learning_rate": 4.972439612737152e-06,
+      "loss": 0.3554,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4789674952198853,
+      "grad_norm": 1.327844262123108,
+      "learning_rate": 4.972693317347518e-06,
+      "loss": 0.0942,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4791586998087954,
+      "grad_norm": 1.6837940216064453,
+      "learning_rate": 4.9729469206988075e-06,
+      "loss": 0.0878,
+      "step": 2506
+    },
+    {
+      "epoch": 0.47934990439770553,
+      "grad_norm": 2.4635088443756104,
+      "learning_rate": 4.973200422871818e-06,
+      "loss": 0.1642,
+      "step": 2507
+    },
+    {
+      "epoch": 0.47954110898661567,
+      "grad_norm": 2.1021690368652344,
+      "learning_rate": 4.973453823947249e-06,
+      "loss": 0.3213,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4797323135755258,
+      "grad_norm": 1.8503338098526,
+      "learning_rate": 4.973707124005704e-06,
+      "loss": 0.1949,
+      "step": 2509
+    },
+    {
+      "epoch": 0.47992351816443596,
+      "grad_norm": 1.9696707725524902,
+      "learning_rate": 4.973960323127691e-06,
+      "loss": 0.3777,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4801147227533461,
+      "grad_norm": 1.7098878622055054,
+      "learning_rate": 4.974213421393625e-06,
+      "loss": 0.1056,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4803059273422562,
+      "grad_norm": 2.5690712928771973,
+      "learning_rate": 4.974466418883816e-06,
+      "loss": 0.1538,
+      "step": 2512
+    },
+    {
+      "epoch": 0.48049713193116633,
+      "grad_norm": 2.4078571796417236,
+      "learning_rate": 4.974719315678486e-06,
+      "loss": 0.4023,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4806883365200765,
+      "grad_norm": 1.6684014797210693,
+      "learning_rate": 4.974972111857759e-06,
+      "loss": 0.1205,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4808795411089866,
+      "grad_norm": 2.405083417892456,
+      "learning_rate": 4.975224807501662e-06,
+      "loss": 0.3595,
+      "step": 2515
+    },
+    {
+      "epoch": 0.48107074569789676,
+      "grad_norm": 1.4291092157363892,
+      "learning_rate": 4.975477402690129e-06,
+      "loss": 0.0913,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4812619502868069,
+      "grad_norm": 6.145540237426758,
+      "learning_rate": 4.975729897502997e-06,
+      "loss": 0.1871,
+      "step": 2517
+    },
+    {
+      "epoch": 0.481453154875717,
+      "grad_norm": 1.5036752223968506,
+      "learning_rate": 4.975982292020009e-06,
+      "loss": 0.1577,
+      "step": 2518
+    },
+    {
+      "epoch": 0.48164435946462714,
+      "grad_norm": 2.249161720275879,
+      "learning_rate": 4.97623458632081e-06,
+      "loss": 0.1663,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4818355640535373,
+      "grad_norm": 1.631930947303772,
+      "learning_rate": 4.976486780484955e-06,
+      "loss": 0.1467,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4820267686424474,
+      "grad_norm": 1.7321642637252808,
+      "learning_rate": 4.9767388745919e-06,
+      "loss": 0.2776,
+      "step": 2521
+    },
+    {
+      "epoch": 0.48221797323135757,
+      "grad_norm": 3.218299388885498,
+      "learning_rate": 4.976990868721009e-06,
+      "loss": 0.5657,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4824091778202677,
+      "grad_norm": 2.6193222999572754,
+      "learning_rate": 4.977242762951551e-06,
+      "loss": 0.2718,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4826003824091778,
+      "grad_norm": 2.339751720428467,
+      "learning_rate": 4.977494557362701e-06,
+      "loss": 0.1729,
+      "step": 2524
+    },
+    {
+      "epoch": 0.48279158699808794,
+      "grad_norm": 1.910982370376587,
+      "learning_rate": 4.9777462520335386e-06,
+      "loss": 0.0808,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4829827915869981,
+      "grad_norm": 1.3816866874694824,
+      "learning_rate": 4.97799784704305e-06,
+      "loss": 0.0891,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4831739961759082,
+      "grad_norm": 1.6029402017593384,
+      "learning_rate": 4.97824934247013e-06,
+      "loss": 0.1892,
+      "step": 2527
+    },
+    {
+      "epoch": 0.48336520076481837,
+      "grad_norm": 1.5500452518463135,
+      "learning_rate": 4.978500738393576e-06,
+      "loss": 0.143,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4835564053537285,
+      "grad_norm": 2.6044082641601562,
+      "learning_rate": 4.978752034892094e-06,
+      "loss": 0.132,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4837476099426386,
+      "grad_norm": 1.7775001525878906,
+      "learning_rate": 4.979003232044297e-06,
+      "loss": 0.1912,
+      "step": 2530
+    },
+    {
+      "epoch": 0.48393881453154874,
+      "grad_norm": 1.5503368377685547,
+      "learning_rate": 4.979254329928704e-06,
+      "loss": 0.0685,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4841300191204589,
+      "grad_norm": 1.6482033729553223,
+      "learning_rate": 4.979505328623739e-06,
+      "loss": 0.2095,
+      "step": 2532
+    },
+    {
+      "epoch": 0.48432122370936903,
+      "grad_norm": 3.1637353897094727,
+      "learning_rate": 4.9797562282077376e-06,
+      "loss": 0.4826,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4845124282982792,
+      "grad_norm": 1.1990470886230469,
+      "learning_rate": 4.98000702875894e-06,
+      "loss": 0.0619,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4847036328871893,
+      "grad_norm": 1.7547122240066528,
+      "learning_rate": 4.980257730355493e-06,
+      "loss": 0.1307,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4848948374760994,
+      "grad_norm": 1.7956041097640991,
+      "learning_rate": 4.9805083330754525e-06,
+      "loss": 0.1256,
+      "step": 2536
+    },
+    {
+      "epoch": 0.48508604206500955,
+      "grad_norm": 1.7650172710418701,
+      "learning_rate": 4.980758836996782e-06,
+      "loss": 0.1144,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4852772466539197,
+      "grad_norm": 3.838001251220703,
+      "learning_rate": 4.981009242197351e-06,
+      "loss": 0.252,
+      "step": 2538
+    },
+    {
+      "epoch": 0.48546845124282983,
+      "grad_norm": 2.8733246326446533,
+      "learning_rate": 4.981259548754939e-06,
+      "loss": 0.3944,
+      "step": 2539
+    },
+    {
+      "epoch": 0.48565965583174,
+      "grad_norm": 1.7851463556289673,
+      "learning_rate": 4.981509756747234e-06,
+      "loss": 0.1834,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4858508604206501,
+      "grad_norm": 1.990270972251892,
+      "learning_rate": 4.981759866251829e-06,
+      "loss": 0.2053,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4860420650095602,
+      "grad_norm": 1.489159107208252,
+      "learning_rate": 4.98200987734623e-06,
+      "loss": 0.1285,
+      "step": 2542
+    },
+    {
+      "epoch": 0.48623326959847035,
+      "grad_norm": 1.8262524604797363,
+      "learning_rate": 4.9822597901078465e-06,
+      "loss": 0.1314,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4864244741873805,
+      "grad_norm": 1.9886893033981323,
+      "learning_rate": 4.982509604614002e-06,
+      "loss": 0.1196,
+      "step": 2544
+    },
+    {
+      "epoch": 0.48661567877629064,
+      "grad_norm": 1.8472392559051514,
+      "learning_rate": 4.982759320941924e-06,
+      "loss": 0.1214,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4868068833652008,
+      "grad_norm": 1.5206496715545654,
+      "learning_rate": 4.9830089391687534e-06,
+      "loss": 0.1371,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4869980879541109,
+      "grad_norm": 2.1726207733154297,
+      "learning_rate": 4.983258459371536e-06,
+      "loss": 0.1734,
+      "step": 2547
+    },
+    {
+      "epoch": 0.487189292543021,
+      "grad_norm": 1.3305667638778687,
+      "learning_rate": 4.9835078816272295e-06,
+      "loss": 0.0642,
+      "step": 2548
+    },
+    {
+      "epoch": 0.48738049713193116,
+      "grad_norm": 1.633368968963623,
+      "learning_rate": 4.983757206012702e-06,
+      "loss": 0.0843,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4875717017208413,
+      "grad_norm": 2.334226608276367,
+      "learning_rate": 4.984006432604726e-06,
+      "loss": 0.1321,
+      "step": 2550
+    },
+    {
+      "epoch": 0.48776290630975144,
+      "grad_norm": 1.615264654159546,
+      "learning_rate": 4.98425556147999e-06,
+      "loss": 0.1884,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4879541108986616,
+      "grad_norm": 2.8392958641052246,
+      "learning_rate": 4.984504592715089e-06,
+      "loss": 0.5023,
+      "step": 2552
+    },
+    {
+      "epoch": 0.48814531548757173,
+      "grad_norm": 2.3064663410186768,
+      "learning_rate": 4.984753526386528e-06,
+      "loss": 0.2044,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4883365200764818,
+      "grad_norm": 2.2431957721710205,
+      "learning_rate": 4.985002362570723e-06,
+      "loss": 0.2858,
+      "step": 2554
+    },
+    {
+      "epoch": 0.48852772466539196,
+      "grad_norm": 2.3659920692443848,
+      "learning_rate": 4.9852511013439995e-06,
+      "loss": 0.1595,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4887189292543021,
+      "grad_norm": 2.477073907852173,
+      "learning_rate": 4.985499742782594e-06,
+      "loss": 0.0791,
+      "step": 2556
+    },
+    {
+      "epoch": 0.48891013384321225,
+      "grad_norm": 3.7777740955352783,
+      "learning_rate": 4.985748286962654e-06,
+      "loss": 0.6394,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4891013384321224,
+      "grad_norm": 2.1839075088500977,
+      "learning_rate": 4.9859967339602365e-06,
+      "loss": 0.2495,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4892925430210325,
+      "grad_norm": 2.8904454708099365,
+      "learning_rate": 4.986245083851309e-06,
+      "loss": 0.419,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4894837476099426,
+      "grad_norm": 1.4429762363433838,
+      "learning_rate": 4.986493336711752e-06,
+      "loss": 0.2267,
+      "step": 2560
+    },
+    {
+      "epoch": 0.48967495219885276,
+      "grad_norm": 5.334555149078369,
+      "learning_rate": 4.986741492617356e-06,
+      "loss": 0.2136,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4898661567877629,
+      "grad_norm": 1.6938073635101318,
+      "learning_rate": 4.98698955164382e-06,
+      "loss": 0.0614,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49005736137667305,
+      "grad_norm": 4.193806171417236,
+      "learning_rate": 4.9872375138667615e-06,
+      "loss": 0.2824,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4902485659655832,
+      "grad_norm": 2.4448256492614746,
+      "learning_rate": 4.987485379361701e-06,
+      "loss": 0.362,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4904397705544933,
+      "grad_norm": 2.175049066543579,
+      "learning_rate": 4.987733148204077e-06,
+      "loss": 0.159,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4906309751434034,
+      "grad_norm": 1.4821659326553345,
+      "learning_rate": 4.987980820469236e-06,
+      "loss": 0.1167,
+      "step": 2566
+    },
+    {
+      "epoch": 0.49082217973231357,
+      "grad_norm": 1.6503522396087646,
+      "learning_rate": 4.988228396232439e-06,
+      "loss": 0.0725,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4910133843212237,
+      "grad_norm": 3.942607879638672,
+      "learning_rate": 4.988475875568857e-06,
+      "loss": 0.136,
+      "step": 2568
+    },
+    {
+      "epoch": 0.49120458891013385,
+      "grad_norm": 1.6592808961868286,
+      "learning_rate": 4.988723258553574e-06,
+      "loss": 0.1298,
+      "step": 2569
+    },
+    {
+      "epoch": 0.491395793499044,
+      "grad_norm": 1.752287745475769,
+      "learning_rate": 4.988970545261588e-06,
+      "loss": 0.311,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4915869980879541,
+      "grad_norm": 2.297788143157959,
+      "learning_rate": 4.989217735767806e-06,
+      "loss": 0.2016,
+      "step": 2571
+    },
+    {
+      "epoch": 0.49177820267686423,
+      "grad_norm": 1.5098553895950317,
+      "learning_rate": 4.989464830147051e-06,
+      "loss": 0.1153,
+      "step": 2572
+    },
+    {
+      "epoch": 0.49196940726577437,
+      "grad_norm": 1.6489616632461548,
+      "learning_rate": 4.989711828474057e-06,
+      "loss": 0.0871,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4921606118546845,
+      "grad_norm": 1.4597628116607666,
+      "learning_rate": 4.989958730823471e-06,
+      "loss": 0.1124,
+      "step": 2574
+    },
+    {
+      "epoch": 0.49235181644359466,
+      "grad_norm": 1.9017386436462402,
+      "learning_rate": 4.990205537269853e-06,
+      "loss": 0.1342,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4925430210325048,
+      "grad_norm": 2.946324586868286,
+      "learning_rate": 4.990452247887675e-06,
+      "loss": 0.3603,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4927342256214149,
+      "grad_norm": 2.808943510055542,
+      "learning_rate": 4.9906988627513265e-06,
+      "loss": 0.3344,
+      "step": 2577
+    },
+    {
+      "epoch": 0.49292543021032503,
+      "grad_norm": 2.313256025314331,
+      "learning_rate": 4.990945381935106e-06,
+      "loss": 0.336,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4931166347992352,
+      "grad_norm": 1.7406623363494873,
+      "learning_rate": 4.991191805513227e-06,
+      "loss": 0.1274,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4933078393881453,
+      "grad_norm": 2.546191692352295,
+      "learning_rate": 4.991438133559817e-06,
+      "loss": 0.3057,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49349904397705546,
+      "grad_norm": 3.7837564945220947,
+      "learning_rate": 4.991684366148917e-06,
+      "loss": 0.1454,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4936902485659656,
+      "grad_norm": 3.7757515907287598,
+      "learning_rate": 4.991930503354482e-06,
+      "loss": 0.3576,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4938814531548757,
+      "grad_norm": 2.6701161861419678,
+      "learning_rate": 4.9921765452503814e-06,
+      "loss": 0.4054,
+      "step": 2583
+    },
+    {
+      "epoch": 0.49407265774378584,
+      "grad_norm": 3.300074577331543,
+      "learning_rate": 4.992422491910399e-06,
+      "loss": 0.3313,
+      "step": 2584
+    },
+    {
+      "epoch": 0.494263862332696,
+      "grad_norm": 1.8099370002746582,
+      "learning_rate": 4.9926683434082315e-06,
+      "loss": 0.1173,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4944550669216061,
+      "grad_norm": 1.8686485290527344,
+      "learning_rate": 4.9929140998174915e-06,
+      "loss": 0.1983,
+      "step": 2586
+    },
+    {
+      "epoch": 0.49464627151051627,
+      "grad_norm": 1.4938725233078003,
+      "learning_rate": 4.9931597612117065e-06,
+      "loss": 0.0842,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4948374760994264,
+      "grad_norm": 1.6325137615203857,
+      "learning_rate": 4.993405327664316e-06,
+      "loss": 0.093,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4950286806883365,
+      "grad_norm": 1.7486827373504639,
+      "learning_rate": 4.99365079924868e-06,
+      "loss": 0.2226,
+      "step": 2589
+    },
+    {
+      "epoch": 0.49521988527724664,
+      "grad_norm": 1.7181081771850586,
+      "learning_rate": 4.993896176038066e-06,
+      "loss": 0.1983,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4954110898661568,
+      "grad_norm": 1.6922866106033325,
+      "learning_rate": 4.9941414581056636e-06,
+      "loss": 0.3837,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4956022944550669,
+      "grad_norm": 1.5740710496902466,
+      "learning_rate": 4.994386645524574e-06,
+      "loss": 0.1492,
+      "step": 2592
+    },
+    {
+      "epoch": 0.49579349904397707,
+      "grad_norm": 1.319801926612854,
+      "learning_rate": 4.994631738367814e-06,
+      "loss": 0.1124,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4959847036328872,
+      "grad_norm": 1.6975462436676025,
+      "learning_rate": 4.994876736708317e-06,
+      "loss": 0.1612,
+      "step": 2594
+    },
+    {
+      "epoch": 0.4961759082217973,
+      "grad_norm": 2.0203516483306885,
+      "learning_rate": 4.99512164061893e-06,
+      "loss": 0.2532,
+      "step": 2595
+    },
+    {
+      "epoch": 0.49636711281070744,
+      "grad_norm": 1.9539330005645752,
+      "learning_rate": 4.995366450172418e-06,
+      "loss": 0.2315,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4965583173996176,
+      "grad_norm": 1.986435890197754,
+      "learning_rate": 4.995611165441463e-06,
+      "loss": 0.1917,
+      "step": 2597
+    },
+    {
+      "epoch": 0.49674952198852773,
+      "grad_norm": 1.494219422340393,
+      "learning_rate": 4.99585578649866e-06,
+      "loss": 0.096,
+      "step": 2598
+    },
+    {
+      "epoch": 0.4969407265774379,
+      "grad_norm": 1.1928995847702026,
+      "learning_rate": 4.996100313416522e-06,
+      "loss": 0.0925,
+      "step": 2599
+    },
+    {
+      "epoch": 0.497131931166348,
+      "grad_norm": 2.613600492477417,
+      "learning_rate": 4.996344746267477e-06,
+      "loss": 0.1606,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4973231357552581,
+      "grad_norm": 1.8623335361480713,
+      "learning_rate": 4.99658908512387e-06,
+      "loss": 0.1377,
+      "step": 2601
+    },
+    {
+      "epoch": 0.49751434034416825,
+      "grad_norm": 2.6297202110290527,
+      "learning_rate": 4.996833330057964e-06,
+      "loss": 0.3241,
+      "step": 2602
+    },
+    {
+      "epoch": 0.4977055449330784,
+      "grad_norm": 1.7815700769424438,
+      "learning_rate": 4.997077481141936e-06,
+      "loss": 0.2196,
+      "step": 2603
+    },
+    {
+      "epoch": 0.49789674952198854,
+      "grad_norm": 3.413189172744751,
+      "learning_rate": 4.9973215384478835e-06,
+      "loss": 0.1373,
+      "step": 2604
+    },
+    {
+      "epoch": 0.4980879541108987,
+      "grad_norm": 4.574033737182617,
+      "learning_rate": 4.997565502047817e-06,
+      "loss": 0.2807,
+      "step": 2605
+    },
+    {
+      "epoch": 0.4982791586998088,
+      "grad_norm": 3.2377426624298096,
+      "learning_rate": 4.997809372013666e-06,
+      "loss": 0.3188,
+      "step": 2606
+    },
+    {
+      "epoch": 0.4984703632887189,
+      "grad_norm": 1.8162243366241455,
+      "learning_rate": 4.998053148417279e-06,
+      "loss": 0.2634,
+      "step": 2607
+    },
+    {
+      "epoch": 0.49866156787762905,
+      "grad_norm": 3.267460823059082,
+      "learning_rate": 4.998296831330417e-06,
+      "loss": 0.2784,
+      "step": 2608
+    },
+    {
+      "epoch": 0.4988527724665392,
+      "grad_norm": 1.5525761842727661,
+      "learning_rate": 4.998540420824764e-06,
+      "loss": 0.1681,
+      "step": 2609
+    },
+    {
+      "epoch": 0.49904397705544934,
+      "grad_norm": 2.893181324005127,
+      "learning_rate": 4.998783916971917e-06,
+      "loss": 0.3174,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4992351816443595,
+      "grad_norm": 1.829892873764038,
+      "learning_rate": 4.999027319843394e-06,
+      "loss": 0.1275,
+      "step": 2611
+    },
+    {
+      "epoch": 0.4994263862332696,
+      "grad_norm": 2.012143611907959,
+      "learning_rate": 4.999270629510629e-06,
+      "loss": 0.1867,
+      "step": 2612
+    },
+    {
+      "epoch": 0.4996175908221797,
+      "grad_norm": 2.164092779159546,
+      "learning_rate": 4.9995138460449755e-06,
+      "loss": 0.2112,
+      "step": 2613
+    },
+    {
+      "epoch": 0.49980879541108986,
+      "grad_norm": 1.7680532932281494,
+      "learning_rate": 4.999756969517703e-06,
+      "loss": 0.1469,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.590065002441406,
+      "learning_rate": 5e-06,
+      "loss": 0.2257,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5001912045889101,
+      "grad_norm": 1.274849772453308,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5003824091778203,
+      "grad_norm": 2.5290040969848633,
+      "learning_rate": 5e-06,
+      "loss": 0.0899,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5005736137667304,
+      "grad_norm": 2.3508877754211426,
+      "learning_rate": 5e-06,
+      "loss": 0.165,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5007648183556406,
+      "grad_norm": 2.5142104625701904,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5009560229445507,
+      "grad_norm": 2.028837203979492,
+      "learning_rate": 5e-06,
+      "loss": 0.265,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5011472275334607,
+      "grad_norm": 1.2833895683288574,
+      "learning_rate": 5e-06,
+      "loss": 0.1018,
+      "step": 2621
+    },
+    {
+      "epoch": 0.501338432122371,
+      "grad_norm": 1.112088918685913,
+      "learning_rate": 5e-06,
+      "loss": 0.09,
+      "step": 2622
+    },
+    {
+      "epoch": 0.501529636711281,
+      "grad_norm": 1.1832149028778076,
+      "learning_rate": 5e-06,
+      "loss": 0.096,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5017208413001912,
+      "grad_norm": 1.8202910423278809,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5019120458891013,
+      "grad_norm": 2.20587420463562,
+      "learning_rate": 5e-06,
+      "loss": 0.0921,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5021032504780115,
+      "grad_norm": 2.4090912342071533,
+      "learning_rate": 5e-06,
+      "loss": 0.401,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5022944550669216,
+      "grad_norm": 2.0977654457092285,
+      "learning_rate": 5e-06,
+      "loss": 0.3296,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5024856596558317,
+      "grad_norm": 1.8550901412963867,
+      "learning_rate": 5e-06,
+      "loss": 0.1836,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5026768642447419,
+      "grad_norm": 0.8883583545684814,
+      "learning_rate": 5e-06,
+      "loss": 0.0655,
+      "step": 2629
+    },
+    {
+      "epoch": 0.502868068833652,
+      "grad_norm": 1.4066579341888428,
+      "learning_rate": 5e-06,
+      "loss": 0.0553,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5030592734225622,
+      "grad_norm": 1.9968935251235962,
+      "learning_rate": 5e-06,
+      "loss": 0.0832,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5032504780114723,
+      "grad_norm": 1.5393012762069702,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5034416826003824,
+      "grad_norm": 2.020643949508667,
+      "learning_rate": 5e-06,
+      "loss": 0.2575,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5036328871892926,
+      "grad_norm": 1.2440476417541504,
+      "learning_rate": 5e-06,
+      "loss": 0.0954,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5038240917782026,
+      "grad_norm": 1.243212103843689,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5040152963671128,
+      "grad_norm": 2.4254207611083984,
+      "learning_rate": 5e-06,
+      "loss": 0.2976,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5042065009560229,
+      "grad_norm": 1.2954270839691162,
+      "learning_rate": 5e-06,
+      "loss": 0.0682,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5043977055449331,
+      "grad_norm": 1.661173701286316,
+      "learning_rate": 5e-06,
+      "loss": 0.2638,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5045889101338432,
+      "grad_norm": 2.2583632469177246,
+      "learning_rate": 5e-06,
+      "loss": 0.2787,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5047801147227533,
+      "grad_norm": 3.421473979949951,
+      "learning_rate": 5e-06,
+      "loss": 0.4827,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5049713193116635,
+      "grad_norm": 1.8076733350753784,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5051625239005736,
+      "grad_norm": 2.341878652572632,
+      "learning_rate": 5e-06,
+      "loss": 0.2703,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5053537284894838,
+      "grad_norm": 1.1928160190582275,
+      "learning_rate": 5e-06,
+      "loss": 0.1011,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5055449330783939,
+      "grad_norm": 1.0467259883880615,
+      "learning_rate": 5e-06,
+      "loss": 0.0831,
+      "step": 2644
+    },
+    {
+      "epoch": 0.505736137667304,
+      "grad_norm": 1.8891057968139648,
+      "learning_rate": 5e-06,
+      "loss": 0.3587,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5059273422562142,
+      "grad_norm": 2.7861201763153076,
+      "learning_rate": 5e-06,
+      "loss": 0.4989,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5061185468451243,
+      "grad_norm": 1.5840612649917603,
+      "learning_rate": 5e-06,
+      "loss": 0.1864,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5063097514340344,
+      "grad_norm": 3.3611576557159424,
+      "learning_rate": 5e-06,
+      "loss": 0.3172,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5065009560229445,
+      "grad_norm": 1.371387004852295,
+      "learning_rate": 5e-06,
+      "loss": 0.0707,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5066921606118547,
+      "grad_norm": 1.2657071352005005,
+      "learning_rate": 5e-06,
+      "loss": 0.0526,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5068833652007648,
+      "grad_norm": 3.1202003955841064,
+      "learning_rate": 5e-06,
+      "loss": 0.3824,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5070745697896749,
+      "grad_norm": 1.1883548498153687,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5072657743785851,
+      "grad_norm": 2.3430683612823486,
+      "learning_rate": 5e-06,
+      "loss": 0.3705,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5074569789674952,
+      "grad_norm": 1.8830441236495972,
+      "learning_rate": 5e-06,
+      "loss": 0.0895,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5076481835564054,
+      "grad_norm": 2.3886196613311768,
+      "learning_rate": 5e-06,
+      "loss": 0.257,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5078393881453155,
+      "grad_norm": 1.6566410064697266,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5080305927342256,
+      "grad_norm": 1.7959635257720947,
+      "learning_rate": 5e-06,
+      "loss": 0.1833,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5082217973231358,
+      "grad_norm": 1.3540290594100952,
+      "learning_rate": 5e-06,
+      "loss": 0.1589,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5084130019120459,
+      "grad_norm": 1.84811532497406,
+      "learning_rate": 5e-06,
+      "loss": 0.1614,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5086042065009561,
+      "grad_norm": 1.4192479848861694,
+      "learning_rate": 5e-06,
+      "loss": 0.1005,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5087954110898661,
+      "grad_norm": 2.123854160308838,
+      "learning_rate": 5e-06,
+      "loss": 0.1387,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5089866156787763,
+      "grad_norm": 3.7756459712982178,
+      "learning_rate": 5e-06,
+      "loss": 0.215,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5091778202676864,
+      "grad_norm": 1.6984196901321411,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5093690248565965,
+      "grad_norm": 2.247396469116211,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5095602294455067,
+      "grad_norm": 2.6015279293060303,
+      "learning_rate": 5e-06,
+      "loss": 0.3206,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5097514340344168,
+      "grad_norm": 1.4520198106765747,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 2666
+    },
+    {
+      "epoch": 0.509942638623327,
+      "grad_norm": 2.357475757598877,
+      "learning_rate": 5e-06,
+      "loss": 0.1052,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5101338432122371,
+      "grad_norm": 1.6518830060958862,
+      "learning_rate": 5e-06,
+      "loss": 0.0741,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5103250478011472,
+      "grad_norm": 3.9007205963134766,
+      "learning_rate": 5e-06,
+      "loss": 0.3746,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5105162523900574,
+      "grad_norm": 2.126702308654785,
+      "learning_rate": 5e-06,
+      "loss": 0.332,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5107074569789675,
+      "grad_norm": 1.7568902969360352,
+      "learning_rate": 5e-06,
+      "loss": 0.1567,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5108986615678777,
+      "grad_norm": 1.939828634262085,
+      "learning_rate": 5e-06,
+      "loss": 0.2383,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5110898661567878,
+      "grad_norm": 1.8241344690322876,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 2673
+    },
+    {
+      "epoch": 0.511281070745698,
+      "grad_norm": 2.5553126335144043,
+      "learning_rate": 5e-06,
+      "loss": 0.2161,
+      "step": 2674
+    },
+    {
+      "epoch": 0.511472275334608,
+      "grad_norm": 1.7424609661102295,
+      "learning_rate": 5e-06,
+      "loss": 0.091,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5116634799235181,
+      "grad_norm": 1.8869847059249878,
+      "learning_rate": 5e-06,
+      "loss": 0.2992,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5118546845124283,
+      "grad_norm": 1.1718688011169434,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5120458891013384,
+      "grad_norm": 1.0170719623565674,
+      "learning_rate": 5e-06,
+      "loss": 0.1267,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5122370936902486,
+      "grad_norm": 1.5140657424926758,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5124282982791587,
+      "grad_norm": 1.8344894647598267,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5126195028680688,
+      "grad_norm": 2.2053725719451904,
+      "learning_rate": 5e-06,
+      "loss": 0.2002,
+      "step": 2681
+    },
+    {
+      "epoch": 0.512810707456979,
+      "grad_norm": 1.968572974205017,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5130019120458891,
+      "grad_norm": 2.810896158218384,
+      "learning_rate": 5e-06,
+      "loss": 0.3554,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5131931166347993,
+      "grad_norm": 1.951485276222229,
+      "learning_rate": 5e-06,
+      "loss": 0.2156,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5133843212237094,
+      "grad_norm": 1.31484055519104,
+      "learning_rate": 5e-06,
+      "loss": 0.1424,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5135755258126194,
+      "grad_norm": 2.028074026107788,
+      "learning_rate": 5e-06,
+      "loss": 0.165,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5137667304015296,
+      "grad_norm": 1.3632593154907227,
+      "learning_rate": 5e-06,
+      "loss": 0.067,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5139579349904397,
+      "grad_norm": 1.8360464572906494,
+      "learning_rate": 5e-06,
+      "loss": 0.1073,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5141491395793499,
+      "grad_norm": 2.7221839427948,
+      "learning_rate": 5e-06,
+      "loss": 0.536,
+      "step": 2689
+    },
+    {
+      "epoch": 0.51434034416826,
+      "grad_norm": 1.6568959951400757,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5145315487571702,
+      "grad_norm": 2.4238178730010986,
+      "learning_rate": 5e-06,
+      "loss": 0.2072,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5147227533460803,
+      "grad_norm": 1.4915567636489868,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5149139579349904,
+      "grad_norm": 1.5793148279190063,
+      "learning_rate": 5e-06,
+      "loss": 0.1,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5151051625239006,
+      "grad_norm": 1.486043930053711,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5152963671128107,
+      "grad_norm": 2.671621561050415,
+      "learning_rate": 5e-06,
+      "loss": 0.6114,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5154875717017209,
+      "grad_norm": 1.9827364683151245,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 2696
+    },
+    {
+      "epoch": 0.515678776290631,
+      "grad_norm": 1.608793020248413,
+      "learning_rate": 5e-06,
+      "loss": 0.1065,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5158699808795411,
+      "grad_norm": 1.5601345300674438,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5160611854684513,
+      "grad_norm": 2.0068085193634033,
+      "learning_rate": 5e-06,
+      "loss": 0.1,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5162523900573613,
+      "grad_norm": 1.006656289100647,
+      "learning_rate": 5e-06,
+      "loss": 0.0425,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5164435946462715,
+      "grad_norm": 4.04417085647583,
+      "learning_rate": 5e-06,
+      "loss": 0.7525,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5166347992351816,
+      "grad_norm": 2.4919166564941406,
+      "learning_rate": 5e-06,
+      "loss": 0.3474,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5168260038240918,
+      "grad_norm": 1.7118905782699585,
+      "learning_rate": 5e-06,
+      "loss": 0.1593,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5170172084130019,
+      "grad_norm": 2.105720043182373,
+      "learning_rate": 5e-06,
+      "loss": 0.2862,
+      "step": 2704
+    },
+    {
+      "epoch": 0.517208413001912,
+      "grad_norm": 2.6111350059509277,
+      "learning_rate": 5e-06,
+      "loss": 0.1986,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5173996175908222,
+      "grad_norm": 1.308851718902588,
+      "learning_rate": 5e-06,
+      "loss": 0.0404,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5175908221797323,
+      "grad_norm": 2.451007127761841,
+      "learning_rate": 5e-06,
+      "loss": 0.241,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5177820267686425,
+      "grad_norm": 1.6604572534561157,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5179732313575526,
+      "grad_norm": 1.995052456855774,
+      "learning_rate": 5e-06,
+      "loss": 0.2343,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5181644359464627,
+      "grad_norm": 1.6679072380065918,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5183556405353729,
+      "grad_norm": 1.284263253211975,
+      "learning_rate": 5e-06,
+      "loss": 0.0664,
+      "step": 2711
+    },
+    {
+      "epoch": 0.518546845124283,
+      "grad_norm": 1.0988402366638184,
+      "learning_rate": 5e-06,
+      "loss": 0.0266,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5187380497131932,
+      "grad_norm": 1.9333552122116089,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5189292543021032,
+      "grad_norm": 2.2870166301727295,
+      "learning_rate": 5e-06,
+      "loss": 0.4859,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5191204588910134,
+      "grad_norm": 2.012103319168091,
+      "learning_rate": 5e-06,
+      "loss": 0.1198,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5193116634799235,
+      "grad_norm": 1.6784409284591675,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5195028680688336,
+      "grad_norm": 2.0358691215515137,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5196940726577438,
+      "grad_norm": 1.3582711219787598,
+      "learning_rate": 5e-06,
+      "loss": 0.0842,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5198852772466539,
+      "grad_norm": 2.2164931297302246,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5200764818355641,
+      "grad_norm": 3.631624698638916,
+      "learning_rate": 5e-06,
+      "loss": 0.2526,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5202676864244742,
+      "grad_norm": 2.41479229927063,
+      "learning_rate": 5e-06,
+      "loss": 0.3641,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5204588910133843,
+      "grad_norm": 1.9997187852859497,
+      "learning_rate": 5e-06,
+      "loss": 0.1708,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5206500956022945,
+      "grad_norm": 3.5295379161834717,
+      "learning_rate": 5e-06,
+      "loss": 0.2561,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5208413001912046,
+      "grad_norm": 4.27695369720459,
+      "learning_rate": 5e-06,
+      "loss": 0.2081,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5210325047801148,
+      "grad_norm": 1.7342065572738647,
+      "learning_rate": 5e-06,
+      "loss": 0.0791,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5212237093690248,
+      "grad_norm": 1.875627040863037,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 2726
+    },
+    {
+      "epoch": 0.521414913957935,
+      "grad_norm": 1.8105803728103638,
+      "learning_rate": 5e-06,
+      "loss": 0.2063,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5216061185468451,
+      "grad_norm": 1.5556526184082031,
+      "learning_rate": 5e-06,
+      "loss": 0.205,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5217973231357552,
+      "grad_norm": 1.5404901504516602,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5219885277246654,
+      "grad_norm": 1.3888638019561768,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5221797323135755,
+      "grad_norm": 1.3965755701065063,
+      "learning_rate": 5e-06,
+      "loss": 0.0549,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5223709369024857,
+      "grad_norm": 2.155271530151367,
+      "learning_rate": 5e-06,
+      "loss": 0.2185,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5225621414913958,
+      "grad_norm": 3.022123336791992,
+      "learning_rate": 5e-06,
+      "loss": 0.3292,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5227533460803059,
+      "grad_norm": 1.5752053260803223,
+      "learning_rate": 5e-06,
+      "loss": 0.1342,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5229445506692161,
+      "grad_norm": 1.3086416721343994,
+      "learning_rate": 5e-06,
+      "loss": 0.0672,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5231357552581262,
+      "grad_norm": 1.8340917825698853,
+      "learning_rate": 5e-06,
+      "loss": 0.0859,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5233269598470364,
+      "grad_norm": 4.185413360595703,
+      "learning_rate": 5e-06,
+      "loss": 0.1021,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5235181644359465,
+      "grad_norm": 3.0778019428253174,
+      "learning_rate": 5e-06,
+      "loss": 0.3103,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5237093690248565,
+      "grad_norm": 2.644026041030884,
+      "learning_rate": 5e-06,
+      "loss": 0.5049,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5239005736137667,
+      "grad_norm": 1.7256097793579102,
+      "learning_rate": 5e-06,
+      "loss": 0.2243,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5240917782026768,
+      "grad_norm": 2.3407437801361084,
+      "learning_rate": 5e-06,
+      "loss": 0.1834,
+      "step": 2741
+    },
+    {
+      "epoch": 0.524282982791587,
+      "grad_norm": 1.885779619216919,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5244741873804971,
+      "grad_norm": 2.010530710220337,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5246653919694073,
+      "grad_norm": 2.1776020526885986,
+      "learning_rate": 5e-06,
+      "loss": 0.131,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5248565965583174,
+      "grad_norm": 2.62050199508667,
+      "learning_rate": 5e-06,
+      "loss": 0.3972,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5250478011472275,
+      "grad_norm": 2.356659173965454,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5252390057361377,
+      "grad_norm": 1.9196159839630127,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5254302103250478,
+      "grad_norm": 2.2066104412078857,
+      "learning_rate": 5e-06,
+      "loss": 0.1625,
+      "step": 2748
+    },
+    {
+      "epoch": 0.525621414913958,
+      "grad_norm": 1.1974610090255737,
+      "learning_rate": 5e-06,
+      "loss": 0.0551,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5258126195028681,
+      "grad_norm": 2.4025866985321045,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5260038240917781,
+      "grad_norm": 2.276185989379883,
+      "learning_rate": 5e-06,
+      "loss": 0.2141,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5261950286806883,
+      "grad_norm": 1.9449247121810913,
+      "learning_rate": 5e-06,
+      "loss": 0.2303,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5263862332695984,
+      "grad_norm": 1.2716035842895508,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5265774378585086,
+      "grad_norm": 1.2771272659301758,
+      "learning_rate": 5e-06,
+      "loss": 0.088,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5267686424474187,
+      "grad_norm": 2.0387823581695557,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5269598470363289,
+      "grad_norm": 2.576063394546509,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 2756
+    },
+    {
+      "epoch": 0.527151051625239,
+      "grad_norm": 2.7884414196014404,
+      "learning_rate": 5e-06,
+      "loss": 0.3298,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5273422562141491,
+      "grad_norm": 3.2937912940979004,
+      "learning_rate": 5e-06,
+      "loss": 0.5074,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5275334608030593,
+      "grad_norm": 1.2143149375915527,
+      "learning_rate": 5e-06,
+      "loss": 0.0907,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5277246653919694,
+      "grad_norm": 1.9609935283660889,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5279158699808796,
+      "grad_norm": 2.3065104484558105,
+      "learning_rate": 5e-06,
+      "loss": 0.2394,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5281070745697897,
+      "grad_norm": 1.3439974784851074,
+      "learning_rate": 5e-06,
+      "loss": 0.0567,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5282982791586998,
+      "grad_norm": 2.701871395111084,
+      "learning_rate": 5e-06,
+      "loss": 0.2844,
+      "step": 2763
+    },
+    {
+      "epoch": 0.52848948374761,
+      "grad_norm": 2.580606460571289,
+      "learning_rate": 5e-06,
+      "loss": 0.3827,
+      "step": 2764
+    },
+    {
+      "epoch": 0.52868068833652,
+      "grad_norm": 2.0490102767944336,
+      "learning_rate": 5e-06,
+      "loss": 0.164,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5288718929254302,
+      "grad_norm": 2.0676021575927734,
+      "learning_rate": 5e-06,
+      "loss": 0.3342,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5290630975143403,
+      "grad_norm": 1.146315097808838,
+      "learning_rate": 5e-06,
+      "loss": 0.0772,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5292543021032505,
+      "grad_norm": 1.5692076683044434,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5294455066921606,
+      "grad_norm": 2.6447064876556396,
+      "learning_rate": 5e-06,
+      "loss": 0.2929,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5296367112810707,
+      "grad_norm": 3.0837621688842773,
+      "learning_rate": 5e-06,
+      "loss": 0.4175,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5298279158699809,
+      "grad_norm": 2.180725574493408,
+      "learning_rate": 5e-06,
+      "loss": 0.2276,
+      "step": 2771
+    },
+    {
+      "epoch": 0.530019120458891,
+      "grad_norm": 1.3501646518707275,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5302103250478012,
+      "grad_norm": 2.3673691749572754,
+      "learning_rate": 5e-06,
+      "loss": 0.2671,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5304015296367113,
+      "grad_norm": 2.891281843185425,
+      "learning_rate": 5e-06,
+      "loss": 0.2122,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5305927342256214,
+      "grad_norm": 2.67206072807312,
+      "learning_rate": 5e-06,
+      "loss": 0.112,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5307839388145316,
+      "grad_norm": 2.2440712451934814,
+      "learning_rate": 5e-06,
+      "loss": 0.3392,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5309751434034417,
+      "grad_norm": 1.790481686592102,
+      "learning_rate": 5e-06,
+      "loss": 0.2133,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5311663479923519,
+      "grad_norm": 1.2576000690460205,
+      "learning_rate": 5e-06,
+      "loss": 0.0937,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5313575525812619,
+      "grad_norm": 1.711289405822754,
+      "learning_rate": 5e-06,
+      "loss": 0.1502,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5315487571701721,
+      "grad_norm": 1.4342141151428223,
+      "learning_rate": 5e-06,
+      "loss": 0.1056,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5317399617590822,
+      "grad_norm": 2.1560006141662598,
+      "learning_rate": 5e-06,
+      "loss": 0.1531,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5319311663479923,
+      "grad_norm": 2.112128734588623,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5321223709369025,
+      "grad_norm": 2.3116440773010254,
+      "learning_rate": 5e-06,
+      "loss": 0.2867,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5323135755258126,
+      "grad_norm": 2.4027740955352783,
+      "learning_rate": 5e-06,
+      "loss": 0.2486,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5325047801147228,
+      "grad_norm": 1.1170294284820557,
+      "learning_rate": 5e-06,
+      "loss": 0.0558,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5326959847036329,
+      "grad_norm": 2.2636983394622803,
+      "learning_rate": 5e-06,
+      "loss": 0.0858,
+      "step": 2786
+    },
+    {
+      "epoch": 0.532887189292543,
+      "grad_norm": 1.4869250059127808,
+      "learning_rate": 5e-06,
+      "loss": 0.0778,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5330783938814532,
+      "grad_norm": 2.6340017318725586,
+      "learning_rate": 5e-06,
+      "loss": 0.1319,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5332695984703633,
+      "grad_norm": 2.0733349323272705,
+      "learning_rate": 5e-06,
+      "loss": 0.2641,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5334608030592735,
+      "grad_norm": 1.8458443880081177,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5336520076481835,
+      "grad_norm": 0.9659456610679626,
+      "learning_rate": 5e-06,
+      "loss": 0.0781,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5338432122370937,
+      "grad_norm": 2.7698538303375244,
+      "learning_rate": 5e-06,
+      "loss": 0.2659,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5340344168260038,
+      "grad_norm": 0.9814829230308533,
+      "learning_rate": 5e-06,
+      "loss": 0.0433,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5342256214149139,
+      "grad_norm": 2.5287699699401855,
+      "learning_rate": 5e-06,
+      "loss": 0.1128,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5344168260038241,
+      "grad_norm": 2.2758593559265137,
+      "learning_rate": 5e-06,
+      "loss": 0.2237,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5346080305927342,
+      "grad_norm": 3.0711472034454346,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5347992351816444,
+      "grad_norm": 2.538210153579712,
+      "learning_rate": 5e-06,
+      "loss": 0.3324,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5349904397705545,
+      "grad_norm": 1.7886542081832886,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5351816443594646,
+      "grad_norm": 1.5378341674804688,
+      "learning_rate": 5e-06,
+      "loss": 0.1657,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5353728489483748,
+      "grad_norm": 1.748417615890503,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5355640535372849,
+      "grad_norm": 1.9090436697006226,
+      "learning_rate": 5e-06,
+      "loss": 0.3028,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5357552581261951,
+      "grad_norm": 2.4026999473571777,
+      "learning_rate": 5e-06,
+      "loss": 0.2898,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5359464627151052,
+      "grad_norm": 1.4179044961929321,
+      "learning_rate": 5e-06,
+      "loss": 0.1178,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5361376673040152,
+      "grad_norm": 1.6482465267181396,
+      "learning_rate": 5e-06,
+      "loss": 0.0878,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5363288718929254,
+      "grad_norm": 2.1801981925964355,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5365200764818355,
+      "grad_norm": 1.33127760887146,
+      "learning_rate": 5e-06,
+      "loss": 0.0881,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5367112810707457,
+      "grad_norm": 3.296818971633911,
+      "learning_rate": 5e-06,
+      "loss": 0.3499,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5369024856596558,
+      "grad_norm": 2.183624267578125,
+      "learning_rate": 5e-06,
+      "loss": 0.303,
+      "step": 2808
+    },
+    {
+      "epoch": 0.537093690248566,
+      "grad_norm": 2.4392447471618652,
+      "learning_rate": 5e-06,
+      "loss": 0.3532,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5372848948374761,
+      "grad_norm": 1.9336990118026733,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5374760994263862,
+      "grad_norm": 1.5660754442214966,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5376673040152964,
+      "grad_norm": 2.9620461463928223,
+      "learning_rate": 5e-06,
+      "loss": 0.2485,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5378585086042065,
+      "grad_norm": 1.6001290082931519,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5380497131931167,
+      "grad_norm": 1.5141234397888184,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5382409177820268,
+      "grad_norm": 1.9983782768249512,
+      "learning_rate": 5e-06,
+      "loss": 0.24,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5384321223709368,
+      "grad_norm": 1.7045232057571411,
+      "learning_rate": 5e-06,
+      "loss": 0.1267,
+      "step": 2816
+    },
+    {
+      "epoch": 0.538623326959847,
+      "grad_norm": 1.1079213619232178,
+      "learning_rate": 5e-06,
+      "loss": 0.1043,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5388145315487571,
+      "grad_norm": 2.0188820362091064,
+      "learning_rate": 5e-06,
+      "loss": 0.314,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5390057361376673,
+      "grad_norm": 1.7051680088043213,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5391969407265774,
+      "grad_norm": 2.6241707801818848,
+      "learning_rate": 5e-06,
+      "loss": 0.3201,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5393881453154876,
+      "grad_norm": 1.4381252527236938,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5395793499043977,
+      "grad_norm": 2.13104248046875,
+      "learning_rate": 5e-06,
+      "loss": 0.3106,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5397705544933078,
+      "grad_norm": 2.429593086242676,
+      "learning_rate": 5e-06,
+      "loss": 0.1502,
+      "step": 2823
+    },
+    {
+      "epoch": 0.539961759082218,
+      "grad_norm": 1.9471462965011597,
+      "learning_rate": 5e-06,
+      "loss": 0.1125,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5401529636711281,
+      "grad_norm": 2.1036086082458496,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5403441682600383,
+      "grad_norm": 3.6985058784484863,
+      "learning_rate": 5e-06,
+      "loss": 0.6219,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5405353728489484,
+      "grad_norm": 2.2251152992248535,
+      "learning_rate": 5e-06,
+      "loss": 0.2841,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5407265774378585,
+      "grad_norm": 1.7809268236160278,
+      "learning_rate": 5e-06,
+      "loss": 0.3284,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5409177820267687,
+      "grad_norm": 2.0658373832702637,
+      "learning_rate": 5e-06,
+      "loss": 0.1905,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5411089866156787,
+      "grad_norm": 2.7861831188201904,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5413001912045889,
+      "grad_norm": 2.1867973804473877,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 2831
+    },
+    {
+      "epoch": 0.541491395793499,
+      "grad_norm": 2.6366775035858154,
+      "learning_rate": 5e-06,
+      "loss": 0.4594,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5416826003824092,
+      "grad_norm": 2.0173962116241455,
+      "learning_rate": 5e-06,
+      "loss": 0.1957,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5418738049713193,
+      "grad_norm": 2.2647476196289062,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5420650095602294,
+      "grad_norm": 1.8371925354003906,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5422562141491396,
+      "grad_norm": 2.1972312927246094,
+      "learning_rate": 5e-06,
+      "loss": 0.22,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5424474187380497,
+      "grad_norm": 1.5029915571212769,
+      "learning_rate": 5e-06,
+      "loss": 0.0747,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5426386233269599,
+      "grad_norm": 1.6815667152404785,
+      "learning_rate": 5e-06,
+      "loss": 0.1455,
+      "step": 2838
+    },
+    {
+      "epoch": 0.54282982791587,
+      "grad_norm": 2.7903823852539062,
+      "learning_rate": 5e-06,
+      "loss": 0.5205,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5430210325047801,
+      "grad_norm": 2.121938943862915,
+      "learning_rate": 5e-06,
+      "loss": 0.374,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5432122370936903,
+      "grad_norm": 2.8478827476501465,
+      "learning_rate": 5e-06,
+      "loss": 0.3739,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5434034416826004,
+      "grad_norm": 2.33632755279541,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5435946462715106,
+      "grad_norm": 1.6369229555130005,
+      "learning_rate": 5e-06,
+      "loss": 0.0817,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5437858508604206,
+      "grad_norm": 2.857534885406494,
+      "learning_rate": 5e-06,
+      "loss": 0.4311,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5439770554493308,
+      "grad_norm": 2.2772583961486816,
+      "learning_rate": 5e-06,
+      "loss": 0.3075,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5441682600382409,
+      "grad_norm": 1.84805166721344,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 2846
+    },
+    {
+      "epoch": 0.544359464627151,
+      "grad_norm": 3.0901503562927246,
+      "learning_rate": 5e-06,
+      "loss": 0.2931,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5445506692160612,
+      "grad_norm": 1.3856008052825928,
+      "learning_rate": 5e-06,
+      "loss": 0.0772,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5447418738049713,
+      "grad_norm": 1.3322361707687378,
+      "learning_rate": 5e-06,
+      "loss": 0.0977,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5449330783938815,
+      "grad_norm": 2.8700478076934814,
+      "learning_rate": 5e-06,
+      "loss": 0.1916,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5451242829827916,
+      "grad_norm": 2.482931137084961,
+      "learning_rate": 5e-06,
+      "loss": 0.3922,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5453154875717017,
+      "grad_norm": 2.8786845207214355,
+      "learning_rate": 5e-06,
+      "loss": 0.3643,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5455066921606119,
+      "grad_norm": 1.5687427520751953,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 2853
+    },
+    {
+      "epoch": 0.545697896749522,
+      "grad_norm": 1.8267560005187988,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5458891013384322,
+      "grad_norm": 2.0420382022857666,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5460803059273422,
+      "grad_norm": 2.005113124847412,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5462715105162524,
+      "grad_norm": 2.4532604217529297,
+      "learning_rate": 5e-06,
+      "loss": 0.4297,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5464627151051625,
+      "grad_norm": 1.8568321466445923,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5466539196940726,
+      "grad_norm": 2.2571911811828613,
+      "learning_rate": 5e-06,
+      "loss": 0.2347,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5468451242829828,
+      "grad_norm": 1.288896918296814,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5470363288718929,
+      "grad_norm": 1.4428579807281494,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5472275334608031,
+      "grad_norm": 1.6780493259429932,
+      "learning_rate": 5e-06,
+      "loss": 0.1161,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5474187380497132,
+      "grad_norm": 2.8386378288269043,
+      "learning_rate": 5e-06,
+      "loss": 0.3605,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5476099426386233,
+      "grad_norm": 2.42722487449646,
+      "learning_rate": 5e-06,
+      "loss": 0.2646,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5478011472275335,
+      "grad_norm": 1.9471642971038818,
+      "learning_rate": 5e-06,
+      "loss": 0.4178,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5479923518164436,
+      "grad_norm": 1.1738109588623047,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5481835564053538,
+      "grad_norm": 1.1273823976516724,
+      "learning_rate": 5e-06,
+      "loss": 0.052,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5483747609942639,
+      "grad_norm": 0.9076278805732727,
+      "learning_rate": 5e-06,
+      "loss": 0.0353,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5485659655831739,
+      "grad_norm": 1.6331543922424316,
+      "learning_rate": 5e-06,
+      "loss": 0.1101,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5487571701720841,
+      "grad_norm": 4.053876876831055,
+      "learning_rate": 5e-06,
+      "loss": 0.6708,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5489483747609942,
+      "grad_norm": 2.119873523712158,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5491395793499044,
+      "grad_norm": 1.7409799098968506,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5493307839388145,
+      "grad_norm": 0.972379207611084,
+      "learning_rate": 5e-06,
+      "loss": 0.0649,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5495219885277247,
+      "grad_norm": 3.567614793777466,
+      "learning_rate": 5e-06,
+      "loss": 0.089,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5497131931166348,
+      "grad_norm": 1.5383042097091675,
+      "learning_rate": 5e-06,
+      "loss": 0.0641,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5499043977055449,
+      "grad_norm": 2.714123010635376,
+      "learning_rate": 5e-06,
+      "loss": 0.3569,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5500956022944551,
+      "grad_norm": 2.086832046508789,
+      "learning_rate": 5e-06,
+      "loss": 0.296,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5502868068833652,
+      "grad_norm": 2.510014295578003,
+      "learning_rate": 5e-06,
+      "loss": 0.2874,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5504780114722754,
+      "grad_norm": 1.494666576385498,
+      "learning_rate": 5e-06,
+      "loss": 0.1435,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5506692160611855,
+      "grad_norm": 2.3554635047912598,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5508604206500956,
+      "grad_norm": 1.758353352546692,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5510516252390057,
+      "grad_norm": 1.8508045673370361,
+      "learning_rate": 5e-06,
+      "loss": 0.2355,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5512428298279158,
+      "grad_norm": 1.8111543655395508,
+      "learning_rate": 5e-06,
+      "loss": 0.2789,
+      "step": 2883
+    },
+    {
+      "epoch": 0.551434034416826,
+      "grad_norm": 2.13771915435791,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5516252390057361,
+      "grad_norm": 1.0804390907287598,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5518164435946463,
+      "grad_norm": 1.9480468034744263,
+      "learning_rate": 5e-06,
+      "loss": 0.1915,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5520076481835564,
+      "grad_norm": 1.979048252105713,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5521988527724665,
+      "grad_norm": 2.188206434249878,
+      "learning_rate": 5e-06,
+      "loss": 0.2204,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5523900573613767,
+      "grad_norm": 4.077398777008057,
+      "learning_rate": 5e-06,
+      "loss": 0.4834,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5525812619502868,
+      "grad_norm": 1.1730834245681763,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 2890
+    },
+    {
+      "epoch": 0.552772466539197,
+      "grad_norm": 2.131211280822754,
+      "learning_rate": 5e-06,
+      "loss": 0.2714,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5529636711281071,
+      "grad_norm": 0.8780372738838196,
+      "learning_rate": 5e-06,
+      "loss": 0.0451,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5531548757170172,
+      "grad_norm": 1.3180248737335205,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5533460803059274,
+      "grad_norm": 2.865992784500122,
+      "learning_rate": 5e-06,
+      "loss": 0.0841,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5535372848948374,
+      "grad_norm": 2.895068407058716,
+      "learning_rate": 5e-06,
+      "loss": 0.6579,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5537284894837476,
+      "grad_norm": 1.6795144081115723,
+      "learning_rate": 5e-06,
+      "loss": 0.3405,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5539196940726577,
+      "grad_norm": 2.2790634632110596,
+      "learning_rate": 5e-06,
+      "loss": 0.189,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5541108986615679,
+      "grad_norm": 2.1477177143096924,
+      "learning_rate": 5e-06,
+      "loss": 0.3104,
+      "step": 2898
+    },
+    {
+      "epoch": 0.554302103250478,
+      "grad_norm": 1.7578496932983398,
+      "learning_rate": 5e-06,
+      "loss": 0.1977,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5544933078393881,
+      "grad_norm": 1.507967233657837,
+      "learning_rate": 5e-06,
+      "loss": 0.0621,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5546845124282983,
+      "grad_norm": 2.4397711753845215,
+      "learning_rate": 5e-06,
+      "loss": 0.3803,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5548757170172084,
+      "grad_norm": 1.9248405694961548,
+      "learning_rate": 5e-06,
+      "loss": 0.1198,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5550669216061186,
+      "grad_norm": 2.194777011871338,
+      "learning_rate": 5e-06,
+      "loss": 0.2595,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5552581261950287,
+      "grad_norm": 1.9272375106811523,
+      "learning_rate": 5e-06,
+      "loss": 0.1686,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5554493307839388,
+      "grad_norm": 1.9887967109680176,
+      "learning_rate": 5e-06,
+      "loss": 0.0554,
+      "step": 2905
+    },
+    {
+      "epoch": 0.555640535372849,
+      "grad_norm": 2.538315773010254,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 2906
+    },
+    {
+      "epoch": 0.555831739961759,
+      "grad_norm": 2.3389220237731934,
+      "learning_rate": 5e-06,
+      "loss": 0.2485,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5560229445506693,
+      "grad_norm": 4.87516450881958,
+      "learning_rate": 5e-06,
+      "loss": 0.3167,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5562141491395793,
+      "grad_norm": 2.524181365966797,
+      "learning_rate": 5e-06,
+      "loss": 0.3042,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5564053537284895,
+      "grad_norm": 1.860914707183838,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5565965583173996,
+      "grad_norm": 1.0398834943771362,
+      "learning_rate": 5e-06,
+      "loss": 0.1021,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5567877629063097,
+      "grad_norm": 2.4790728092193604,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5569789674952199,
+      "grad_norm": 1.9124921560287476,
+      "learning_rate": 5e-06,
+      "loss": 0.1495,
+      "step": 2913
+    },
+    {
+      "epoch": 0.55717017208413,
+      "grad_norm": 3.33526349067688,
+      "learning_rate": 5e-06,
+      "loss": 0.7095,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5573613766730402,
+      "grad_norm": 1.0180929899215698,
+      "learning_rate": 5e-06,
+      "loss": 0.1051,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5575525812619503,
+      "grad_norm": 3.9493212699890137,
+      "learning_rate": 5e-06,
+      "loss": 0.335,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5577437858508604,
+      "grad_norm": 3.441434621810913,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5579349904397706,
+      "grad_norm": 1.9761435985565186,
+      "learning_rate": 5e-06,
+      "loss": 0.1083,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5581261950286807,
+      "grad_norm": 1.1638832092285156,
+      "learning_rate": 5e-06,
+      "loss": 0.0597,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5583173996175909,
+      "grad_norm": 3.2977254390716553,
+      "learning_rate": 5e-06,
+      "loss": 0.7461,
+      "step": 2920
+    },
+    {
+      "epoch": 0.558508604206501,
+      "grad_norm": 1.3047654628753662,
+      "learning_rate": 5e-06,
+      "loss": 0.0716,
+      "step": 2921
+    },
+    {
+      "epoch": 0.558699808795411,
+      "grad_norm": 2.069331169128418,
+      "learning_rate": 5e-06,
+      "loss": 0.2381,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5588910133843212,
+      "grad_norm": 1.0748863220214844,
+      "learning_rate": 5e-06,
+      "loss": 0.0702,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5590822179732313,
+      "grad_norm": 1.0665748119354248,
+      "learning_rate": 5e-06,
+      "loss": 0.0602,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5592734225621415,
+      "grad_norm": 0.8832602500915527,
+      "learning_rate": 5e-06,
+      "loss": 0.0312,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5594646271510516,
+      "grad_norm": 2.666064739227295,
+      "learning_rate": 5e-06,
+      "loss": 0.4669,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5596558317399618,
+      "grad_norm": 1.9858882427215576,
+      "learning_rate": 5e-06,
+      "loss": 0.1749,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5598470363288719,
+      "grad_norm": 2.102036237716675,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 2928
+    },
+    {
+      "epoch": 0.560038240917782,
+      "grad_norm": 1.3998578786849976,
+      "learning_rate": 5e-06,
+      "loss": 0.1824,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5602294455066922,
+      "grad_norm": 1.6715340614318848,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5604206500956023,
+      "grad_norm": 1.6286790370941162,
+      "learning_rate": 5e-06,
+      "loss": 0.1024,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5606118546845125,
+      "grad_norm": 2.3067750930786133,
+      "learning_rate": 5e-06,
+      "loss": 0.3355,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5608030592734226,
+      "grad_norm": 2.1875245571136475,
+      "learning_rate": 5e-06,
+      "loss": 0.3912,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5609942638623326,
+      "grad_norm": 3.647351026535034,
+      "learning_rate": 5e-06,
+      "loss": 0.3519,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5611854684512428,
+      "grad_norm": 1.6276153326034546,
+      "learning_rate": 5e-06,
+      "loss": 0.1034,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5613766730401529,
+      "grad_norm": 1.9249351024627686,
+      "learning_rate": 5e-06,
+      "loss": 0.3196,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5615678776290631,
+      "grad_norm": 2.1490015983581543,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5617590822179732,
+      "grad_norm": 1.8144192695617676,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5619502868068834,
+      "grad_norm": 1.4243121147155762,
+      "learning_rate": 5e-06,
+      "loss": 0.1824,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5621414913957935,
+      "grad_norm": 2.2722017765045166,
+      "learning_rate": 5e-06,
+      "loss": 0.3895,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5623326959847036,
+      "grad_norm": 1.8516919612884521,
+      "learning_rate": 5e-06,
+      "loss": 0.0461,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5625239005736138,
+      "grad_norm": 2.4373631477355957,
+      "learning_rate": 5e-06,
+      "loss": 0.2125,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5627151051625239,
+      "grad_norm": 2.4550974369049072,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5629063097514341,
+      "grad_norm": 1.8102632761001587,
+      "learning_rate": 5e-06,
+      "loss": 0.0701,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5630975143403442,
+      "grad_norm": 2.8042306900024414,
+      "learning_rate": 5e-06,
+      "loss": 0.5347,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5632887189292543,
+      "grad_norm": 1.6445013284683228,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5634799235181644,
+      "grad_norm": 1.7900830507278442,
+      "learning_rate": 5e-06,
+      "loss": 0.2208,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5636711281070745,
+      "grad_norm": 2.265395164489746,
+      "learning_rate": 5e-06,
+      "loss": 0.2767,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5638623326959847,
+      "grad_norm": 0.9388930201530457,
+      "learning_rate": 5e-06,
+      "loss": 0.0764,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5640535372848948,
+      "grad_norm": 2.3902626037597656,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 2950
+    },
+    {
+      "epoch": 0.564244741873805,
+      "grad_norm": 2.793389320373535,
+      "learning_rate": 5e-06,
+      "loss": 0.5114,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5644359464627151,
+      "grad_norm": 1.428493618965149,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5646271510516252,
+      "grad_norm": 2.532050848007202,
+      "learning_rate": 5e-06,
+      "loss": 0.2788,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5648183556405354,
+      "grad_norm": 2.010270357131958,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5650095602294455,
+      "grad_norm": 2.6057164669036865,
+      "learning_rate": 5e-06,
+      "loss": 0.2012,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5652007648183557,
+      "grad_norm": 1.232944130897522,
+      "learning_rate": 5e-06,
+      "loss": 0.0479,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5653919694072658,
+      "grad_norm": 2.149996519088745,
+      "learning_rate": 5e-06,
+      "loss": 0.2825,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5655831739961759,
+      "grad_norm": 1.599258303642273,
+      "learning_rate": 5e-06,
+      "loss": 0.2824,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5657743785850861,
+      "grad_norm": 2.3001720905303955,
+      "learning_rate": 5e-06,
+      "loss": 0.4862,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5659655831739961,
+      "grad_norm": 3.394070863723755,
+      "learning_rate": 5e-06,
+      "loss": 0.3695,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5661567877629063,
+      "grad_norm": 2.2293701171875,
+      "learning_rate": 5e-06,
+      "loss": 0.0652,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5663479923518164,
+      "grad_norm": 2.403902769088745,
+      "learning_rate": 5e-06,
+      "loss": 0.17,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5665391969407266,
+      "grad_norm": 1.5070533752441406,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5667304015296367,
+      "grad_norm": 3.1238303184509277,
+      "learning_rate": 5e-06,
+      "loss": 0.3401,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5669216061185468,
+      "grad_norm": 2.93849778175354,
+      "learning_rate": 5e-06,
+      "loss": 0.3098,
+      "step": 2965
+    },
+    {
+      "epoch": 0.567112810707457,
+      "grad_norm": 1.2622096538543701,
+      "learning_rate": 5e-06,
+      "loss": 0.0951,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5673040152963671,
+      "grad_norm": 1.6056170463562012,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5674952198852773,
+      "grad_norm": 1.6788007020950317,
+      "learning_rate": 5e-06,
+      "loss": 0.128,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5676864244741874,
+      "grad_norm": 2.0569911003112793,
+      "learning_rate": 5e-06,
+      "loss": 0.1404,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5678776290630975,
+      "grad_norm": 3.8002560138702393,
+      "learning_rate": 5e-06,
+      "loss": 1.0882,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5680688336520077,
+      "grad_norm": 1.7185357809066772,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5682600382409178,
+      "grad_norm": 1.3213095664978027,
+      "learning_rate": 5e-06,
+      "loss": 0.104,
+      "step": 2972
+    },
+    {
+      "epoch": 0.568451242829828,
+      "grad_norm": 1.4530354738235474,
+      "learning_rate": 5e-06,
+      "loss": 0.1042,
+      "step": 2973
+    },
+    {
+      "epoch": 0.568642447418738,
+      "grad_norm": 3.0464630126953125,
+      "learning_rate": 5e-06,
+      "loss": 0.2976,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5688336520076482,
+      "grad_norm": 2.337543249130249,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5690248565965583,
+      "grad_norm": 2.6473655700683594,
+      "learning_rate": 5e-06,
+      "loss": 0.4181,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5692160611854684,
+      "grad_norm": 2.9804956912994385,
+      "learning_rate": 5e-06,
+      "loss": 0.5121,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5694072657743786,
+      "grad_norm": 1.9707882404327393,
+      "learning_rate": 5e-06,
+      "loss": 0.2911,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5695984703632887,
+      "grad_norm": 1.867041826248169,
+      "learning_rate": 5e-06,
+      "loss": 0.2345,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5697896749521989,
+      "grad_norm": 1.9662240743637085,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 2980
+    },
+    {
+      "epoch": 0.569980879541109,
+      "grad_norm": 1.9298244714736938,
+      "learning_rate": 5e-06,
+      "loss": 0.0948,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5701720841300191,
+      "grad_norm": 2.791269063949585,
+      "learning_rate": 5e-06,
+      "loss": 0.3562,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5703632887189293,
+      "grad_norm": 1.7784298658370972,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5705544933078394,
+      "grad_norm": 2.0003790855407715,
+      "learning_rate": 5e-06,
+      "loss": 0.3397,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5707456978967496,
+      "grad_norm": 1.470927119255066,
+      "learning_rate": 5e-06,
+      "loss": 0.1265,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5709369024856596,
+      "grad_norm": 2.981893539428711,
+      "learning_rate": 5e-06,
+      "loss": 0.3084,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5711281070745697,
+      "grad_norm": 1.706616997718811,
+      "learning_rate": 5e-06,
+      "loss": 0.1126,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5713193116634799,
+      "grad_norm": 1.6704820394515991,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 2988
+    },
+    {
+      "epoch": 0.57151051625239,
+      "grad_norm": 1.1496891975402832,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5717017208413002,
+      "grad_norm": 2.2269532680511475,
+      "learning_rate": 5e-06,
+      "loss": 0.1694,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5718929254302103,
+      "grad_norm": 0.9735349416732788,
+      "learning_rate": 5e-06,
+      "loss": 0.0651,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5720841300191205,
+      "grad_norm": 1.315924048423767,
+      "learning_rate": 5e-06,
+      "loss": 0.0818,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5722753346080306,
+      "grad_norm": 2.6886305809020996,
+      "learning_rate": 5e-06,
+      "loss": 0.4085,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5724665391969407,
+      "grad_norm": 1.624668836593628,
+      "learning_rate": 5e-06,
+      "loss": 0.1512,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5726577437858509,
+      "grad_norm": 2.3353240489959717,
+      "learning_rate": 5e-06,
+      "loss": 0.2948,
+      "step": 2995
+    },
+    {
+      "epoch": 0.572848948374761,
+      "grad_norm": 2.4698197841644287,
+      "learning_rate": 5e-06,
+      "loss": 0.2742,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5730401529636712,
+      "grad_norm": 2.138054370880127,
+      "learning_rate": 5e-06,
+      "loss": 0.3292,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5732313575525813,
+      "grad_norm": 2.4792983531951904,
+      "learning_rate": 5e-06,
+      "loss": 0.2956,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5734225621414913,
+      "grad_norm": 1.899868130683899,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5736137667304015,
+      "grad_norm": 3.1172773838043213,
+      "learning_rate": 5e-06,
+      "loss": 0.2009,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5736137667304015,
+      "eval_runtime": 817.4268,
+      "eval_samples_per_second": 1.877,
+      "eval_steps_per_second": 0.235,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5738049713193116,
+      "grad_norm": 2.6903367042541504,
+      "learning_rate": 5e-06,
+      "loss": 0.1569,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5739961759082218,
+      "grad_norm": 1.9376270771026611,
+      "learning_rate": 5e-06,
+      "loss": 0.2842,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5741873804971319,
+      "grad_norm": 1.1127880811691284,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5743785850860421,
+      "grad_norm": 1.381749153137207,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5745697896749522,
+      "grad_norm": 3.1074182987213135,
+      "learning_rate": 5e-06,
+      "loss": 0.1865,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5747609942638623,
+      "grad_norm": 1.8517677783966064,
+      "learning_rate": 5e-06,
+      "loss": 0.0811,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5749521988527725,
+      "grad_norm": 2.8040478229522705,
+      "learning_rate": 5e-06,
+      "loss": 0.4882,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5751434034416826,
+      "grad_norm": 1.2875118255615234,
+      "learning_rate": 5e-06,
+      "loss": 0.1718,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5753346080305928,
+      "grad_norm": 1.20366370677948,
+      "learning_rate": 5e-06,
+      "loss": 0.0758,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5755258126195029,
+      "grad_norm": 1.6420012712478638,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 3010
+    },
+    {
+      "epoch": 0.575717017208413,
+      "grad_norm": 1.7481510639190674,
+      "learning_rate": 5e-06,
+      "loss": 0.0927,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5759082217973231,
+      "grad_norm": 2.993232011795044,
+      "learning_rate": 5e-06,
+      "loss": 0.2496,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5760994263862332,
+      "grad_norm": 1.9660049676895142,
+      "learning_rate": 5e-06,
+      "loss": 0.2413,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5762906309751434,
+      "grad_norm": 1.9961297512054443,
+      "learning_rate": 5e-06,
+      "loss": 0.3713,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5764818355640535,
+      "grad_norm": 1.5734777450561523,
+      "learning_rate": 5e-06,
+      "loss": 0.1268,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5766730401529637,
+      "grad_norm": 1.7662030458450317,
+      "learning_rate": 5e-06,
+      "loss": 0.2618,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5768642447418738,
+      "grad_norm": 1.6102837324142456,
+      "learning_rate": 5e-06,
+      "loss": 0.1833,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5770554493307839,
+      "grad_norm": 2.5440475940704346,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5772466539196941,
+      "grad_norm": 2.081432580947876,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5774378585086042,
+      "grad_norm": 5.738420009613037,
+      "learning_rate": 5e-06,
+      "loss": 0.2006,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5776290630975144,
+      "grad_norm": 1.8196947574615479,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5778202676864245,
+      "grad_norm": 1.6107805967330933,
+      "learning_rate": 5e-06,
+      "loss": 0.2009,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5780114722753346,
+      "grad_norm": 1.1476129293441772,
+      "learning_rate": 5e-06,
+      "loss": 0.0376,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5782026768642448,
+      "grad_norm": 2.0225865840911865,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5783938814531548,
+      "grad_norm": 2.241004228591919,
+      "learning_rate": 5e-06,
+      "loss": 0.2727,
+      "step": 3025
+    },
+    {
+      "epoch": 0.578585086042065,
+      "grad_norm": 2.0061728954315186,
+      "learning_rate": 5e-06,
+      "loss": 0.1177,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5787762906309751,
+      "grad_norm": 2.0233190059661865,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5789674952198853,
+      "grad_norm": 1.3441364765167236,
+      "learning_rate": 5e-06,
+      "loss": 0.0912,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5791586998087954,
+      "grad_norm": 1.7095820903778076,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5793499043977055,
+      "grad_norm": 1.2957032918930054,
+      "learning_rate": 5e-06,
+      "loss": 0.0431,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5795411089866157,
+      "grad_norm": 2.276420831680298,
+      "learning_rate": 5e-06,
+      "loss": 0.2461,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5797323135755258,
+      "grad_norm": 2.8638715744018555,
+      "learning_rate": 5e-06,
+      "loss": 0.4221,
+      "step": 3032
+    },
+    {
+      "epoch": 0.579923518164436,
+      "grad_norm": 1.7681334018707275,
+      "learning_rate": 5e-06,
+      "loss": 0.218,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5801147227533461,
+      "grad_norm": 2.0881717205047607,
+      "learning_rate": 5e-06,
+      "loss": 0.3088,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5803059273422562,
+      "grad_norm": 1.6630779504776,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5804971319311664,
+      "grad_norm": 1.7102560997009277,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5806883365200765,
+      "grad_norm": 1.850621223449707,
+      "learning_rate": 5e-06,
+      "loss": 0.0667,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5808795411089867,
+      "grad_norm": 2.2591307163238525,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5810707456978967,
+      "grad_norm": 3.2868549823760986,
+      "learning_rate": 5e-06,
+      "loss": 0.7069,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5812619502868069,
+      "grad_norm": 2.1534550189971924,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 3040
+    },
+    {
+      "epoch": 0.581453154875717,
+      "grad_norm": 2.671644687652588,
+      "learning_rate": 5e-06,
+      "loss": 0.6401,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5816443594646271,
+      "grad_norm": 1.346547245979309,
+      "learning_rate": 5e-06,
+      "loss": 0.0957,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5818355640535373,
+      "grad_norm": 1.1658374071121216,
+      "learning_rate": 5e-06,
+      "loss": 0.1317,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5820267686424474,
+      "grad_norm": 1.5041284561157227,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5822179732313576,
+      "grad_norm": 2.461108922958374,
+      "learning_rate": 5e-06,
+      "loss": 0.4222,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5824091778202677,
+      "grad_norm": 1.9935189485549927,
+      "learning_rate": 5e-06,
+      "loss": 0.2012,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5826003824091778,
+      "grad_norm": 3.317734718322754,
+      "learning_rate": 5e-06,
+      "loss": 0.2825,
+      "step": 3047
+    },
+    {
+      "epoch": 0.582791586998088,
+      "grad_norm": 2.7666308879852295,
+      "learning_rate": 5e-06,
+      "loss": 0.1017,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5829827915869981,
+      "grad_norm": 1.4334791898727417,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5831739961759083,
+      "grad_norm": 2.323887825012207,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5833652007648183,
+      "grad_norm": 2.63179874420166,
+      "learning_rate": 5e-06,
+      "loss": 0.5225,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5835564053537284,
+      "grad_norm": 2.2976789474487305,
+      "learning_rate": 5e-06,
+      "loss": 0.3479,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5837476099426386,
+      "grad_norm": 1.7970479726791382,
+      "learning_rate": 5e-06,
+      "loss": 0.0878,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5839388145315487,
+      "grad_norm": 2.2620391845703125,
+      "learning_rate": 5e-06,
+      "loss": 0.2336,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5841300191204589,
+      "grad_norm": 1.0997439622879028,
+      "learning_rate": 5e-06,
+      "loss": 0.0781,
+      "step": 3055
+    },
+    {
+      "epoch": 0.584321223709369,
+      "grad_norm": 1.8581743240356445,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5845124282982792,
+      "grad_norm": 2.1834423542022705,
+      "learning_rate": 5e-06,
+      "loss": 0.1402,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5847036328871893,
+      "grad_norm": 1.7100716829299927,
+      "learning_rate": 5e-06,
+      "loss": 0.194,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5848948374760994,
+      "grad_norm": 3.5780153274536133,
+      "learning_rate": 5e-06,
+      "loss": 0.1706,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5850860420650096,
+      "grad_norm": 1.2528570890426636,
+      "learning_rate": 5e-06,
+      "loss": 0.0991,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5852772466539197,
+      "grad_norm": 1.7608826160430908,
+      "learning_rate": 5e-06,
+      "loss": 0.1739,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5854684512428299,
+      "grad_norm": 2.2554149627685547,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 3062
+    },
+    {
+      "epoch": 0.58565965583174,
+      "grad_norm": 1.9988476037979126,
+      "learning_rate": 5e-06,
+      "loss": 0.3087,
+      "step": 3063
+    },
+    {
+      "epoch": 0.58585086042065,
+      "grad_norm": 2.21211314201355,
+      "learning_rate": 5e-06,
+      "loss": 0.273,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5860420650095602,
+      "grad_norm": 1.947128415107727,
+      "learning_rate": 5e-06,
+      "loss": 0.2207,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5862332695984703,
+      "grad_norm": 3.2861714363098145,
+      "learning_rate": 5e-06,
+      "loss": 0.4373,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5864244741873805,
+      "grad_norm": 1.3944896459579468,
+      "learning_rate": 5e-06,
+      "loss": 0.0627,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5866156787762906,
+      "grad_norm": 2.653266429901123,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5868068833652008,
+      "grad_norm": 9.50674819946289,
+      "learning_rate": 5e-06,
+      "loss": 0.2661,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5869980879541109,
+      "grad_norm": 2.211970329284668,
+      "learning_rate": 5e-06,
+      "loss": 0.2697,
+      "step": 3070
+    },
+    {
+      "epoch": 0.587189292543021,
+      "grad_norm": 1.0039657354354858,
+      "learning_rate": 5e-06,
+      "loss": 0.0918,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5873804971319312,
+      "grad_norm": 2.002715587615967,
+      "learning_rate": 5e-06,
+      "loss": 0.2972,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5875717017208413,
+      "grad_norm": 1.1226190328598022,
+      "learning_rate": 5e-06,
+      "loss": 0.0923,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5877629063097515,
+      "grad_norm": 1.6911029815673828,
+      "learning_rate": 5e-06,
+      "loss": 0.0904,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5879541108986616,
+      "grad_norm": 2.079878568649292,
+      "learning_rate": 5e-06,
+      "loss": 0.0876,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5881453154875717,
+      "grad_norm": 2.1699066162109375,
+      "learning_rate": 5e-06,
+      "loss": 0.4323,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5883365200764819,
+      "grad_norm": 1.8481723070144653,
+      "learning_rate": 5e-06,
+      "loss": 0.208,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5885277246653919,
+      "grad_norm": 1.2353150844573975,
+      "learning_rate": 5e-06,
+      "loss": 0.0871,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5887189292543021,
+      "grad_norm": 3.055333375930786,
+      "learning_rate": 5e-06,
+      "loss": 0.1118,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5889101338432122,
+      "grad_norm": 1.535212755203247,
+      "learning_rate": 5e-06,
+      "loss": 0.2145,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5891013384321224,
+      "grad_norm": 3.490246295928955,
+      "learning_rate": 5e-06,
+      "loss": 0.0861,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5892925430210325,
+      "grad_norm": 1.9713093042373657,
+      "learning_rate": 5e-06,
+      "loss": 0.2104,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5894837476099426,
+      "grad_norm": 2.162919282913208,
+      "learning_rate": 5e-06,
+      "loss": 0.3863,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5896749521988528,
+      "grad_norm": 1.0454967021942139,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5898661567877629,
+      "grad_norm": 1.190918207168579,
+      "learning_rate": 5e-06,
+      "loss": 0.1009,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5900573613766731,
+      "grad_norm": 1.4480558633804321,
+      "learning_rate": 5e-06,
+      "loss": 0.0612,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5902485659655832,
+      "grad_norm": 2.896592140197754,
+      "learning_rate": 5e-06,
+      "loss": 0.121,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5904397705544933,
+      "grad_norm": 1.869307279586792,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5906309751434035,
+      "grad_norm": 1.9074268341064453,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5908221797323135,
+      "grad_norm": 2.827545166015625,
+      "learning_rate": 5e-06,
+      "loss": 0.5372,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5910133843212237,
+      "grad_norm": 3.0285253524780273,
+      "learning_rate": 5e-06,
+      "loss": 0.1217,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5912045889101338,
+      "grad_norm": 1.5774853229522705,
+      "learning_rate": 5e-06,
+      "loss": 0.2967,
+      "step": 3092
+    },
+    {
+      "epoch": 0.591395793499044,
+      "grad_norm": 1.55910062789917,
+      "learning_rate": 5e-06,
+      "loss": 0.0962,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5915869980879541,
+      "grad_norm": 2.466980457305908,
+      "learning_rate": 5e-06,
+      "loss": 0.1794,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5917782026768642,
+      "grad_norm": 1.3452435731887817,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5919694072657744,
+      "grad_norm": 1.193184494972229,
+      "learning_rate": 5e-06,
+      "loss": 0.1735,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5921606118546845,
+      "grad_norm": 4.413208961486816,
+      "learning_rate": 5e-06,
+      "loss": 0.1679,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5923518164435947,
+      "grad_norm": 1.8981341123580933,
+      "learning_rate": 5e-06,
+      "loss": 0.1866,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5925430210325048,
+      "grad_norm": 1.924362301826477,
+      "learning_rate": 5e-06,
+      "loss": 0.1081,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5927342256214149,
+      "grad_norm": 3.043461561203003,
+      "learning_rate": 5e-06,
+      "loss": 0.1592,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5929254302103251,
+      "grad_norm": 2.672950506210327,
+      "learning_rate": 5e-06,
+      "loss": 0.5285,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5931166347992352,
+      "grad_norm": 2.3183720111846924,
+      "learning_rate": 5e-06,
+      "loss": 0.5674,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5933078393881454,
+      "grad_norm": 2.15521502494812,
+      "learning_rate": 5e-06,
+      "loss": 0.23,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5934990439770554,
+      "grad_norm": 2.041330575942993,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5936902485659655,
+      "grad_norm": 2.0473520755767822,
+      "learning_rate": 5e-06,
+      "loss": 0.3554,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5938814531548757,
+      "grad_norm": 1.3054406642913818,
+      "learning_rate": 5e-06,
+      "loss": 0.1113,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5940726577437858,
+      "grad_norm": 2.71018385887146,
+      "learning_rate": 5e-06,
+      "loss": 0.3178,
+      "step": 3107
+    },
+    {
+      "epoch": 0.594263862332696,
+      "grad_norm": 1.7151776552200317,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5944550669216061,
+      "grad_norm": 2.065749168395996,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5946462715105163,
+      "grad_norm": 0.8944826126098633,
+      "learning_rate": 5e-06,
+      "loss": 0.0657,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5948374760994264,
+      "grad_norm": 1.4383631944656372,
+      "learning_rate": 5e-06,
+      "loss": 0.0874,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5950286806883365,
+      "grad_norm": 1.2038031816482544,
+      "learning_rate": 5e-06,
+      "loss": 0.0653,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5952198852772467,
+      "grad_norm": 1.0844436883926392,
+      "learning_rate": 5e-06,
+      "loss": 0.0806,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5954110898661568,
+      "grad_norm": 1.6793997287750244,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 3114
+    },
+    {
+      "epoch": 0.595602294455067,
+      "grad_norm": 2.1270453929901123,
+      "learning_rate": 5e-06,
+      "loss": 0.3629,
+      "step": 3115
+    },
+    {
+      "epoch": 0.595793499043977,
+      "grad_norm": 2.146496295928955,
+      "learning_rate": 5e-06,
+      "loss": 0.317,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5959847036328871,
+      "grad_norm": 1.409468173980713,
+      "learning_rate": 5e-06,
+      "loss": 0.0948,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5961759082217973,
+      "grad_norm": 1.9205509424209595,
+      "learning_rate": 5e-06,
+      "loss": 0.2199,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5963671128107074,
+      "grad_norm": 1.7777714729309082,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5965583173996176,
+      "grad_norm": 3.087475061416626,
+      "learning_rate": 5e-06,
+      "loss": 0.3876,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5967495219885277,
+      "grad_norm": 2.7254316806793213,
+      "learning_rate": 5e-06,
+      "loss": 0.5461,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5969407265774379,
+      "grad_norm": 1.9217915534973145,
+      "learning_rate": 5e-06,
+      "loss": 0.2579,
+      "step": 3122
+    },
+    {
+      "epoch": 0.597131931166348,
+      "grad_norm": 2.051192283630371,
+      "learning_rate": 5e-06,
+      "loss": 0.2243,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5973231357552581,
+      "grad_norm": 1.4439209699630737,
+      "learning_rate": 5e-06,
+      "loss": 0.0946,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5975143403441683,
+      "grad_norm": 0.8825449347496033,
+      "learning_rate": 5e-06,
+      "loss": 0.0568,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5977055449330784,
+      "grad_norm": 13.12146282196045,
+      "learning_rate": 5e-06,
+      "loss": 0.5537,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5978967495219886,
+      "grad_norm": 1.4262953996658325,
+      "learning_rate": 5e-06,
+      "loss": 0.1899,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5980879541108987,
+      "grad_norm": 2.342287063598633,
+      "learning_rate": 5e-06,
+      "loss": 0.312,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5982791586998087,
+      "grad_norm": 2.072007179260254,
+      "learning_rate": 5e-06,
+      "loss": 0.1151,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5984703632887189,
+      "grad_norm": 4.116969108581543,
+      "learning_rate": 5e-06,
+      "loss": 0.257,
+      "step": 3130
+    },
+    {
+      "epoch": 0.598661567877629,
+      "grad_norm": 1.1240900754928589,
+      "learning_rate": 5e-06,
+      "loss": 0.0532,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5988527724665392,
+      "grad_norm": 1.6373003721237183,
+      "learning_rate": 5e-06,
+      "loss": 0.177,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5990439770554493,
+      "grad_norm": 2.227210521697998,
+      "learning_rate": 5e-06,
+      "loss": 0.1885,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5992351816443595,
+      "grad_norm": 1.2359378337860107,
+      "learning_rate": 5e-06,
+      "loss": 0.1457,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5994263862332696,
+      "grad_norm": 1.923424482345581,
+      "learning_rate": 5e-06,
+      "loss": 0.247,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5996175908221797,
+      "grad_norm": 1.0987998247146606,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5998087954110899,
+      "grad_norm": 2.2175567150115967,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.118872880935669,
+      "learning_rate": 5e-06,
+      "loss": 0.1863,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6001912045889102,
+      "grad_norm": 2.241590738296509,
+      "learning_rate": 5e-06,
+      "loss": 0.3768,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6003824091778203,
+      "grad_norm": 2.5006096363067627,
+      "learning_rate": 5e-06,
+      "loss": 0.3921,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6005736137667304,
+      "grad_norm": 1.5924909114837646,
+      "learning_rate": 5e-06,
+      "loss": 0.1555,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6007648183556406,
+      "grad_norm": 2.0221755504608154,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6009560229445506,
+      "grad_norm": 2.870837450027466,
+      "learning_rate": 5e-06,
+      "loss": 0.0977,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6011472275334608,
+      "grad_norm": 2.1391944885253906,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6013384321223709,
+      "grad_norm": 2.3611199855804443,
+      "learning_rate": 5e-06,
+      "loss": 0.5372,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6015296367112811,
+      "grad_norm": 2.637998104095459,
+      "learning_rate": 5e-06,
+      "loss": 0.4384,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6017208413001912,
+      "grad_norm": 3.041144609451294,
+      "learning_rate": 5e-06,
+      "loss": 0.2068,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6019120458891013,
+      "grad_norm": 2.0565388202667236,
+      "learning_rate": 5e-06,
+      "loss": 0.2483,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6021032504780115,
+      "grad_norm": 2.326119899749756,
+      "learning_rate": 5e-06,
+      "loss": 0.3495,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6022944550669216,
+      "grad_norm": 2.5809388160705566,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6024856596558318,
+      "grad_norm": 2.951171398162842,
+      "learning_rate": 5e-06,
+      "loss": 0.4543,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6026768642447419,
+      "grad_norm": 1.1240613460540771,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 3152
+    },
+    {
+      "epoch": 0.602868068833652,
+      "grad_norm": 1.6378625631332397,
+      "learning_rate": 5e-06,
+      "loss": 0.1978,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6030592734225622,
+      "grad_norm": 2.002222776412964,
+      "learning_rate": 5e-06,
+      "loss": 0.1737,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6032504780114722,
+      "grad_norm": 1.2147256135940552,
+      "learning_rate": 5e-06,
+      "loss": 0.0322,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6034416826003824,
+      "grad_norm": 1.5315823554992676,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6036328871892925,
+      "grad_norm": 2.548656463623047,
+      "learning_rate": 5e-06,
+      "loss": 0.376,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6038240917782027,
+      "grad_norm": 2.330667495727539,
+      "learning_rate": 5e-06,
+      "loss": 0.3136,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6040152963671128,
+      "grad_norm": 3.249701738357544,
+      "learning_rate": 5e-06,
+      "loss": 0.7531,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6042065009560229,
+      "grad_norm": 1.688836932182312,
+      "learning_rate": 5e-06,
+      "loss": 0.1473,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6043977055449331,
+      "grad_norm": 2.461451530456543,
+      "learning_rate": 5e-06,
+      "loss": 0.246,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6045889101338432,
+      "grad_norm": 1.6401633024215698,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6047801147227534,
+      "grad_norm": 2.8850514888763428,
+      "learning_rate": 5e-06,
+      "loss": 0.466,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6049713193116635,
+      "grad_norm": 3.2717480659484863,
+      "learning_rate": 5e-06,
+      "loss": 0.5378,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6051625239005736,
+      "grad_norm": 3.4596874713897705,
+      "learning_rate": 5e-06,
+      "loss": 0.3068,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6053537284894838,
+      "grad_norm": 1.0679250955581665,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6055449330783939,
+      "grad_norm": 1.2040866613388062,
+      "learning_rate": 5e-06,
+      "loss": 0.0847,
+      "step": 3167
+    },
+    {
+      "epoch": 0.605736137667304,
+      "grad_norm": 3.1101861000061035,
+      "learning_rate": 5e-06,
+      "loss": 0.2559,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6059273422562141,
+      "grad_norm": 2.351142168045044,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6061185468451242,
+      "grad_norm": 1.722732663154602,
+      "learning_rate": 5e-06,
+      "loss": 0.2063,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6063097514340344,
+      "grad_norm": 2.38865327835083,
+      "learning_rate": 5e-06,
+      "loss": 0.4377,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6065009560229445,
+      "grad_norm": 2.085334300994873,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6066921606118547,
+      "grad_norm": 1.9132529497146606,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6068833652007648,
+      "grad_norm": 2.437124013900757,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 3174
+    },
+    {
+      "epoch": 0.607074569789675,
+      "grad_norm": 2.508739471435547,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6072657743785851,
+      "grad_norm": 2.6360321044921875,
+      "learning_rate": 5e-06,
+      "loss": 0.2388,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6074569789674952,
+      "grad_norm": 1.8685966730117798,
+      "learning_rate": 5e-06,
+      "loss": 0.2663,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6076481835564054,
+      "grad_norm": 1.755370020866394,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6078393881453155,
+      "grad_norm": 1.2698965072631836,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6080305927342257,
+      "grad_norm": 3.567136764526367,
+      "learning_rate": 5e-06,
+      "loss": 0.2535,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6082217973231357,
+      "grad_norm": 1.5889770984649658,
+      "learning_rate": 5e-06,
+      "loss": 0.0809,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6084130019120458,
+      "grad_norm": 1.9720900058746338,
+      "learning_rate": 5e-06,
+      "loss": 0.1599,
+      "step": 3182
+    },
+    {
+      "epoch": 0.608604206500956,
+      "grad_norm": 1.6620185375213623,
+      "learning_rate": 5e-06,
+      "loss": 0.2844,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6087954110898661,
+      "grad_norm": 2.424464464187622,
+      "learning_rate": 5e-06,
+      "loss": 0.2883,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6089866156787763,
+      "grad_norm": 3.4349234104156494,
+      "learning_rate": 5e-06,
+      "loss": 0.4843,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6091778202676864,
+      "grad_norm": 1.0751681327819824,
+      "learning_rate": 5e-06,
+      "loss": 0.0913,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6093690248565966,
+      "grad_norm": 1.671789526939392,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6095602294455067,
+      "grad_norm": 2.0283384323120117,
+      "learning_rate": 5e-06,
+      "loss": 0.1114,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6097514340344168,
+      "grad_norm": 2.488652229309082,
+      "learning_rate": 5e-06,
+      "loss": 0.3425,
+      "step": 3189
+    },
+    {
+      "epoch": 0.609942638623327,
+      "grad_norm": 1.1259547472000122,
+      "learning_rate": 5e-06,
+      "loss": 0.1119,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6101338432122371,
+      "grad_norm": 5.288424968719482,
+      "learning_rate": 5e-06,
+      "loss": 0.3013,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6103250478011473,
+      "grad_norm": 1.1626574993133545,
+      "learning_rate": 5e-06,
+      "loss": 0.1034,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6105162523900574,
+      "grad_norm": 2.4435818195343018,
+      "learning_rate": 5e-06,
+      "loss": 0.2786,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6107074569789674,
+      "grad_norm": 1.540985107421875,
+      "learning_rate": 5e-06,
+      "loss": 0.0953,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6108986615678776,
+      "grad_norm": 2.901230812072754,
+      "learning_rate": 5e-06,
+      "loss": 0.4028,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6110898661567877,
+      "grad_norm": 1.308781385421753,
+      "learning_rate": 5e-06,
+      "loss": 0.1013,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6112810707456979,
+      "grad_norm": 1.9014428853988647,
+      "learning_rate": 5e-06,
+      "loss": 0.2525,
+      "step": 3197
+    },
+    {
+      "epoch": 0.611472275334608,
+      "grad_norm": 1.3016637563705444,
+      "learning_rate": 5e-06,
+      "loss": 0.1119,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6116634799235182,
+      "grad_norm": 1.2226108312606812,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6118546845124283,
+      "grad_norm": 1.6033170223236084,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6120458891013384,
+      "grad_norm": 1.5445128679275513,
+      "learning_rate": 5e-06,
+      "loss": 0.1334,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6122370936902486,
+      "grad_norm": 2.2498819828033447,
+      "learning_rate": 5e-06,
+      "loss": 0.3817,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6124282982791587,
+      "grad_norm": 1.170806646347046,
+      "learning_rate": 5e-06,
+      "loss": 0.0951,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6126195028680689,
+      "grad_norm": 1.942797303199768,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 3204
+    },
+    {
+      "epoch": 0.612810707456979,
+      "grad_norm": 1.344512701034546,
+      "learning_rate": 5e-06,
+      "loss": 0.081,
+      "step": 3205
+    },
+    {
+      "epoch": 0.613001912045889,
+      "grad_norm": 1.1535552740097046,
+      "learning_rate": 5e-06,
+      "loss": 0.0581,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6131931166347993,
+      "grad_norm": 1.867598056793213,
+      "learning_rate": 5e-06,
+      "loss": 0.1903,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6133843212237093,
+      "grad_norm": 2.3806843757629395,
+      "learning_rate": 5e-06,
+      "loss": 0.3296,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6135755258126195,
+      "grad_norm": 1.6405065059661865,
+      "learning_rate": 5e-06,
+      "loss": 0.161,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6137667304015296,
+      "grad_norm": 2.7260189056396484,
+      "learning_rate": 5e-06,
+      "loss": 0.3575,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6139579349904398,
+      "grad_norm": 2.125488519668579,
+      "learning_rate": 5e-06,
+      "loss": 0.339,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6141491395793499,
+      "grad_norm": 2.283906936645508,
+      "learning_rate": 5e-06,
+      "loss": 0.1835,
+      "step": 3212
+    },
+    {
+      "epoch": 0.61434034416826,
+      "grad_norm": 2.345285654067993,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6145315487571702,
+      "grad_norm": 2.2793378829956055,
+      "learning_rate": 5e-06,
+      "loss": 0.3886,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6147227533460803,
+      "grad_norm": 2.4484992027282715,
+      "learning_rate": 5e-06,
+      "loss": 0.4042,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6149139579349905,
+      "grad_norm": 2.2618656158447266,
+      "learning_rate": 5e-06,
+      "loss": 0.2397,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6151051625239006,
+      "grad_norm": 2.3636066913604736,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6152963671128107,
+      "grad_norm": 1.7464470863342285,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6154875717017209,
+      "grad_norm": 3.1975414752960205,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 3219
+    },
+    {
+      "epoch": 0.615678776290631,
+      "grad_norm": 2.530433416366577,
+      "learning_rate": 5e-06,
+      "loss": 0.3814,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6158699808795411,
+      "grad_norm": 1.9266382455825806,
+      "learning_rate": 5e-06,
+      "loss": 0.253,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6160611854684512,
+      "grad_norm": 2.1497132778167725,
+      "learning_rate": 5e-06,
+      "loss": 0.2549,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6162523900573614,
+      "grad_norm": 1.5742179155349731,
+      "learning_rate": 5e-06,
+      "loss": 0.1135,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6164435946462715,
+      "grad_norm": 1.3919093608856201,
+      "learning_rate": 5e-06,
+      "loss": 0.0582,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6166347992351816,
+      "grad_norm": 2.1460893154144287,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6168260038240918,
+      "grad_norm": 3.082763671875,
+      "learning_rate": 5e-06,
+      "loss": 0.3345,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6170172084130019,
+      "grad_norm": 2.1403768062591553,
+      "learning_rate": 5e-06,
+      "loss": 0.1945,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6172084130019121,
+      "grad_norm": 2.1468288898468018,
+      "learning_rate": 5e-06,
+      "loss": 0.1464,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6173996175908222,
+      "grad_norm": 1.821694254875183,
+      "learning_rate": 5e-06,
+      "loss": 0.1214,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6175908221797323,
+      "grad_norm": 0.9342910051345825,
+      "learning_rate": 5e-06,
+      "loss": 0.0562,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6177820267686425,
+      "grad_norm": 1.4590644836425781,
+      "learning_rate": 5e-06,
+      "loss": 0.0798,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6179732313575526,
+      "grad_norm": 2.7206311225891113,
+      "learning_rate": 5e-06,
+      "loss": 0.3394,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6181644359464628,
+      "grad_norm": 2.2347638607025146,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6183556405353728,
+      "grad_norm": 2.2184712886810303,
+      "learning_rate": 5e-06,
+      "loss": 0.3338,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6185468451242829,
+      "grad_norm": 1.4767361879348755,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6187380497131931,
+      "grad_norm": 1.7414273023605347,
+      "learning_rate": 5e-06,
+      "loss": 0.108,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6189292543021032,
+      "grad_norm": 2.4318485260009766,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6191204588910134,
+      "grad_norm": 2.138657808303833,
+      "learning_rate": 5e-06,
+      "loss": 0.2567,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6193116634799235,
+      "grad_norm": 1.6914198398590088,
+      "learning_rate": 5e-06,
+      "loss": 0.1906,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6195028680688337,
+      "grad_norm": 2.306057929992676,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6196940726577438,
+      "grad_norm": 2.9438154697418213,
+      "learning_rate": 5e-06,
+      "loss": 0.2681,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6198852772466539,
+      "grad_norm": 0.8288871049880981,
+      "learning_rate": 5e-06,
+      "loss": 0.043,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6200764818355641,
+      "grad_norm": 2.2429580688476562,
+      "learning_rate": 5e-06,
+      "loss": 0.2368,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6202676864244742,
+      "grad_norm": 2.557755708694458,
+      "learning_rate": 5e-06,
+      "loss": 0.3005,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6204588910133844,
+      "grad_norm": 1.9078497886657715,
+      "learning_rate": 5e-06,
+      "loss": 0.1692,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6206500956022944,
+      "grad_norm": 2.475344657897949,
+      "learning_rate": 5e-06,
+      "loss": 0.2898,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6208413001912045,
+      "grad_norm": 2.7360498905181885,
+      "learning_rate": 5e-06,
+      "loss": 0.283,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6210325047801147,
+      "grad_norm": 1.3334991931915283,
+      "learning_rate": 5e-06,
+      "loss": 0.1089,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6212237093690248,
+      "grad_norm": 1.7204349040985107,
+      "learning_rate": 5e-06,
+      "loss": 0.1098,
+      "step": 3249
+    },
+    {
+      "epoch": 0.621414913957935,
+      "grad_norm": 4.411588191986084,
+      "learning_rate": 5e-06,
+      "loss": 0.1599,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6216061185468451,
+      "grad_norm": 1.615220308303833,
+      "learning_rate": 5e-06,
+      "loss": 0.128,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6217973231357553,
+      "grad_norm": 2.4545962810516357,
+      "learning_rate": 5e-06,
+      "loss": 0.3416,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6219885277246654,
+      "grad_norm": 1.9672613143920898,
+      "learning_rate": 5e-06,
+      "loss": 0.1455,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6221797323135755,
+      "grad_norm": 2.6472153663635254,
+      "learning_rate": 5e-06,
+      "loss": 0.3045,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6223709369024857,
+      "grad_norm": 1.7558777332305908,
+      "learning_rate": 5e-06,
+      "loss": 0.1715,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6225621414913958,
+      "grad_norm": 1.1533193588256836,
+      "learning_rate": 5e-06,
+      "loss": 0.0723,
+      "step": 3256
+    },
+    {
+      "epoch": 0.622753346080306,
+      "grad_norm": 2.2439403533935547,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6229445506692161,
+      "grad_norm": 2.317619800567627,
+      "learning_rate": 5e-06,
+      "loss": 0.2825,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6231357552581261,
+      "grad_norm": 1.0267956256866455,
+      "learning_rate": 5e-06,
+      "loss": 0.0867,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6233269598470363,
+      "grad_norm": 1.7949508428573608,
+      "learning_rate": 5e-06,
+      "loss": 0.1758,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6235181644359464,
+      "grad_norm": 0.780097246170044,
+      "learning_rate": 5e-06,
+      "loss": 0.0627,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6237093690248566,
+      "grad_norm": 3.0155014991760254,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6239005736137667,
+      "grad_norm": 1.4573692083358765,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6240917782026769,
+      "grad_norm": 2.20967173576355,
+      "learning_rate": 5e-06,
+      "loss": 0.265,
+      "step": 3264
+    },
+    {
+      "epoch": 0.624282982791587,
+      "grad_norm": 2.5910797119140625,
+      "learning_rate": 5e-06,
+      "loss": 0.2157,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6244741873804971,
+      "grad_norm": 1.6436420679092407,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6246653919694073,
+      "grad_norm": 1.099359154701233,
+      "learning_rate": 5e-06,
+      "loss": 0.0784,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6248565965583174,
+      "grad_norm": 2.09850811958313,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6250478011472276,
+      "grad_norm": 2.030501365661621,
+      "learning_rate": 5e-06,
+      "loss": 0.1151,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6252390057361377,
+      "grad_norm": 1.77583646774292,
+      "learning_rate": 5e-06,
+      "loss": 0.2257,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6254302103250478,
+      "grad_norm": 2.196660280227661,
+      "learning_rate": 5e-06,
+      "loss": 0.2526,
+      "step": 3271
+    },
+    {
+      "epoch": 0.625621414913958,
+      "grad_norm": 2.8073387145996094,
+      "learning_rate": 5e-06,
+      "loss": 0.1847,
+      "step": 3272
+    },
+    {
+      "epoch": 0.625812619502868,
+      "grad_norm": 1.610040307044983,
+      "learning_rate": 5e-06,
+      "loss": 0.0844,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6260038240917782,
+      "grad_norm": 4.642181396484375,
+      "learning_rate": 5e-06,
+      "loss": 0.1936,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6261950286806883,
+      "grad_norm": 1.8837900161743164,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6263862332695985,
+      "grad_norm": 2.7366063594818115,
+      "learning_rate": 5e-06,
+      "loss": 0.3408,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6265774378585086,
+      "grad_norm": 2.5217320919036865,
+      "learning_rate": 5e-06,
+      "loss": 0.3369,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6267686424474187,
+      "grad_norm": 2.0307939052581787,
+      "learning_rate": 5e-06,
+      "loss": 0.2868,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6269598470363289,
+      "grad_norm": 2.022230386734009,
+      "learning_rate": 5e-06,
+      "loss": 0.2864,
+      "step": 3279
+    },
+    {
+      "epoch": 0.627151051625239,
+      "grad_norm": 1.771817684173584,
+      "learning_rate": 5e-06,
+      "loss": 0.086,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6273422562141492,
+      "grad_norm": 1.1006555557250977,
+      "learning_rate": 5e-06,
+      "loss": 0.036,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6275334608030593,
+      "grad_norm": 1.6326366662979126,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6277246653919694,
+      "grad_norm": 2.166058301925659,
+      "learning_rate": 5e-06,
+      "loss": 0.3915,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6279158699808796,
+      "grad_norm": 1.752289056777954,
+      "learning_rate": 5e-06,
+      "loss": 0.1433,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6281070745697896,
+      "grad_norm": 1.6301438808441162,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6282982791586998,
+      "grad_norm": 3.006061553955078,
+      "learning_rate": 5e-06,
+      "loss": 0.0893,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6284894837476099,
+      "grad_norm": 1.6843498945236206,
+      "learning_rate": 5e-06,
+      "loss": 0.0768,
+      "step": 3287
+    },
+    {
+      "epoch": 0.62868068833652,
+      "grad_norm": 2.755248546600342,
+      "learning_rate": 5e-06,
+      "loss": 0.4565,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6288718929254302,
+      "grad_norm": 1.122692584991455,
+      "learning_rate": 5e-06,
+      "loss": 0.1002,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6290630975143403,
+      "grad_norm": 1.585771918296814,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6292543021032505,
+      "grad_norm": 1.6943069696426392,
+      "learning_rate": 5e-06,
+      "loss": 0.0915,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6294455066921606,
+      "grad_norm": 1.3940718173980713,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6296367112810708,
+      "grad_norm": 1.1676661968231201,
+      "learning_rate": 5e-06,
+      "loss": 0.0367,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6298279158699809,
+      "grad_norm": 2.5931499004364014,
+      "learning_rate": 5e-06,
+      "loss": 0.2025,
+      "step": 3294
+    },
+    {
+      "epoch": 0.630019120458891,
+      "grad_norm": 1.825156331062317,
+      "learning_rate": 5e-06,
+      "loss": 0.1334,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6302103250478012,
+      "grad_norm": 0.9374414086341858,
+      "learning_rate": 5e-06,
+      "loss": 0.0959,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6304015296367113,
+      "grad_norm": 1.9733580350875854,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6305927342256215,
+      "grad_norm": 1.1101990938186646,
+      "learning_rate": 5e-06,
+      "loss": 0.0681,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6307839388145315,
+      "grad_norm": 1.4817825555801392,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6309751434034416,
+      "grad_norm": 1.6432628631591797,
+      "learning_rate": 5e-06,
+      "loss": 0.0721,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6311663479923518,
+      "grad_norm": 3.0500617027282715,
+      "learning_rate": 5e-06,
+      "loss": 0.4934,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6313575525812619,
+      "grad_norm": 1.2397555112838745,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6315487571701721,
+      "grad_norm": 1.4868794679641724,
+      "learning_rate": 5e-06,
+      "loss": 0.251,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6317399617590822,
+      "grad_norm": 1.124295711517334,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6319311663479924,
+      "grad_norm": 1.9603227376937866,
+      "learning_rate": 5e-06,
+      "loss": 0.2923,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6321223709369025,
+      "grad_norm": 1.851251482963562,
+      "learning_rate": 5e-06,
+      "loss": 0.1242,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6323135755258126,
+      "grad_norm": 2.5106050968170166,
+      "learning_rate": 5e-06,
+      "loss": 0.2472,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6325047801147228,
+      "grad_norm": 1.8182765245437622,
+      "learning_rate": 5e-06,
+      "loss": 0.221,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6326959847036329,
+      "grad_norm": 1.9880139827728271,
+      "learning_rate": 5e-06,
+      "loss": 0.075,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6328871892925431,
+      "grad_norm": 1.8051013946533203,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6330783938814531,
+      "grad_norm": 1.6434516906738281,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6332695984703632,
+      "grad_norm": 1.805076241493225,
+      "learning_rate": 5e-06,
+      "loss": 0.0783,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6334608030592734,
+      "grad_norm": 2.02512788772583,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6336520076481835,
+      "grad_norm": 1.037001132965088,
+      "learning_rate": 5e-06,
+      "loss": 0.0982,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6338432122370937,
+      "grad_norm": 1.426772117614746,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6340344168260038,
+      "grad_norm": 1.1725845336914062,
+      "learning_rate": 5e-06,
+      "loss": 0.0841,
+      "step": 3316
+    },
+    {
+      "epoch": 0.634225621414914,
+      "grad_norm": 1.404199242591858,
+      "learning_rate": 5e-06,
+      "loss": 0.0808,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6344168260038241,
+      "grad_norm": 1.5793739557266235,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6346080305927342,
+      "grad_norm": 2.362203359603882,
+      "learning_rate": 5e-06,
+      "loss": 0.1639,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6347992351816444,
+      "grad_norm": 2.5650269985198975,
+      "learning_rate": 5e-06,
+      "loss": 0.3538,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6349904397705545,
+      "grad_norm": 1.7669709920883179,
+      "learning_rate": 5e-06,
+      "loss": 0.1892,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6351816443594647,
+      "grad_norm": 3.3054776191711426,
+      "learning_rate": 5e-06,
+      "loss": 0.5275,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6353728489483748,
+      "grad_norm": 1.7212293148040771,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6355640535372848,
+      "grad_norm": 2.0429680347442627,
+      "learning_rate": 5e-06,
+      "loss": 0.0866,
+      "step": 3324
+    },
+    {
+      "epoch": 0.635755258126195,
+      "grad_norm": 1.428225040435791,
+      "learning_rate": 5e-06,
+      "loss": 0.0705,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6359464627151051,
+      "grad_norm": 2.1418018341064453,
+      "learning_rate": 5e-06,
+      "loss": 0.3357,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6361376673040153,
+      "grad_norm": 1.53099524974823,
+      "learning_rate": 5e-06,
+      "loss": 0.1017,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6363288718929254,
+      "grad_norm": 1.5691248178482056,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6365200764818356,
+      "grad_norm": 2.0633819103240967,
+      "learning_rate": 5e-06,
+      "loss": 0.2149,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6367112810707457,
+      "grad_norm": 2.932884693145752,
+      "learning_rate": 5e-06,
+      "loss": 0.3623,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6369024856596558,
+      "grad_norm": 1.8070765733718872,
+      "learning_rate": 5e-06,
+      "loss": 0.097,
+      "step": 3331
+    },
+    {
+      "epoch": 0.637093690248566,
+      "grad_norm": 2.7510499954223633,
+      "learning_rate": 5e-06,
+      "loss": 0.3189,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6372848948374761,
+      "grad_norm": 2.5223939418792725,
+      "learning_rate": 5e-06,
+      "loss": 0.4771,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6374760994263863,
+      "grad_norm": 2.9238736629486084,
+      "learning_rate": 5e-06,
+      "loss": 0.2895,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6376673040152964,
+      "grad_norm": 0.8285096883773804,
+      "learning_rate": 5e-06,
+      "loss": 0.0522,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6378585086042065,
+      "grad_norm": 1.8238945007324219,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6380497131931167,
+      "grad_norm": 2.4051716327667236,
+      "learning_rate": 5e-06,
+      "loss": 0.1254,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6382409177820267,
+      "grad_norm": 3.326117515563965,
+      "learning_rate": 5e-06,
+      "loss": 0.5696,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6384321223709369,
+      "grad_norm": 2.6902754306793213,
+      "learning_rate": 5e-06,
+      "loss": 0.211,
+      "step": 3339
+    },
+    {
+      "epoch": 0.638623326959847,
+      "grad_norm": 2.737630605697632,
+      "learning_rate": 5e-06,
+      "loss": 0.3434,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6388145315487572,
+      "grad_norm": 1.2848268747329712,
+      "learning_rate": 5e-06,
+      "loss": 0.0847,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6390057361376673,
+      "grad_norm": 2.3309013843536377,
+      "learning_rate": 5e-06,
+      "loss": 0.1437,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6391969407265774,
+      "grad_norm": 2.1647744178771973,
+      "learning_rate": 5e-06,
+      "loss": 0.0887,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6393881453154876,
+      "grad_norm": 1.908402681350708,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6395793499043977,
+      "grad_norm": 1.631221055984497,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6397705544933079,
+      "grad_norm": 3.7973790168762207,
+      "learning_rate": 5e-06,
+      "loss": 0.5618,
+      "step": 3346
+    },
+    {
+      "epoch": 0.639961759082218,
+      "grad_norm": 1.2875862121582031,
+      "learning_rate": 5e-06,
+      "loss": 0.0914,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6401529636711281,
+      "grad_norm": 1.294562578201294,
+      "learning_rate": 5e-06,
+      "loss": 0.0698,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6403441682600383,
+      "grad_norm": 2.602620840072632,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6405353728489483,
+      "grad_norm": 2.0691466331481934,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6407265774378585,
+      "grad_norm": 1.8351796865463257,
+      "learning_rate": 5e-06,
+      "loss": 0.2551,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6409177820267686,
+      "grad_norm": 2.922478199005127,
+      "learning_rate": 5e-06,
+      "loss": 0.3711,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6411089866156787,
+      "grad_norm": 1.2165088653564453,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6413001912045889,
+      "grad_norm": 1.7305443286895752,
+      "learning_rate": 5e-06,
+      "loss": 0.0971,
+      "step": 3354
+    },
+    {
+      "epoch": 0.641491395793499,
+      "grad_norm": 2.2913918495178223,
+      "learning_rate": 5e-06,
+      "loss": 0.0785,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6416826003824092,
+      "grad_norm": 2.2279272079467773,
+      "learning_rate": 5e-06,
+      "loss": 0.143,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6418738049713193,
+      "grad_norm": 2.924514055252075,
+      "learning_rate": 5e-06,
+      "loss": 0.4447,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6420650095602295,
+      "grad_norm": 2.028049945831299,
+      "learning_rate": 5e-06,
+      "loss": 0.2733,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6422562141491396,
+      "grad_norm": 1.2572370767593384,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6424474187380497,
+      "grad_norm": 1.216548204421997,
+      "learning_rate": 5e-06,
+      "loss": 0.0975,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6426386233269599,
+      "grad_norm": 1.6940453052520752,
+      "learning_rate": 5e-06,
+      "loss": 0.0758,
+      "step": 3361
+    },
+    {
+      "epoch": 0.64282982791587,
+      "grad_norm": 1.7871954441070557,
+      "learning_rate": 5e-06,
+      "loss": 0.0795,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6430210325047802,
+      "grad_norm": 1.6503570079803467,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6432122370936902,
+      "grad_norm": 2.5519793033599854,
+      "learning_rate": 5e-06,
+      "loss": 0.4499,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6434034416826003,
+      "grad_norm": 1.5863988399505615,
+      "learning_rate": 5e-06,
+      "loss": 0.085,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6435946462715105,
+      "grad_norm": 0.7910031080245972,
+      "learning_rate": 5e-06,
+      "loss": 0.068,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6437858508604206,
+      "grad_norm": 1.732580304145813,
+      "learning_rate": 5e-06,
+      "loss": 0.1334,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6439770554493308,
+      "grad_norm": 2.4041748046875,
+      "learning_rate": 5e-06,
+      "loss": 0.3406,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6441682600382409,
+      "grad_norm": 1.4867603778839111,
+      "learning_rate": 5e-06,
+      "loss": 0.0677,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6443594646271511,
+      "grad_norm": 1.6989647150039673,
+      "learning_rate": 5e-06,
+      "loss": 0.1288,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6445506692160612,
+      "grad_norm": 1.4293967485427856,
+      "learning_rate": 5e-06,
+      "loss": 0.1531,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6447418738049713,
+      "grad_norm": 1.3375941514968872,
+      "learning_rate": 5e-06,
+      "loss": 0.1033,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6449330783938815,
+      "grad_norm": 0.9712766408920288,
+      "learning_rate": 5e-06,
+      "loss": 0.0382,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6451242829827916,
+      "grad_norm": 2.1650278568267822,
+      "learning_rate": 5e-06,
+      "loss": 0.2237,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6453154875717018,
+      "grad_norm": 1.0781400203704834,
+      "learning_rate": 5e-06,
+      "loss": 0.0441,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6455066921606119,
+      "grad_norm": 2.823831081390381,
+      "learning_rate": 5e-06,
+      "loss": 0.3515,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6456978967495219,
+      "grad_norm": 1.8595666885375977,
+      "learning_rate": 5e-06,
+      "loss": 0.3496,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6458891013384321,
+      "grad_norm": 2.762394905090332,
+      "learning_rate": 5e-06,
+      "loss": 0.4007,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6460803059273422,
+      "grad_norm": 1.8678456544876099,
+      "learning_rate": 5e-06,
+      "loss": 0.2212,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6462715105162524,
+      "grad_norm": 1.9255629777908325,
+      "learning_rate": 5e-06,
+      "loss": 0.2845,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6464627151051625,
+      "grad_norm": 2.22768497467041,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6466539196940727,
+      "grad_norm": 2.3426296710968018,
+      "learning_rate": 5e-06,
+      "loss": 0.4442,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6468451242829828,
+      "grad_norm": 2.8555681705474854,
+      "learning_rate": 5e-06,
+      "loss": 0.1754,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6470363288718929,
+      "grad_norm": 2.99185848236084,
+      "learning_rate": 5e-06,
+      "loss": 0.2644,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6472275334608031,
+      "grad_norm": 2.5336363315582275,
+      "learning_rate": 5e-06,
+      "loss": 0.4541,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6474187380497132,
+      "grad_norm": 2.059492826461792,
+      "learning_rate": 5e-06,
+      "loss": 0.2929,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6476099426386234,
+      "grad_norm": 1.113668441772461,
+      "learning_rate": 5e-06,
+      "loss": 0.1024,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6478011472275335,
+      "grad_norm": 2.365506649017334,
+      "learning_rate": 5e-06,
+      "loss": 0.3054,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6479923518164435,
+      "grad_norm": 2.7506308555603027,
+      "learning_rate": 5e-06,
+      "loss": 0.4685,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6481835564053537,
+      "grad_norm": 1.8486632108688354,
+      "learning_rate": 5e-06,
+      "loss": 0.1623,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6483747609942638,
+      "grad_norm": 3.001056432723999,
+      "learning_rate": 5e-06,
+      "loss": 0.1733,
+      "step": 3391
+    },
+    {
+      "epoch": 0.648565965583174,
+      "grad_norm": 1.8992429971694946,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6487571701720841,
+      "grad_norm": 0.8879466652870178,
+      "learning_rate": 5e-06,
+      "loss": 0.0552,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6489483747609943,
+      "grad_norm": 1.9001550674438477,
+      "learning_rate": 5e-06,
+      "loss": 0.1817,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6491395793499044,
+      "grad_norm": 2.546579122543335,
+      "learning_rate": 5e-06,
+      "loss": 0.2765,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6493307839388145,
+      "grad_norm": 1.5954774618148804,
+      "learning_rate": 5e-06,
+      "loss": 0.3091,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6495219885277247,
+      "grad_norm": 2.02475905418396,
+      "learning_rate": 5e-06,
+      "loss": 0.2654,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6497131931166348,
+      "grad_norm": 2.6336565017700195,
+      "learning_rate": 5e-06,
+      "loss": 0.1651,
+      "step": 3398
+    },
+    {
+      "epoch": 0.649904397705545,
+      "grad_norm": 2.454435348510742,
+      "learning_rate": 5e-06,
+      "loss": 0.0972,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6500956022944551,
+      "grad_norm": 1.3472257852554321,
+      "learning_rate": 5e-06,
+      "loss": 0.0623,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6502868068833652,
+      "grad_norm": 2.588073492050171,
+      "learning_rate": 5e-06,
+      "loss": 0.3797,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6504780114722754,
+      "grad_norm": 1.4676856994628906,
+      "learning_rate": 5e-06,
+      "loss": 0.1556,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6506692160611854,
+      "grad_norm": 1.6890233755111694,
+      "learning_rate": 5e-06,
+      "loss": 0.1048,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6508604206500956,
+      "grad_norm": 2.6827495098114014,
+      "learning_rate": 5e-06,
+      "loss": 0.223,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6510516252390057,
+      "grad_norm": 2.6940207481384277,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6512428298279158,
+      "grad_norm": 2.2096269130706787,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 3406
+    },
+    {
+      "epoch": 0.651434034416826,
+      "grad_norm": 1.9896823167800903,
+      "learning_rate": 5e-06,
+      "loss": 0.3026,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6516252390057361,
+      "grad_norm": 2.7920219898223877,
+      "learning_rate": 5e-06,
+      "loss": 0.3761,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6518164435946463,
+      "grad_norm": 1.8184442520141602,
+      "learning_rate": 5e-06,
+      "loss": 0.126,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6520076481835564,
+      "grad_norm": 1.1374013423919678,
+      "learning_rate": 5e-06,
+      "loss": 0.0662,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6521988527724666,
+      "grad_norm": 2.6526119709014893,
+      "learning_rate": 5e-06,
+      "loss": 0.3993,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6523900573613767,
+      "grad_norm": 2.8433001041412354,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6525812619502868,
+      "grad_norm": 0.9985430836677551,
+      "learning_rate": 5e-06,
+      "loss": 0.08,
+      "step": 3413
+    },
+    {
+      "epoch": 0.652772466539197,
+      "grad_norm": 2.4228732585906982,
+      "learning_rate": 5e-06,
+      "loss": 0.3744,
+      "step": 3414
+    },
+    {
+      "epoch": 0.652963671128107,
+      "grad_norm": 1.5509470701217651,
+      "learning_rate": 5e-06,
+      "loss": 0.2037,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6531548757170172,
+      "grad_norm": 1.4869673252105713,
+      "learning_rate": 5e-06,
+      "loss": 0.093,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6533460803059273,
+      "grad_norm": 0.611136257648468,
+      "learning_rate": 5e-06,
+      "loss": 0.0593,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6535372848948374,
+      "grad_norm": 1.8549314737319946,
+      "learning_rate": 5e-06,
+      "loss": 0.0853,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6537284894837476,
+      "grad_norm": 1.9247181415557861,
+      "learning_rate": 5e-06,
+      "loss": 0.2044,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6539196940726577,
+      "grad_norm": 2.686331033706665,
+      "learning_rate": 5e-06,
+      "loss": 0.18,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6541108986615679,
+      "grad_norm": 1.9724444150924683,
+      "learning_rate": 5e-06,
+      "loss": 0.2007,
+      "step": 3421
+    },
+    {
+      "epoch": 0.654302103250478,
+      "grad_norm": 1.914251685142517,
+      "learning_rate": 5e-06,
+      "loss": 0.2638,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6544933078393882,
+      "grad_norm": 1.8022854328155518,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6546845124282983,
+      "grad_norm": 2.4049770832061768,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6548757170172084,
+      "grad_norm": 1.664724349975586,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6550669216061186,
+      "grad_norm": 2.4440340995788574,
+      "learning_rate": 5e-06,
+      "loss": 0.2504,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6552581261950287,
+      "grad_norm": 1.7133796215057373,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6554493307839389,
+      "grad_norm": 2.668142080307007,
+      "learning_rate": 5e-06,
+      "loss": 0.2902,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6556405353728489,
+      "grad_norm": 2.5641353130340576,
+      "learning_rate": 5e-06,
+      "loss": 0.3005,
+      "step": 3429
+    },
+    {
+      "epoch": 0.655831739961759,
+      "grad_norm": 2.3122782707214355,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6560229445506692,
+      "grad_norm": 1.7354764938354492,
+      "learning_rate": 5e-06,
+      "loss": 0.0937,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6562141491395793,
+      "grad_norm": 2.0829017162323,
+      "learning_rate": 5e-06,
+      "loss": 0.3719,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6564053537284895,
+      "grad_norm": 1.4630341529846191,
+      "learning_rate": 5e-06,
+      "loss": 0.1413,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6565965583173996,
+      "grad_norm": 1.0866786241531372,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6567877629063098,
+      "grad_norm": 2.1426541805267334,
+      "learning_rate": 5e-06,
+      "loss": 0.4141,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6569789674952199,
+      "grad_norm": 3.4362006187438965,
+      "learning_rate": 5e-06,
+      "loss": 0.2099,
+      "step": 3436
+    },
+    {
+      "epoch": 0.65717017208413,
+      "grad_norm": 2.1101267337799072,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6573613766730402,
+      "grad_norm": 1.7151868343353271,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6575525812619503,
+      "grad_norm": 1.1048760414123535,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6577437858508605,
+      "grad_norm": 2.037752628326416,
+      "learning_rate": 5e-06,
+      "loss": 0.2758,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6579349904397706,
+      "grad_norm": 1.7046453952789307,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6581261950286806,
+      "grad_norm": 1.204085350036621,
+      "learning_rate": 5e-06,
+      "loss": 0.0877,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6583173996175908,
+      "grad_norm": 1.8280205726623535,
+      "learning_rate": 5e-06,
+      "loss": 0.1489,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6585086042065009,
+      "grad_norm": 1.589545726776123,
+      "learning_rate": 5e-06,
+      "loss": 0.0657,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6586998087954111,
+      "grad_norm": 2.3866069316864014,
+      "learning_rate": 5e-06,
+      "loss": 0.2623,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6588910133843212,
+      "grad_norm": 1.6342370510101318,
+      "learning_rate": 5e-06,
+      "loss": 0.336,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6590822179732314,
+      "grad_norm": 1.841517448425293,
+      "learning_rate": 5e-06,
+      "loss": 0.1794,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6592734225621415,
+      "grad_norm": 1.5720770359039307,
+      "learning_rate": 5e-06,
+      "loss": 0.0911,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6594646271510516,
+      "grad_norm": 3.1381449699401855,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6596558317399618,
+      "grad_norm": 2.0292165279388428,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6598470363288719,
+      "grad_norm": 2.1982815265655518,
+      "learning_rate": 5e-06,
+      "loss": 0.2026,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6600382409177821,
+      "grad_norm": 3.106961727142334,
+      "learning_rate": 5e-06,
+      "loss": 0.0886,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6602294455066922,
+      "grad_norm": 2.6791203022003174,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6604206500956022,
+      "grad_norm": 0.9955388903617859,
+      "learning_rate": 5e-06,
+      "loss": 0.0829,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6606118546845124,
+      "grad_norm": 1.959532618522644,
+      "learning_rate": 5e-06,
+      "loss": 0.1379,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6608030592734225,
+      "grad_norm": 1.4918891191482544,
+      "learning_rate": 5e-06,
+      "loss": 0.0437,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6609942638623327,
+      "grad_norm": 2.387288808822632,
+      "learning_rate": 5e-06,
+      "loss": 0.2389,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6611854684512428,
+      "grad_norm": 2.7315800189971924,
+      "learning_rate": 5e-06,
+      "loss": 0.4122,
+      "step": 3458
+    },
+    {
+      "epoch": 0.661376673040153,
+      "grad_norm": 3.008780002593994,
+      "learning_rate": 5e-06,
+      "loss": 0.5461,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6615678776290631,
+      "grad_norm": 1.4275813102722168,
+      "learning_rate": 5e-06,
+      "loss": 0.0968,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6617590822179732,
+      "grad_norm": 2.498218297958374,
+      "learning_rate": 5e-06,
+      "loss": 0.176,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6619502868068834,
+      "grad_norm": 2.3589062690734863,
+      "learning_rate": 5e-06,
+      "loss": 0.1207,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6621414913957935,
+      "grad_norm": 1.8128048181533813,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6623326959847037,
+      "grad_norm": 2.6147453784942627,
+      "learning_rate": 5e-06,
+      "loss": 0.4276,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6625239005736138,
+      "grad_norm": 1.5070366859436035,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6627151051625239,
+      "grad_norm": 1.6204683780670166,
+      "learning_rate": 5e-06,
+      "loss": 0.0838,
+      "step": 3466
+    },
+    {
+      "epoch": 0.662906309751434,
+      "grad_norm": 0.9713101983070374,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6630975143403441,
+      "grad_norm": 2.181532621383667,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6632887189292543,
+      "grad_norm": 2.44734787940979,
+      "learning_rate": 5e-06,
+      "loss": 0.1991,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6634799235181644,
+      "grad_norm": 3.1234562397003174,
+      "learning_rate": 5e-06,
+      "loss": 0.6639,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6636711281070745,
+      "grad_norm": 1.2073078155517578,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6638623326959847,
+      "grad_norm": 2.172306537628174,
+      "learning_rate": 5e-06,
+      "loss": 0.2327,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6640535372848948,
+      "grad_norm": 1.4012967348098755,
+      "learning_rate": 5e-06,
+      "loss": 0.1017,
+      "step": 3473
+    },
+    {
+      "epoch": 0.664244741873805,
+      "grad_norm": 1.9511357545852661,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6644359464627151,
+      "grad_norm": 4.60706090927124,
+      "learning_rate": 5e-06,
+      "loss": 0.1754,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6646271510516253,
+      "grad_norm": 2.3865597248077393,
+      "learning_rate": 5e-06,
+      "loss": 0.3496,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6648183556405354,
+      "grad_norm": 2.6536009311676025,
+      "learning_rate": 5e-06,
+      "loss": 0.2316,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6650095602294455,
+      "grad_norm": 1.3826388120651245,
+      "learning_rate": 5e-06,
+      "loss": 0.159,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6652007648183557,
+      "grad_norm": 1.5564355850219727,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6653919694072657,
+      "grad_norm": 1.9521403312683105,
+      "learning_rate": 5e-06,
+      "loss": 0.1619,
+      "step": 3480
+    },
+    {
+      "epoch": 0.665583173996176,
+      "grad_norm": 1.7389386892318726,
+      "learning_rate": 5e-06,
+      "loss": 0.0573,
+      "step": 3481
+    },
+    {
+      "epoch": 0.665774378585086,
+      "grad_norm": 1.4144505262374878,
+      "learning_rate": 5e-06,
+      "loss": 0.245,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6659655831739961,
+      "grad_norm": 1.4570914506912231,
+      "learning_rate": 5e-06,
+      "loss": 0.1495,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6661567877629063,
+      "grad_norm": 1.8888745307922363,
+      "learning_rate": 5e-06,
+      "loss": 0.2384,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6663479923518164,
+      "grad_norm": 2.382516860961914,
+      "learning_rate": 5e-06,
+      "loss": 0.2814,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6665391969407266,
+      "grad_norm": 1.7876882553100586,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6667304015296367,
+      "grad_norm": 1.878159999847412,
+      "learning_rate": 5e-06,
+      "loss": 0.1178,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6669216061185469,
+      "grad_norm": 2.022584915161133,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 3488
+    },
+    {
+      "epoch": 0.667112810707457,
+      "grad_norm": 2.123523473739624,
+      "learning_rate": 5e-06,
+      "loss": 0.1952,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6673040152963671,
+      "grad_norm": 2.8303143978118896,
+      "learning_rate": 5e-06,
+      "loss": 0.3464,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6674952198852773,
+      "grad_norm": 1.546257495880127,
+      "learning_rate": 5e-06,
+      "loss": 0.1826,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6676864244741874,
+      "grad_norm": 1.338302731513977,
+      "learning_rate": 5e-06,
+      "loss": 0.109,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6678776290630976,
+      "grad_norm": 1.4451181888580322,
+      "learning_rate": 5e-06,
+      "loss": 0.1015,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6680688336520076,
+      "grad_norm": 1.242767095565796,
+      "learning_rate": 5e-06,
+      "loss": 0.0837,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6682600382409177,
+      "grad_norm": 2.2983875274658203,
+      "learning_rate": 5e-06,
+      "loss": 0.3422,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6684512428298279,
+      "grad_norm": 2.008904218673706,
+      "learning_rate": 5e-06,
+      "loss": 0.2969,
+      "step": 3496
+    },
+    {
+      "epoch": 0.668642447418738,
+      "grad_norm": 2.3049376010894775,
+      "learning_rate": 5e-06,
+      "loss": 0.3266,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6688336520076482,
+      "grad_norm": 2.297036647796631,
+      "learning_rate": 5e-06,
+      "loss": 0.2258,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6690248565965583,
+      "grad_norm": 1.6770265102386475,
+      "learning_rate": 5e-06,
+      "loss": 0.0976,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6692160611854685,
+      "grad_norm": 0.9409367442131042,
+      "learning_rate": 5e-06,
+      "loss": 0.034,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6692160611854685,
+      "eval_runtime": 786.6312,
+      "eval_samples_per_second": 1.95,
+      "eval_steps_per_second": 0.244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6694072657743786,
+      "grad_norm": 1.6677579879760742,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6695984703632887,
+      "grad_norm": 1.5019170045852661,
+      "learning_rate": 5e-06,
+      "loss": 0.2266,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6697896749521989,
+      "grad_norm": 2.645740509033203,
+      "learning_rate": 5e-06,
+      "loss": 0.3297,
+      "step": 3503
+    },
+    {
+      "epoch": 0.669980879541109,
+      "grad_norm": 2.4888534545898438,
+      "learning_rate": 5e-06,
+      "loss": 0.2745,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6701720841300192,
+      "grad_norm": 1.4102755784988403,
+      "learning_rate": 5e-06,
+      "loss": 0.0735,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6703632887189293,
+      "grad_norm": 11.278008460998535,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6705544933078393,
+      "grad_norm": 2.274052381515503,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6707456978967495,
+      "grad_norm": 1.99003005027771,
+      "learning_rate": 5e-06,
+      "loss": 0.268,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6709369024856596,
+      "grad_norm": 1.6028283834457397,
+      "learning_rate": 5e-06,
+      "loss": 0.1354,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6711281070745698,
+      "grad_norm": 2.1967668533325195,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6713193116634799,
+      "grad_norm": 2.1785519123077393,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6715105162523901,
+      "grad_norm": 1.8150266408920288,
+      "learning_rate": 5e-06,
+      "loss": 0.0938,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6717017208413002,
+      "grad_norm": 1.4618399143218994,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6718929254302103,
+      "grad_norm": 3.069847822189331,
+      "learning_rate": 5e-06,
+      "loss": 0.5254,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6720841300191205,
+      "grad_norm": 1.6214332580566406,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6722753346080306,
+      "grad_norm": 2.7934930324554443,
+      "learning_rate": 5e-06,
+      "loss": 0.2811,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6724665391969408,
+      "grad_norm": 1.2103495597839355,
+      "learning_rate": 5e-06,
+      "loss": 0.075,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6726577437858509,
+      "grad_norm": 1.6577422618865967,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 3518
+    },
+    {
+      "epoch": 0.672848948374761,
+      "grad_norm": 1.5630927085876465,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6730401529636711,
+      "grad_norm": 3.0157723426818848,
+      "learning_rate": 5e-06,
+      "loss": 0.5516,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6732313575525812,
+      "grad_norm": 2.152554988861084,
+      "learning_rate": 5e-06,
+      "loss": 0.3016,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6734225621414914,
+      "grad_norm": 2.146681547164917,
+      "learning_rate": 5e-06,
+      "loss": 0.4707,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6736137667304015,
+      "grad_norm": 2.0331432819366455,
+      "learning_rate": 5e-06,
+      "loss": 0.1046,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6738049713193117,
+      "grad_norm": 1.3789559602737427,
+      "learning_rate": 5e-06,
+      "loss": 0.076,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6739961759082218,
+      "grad_norm": 0.8906503319740295,
+      "learning_rate": 5e-06,
+      "loss": 0.0667,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6741873804971319,
+      "grad_norm": 2.7111520767211914,
+      "learning_rate": 5e-06,
+      "loss": 0.3719,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6743785850860421,
+      "grad_norm": 2.096019744873047,
+      "learning_rate": 5e-06,
+      "loss": 0.3089,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6745697896749522,
+      "grad_norm": 1.231835961341858,
+      "learning_rate": 5e-06,
+      "loss": 0.0677,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6747609942638624,
+      "grad_norm": 1.7469233274459839,
+      "learning_rate": 5e-06,
+      "loss": 0.0908,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6749521988527725,
+      "grad_norm": 0.5927736163139343,
+      "learning_rate": 5e-06,
+      "loss": 0.0342,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6751434034416826,
+      "grad_norm": 3.6534321308135986,
+      "learning_rate": 5e-06,
+      "loss": 0.1764,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6753346080305928,
+      "grad_norm": 1.5184030532836914,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6755258126195028,
+      "grad_norm": 1.760200023651123,
+      "learning_rate": 5e-06,
+      "loss": 0.1343,
+      "step": 3533
+    },
+    {
+      "epoch": 0.675717017208413,
+      "grad_norm": 2.823066234588623,
+      "learning_rate": 5e-06,
+      "loss": 0.3365,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6759082217973231,
+      "grad_norm": 1.697743535041809,
+      "learning_rate": 5e-06,
+      "loss": 0.1053,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6760994263862332,
+      "grad_norm": 1.5451792478561401,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6762906309751434,
+      "grad_norm": 2.006969690322876,
+      "learning_rate": 5e-06,
+      "loss": 0.1252,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6764818355640535,
+      "grad_norm": 2.5118765830993652,
+      "learning_rate": 5e-06,
+      "loss": 0.3674,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6766730401529637,
+      "grad_norm": 2.0223960876464844,
+      "learning_rate": 5e-06,
+      "loss": 0.2248,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6768642447418738,
+      "grad_norm": 1.5017013549804688,
+      "learning_rate": 5e-06,
+      "loss": 0.1731,
+      "step": 3540
+    },
+    {
+      "epoch": 0.677055449330784,
+      "grad_norm": 1.1571253538131714,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6772466539196941,
+      "grad_norm": 0.9846636652946472,
+      "learning_rate": 5e-06,
+      "loss": 0.092,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6774378585086042,
+      "grad_norm": 1.6082611083984375,
+      "learning_rate": 5e-06,
+      "loss": 0.0561,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6776290630975144,
+      "grad_norm": 2.4058167934417725,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6778202676864244,
+      "grad_norm": 1.8551628589630127,
+      "learning_rate": 5e-06,
+      "loss": 0.227,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6780114722753346,
+      "grad_norm": 2.5418436527252197,
+      "learning_rate": 5e-06,
+      "loss": 0.4509,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6782026768642447,
+      "grad_norm": 1.8027793169021606,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6783938814531548,
+      "grad_norm": 1.5673744678497314,
+      "learning_rate": 5e-06,
+      "loss": 0.2043,
+      "step": 3548
+    },
+    {
+      "epoch": 0.678585086042065,
+      "grad_norm": 1.2510675191879272,
+      "learning_rate": 5e-06,
+      "loss": 0.0827,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6787762906309751,
+      "grad_norm": 1.4993951320648193,
+      "learning_rate": 5e-06,
+      "loss": 0.0514,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6789674952198853,
+      "grad_norm": 2.15187931060791,
+      "learning_rate": 5e-06,
+      "loss": 0.2552,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6791586998087954,
+      "grad_norm": 2.6083855628967285,
+      "learning_rate": 5e-06,
+      "loss": 0.4038,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6793499043977056,
+      "grad_norm": 2.16745924949646,
+      "learning_rate": 5e-06,
+      "loss": 0.2588,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6795411089866157,
+      "grad_norm": 1.6964585781097412,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6797323135755258,
+      "grad_norm": 2.437232732772827,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 3555
+    },
+    {
+      "epoch": 0.679923518164436,
+      "grad_norm": 1.8153268098831177,
+      "learning_rate": 5e-06,
+      "loss": 0.095,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6801147227533461,
+      "grad_norm": 2.4039676189422607,
+      "learning_rate": 5e-06,
+      "loss": 0.2803,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6803059273422563,
+      "grad_norm": 3.8345534801483154,
+      "learning_rate": 5e-06,
+      "loss": 0.2945,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6804971319311663,
+      "grad_norm": 2.759597063064575,
+      "learning_rate": 5e-06,
+      "loss": 0.2355,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6806883365200764,
+      "grad_norm": 1.6006584167480469,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6808795411089866,
+      "grad_norm": 1.7173149585723877,
+      "learning_rate": 5e-06,
+      "loss": 0.1152,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6810707456978967,
+      "grad_norm": 1.0151087045669556,
+      "learning_rate": 5e-06,
+      "loss": 0.0712,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6812619502868069,
+      "grad_norm": 1.4937494993209839,
+      "learning_rate": 5e-06,
+      "loss": 0.0642,
+      "step": 3563
+    },
+    {
+      "epoch": 0.681453154875717,
+      "grad_norm": 1.3477442264556885,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6816443594646272,
+      "grad_norm": 1.9342656135559082,
+      "learning_rate": 5e-06,
+      "loss": 0.4498,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6818355640535373,
+      "grad_norm": 2.319324254989624,
+      "learning_rate": 5e-06,
+      "loss": 0.3148,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6820267686424474,
+      "grad_norm": 2.8179759979248047,
+      "learning_rate": 5e-06,
+      "loss": 0.362,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6822179732313576,
+      "grad_norm": 2.6950948238372803,
+      "learning_rate": 5e-06,
+      "loss": 0.0819,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6824091778202677,
+      "grad_norm": 2.1264123916625977,
+      "learning_rate": 5e-06,
+      "loss": 0.1911,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6826003824091779,
+      "grad_norm": 2.7540876865386963,
+      "learning_rate": 5e-06,
+      "loss": 0.6599,
+      "step": 3570
+    },
+    {
+      "epoch": 0.682791586998088,
+      "grad_norm": 2.1270487308502197,
+      "learning_rate": 5e-06,
+      "loss": 0.3011,
+      "step": 3571
+    },
+    {
+      "epoch": 0.682982791586998,
+      "grad_norm": 2.540781021118164,
+      "learning_rate": 5e-06,
+      "loss": 0.4374,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6831739961759082,
+      "grad_norm": 1.4416691064834595,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6833652007648183,
+      "grad_norm": 1.5677932500839233,
+      "learning_rate": 5e-06,
+      "loss": 0.0783,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6835564053537285,
+      "grad_norm": 2.3644917011260986,
+      "learning_rate": 5e-06,
+      "loss": 0.0778,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6837476099426386,
+      "grad_norm": 2.1362829208374023,
+      "learning_rate": 5e-06,
+      "loss": 0.2909,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6839388145315488,
+      "grad_norm": 3.296583890914917,
+      "learning_rate": 5e-06,
+      "loss": 0.4958,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6841300191204589,
+      "grad_norm": 1.5384232997894287,
+      "learning_rate": 5e-06,
+      "loss": 0.1698,
+      "step": 3578
+    },
+    {
+      "epoch": 0.684321223709369,
+      "grad_norm": 2.5473268032073975,
+      "learning_rate": 5e-06,
+      "loss": 0.2099,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6845124282982792,
+      "grad_norm": 1.863061547279358,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6847036328871893,
+      "grad_norm": 2.5463814735412598,
+      "learning_rate": 5e-06,
+      "loss": 0.0986,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6848948374760995,
+      "grad_norm": 1.9547677040100098,
+      "learning_rate": 5e-06,
+      "loss": 0.2291,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6850860420650096,
+      "grad_norm": 1.2805721759796143,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6852772466539196,
+      "grad_norm": 2.379103183746338,
+      "learning_rate": 5e-06,
+      "loss": 0.3174,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6854684512428298,
+      "grad_norm": 1.308780312538147,
+      "learning_rate": 5e-06,
+      "loss": 0.1225,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6856596558317399,
+      "grad_norm": 2.074080228805542,
+      "learning_rate": 5e-06,
+      "loss": 0.1967,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6858508604206501,
+      "grad_norm": 1.9603567123413086,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6860420650095602,
+      "grad_norm": 1.7865123748779297,
+      "learning_rate": 5e-06,
+      "loss": 0.1054,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6862332695984703,
+      "grad_norm": 3.6250503063201904,
+      "learning_rate": 5e-06,
+      "loss": 0.4375,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6864244741873805,
+      "grad_norm": 2.0481324195861816,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6866156787762906,
+      "grad_norm": 2.7558176517486572,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6868068833652008,
+      "grad_norm": 1.8074321746826172,
+      "learning_rate": 5e-06,
+      "loss": 0.1178,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6869980879541109,
+      "grad_norm": 1.997239351272583,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6871892925430211,
+      "grad_norm": 2.378185987472534,
+      "learning_rate": 5e-06,
+      "loss": 0.3194,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6873804971319312,
+      "grad_norm": 2.6446828842163086,
+      "learning_rate": 5e-06,
+      "loss": 0.4638,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6875717017208413,
+      "grad_norm": 2.4682164192199707,
+      "learning_rate": 5e-06,
+      "loss": 0.3686,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6877629063097515,
+      "grad_norm": 1.9417331218719482,
+      "learning_rate": 5e-06,
+      "loss": 0.2668,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6879541108986615,
+      "grad_norm": 1.9014066457748413,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6881453154875717,
+      "grad_norm": 1.8162672519683838,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6883365200764818,
+      "grad_norm": 1.8843580484390259,
+      "learning_rate": 5e-06,
+      "loss": 0.1089,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6885277246653919,
+      "grad_norm": 1.4576982259750366,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6887189292543021,
+      "grad_norm": 3.355142831802368,
+      "learning_rate": 5e-06,
+      "loss": 0.2791,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6889101338432122,
+      "grad_norm": 1.7258144617080688,
+      "learning_rate": 5e-06,
+      "loss": 0.2368,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6891013384321224,
+      "grad_norm": 1.7668720483779907,
+      "learning_rate": 5e-06,
+      "loss": 0.132,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6892925430210325,
+      "grad_norm": 1.8233599662780762,
+      "learning_rate": 5e-06,
+      "loss": 0.0762,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6894837476099427,
+      "grad_norm": 2.0434274673461914,
+      "learning_rate": 5e-06,
+      "loss": 0.071,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6896749521988528,
+      "grad_norm": 2.575824022293091,
+      "learning_rate": 5e-06,
+      "loss": 0.4385,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6898661567877629,
+      "grad_norm": 3.179419755935669,
+      "learning_rate": 5e-06,
+      "loss": 0.4052,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6900573613766731,
+      "grad_norm": 2.368760347366333,
+      "learning_rate": 5e-06,
+      "loss": 0.4097,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6902485659655831,
+      "grad_norm": 1.6637201309204102,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6904397705544933,
+      "grad_norm": 1.4172734022140503,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6906309751434034,
+      "grad_norm": 2.328533411026001,
+      "learning_rate": 5e-06,
+      "loss": 0.2946,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6908221797323135,
+      "grad_norm": 1.887550711631775,
+      "learning_rate": 5e-06,
+      "loss": 0.1478,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6910133843212237,
+      "grad_norm": 1.8322516679763794,
+      "learning_rate": 5e-06,
+      "loss": 0.1951,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6912045889101338,
+      "grad_norm": 1.0440239906311035,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 3615
+    },
+    {
+      "epoch": 0.691395793499044,
+      "grad_norm": 2.5914018154144287,
+      "learning_rate": 5e-06,
+      "loss": 0.3502,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6915869980879541,
+      "grad_norm": 2.4446136951446533,
+      "learning_rate": 5e-06,
+      "loss": 0.1806,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6917782026768643,
+      "grad_norm": 1.4871900081634521,
+      "learning_rate": 5e-06,
+      "loss": 0.0797,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6919694072657744,
+      "grad_norm": 2.3120055198669434,
+      "learning_rate": 5e-06,
+      "loss": 0.1075,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6921606118546845,
+      "grad_norm": 1.4232051372528076,
+      "learning_rate": 5e-06,
+      "loss": 0.1484,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6923518164435947,
+      "grad_norm": 3.1616592407226562,
+      "learning_rate": 5e-06,
+      "loss": 0.5661,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6925430210325048,
+      "grad_norm": 0.8112595081329346,
+      "learning_rate": 5e-06,
+      "loss": 0.0917,
+      "step": 3622
+    },
+    {
+      "epoch": 0.692734225621415,
+      "grad_norm": 1.0516867637634277,
+      "learning_rate": 5e-06,
+      "loss": 0.0633,
+      "step": 3623
+    },
+    {
+      "epoch": 0.692925430210325,
+      "grad_norm": 1.4513195753097534,
+      "learning_rate": 5e-06,
+      "loss": 0.0825,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6931166347992351,
+      "grad_norm": 2.015878200531006,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6933078393881453,
+      "grad_norm": 2.6140496730804443,
+      "learning_rate": 5e-06,
+      "loss": 0.5673,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6934990439770554,
+      "grad_norm": 2.6710994243621826,
+      "learning_rate": 5e-06,
+      "loss": 0.1457,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6936902485659656,
+      "grad_norm": 1.289347529411316,
+      "learning_rate": 5e-06,
+      "loss": 0.0971,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6938814531548757,
+      "grad_norm": 1.6329401731491089,
+      "learning_rate": 5e-06,
+      "loss": 0.1046,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6940726577437859,
+      "grad_norm": 1.3376904726028442,
+      "learning_rate": 5e-06,
+      "loss": 0.0794,
+      "step": 3630
+    },
+    {
+      "epoch": 0.694263862332696,
+      "grad_norm": 2.8650128841400146,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6944550669216061,
+      "grad_norm": 3.3484082221984863,
+      "learning_rate": 5e-06,
+      "loss": 0.5369,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6946462715105163,
+      "grad_norm": 2.710831880569458,
+      "learning_rate": 5e-06,
+      "loss": 0.5368,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6948374760994264,
+      "grad_norm": 1.1810380220413208,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6950286806883366,
+      "grad_norm": 0.5851054787635803,
+      "learning_rate": 5e-06,
+      "loss": 0.0376,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6952198852772467,
+      "grad_norm": 3.097205400466919,
+      "learning_rate": 5e-06,
+      "loss": 0.2126,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6954110898661567,
+      "grad_norm": 1.8246523141860962,
+      "learning_rate": 5e-06,
+      "loss": 0.1213,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6956022944550669,
+      "grad_norm": 1.682036280632019,
+      "learning_rate": 5e-06,
+      "loss": 0.0875,
+      "step": 3638
+    },
+    {
+      "epoch": 0.695793499043977,
+      "grad_norm": 3.0946667194366455,
+      "learning_rate": 5e-06,
+      "loss": 0.6476,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6959847036328872,
+      "grad_norm": 1.0556777715682983,
+      "learning_rate": 5e-06,
+      "loss": 0.0874,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6961759082217973,
+      "grad_norm": 1.5672385692596436,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6963671128107075,
+      "grad_norm": 1.1081948280334473,
+      "learning_rate": 5e-06,
+      "loss": 0.1039,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6965583173996176,
+      "grad_norm": 3.239100694656372,
+      "learning_rate": 5e-06,
+      "loss": 0.3624,
+      "step": 3643
+    },
+    {
+      "epoch": 0.6967495219885277,
+      "grad_norm": 1.96116304397583,
+      "learning_rate": 5e-06,
+      "loss": 0.1417,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6969407265774379,
+      "grad_norm": 2.3989415168762207,
+      "learning_rate": 5e-06,
+      "loss": 0.3667,
+      "step": 3645
+    },
+    {
+      "epoch": 0.697131931166348,
+      "grad_norm": 1.7304235696792603,
+      "learning_rate": 5e-06,
+      "loss": 0.3046,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6973231357552582,
+      "grad_norm": 1.469014286994934,
+      "learning_rate": 5e-06,
+      "loss": 0.1986,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6975143403441683,
+      "grad_norm": 2.6511895656585693,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6977055449330783,
+      "grad_norm": 4.76706600189209,
+      "learning_rate": 5e-06,
+      "loss": 0.3297,
+      "step": 3649
+    },
+    {
+      "epoch": 0.6978967495219885,
+      "grad_norm": 1.9821535348892212,
+      "learning_rate": 5e-06,
+      "loss": 0.0866,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6980879541108986,
+      "grad_norm": 1.3378146886825562,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6982791586998088,
+      "grad_norm": 1.85330069065094,
+      "learning_rate": 5e-06,
+      "loss": 0.2699,
+      "step": 3652
+    },
+    {
+      "epoch": 0.6984703632887189,
+      "grad_norm": 2.159931182861328,
+      "learning_rate": 5e-06,
+      "loss": 0.3349,
+      "step": 3653
+    },
+    {
+      "epoch": 0.698661567877629,
+      "grad_norm": 2.206817150115967,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6988527724665392,
+      "grad_norm": 2.111424446105957,
+      "learning_rate": 5e-06,
+      "loss": 0.2197,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6990439770554493,
+      "grad_norm": 2.3077478408813477,
+      "learning_rate": 5e-06,
+      "loss": 0.3232,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6992351816443595,
+      "grad_norm": 2.5184547901153564,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6994263862332696,
+      "grad_norm": 2.8431849479675293,
+      "learning_rate": 5e-06,
+      "loss": 0.5781,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6996175908221798,
+      "grad_norm": 2.026881694793701,
+      "learning_rate": 5e-06,
+      "loss": 0.3044,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6998087954110899,
+      "grad_norm": 1.5575225353240967,
+      "learning_rate": 5e-06,
+      "loss": 0.0619,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6749697923660278,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7001912045889102,
+      "grad_norm": 1.1908769607543945,
+      "learning_rate": 5e-06,
+      "loss": 0.0485,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7003824091778202,
+      "grad_norm": 2.0859763622283936,
+      "learning_rate": 5e-06,
+      "loss": 0.2337,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7005736137667304,
+      "grad_norm": 1.7970099449157715,
+      "learning_rate": 5e-06,
+      "loss": 0.2513,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7007648183556405,
+      "grad_norm": 2.7064900398254395,
+      "learning_rate": 5e-06,
+      "loss": 0.3637,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7009560229445506,
+      "grad_norm": 3.3071577548980713,
+      "learning_rate": 5e-06,
+      "loss": 0.3557,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7011472275334608,
+      "grad_norm": 1.6861001253128052,
+      "learning_rate": 5e-06,
+      "loss": 0.1279,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7013384321223709,
+      "grad_norm": 1.3847708702087402,
+      "learning_rate": 5e-06,
+      "loss": 0.0656,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7015296367112811,
+      "grad_norm": 3.5664327144622803,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7017208413001912,
+      "grad_norm": 3.75789213180542,
+      "learning_rate": 5e-06,
+      "loss": 0.5611,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7019120458891014,
+      "grad_norm": 2.001065254211426,
+      "learning_rate": 5e-06,
+      "loss": 0.2467,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7021032504780115,
+      "grad_norm": 2.242720603942871,
+      "learning_rate": 5e-06,
+      "loss": 0.1748,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7022944550669216,
+      "grad_norm": 1.567443609237671,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7024856596558318,
+      "grad_norm": 1.640781044960022,
+      "learning_rate": 5e-06,
+      "loss": 0.0745,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7026768642447419,
+      "grad_norm": 1.6443709135055542,
+      "learning_rate": 5e-06,
+      "loss": 0.1,
+      "step": 3675
+    },
+    {
+      "epoch": 0.702868068833652,
+      "grad_norm": 2.157893419265747,
+      "learning_rate": 5e-06,
+      "loss": 0.2708,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7030592734225621,
+      "grad_norm": 1.9968886375427246,
+      "learning_rate": 5e-06,
+      "loss": 0.1628,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7032504780114722,
+      "grad_norm": 1.2432177066802979,
+      "learning_rate": 5e-06,
+      "loss": 0.0915,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7034416826003824,
+      "grad_norm": 1.2381335496902466,
+      "learning_rate": 5e-06,
+      "loss": 0.1092,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7036328871892925,
+      "grad_norm": 1.1749067306518555,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7038240917782027,
+      "grad_norm": 3.458665370941162,
+      "learning_rate": 5e-06,
+      "loss": 0.1514,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7040152963671128,
+      "grad_norm": 2.4206576347351074,
+      "learning_rate": 5e-06,
+      "loss": 0.2704,
+      "step": 3682
+    },
+    {
+      "epoch": 0.704206500956023,
+      "grad_norm": 1.8550755977630615,
+      "learning_rate": 5e-06,
+      "loss": 0.2404,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7043977055449331,
+      "grad_norm": 1.5968091487884521,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7045889101338432,
+      "grad_norm": 2.3588109016418457,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7047801147227534,
+      "grad_norm": 1.6067357063293457,
+      "learning_rate": 5e-06,
+      "loss": 0.0753,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7049713193116635,
+      "grad_norm": 1.9464937448501587,
+      "learning_rate": 5e-06,
+      "loss": 0.0893,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7051625239005737,
+      "grad_norm": 2.355674982070923,
+      "learning_rate": 5e-06,
+      "loss": 0.1453,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7053537284894837,
+      "grad_norm": 3.20164155960083,
+      "learning_rate": 5e-06,
+      "loss": 0.4407,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7055449330783938,
+      "grad_norm": 2.1521313190460205,
+      "learning_rate": 5e-06,
+      "loss": 0.1517,
+      "step": 3690
+    },
+    {
+      "epoch": 0.705736137667304,
+      "grad_norm": 1.2731857299804688,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7059273422562141,
+      "grad_norm": 1.5699528455734253,
+      "learning_rate": 5e-06,
+      "loss": 0.0717,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7061185468451243,
+      "grad_norm": 1.5166230201721191,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7063097514340344,
+      "grad_norm": 1.5976234674453735,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7065009560229446,
+      "grad_norm": 2.420431137084961,
+      "learning_rate": 5e-06,
+      "loss": 0.3126,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7066921606118547,
+      "grad_norm": 2.7159039974212646,
+      "learning_rate": 5e-06,
+      "loss": 0.5264,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7068833652007648,
+      "grad_norm": 2.4235072135925293,
+      "learning_rate": 5e-06,
+      "loss": 0.4054,
+      "step": 3697
+    },
+    {
+      "epoch": 0.707074569789675,
+      "grad_norm": 1.8437212705612183,
+      "learning_rate": 5e-06,
+      "loss": 0.2117,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7072657743785851,
+      "grad_norm": 1.3941651582717896,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7074569789674953,
+      "grad_norm": 1.5819287300109863,
+      "learning_rate": 5e-06,
+      "loss": 0.1037,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7076481835564054,
+      "grad_norm": 3.036449670791626,
+      "learning_rate": 5e-06,
+      "loss": 0.4601,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7078393881453154,
+      "grad_norm": 1.64691960811615,
+      "learning_rate": 5e-06,
+      "loss": 0.2143,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7080305927342256,
+      "grad_norm": 1.3194284439086914,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7082217973231357,
+      "grad_norm": 1.3173198699951172,
+      "learning_rate": 5e-06,
+      "loss": 0.1449,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7084130019120459,
+      "grad_norm": 2.3822851181030273,
+      "learning_rate": 5e-06,
+      "loss": 0.2857,
+      "step": 3705
+    },
+    {
+      "epoch": 0.708604206500956,
+      "grad_norm": 0.9149222373962402,
+      "learning_rate": 5e-06,
+      "loss": 0.0497,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7087954110898662,
+      "grad_norm": 2.2636375427246094,
+      "learning_rate": 5e-06,
+      "loss": 0.2972,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7089866156787763,
+      "grad_norm": 1.518246054649353,
+      "learning_rate": 5e-06,
+      "loss": 0.1799,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7091778202676864,
+      "grad_norm": 1.2613885402679443,
+      "learning_rate": 5e-06,
+      "loss": 0.1001,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7093690248565966,
+      "grad_norm": 2.623847723007202,
+      "learning_rate": 5e-06,
+      "loss": 0.2827,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7095602294455067,
+      "grad_norm": 1.9074766635894775,
+      "learning_rate": 5e-06,
+      "loss": 0.0718,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7097514340344169,
+      "grad_norm": 2.9987568855285645,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 3712
+    },
+    {
+      "epoch": 0.709942638623327,
+      "grad_norm": 1.0691415071487427,
+      "learning_rate": 5e-06,
+      "loss": 0.0659,
+      "step": 3713
+    },
+    {
+      "epoch": 0.710133843212237,
+      "grad_norm": 3.7600460052490234,
+      "learning_rate": 5e-06,
+      "loss": 0.86,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7103250478011472,
+      "grad_norm": 1.6110256910324097,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7105162523900573,
+      "grad_norm": 2.0216493606567383,
+      "learning_rate": 5e-06,
+      "loss": 0.2925,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7107074569789675,
+      "grad_norm": 1.7469826936721802,
+      "learning_rate": 5e-06,
+      "loss": 0.2517,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7108986615678776,
+      "grad_norm": 1.7995054721832275,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7110898661567877,
+      "grad_norm": 2.31731915473938,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7112810707456979,
+      "grad_norm": 3.077112913131714,
+      "learning_rate": 5e-06,
+      "loss": 0.7406,
+      "step": 3720
+    },
+    {
+      "epoch": 0.711472275334608,
+      "grad_norm": 3.16402530670166,
+      "learning_rate": 5e-06,
+      "loss": 0.465,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7116634799235182,
+      "grad_norm": 2.755120038986206,
+      "learning_rate": 5e-06,
+      "loss": 0.2713,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7118546845124283,
+      "grad_norm": 1.240608811378479,
+      "learning_rate": 5e-06,
+      "loss": 0.0644,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7120458891013385,
+      "grad_norm": 1.4552818536758423,
+      "learning_rate": 5e-06,
+      "loss": 0.1006,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7122370936902486,
+      "grad_norm": 1.733351707458496,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7124282982791587,
+      "grad_norm": 2.0551533699035645,
+      "learning_rate": 5e-06,
+      "loss": 0.3772,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7126195028680689,
+      "grad_norm": 3.2883100509643555,
+      "learning_rate": 5e-06,
+      "loss": 0.5013,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7128107074569789,
+      "grad_norm": 1.076749563217163,
+      "learning_rate": 5e-06,
+      "loss": 0.1152,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7130019120458891,
+      "grad_norm": 2.451246976852417,
+      "learning_rate": 5e-06,
+      "loss": 0.2865,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7131931166347992,
+      "grad_norm": 1.9702866077423096,
+      "learning_rate": 5e-06,
+      "loss": 0.2526,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7133843212237093,
+      "grad_norm": 2.362886667251587,
+      "learning_rate": 5e-06,
+      "loss": 0.1457,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7135755258126195,
+      "grad_norm": 2.858502149581909,
+      "learning_rate": 5e-06,
+      "loss": 0.4665,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7137667304015296,
+      "grad_norm": 1.5832735300064087,
+      "learning_rate": 5e-06,
+      "loss": 0.1224,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7139579349904398,
+      "grad_norm": 1.0978338718414307,
+      "learning_rate": 5e-06,
+      "loss": 0.1259,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7141491395793499,
+      "grad_norm": 2.182143211364746,
+      "learning_rate": 5e-06,
+      "loss": 0.2562,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7143403441682601,
+      "grad_norm": 2.5188608169555664,
+      "learning_rate": 5e-06,
+      "loss": 0.2833,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7145315487571702,
+      "grad_norm": 2.875762701034546,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7147227533460803,
+      "grad_norm": 1.844509243965149,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7149139579349905,
+      "grad_norm": 2.9208295345306396,
+      "learning_rate": 5e-06,
+      "loss": 0.3887,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7151051625239006,
+      "grad_norm": 2.217679977416992,
+      "learning_rate": 5e-06,
+      "loss": 0.2439,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7152963671128107,
+      "grad_norm": 2.5245535373687744,
+      "learning_rate": 5e-06,
+      "loss": 0.3163,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7154875717017208,
+      "grad_norm": 1.9459753036499023,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7156787762906309,
+      "grad_norm": 2.5548386573791504,
+      "learning_rate": 5e-06,
+      "loss": 0.0823,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7158699808795411,
+      "grad_norm": 1.1367732286453247,
+      "learning_rate": 5e-06,
+      "loss": 0.0603,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7160611854684512,
+      "grad_norm": 1.7649719715118408,
+      "learning_rate": 5e-06,
+      "loss": 0.2555,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7162523900573614,
+      "grad_norm": 2.476152181625366,
+      "learning_rate": 5e-06,
+      "loss": 0.2607,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7164435946462715,
+      "grad_norm": 1.505000114440918,
+      "learning_rate": 5e-06,
+      "loss": 0.1498,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7166347992351817,
+      "grad_norm": 1.3874759674072266,
+      "learning_rate": 5e-06,
+      "loss": 0.0955,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7168260038240918,
+      "grad_norm": 1.742844581604004,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7170172084130019,
+      "grad_norm": 2.4118449687957764,
+      "learning_rate": 5e-06,
+      "loss": 0.176,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7172084130019121,
+      "grad_norm": 1.4940874576568604,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7173996175908222,
+      "grad_norm": 1.963732361793518,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7175908221797324,
+      "grad_norm": 1.9962869882583618,
+      "learning_rate": 5e-06,
+      "loss": 0.2072,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7177820267686424,
+      "grad_norm": 1.2074509859085083,
+      "learning_rate": 5e-06,
+      "loss": 0.1099,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7179732313575525,
+      "grad_norm": 1.593335509300232,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7181644359464627,
+      "grad_norm": 1.2521311044692993,
+      "learning_rate": 5e-06,
+      "loss": 0.0686,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7183556405353728,
+      "grad_norm": 2.19175124168396,
+      "learning_rate": 5e-06,
+      "loss": 0.3097,
+      "step": 3757
+    },
+    {
+      "epoch": 0.718546845124283,
+      "grad_norm": 1.7197211980819702,
+      "learning_rate": 5e-06,
+      "loss": 0.1602,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7187380497131931,
+      "grad_norm": 2.827324867248535,
+      "learning_rate": 5e-06,
+      "loss": 0.5046,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7189292543021033,
+      "grad_norm": 2.1176669597625732,
+      "learning_rate": 5e-06,
+      "loss": 0.0945,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7191204588910134,
+      "grad_norm": 1.07952880859375,
+      "learning_rate": 5e-06,
+      "loss": 0.0495,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7193116634799235,
+      "grad_norm": 2.3675689697265625,
+      "learning_rate": 5e-06,
+      "loss": 0.2094,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7195028680688337,
+      "grad_norm": 2.8417303562164307,
+      "learning_rate": 5e-06,
+      "loss": 0.4245,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7196940726577438,
+      "grad_norm": 3.0691707134246826,
+      "learning_rate": 5e-06,
+      "loss": 0.5019,
+      "step": 3764
+    },
+    {
+      "epoch": 0.719885277246654,
+      "grad_norm": 1.418975591659546,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 3765
+    },
+    {
+      "epoch": 0.720076481835564,
+      "grad_norm": 2.2532196044921875,
+      "learning_rate": 5e-06,
+      "loss": 0.1952,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7202676864244741,
+      "grad_norm": 1.7203007936477661,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7204588910133843,
+      "grad_norm": 2.12467885017395,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7206500956022944,
+      "grad_norm": 1.4739885330200195,
+      "learning_rate": 5e-06,
+      "loss": 0.0862,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7208413001912046,
+      "grad_norm": 3.242621898651123,
+      "learning_rate": 5e-06,
+      "loss": 0.6335,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7210325047801147,
+      "grad_norm": 2.5275886058807373,
+      "learning_rate": 5e-06,
+      "loss": 0.4305,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7212237093690248,
+      "grad_norm": 1.7171810865402222,
+      "learning_rate": 5e-06,
+      "loss": 0.1955,
+      "step": 3772
+    },
+    {
+      "epoch": 0.721414913957935,
+      "grad_norm": 1.231096625328064,
+      "learning_rate": 5e-06,
+      "loss": 0.092,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7216061185468451,
+      "grad_norm": 1.7485922574996948,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7217973231357553,
+      "grad_norm": 2.117539882659912,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7219885277246654,
+      "grad_norm": 3.3793411254882812,
+      "learning_rate": 5e-06,
+      "loss": 0.5776,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7221797323135756,
+      "grad_norm": 1.7460490465164185,
+      "learning_rate": 5e-06,
+      "loss": 0.1292,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7223709369024857,
+      "grad_norm": 2.388984441757202,
+      "learning_rate": 5e-06,
+      "loss": 0.3883,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7225621414913957,
+      "grad_norm": 2.096761703491211,
+      "learning_rate": 5e-06,
+      "loss": 0.1509,
+      "step": 3779
+    },
+    {
+      "epoch": 0.722753346080306,
+      "grad_norm": 1.783759593963623,
+      "learning_rate": 5e-06,
+      "loss": 0.2602,
+      "step": 3780
+    },
+    {
+      "epoch": 0.722944550669216,
+      "grad_norm": 1.4529781341552734,
+      "learning_rate": 5e-06,
+      "loss": 0.0799,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7231357552581262,
+      "grad_norm": 2.474672317504883,
+      "learning_rate": 5e-06,
+      "loss": 0.2724,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7233269598470363,
+      "grad_norm": 2.139212131500244,
+      "learning_rate": 5e-06,
+      "loss": 0.236,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7235181644359464,
+      "grad_norm": 2.131075859069824,
+      "learning_rate": 5e-06,
+      "loss": 0.3092,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7237093690248566,
+      "grad_norm": 1.3767306804656982,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7239005736137667,
+      "grad_norm": 1.0231573581695557,
+      "learning_rate": 5e-06,
+      "loss": 0.0584,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7240917782026769,
+      "grad_norm": 1.745872139930725,
+      "learning_rate": 5e-06,
+      "loss": 0.132,
+      "step": 3787
+    },
+    {
+      "epoch": 0.724282982791587,
+      "grad_norm": 2.5864417552948,
+      "learning_rate": 5e-06,
+      "loss": 0.4019,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7244741873804972,
+      "grad_norm": 1.4365683794021606,
+      "learning_rate": 5e-06,
+      "loss": 0.089,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7246653919694073,
+      "grad_norm": 2.468015193939209,
+      "learning_rate": 5e-06,
+      "loss": 0.3741,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7248565965583174,
+      "grad_norm": 3.0470778942108154,
+      "learning_rate": 5e-06,
+      "loss": 0.4877,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7250478011472276,
+      "grad_norm": 1.820553183555603,
+      "learning_rate": 5e-06,
+      "loss": 0.1056,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7252390057361376,
+      "grad_norm": 1.5439107418060303,
+      "learning_rate": 5e-06,
+      "loss": 0.0992,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7254302103250478,
+      "grad_norm": 2.6746387481689453,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7256214149139579,
+      "grad_norm": 2.123269557952881,
+      "learning_rate": 5e-06,
+      "loss": 0.2376,
+      "step": 3795
+    },
+    {
+      "epoch": 0.725812619502868,
+      "grad_norm": 1.2966785430908203,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7260038240917782,
+      "grad_norm": 1.3482623100280762,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7261950286806883,
+      "grad_norm": 1.8395178318023682,
+      "learning_rate": 5e-06,
+      "loss": 0.2555,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7263862332695985,
+      "grad_norm": 1.2841962575912476,
+      "learning_rate": 5e-06,
+      "loss": 0.0857,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7265774378585086,
+      "grad_norm": 2.664163827896118,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7267686424474188,
+      "grad_norm": 2.564634323120117,
+      "learning_rate": 5e-06,
+      "loss": 0.4485,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7269598470363289,
+      "grad_norm": 2.6707558631896973,
+      "learning_rate": 5e-06,
+      "loss": 0.4055,
+      "step": 3802
+    },
+    {
+      "epoch": 0.727151051625239,
+      "grad_norm": 1.4090027809143066,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7273422562141492,
+      "grad_norm": 1.7918213605880737,
+      "learning_rate": 5e-06,
+      "loss": 0.1137,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7275334608030593,
+      "grad_norm": 1.4990943670272827,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7277246653919694,
+      "grad_norm": 2.7204136848449707,
+      "learning_rate": 5e-06,
+      "loss": 0.1131,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7279158699808795,
+      "grad_norm": 2.5874335765838623,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7281070745697896,
+      "grad_norm": 2.7016761302948,
+      "learning_rate": 5e-06,
+      "loss": 0.3025,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7282982791586998,
+      "grad_norm": 1.1666982173919678,
+      "learning_rate": 5e-06,
+      "loss": 0.1359,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7284894837476099,
+      "grad_norm": 1.7149765491485596,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7286806883365201,
+      "grad_norm": 2.1318461894989014,
+      "learning_rate": 5e-06,
+      "loss": 0.1363,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7288718929254302,
+      "grad_norm": 1.7442179918289185,
+      "learning_rate": 5e-06,
+      "loss": 0.1217,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7290630975143404,
+      "grad_norm": 2.2777628898620605,
+      "learning_rate": 5e-06,
+      "loss": 0.1651,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7292543021032505,
+      "grad_norm": 6.601057052612305,
+      "learning_rate": 5e-06,
+      "loss": 0.5386,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7294455066921606,
+      "grad_norm": 1.6855504512786865,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7296367112810708,
+      "grad_norm": 1.998500943183899,
+      "learning_rate": 5e-06,
+      "loss": 0.2681,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7298279158699809,
+      "grad_norm": 1.0050654411315918,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7300191204588911,
+      "grad_norm": 2.510183811187744,
+      "learning_rate": 5e-06,
+      "loss": 0.2464,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7302103250478011,
+      "grad_norm": 1.5420417785644531,
+      "learning_rate": 5e-06,
+      "loss": 0.0933,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7304015296367112,
+      "grad_norm": 2.9417569637298584,
+      "learning_rate": 5e-06,
+      "loss": 0.5287,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7305927342256214,
+      "grad_norm": 2.2762291431427,
+      "learning_rate": 5e-06,
+      "loss": 0.2584,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7307839388145315,
+      "grad_norm": 3.564465045928955,
+      "learning_rate": 5e-06,
+      "loss": 0.2454,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7309751434034417,
+      "grad_norm": 1.3632512092590332,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7311663479923518,
+      "grad_norm": 2.418616771697998,
+      "learning_rate": 5e-06,
+      "loss": 0.1328,
+      "step": 3824
+    },
+    {
+      "epoch": 0.731357552581262,
+      "grad_norm": 1.0035635232925415,
+      "learning_rate": 5e-06,
+      "loss": 0.0504,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7315487571701721,
+      "grad_norm": 2.8982577323913574,
+      "learning_rate": 5e-06,
+      "loss": 0.5397,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7317399617590822,
+      "grad_norm": 2.096374273300171,
+      "learning_rate": 5e-06,
+      "loss": 0.3442,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7319311663479924,
+      "grad_norm": 2.2000863552093506,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7321223709369025,
+      "grad_norm": 2.5556576251983643,
+      "learning_rate": 5e-06,
+      "loss": 0.2146,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7323135755258127,
+      "grad_norm": 1.023587942123413,
+      "learning_rate": 5e-06,
+      "loss": 0.0599,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7325047801147228,
+      "grad_norm": 2.018570899963379,
+      "learning_rate": 5e-06,
+      "loss": 0.1337,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7326959847036328,
+      "grad_norm": 1.6819159984588623,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 3832
+    },
+    {
+      "epoch": 0.732887189292543,
+      "grad_norm": 10.953661918640137,
+      "learning_rate": 5e-06,
+      "loss": 0.1749,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7330783938814531,
+      "grad_norm": 2.3553245067596436,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7332695984703633,
+      "grad_norm": 3.415651798248291,
+      "learning_rate": 5e-06,
+      "loss": 0.6602,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7334608030592734,
+      "grad_norm": 1.051983118057251,
+      "learning_rate": 5e-06,
+      "loss": 0.0589,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7336520076481835,
+      "grad_norm": 1.977689504623413,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7338432122370937,
+      "grad_norm": 2.170513391494751,
+      "learning_rate": 5e-06,
+      "loss": 0.2598,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7340344168260038,
+      "grad_norm": 2.0110361576080322,
+      "learning_rate": 5e-06,
+      "loss": 0.3082,
+      "step": 3839
+    },
+    {
+      "epoch": 0.734225621414914,
+      "grad_norm": 1.711444616317749,
+      "learning_rate": 5e-06,
+      "loss": 0.159,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7344168260038241,
+      "grad_norm": 1.1422127485275269,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7346080305927343,
+      "grad_norm": 1.628873586654663,
+      "learning_rate": 5e-06,
+      "loss": 0.0867,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7347992351816444,
+      "grad_norm": 2.754671335220337,
+      "learning_rate": 5e-06,
+      "loss": 0.2112,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7349904397705544,
+      "grad_norm": 2.050431728363037,
+      "learning_rate": 5e-06,
+      "loss": 0.2173,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7351816443594646,
+      "grad_norm": 1.6508783102035522,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7353728489483747,
+      "grad_norm": 1.9832112789154053,
+      "learning_rate": 5e-06,
+      "loss": 0.2658,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7355640535372849,
+      "grad_norm": 1.4163705110549927,
+      "learning_rate": 5e-06,
+      "loss": 0.0685,
+      "step": 3847
+    },
+    {
+      "epoch": 0.735755258126195,
+      "grad_norm": 2.1025025844573975,
+      "learning_rate": 5e-06,
+      "loss": 0.0973,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7359464627151051,
+      "grad_norm": 1.321691632270813,
+      "learning_rate": 5e-06,
+      "loss": 0.0648,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7361376673040153,
+      "grad_norm": 1.1205233335494995,
+      "learning_rate": 5e-06,
+      "loss": 0.0491,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7363288718929254,
+      "grad_norm": 1.8860313892364502,
+      "learning_rate": 5e-06,
+      "loss": 0.246,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7365200764818356,
+      "grad_norm": 2.046706438064575,
+      "learning_rate": 5e-06,
+      "loss": 0.3516,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7367112810707457,
+      "grad_norm": 1.3845429420471191,
+      "learning_rate": 5e-06,
+      "loss": 0.115,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7369024856596559,
+      "grad_norm": 1.796648383140564,
+      "learning_rate": 5e-06,
+      "loss": 0.1873,
+      "step": 3854
+    },
+    {
+      "epoch": 0.737093690248566,
+      "grad_norm": 1.5792717933654785,
+      "learning_rate": 5e-06,
+      "loss": 0.2649,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7372848948374761,
+      "grad_norm": 1.4286028146743774,
+      "learning_rate": 5e-06,
+      "loss": 0.0831,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7374760994263863,
+      "grad_norm": 1.319799780845642,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7376673040152963,
+      "grad_norm": 1.7097197771072388,
+      "learning_rate": 5e-06,
+      "loss": 0.1823,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7378585086042065,
+      "grad_norm": 2.7568390369415283,
+      "learning_rate": 5e-06,
+      "loss": 0.1829,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7380497131931166,
+      "grad_norm": 2.267392873764038,
+      "learning_rate": 5e-06,
+      "loss": 0.3423,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7382409177820267,
+      "grad_norm": 1.5888205766677856,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 3861
+    },
+    {
+      "epoch": 0.7384321223709369,
+      "grad_norm": 0.9303539991378784,
+      "learning_rate": 5e-06,
+      "loss": 0.0497,
+      "step": 3862
+    },
+    {
+      "epoch": 0.738623326959847,
+      "grad_norm": 3.3895652294158936,
+      "learning_rate": 5e-06,
+      "loss": 0.1808,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7388145315487572,
+      "grad_norm": 3.354686975479126,
+      "learning_rate": 5e-06,
+      "loss": 0.3795,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7390057361376673,
+      "grad_norm": 2.8291988372802734,
+      "learning_rate": 5e-06,
+      "loss": 0.1875,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7391969407265775,
+      "grad_norm": 1.4242098331451416,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7393881453154876,
+      "grad_norm": 0.8613669276237488,
+      "learning_rate": 5e-06,
+      "loss": 0.0874,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7395793499043977,
+      "grad_norm": 1.4547524452209473,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7397705544933079,
+      "grad_norm": 1.1557694673538208,
+      "learning_rate": 5e-06,
+      "loss": 0.0494,
+      "step": 3869
+    },
+    {
+      "epoch": 0.739961759082218,
+      "grad_norm": 2.8671276569366455,
+      "learning_rate": 5e-06,
+      "loss": 0.3717,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7401529636711282,
+      "grad_norm": 1.4811517000198364,
+      "learning_rate": 5e-06,
+      "loss": 0.1741,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7403441682600382,
+      "grad_norm": 1.3373982906341553,
+      "learning_rate": 5e-06,
+      "loss": 0.1384,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7405353728489483,
+      "grad_norm": 0.8785936832427979,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7407265774378585,
+      "grad_norm": 0.9371583461761475,
+      "learning_rate": 5e-06,
+      "loss": 0.0445,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7409177820267686,
+      "grad_norm": 1.681098222732544,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7411089866156788,
+      "grad_norm": 2.265328884124756,
+      "learning_rate": 5e-06,
+      "loss": 0.4473,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7413001912045889,
+      "grad_norm": 1.790353536605835,
+      "learning_rate": 5e-06,
+      "loss": 0.0827,
+      "step": 3877
+    },
+    {
+      "epoch": 0.7414913957934991,
+      "grad_norm": 2.781074047088623,
+      "learning_rate": 5e-06,
+      "loss": 0.3687,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7416826003824092,
+      "grad_norm": 1.5163137912750244,
+      "learning_rate": 5e-06,
+      "loss": 0.0667,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7418738049713193,
+      "grad_norm": 1.391534686088562,
+      "learning_rate": 5e-06,
+      "loss": 0.0774,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7420650095602295,
+      "grad_norm": 1.9660178422927856,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7422562141491396,
+      "grad_norm": 1.3986618518829346,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7424474187380498,
+      "grad_norm": 2.422389507293701,
+      "learning_rate": 5e-06,
+      "loss": 0.3266,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7426386233269598,
+      "grad_norm": 1.2864700555801392,
+      "learning_rate": 5e-06,
+      "loss": 0.2191,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7428298279158699,
+      "grad_norm": 1.7133296728134155,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7430210325047801,
+      "grad_norm": 2.243800163269043,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7432122370936902,
+      "grad_norm": 2.1287524700164795,
+      "learning_rate": 5e-06,
+      "loss": 0.3237,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7434034416826004,
+      "grad_norm": 3.481367588043213,
+      "learning_rate": 5e-06,
+      "loss": 0.5053,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7435946462715105,
+      "grad_norm": 2.098524332046509,
+      "learning_rate": 5e-06,
+      "loss": 0.2776,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7437858508604207,
+      "grad_norm": 0.9846514463424683,
+      "learning_rate": 5e-06,
+      "loss": 0.0791,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7439770554493308,
+      "grad_norm": 1.8553352355957031,
+      "learning_rate": 5e-06,
+      "loss": 0.2831,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7441682600382409,
+      "grad_norm": 2.703209638595581,
+      "learning_rate": 5e-06,
+      "loss": 0.1956,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7443594646271511,
+      "grad_norm": 0.9680504202842712,
+      "learning_rate": 5e-06,
+      "loss": 0.0764,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7445506692160612,
+      "grad_norm": 1.8443779945373535,
+      "learning_rate": 5e-06,
+      "loss": 0.2195,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7447418738049714,
+      "grad_norm": 2.6461219787597656,
+      "learning_rate": 5e-06,
+      "loss": 0.4909,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7449330783938815,
+      "grad_norm": 1.3912622928619385,
+      "learning_rate": 5e-06,
+      "loss": 0.1268,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7451242829827915,
+      "grad_norm": 1.7276599407196045,
+      "learning_rate": 5e-06,
+      "loss": 0.226,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7453154875717017,
+      "grad_norm": 2.662609815597534,
+      "learning_rate": 5e-06,
+      "loss": 0.1942,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7455066921606118,
+      "grad_norm": 1.6807820796966553,
+      "learning_rate": 5e-06,
+      "loss": 0.1768,
+      "step": 3899
+    },
+    {
+      "epoch": 0.745697896749522,
+      "grad_norm": 1.3572360277175903,
+      "learning_rate": 5e-06,
+      "loss": 0.0603,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7458891013384321,
+      "grad_norm": 1.2281373739242554,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7460803059273422,
+      "grad_norm": 1.9232803583145142,
+      "learning_rate": 5e-06,
+      "loss": 0.2239,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7462715105162524,
+      "grad_norm": 2.2120046615600586,
+      "learning_rate": 5e-06,
+      "loss": 0.2476,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7464627151051625,
+      "grad_norm": 2.8236958980560303,
+      "learning_rate": 5e-06,
+      "loss": 0.5103,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7466539196940727,
+      "grad_norm": 2.782604694366455,
+      "learning_rate": 5e-06,
+      "loss": 0.3232,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7468451242829828,
+      "grad_norm": 1.3832602500915527,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 3906
+    },
+    {
+      "epoch": 0.747036328871893,
+      "grad_norm": 3.4751503467559814,
+      "learning_rate": 5e-06,
+      "loss": 0.3804,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7472275334608031,
+      "grad_norm": 1.4347681999206543,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7474187380497131,
+      "grad_norm": 1.5916887521743774,
+      "learning_rate": 5e-06,
+      "loss": 0.2899,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7476099426386233,
+      "grad_norm": 1.5356292724609375,
+      "learning_rate": 5e-06,
+      "loss": 0.2278,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7478011472275334,
+      "grad_norm": 1.4358553886413574,
+      "learning_rate": 5e-06,
+      "loss": 0.0868,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7479923518164436,
+      "grad_norm": 2.5048086643218994,
+      "learning_rate": 5e-06,
+      "loss": 0.227,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7481835564053537,
+      "grad_norm": 1.7658495903015137,
+      "learning_rate": 5e-06,
+      "loss": 0.1685,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7483747609942638,
+      "grad_norm": 2.723329782485962,
+      "learning_rate": 5e-06,
+      "loss": 0.3353,
+      "step": 3914
+    },
+    {
+      "epoch": 0.748565965583174,
+      "grad_norm": 2.999716281890869,
+      "learning_rate": 5e-06,
+      "loss": 0.3282,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7487571701720841,
+      "grad_norm": 3.776826858520508,
+      "learning_rate": 5e-06,
+      "loss": 0.2084,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7489483747609943,
+      "grad_norm": 1.5918197631835938,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7491395793499044,
+      "grad_norm": 1.6553537845611572,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7493307839388146,
+      "grad_norm": 1.9501301050186157,
+      "learning_rate": 5e-06,
+      "loss": 0.1137,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7495219885277247,
+      "grad_norm": 3.1001572608947754,
+      "learning_rate": 5e-06,
+      "loss": 0.7937,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7497131931166348,
+      "grad_norm": 1.8700470924377441,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 3921
+    },
+    {
+      "epoch": 0.749904397705545,
+      "grad_norm": 1.2993032932281494,
+      "learning_rate": 5e-06,
+      "loss": 0.1429,
+      "step": 3922
+    },
+    {
+      "epoch": 0.750095602294455,
+      "grad_norm": 2.387141466140747,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7502868068833652,
+      "grad_norm": 1.0744256973266602,
+      "learning_rate": 5e-06,
+      "loss": 0.0685,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7504780114722753,
+      "grad_norm": 3.1105291843414307,
+      "learning_rate": 5e-06,
+      "loss": 0.2699,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7506692160611854,
+      "grad_norm": 3.3172290325164795,
+      "learning_rate": 5e-06,
+      "loss": 0.6928,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7508604206500956,
+      "grad_norm": 1.4145816564559937,
+      "learning_rate": 5e-06,
+      "loss": 0.1114,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7510516252390057,
+      "grad_norm": 2.017582893371582,
+      "learning_rate": 5e-06,
+      "loss": 0.3379,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7512428298279159,
+      "grad_norm": 2.438030481338501,
+      "learning_rate": 5e-06,
+      "loss": 0.273,
+      "step": 3929
+    },
+    {
+      "epoch": 0.751434034416826,
+      "grad_norm": 2.810269832611084,
+      "learning_rate": 5e-06,
+      "loss": 0.1784,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7516252390057362,
+      "grad_norm": 2.189603328704834,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7518164435946463,
+      "grad_norm": 1.3899329900741577,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7520076481835564,
+      "grad_norm": 2.109938144683838,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7521988527724666,
+      "grad_norm": 2.4114205837249756,
+      "learning_rate": 5e-06,
+      "loss": 0.3787,
+      "step": 3934
+    },
+    {
+      "epoch": 0.7523900573613767,
+      "grad_norm": 1.5396156311035156,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7525812619502869,
+      "grad_norm": 1.966928482055664,
+      "learning_rate": 5e-06,
+      "loss": 0.2534,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7527724665391969,
+      "grad_norm": 1.3008625507354736,
+      "learning_rate": 5e-06,
+      "loss": 0.0839,
+      "step": 3937
+    },
+    {
+      "epoch": 0.752963671128107,
+      "grad_norm": 3.0656418800354004,
+      "learning_rate": 5e-06,
+      "loss": 0.3491,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7531548757170172,
+      "grad_norm": 2.63382887840271,
+      "learning_rate": 5e-06,
+      "loss": 0.3588,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7533460803059273,
+      "grad_norm": 1.6612306833267212,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7535372848948375,
+      "grad_norm": 2.234178304672241,
+      "learning_rate": 5e-06,
+      "loss": 0.1589,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7537284894837476,
+      "grad_norm": 3.02970814704895,
+      "learning_rate": 5e-06,
+      "loss": 0.4124,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7539196940726578,
+      "grad_norm": 2.506859540939331,
+      "learning_rate": 5e-06,
+      "loss": 0.2223,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7541108986615679,
+      "grad_norm": 1.3613660335540771,
+      "learning_rate": 5e-06,
+      "loss": 0.1406,
+      "step": 3944
+    },
+    {
+      "epoch": 0.754302103250478,
+      "grad_norm": 2.071458578109741,
+      "learning_rate": 5e-06,
+      "loss": 0.2519,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7544933078393882,
+      "grad_norm": 1.1499487161636353,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7546845124282983,
+      "grad_norm": 2.7436764240264893,
+      "learning_rate": 5e-06,
+      "loss": 0.1509,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7548757170172085,
+      "grad_norm": 2.140192985534668,
+      "learning_rate": 5e-06,
+      "loss": 0.2268,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7550669216061185,
+      "grad_norm": 1.5796760320663452,
+      "learning_rate": 5e-06,
+      "loss": 0.113,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7552581261950286,
+      "grad_norm": 1.0383970737457275,
+      "learning_rate": 5e-06,
+      "loss": 0.0439,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7554493307839388,
+      "grad_norm": 2.50980281829834,
+      "learning_rate": 5e-06,
+      "loss": 0.4777,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7556405353728489,
+      "grad_norm": 1.7971466779708862,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7558317399617591,
+      "grad_norm": 1.9320497512817383,
+      "learning_rate": 5e-06,
+      "loss": 0.1991,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7560229445506692,
+      "grad_norm": 1.246254324913025,
+      "learning_rate": 5e-06,
+      "loss": 0.0755,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7562141491395793,
+      "grad_norm": 1.7969180345535278,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7564053537284895,
+      "grad_norm": 1.317395567893982,
+      "learning_rate": 5e-06,
+      "loss": 0.0401,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7565965583173996,
+      "grad_norm": 1.8792531490325928,
+      "learning_rate": 5e-06,
+      "loss": 0.308,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7567877629063098,
+      "grad_norm": 1.601319670677185,
+      "learning_rate": 5e-06,
+      "loss": 0.1683,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7569789674952199,
+      "grad_norm": 1.1691665649414062,
+      "learning_rate": 5e-06,
+      "loss": 0.0809,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7571701720841301,
+      "grad_norm": 1.8878777027130127,
+      "learning_rate": 5e-06,
+      "loss": 0.1181,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7573613766730402,
+      "grad_norm": 1.6922614574432373,
+      "learning_rate": 5e-06,
+      "loss": 0.1016,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7575525812619502,
+      "grad_norm": 2.844416379928589,
+      "learning_rate": 5e-06,
+      "loss": 0.1851,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7577437858508604,
+      "grad_norm": 2.265589952468872,
+      "learning_rate": 5e-06,
+      "loss": 0.2272,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7579349904397705,
+      "grad_norm": 2.574376344680786,
+      "learning_rate": 5e-06,
+      "loss": 0.4007,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7581261950286807,
+      "grad_norm": 2.0190422534942627,
+      "learning_rate": 5e-06,
+      "loss": 0.2131,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7583173996175908,
+      "grad_norm": 1.53504478931427,
+      "learning_rate": 5e-06,
+      "loss": 0.1167,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7585086042065009,
+      "grad_norm": 1.4824988842010498,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7586998087954111,
+      "grad_norm": 2.1277225017547607,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7588910133843212,
+      "grad_norm": 0.9543938636779785,
+      "learning_rate": 5e-06,
+      "loss": 0.0364,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7590822179732314,
+      "grad_norm": 3.176586151123047,
+      "learning_rate": 5e-06,
+      "loss": 0.5473,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7592734225621415,
+      "grad_norm": 1.5333970785140991,
+      "learning_rate": 5e-06,
+      "loss": 0.1819,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7594646271510517,
+      "grad_norm": 1.2855944633483887,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7596558317399618,
+      "grad_norm": 1.5741631984710693,
+      "learning_rate": 5e-06,
+      "loss": 0.0947,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7598470363288718,
+      "grad_norm": 1.533542513847351,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 3974
+    },
+    {
+      "epoch": 0.760038240917782,
+      "grad_norm": 2.86518931388855,
+      "learning_rate": 5e-06,
+      "loss": 0.2103,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7602294455066921,
+      "grad_norm": 2.7985222339630127,
+      "learning_rate": 5e-06,
+      "loss": 0.2571,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7604206500956023,
+      "grad_norm": 2.5995070934295654,
+      "learning_rate": 5e-06,
+      "loss": 0.4758,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7606118546845124,
+      "grad_norm": 1.5301446914672852,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7608030592734225,
+      "grad_norm": 0.876580536365509,
+      "learning_rate": 5e-06,
+      "loss": 0.0367,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7609942638623327,
+      "grad_norm": 1.7435474395751953,
+      "learning_rate": 5e-06,
+      "loss": 0.1279,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7611854684512428,
+      "grad_norm": 1.7620370388031006,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 3981
+    },
+    {
+      "epoch": 0.761376673040153,
+      "grad_norm": 1.7652002573013306,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 3982
+    },
+    {
+      "epoch": 0.7615678776290631,
+      "grad_norm": 3.072636127471924,
+      "learning_rate": 5e-06,
+      "loss": 0.4287,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7617590822179733,
+      "grad_norm": 2.880697727203369,
+      "learning_rate": 5e-06,
+      "loss": 0.3342,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7619502868068834,
+      "grad_norm": 1.7110553979873657,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7621414913957935,
+      "grad_norm": 1.7476038932800293,
+      "learning_rate": 5e-06,
+      "loss": 0.0955,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7623326959847037,
+      "grad_norm": 2.2482314109802246,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7625239005736137,
+      "grad_norm": 1.754885196685791,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7627151051625239,
+      "grad_norm": 2.0199167728424072,
+      "learning_rate": 5e-06,
+      "loss": 0.2003,
+      "step": 3989
+    },
+    {
+      "epoch": 0.762906309751434,
+      "grad_norm": 2.1370720863342285,
+      "learning_rate": 5e-06,
+      "loss": 0.2559,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7630975143403441,
+      "grad_norm": 1.4818934202194214,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7632887189292543,
+      "grad_norm": 2.2542731761932373,
+      "learning_rate": 5e-06,
+      "loss": 0.1821,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7634799235181644,
+      "grad_norm": 2.6181223392486572,
+      "learning_rate": 5e-06,
+      "loss": 0.2197,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7636711281070746,
+      "grad_norm": 2.795283317565918,
+      "learning_rate": 5e-06,
+      "loss": 0.2661,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7638623326959847,
+      "grad_norm": 4.159665107727051,
+      "learning_rate": 5e-06,
+      "loss": 0.8165,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7640535372848949,
+      "grad_norm": 1.8209826946258545,
+      "learning_rate": 5e-06,
+      "loss": 0.221,
+      "step": 3996
+    },
+    {
+      "epoch": 0.764244741873805,
+      "grad_norm": 1.490784764289856,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7644359464627151,
+      "grad_norm": 1.8983914852142334,
+      "learning_rate": 5e-06,
+      "loss": 0.2311,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7646271510516253,
+      "grad_norm": 2.7288661003112793,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7648183556405354,
+      "grad_norm": 1.5600886344909668,
+      "learning_rate": 5e-06,
+      "loss": 0.0708,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7648183556405354,
+      "eval_runtime": 831.7494,
+      "eval_samples_per_second": 1.844,
+      "eval_steps_per_second": 0.231,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7650095602294456,
+      "grad_norm": 1.5232917070388794,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7652007648183556,
+      "grad_norm": 1.4118472337722778,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7653919694072657,
+      "grad_norm": 2.754512310028076,
+      "learning_rate": 5e-06,
+      "loss": 0.3549,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7655831739961759,
+      "grad_norm": 1.406946063041687,
+      "learning_rate": 5e-06,
+      "loss": 0.136,
+      "step": 4004
+    },
+    {
+      "epoch": 0.765774378585086,
+      "grad_norm": 2.2281954288482666,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7659655831739962,
+      "grad_norm": 1.819032907485962,
+      "learning_rate": 5e-06,
+      "loss": 0.1981,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7661567877629063,
+      "grad_norm": 2.1007330417633057,
+      "learning_rate": 5e-06,
+      "loss": 0.3524,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7663479923518165,
+      "grad_norm": 2.4521172046661377,
+      "learning_rate": 5e-06,
+      "loss": 0.4039,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7665391969407266,
+      "grad_norm": 2.598778247833252,
+      "learning_rate": 5e-06,
+      "loss": 0.2918,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7667304015296367,
+      "grad_norm": 2.3652114868164062,
+      "learning_rate": 5e-06,
+      "loss": 0.2555,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7669216061185469,
+      "grad_norm": 1.7355620861053467,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 4011
+    },
+    {
+      "epoch": 0.767112810707457,
+      "grad_norm": 3.2190678119659424,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7673040152963672,
+      "grad_norm": 2.945258617401123,
+      "learning_rate": 5e-06,
+      "loss": 0.261,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7674952198852772,
+      "grad_norm": 4.270470142364502,
+      "learning_rate": 5e-06,
+      "loss": 0.7201,
+      "step": 4014
+    },
+    {
+      "epoch": 0.7676864244741873,
+      "grad_norm": 1.7441529035568237,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7678776290630975,
+      "grad_norm": 1.0526031255722046,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7680688336520076,
+      "grad_norm": 2.0954020023345947,
+      "learning_rate": 5e-06,
+      "loss": 0.2734,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7682600382409178,
+      "grad_norm": 1.8673603534698486,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7684512428298279,
+      "grad_norm": 1.6339843273162842,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 4019
+    },
+    {
+      "epoch": 0.768642447418738,
+      "grad_norm": 1.8186370134353638,
+      "learning_rate": 5e-06,
+      "loss": 0.2133,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7688336520076482,
+      "grad_norm": 2.424779176712036,
+      "learning_rate": 5e-06,
+      "loss": 0.2374,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7690248565965583,
+      "grad_norm": 1.2409683465957642,
+      "learning_rate": 5e-06,
+      "loss": 0.0816,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7692160611854685,
+      "grad_norm": 2.5901854038238525,
+      "learning_rate": 5e-06,
+      "loss": 0.134,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7694072657743786,
+      "grad_norm": 1.4344134330749512,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7695984703632888,
+      "grad_norm": 1.9438978433609009,
+      "learning_rate": 5e-06,
+      "loss": 0.0898,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7697896749521989,
+      "grad_norm": 3.3850362300872803,
+      "learning_rate": 5e-06,
+      "loss": 0.6324,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7699808795411089,
+      "grad_norm": 1.7402697801589966,
+      "learning_rate": 5e-06,
+      "loss": 0.2097,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7701720841300191,
+      "grad_norm": 1.9264459609985352,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7703632887189292,
+      "grad_norm": 1.2523527145385742,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7705544933078394,
+      "grad_norm": 2.009114980697632,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7707456978967495,
+      "grad_norm": 1.3284012079238892,
+      "learning_rate": 5e-06,
+      "loss": 0.0669,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7709369024856596,
+      "grad_norm": 3.460033893585205,
+      "learning_rate": 5e-06,
+      "loss": 0.653,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7711281070745698,
+      "grad_norm": 1.1250536441802979,
+      "learning_rate": 5e-06,
+      "loss": 0.1061,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7713193116634799,
+      "grad_norm": 2.0778872966766357,
+      "learning_rate": 5e-06,
+      "loss": 0.4068,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7715105162523901,
+      "grad_norm": 2.448514461517334,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7717017208413002,
+      "grad_norm": 1.0364723205566406,
+      "learning_rate": 5e-06,
+      "loss": 0.0746,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7718929254302104,
+      "grad_norm": 4.103033065795898,
+      "learning_rate": 5e-06,
+      "loss": 0.3191,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7720841300191205,
+      "grad_norm": 2.208747148513794,
+      "learning_rate": 5e-06,
+      "loss": 0.2741,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7722753346080306,
+      "grad_norm": 3.062417984008789,
+      "learning_rate": 5e-06,
+      "loss": 0.6122,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7724665391969407,
+      "grad_norm": 1.5257611274719238,
+      "learning_rate": 5e-06,
+      "loss": 0.1512,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7726577437858508,
+      "grad_norm": 1.7130473852157593,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 4041
+    },
+    {
+      "epoch": 0.772848948374761,
+      "grad_norm": 2.802297353744507,
+      "learning_rate": 5e-06,
+      "loss": 0.0933,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7730401529636711,
+      "grad_norm": 1.7665363550186157,
+      "learning_rate": 5e-06,
+      "loss": 0.1321,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7732313575525812,
+      "grad_norm": 1.023848533630371,
+      "learning_rate": 5e-06,
+      "loss": 0.0549,
+      "step": 4044
+    },
+    {
+      "epoch": 0.7734225621414914,
+      "grad_norm": 2.230396270751953,
+      "learning_rate": 5e-06,
+      "loss": 0.3784,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7736137667304015,
+      "grad_norm": 1.3800081014633179,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7738049713193117,
+      "grad_norm": 2.2208352088928223,
+      "learning_rate": 5e-06,
+      "loss": 0.2697,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7739961759082218,
+      "grad_norm": 2.5575742721557617,
+      "learning_rate": 5e-06,
+      "loss": 0.2564,
+      "step": 4048
+    },
+    {
+      "epoch": 0.774187380497132,
+      "grad_norm": 1.4058862924575806,
+      "learning_rate": 5e-06,
+      "loss": 0.0901,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7743785850860421,
+      "grad_norm": 2.812058687210083,
+      "learning_rate": 5e-06,
+      "loss": 0.1837,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7745697896749522,
+      "grad_norm": 2.573519706726074,
+      "learning_rate": 5e-06,
+      "loss": 0.483,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7747609942638624,
+      "grad_norm": 3.118896961212158,
+      "learning_rate": 5e-06,
+      "loss": 0.3858,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7749521988527724,
+      "grad_norm": 1.3650435209274292,
+      "learning_rate": 5e-06,
+      "loss": 0.0917,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7751434034416826,
+      "grad_norm": 3.2142133712768555,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7753346080305927,
+      "grad_norm": 1.4311920404434204,
+      "learning_rate": 5e-06,
+      "loss": 0.095,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7755258126195028,
+      "grad_norm": 1.5923914909362793,
+      "learning_rate": 5e-06,
+      "loss": 0.0758,
+      "step": 4056
+    },
+    {
+      "epoch": 0.775717017208413,
+      "grad_norm": 1.7863245010375977,
+      "learning_rate": 5e-06,
+      "loss": 0.1589,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7759082217973231,
+      "grad_norm": 2.2520720958709717,
+      "learning_rate": 5e-06,
+      "loss": 0.2503,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7760994263862333,
+      "grad_norm": 3.211054563522339,
+      "learning_rate": 5e-06,
+      "loss": 0.6649,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7762906309751434,
+      "grad_norm": 0.8572031855583191,
+      "learning_rate": 5e-06,
+      "loss": 0.0593,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7764818355640536,
+      "grad_norm": 1.4232878684997559,
+      "learning_rate": 5e-06,
+      "loss": 0.0751,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7766730401529637,
+      "grad_norm": 1.295220971107483,
+      "learning_rate": 5e-06,
+      "loss": 0.0767,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7768642447418738,
+      "grad_norm": 2.2189230918884277,
+      "learning_rate": 5e-06,
+      "loss": 0.1053,
+      "step": 4063
+    },
+    {
+      "epoch": 0.777055449330784,
+      "grad_norm": 2.4356725215911865,
+      "learning_rate": 5e-06,
+      "loss": 0.3199,
+      "step": 4064
+    },
+    {
+      "epoch": 0.777246653919694,
+      "grad_norm": 1.6857635974884033,
+      "learning_rate": 5e-06,
+      "loss": 0.1233,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7774378585086043,
+      "grad_norm": 2.9625089168548584,
+      "learning_rate": 5e-06,
+      "loss": 0.3591,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7776290630975143,
+      "grad_norm": 1.5815578699111938,
+      "learning_rate": 5e-06,
+      "loss": 0.0995,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7778202676864244,
+      "grad_norm": 1.3760156631469727,
+      "learning_rate": 5e-06,
+      "loss": 0.1136,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7780114722753346,
+      "grad_norm": 3.329010486602783,
+      "learning_rate": 5e-06,
+      "loss": 0.167,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7782026768642447,
+      "grad_norm": 2.279073715209961,
+      "learning_rate": 5e-06,
+      "loss": 0.3059,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7783938814531549,
+      "grad_norm": 1.5212337970733643,
+      "learning_rate": 5e-06,
+      "loss": 0.1047,
+      "step": 4071
+    },
+    {
+      "epoch": 0.778585086042065,
+      "grad_norm": 1.8684823513031006,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7787762906309752,
+      "grad_norm": 1.487801432609558,
+      "learning_rate": 5e-06,
+      "loss": 0.0738,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7789674952198853,
+      "grad_norm": 2.3601162433624268,
+      "learning_rate": 5e-06,
+      "loss": 0.0753,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7791586998087954,
+      "grad_norm": 1.5325002670288086,
+      "learning_rate": 5e-06,
+      "loss": 0.0589,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7793499043977056,
+      "grad_norm": 2.376420497894287,
+      "learning_rate": 5e-06,
+      "loss": 0.3178,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7795411089866157,
+      "grad_norm": 1.3300342559814453,
+      "learning_rate": 5e-06,
+      "loss": 0.1267,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7797323135755259,
+      "grad_norm": 1.2323758602142334,
+      "learning_rate": 5e-06,
+      "loss": 0.1096,
+      "step": 4078
+    },
+    {
+      "epoch": 0.779923518164436,
+      "grad_norm": 2.7862613201141357,
+      "learning_rate": 5e-06,
+      "loss": 0.4678,
+      "step": 4079
+    },
+    {
+      "epoch": 0.780114722753346,
+      "grad_norm": 1.2939293384552002,
+      "learning_rate": 5e-06,
+      "loss": 0.0784,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7803059273422562,
+      "grad_norm": 2.031770944595337,
+      "learning_rate": 5e-06,
+      "loss": 0.076,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7804971319311663,
+      "grad_norm": 1.987504243850708,
+      "learning_rate": 5e-06,
+      "loss": 0.2646,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7806883365200765,
+      "grad_norm": 3.281132936477661,
+      "learning_rate": 5e-06,
+      "loss": 0.4443,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7808795411089866,
+      "grad_norm": 1.6435940265655518,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7810707456978967,
+      "grad_norm": 2.0128467082977295,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7812619502868069,
+      "grad_norm": 1.966070532798767,
+      "learning_rate": 5e-06,
+      "loss": 0.1004,
+      "step": 4086
+    },
+    {
+      "epoch": 0.781453154875717,
+      "grad_norm": 2.714599370956421,
+      "learning_rate": 5e-06,
+      "loss": 0.1898,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7816443594646272,
+      "grad_norm": 2.605646848678589,
+      "learning_rate": 5e-06,
+      "loss": 0.5377,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7818355640535373,
+      "grad_norm": 2.5793659687042236,
+      "learning_rate": 5e-06,
+      "loss": 0.3054,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7820267686424475,
+      "grad_norm": 1.0576157569885254,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7822179732313576,
+      "grad_norm": 2.084843397140503,
+      "learning_rate": 5e-06,
+      "loss": 0.1685,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7824091778202676,
+      "grad_norm": 1.4531314373016357,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7826003824091778,
+      "grad_norm": 1.2267307043075562,
+      "learning_rate": 5e-06,
+      "loss": 0.0672,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7827915869980879,
+      "grad_norm": 2.5020968914031982,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7829827915869981,
+      "grad_norm": 2.856722354888916,
+      "learning_rate": 5e-06,
+      "loss": 0.6113,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7831739961759082,
+      "grad_norm": 3.0301995277404785,
+      "learning_rate": 5e-06,
+      "loss": 0.3728,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7833652007648183,
+      "grad_norm": 2.254077911376953,
+      "learning_rate": 5e-06,
+      "loss": 0.3342,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7835564053537285,
+      "grad_norm": 1.5700175762176514,
+      "learning_rate": 5e-06,
+      "loss": 0.0846,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7837476099426386,
+      "grad_norm": 1.795477032661438,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7839388145315488,
+      "grad_norm": 3.384065866470337,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7841300191204589,
+      "grad_norm": 1.4682124853134155,
+      "learning_rate": 5e-06,
+      "loss": 0.185,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7843212237093691,
+      "grad_norm": 1.558057188987732,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7845124282982792,
+      "grad_norm": 1.5690349340438843,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7847036328871893,
+      "grad_norm": 1.0313045978546143,
+      "learning_rate": 5e-06,
+      "loss": 0.0571,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7848948374760994,
+      "grad_norm": 1.4557856321334839,
+      "learning_rate": 5e-06,
+      "loss": 0.0803,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7850860420650095,
+      "grad_norm": 1.7461309432983398,
+      "learning_rate": 5e-06,
+      "loss": 0.0582,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7852772466539197,
+      "grad_norm": 3.133112668991089,
+      "learning_rate": 5e-06,
+      "loss": 0.4884,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7854684512428298,
+      "grad_norm": 2.038489580154419,
+      "learning_rate": 5e-06,
+      "loss": 0.2294,
+      "step": 4108
+    },
+    {
+      "epoch": 0.7856596558317399,
+      "grad_norm": 1.714369535446167,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7858508604206501,
+      "grad_norm": 2.173279047012329,
+      "learning_rate": 5e-06,
+      "loss": 0.1365,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7860420650095602,
+      "grad_norm": 1.4787670373916626,
+      "learning_rate": 5e-06,
+      "loss": 0.0929,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7862332695984704,
+      "grad_norm": 2.0035345554351807,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7864244741873805,
+      "grad_norm": 2.027458906173706,
+      "learning_rate": 5e-06,
+      "loss": 0.3061,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7866156787762907,
+      "grad_norm": 3.2767152786254883,
+      "learning_rate": 5e-06,
+      "loss": 0.372,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7868068833652008,
+      "grad_norm": 2.2546582221984863,
+      "learning_rate": 5e-06,
+      "loss": 0.2883,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7869980879541109,
+      "grad_norm": 1.9276899099349976,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7871892925430211,
+      "grad_norm": 1.4487714767456055,
+      "learning_rate": 5e-06,
+      "loss": 0.1152,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7873804971319311,
+      "grad_norm": 0.9173452854156494,
+      "learning_rate": 5e-06,
+      "loss": 0.0181,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7875717017208413,
+      "grad_norm": 2.420795440673828,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7877629063097514,
+      "grad_norm": 1.8025798797607422,
+      "learning_rate": 5e-06,
+      "loss": 0.2264,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7879541108986615,
+      "grad_norm": 2.618238925933838,
+      "learning_rate": 5e-06,
+      "loss": 0.4072,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7881453154875717,
+      "grad_norm": 2.343017101287842,
+      "learning_rate": 5e-06,
+      "loss": 0.195,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7883365200764818,
+      "grad_norm": 1.8181397914886475,
+      "learning_rate": 5e-06,
+      "loss": 0.0863,
+      "step": 4123
+    },
+    {
+      "epoch": 0.788527724665392,
+      "grad_norm": 1.8634696006774902,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7887189292543021,
+      "grad_norm": 2.9761476516723633,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7889101338432123,
+      "grad_norm": 2.6176702976226807,
+      "learning_rate": 5e-06,
+      "loss": 0.3955,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7891013384321224,
+      "grad_norm": 2.2697925567626953,
+      "learning_rate": 5e-06,
+      "loss": 0.3743,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7892925430210325,
+      "grad_norm": 2.294552803039551,
+      "learning_rate": 5e-06,
+      "loss": 0.2497,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7894837476099427,
+      "grad_norm": 3.0859687328338623,
+      "learning_rate": 5e-06,
+      "loss": 0.3342,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7896749521988528,
+      "grad_norm": 3.8342316150665283,
+      "learning_rate": 5e-06,
+      "loss": 0.2451,
+      "step": 4130
+    },
+    {
+      "epoch": 0.789866156787763,
+      "grad_norm": 1.9949150085449219,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 4131
+    },
+    {
+      "epoch": 0.790057361376673,
+      "grad_norm": 1.5915290117263794,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7902485659655831,
+      "grad_norm": 2.4300689697265625,
+      "learning_rate": 5e-06,
+      "loss": 0.3243,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7904397705544933,
+      "grad_norm": 2.693277597427368,
+      "learning_rate": 5e-06,
+      "loss": 0.2611,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7906309751434034,
+      "grad_norm": 1.841639518737793,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7908221797323136,
+      "grad_norm": 3.0559420585632324,
+      "learning_rate": 5e-06,
+      "loss": 0.25,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7910133843212237,
+      "grad_norm": 2.3945229053497314,
+      "learning_rate": 5e-06,
+      "loss": 0.1167,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7912045889101338,
+      "grad_norm": 1.6890507936477661,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 4138
+    },
+    {
+      "epoch": 0.791395793499044,
+      "grad_norm": 4.068086624145508,
+      "learning_rate": 5e-06,
+      "loss": 0.4148,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7915869980879541,
+      "grad_norm": 1.1503350734710693,
+      "learning_rate": 5e-06,
+      "loss": 0.1057,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7917782026768643,
+      "grad_norm": 1.6826403141021729,
+      "learning_rate": 5e-06,
+      "loss": 0.2092,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7919694072657744,
+      "grad_norm": 2.374329090118408,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7921606118546846,
+      "grad_norm": 1.413367509841919,
+      "learning_rate": 5e-06,
+      "loss": 0.0968,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7923518164435946,
+      "grad_norm": 0.6716412305831909,
+      "learning_rate": 5e-06,
+      "loss": 0.0294,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7925430210325047,
+      "grad_norm": 1.7345741987228394,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7927342256214149,
+      "grad_norm": 2.5227432250976562,
+      "learning_rate": 5e-06,
+      "loss": 0.2367,
+      "step": 4146
+    },
+    {
+      "epoch": 0.792925430210325,
+      "grad_norm": 1.6562448740005493,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7931166347992352,
+      "grad_norm": 1.9978289604187012,
+      "learning_rate": 5e-06,
+      "loss": 0.0856,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7933078393881453,
+      "grad_norm": 1.263630747795105,
+      "learning_rate": 5e-06,
+      "loss": 0.0879,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7934990439770554,
+      "grad_norm": 1.8136980533599854,
+      "learning_rate": 5e-06,
+      "loss": 0.1034,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7936902485659656,
+      "grad_norm": 2.311706781387329,
+      "learning_rate": 5e-06,
+      "loss": 0.1771,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7938814531548757,
+      "grad_norm": 2.5121023654937744,
+      "learning_rate": 5e-06,
+      "loss": 0.2529,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7940726577437859,
+      "grad_norm": 2.3066329956054688,
+      "learning_rate": 5e-06,
+      "loss": 0.3744,
+      "step": 4153
+    },
+    {
+      "epoch": 0.794263862332696,
+      "grad_norm": 1.4853479862213135,
+      "learning_rate": 5e-06,
+      "loss": 0.1752,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7944550669216062,
+      "grad_norm": 1.425007939338684,
+      "learning_rate": 5e-06,
+      "loss": 0.0867,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7946462715105163,
+      "grad_norm": 2.0261852741241455,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7948374760994263,
+      "grad_norm": 2.156608819961548,
+      "learning_rate": 5e-06,
+      "loss": 0.3367,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7950286806883365,
+      "grad_norm": 1.3627783060073853,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 4158
+    },
+    {
+      "epoch": 0.7952198852772466,
+      "grad_norm": 2.0210812091827393,
+      "learning_rate": 5e-06,
+      "loss": 0.2219,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7954110898661568,
+      "grad_norm": 1.251874327659607,
+      "learning_rate": 5e-06,
+      "loss": 0.0726,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7956022944550669,
+      "grad_norm": 1.1710023880004883,
+      "learning_rate": 5e-06,
+      "loss": 0.0736,
+      "step": 4161
+    },
+    {
+      "epoch": 0.795793499043977,
+      "grad_norm": 1.9900565147399902,
+      "learning_rate": 5e-06,
+      "loss": 0.2058,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7959847036328872,
+      "grad_norm": 1.1456722021102905,
+      "learning_rate": 5e-06,
+      "loss": 0.0785,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7961759082217973,
+      "grad_norm": 2.3891546726226807,
+      "learning_rate": 5e-06,
+      "loss": 0.3063,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7963671128107075,
+      "grad_norm": 1.7114723920822144,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7965583173996176,
+      "grad_norm": 3.918834686279297,
+      "learning_rate": 5e-06,
+      "loss": 0.4596,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7967495219885278,
+      "grad_norm": 1.0303666591644287,
+      "learning_rate": 5e-06,
+      "loss": 0.0694,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7969407265774379,
+      "grad_norm": 1.4666823148727417,
+      "learning_rate": 5e-06,
+      "loss": 0.0555,
+      "step": 4168
+    },
+    {
+      "epoch": 0.797131931166348,
+      "grad_norm": 1.776019811630249,
+      "learning_rate": 5e-06,
+      "loss": 0.1054,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7973231357552581,
+      "grad_norm": 3.0186281204223633,
+      "learning_rate": 5e-06,
+      "loss": 0.4206,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7975143403441682,
+      "grad_norm": 1.698807716369629,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7977055449330784,
+      "grad_norm": 3.590824604034424,
+      "learning_rate": 5e-06,
+      "loss": 0.5016,
+      "step": 4172
+    },
+    {
+      "epoch": 0.7978967495219885,
+      "grad_norm": 1.6603933572769165,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 4173
+    },
+    {
+      "epoch": 0.7980879541108986,
+      "grad_norm": 2.0738401412963867,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 4174
+    },
+    {
+      "epoch": 0.7982791586998088,
+      "grad_norm": 1.8535012006759644,
+      "learning_rate": 5e-06,
+      "loss": 0.1028,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7984703632887189,
+      "grad_norm": 2.4131641387939453,
+      "learning_rate": 5e-06,
+      "loss": 0.3586,
+      "step": 4176
+    },
+    {
+      "epoch": 0.7986615678776291,
+      "grad_norm": 3.5233068466186523,
+      "learning_rate": 5e-06,
+      "loss": 0.6267,
+      "step": 4177
+    },
+    {
+      "epoch": 0.7988527724665392,
+      "grad_norm": 1.7418206930160522,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 4178
+    },
+    {
+      "epoch": 0.7990439770554494,
+      "grad_norm": 1.813206672668457,
+      "learning_rate": 5e-06,
+      "loss": 0.1225,
+      "step": 4179
+    },
+    {
+      "epoch": 0.7992351816443595,
+      "grad_norm": 1.5349483489990234,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7994263862332696,
+      "grad_norm": 1.7856202125549316,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 4181
+    },
+    {
+      "epoch": 0.7996175908221798,
+      "grad_norm": 1.9936387538909912,
+      "learning_rate": 5e-06,
+      "loss": 0.2355,
+      "step": 4182
+    },
+    {
+      "epoch": 0.7998087954110898,
+      "grad_norm": 1.6209287643432617,
+      "learning_rate": 5e-06,
+      "loss": 0.1824,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2457656860351562,
+      "learning_rate": 5e-06,
+      "loss": 0.3357,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8001912045889101,
+      "grad_norm": 2.640838861465454,
+      "learning_rate": 5e-06,
+      "loss": 0.3455,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8003824091778202,
+      "grad_norm": 2.3119945526123047,
+      "learning_rate": 5e-06,
+      "loss": 0.2474,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8005736137667304,
+      "grad_norm": 3.9085891246795654,
+      "learning_rate": 5e-06,
+      "loss": 0.3996,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8007648183556405,
+      "grad_norm": 2.8617188930511475,
+      "learning_rate": 5e-06,
+      "loss": 0.4351,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8009560229445507,
+      "grad_norm": 0.908932089805603,
+      "learning_rate": 5e-06,
+      "loss": 0.0602,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8011472275334608,
+      "grad_norm": 1.0495219230651855,
+      "learning_rate": 5e-06,
+      "loss": 0.0867,
+      "step": 4190
+    },
+    {
+      "epoch": 0.801338432122371,
+      "grad_norm": 2.866431951522827,
+      "learning_rate": 5e-06,
+      "loss": 0.4275,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8015296367112811,
+      "grad_norm": 1.4292815923690796,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8017208413001912,
+      "grad_norm": 1.7318116426467896,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8019120458891014,
+      "grad_norm": 1.6380189657211304,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8021032504780115,
+      "grad_norm": 2.122971534729004,
+      "learning_rate": 5e-06,
+      "loss": 0.2761,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8022944550669217,
+      "grad_norm": 2.2278707027435303,
+      "learning_rate": 5e-06,
+      "loss": 0.1824,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8024856596558317,
+      "grad_norm": 1.6206939220428467,
+      "learning_rate": 5e-06,
+      "loss": 0.2697,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8026768642447418,
+      "grad_norm": 2.017160654067993,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 4198
+    },
+    {
+      "epoch": 0.802868068833652,
+      "grad_norm": 2.2551474571228027,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8030592734225621,
+      "grad_norm": 0.9970536231994629,
+      "learning_rate": 5e-06,
+      "loss": 0.0528,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8032504780114723,
+      "grad_norm": 1.8154901266098022,
+      "learning_rate": 5e-06,
+      "loss": 0.2272,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8034416826003824,
+      "grad_norm": 3.230262041091919,
+      "learning_rate": 5e-06,
+      "loss": 0.5354,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8036328871892925,
+      "grad_norm": 2.7765953540802,
+      "learning_rate": 5e-06,
+      "loss": 0.247,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8038240917782027,
+      "grad_norm": 1.6847764253616333,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8040152963671128,
+      "grad_norm": 1.6412423849105835,
+      "learning_rate": 5e-06,
+      "loss": 0.2218,
+      "step": 4205
+    },
+    {
+      "epoch": 0.804206500956023,
+      "grad_norm": 1.1843851804733276,
+      "learning_rate": 5e-06,
+      "loss": 0.0497,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8043977055449331,
+      "grad_norm": 2.4135520458221436,
+      "learning_rate": 5e-06,
+      "loss": 0.3082,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8045889101338433,
+      "grad_norm": 2.1240572929382324,
+      "learning_rate": 5e-06,
+      "loss": 0.3888,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8047801147227533,
+      "grad_norm": 1.9531352519989014,
+      "learning_rate": 5e-06,
+      "loss": 0.183,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8049713193116634,
+      "grad_norm": 1.7559449672698975,
+      "learning_rate": 5e-06,
+      "loss": 0.1177,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8051625239005736,
+      "grad_norm": 1.8385118246078491,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8053537284894837,
+      "grad_norm": 2.0431954860687256,
+      "learning_rate": 5e-06,
+      "loss": 0.1699,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8055449330783939,
+      "grad_norm": 1.8413811922073364,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 4213
+    },
+    {
+      "epoch": 0.805736137667304,
+      "grad_norm": 2.166095018386841,
+      "learning_rate": 5e-06,
+      "loss": 0.2953,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8059273422562141,
+      "grad_norm": 1.8633686304092407,
+      "learning_rate": 5e-06,
+      "loss": 0.1473,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8061185468451243,
+      "grad_norm": 6.3489861488342285,
+      "learning_rate": 5e-06,
+      "loss": 0.417,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8063097514340344,
+      "grad_norm": 1.1961588859558105,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8065009560229446,
+      "grad_norm": 2.456892728805542,
+      "learning_rate": 5e-06,
+      "loss": 0.2194,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8066921606118547,
+      "grad_norm": 1.9836076498031616,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8068833652007649,
+      "grad_norm": 2.320347785949707,
+      "learning_rate": 5e-06,
+      "loss": 0.3102,
+      "step": 4220
+    },
+    {
+      "epoch": 0.807074569789675,
+      "grad_norm": 1.4314194917678833,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 4221
+    },
+    {
+      "epoch": 0.807265774378585,
+      "grad_norm": 2.4354302883148193,
+      "learning_rate": 5e-06,
+      "loss": 0.3198,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8074569789674952,
+      "grad_norm": 3.135202407836914,
+      "learning_rate": 5e-06,
+      "loss": 0.2219,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8076481835564053,
+      "grad_norm": 1.7042864561080933,
+      "learning_rate": 5e-06,
+      "loss": 0.1439,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8078393881453155,
+      "grad_norm": 2.076150417327881,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8080305927342256,
+      "grad_norm": 2.117445707321167,
+      "learning_rate": 5e-06,
+      "loss": 0.223,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8082217973231357,
+      "grad_norm": 2.044095993041992,
+      "learning_rate": 5e-06,
+      "loss": 0.2822,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8084130019120459,
+      "grad_norm": 2.046536684036255,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 4228
+    },
+    {
+      "epoch": 0.808604206500956,
+      "grad_norm": 1.4895386695861816,
+      "learning_rate": 5e-06,
+      "loss": 0.0805,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8087954110898662,
+      "grad_norm": 0.9864174723625183,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8089866156787763,
+      "grad_norm": 1.7758212089538574,
+      "learning_rate": 5e-06,
+      "loss": 0.2118,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8091778202676865,
+      "grad_norm": 2.2796030044555664,
+      "learning_rate": 5e-06,
+      "loss": 0.2636,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8093690248565966,
+      "grad_norm": 1.5463277101516724,
+      "learning_rate": 5e-06,
+      "loss": 0.1616,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8095602294455067,
+      "grad_norm": 1.383233904838562,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8097514340344169,
+      "grad_norm": 1.3084925413131714,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8099426386233269,
+      "grad_norm": 1.4625215530395508,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8101338432122371,
+      "grad_norm": 1.1085083484649658,
+      "learning_rate": 5e-06,
+      "loss": 0.0566,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8103250478011472,
+      "grad_norm": 2.045025110244751,
+      "learning_rate": 5e-06,
+      "loss": 0.1817,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8105162523900573,
+      "grad_norm": 2.7750654220581055,
+      "learning_rate": 5e-06,
+      "loss": 0.3498,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8107074569789675,
+      "grad_norm": 1.906060814857483,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8108986615678776,
+      "grad_norm": 3.07199764251709,
+      "learning_rate": 5e-06,
+      "loss": 0.3699,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8110898661567878,
+      "grad_norm": 1.987126350402832,
+      "learning_rate": 5e-06,
+      "loss": 0.2444,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8112810707456979,
+      "grad_norm": 1.435240387916565,
+      "learning_rate": 5e-06,
+      "loss": 0.0708,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8114722753346081,
+      "grad_norm": 1.3612838983535767,
+      "learning_rate": 5e-06,
+      "loss": 0.0594,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8116634799235182,
+      "grad_norm": 2.586632013320923,
+      "learning_rate": 5e-06,
+      "loss": 0.3922,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8118546845124283,
+      "grad_norm": 1.353735327720642,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8120458891013385,
+      "grad_norm": 1.7560728788375854,
+      "learning_rate": 5e-06,
+      "loss": 0.2315,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8122370936902485,
+      "grad_norm": 1.8247698545455933,
+      "learning_rate": 5e-06,
+      "loss": 0.1895,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8124282982791587,
+      "grad_norm": 1.327373743057251,
+      "learning_rate": 5e-06,
+      "loss": 0.0454,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8126195028680688,
+      "grad_norm": 1.8750287294387817,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8128107074569789,
+      "grad_norm": 2.295496702194214,
+      "learning_rate": 5e-06,
+      "loss": 0.2928,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8130019120458891,
+      "grad_norm": 2.422713041305542,
+      "learning_rate": 5e-06,
+      "loss": 0.283,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8131931166347992,
+      "grad_norm": 1.8091152906417847,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8133843212237094,
+      "grad_norm": 1.2801040410995483,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8135755258126195,
+      "grad_norm": 1.0623581409454346,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8137667304015297,
+      "grad_norm": 2.4953532218933105,
+      "learning_rate": 5e-06,
+      "loss": 0.1855,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8139579349904398,
+      "grad_norm": 2.282299757003784,
+      "learning_rate": 5e-06,
+      "loss": 0.4767,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8141491395793499,
+      "grad_norm": 1.9749537706375122,
+      "learning_rate": 5e-06,
+      "loss": 0.1855,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8143403441682601,
+      "grad_norm": 1.3995250463485718,
+      "learning_rate": 5e-06,
+      "loss": 0.1177,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8145315487571702,
+      "grad_norm": 2.201573133468628,
+      "learning_rate": 5e-06,
+      "loss": 0.1921,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8147227533460804,
+      "grad_norm": 1.34676194190979,
+      "learning_rate": 5e-06,
+      "loss": 0.0499,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8149139579349904,
+      "grad_norm": 1.8636780977249146,
+      "learning_rate": 5e-06,
+      "loss": 0.2137,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8151051625239005,
+      "grad_norm": 1.9577280282974243,
+      "learning_rate": 5e-06,
+      "loss": 0.3015,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8152963671128107,
+      "grad_norm": 2.576096296310425,
+      "learning_rate": 5e-06,
+      "loss": 0.2557,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8154875717017208,
+      "grad_norm": 1.2772091627120972,
+      "learning_rate": 5e-06,
+      "loss": 0.1053,
+      "step": 4265
+    },
+    {
+      "epoch": 0.815678776290631,
+      "grad_norm": 1.6483020782470703,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8158699808795411,
+      "grad_norm": 2.198559522628784,
+      "learning_rate": 5e-06,
+      "loss": 0.1151,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8160611854684512,
+      "grad_norm": 0.5627172589302063,
+      "learning_rate": 5e-06,
+      "loss": 0.021,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8162523900573614,
+      "grad_norm": 1.8419286012649536,
+      "learning_rate": 5e-06,
+      "loss": 0.1135,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8164435946462715,
+      "grad_norm": 2.3597888946533203,
+      "learning_rate": 5e-06,
+      "loss": 0.5168,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8166347992351817,
+      "grad_norm": 1.2093944549560547,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8168260038240918,
+      "grad_norm": 2.0463786125183105,
+      "learning_rate": 5e-06,
+      "loss": 0.2085,
+      "step": 4272
+    },
+    {
+      "epoch": 0.817017208413002,
+      "grad_norm": 2.7457947731018066,
+      "learning_rate": 5e-06,
+      "loss": 0.246,
+      "step": 4273
+    },
+    {
+      "epoch": 0.817208413001912,
+      "grad_norm": 1.0748988389968872,
+      "learning_rate": 5e-06,
+      "loss": 0.0544,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8173996175908221,
+      "grad_norm": 3.576910972595215,
+      "learning_rate": 5e-06,
+      "loss": 0.3724,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8175908221797323,
+      "grad_norm": 2.98368763923645,
+      "learning_rate": 5e-06,
+      "loss": 0.5464,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8177820267686424,
+      "grad_norm": 2.850346565246582,
+      "learning_rate": 5e-06,
+      "loss": 0.4558,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8179732313575526,
+      "grad_norm": 2.513153553009033,
+      "learning_rate": 5e-06,
+      "loss": 0.2441,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8181644359464627,
+      "grad_norm": 0.7250884175300598,
+      "learning_rate": 5e-06,
+      "loss": 0.0649,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8183556405353728,
+      "grad_norm": 1.4739652872085571,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 4280
+    },
+    {
+      "epoch": 0.818546845124283,
+      "grad_norm": 2.511843681335449,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8187380497131931,
+      "grad_norm": 2.0406453609466553,
+      "learning_rate": 5e-06,
+      "loss": 0.3064,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8189292543021033,
+      "grad_norm": 2.142224073410034,
+      "learning_rate": 5e-06,
+      "loss": 0.2985,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8191204588910134,
+      "grad_norm": 1.9565352201461792,
+      "learning_rate": 5e-06,
+      "loss": 0.2398,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8193116634799236,
+      "grad_norm": 1.1692537069320679,
+      "learning_rate": 5e-06,
+      "loss": 0.1201,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8195028680688337,
+      "grad_norm": 0.9735251665115356,
+      "learning_rate": 5e-06,
+      "loss": 0.0589,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8196940726577437,
+      "grad_norm": 1.0305509567260742,
+      "learning_rate": 5e-06,
+      "loss": 0.0485,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8198852772466539,
+      "grad_norm": 2.014246940612793,
+      "learning_rate": 5e-06,
+      "loss": 0.2368,
+      "step": 4288
+    },
+    {
+      "epoch": 0.820076481835564,
+      "grad_norm": 3.3924779891967773,
+      "learning_rate": 5e-06,
+      "loss": 0.381,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8202676864244742,
+      "grad_norm": 2.973250389099121,
+      "learning_rate": 5e-06,
+      "loss": 0.4036,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8204588910133843,
+      "grad_norm": 1.6813417673110962,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8206500956022944,
+      "grad_norm": 1.9835141897201538,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8208413001912046,
+      "grad_norm": 2.34795880317688,
+      "learning_rate": 5e-06,
+      "loss": 0.2726,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8210325047801147,
+      "grad_norm": 2.353971004486084,
+      "learning_rate": 5e-06,
+      "loss": 0.2274,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8212237093690249,
+      "grad_norm": 2.977048635482788,
+      "learning_rate": 5e-06,
+      "loss": 0.4693,
+      "step": 4295
+    },
+    {
+      "epoch": 0.821414913957935,
+      "grad_norm": 1.9630651473999023,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8216061185468452,
+      "grad_norm": 1.8660008907318115,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8217973231357553,
+      "grad_norm": 2.163969039916992,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8219885277246654,
+      "grad_norm": 1.4228352308273315,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8221797323135756,
+      "grad_norm": 2.8689212799072266,
+      "learning_rate": 5e-06,
+      "loss": 0.3061,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8223709369024856,
+      "grad_norm": 2.3054051399230957,
+      "learning_rate": 5e-06,
+      "loss": 0.3133,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8225621414913958,
+      "grad_norm": 2.6142942905426025,
+      "learning_rate": 5e-06,
+      "loss": 0.5324,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8227533460803059,
+      "grad_norm": 1.669835090637207,
+      "learning_rate": 5e-06,
+      "loss": 0.1098,
+      "step": 4303
+    },
+    {
+      "epoch": 0.822944550669216,
+      "grad_norm": 1.9401994943618774,
+      "learning_rate": 5e-06,
+      "loss": 0.2461,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8231357552581262,
+      "grad_norm": 3.1794354915618896,
+      "learning_rate": 5e-06,
+      "loss": 0.3799,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8233269598470363,
+      "grad_norm": 2.4734315872192383,
+      "learning_rate": 5e-06,
+      "loss": 0.0708,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8235181644359465,
+      "grad_norm": 2.165508508682251,
+      "learning_rate": 5e-06,
+      "loss": 0.1704,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8237093690248566,
+      "grad_norm": 3.454468250274658,
+      "learning_rate": 5e-06,
+      "loss": 0.4224,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8239005736137668,
+      "grad_norm": 1.4881060123443604,
+      "learning_rate": 5e-06,
+      "loss": 0.1237,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8240917782026769,
+      "grad_norm": 2.7141613960266113,
+      "learning_rate": 5e-06,
+      "loss": 0.4328,
+      "step": 4310
+    },
+    {
+      "epoch": 0.824282982791587,
+      "grad_norm": 1.3264292478561401,
+      "learning_rate": 5e-06,
+      "loss": 0.0853,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8244741873804972,
+      "grad_norm": 1.4157055616378784,
+      "learning_rate": 5e-06,
+      "loss": 0.0634,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8246653919694072,
+      "grad_norm": 2.11893630027771,
+      "learning_rate": 5e-06,
+      "loss": 0.2645,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8248565965583174,
+      "grad_norm": 2.659165620803833,
+      "learning_rate": 5e-06,
+      "loss": 0.4391,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8250478011472275,
+      "grad_norm": 3.263123035430908,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8252390057361376,
+      "grad_norm": 1.4275321960449219,
+      "learning_rate": 5e-06,
+      "loss": 0.1575,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8254302103250478,
+      "grad_norm": 1.7727454900741577,
+      "learning_rate": 5e-06,
+      "loss": 0.0899,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8256214149139579,
+      "grad_norm": 2.1350901126861572,
+      "learning_rate": 5e-06,
+      "loss": 0.1986,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8258126195028681,
+      "grad_norm": 3.224039316177368,
+      "learning_rate": 5e-06,
+      "loss": 0.2131,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8260038240917782,
+      "grad_norm": 2.624913215637207,
+      "learning_rate": 5e-06,
+      "loss": 0.4338,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8261950286806883,
+      "grad_norm": 1.3268989324569702,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8263862332695985,
+      "grad_norm": 1.9256988763809204,
+      "learning_rate": 5e-06,
+      "loss": 0.177,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8265774378585086,
+      "grad_norm": 3.2375950813293457,
+      "learning_rate": 5e-06,
+      "loss": 0.2375,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8267686424474188,
+      "grad_norm": 3.528641700744629,
+      "learning_rate": 5e-06,
+      "loss": 0.3476,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8269598470363289,
+      "grad_norm": 5.507506847381592,
+      "learning_rate": 5e-06,
+      "loss": 0.3687,
+      "step": 4325
+    },
+    {
+      "epoch": 0.827151051625239,
+      "grad_norm": 2.442595958709717,
+      "learning_rate": 5e-06,
+      "loss": 0.3488,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8273422562141491,
+      "grad_norm": 4.454098224639893,
+      "learning_rate": 5e-06,
+      "loss": 0.1144,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8275334608030592,
+      "grad_norm": 2.602275848388672,
+      "learning_rate": 5e-06,
+      "loss": 0.3673,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8277246653919694,
+      "grad_norm": 1.1066206693649292,
+      "learning_rate": 5e-06,
+      "loss": 0.0566,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8279158699808795,
+      "grad_norm": 2.778461217880249,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8281070745697897,
+      "grad_norm": 1.5432476997375488,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8282982791586998,
+      "grad_norm": 1.3616504669189453,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8284894837476099,
+      "grad_norm": 2.906709671020508,
+      "learning_rate": 5e-06,
+      "loss": 0.2886,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8286806883365201,
+      "grad_norm": 3.554781913757324,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8288718929254302,
+      "grad_norm": 2.459982395172119,
+      "learning_rate": 5e-06,
+      "loss": 0.1131,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8290630975143404,
+      "grad_norm": 2.0762665271759033,
+      "learning_rate": 5e-06,
+      "loss": 0.0706,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8292543021032505,
+      "grad_norm": 1.5950919389724731,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8294455066921607,
+      "grad_norm": 2.4426164627075195,
+      "learning_rate": 5e-06,
+      "loss": 0.4334,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8296367112810707,
+      "grad_norm": 2.9879558086395264,
+      "learning_rate": 5e-06,
+      "loss": 0.5294,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8298279158699808,
+      "grad_norm": 1.552777886390686,
+      "learning_rate": 5e-06,
+      "loss": 0.1626,
+      "step": 4340
+    },
+    {
+      "epoch": 0.830019120458891,
+      "grad_norm": 2.607077121734619,
+      "learning_rate": 5e-06,
+      "loss": 0.1706,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8302103250478011,
+      "grad_norm": 1.6359832286834717,
+      "learning_rate": 5e-06,
+      "loss": 0.079,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8304015296367113,
+      "grad_norm": 1.792439341545105,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8305927342256214,
+      "grad_norm": 1.6875145435333252,
+      "learning_rate": 5e-06,
+      "loss": 0.0992,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8307839388145315,
+      "grad_norm": 1.9896464347839355,
+      "learning_rate": 5e-06,
+      "loss": 0.2902,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8309751434034417,
+      "grad_norm": 1.053695797920227,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8311663479923518,
+      "grad_norm": 2.7914059162139893,
+      "learning_rate": 5e-06,
+      "loss": 0.3776,
+      "step": 4347
+    },
+    {
+      "epoch": 0.831357552581262,
+      "grad_norm": 2.2700257301330566,
+      "learning_rate": 5e-06,
+      "loss": 0.2601,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8315487571701721,
+      "grad_norm": 2.1097259521484375,
+      "learning_rate": 5e-06,
+      "loss": 0.1924,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8317399617590823,
+      "grad_norm": 2.5981993675231934,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8319311663479924,
+      "grad_norm": 2.368746280670166,
+      "learning_rate": 5e-06,
+      "loss": 0.3906,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8321223709369024,
+      "grad_norm": 3.390716075897217,
+      "learning_rate": 5e-06,
+      "loss": 0.3522,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8323135755258126,
+      "grad_norm": 2.0879626274108887,
+      "learning_rate": 5e-06,
+      "loss": 0.3756,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8325047801147227,
+      "grad_norm": 1.396887183189392,
+      "learning_rate": 5e-06,
+      "loss": 0.1102,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8326959847036329,
+      "grad_norm": 1.5137981176376343,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 4355
+    },
+    {
+      "epoch": 0.832887189292543,
+      "grad_norm": 2.094395637512207,
+      "learning_rate": 5e-06,
+      "loss": 0.127,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8330783938814531,
+      "grad_norm": 2.304523229598999,
+      "learning_rate": 5e-06,
+      "loss": 0.309,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8332695984703633,
+      "grad_norm": 1.592523455619812,
+      "learning_rate": 5e-06,
+      "loss": 0.2083,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8334608030592734,
+      "grad_norm": 1.6386781930923462,
+      "learning_rate": 5e-06,
+      "loss": 0.1019,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8336520076481836,
+      "grad_norm": 1.4175693988800049,
+      "learning_rate": 5e-06,
+      "loss": 0.0789,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8338432122370937,
+      "grad_norm": 1.1679911613464355,
+      "learning_rate": 5e-06,
+      "loss": 0.068,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8340344168260039,
+      "grad_norm": 2.1180410385131836,
+      "learning_rate": 5e-06,
+      "loss": 0.1132,
+      "step": 4362
+    },
+    {
+      "epoch": 0.834225621414914,
+      "grad_norm": 1.151820182800293,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 4363
+    },
+    {
+      "epoch": 0.834416826003824,
+      "grad_norm": 2.254429817199707,
+      "learning_rate": 5e-06,
+      "loss": 0.3339,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8346080305927343,
+      "grad_norm": 1.0687077045440674,
+      "learning_rate": 5e-06,
+      "loss": 0.1122,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8347992351816443,
+      "grad_norm": 3.133803129196167,
+      "learning_rate": 5e-06,
+      "loss": 0.32,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8349904397705545,
+      "grad_norm": 2.1867737770080566,
+      "learning_rate": 5e-06,
+      "loss": 0.2895,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8351816443594646,
+      "grad_norm": 1.6033610105514526,
+      "learning_rate": 5e-06,
+      "loss": 0.2681,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8353728489483747,
+      "grad_norm": 1.4587613344192505,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8355640535372849,
+      "grad_norm": 2.3070292472839355,
+      "learning_rate": 5e-06,
+      "loss": 0.5036,
+      "step": 4370
+    },
+    {
+      "epoch": 0.835755258126195,
+      "grad_norm": 1.722882628440857,
+      "learning_rate": 5e-06,
+      "loss": 0.2494,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8359464627151052,
+      "grad_norm": 1.6158150434494019,
+      "learning_rate": 5e-06,
+      "loss": 0.1648,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8361376673040153,
+      "grad_norm": 1.1540052890777588,
+      "learning_rate": 5e-06,
+      "loss": 0.0811,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8363288718929255,
+      "grad_norm": 0.9952983260154724,
+      "learning_rate": 5e-06,
+      "loss": 0.0439,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8365200764818356,
+      "grad_norm": 1.722088098526001,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8367112810707457,
+      "grad_norm": 1.8059070110321045,
+      "learning_rate": 5e-06,
+      "loss": 0.3103,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8369024856596559,
+      "grad_norm": 1.7783750295639038,
+      "learning_rate": 5e-06,
+      "loss": 0.2965,
+      "step": 4377
+    },
+    {
+      "epoch": 0.837093690248566,
+      "grad_norm": 1.6145813465118408,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8372848948374761,
+      "grad_norm": 3.1882593631744385,
+      "learning_rate": 5e-06,
+      "loss": 0.3732,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8374760994263862,
+      "grad_norm": 1.8107242584228516,
+      "learning_rate": 5e-06,
+      "loss": 0.2011,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8376673040152963,
+      "grad_norm": 2.4885756969451904,
+      "learning_rate": 5e-06,
+      "loss": 0.1237,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8378585086042065,
+      "grad_norm": 3.0721991062164307,
+      "learning_rate": 5e-06,
+      "loss": 0.2544,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8380497131931166,
+      "grad_norm": 2.654618740081787,
+      "learning_rate": 5e-06,
+      "loss": 0.3133,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8382409177820268,
+      "grad_norm": 1.3207848072052002,
+      "learning_rate": 5e-06,
+      "loss": 0.2198,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8384321223709369,
+      "grad_norm": 1.5825202465057373,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 4385
+    },
+    {
+      "epoch": 0.838623326959847,
+      "grad_norm": 3.0859808921813965,
+      "learning_rate": 5e-06,
+      "loss": 0.2396,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8388145315487572,
+      "grad_norm": 1.6806787252426147,
+      "learning_rate": 5e-06,
+      "loss": 0.0929,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8390057361376673,
+      "grad_norm": 2.5619802474975586,
+      "learning_rate": 5e-06,
+      "loss": 0.164,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8391969407265775,
+      "grad_norm": 2.1878180503845215,
+      "learning_rate": 5e-06,
+      "loss": 0.2702,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8393881453154876,
+      "grad_norm": 1.7535722255706787,
+      "learning_rate": 5e-06,
+      "loss": 0.1757,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8395793499043978,
+      "grad_norm": 2.350954055786133,
+      "learning_rate": 5e-06,
+      "loss": 0.1784,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8397705544933078,
+      "grad_norm": 1.941375494003296,
+      "learning_rate": 5e-06,
+      "loss": 0.0602,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8399617590822179,
+      "grad_norm": 0.9962285161018372,
+      "learning_rate": 5e-06,
+      "loss": 0.0257,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8401529636711281,
+      "grad_norm": 1.4986315965652466,
+      "learning_rate": 5e-06,
+      "loss": 0.0806,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8403441682600382,
+      "grad_norm": 2.466259717941284,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8405353728489484,
+      "grad_norm": 1.511971116065979,
+      "learning_rate": 5e-06,
+      "loss": 0.146,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8407265774378585,
+      "grad_norm": 3.6628222465515137,
+      "learning_rate": 5e-06,
+      "loss": 0.4908,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8409177820267686,
+      "grad_norm": 1.4136232137680054,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8411089866156788,
+      "grad_norm": 1.857986569404602,
+      "learning_rate": 5e-06,
+      "loss": 0.1274,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8413001912045889,
+      "grad_norm": 1.0941559076309204,
+      "learning_rate": 5e-06,
+      "loss": 0.0448,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8414913957934991,
+      "grad_norm": 2.9076249599456787,
+      "learning_rate": 5e-06,
+      "loss": 0.4855,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8416826003824092,
+      "grad_norm": 2.237715721130371,
+      "learning_rate": 5e-06,
+      "loss": 0.2907,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8418738049713194,
+      "grad_norm": 1.244497537612915,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8420650095602294,
+      "grad_norm": 1.071437120437622,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8422562141491395,
+      "grad_norm": 1.5743379592895508,
+      "learning_rate": 5e-06,
+      "loss": 0.0687,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8424474187380497,
+      "grad_norm": 2.162822723388672,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8426386233269598,
+      "grad_norm": 1.7346101999282837,
+      "learning_rate": 5e-06,
+      "loss": 0.2139,
+      "step": 4407
+    },
+    {
+      "epoch": 0.84282982791587,
+      "grad_norm": 1.8639531135559082,
+      "learning_rate": 5e-06,
+      "loss": 0.2507,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8430210325047801,
+      "grad_norm": 3.444169044494629,
+      "learning_rate": 5e-06,
+      "loss": 0.4141,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8432122370936902,
+      "grad_norm": 2.168269395828247,
+      "learning_rate": 5e-06,
+      "loss": 0.2911,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8434034416826004,
+      "grad_norm": 1.8232005834579468,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8435946462715105,
+      "grad_norm": 2.408576488494873,
+      "learning_rate": 5e-06,
+      "loss": 0.121,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8437858508604207,
+      "grad_norm": 3.882044792175293,
+      "learning_rate": 5e-06,
+      "loss": 0.3033,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8439770554493308,
+      "grad_norm": 1.6000193357467651,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 4414
+    },
+    {
+      "epoch": 0.844168260038241,
+      "grad_norm": 1.694844126701355,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8443594646271511,
+      "grad_norm": 2.597773551940918,
+      "learning_rate": 5e-06,
+      "loss": 0.4699,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8445506692160611,
+      "grad_norm": 1.8982315063476562,
+      "learning_rate": 5e-06,
+      "loss": 0.0863,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8447418738049713,
+      "grad_norm": 2.2443645000457764,
+      "learning_rate": 5e-06,
+      "loss": 0.0798,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8449330783938814,
+      "grad_norm": 1.6703152656555176,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8451242829827916,
+      "grad_norm": 2.743288993835449,
+      "learning_rate": 5e-06,
+      "loss": 0.3322,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8453154875717017,
+      "grad_norm": 1.3997440338134766,
+      "learning_rate": 5e-06,
+      "loss": 0.1167,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8455066921606118,
+      "grad_norm": 1.7278668880462646,
+      "learning_rate": 5e-06,
+      "loss": 0.2425,
+      "step": 4422
+    },
+    {
+      "epoch": 0.845697896749522,
+      "grad_norm": 2.021827459335327,
+      "learning_rate": 5e-06,
+      "loss": 0.1857,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8458891013384321,
+      "grad_norm": 1.941655158996582,
+      "learning_rate": 5e-06,
+      "loss": 0.2009,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8460803059273423,
+      "grad_norm": 2.000415563583374,
+      "learning_rate": 5e-06,
+      "loss": 0.1906,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8462715105162524,
+      "grad_norm": 2.7244679927825928,
+      "learning_rate": 5e-06,
+      "loss": 0.3982,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8464627151051626,
+      "grad_norm": 3.001358985900879,
+      "learning_rate": 5e-06,
+      "loss": 0.5675,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8466539196940727,
+      "grad_norm": 1.9397376775741577,
+      "learning_rate": 5e-06,
+      "loss": 0.3761,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8468451242829828,
+      "grad_norm": 2.951451539993286,
+      "learning_rate": 5e-06,
+      "loss": 0.3985,
+      "step": 4429
+    },
+    {
+      "epoch": 0.847036328871893,
+      "grad_norm": 0.8639160990715027,
+      "learning_rate": 5e-06,
+      "loss": 0.0371,
+      "step": 4430
+    },
+    {
+      "epoch": 0.847227533460803,
+      "grad_norm": 1.7857511043548584,
+      "learning_rate": 5e-06,
+      "loss": 0.0694,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8474187380497132,
+      "grad_norm": 1.9648922681808472,
+      "learning_rate": 5e-06,
+      "loss": 0.2294,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8476099426386233,
+      "grad_norm": 1.5596647262573242,
+      "learning_rate": 5e-06,
+      "loss": 0.2072,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8478011472275334,
+      "grad_norm": 1.354530692100525,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8479923518164436,
+      "grad_norm": 2.014345407485962,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8481835564053537,
+      "grad_norm": 0.9619423747062683,
+      "learning_rate": 5e-06,
+      "loss": 0.0551,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8483747609942639,
+      "grad_norm": 2.9514081478118896,
+      "learning_rate": 5e-06,
+      "loss": 0.2785,
+      "step": 4437
+    },
+    {
+      "epoch": 0.848565965583174,
+      "grad_norm": 3.460991859436035,
+      "learning_rate": 5e-06,
+      "loss": 0.5416,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8487571701720842,
+      "grad_norm": 1.313177227973938,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8489483747609943,
+      "grad_norm": 2.4580676555633545,
+      "learning_rate": 5e-06,
+      "loss": 0.3445,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8491395793499044,
+      "grad_norm": 1.3704016208648682,
+      "learning_rate": 5e-06,
+      "loss": 0.0929,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8493307839388146,
+      "grad_norm": 1.3280479907989502,
+      "learning_rate": 5e-06,
+      "loss": 0.0958,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8495219885277246,
+      "grad_norm": 2.1530704498291016,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8497131931166348,
+      "grad_norm": 1.4062306880950928,
+      "learning_rate": 5e-06,
+      "loss": 0.0833,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8499043977055449,
+      "grad_norm": 1.9014475345611572,
+      "learning_rate": 5e-06,
+      "loss": 0.2039,
+      "step": 4445
+    },
+    {
+      "epoch": 0.850095602294455,
+      "grad_norm": 1.7016692161560059,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8502868068833652,
+      "grad_norm": 1.8992486000061035,
+      "learning_rate": 5e-06,
+      "loss": 0.2124,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8504780114722753,
+      "grad_norm": 2.011763095855713,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8506692160611855,
+      "grad_norm": 0.7455622553825378,
+      "learning_rate": 5e-06,
+      "loss": 0.0826,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8508604206500956,
+      "grad_norm": 1.7753809690475464,
+      "learning_rate": 5e-06,
+      "loss": 0.1413,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8510516252390057,
+      "grad_norm": 1.924593448638916,
+      "learning_rate": 5e-06,
+      "loss": 0.3003,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8512428298279159,
+      "grad_norm": 1.3443734645843506,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 4452
+    },
+    {
+      "epoch": 0.851434034416826,
+      "grad_norm": 1.712401032447815,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8516252390057362,
+      "grad_norm": 2.6731860637664795,
+      "learning_rate": 5e-06,
+      "loss": 0.1451,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8518164435946463,
+      "grad_norm": 1.2535629272460938,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8520076481835565,
+      "grad_norm": 1.3341909646987915,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8521988527724665,
+      "grad_norm": 2.0949151515960693,
+      "learning_rate": 5e-06,
+      "loss": 0.3415,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8523900573613766,
+      "grad_norm": 2.8425564765930176,
+      "learning_rate": 5e-06,
+      "loss": 0.6229,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8525812619502868,
+      "grad_norm": 2.830179214477539,
+      "learning_rate": 5e-06,
+      "loss": 0.4966,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8527724665391969,
+      "grad_norm": 3.527388334274292,
+      "learning_rate": 5e-06,
+      "loss": 0.5026,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8529636711281071,
+      "grad_norm": 1.1000351905822754,
+      "learning_rate": 5e-06,
+      "loss": 0.0993,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8531548757170172,
+      "grad_norm": 1.7830493450164795,
+      "learning_rate": 5e-06,
+      "loss": 0.0735,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8533460803059273,
+      "grad_norm": 2.3873448371887207,
+      "learning_rate": 5e-06,
+      "loss": 0.271,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8535372848948375,
+      "grad_norm": 2.0163981914520264,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8537284894837476,
+      "grad_norm": 1.701276183128357,
+      "learning_rate": 5e-06,
+      "loss": 0.2418,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8539196940726578,
+      "grad_norm": 1.9887462854385376,
+      "learning_rate": 5e-06,
+      "loss": 0.2069,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8541108986615679,
+      "grad_norm": 1.4311776161193848,
+      "learning_rate": 5e-06,
+      "loss": 0.0844,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8543021032504781,
+      "grad_norm": 1.4655697345733643,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8544933078393881,
+      "grad_norm": 2.374417781829834,
+      "learning_rate": 5e-06,
+      "loss": 0.1252,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8546845124282982,
+      "grad_norm": 1.744074821472168,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8548757170172084,
+      "grad_norm": 1.7160265445709229,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8550669216061185,
+      "grad_norm": 2.1769425868988037,
+      "learning_rate": 5e-06,
+      "loss": 0.2596,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8552581261950287,
+      "grad_norm": 2.5814943313598633,
+      "learning_rate": 5e-06,
+      "loss": 0.2571,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8554493307839388,
+      "grad_norm": 1.3457120656967163,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 4474
+    },
+    {
+      "epoch": 0.8556405353728489,
+      "grad_norm": 1.8106434345245361,
+      "learning_rate": 5e-06,
+      "loss": 0.1167,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8558317399617591,
+      "grad_norm": 1.8618403673171997,
+      "learning_rate": 5e-06,
+      "loss": 0.1876,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8560229445506692,
+      "grad_norm": 1.3731857538223267,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8562141491395794,
+      "grad_norm": 3.5358798503875732,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8564053537284895,
+      "grad_norm": 2.0287272930145264,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8565965583173997,
+      "grad_norm": 1.2044274806976318,
+      "learning_rate": 5e-06,
+      "loss": 0.0622,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8567877629063098,
+      "grad_norm": 1.2688884735107422,
+      "learning_rate": 5e-06,
+      "loss": 0.0831,
+      "step": 4481
+    },
+    {
+      "epoch": 0.8569789674952198,
+      "grad_norm": 3.058166980743408,
+      "learning_rate": 5e-06,
+      "loss": 0.5819,
+      "step": 4482
+    },
+    {
+      "epoch": 0.85717017208413,
+      "grad_norm": 2.472184896469116,
+      "learning_rate": 5e-06,
+      "loss": 0.3191,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8573613766730401,
+      "grad_norm": 2.22625732421875,
+      "learning_rate": 5e-06,
+      "loss": 0.2619,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8575525812619503,
+      "grad_norm": 1.0253074169158936,
+      "learning_rate": 5e-06,
+      "loss": 0.0643,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8577437858508604,
+      "grad_norm": 1.16357421875,
+      "learning_rate": 5e-06,
+      "loss": 0.0652,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8579349904397705,
+      "grad_norm": 1.9855798482894897,
+      "learning_rate": 5e-06,
+      "loss": 0.136,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8581261950286807,
+      "grad_norm": 2.942619800567627,
+      "learning_rate": 5e-06,
+      "loss": 0.1555,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8583173996175908,
+      "grad_norm": 1.8146679401397705,
+      "learning_rate": 5e-06,
+      "loss": 0.136,
+      "step": 4489
+    },
+    {
+      "epoch": 0.858508604206501,
+      "grad_norm": 1.2012311220169067,
+      "learning_rate": 5e-06,
+      "loss": 0.0923,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8586998087954111,
+      "grad_norm": 2.180446147918701,
+      "learning_rate": 5e-06,
+      "loss": 0.4023,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8588910133843213,
+      "grad_norm": 2.272432327270508,
+      "learning_rate": 5e-06,
+      "loss": 0.2363,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8590822179732314,
+      "grad_norm": 1.4137390851974487,
+      "learning_rate": 5e-06,
+      "loss": 0.1623,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8592734225621415,
+      "grad_norm": 3.289259433746338,
+      "learning_rate": 5e-06,
+      "loss": 0.3883,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8594646271510517,
+      "grad_norm": 1.947654128074646,
+      "learning_rate": 5e-06,
+      "loss": 0.2005,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8596558317399617,
+      "grad_norm": 1.1897157430648804,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8598470363288719,
+      "grad_norm": 1.1473478078842163,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 4497
+    },
+    {
+      "epoch": 0.860038240917782,
+      "grad_norm": 1.4698823690414429,
+      "learning_rate": 5e-06,
+      "loss": 0.0852,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8602294455066921,
+      "grad_norm": 3.318509578704834,
+      "learning_rate": 5e-06,
+      "loss": 0.3434,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8604206500956023,
+      "grad_norm": 1.8506419658660889,
+      "learning_rate": 5e-06,
+      "loss": 0.1088,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8604206500956023,
+      "eval_runtime": 769.0971,
+      "eval_samples_per_second": 1.995,
+      "eval_steps_per_second": 0.25,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8606118546845124,
+      "grad_norm": 1.9987014532089233,
+      "learning_rate": 5e-06,
+      "loss": 0.276,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8608030592734226,
+      "grad_norm": 14.492589950561523,
+      "learning_rate": 5e-06,
+      "loss": 0.4812,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8609942638623327,
+      "grad_norm": 2.069856882095337,
+      "learning_rate": 5e-06,
+      "loss": 0.2254,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8611854684512428,
+      "grad_norm": 2.994016647338867,
+      "learning_rate": 5e-06,
+      "loss": 0.291,
+      "step": 4504
+    },
+    {
+      "epoch": 0.861376673040153,
+      "grad_norm": 1.7594068050384521,
+      "learning_rate": 5e-06,
+      "loss": 0.0717,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8615678776290631,
+      "grad_norm": 2.7324862480163574,
+      "learning_rate": 5e-06,
+      "loss": 0.0629,
+      "step": 4506
+    },
+    {
+      "epoch": 0.8617590822179733,
+      "grad_norm": 3.193103075027466,
+      "learning_rate": 5e-06,
+      "loss": 0.6214,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8619502868068833,
+      "grad_norm": 2.8593437671661377,
+      "learning_rate": 5e-06,
+      "loss": 0.3593,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8621414913957935,
+      "grad_norm": 0.9549627304077148,
+      "learning_rate": 5e-06,
+      "loss": 0.0506,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8623326959847036,
+      "grad_norm": 0.934559166431427,
+      "learning_rate": 5e-06,
+      "loss": 0.0455,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8625239005736137,
+      "grad_norm": 1.489575743675232,
+      "learning_rate": 5e-06,
+      "loss": 0.1217,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8627151051625239,
+      "grad_norm": 2.525219202041626,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 4512
+    },
+    {
+      "epoch": 0.862906309751434,
+      "grad_norm": 2.3394064903259277,
+      "learning_rate": 5e-06,
+      "loss": 0.2619,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8630975143403442,
+      "grad_norm": 3.068735361099243,
+      "learning_rate": 5e-06,
+      "loss": 0.497,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8632887189292543,
+      "grad_norm": 1.4633656740188599,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8634799235181644,
+      "grad_norm": 1.276939868927002,
+      "learning_rate": 5e-06,
+      "loss": 0.071,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8636711281070746,
+      "grad_norm": 1.701491355895996,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8638623326959847,
+      "grad_norm": 1.7778617143630981,
+      "learning_rate": 5e-06,
+      "loss": 0.158,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8640535372848949,
+      "grad_norm": 2.2561211585998535,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 4519
+    },
+    {
+      "epoch": 0.864244741873805,
+      "grad_norm": 1.6372724771499634,
+      "learning_rate": 5e-06,
+      "loss": 0.165,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8644359464627152,
+      "grad_norm": 1.8756248950958252,
+      "learning_rate": 5e-06,
+      "loss": 0.1654,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8646271510516252,
+      "grad_norm": 1.4479553699493408,
+      "learning_rate": 5e-06,
+      "loss": 0.2137,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8648183556405353,
+      "grad_norm": 1.0815504789352417,
+      "learning_rate": 5e-06,
+      "loss": 0.0721,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8650095602294455,
+      "grad_norm": 1.7075153589248657,
+      "learning_rate": 5e-06,
+      "loss": 0.0744,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8652007648183556,
+      "grad_norm": 2.4414803981781006,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8653919694072658,
+      "grad_norm": 2.04904842376709,
+      "learning_rate": 5e-06,
+      "loss": 0.2584,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8655831739961759,
+      "grad_norm": 2.7054378986358643,
+      "learning_rate": 5e-06,
+      "loss": 0.4034,
+      "step": 4527
+    },
+    {
+      "epoch": 0.865774378585086,
+      "grad_norm": 1.9329419136047363,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8659655831739962,
+      "grad_norm": 2.828913450241089,
+      "learning_rate": 5e-06,
+      "loss": 0.4183,
+      "step": 4529
+    },
+    {
+      "epoch": 0.8661567877629063,
+      "grad_norm": 2.64652419090271,
+      "learning_rate": 5e-06,
+      "loss": 0.0719,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8663479923518165,
+      "grad_norm": 1.249855399131775,
+      "learning_rate": 5e-06,
+      "loss": 0.0515,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8665391969407266,
+      "grad_norm": 2.1562163829803467,
+      "learning_rate": 5e-06,
+      "loss": 0.1446,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8667304015296368,
+      "grad_norm": 2.4176135063171387,
+      "learning_rate": 5e-06,
+      "loss": 0.4205,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8669216061185469,
+      "grad_norm": 1.6887050867080688,
+      "learning_rate": 5e-06,
+      "loss": 0.1691,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8671128107074569,
+      "grad_norm": 3.1317782402038574,
+      "learning_rate": 5e-06,
+      "loss": 0.2806,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8673040152963671,
+      "grad_norm": 2.647075891494751,
+      "learning_rate": 5e-06,
+      "loss": 0.3814,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8674952198852772,
+      "grad_norm": 1.211713194847107,
+      "learning_rate": 5e-06,
+      "loss": 0.0743,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8676864244741874,
+      "grad_norm": 1.462576150894165,
+      "learning_rate": 5e-06,
+      "loss": 0.1225,
+      "step": 4538
+    },
+    {
+      "epoch": 0.8678776290630975,
+      "grad_norm": 2.324312925338745,
+      "learning_rate": 5e-06,
+      "loss": 0.2452,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8680688336520076,
+      "grad_norm": 1.4178894758224487,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8682600382409178,
+      "grad_norm": 4.281589508056641,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8684512428298279,
+      "grad_norm": 2.9054739475250244,
+      "learning_rate": 5e-06,
+      "loss": 0.1756,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8686424474187381,
+      "grad_norm": 1.1191303730010986,
+      "learning_rate": 5e-06,
+      "loss": 0.0303,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8688336520076482,
+      "grad_norm": 4.550809383392334,
+      "learning_rate": 5e-06,
+      "loss": 0.2272,
+      "step": 4544
+    },
+    {
+      "epoch": 0.8690248565965584,
+      "grad_norm": 2.311413526535034,
+      "learning_rate": 5e-06,
+      "loss": 0.2753,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8692160611854685,
+      "grad_norm": 2.0396246910095215,
+      "learning_rate": 5e-06,
+      "loss": 0.2035,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8694072657743785,
+      "grad_norm": 3.7515199184417725,
+      "learning_rate": 5e-06,
+      "loss": 0.3386,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8695984703632887,
+      "grad_norm": 2.69758939743042,
+      "learning_rate": 5e-06,
+      "loss": 0.3004,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8697896749521988,
+      "grad_norm": 1.6221609115600586,
+      "learning_rate": 5e-06,
+      "loss": 0.1011,
+      "step": 4549
+    },
+    {
+      "epoch": 0.869980879541109,
+      "grad_norm": 1.3611334562301636,
+      "learning_rate": 5e-06,
+      "loss": 0.0866,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8701720841300191,
+      "grad_norm": 2.460482120513916,
+      "learning_rate": 5e-06,
+      "loss": 0.3597,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8703632887189292,
+      "grad_norm": 1.2450191974639893,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8705544933078394,
+      "grad_norm": 1.1274508237838745,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8707456978967495,
+      "grad_norm": 1.571494221687317,
+      "learning_rate": 5e-06,
+      "loss": 0.2236,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8709369024856597,
+      "grad_norm": 1.62301504611969,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8711281070745698,
+      "grad_norm": 0.9670308828353882,
+      "learning_rate": 5e-06,
+      "loss": 0.0481,
+      "step": 4556
+    },
+    {
+      "epoch": 0.87131931166348,
+      "grad_norm": 1.0194993019104004,
+      "learning_rate": 5e-06,
+      "loss": 0.0995,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8715105162523901,
+      "grad_norm": 2.630286455154419,
+      "learning_rate": 5e-06,
+      "loss": 0.526,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8717017208413002,
+      "grad_norm": 2.8174970149993896,
+      "learning_rate": 5e-06,
+      "loss": 0.1966,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8718929254302104,
+      "grad_norm": 1.6453908681869507,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8720841300191204,
+      "grad_norm": 1.6050821542739868,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 4561
+    },
+    {
+      "epoch": 0.8722753346080306,
+      "grad_norm": 2.457448959350586,
+      "learning_rate": 5e-06,
+      "loss": 0.2686,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8724665391969407,
+      "grad_norm": 1.3800158500671387,
+      "learning_rate": 5e-06,
+      "loss": 0.0852,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8726577437858508,
+      "grad_norm": 3.040879726409912,
+      "learning_rate": 5e-06,
+      "loss": 0.4089,
+      "step": 4564
+    },
+    {
+      "epoch": 0.872848948374761,
+      "grad_norm": 1.8485065698623657,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8730401529636711,
+      "grad_norm": 1.5646363496780396,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8732313575525813,
+      "grad_norm": 2.561878204345703,
+      "learning_rate": 5e-06,
+      "loss": 0.1572,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8734225621414914,
+      "grad_norm": 1.187648892402649,
+      "learning_rate": 5e-06,
+      "loss": 0.0962,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8736137667304015,
+      "grad_norm": 2.9988815784454346,
+      "learning_rate": 5e-06,
+      "loss": 0.2014,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8738049713193117,
+      "grad_norm": 1.185423731803894,
+      "learning_rate": 5e-06,
+      "loss": 0.1753,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8739961759082218,
+      "grad_norm": 1.1318641901016235,
+      "learning_rate": 5e-06,
+      "loss": 0.0889,
+      "step": 4571
+    },
+    {
+      "epoch": 0.874187380497132,
+      "grad_norm": 1.3041385412216187,
+      "learning_rate": 5e-06,
+      "loss": 0.0902,
+      "step": 4572
+    },
+    {
+      "epoch": 0.874378585086042,
+      "grad_norm": 2.061596632003784,
+      "learning_rate": 5e-06,
+      "loss": 0.3094,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8745697896749522,
+      "grad_norm": 1.7980538606643677,
+      "learning_rate": 5e-06,
+      "loss": 0.2838,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8747609942638623,
+      "grad_norm": 4.6933135986328125,
+      "learning_rate": 5e-06,
+      "loss": 0.1503,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8749521988527724,
+      "grad_norm": 2.4895479679107666,
+      "learning_rate": 5e-06,
+      "loss": 0.385,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8751434034416826,
+      "grad_norm": 2.2268948554992676,
+      "learning_rate": 5e-06,
+      "loss": 0.2536,
+      "step": 4577
+    },
+    {
+      "epoch": 0.8753346080305927,
+      "grad_norm": 1.4584846496582031,
+      "learning_rate": 5e-06,
+      "loss": 0.0736,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8755258126195029,
+      "grad_norm": 1.1387372016906738,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 4579
+    },
+    {
+      "epoch": 0.875717017208413,
+      "grad_norm": 1.2333475351333618,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8759082217973231,
+      "grad_norm": 1.2063689231872559,
+      "learning_rate": 5e-06,
+      "loss": 0.0427,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8760994263862333,
+      "grad_norm": 2.055612087249756,
+      "learning_rate": 5e-06,
+      "loss": 0.3624,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8762906309751434,
+      "grad_norm": 1.719244360923767,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8764818355640536,
+      "grad_norm": 1.0476865768432617,
+      "learning_rate": 5e-06,
+      "loss": 0.0592,
+      "step": 4584
+    },
+    {
+      "epoch": 0.8766730401529637,
+      "grad_norm": 1.689634919166565,
+      "learning_rate": 5e-06,
+      "loss": 0.2333,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8768642447418739,
+      "grad_norm": 1.7837719917297363,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 4586
+    },
+    {
+      "epoch": 0.8770554493307839,
+      "grad_norm": 1.6960902214050293,
+      "learning_rate": 5e-06,
+      "loss": 0.0815,
+      "step": 4587
+    },
+    {
+      "epoch": 0.877246653919694,
+      "grad_norm": 3.9408388137817383,
+      "learning_rate": 5e-06,
+      "loss": 0.3912,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8774378585086042,
+      "grad_norm": 2.848705530166626,
+      "learning_rate": 5e-06,
+      "loss": 0.2726,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8776290630975143,
+      "grad_norm": 1.9206777811050415,
+      "learning_rate": 5e-06,
+      "loss": 0.1705,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8778202676864245,
+      "grad_norm": 1.366650104522705,
+      "learning_rate": 5e-06,
+      "loss": 0.1402,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8780114722753346,
+      "grad_norm": 2.2299587726593018,
+      "learning_rate": 5e-06,
+      "loss": 0.1416,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8782026768642447,
+      "grad_norm": 1.1620739698410034,
+      "learning_rate": 5e-06,
+      "loss": 0.0649,
+      "step": 4593
+    },
+    {
+      "epoch": 0.8783938814531549,
+      "grad_norm": 1.5828089714050293,
+      "learning_rate": 5e-06,
+      "loss": 0.129,
+      "step": 4594
+    },
+    {
+      "epoch": 0.878585086042065,
+      "grad_norm": 1.6714706420898438,
+      "learning_rate": 5e-06,
+      "loss": 0.213,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8787762906309752,
+      "grad_norm": 2.159733533859253,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8789674952198853,
+      "grad_norm": 4.393616676330566,
+      "learning_rate": 5e-06,
+      "loss": 0.2552,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8791586998087955,
+      "grad_norm": 1.0370787382125854,
+      "learning_rate": 5e-06,
+      "loss": 0.0785,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8793499043977056,
+      "grad_norm": 1.0146803855895996,
+      "learning_rate": 5e-06,
+      "loss": 0.0634,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8795411089866156,
+      "grad_norm": 2.365654945373535,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8797323135755258,
+      "grad_norm": 2.3569753170013428,
+      "learning_rate": 5e-06,
+      "loss": 0.4793,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8799235181644359,
+      "grad_norm": 3.6621952056884766,
+      "learning_rate": 5e-06,
+      "loss": 0.6331,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8801147227533461,
+      "grad_norm": 2.0783820152282715,
+      "learning_rate": 5e-06,
+      "loss": 0.3076,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8803059273422562,
+      "grad_norm": 1.9632409811019897,
+      "learning_rate": 5e-06,
+      "loss": 0.0954,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8804971319311663,
+      "grad_norm": 1.1970844268798828,
+      "learning_rate": 5e-06,
+      "loss": 0.0962,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8806883365200765,
+      "grad_norm": 1.9226810932159424,
+      "learning_rate": 5e-06,
+      "loss": 0.0956,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8808795411089866,
+      "grad_norm": 2.701702833175659,
+      "learning_rate": 5e-06,
+      "loss": 0.4784,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8810707456978968,
+      "grad_norm": 1.3243826627731323,
+      "learning_rate": 5e-06,
+      "loss": 0.0847,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8812619502868069,
+      "grad_norm": 1.84163236618042,
+      "learning_rate": 5e-06,
+      "loss": 0.1919,
+      "step": 4609
+    },
+    {
+      "epoch": 0.8814531548757171,
+      "grad_norm": 2.9473471641540527,
+      "learning_rate": 5e-06,
+      "loss": 0.2047,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8816443594646272,
+      "grad_norm": 1.4090523719787598,
+      "learning_rate": 5e-06,
+      "loss": 0.1088,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8818355640535372,
+      "grad_norm": 1.92579185962677,
+      "learning_rate": 5e-06,
+      "loss": 0.0899,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8820267686424474,
+      "grad_norm": 2.1861867904663086,
+      "learning_rate": 5e-06,
+      "loss": 0.283,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8822179732313575,
+      "grad_norm": 1.9572925567626953,
+      "learning_rate": 5e-06,
+      "loss": 0.2118,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8824091778202677,
+      "grad_norm": 1.4317727088928223,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8826003824091778,
+      "grad_norm": 0.7707310914993286,
+      "learning_rate": 5e-06,
+      "loss": 0.0597,
+      "step": 4616
+    },
+    {
+      "epoch": 0.8827915869980879,
+      "grad_norm": 3.6430459022521973,
+      "learning_rate": 5e-06,
+      "loss": 0.5254,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8829827915869981,
+      "grad_norm": 0.7240305542945862,
+      "learning_rate": 5e-06,
+      "loss": 0.035,
+      "step": 4618
+    },
+    {
+      "epoch": 0.8831739961759082,
+      "grad_norm": 1.372766137123108,
+      "learning_rate": 5e-06,
+      "loss": 0.0557,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8833652007648184,
+      "grad_norm": 2.592245578765869,
+      "learning_rate": 5e-06,
+      "loss": 0.3585,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8835564053537285,
+      "grad_norm": 1.7263092994689941,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8837476099426386,
+      "grad_norm": 1.6172292232513428,
+      "learning_rate": 5e-06,
+      "loss": 0.1043,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8839388145315488,
+      "grad_norm": 1.8567088842391968,
+      "learning_rate": 5e-06,
+      "loss": 0.0783,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8841300191204589,
+      "grad_norm": 1.0310090780258179,
+      "learning_rate": 5e-06,
+      "loss": 0.0906,
+      "step": 4624
+    },
+    {
+      "epoch": 0.884321223709369,
+      "grad_norm": 2.3422353267669678,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8845124282982791,
+      "grad_norm": 2.5427935123443604,
+      "learning_rate": 5e-06,
+      "loss": 0.4579,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8847036328871893,
+      "grad_norm": 1.8456017971038818,
+      "learning_rate": 5e-06,
+      "loss": 0.1279,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8848948374760994,
+      "grad_norm": 1.4140512943267822,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8850860420650095,
+      "grad_norm": 1.278416395187378,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8852772466539197,
+      "grad_norm": 0.722949206829071,
+      "learning_rate": 5e-06,
+      "loss": 0.0267,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8854684512428298,
+      "grad_norm": 1.8881372213363647,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 4631
+    },
+    {
+      "epoch": 0.88565965583174,
+      "grad_norm": 2.0543901920318604,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8858508604206501,
+      "grad_norm": 2.746734619140625,
+      "learning_rate": 5e-06,
+      "loss": 0.4295,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8860420650095602,
+      "grad_norm": 1.3126775026321411,
+      "learning_rate": 5e-06,
+      "loss": 0.072,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8862332695984704,
+      "grad_norm": 2.1847662925720215,
+      "learning_rate": 5e-06,
+      "loss": 0.193,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8864244741873805,
+      "grad_norm": 2.4903340339660645,
+      "learning_rate": 5e-06,
+      "loss": 0.0869,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8866156787762907,
+      "grad_norm": 2.261547803878784,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8868068833652007,
+      "grad_norm": 3.0043082237243652,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 4638
+    },
+    {
+      "epoch": 0.886998087954111,
+      "grad_norm": 1.8434854745864868,
+      "learning_rate": 5e-06,
+      "loss": 0.2356,
+      "step": 4639
+    },
+    {
+      "epoch": 0.887189292543021,
+      "grad_norm": 1.221226453781128,
+      "learning_rate": 5e-06,
+      "loss": 0.0681,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8873804971319311,
+      "grad_norm": 1.7801470756530762,
+      "learning_rate": 5e-06,
+      "loss": 0.0785,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8875717017208413,
+      "grad_norm": 1.2522605657577515,
+      "learning_rate": 5e-06,
+      "loss": 0.0611,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8877629063097514,
+      "grad_norm": 1.8315064907073975,
+      "learning_rate": 5e-06,
+      "loss": 0.0613,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8879541108986616,
+      "grad_norm": 1.2048170566558838,
+      "learning_rate": 5e-06,
+      "loss": 0.0696,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8881453154875717,
+      "grad_norm": 2.3258895874023438,
+      "learning_rate": 5e-06,
+      "loss": 0.3081,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8883365200764818,
+      "grad_norm": 1.9494102001190186,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 4646
+    },
+    {
+      "epoch": 0.888527724665392,
+      "grad_norm": 3.9364311695098877,
+      "learning_rate": 5e-06,
+      "loss": 0.5175,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8887189292543021,
+      "grad_norm": 2.00209641456604,
+      "learning_rate": 5e-06,
+      "loss": 0.2127,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8889101338432123,
+      "grad_norm": 1.54474675655365,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8891013384321224,
+      "grad_norm": 2.311948537826538,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8892925430210326,
+      "grad_norm": 2.3398048877716064,
+      "learning_rate": 5e-06,
+      "loss": 0.2935,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8894837476099426,
+      "grad_norm": 3.875178575515747,
+      "learning_rate": 5e-06,
+      "loss": 0.4512,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8896749521988527,
+      "grad_norm": 1.4064953327178955,
+      "learning_rate": 5e-06,
+      "loss": 0.1605,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8898661567877629,
+      "grad_norm": 2.825348138809204,
+      "learning_rate": 5e-06,
+      "loss": 0.2991,
+      "step": 4654
+    },
+    {
+      "epoch": 0.890057361376673,
+      "grad_norm": 1.4716694355010986,
+      "learning_rate": 5e-06,
+      "loss": 0.092,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8902485659655832,
+      "grad_norm": 1.371217966079712,
+      "learning_rate": 5e-06,
+      "loss": 0.1255,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8904397705544933,
+      "grad_norm": 2.445219039916992,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8906309751434034,
+      "grad_norm": 2.0857105255126953,
+      "learning_rate": 5e-06,
+      "loss": 0.33,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8908221797323136,
+      "grad_norm": 1.8164504766464233,
+      "learning_rate": 5e-06,
+      "loss": 0.256,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8910133843212237,
+      "grad_norm": 1.9205447435379028,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8912045889101339,
+      "grad_norm": 1.0629031658172607,
+      "learning_rate": 5e-06,
+      "loss": 0.054,
+      "step": 4661
+    },
+    {
+      "epoch": 0.891395793499044,
+      "grad_norm": 2.1165101528167725,
+      "learning_rate": 5e-06,
+      "loss": 0.2257,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8915869980879542,
+      "grad_norm": 2.4531354904174805,
+      "learning_rate": 5e-06,
+      "loss": 0.2545,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8917782026768643,
+      "grad_norm": 1.8205480575561523,
+      "learning_rate": 5e-06,
+      "loss": 0.2478,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8919694072657743,
+      "grad_norm": 1.4492504596710205,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8921606118546845,
+      "grad_norm": 2.0282204151153564,
+      "learning_rate": 5e-06,
+      "loss": 0.109,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8923518164435946,
+      "grad_norm": 1.1359578371047974,
+      "learning_rate": 5e-06,
+      "loss": 0.0532,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8925430210325048,
+      "grad_norm": 1.3664461374282837,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8927342256214149,
+      "grad_norm": 2.3731942176818848,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 4669
+    },
+    {
+      "epoch": 0.892925430210325,
+      "grad_norm": 1.5064061880111694,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8931166347992352,
+      "grad_norm": 3.036389112472534,
+      "learning_rate": 5e-06,
+      "loss": 0.3978,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8933078393881453,
+      "grad_norm": 1.1229474544525146,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8934990439770555,
+      "grad_norm": 1.6844687461853027,
+      "learning_rate": 5e-06,
+      "loss": 0.2174,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8936902485659656,
+      "grad_norm": 1.0797251462936401,
+      "learning_rate": 5e-06,
+      "loss": 0.0696,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8938814531548758,
+      "grad_norm": 1.7389681339263916,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8940726577437859,
+      "grad_norm": 1.4130902290344238,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 4676
+    },
+    {
+      "epoch": 0.894263862332696,
+      "grad_norm": 1.8311725854873657,
+      "learning_rate": 5e-06,
+      "loss": 0.2501,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8944550669216061,
+      "grad_norm": 3.1457982063293457,
+      "learning_rate": 5e-06,
+      "loss": 0.2693,
+      "step": 4678
+    },
+    {
+      "epoch": 0.8946462715105162,
+      "grad_norm": 1.2318836450576782,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 4679
+    },
+    {
+      "epoch": 0.8948374760994264,
+      "grad_norm": 1.2631968259811401,
+      "learning_rate": 5e-06,
+      "loss": 0.0692,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8950286806883365,
+      "grad_norm": 1.3267889022827148,
+      "learning_rate": 5e-06,
+      "loss": 0.1024,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8952198852772466,
+      "grad_norm": 3.25459361076355,
+      "learning_rate": 5e-06,
+      "loss": 0.4319,
+      "step": 4682
+    },
+    {
+      "epoch": 0.8954110898661568,
+      "grad_norm": 3.177501678466797,
+      "learning_rate": 5e-06,
+      "loss": 0.7296,
+      "step": 4683
+    },
+    {
+      "epoch": 0.8956022944550669,
+      "grad_norm": 2.4800193309783936,
+      "learning_rate": 5e-06,
+      "loss": 0.309,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8957934990439771,
+      "grad_norm": 1.386919379234314,
+      "learning_rate": 5e-06,
+      "loss": 0.1178,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8959847036328872,
+      "grad_norm": 0.9860485196113586,
+      "learning_rate": 5e-06,
+      "loss": 0.0742,
+      "step": 4686
+    },
+    {
+      "epoch": 0.8961759082217973,
+      "grad_norm": 1.5006732940673828,
+      "learning_rate": 5e-06,
+      "loss": 0.0625,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8963671128107075,
+      "grad_norm": 1.8717820644378662,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8965583173996176,
+      "grad_norm": 2.3897619247436523,
+      "learning_rate": 5e-06,
+      "loss": 0.6113,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8967495219885278,
+      "grad_norm": 1.6034753322601318,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8969407265774378,
+      "grad_norm": 2.3642208576202393,
+      "learning_rate": 5e-06,
+      "loss": 0.3357,
+      "step": 4691
+    },
+    {
+      "epoch": 0.897131931166348,
+      "grad_norm": 2.514803409576416,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 4692
+    },
+    {
+      "epoch": 0.8973231357552581,
+      "grad_norm": 1.097195029258728,
+      "learning_rate": 5e-06,
+      "loss": 0.0548,
+      "step": 4693
+    },
+    {
+      "epoch": 0.8975143403441682,
+      "grad_norm": 1.1098480224609375,
+      "learning_rate": 5e-06,
+      "loss": 0.0912,
+      "step": 4694
+    },
+    {
+      "epoch": 0.8977055449330784,
+      "grad_norm": 2.1470210552215576,
+      "learning_rate": 5e-06,
+      "loss": 0.3154,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8978967495219885,
+      "grad_norm": 1.509371042251587,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 4696
+    },
+    {
+      "epoch": 0.8980879541108987,
+      "grad_norm": 1.2485581636428833,
+      "learning_rate": 5e-06,
+      "loss": 0.1053,
+      "step": 4697
+    },
+    {
+      "epoch": 0.8982791586998088,
+      "grad_norm": 1.2133026123046875,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 4698
+    },
+    {
+      "epoch": 0.8984703632887189,
+      "grad_norm": 1.5444940328598022,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 4699
+    },
+    {
+      "epoch": 0.8986615678776291,
+      "grad_norm": 2.6265218257904053,
+      "learning_rate": 5e-06,
+      "loss": 0.2122,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8988527724665392,
+      "grad_norm": 2.756988286972046,
+      "learning_rate": 5e-06,
+      "loss": 0.5817,
+      "step": 4701
+    },
+    {
+      "epoch": 0.8990439770554494,
+      "grad_norm": 2.6874217987060547,
+      "learning_rate": 5e-06,
+      "loss": 0.3232,
+      "step": 4702
+    },
+    {
+      "epoch": 0.8992351816443594,
+      "grad_norm": 2.7615904808044434,
+      "learning_rate": 5e-06,
+      "loss": 0.3782,
+      "step": 4703
+    },
+    {
+      "epoch": 0.8994263862332696,
+      "grad_norm": 2.0286366939544678,
+      "learning_rate": 5e-06,
+      "loss": 0.2288,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8996175908221797,
+      "grad_norm": 1.6421332359313965,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8998087954110898,
+      "grad_norm": 1.2805842161178589,
+      "learning_rate": 5e-06,
+      "loss": 0.0755,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.985251545906067,
+      "learning_rate": 5e-06,
+      "loss": 0.3243,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9001912045889101,
+      "grad_norm": 1.3914923667907715,
+      "learning_rate": 5e-06,
+      "loss": 0.2003,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9003824091778203,
+      "grad_norm": 2.9044227600097656,
+      "learning_rate": 5e-06,
+      "loss": 0.3824,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9005736137667304,
+      "grad_norm": 1.989436149597168,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9007648183556405,
+      "grad_norm": 3.248147487640381,
+      "learning_rate": 5e-06,
+      "loss": 0.1397,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9009560229445507,
+      "grad_norm": 3.772764205932617,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9011472275334608,
+      "grad_norm": 1.53439199924469,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 4713
+    },
+    {
+      "epoch": 0.901338432122371,
+      "grad_norm": 2.0962002277374268,
+      "learning_rate": 5e-06,
+      "loss": 0.2381,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9015296367112811,
+      "grad_norm": 1.7860227823257446,
+      "learning_rate": 5e-06,
+      "loss": 0.0828,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9017208413001913,
+      "grad_norm": 2.519381284713745,
+      "learning_rate": 5e-06,
+      "loss": 0.2643,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9019120458891013,
+      "grad_norm": 2.8200933933258057,
+      "learning_rate": 5e-06,
+      "loss": 0.3501,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9021032504780114,
+      "grad_norm": 1.7289233207702637,
+      "learning_rate": 5e-06,
+      "loss": 0.1181,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9022944550669216,
+      "grad_norm": 2.348267078399658,
+      "learning_rate": 5e-06,
+      "loss": 0.2862,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9024856596558317,
+      "grad_norm": 1.9853914976119995,
+      "learning_rate": 5e-06,
+      "loss": 0.3011,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9026768642447419,
+      "grad_norm": 1.2755216360092163,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 4721
+    },
+    {
+      "epoch": 0.902868068833652,
+      "grad_norm": 2.737816095352173,
+      "learning_rate": 5e-06,
+      "loss": 0.3493,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9030592734225621,
+      "grad_norm": 2.026895523071289,
+      "learning_rate": 5e-06,
+      "loss": 0.2523,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9032504780114723,
+      "grad_norm": 1.8737328052520752,
+      "learning_rate": 5e-06,
+      "loss": 0.1153,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9034416826003824,
+      "grad_norm": 2.0649120807647705,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9036328871892926,
+      "grad_norm": 2.1456220149993896,
+      "learning_rate": 5e-06,
+      "loss": 0.1975,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9038240917782027,
+      "grad_norm": 1.953724980354309,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9040152963671129,
+      "grad_norm": 1.8601813316345215,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 4728
+    },
+    {
+      "epoch": 0.904206500956023,
+      "grad_norm": 2.1330983638763428,
+      "learning_rate": 5e-06,
+      "loss": 0.1379,
+      "step": 4729
+    },
+    {
+      "epoch": 0.904397705544933,
+      "grad_norm": 2.6561925411224365,
+      "learning_rate": 5e-06,
+      "loss": 0.277,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9045889101338432,
+      "grad_norm": 2.5746819972991943,
+      "learning_rate": 5e-06,
+      "loss": 0.0965,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9047801147227533,
+      "grad_norm": 2.185016632080078,
+      "learning_rate": 5e-06,
+      "loss": 0.2095,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9049713193116635,
+      "grad_norm": 2.8116836547851562,
+      "learning_rate": 5e-06,
+      "loss": 0.5521,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9051625239005736,
+      "grad_norm": 2.427065134048462,
+      "learning_rate": 5e-06,
+      "loss": 0.2564,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9053537284894837,
+      "grad_norm": 2.053335666656494,
+      "learning_rate": 5e-06,
+      "loss": 0.272,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9055449330783939,
+      "grad_norm": 2.481733798980713,
+      "learning_rate": 5e-06,
+      "loss": 0.3331,
+      "step": 4736
+    },
+    {
+      "epoch": 0.905736137667304,
+      "grad_norm": 2.1682722568511963,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9059273422562142,
+      "grad_norm": 2.0562095642089844,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9061185468451243,
+      "grad_norm": 1.898068904876709,
+      "learning_rate": 5e-06,
+      "loss": 0.2335,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9063097514340345,
+      "grad_norm": 1.218976378440857,
+      "learning_rate": 5e-06,
+      "loss": 0.085,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9065009560229446,
+      "grad_norm": 1.4753073453903198,
+      "learning_rate": 5e-06,
+      "loss": 0.181,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9066921606118546,
+      "grad_norm": 1.8766474723815918,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9068833652007648,
+      "grad_norm": 1.521669626235962,
+      "learning_rate": 5e-06,
+      "loss": 0.0619,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9070745697896749,
+      "grad_norm": 1.3307064771652222,
+      "learning_rate": 5e-06,
+      "loss": 0.1082,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9072657743785851,
+      "grad_norm": 2.4466915130615234,
+      "learning_rate": 5e-06,
+      "loss": 0.6762,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9074569789674952,
+      "grad_norm": 1.5775731801986694,
+      "learning_rate": 5e-06,
+      "loss": 0.261,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9076481835564053,
+      "grad_norm": 1.7367063760757446,
+      "learning_rate": 5e-06,
+      "loss": 0.217,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9078393881453155,
+      "grad_norm": 2.7421231269836426,
+      "learning_rate": 5e-06,
+      "loss": 0.2712,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9080305927342256,
+      "grad_norm": 1.4664653539657593,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9082217973231358,
+      "grad_norm": 6.154154300689697,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9084130019120459,
+      "grad_norm": 1.722429871559143,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 4751
+    },
+    {
+      "epoch": 0.908604206500956,
+      "grad_norm": 3.028075695037842,
+      "learning_rate": 5e-06,
+      "loss": 0.5589,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9087954110898662,
+      "grad_norm": 1.7719171047210693,
+      "learning_rate": 5e-06,
+      "loss": 0.2481,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9089866156787763,
+      "grad_norm": 2.284649610519409,
+      "learning_rate": 5e-06,
+      "loss": 0.3038,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9091778202676865,
+      "grad_norm": 1.697432041168213,
+      "learning_rate": 5e-06,
+      "loss": 0.093,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9093690248565965,
+      "grad_norm": 1.257247805595398,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9095602294455067,
+      "grad_norm": 4.2492523193359375,
+      "learning_rate": 5e-06,
+      "loss": 0.2155,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9097514340344168,
+      "grad_norm": 1.3657695055007935,
+      "learning_rate": 5e-06,
+      "loss": 0.2232,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9099426386233269,
+      "grad_norm": 1.7062994241714478,
+      "learning_rate": 5e-06,
+      "loss": 0.2433,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9101338432122371,
+      "grad_norm": 1.1820368766784668,
+      "learning_rate": 5e-06,
+      "loss": 0.0536,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9103250478011472,
+      "grad_norm": 1.561043620109558,
+      "learning_rate": 5e-06,
+      "loss": 0.0761,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9105162523900574,
+      "grad_norm": 2.44998836517334,
+      "learning_rate": 5e-06,
+      "loss": 0.2631,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9107074569789675,
+      "grad_norm": 2.2695345878601074,
+      "learning_rate": 5e-06,
+      "loss": 0.2851,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9108986615678776,
+      "grad_norm": 2.3107171058654785,
+      "learning_rate": 5e-06,
+      "loss": 0.3215,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9110898661567878,
+      "grad_norm": 2.1133556365966797,
+      "learning_rate": 5e-06,
+      "loss": 0.2423,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9112810707456979,
+      "grad_norm": 1.5830267667770386,
+      "learning_rate": 5e-06,
+      "loss": 0.1177,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9114722753346081,
+      "grad_norm": 2.6378262042999268,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9116634799235181,
+      "grad_norm": 1.1338980197906494,
+      "learning_rate": 5e-06,
+      "loss": 0.0795,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9118546845124283,
+      "grad_norm": 2.1762497425079346,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9120458891013384,
+      "grad_norm": 2.5121426582336426,
+      "learning_rate": 5e-06,
+      "loss": 0.3784,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9122370936902485,
+      "grad_norm": 2.436002731323242,
+      "learning_rate": 5e-06,
+      "loss": 0.3143,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9124282982791587,
+      "grad_norm": 2.774428129196167,
+      "learning_rate": 5e-06,
+      "loss": 0.3945,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9126195028680688,
+      "grad_norm": 1.4498441219329834,
+      "learning_rate": 5e-06,
+      "loss": 0.1016,
+      "step": 4773
+    },
+    {
+      "epoch": 0.912810707456979,
+      "grad_norm": 2.4954347610473633,
+      "learning_rate": 5e-06,
+      "loss": 0.0956,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9130019120458891,
+      "grad_norm": 1.8431687355041504,
+      "learning_rate": 5e-06,
+      "loss": 0.0724,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9131931166347992,
+      "grad_norm": 1.2787989377975464,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9133843212237094,
+      "grad_norm": 1.2174687385559082,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9135755258126195,
+      "grad_norm": 2.3031578063964844,
+      "learning_rate": 5e-06,
+      "loss": 0.3992,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9137667304015297,
+      "grad_norm": 2.224651336669922,
+      "learning_rate": 5e-06,
+      "loss": 0.3414,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9139579349904398,
+      "grad_norm": 1.978184700012207,
+      "learning_rate": 5e-06,
+      "loss": 0.2514,
+      "step": 4780
+    },
+    {
+      "epoch": 0.91414913957935,
+      "grad_norm": 1.6605628728866577,
+      "learning_rate": 5e-06,
+      "loss": 0.0844,
+      "step": 4781
+    },
+    {
+      "epoch": 0.91434034416826,
+      "grad_norm": 1.5570906400680542,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9145315487571701,
+      "grad_norm": 1.553684949874878,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9147227533460803,
+      "grad_norm": 1.4999451637268066,
+      "learning_rate": 5e-06,
+      "loss": 0.0977,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9149139579349904,
+      "grad_norm": 1.102944254875183,
+      "learning_rate": 5e-06,
+      "loss": 0.0922,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9151051625239006,
+      "grad_norm": 1.5913718938827515,
+      "learning_rate": 5e-06,
+      "loss": 0.0704,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9152963671128107,
+      "grad_norm": 1.003926396369934,
+      "learning_rate": 5e-06,
+      "loss": 0.0447,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9154875717017208,
+      "grad_norm": 3.418393135070801,
+      "learning_rate": 5e-06,
+      "loss": 0.5752,
+      "step": 4788
+    },
+    {
+      "epoch": 0.915678776290631,
+      "grad_norm": 1.964396595954895,
+      "learning_rate": 5e-06,
+      "loss": 0.374,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9158699808795411,
+      "grad_norm": 2.352846145629883,
+      "learning_rate": 5e-06,
+      "loss": 0.2557,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9160611854684513,
+      "grad_norm": 1.851499319076538,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9162523900573614,
+      "grad_norm": 2.526007652282715,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9164435946462716,
+      "grad_norm": 2.6899187564849854,
+      "learning_rate": 5e-06,
+      "loss": 0.2373,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9166347992351817,
+      "grad_norm": 1.4613804817199707,
+      "learning_rate": 5e-06,
+      "loss": 0.0741,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9168260038240917,
+      "grad_norm": 1.7826400995254517,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9170172084130019,
+      "grad_norm": 2.50441575050354,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 4796
+    },
+    {
+      "epoch": 0.917208413001912,
+      "grad_norm": 1.1559553146362305,
+      "learning_rate": 5e-06,
+      "loss": 0.0645,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9173996175908222,
+      "grad_norm": 1.493688702583313,
+      "learning_rate": 5e-06,
+      "loss": 0.1094,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9175908221797323,
+      "grad_norm": 2.530729055404663,
+      "learning_rate": 5e-06,
+      "loss": 0.127,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9177820267686424,
+      "grad_norm": 2.9667983055114746,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9179732313575526,
+      "grad_norm": 1.1197313070297241,
+      "learning_rate": 5e-06,
+      "loss": 0.0769,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9181644359464627,
+      "grad_norm": 1.4685611724853516,
+      "learning_rate": 5e-06,
+      "loss": 0.1881,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9183556405353729,
+      "grad_norm": 1.4080411195755005,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 4803
+    },
+    {
+      "epoch": 0.918546845124283,
+      "grad_norm": 3.9118635654449463,
+      "learning_rate": 5e-06,
+      "loss": 0.3449,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9187380497131931,
+      "grad_norm": 1.7937930822372437,
+      "learning_rate": 5e-06,
+      "loss": 0.2372,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9189292543021033,
+      "grad_norm": 1.3490447998046875,
+      "learning_rate": 5e-06,
+      "loss": 0.0773,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9191204588910133,
+      "grad_norm": 2.1110494136810303,
+      "learning_rate": 5e-06,
+      "loss": 0.2215,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9193116634799235,
+      "grad_norm": 2.0199520587921143,
+      "learning_rate": 5e-06,
+      "loss": 0.2441,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9195028680688336,
+      "grad_norm": 1.7356966733932495,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9196940726577438,
+      "grad_norm": 1.952366828918457,
+      "learning_rate": 5e-06,
+      "loss": 0.1433,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9198852772466539,
+      "grad_norm": 2.1498913764953613,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 4811
+    },
+    {
+      "epoch": 0.920076481835564,
+      "grad_norm": 1.1379492282867432,
+      "learning_rate": 5e-06,
+      "loss": 0.0524,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9202676864244742,
+      "grad_norm": 2.057628870010376,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9204588910133843,
+      "grad_norm": 2.6569430828094482,
+      "learning_rate": 5e-06,
+      "loss": 0.4691,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9206500956022945,
+      "grad_norm": 2.3530871868133545,
+      "learning_rate": 5e-06,
+      "loss": 0.2668,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9208413001912046,
+      "grad_norm": 2.5906331539154053,
+      "learning_rate": 5e-06,
+      "loss": 0.2775,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9210325047801147,
+      "grad_norm": 1.04031503200531,
+      "learning_rate": 5e-06,
+      "loss": 0.0453,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9212237093690249,
+      "grad_norm": 1.8329519033432007,
+      "learning_rate": 5e-06,
+      "loss": 0.0703,
+      "step": 4818
+    },
+    {
+      "epoch": 0.921414913957935,
+      "grad_norm": 1.4083631038665771,
+      "learning_rate": 5e-06,
+      "loss": 0.0891,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9216061185468452,
+      "grad_norm": 2.5658862590789795,
+      "learning_rate": 5e-06,
+      "loss": 0.4482,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9217973231357552,
+      "grad_norm": 2.2376391887664795,
+      "learning_rate": 5e-06,
+      "loss": 0.2053,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9219885277246654,
+      "grad_norm": 1.5132383108139038,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9221797323135755,
+      "grad_norm": 1.9439269304275513,
+      "learning_rate": 5e-06,
+      "loss": 0.2446,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9223709369024856,
+      "grad_norm": 1.8772526979446411,
+      "learning_rate": 5e-06,
+      "loss": 0.0855,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9225621414913958,
+      "grad_norm": 0.8960617184638977,
+      "learning_rate": 5e-06,
+      "loss": 0.0342,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9227533460803059,
+      "grad_norm": 2.050640106201172,
+      "learning_rate": 5e-06,
+      "loss": 0.3478,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9229445506692161,
+      "grad_norm": 1.714992880821228,
+      "learning_rate": 5e-06,
+      "loss": 0.1056,
+      "step": 4827
+    },
+    {
+      "epoch": 0.9231357552581262,
+      "grad_norm": 1.1834768056869507,
+      "learning_rate": 5e-06,
+      "loss": 0.0615,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9233269598470363,
+      "grad_norm": 1.3316147327423096,
+      "learning_rate": 5e-06,
+      "loss": 0.0867,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9235181644359465,
+      "grad_norm": 2.6052935123443604,
+      "learning_rate": 5e-06,
+      "loss": 0.1191,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9237093690248566,
+      "grad_norm": 4.509915828704834,
+      "learning_rate": 5e-06,
+      "loss": 0.3938,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9239005736137668,
+      "grad_norm": 2.1398138999938965,
+      "learning_rate": 5e-06,
+      "loss": 0.3432,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9240917782026769,
+      "grad_norm": 2.1571106910705566,
+      "learning_rate": 5e-06,
+      "loss": 0.3267,
+      "step": 4833
+    },
+    {
+      "epoch": 0.924282982791587,
+      "grad_norm": 1.6129310131072998,
+      "learning_rate": 5e-06,
+      "loss": 0.1886,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9244741873804971,
+      "grad_norm": 1.8231936693191528,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9246653919694072,
+      "grad_norm": 1.4259830713272095,
+      "learning_rate": 5e-06,
+      "loss": 0.065,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9248565965583174,
+      "grad_norm": 1.6335722208023071,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9250478011472275,
+      "grad_norm": 2.2979252338409424,
+      "learning_rate": 5e-06,
+      "loss": 0.0973,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9252390057361377,
+      "grad_norm": 2.0486056804656982,
+      "learning_rate": 5e-06,
+      "loss": 0.4276,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9254302103250478,
+      "grad_norm": 0.9143030047416687,
+      "learning_rate": 5e-06,
+      "loss": 0.0769,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9256214149139579,
+      "grad_norm": 2.2786619663238525,
+      "learning_rate": 5e-06,
+      "loss": 0.2726,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9258126195028681,
+      "grad_norm": 0.9805214405059814,
+      "learning_rate": 5e-06,
+      "loss": 0.0744,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9260038240917782,
+      "grad_norm": 1.4829126596450806,
+      "learning_rate": 5e-06,
+      "loss": 0.0954,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9261950286806884,
+      "grad_norm": 1.7729511260986328,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9263862332695985,
+      "grad_norm": 2.1169512271881104,
+      "learning_rate": 5e-06,
+      "loss": 0.2112,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9265774378585087,
+      "grad_norm": 3.1143903732299805,
+      "learning_rate": 5e-06,
+      "loss": 0.1733,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9267686424474187,
+      "grad_norm": 1.8943371772766113,
+      "learning_rate": 5e-06,
+      "loss": 0.3117,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9269598470363288,
+      "grad_norm": 1.6305577754974365,
+      "learning_rate": 5e-06,
+      "loss": 0.169,
+      "step": 4848
+    },
+    {
+      "epoch": 0.927151051625239,
+      "grad_norm": 0.9189224243164062,
+      "learning_rate": 5e-06,
+      "loss": 0.0763,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9273422562141491,
+      "grad_norm": 2.8332042694091797,
+      "learning_rate": 5e-06,
+      "loss": 0.2152,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9275334608030593,
+      "grad_norm": 5.684455394744873,
+      "learning_rate": 5e-06,
+      "loss": 0.2779,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9277246653919694,
+      "grad_norm": 1.2276694774627686,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9279158699808795,
+      "grad_norm": 2.8665270805358887,
+      "learning_rate": 5e-06,
+      "loss": 0.3275,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9281070745697897,
+      "grad_norm": 1.772354245185852,
+      "learning_rate": 5e-06,
+      "loss": 0.1607,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9282982791586998,
+      "grad_norm": 2.3981382846832275,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 4855
+    },
+    {
+      "epoch": 0.92848948374761,
+      "grad_norm": 1.4854987859725952,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9286806883365201,
+      "grad_norm": 1.3724586963653564,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9288718929254303,
+      "grad_norm": 1.819861650466919,
+      "learning_rate": 5e-06,
+      "loss": 0.2387,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9290630975143404,
+      "grad_norm": 1.742345929145813,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9292543021032504,
+      "grad_norm": 1.5727949142456055,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9294455066921606,
+      "grad_norm": 1.9011776447296143,
+      "learning_rate": 5e-06,
+      "loss": 0.2739,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9296367112810707,
+      "grad_norm": 1.4754587411880493,
+      "learning_rate": 5e-06,
+      "loss": 0.111,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9298279158699809,
+      "grad_norm": 1.5760308504104614,
+      "learning_rate": 5e-06,
+      "loss": 0.1089,
+      "step": 4863
+    },
+    {
+      "epoch": 0.930019120458891,
+      "grad_norm": 2.719432830810547,
+      "learning_rate": 5e-06,
+      "loss": 0.4185,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9302103250478011,
+      "grad_norm": 2.2570371627807617,
+      "learning_rate": 5e-06,
+      "loss": 0.1839,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9304015296367113,
+      "grad_norm": 1.3954803943634033,
+      "learning_rate": 5e-06,
+      "loss": 0.099,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9305927342256214,
+      "grad_norm": 1.2225748300552368,
+      "learning_rate": 5e-06,
+      "loss": 0.1137,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9307839388145316,
+      "grad_norm": 0.926919162273407,
+      "learning_rate": 5e-06,
+      "loss": 0.0477,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9309751434034417,
+      "grad_norm": 1.8387597799301147,
+      "learning_rate": 5e-06,
+      "loss": 0.0733,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9311663479923518,
+      "grad_norm": 1.1521694660186768,
+      "learning_rate": 5e-06,
+      "loss": 0.1333,
+      "step": 4870
+    },
+    {
+      "epoch": 0.931357552581262,
+      "grad_norm": 1.0635244846343994,
+      "learning_rate": 5e-06,
+      "loss": 0.1047,
+      "step": 4871
+    },
+    {
+      "epoch": 0.931548757170172,
+      "grad_norm": 1.6397044658660889,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9317399617590822,
+      "grad_norm": 1.338733434677124,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9319311663479923,
+      "grad_norm": 1.307789921760559,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9321223709369025,
+      "grad_norm": 2.0790135860443115,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9323135755258126,
+      "grad_norm": 1.5820248126983643,
+      "learning_rate": 5e-06,
+      "loss": 0.206,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9325047801147227,
+      "grad_norm": 1.5666821002960205,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9326959847036329,
+      "grad_norm": 1.4068495035171509,
+      "learning_rate": 5e-06,
+      "loss": 0.2069,
+      "step": 4878
+    },
+    {
+      "epoch": 0.932887189292543,
+      "grad_norm": 1.985937237739563,
+      "learning_rate": 5e-06,
+      "loss": 0.1667,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9330783938814532,
+      "grad_norm": 1.2681323289871216,
+      "learning_rate": 5e-06,
+      "loss": 0.0669,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9332695984703633,
+      "grad_norm": 1.4360939264297485,
+      "learning_rate": 5e-06,
+      "loss": 0.0561,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9334608030592734,
+      "grad_norm": 1.2678735256195068,
+      "learning_rate": 5e-06,
+      "loss": 0.1019,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9336520076481836,
+      "grad_norm": 2.8025400638580322,
+      "learning_rate": 5e-06,
+      "loss": 0.4316,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9338432122370937,
+      "grad_norm": 2.1860554218292236,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9340344168260039,
+      "grad_norm": 1.1986353397369385,
+      "learning_rate": 5e-06,
+      "loss": 0.099,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9342256214149139,
+      "grad_norm": 0.7431411743164062,
+      "learning_rate": 5e-06,
+      "loss": 0.0325,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9344168260038241,
+      "grad_norm": 1.0779439210891724,
+      "learning_rate": 5e-06,
+      "loss": 0.0742,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9346080305927342,
+      "grad_norm": 1.574372410774231,
+      "learning_rate": 5e-06,
+      "loss": 0.1254,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9347992351816443,
+      "grad_norm": 1.4904820919036865,
+      "learning_rate": 5e-06,
+      "loss": 0.1166,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9349904397705545,
+      "grad_norm": 3.761897325515747,
+      "learning_rate": 5e-06,
+      "loss": 0.5072,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9351816443594646,
+      "grad_norm": 2.5792999267578125,
+      "learning_rate": 5e-06,
+      "loss": 0.3623,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9353728489483748,
+      "grad_norm": 2.6418819427490234,
+      "learning_rate": 5e-06,
+      "loss": 0.1922,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9355640535372849,
+      "grad_norm": 2.635206937789917,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 4893
+    },
+    {
+      "epoch": 0.935755258126195,
+      "grad_norm": 2.728708028793335,
+      "learning_rate": 5e-06,
+      "loss": 0.3097,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9359464627151052,
+      "grad_norm": 2.1992111206054688,
+      "learning_rate": 5e-06,
+      "loss": 0.2703,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9361376673040153,
+      "grad_norm": 1.2282865047454834,
+      "learning_rate": 5e-06,
+      "loss": 0.1054,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9363288718929255,
+      "grad_norm": 1.8844170570373535,
+      "learning_rate": 5e-06,
+      "loss": 0.1161,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9365200764818356,
+      "grad_norm": 1.5336863994598389,
+      "learning_rate": 5e-06,
+      "loss": 0.109,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9367112810707457,
+      "grad_norm": 1.703996181488037,
+      "learning_rate": 5e-06,
+      "loss": 0.081,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9369024856596558,
+      "grad_norm": 1.7110828161239624,
+      "learning_rate": 5e-06,
+      "loss": 0.1016,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9370936902485659,
+      "grad_norm": 1.9495599269866943,
+      "learning_rate": 5e-06,
+      "loss": 0.3473,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9372848948374761,
+      "grad_norm": 2.5797476768493652,
+      "learning_rate": 5e-06,
+      "loss": 0.4117,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9374760994263862,
+      "grad_norm": 2.9593513011932373,
+      "learning_rate": 5e-06,
+      "loss": 0.3348,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9376673040152964,
+      "grad_norm": 1.2573286294937134,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9378585086042065,
+      "grad_norm": 0.9126336574554443,
+      "learning_rate": 5e-06,
+      "loss": 0.0479,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9380497131931166,
+      "grad_norm": 2.2401199340820312,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9382409177820268,
+      "grad_norm": 1.2739428281784058,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9384321223709369,
+      "grad_norm": 1.4765453338623047,
+      "learning_rate": 5e-06,
+      "loss": 0.1207,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9386233269598471,
+      "grad_norm": 2.1147773265838623,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9388145315487572,
+      "grad_norm": 1.5339463949203491,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9390057361376674,
+      "grad_norm": 2.7874417304992676,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9391969407265774,
+      "grad_norm": 1.52443265914917,
+      "learning_rate": 5e-06,
+      "loss": 0.0438,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9393881453154875,
+      "grad_norm": 2.6554689407348633,
+      "learning_rate": 5e-06,
+      "loss": 0.4022,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9395793499043977,
+      "grad_norm": 2.2543184757232666,
+      "learning_rate": 5e-06,
+      "loss": 0.3544,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9397705544933078,
+      "grad_norm": 3.414179563522339,
+      "learning_rate": 5e-06,
+      "loss": 0.5646,
+      "step": 4915
+    },
+    {
+      "epoch": 0.939961759082218,
+      "grad_norm": 2.3218886852264404,
+      "learning_rate": 5e-06,
+      "loss": 0.2771,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9401529636711281,
+      "grad_norm": 1.0181642770767212,
+      "learning_rate": 5e-06,
+      "loss": 0.0543,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9403441682600382,
+      "grad_norm": 2.524704933166504,
+      "learning_rate": 5e-06,
+      "loss": 0.1449,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9405353728489484,
+      "grad_norm": 2.3251535892486572,
+      "learning_rate": 5e-06,
+      "loss": 0.1364,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9407265774378585,
+      "grad_norm": 2.4270453453063965,
+      "learning_rate": 5e-06,
+      "loss": 0.2367,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9409177820267687,
+      "grad_norm": 1.9268070459365845,
+      "learning_rate": 5e-06,
+      "loss": 0.2149,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9411089866156788,
+      "grad_norm": 1.8743796348571777,
+      "learning_rate": 5e-06,
+      "loss": 0.3339,
+      "step": 4922
+    },
+    {
+      "epoch": 0.941300191204589,
+      "grad_norm": 1.756379246711731,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 4923
+    },
+    {
+      "epoch": 0.941491395793499,
+      "grad_norm": 0.8560100793838501,
+      "learning_rate": 5e-06,
+      "loss": 0.0507,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9416826003824091,
+      "grad_norm": 2.884312391281128,
+      "learning_rate": 5e-06,
+      "loss": 0.2991,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9418738049713193,
+      "grad_norm": 1.6877045631408691,
+      "learning_rate": 5e-06,
+      "loss": 0.2015,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9420650095602294,
+      "grad_norm": 1.8372879028320312,
+      "learning_rate": 5e-06,
+      "loss": 0.1594,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9422562141491396,
+      "grad_norm": 1.6219450235366821,
+      "learning_rate": 5e-06,
+      "loss": 0.1066,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9424474187380497,
+      "grad_norm": 3.6380860805511475,
+      "learning_rate": 5e-06,
+      "loss": 0.541,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9426386233269598,
+      "grad_norm": 2.0586066246032715,
+      "learning_rate": 5e-06,
+      "loss": 0.1329,
+      "step": 4930
+    },
+    {
+      "epoch": 0.94282982791587,
+      "grad_norm": 1.7319679260253906,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9430210325047801,
+      "grad_norm": 2.5363359451293945,
+      "learning_rate": 5e-06,
+      "loss": 0.4112,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9432122370936903,
+      "grad_norm": 1.9458513259887695,
+      "learning_rate": 5e-06,
+      "loss": 0.2529,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9434034416826004,
+      "grad_norm": 3.23052716255188,
+      "learning_rate": 5e-06,
+      "loss": 0.5106,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9435946462715105,
+      "grad_norm": 1.5255218744277954,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9437858508604207,
+      "grad_norm": 2.298287868499756,
+      "learning_rate": 5e-06,
+      "loss": 0.3477,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9439770554493307,
+      "grad_norm": 1.1250810623168945,
+      "learning_rate": 5e-06,
+      "loss": 0.054,
+      "step": 4937
+    },
+    {
+      "epoch": 0.944168260038241,
+      "grad_norm": 3.244180679321289,
+      "learning_rate": 5e-06,
+      "loss": 0.2005,
+      "step": 4938
+    },
+    {
+      "epoch": 0.944359464627151,
+      "grad_norm": 3.1849043369293213,
+      "learning_rate": 5e-06,
+      "loss": 0.355,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9445506692160612,
+      "grad_norm": 2.582594156265259,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9447418738049713,
+      "grad_norm": 1.0963287353515625,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9449330783938814,
+      "grad_norm": 2.1446595191955566,
+      "learning_rate": 5e-06,
+      "loss": 0.2562,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9451242829827916,
+      "grad_norm": 1.8604848384857178,
+      "learning_rate": 5e-06,
+      "loss": 0.0994,
+      "step": 4943
+    },
+    {
+      "epoch": 0.9453154875717017,
+      "grad_norm": 1.9539889097213745,
+      "learning_rate": 5e-06,
+      "loss": 0.0991,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9455066921606119,
+      "grad_norm": 2.2747185230255127,
+      "learning_rate": 5e-06,
+      "loss": 0.2482,
+      "step": 4945
+    },
+    {
+      "epoch": 0.945697896749522,
+      "grad_norm": 1.741715669631958,
+      "learning_rate": 5e-06,
+      "loss": 0.2089,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9458891013384321,
+      "grad_norm": 1.2032427787780762,
+      "learning_rate": 5e-06,
+      "loss": 0.095,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9460803059273423,
+      "grad_norm": 1.8849596977233887,
+      "learning_rate": 5e-06,
+      "loss": 0.0805,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9462715105162524,
+      "grad_norm": 1.4403064250946045,
+      "learning_rate": 5e-06,
+      "loss": 0.1019,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9464627151051626,
+      "grad_norm": 1.0823032855987549,
+      "learning_rate": 5e-06,
+      "loss": 0.059,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9466539196940726,
+      "grad_norm": 3.9184749126434326,
+      "learning_rate": 5e-06,
+      "loss": 0.7392,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9468451242829828,
+      "grad_norm": 10.959403991699219,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9470363288718929,
+      "grad_norm": 2.1564931869506836,
+      "learning_rate": 5e-06,
+      "loss": 0.2236,
+      "step": 4953
+    },
+    {
+      "epoch": 0.947227533460803,
+      "grad_norm": 1.5129090547561646,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9474187380497132,
+      "grad_norm": 0.9726706147193909,
+      "learning_rate": 5e-06,
+      "loss": 0.0421,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9476099426386233,
+      "grad_norm": 1.2113652229309082,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9478011472275335,
+      "grad_norm": 1.5728042125701904,
+      "learning_rate": 5e-06,
+      "loss": 0.1368,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9479923518164436,
+      "grad_norm": 1.3641796112060547,
+      "learning_rate": 5e-06,
+      "loss": 0.1675,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9481835564053537,
+      "grad_norm": 1.7657511234283447,
+      "learning_rate": 5e-06,
+      "loss": 0.2036,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9483747609942639,
+      "grad_norm": 0.9892101883888245,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 4960
+    },
+    {
+      "epoch": 0.948565965583174,
+      "grad_norm": 2.8380050659179688,
+      "learning_rate": 5e-06,
+      "loss": 0.3084,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9487571701720842,
+      "grad_norm": 0.8977944850921631,
+      "learning_rate": 5e-06,
+      "loss": 0.0524,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9489483747609943,
+      "grad_norm": 2.5370876789093018,
+      "learning_rate": 5e-06,
+      "loss": 0.407,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9491395793499044,
+      "grad_norm": 2.922842502593994,
+      "learning_rate": 5e-06,
+      "loss": 0.4577,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9493307839388145,
+      "grad_norm": 1.981605887413025,
+      "learning_rate": 5e-06,
+      "loss": 0.2596,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9495219885277246,
+      "grad_norm": 2.6923184394836426,
+      "learning_rate": 5e-06,
+      "loss": 0.4727,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9497131931166348,
+      "grad_norm": 1.4957002401351929,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9499043977055449,
+      "grad_norm": 1.2848210334777832,
+      "learning_rate": 5e-06,
+      "loss": 0.078,
+      "step": 4968
+    },
+    {
+      "epoch": 0.9500956022944551,
+      "grad_norm": 1.8343178033828735,
+      "learning_rate": 5e-06,
+      "loss": 0.1058,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9502868068833652,
+      "grad_norm": 1.764131784439087,
+      "learning_rate": 5e-06,
+      "loss": 0.2088,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9504780114722753,
+      "grad_norm": 1.4648293256759644,
+      "learning_rate": 5e-06,
+      "loss": 0.1017,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9506692160611855,
+      "grad_norm": 1.3409184217453003,
+      "learning_rate": 5e-06,
+      "loss": 0.0881,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9508604206500956,
+      "grad_norm": 1.3016523122787476,
+      "learning_rate": 5e-06,
+      "loss": 0.0451,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9510516252390058,
+      "grad_norm": 0.7061133980751038,
+      "learning_rate": 5e-06,
+      "loss": 0.0428,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9512428298279159,
+      "grad_norm": 3.5251235961914062,
+      "learning_rate": 5e-06,
+      "loss": 0.2068,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9514340344168261,
+      "grad_norm": 2.1218931674957275,
+      "learning_rate": 5e-06,
+      "loss": 0.3303,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9516252390057361,
+      "grad_norm": 2.1984875202178955,
+      "learning_rate": 5e-06,
+      "loss": 0.3954,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9518164435946462,
+      "grad_norm": 1.5826531648635864,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9520076481835564,
+      "grad_norm": 1.824207067489624,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9521988527724665,
+      "grad_norm": 1.0942171812057495,
+      "learning_rate": 5e-06,
+      "loss": 0.0596,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9523900573613767,
+      "grad_norm": 2.183330535888672,
+      "learning_rate": 5e-06,
+      "loss": 0.1098,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9525812619502868,
+      "grad_norm": 1.866705060005188,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9527724665391969,
+      "grad_norm": 2.1280243396759033,
+      "learning_rate": 5e-06,
+      "loss": 0.2009,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9529636711281071,
+      "grad_norm": 1.8956713676452637,
+      "learning_rate": 5e-06,
+      "loss": 0.1752,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9531548757170172,
+      "grad_norm": 2.302192449569702,
+      "learning_rate": 5e-06,
+      "loss": 0.2851,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9533460803059274,
+      "grad_norm": 4.471398830413818,
+      "learning_rate": 5e-06,
+      "loss": 0.1626,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9535372848948375,
+      "grad_norm": 2.243290424346924,
+      "learning_rate": 5e-06,
+      "loss": 0.1037,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9537284894837476,
+      "grad_norm": 2.0457358360290527,
+      "learning_rate": 5e-06,
+      "loss": 0.2751,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9539196940726578,
+      "grad_norm": 1.7639656066894531,
+      "learning_rate": 5e-06,
+      "loss": 0.2408,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9541108986615678,
+      "grad_norm": 2.313894748687744,
+      "learning_rate": 5e-06,
+      "loss": 0.2289,
+      "step": 4990
+    },
+    {
+      "epoch": 0.954302103250478,
+      "grad_norm": 1.8488069772720337,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 4991
+    },
+    {
+      "epoch": 0.9544933078393881,
+      "grad_norm": 1.261997938156128,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9546845124282983,
+      "grad_norm": 0.8408591747283936,
+      "learning_rate": 5e-06,
+      "loss": 0.0461,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9548757170172084,
+      "grad_norm": 2.0298595428466797,
+      "learning_rate": 5e-06,
+      "loss": 0.0984,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9550669216061185,
+      "grad_norm": 2.701509714126587,
+      "learning_rate": 5e-06,
+      "loss": 0.3663,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9552581261950287,
+      "grad_norm": 1.558736801147461,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9554493307839388,
+      "grad_norm": 2.992225170135498,
+      "learning_rate": 5e-06,
+      "loss": 0.4301,
+      "step": 4997
+    },
+    {
+      "epoch": 0.955640535372849,
+      "grad_norm": 1.2962523698806763,
+      "learning_rate": 5e-06,
+      "loss": 0.0935,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9558317399617591,
+      "grad_norm": 1.3460264205932617,
+      "learning_rate": 5e-06,
+      "loss": 0.0787,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9560229445506692,
+      "grad_norm": 2.1929922103881836,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9560229445506692,
+      "eval_runtime": 742.0542,
+      "eval_samples_per_second": 2.067,
+      "eval_steps_per_second": 0.259,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9562141491395794,
+      "grad_norm": 1.8363579511642456,
+      "learning_rate": 5e-06,
+      "loss": 0.1593,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9564053537284894,
+      "grad_norm": 2.0634498596191406,
+      "learning_rate": 5e-06,
+      "loss": 0.1911,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9565965583173996,
+      "grad_norm": 1.5266684293746948,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9567877629063097,
+      "grad_norm": 1.1533805131912231,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9569789674952199,
+      "grad_norm": 2.2362139225006104,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 5005
+    },
+    {
+      "epoch": 0.95717017208413,
+      "grad_norm": 2.0843067169189453,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9573613766730401,
+      "grad_norm": 2.134740114212036,
+      "learning_rate": 5e-06,
+      "loss": 0.3089,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9575525812619503,
+      "grad_norm": 2.9222230911254883,
+      "learning_rate": 5e-06,
+      "loss": 0.4906,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9577437858508604,
+      "grad_norm": 2.486271619796753,
+      "learning_rate": 5e-06,
+      "loss": 0.3661,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9579349904397706,
+      "grad_norm": 0.9269048571586609,
+      "learning_rate": 5e-06,
+      "loss": 0.1277,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9581261950286807,
+      "grad_norm": 2.4355266094207764,
+      "learning_rate": 5e-06,
+      "loss": 0.1951,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9583173996175908,
+      "grad_norm": 2.3811750411987305,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 5012
+    },
+    {
+      "epoch": 0.958508604206501,
+      "grad_norm": 2.0431880950927734,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9586998087954111,
+      "grad_norm": 3.104672908782959,
+      "learning_rate": 5e-06,
+      "loss": 0.4546,
+      "step": 5014
+    },
+    {
+      "epoch": 0.9588910133843213,
+      "grad_norm": 2.163524866104126,
+      "learning_rate": 5e-06,
+      "loss": 0.1894,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9590822179732313,
+      "grad_norm": 2.0325043201446533,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9592734225621415,
+      "grad_norm": 2.2608044147491455,
+      "learning_rate": 5e-06,
+      "loss": 0.2509,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9594646271510516,
+      "grad_norm": 1.2994120121002197,
+      "learning_rate": 5e-06,
+      "loss": 0.0878,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9596558317399617,
+      "grad_norm": 1.0766092538833618,
+      "learning_rate": 5e-06,
+      "loss": 0.0688,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9598470363288719,
+      "grad_norm": 2.5063343048095703,
+      "learning_rate": 5e-06,
+      "loss": 0.3423,
+      "step": 5020
+    },
+    {
+      "epoch": 0.960038240917782,
+      "grad_norm": 2.1994948387145996,
+      "learning_rate": 5e-06,
+      "loss": 0.2923,
+      "step": 5021
+    },
+    {
+      "epoch": 0.9602294455066922,
+      "grad_norm": 3.1960678100585938,
+      "learning_rate": 5e-06,
+      "loss": 0.4548,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9604206500956023,
+      "grad_norm": 1.9123889207839966,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 5023
+    },
+    {
+      "epoch": 0.9606118546845124,
+      "grad_norm": 1.6679797172546387,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9608030592734226,
+      "grad_norm": 1.2722249031066895,
+      "learning_rate": 5e-06,
+      "loss": 0.0643,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9609942638623327,
+      "grad_norm": 1.5316091775894165,
+      "learning_rate": 5e-06,
+      "loss": 0.2274,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9611854684512429,
+      "grad_norm": 1.5301028490066528,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 5027
+    },
+    {
+      "epoch": 0.961376673040153,
+      "grad_norm": 1.963606357574463,
+      "learning_rate": 5e-06,
+      "loss": 0.2119,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9615678776290632,
+      "grad_norm": 1.5402876138687134,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9617590822179732,
+      "grad_norm": 1.1363723278045654,
+      "learning_rate": 5e-06,
+      "loss": 0.0777,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9619502868068833,
+      "grad_norm": 1.909873604774475,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9621414913957935,
+      "grad_norm": 1.3693710565567017,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9623326959847036,
+      "grad_norm": 2.0859200954437256,
+      "learning_rate": 5e-06,
+      "loss": 0.2549,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9625239005736138,
+      "grad_norm": 1.5054364204406738,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 5034
+    },
+    {
+      "epoch": 0.9627151051625239,
+      "grad_norm": 1.8584833145141602,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 5035
+    },
+    {
+      "epoch": 0.962906309751434,
+      "grad_norm": 1.1875016689300537,
+      "learning_rate": 5e-06,
+      "loss": 0.0729,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9630975143403442,
+      "grad_norm": 1.1567950248718262,
+      "learning_rate": 5e-06,
+      "loss": 0.0653,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9632887189292543,
+      "grad_norm": 1.8108049631118774,
+      "learning_rate": 5e-06,
+      "loss": 0.1345,
+      "step": 5038
+    },
+    {
+      "epoch": 0.9634799235181645,
+      "grad_norm": 3.5861775875091553,
+      "learning_rate": 5e-06,
+      "loss": 0.5882,
+      "step": 5039
+    },
+    {
+      "epoch": 0.9636711281070746,
+      "grad_norm": 2.6124978065490723,
+      "learning_rate": 5e-06,
+      "loss": 0.1039,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9638623326959848,
+      "grad_norm": 1.3670365810394287,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9640535372848948,
+      "grad_norm": 1.332395315170288,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 5042
+    },
+    {
+      "epoch": 0.9642447418738049,
+      "grad_norm": 1.1727179288864136,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9644359464627151,
+      "grad_norm": 1.3848689794540405,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 5044
+    },
+    {
+      "epoch": 0.9646271510516252,
+      "grad_norm": 1.4246430397033691,
+      "learning_rate": 5e-06,
+      "loss": 0.099,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9648183556405354,
+      "grad_norm": 2.0380055904388428,
+      "learning_rate": 5e-06,
+      "loss": 0.2963,
+      "step": 5046
+    },
+    {
+      "epoch": 0.9650095602294455,
+      "grad_norm": 2.6432945728302,
+      "learning_rate": 5e-06,
+      "loss": 0.236,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9652007648183556,
+      "grad_norm": 1.9849876165390015,
+      "learning_rate": 5e-06,
+      "loss": 0.2692,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9653919694072658,
+      "grad_norm": 1.9646875858306885,
+      "learning_rate": 5e-06,
+      "loss": 0.3539,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9655831739961759,
+      "grad_norm": 2.007697820663452,
+      "learning_rate": 5e-06,
+      "loss": 0.1226,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9657743785850861,
+      "grad_norm": 1.224379062652588,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9659655831739962,
+      "grad_norm": 0.585542619228363,
+      "learning_rate": 5e-06,
+      "loss": 0.0688,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9661567877629063,
+      "grad_norm": 1.7798622846603394,
+      "learning_rate": 5e-06,
+      "loss": 0.2512,
+      "step": 5053
+    },
+    {
+      "epoch": 0.9663479923518165,
+      "grad_norm": 1.9933887720108032,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9665391969407265,
+      "grad_norm": 1.6793018579483032,
+      "learning_rate": 5e-06,
+      "loss": 0.1758,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9667304015296367,
+      "grad_norm": 1.5060217380523682,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9669216061185468,
+      "grad_norm": 2.19502329826355,
+      "learning_rate": 5e-06,
+      "loss": 0.2314,
+      "step": 5057
+    },
+    {
+      "epoch": 0.967112810707457,
+      "grad_norm": 1.7284197807312012,
+      "learning_rate": 5e-06,
+      "loss": 0.1213,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9673040152963671,
+      "grad_norm": 1.6519883871078491,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9674952198852772,
+      "grad_norm": 2.585042715072632,
+      "learning_rate": 5e-06,
+      "loss": 0.2157,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9676864244741874,
+      "grad_norm": 2.122485637664795,
+      "learning_rate": 5e-06,
+      "loss": 0.3705,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9678776290630975,
+      "grad_norm": 1.6645944118499756,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 5062
+    },
+    {
+      "epoch": 0.9680688336520077,
+      "grad_norm": 2.446328639984131,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9682600382409178,
+      "grad_norm": 3.5966954231262207,
+      "learning_rate": 5e-06,
+      "loss": 0.5335,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9684512428298279,
+      "grad_norm": 1.5934321880340576,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9686424474187381,
+      "grad_norm": 2.7553305625915527,
+      "learning_rate": 5e-06,
+      "loss": 0.2518,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9688336520076481,
+      "grad_norm": 1.3165737390518188,
+      "learning_rate": 5e-06,
+      "loss": 0.0789,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9690248565965583,
+      "grad_norm": 1.7967562675476074,
+      "learning_rate": 5e-06,
+      "loss": 0.1001,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9692160611854684,
+      "grad_norm": 1.7139383554458618,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9694072657743786,
+      "grad_norm": 2.3538708686828613,
+      "learning_rate": 5e-06,
+      "loss": 0.3276,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9695984703632887,
+      "grad_norm": 2.8249590396881104,
+      "learning_rate": 5e-06,
+      "loss": 0.446,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9697896749521988,
+      "grad_norm": 1.2258639335632324,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 5072
+    },
+    {
+      "epoch": 0.969980879541109,
+      "grad_norm": 1.419464111328125,
+      "learning_rate": 5e-06,
+      "loss": 0.0757,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9701720841300191,
+      "grad_norm": 1.7254623174667358,
+      "learning_rate": 5e-06,
+      "loss": 0.0864,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9703632887189293,
+      "grad_norm": 2.9186739921569824,
+      "learning_rate": 5e-06,
+      "loss": 0.2495,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9705544933078394,
+      "grad_norm": 1.3285423517227173,
+      "learning_rate": 5e-06,
+      "loss": 0.1859,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9707456978967495,
+      "grad_norm": 1.7535793781280518,
+      "learning_rate": 5e-06,
+      "loss": 0.2977,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9709369024856597,
+      "grad_norm": 1.533988356590271,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 5078
+    },
+    {
+      "epoch": 0.9711281070745698,
+      "grad_norm": 1.5041394233703613,
+      "learning_rate": 5e-06,
+      "loss": 0.142,
+      "step": 5079
+    },
+    {
+      "epoch": 0.97131931166348,
+      "grad_norm": 1.4372128248214722,
+      "learning_rate": 5e-06,
+      "loss": 0.0904,
+      "step": 5080
+    },
+    {
+      "epoch": 0.97151051625239,
+      "grad_norm": 2.476388454437256,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9717017208413002,
+      "grad_norm": 2.5645976066589355,
+      "learning_rate": 5e-06,
+      "loss": 0.2597,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9718929254302103,
+      "grad_norm": 1.5782102346420288,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9720841300191204,
+      "grad_norm": 1.2604018449783325,
+      "learning_rate": 5e-06,
+      "loss": 0.2281,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9722753346080306,
+      "grad_norm": 1.2508944272994995,
+      "learning_rate": 5e-06,
+      "loss": 0.0719,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9724665391969407,
+      "grad_norm": 2.0485973358154297,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9726577437858509,
+      "grad_norm": 1.1528031826019287,
+      "learning_rate": 5e-06,
+      "loss": 0.0832,
+      "step": 5087
+    },
+    {
+      "epoch": 0.972848948374761,
+      "grad_norm": 1.789041519165039,
+      "learning_rate": 5e-06,
+      "loss": 0.1181,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9730401529636711,
+      "grad_norm": 1.92593514919281,
+      "learning_rate": 5e-06,
+      "loss": 0.2047,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9732313575525813,
+      "grad_norm": 1.1563007831573486,
+      "learning_rate": 5e-06,
+      "loss": 0.1133,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9734225621414914,
+      "grad_norm": 1.1158252954483032,
+      "learning_rate": 5e-06,
+      "loss": 0.0862,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9736137667304016,
+      "grad_norm": 1.5522339344024658,
+      "learning_rate": 5e-06,
+      "loss": 0.103,
+      "step": 5092
+    },
+    {
+      "epoch": 0.9738049713193117,
+      "grad_norm": 2.535926342010498,
+      "learning_rate": 5e-06,
+      "loss": 0.2977,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9739961759082219,
+      "grad_norm": 1.2164846658706665,
+      "learning_rate": 5e-06,
+      "loss": 0.0662,
+      "step": 5094
+    },
+    {
+      "epoch": 0.9741873804971319,
+      "grad_norm": 2.7936418056488037,
+      "learning_rate": 5e-06,
+      "loss": 0.4989,
+      "step": 5095
+    },
+    {
+      "epoch": 0.974378585086042,
+      "grad_norm": 2.6002554893493652,
+      "learning_rate": 5e-06,
+      "loss": 0.3435,
+      "step": 5096
+    },
+    {
+      "epoch": 0.9745697896749522,
+      "grad_norm": 2.3612749576568604,
+      "learning_rate": 5e-06,
+      "loss": 0.2797,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9747609942638623,
+      "grad_norm": 2.0107052326202393,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9749521988527725,
+      "grad_norm": 1.1361372470855713,
+      "learning_rate": 5e-06,
+      "loss": 0.0784,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9751434034416826,
+      "grad_norm": 1.9277843236923218,
+      "learning_rate": 5e-06,
+      "loss": 0.1257,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9753346080305927,
+      "grad_norm": 1.3512511253356934,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 5101
+    },
+    {
+      "epoch": 0.9755258126195029,
+      "grad_norm": 3.0426383018493652,
+      "learning_rate": 5e-06,
+      "loss": 0.4038,
+      "step": 5102
+    },
+    {
+      "epoch": 0.975717017208413,
+      "grad_norm": 2.047166347503662,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9759082217973232,
+      "grad_norm": 1.3602712154388428,
+      "learning_rate": 5e-06,
+      "loss": 0.0953,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9760994263862333,
+      "grad_norm": 2.696289539337158,
+      "learning_rate": 5e-06,
+      "loss": 0.1088,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9762906309751435,
+      "grad_norm": 1.646152138710022,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9764818355640535,
+      "grad_norm": 1.7276148796081543,
+      "learning_rate": 5e-06,
+      "loss": 0.1875,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9766730401529636,
+      "grad_norm": 2.928664445877075,
+      "learning_rate": 5e-06,
+      "loss": 0.1254,
+      "step": 5108
+    },
+    {
+      "epoch": 0.9768642447418738,
+      "grad_norm": 2.494978904724121,
+      "learning_rate": 5e-06,
+      "loss": 0.3259,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9770554493307839,
+      "grad_norm": 1.8363969326019287,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 5110
+    },
+    {
+      "epoch": 0.9772466539196941,
+      "grad_norm": 3.1696293354034424,
+      "learning_rate": 5e-06,
+      "loss": 0.2118,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9774378585086042,
+      "grad_norm": 1.7869844436645508,
+      "learning_rate": 5e-06,
+      "loss": 0.1343,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9776290630975143,
+      "grad_norm": 1.5095562934875488,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9778202676864245,
+      "grad_norm": 1.0771520137786865,
+      "learning_rate": 5e-06,
+      "loss": 0.1536,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9780114722753346,
+      "grad_norm": 1.9261468648910522,
+      "learning_rate": 5e-06,
+      "loss": 0.2512,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9782026768642448,
+      "grad_norm": 2.3080861568450928,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9783938814531549,
+      "grad_norm": 2.382253885269165,
+      "learning_rate": 5e-06,
+      "loss": 0.1637,
+      "step": 5117
+    },
+    {
+      "epoch": 0.978585086042065,
+      "grad_norm": 2.896066665649414,
+      "learning_rate": 5e-06,
+      "loss": 0.3333,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9787762906309752,
+      "grad_norm": 1.2175503969192505,
+      "learning_rate": 5e-06,
+      "loss": 0.0933,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9789674952198852,
+      "grad_norm": 1.7644296884536743,
+      "learning_rate": 5e-06,
+      "loss": 0.1718,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9791586998087954,
+      "grad_norm": 1.8412610292434692,
+      "learning_rate": 5e-06,
+      "loss": 0.1739,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9793499043977055,
+      "grad_norm": 1.3663194179534912,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9795411089866157,
+      "grad_norm": 1.2034631967544556,
+      "learning_rate": 5e-06,
+      "loss": 0.1127,
+      "step": 5123
+    },
+    {
+      "epoch": 0.9797323135755258,
+      "grad_norm": 2.052450180053711,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9799235181644359,
+      "grad_norm": 2.3145527839660645,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9801147227533461,
+      "grad_norm": 1.6712604761123657,
+      "learning_rate": 5e-06,
+      "loss": 0.2101,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9803059273422562,
+      "grad_norm": 2.388617992401123,
+      "learning_rate": 5e-06,
+      "loss": 0.2405,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9804971319311664,
+      "grad_norm": 3.332637071609497,
+      "learning_rate": 5e-06,
+      "loss": 0.286,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9806883365200765,
+      "grad_norm": 2.245945930480957,
+      "learning_rate": 5e-06,
+      "loss": 0.1992,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9808795411089866,
+      "grad_norm": 2.0014476776123047,
+      "learning_rate": 5e-06,
+      "loss": 0.2369,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9810707456978968,
+      "grad_norm": 2.185051441192627,
+      "learning_rate": 5e-06,
+      "loss": 0.124,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9812619502868068,
+      "grad_norm": 3.205618381500244,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 5132
+    },
+    {
+      "epoch": 0.981453154875717,
+      "grad_norm": 2.2920567989349365,
+      "learning_rate": 5e-06,
+      "loss": 0.3075,
+      "step": 5133
+    },
+    {
+      "epoch": 0.9816443594646271,
+      "grad_norm": 1.5987259149551392,
+      "learning_rate": 5e-06,
+      "loss": 0.0987,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9818355640535373,
+      "grad_norm": 1.4586559534072876,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9820267686424474,
+      "grad_norm": 1.3265559673309326,
+      "learning_rate": 5e-06,
+      "loss": 0.0885,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9822179732313575,
+      "grad_norm": 1.8111308813095093,
+      "learning_rate": 5e-06,
+      "loss": 0.098,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9824091778202677,
+      "grad_norm": 1.9559201002120972,
+      "learning_rate": 5e-06,
+      "loss": 0.1827,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9826003824091778,
+      "grad_norm": 2.519045829772949,
+      "learning_rate": 5e-06,
+      "loss": 0.4279,
+      "step": 5139
+    },
+    {
+      "epoch": 0.982791586998088,
+      "grad_norm": 1.0115805864334106,
+      "learning_rate": 5e-06,
+      "loss": 0.0554,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9829827915869981,
+      "grad_norm": 1.5444289445877075,
+      "learning_rate": 5e-06,
+      "loss": 0.1956,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9831739961759082,
+      "grad_norm": 1.1781493425369263,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 5142
+    },
+    {
+      "epoch": 0.9833652007648184,
+      "grad_norm": 1.6144378185272217,
+      "learning_rate": 5e-06,
+      "loss": 0.1011,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9835564053537285,
+      "grad_norm": 1.777444839477539,
+      "learning_rate": 5e-06,
+      "loss": 0.3058,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9837476099426387,
+      "grad_norm": 2.543937921524048,
+      "learning_rate": 5e-06,
+      "loss": 0.3751,
+      "step": 5145
+    },
+    {
+      "epoch": 0.9839388145315487,
+      "grad_norm": 2.032430648803711,
+      "learning_rate": 5e-06,
+      "loss": 0.2294,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9841300191204589,
+      "grad_norm": 1.2935893535614014,
+      "learning_rate": 5e-06,
+      "loss": 0.1531,
+      "step": 5147
+    },
+    {
+      "epoch": 0.984321223709369,
+      "grad_norm": 1.9025208950042725,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9845124282982791,
+      "grad_norm": 3.2536699771881104,
+      "learning_rate": 5e-06,
+      "loss": 0.3226,
+      "step": 5149
+    },
+    {
+      "epoch": 0.9847036328871893,
+      "grad_norm": 2.4361836910247803,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9848948374760994,
+      "grad_norm": 1.2773830890655518,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9850860420650096,
+      "grad_norm": 2.558335542678833,
+      "learning_rate": 5e-06,
+      "loss": 0.2727,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9852772466539197,
+      "grad_norm": 2.8196585178375244,
+      "learning_rate": 5e-06,
+      "loss": 0.2036,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9854684512428298,
+      "grad_norm": 1.9609150886535645,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 5154
+    },
+    {
+      "epoch": 0.98565965583174,
+      "grad_norm": 3.2695369720458984,
+      "learning_rate": 5e-06,
+      "loss": 0.0569,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9858508604206501,
+      "grad_norm": 1.6373162269592285,
+      "learning_rate": 5e-06,
+      "loss": 0.0689,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9860420650095603,
+      "grad_norm": 1.8946233987808228,
+      "learning_rate": 5e-06,
+      "loss": 0.2049,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9862332695984704,
+      "grad_norm": 2.2586464881896973,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9864244741873806,
+      "grad_norm": 2.046656370162964,
+      "learning_rate": 5e-06,
+      "loss": 0.2424,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9866156787762906,
+      "grad_norm": 0.6712673902511597,
+      "learning_rate": 5e-06,
+      "loss": 0.0386,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9868068833652007,
+      "grad_norm": 1.4954816102981567,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9869980879541109,
+      "grad_norm": 2.4143035411834717,
+      "learning_rate": 5e-06,
+      "loss": 0.0619,
+      "step": 5162
+    },
+    {
+      "epoch": 0.987189292543021,
+      "grad_norm": 1.1633849143981934,
+      "learning_rate": 5e-06,
+      "loss": 0.0919,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9873804971319312,
+      "grad_norm": 3.078310966491699,
+      "learning_rate": 5e-06,
+      "loss": 0.5286,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9875717017208413,
+      "grad_norm": 3.7422358989715576,
+      "learning_rate": 5e-06,
+      "loss": 0.7243,
+      "step": 5165
+    },
+    {
+      "epoch": 0.9877629063097514,
+      "grad_norm": 2.2467732429504395,
+      "learning_rate": 5e-06,
+      "loss": 0.2758,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9879541108986616,
+      "grad_norm": 1.4942512512207031,
+      "learning_rate": 5e-06,
+      "loss": 0.1251,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9881453154875717,
+      "grad_norm": 3.4213297367095947,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9883365200764819,
+      "grad_norm": 0.8668385744094849,
+      "learning_rate": 5e-06,
+      "loss": 0.0532,
+      "step": 5169
+    },
+    {
+      "epoch": 0.988527724665392,
+      "grad_norm": 4.07417106628418,
+      "learning_rate": 5e-06,
+      "loss": 0.2916,
+      "step": 5170
+    },
+    {
+      "epoch": 0.988718929254302,
+      "grad_norm": 1.8834352493286133,
+      "learning_rate": 5e-06,
+      "loss": 0.2522,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9889101338432122,
+      "grad_norm": 1.5943543910980225,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 5172
+    },
+    {
+      "epoch": 0.9891013384321223,
+      "grad_norm": 2.434356927871704,
+      "learning_rate": 5e-06,
+      "loss": 0.1855,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9892925430210325,
+      "grad_norm": 2.505683422088623,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 5174
+    },
+    {
+      "epoch": 0.9894837476099426,
+      "grad_norm": 2.152951240539551,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9896749521988528,
+      "grad_norm": 1.6887750625610352,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9898661567877629,
+      "grad_norm": 1.2444119453430176,
+      "learning_rate": 5e-06,
+      "loss": 0.1079,
+      "step": 5177
+    },
+    {
+      "epoch": 0.990057361376673,
+      "grad_norm": 2.2396018505096436,
+      "learning_rate": 5e-06,
+      "loss": 0.3107,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9902485659655832,
+      "grad_norm": 2.357428550720215,
+      "learning_rate": 5e-06,
+      "loss": 0.3262,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9904397705544933,
+      "grad_norm": 1.2109572887420654,
+      "learning_rate": 5e-06,
+      "loss": 0.0421,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9906309751434035,
+      "grad_norm": 3.045706272125244,
+      "learning_rate": 5e-06,
+      "loss": 0.0984,
+      "step": 5181
+    },
+    {
+      "epoch": 0.9908221797323136,
+      "grad_norm": 3.762176990509033,
+      "learning_rate": 5e-06,
+      "loss": 0.2968,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9910133843212237,
+      "grad_norm": 1.7476848363876343,
+      "learning_rate": 5e-06,
+      "loss": 0.1861,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9912045889101339,
+      "grad_norm": 1.974594235420227,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9913957934990439,
+      "grad_norm": 1.9838353395462036,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9915869980879541,
+      "grad_norm": 1.257691502571106,
+      "learning_rate": 5e-06,
+      "loss": 0.0803,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9917782026768642,
+      "grad_norm": 1.7940335273742676,
+      "learning_rate": 5e-06,
+      "loss": 0.0902,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9919694072657744,
+      "grad_norm": 2.434577226638794,
+      "learning_rate": 5e-06,
+      "loss": 0.1678,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9921606118546845,
+      "grad_norm": 2.495274543762207,
+      "learning_rate": 5e-06,
+      "loss": 0.3524,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9923518164435946,
+      "grad_norm": 2.122135639190674,
+      "learning_rate": 5e-06,
+      "loss": 0.334,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9925430210325048,
+      "grad_norm": 3.421670436859131,
+      "learning_rate": 5e-06,
+      "loss": 0.3898,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9927342256214149,
+      "grad_norm": 0.9619971513748169,
+      "learning_rate": 5e-06,
+      "loss": 0.0719,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9929254302103251,
+      "grad_norm": 0.8660413026809692,
+      "learning_rate": 5e-06,
+      "loss": 0.0436,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9931166347992352,
+      "grad_norm": 1.1905746459960938,
+      "learning_rate": 5e-06,
+      "loss": 0.0579,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9933078393881453,
+      "grad_norm": 3.714054822921753,
+      "learning_rate": 5e-06,
+      "loss": 0.4012,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9934990439770555,
+      "grad_norm": 1.5647445917129517,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9936902485659656,
+      "grad_norm": 1.8715081214904785,
+      "learning_rate": 5e-06,
+      "loss": 0.236,
+      "step": 5197
+    },
+    {
+      "epoch": 0.9938814531548757,
+      "grad_norm": 1.3924305438995361,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 5198
+    },
+    {
+      "epoch": 0.9940726577437858,
+      "grad_norm": 1.4491206407546997,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 5199
+    },
+    {
+      "epoch": 0.994263862332696,
+      "grad_norm": 3.2908480167388916,
+      "learning_rate": 5e-06,
+      "loss": 0.241,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9944550669216061,
+      "grad_norm": 1.7284696102142334,
+      "learning_rate": 5e-06,
+      "loss": 0.1932,
+      "step": 5201
+    },
+    {
+      "epoch": 0.9946462715105162,
+      "grad_norm": 3.488659381866455,
+      "learning_rate": 5e-06,
+      "loss": 0.3217,
+      "step": 5202
+    },
+    {
+      "epoch": 0.9948374760994264,
+      "grad_norm": 1.9990845918655396,
+      "learning_rate": 5e-06,
+      "loss": 0.4941,
+      "step": 5203
+    },
+    {
+      "epoch": 0.9950286806883365,
+      "grad_norm": 1.8910921812057495,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 5204
+    },
+    {
+      "epoch": 0.9952198852772467,
+      "grad_norm": 1.4411567449569702,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 5205
+    },
+    {
+      "epoch": 0.9954110898661568,
+      "grad_norm": 1.101347804069519,
+      "learning_rate": 5e-06,
+      "loss": 0.0547,
+      "step": 5206
+    },
+    {
+      "epoch": 0.9956022944550669,
+      "grad_norm": 1.8866088390350342,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 5207
+    },
+    {
+      "epoch": 0.9957934990439771,
+      "grad_norm": 2.1902003288269043,
+      "learning_rate": 5e-06,
+      "loss": 0.281,
+      "step": 5208
+    },
+    {
+      "epoch": 0.9959847036328872,
+      "grad_norm": 2.494462490081787,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 5209
+    },
+    {
+      "epoch": 0.9961759082217974,
+      "grad_norm": 1.2047346830368042,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 5210
+    },
+    {
+      "epoch": 0.9963671128107074,
+      "grad_norm": 0.9425076246261597,
+      "learning_rate": 5e-06,
+      "loss": 0.059,
+      "step": 5211
+    },
+    {
+      "epoch": 0.9965583173996176,
+      "grad_norm": 1.435917615890503,
+      "learning_rate": 5e-06,
+      "loss": 0.0519,
+      "step": 5212
+    },
+    {
+      "epoch": 0.9967495219885277,
+      "grad_norm": 1.1277811527252197,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 5213
+    },
+    {
+      "epoch": 0.9969407265774378,
+      "grad_norm": 2.782651901245117,
+      "learning_rate": 5e-06,
+      "loss": 0.3652,
+      "step": 5214
+    },
+    {
+      "epoch": 0.997131931166348,
+      "grad_norm": 1.5946991443634033,
+      "learning_rate": 5e-06,
+      "loss": 0.1018,
+      "step": 5215
+    },
+    {
+      "epoch": 0.9973231357552581,
+      "grad_norm": 0.9089525938034058,
+      "learning_rate": 5e-06,
+      "loss": 0.0902,
+      "step": 5216
+    },
+    {
+      "epoch": 0.9975143403441683,
+      "grad_norm": 2.2287213802337646,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 5217
+    },
+    {
+      "epoch": 0.9977055449330784,
+      "grad_norm": 2.977839231491089,
+      "learning_rate": 5e-06,
+      "loss": 0.2146,
+      "step": 5218
+    },
+    {
+      "epoch": 0.9978967495219885,
+      "grad_norm": 1.4728442430496216,
+      "learning_rate": 5e-06,
+      "loss": 0.2433,
+      "step": 5219
+    },
+    {
+      "epoch": 0.9980879541108987,
+      "grad_norm": 2.0566630363464355,
+      "learning_rate": 5e-06,
+      "loss": 0.2769,
+      "step": 5220
+    },
+    {
+      "epoch": 0.9982791586998088,
+      "grad_norm": 1.2457760572433472,
+      "learning_rate": 5e-06,
+      "loss": 0.0825,
+      "step": 5221
+    },
+    {
+      "epoch": 0.998470363288719,
+      "grad_norm": 1.6305168867111206,
+      "learning_rate": 5e-06,
+      "loss": 0.1614,
+      "step": 5222
+    },
+    {
+      "epoch": 0.998661567877629,
+      "grad_norm": 1.9912116527557373,
+      "learning_rate": 5e-06,
+      "loss": 0.0982,
+      "step": 5223
+    },
+    {
+      "epoch": 0.9988527724665393,
+      "grad_norm": 4.200954437255859,
+      "learning_rate": 5e-06,
+      "loss": 0.2215,
+      "step": 5224
+    },
+    {
+      "epoch": 0.9990439770554493,
+      "grad_norm": 1.4355394840240479,
+      "learning_rate": 5e-06,
+      "loss": 0.0419,
+      "step": 5225
+    },
+    {
+      "epoch": 0.9992351816443594,
+      "grad_norm": 1.8551381826400757,
+      "learning_rate": 5e-06,
+      "loss": 0.2277,
+      "step": 5226
+    },
+    {
+      "epoch": 0.9994263862332696,
+      "grad_norm": 2.7024409770965576,
+      "learning_rate": 5e-06,
+      "loss": 0.5278,
+      "step": 5227
+    },
+    {
+      "epoch": 0.9996175908221797,
+      "grad_norm": 1.0480506420135498,
+      "learning_rate": 5e-06,
+      "loss": 0.0521,
+      "step": 5228
+    },
+    {
+      "epoch": 0.9998087954110899,
+      "grad_norm": 2.2718355655670166,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 5229
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5123586654663086,
+      "learning_rate": 5e-06,
+      "loss": 0.3213,
+      "step": 5230
+    },
+    {
+      "epoch": 1.00019120458891,
+      "grad_norm": 1.7297927141189575,
+      "learning_rate": 5e-06,
+      "loss": 0.2801,
+      "step": 5231
+    },
+    {
+      "epoch": 1.0003824091778202,
+      "grad_norm": 1.3347065448760986,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 5232
+    },
+    {
+      "epoch": 1.0005736137667305,
+      "grad_norm": 1.1105912923812866,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 5233
+    },
+    {
+      "epoch": 1.0007648183556406,
+      "grad_norm": 0.9613377451896667,
+      "learning_rate": 5e-06,
+      "loss": 0.0577,
+      "step": 5234
+    },
+    {
+      "epoch": 1.0009560229445507,
+      "grad_norm": 1.0919339656829834,
+      "learning_rate": 5e-06,
+      "loss": 0.099,
+      "step": 5235
+    },
+    {
+      "epoch": 1.0011472275334607,
+      "grad_norm": 0.9715074300765991,
+      "learning_rate": 5e-06,
+      "loss": 0.063,
+      "step": 5236
+    },
+    {
+      "epoch": 1.0013384321223708,
+      "grad_norm": 1.8640201091766357,
+      "learning_rate": 5e-06,
+      "loss": 0.2279,
+      "step": 5237
+    },
+    {
+      "epoch": 1.0015296367112811,
+      "grad_norm": 2.2503163814544678,
+      "learning_rate": 5e-06,
+      "loss": 0.2751,
+      "step": 5238
+    },
+    {
+      "epoch": 1.0017208413001912,
+      "grad_norm": 1.2504420280456543,
+      "learning_rate": 5e-06,
+      "loss": 0.0953,
+      "step": 5239
+    },
+    {
+      "epoch": 1.0019120458891013,
+      "grad_norm": 1.6367700099945068,
+      "learning_rate": 5e-06,
+      "loss": 0.103,
+      "step": 5240
+    },
+    {
+      "epoch": 1.0021032504780114,
+      "grad_norm": 1.514311671257019,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 5241
+    },
+    {
+      "epoch": 1.0022944550669215,
+      "grad_norm": 1.054772973060608,
+      "learning_rate": 5e-06,
+      "loss": 0.0299,
+      "step": 5242
+    },
+    {
+      "epoch": 1.0024856596558318,
+      "grad_norm": 2.3778746128082275,
+      "learning_rate": 5e-06,
+      "loss": 0.3053,
+      "step": 5243
+    },
+    {
+      "epoch": 1.002676864244742,
+      "grad_norm": 2.213348150253296,
+      "learning_rate": 5e-06,
+      "loss": 0.2055,
+      "step": 5244
+    },
+    {
+      "epoch": 1.002868068833652,
+      "grad_norm": 1.6743370294570923,
+      "learning_rate": 5e-06,
+      "loss": 0.0935,
+      "step": 5245
+    },
+    {
+      "epoch": 1.003059273422562,
+      "grad_norm": 1.9218884706497192,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 5246
+    },
+    {
+      "epoch": 1.0032504780114724,
+      "grad_norm": 0.5441107153892517,
+      "learning_rate": 5e-06,
+      "loss": 0.0418,
+      "step": 5247
+    },
+    {
+      "epoch": 1.0034416826003825,
+      "grad_norm": 1.4847966432571411,
+      "learning_rate": 5e-06,
+      "loss": 0.0598,
+      "step": 5248
+    },
+    {
+      "epoch": 1.0036328871892926,
+      "grad_norm": 1.8190624713897705,
+      "learning_rate": 5e-06,
+      "loss": 0.0375,
+      "step": 5249
+    },
+    {
+      "epoch": 1.0038240917782026,
+      "grad_norm": 2.3524463176727295,
+      "learning_rate": 5e-06,
+      "loss": 0.1668,
+      "step": 5250
+    },
+    {
+      "epoch": 1.0040152963671127,
+      "grad_norm": 2.548030376434326,
+      "learning_rate": 5e-06,
+      "loss": 0.3332,
+      "step": 5251
+    },
+    {
+      "epoch": 1.004206500956023,
+      "grad_norm": 1.4580950736999512,
+      "learning_rate": 5e-06,
+      "loss": 0.1052,
+      "step": 5252
+    },
+    {
+      "epoch": 1.0043977055449331,
+      "grad_norm": 1.2642977237701416,
+      "learning_rate": 5e-06,
+      "loss": 0.0873,
+      "step": 5253
+    },
+    {
+      "epoch": 1.0045889101338432,
+      "grad_norm": 1.0301834344863892,
+      "learning_rate": 5e-06,
+      "loss": 0.0373,
+      "step": 5254
+    },
+    {
+      "epoch": 1.0047801147227533,
+      "grad_norm": 1.7443534135818481,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 5255
+    },
+    {
+      "epoch": 1.0049713193116634,
+      "grad_norm": 1.4613491296768188,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 5256
+    },
+    {
+      "epoch": 1.0051625239005737,
+      "grad_norm": 1.6461591720581055,
+      "learning_rate": 5e-06,
+      "loss": 0.1837,
+      "step": 5257
+    },
+    {
+      "epoch": 1.0053537284894838,
+      "grad_norm": 1.1870779991149902,
+      "learning_rate": 5e-06,
+      "loss": 0.0809,
+      "step": 5258
+    },
+    {
+      "epoch": 1.0055449330783939,
+      "grad_norm": 0.8499864935874939,
+      "learning_rate": 5e-06,
+      "loss": 0.0478,
+      "step": 5259
+    },
+    {
+      "epoch": 1.005736137667304,
+      "grad_norm": 1.4943770170211792,
+      "learning_rate": 5e-06,
+      "loss": 0.0532,
+      "step": 5260
+    },
+    {
+      "epoch": 1.005927342256214,
+      "grad_norm": 1.3564469814300537,
+      "learning_rate": 5e-06,
+      "loss": 0.0571,
+      "step": 5261
+    },
+    {
+      "epoch": 1.0061185468451244,
+      "grad_norm": 1.2749664783477783,
+      "learning_rate": 5e-06,
+      "loss": 0.066,
+      "step": 5262
+    },
+    {
+      "epoch": 1.0063097514340344,
+      "grad_norm": 2.179372549057007,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 5263
+    },
+    {
+      "epoch": 1.0065009560229445,
+      "grad_norm": 1.2617064714431763,
+      "learning_rate": 5e-06,
+      "loss": 0.0835,
+      "step": 5264
+    },
+    {
+      "epoch": 1.0066921606118546,
+      "grad_norm": 0.9173052310943604,
+      "learning_rate": 5e-06,
+      "loss": 0.0922,
+      "step": 5265
+    },
+    {
+      "epoch": 1.0068833652007647,
+      "grad_norm": 1.5798940658569336,
+      "learning_rate": 5e-06,
+      "loss": 0.0287,
+      "step": 5266
+    },
+    {
+      "epoch": 1.007074569789675,
+      "grad_norm": 1.8742361068725586,
+      "learning_rate": 5e-06,
+      "loss": 0.088,
+      "step": 5267
+    },
+    {
+      "epoch": 1.007265774378585,
+      "grad_norm": 2.5780484676361084,
+      "learning_rate": 5e-06,
+      "loss": 0.087,
+      "step": 5268
+    },
+    {
+      "epoch": 1.0074569789674952,
+      "grad_norm": 2.258584499359131,
+      "learning_rate": 5e-06,
+      "loss": 0.2834,
+      "step": 5269
+    },
+    {
+      "epoch": 1.0076481835564053,
+      "grad_norm": 0.6943368315696716,
+      "learning_rate": 5e-06,
+      "loss": 0.0394,
+      "step": 5270
+    },
+    {
+      "epoch": 1.0078393881453156,
+      "grad_norm": 1.9142544269561768,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 5271
+    },
+    {
+      "epoch": 1.0080305927342257,
+      "grad_norm": 0.8416318297386169,
+      "learning_rate": 5e-06,
+      "loss": 0.0664,
+      "step": 5272
+    },
+    {
+      "epoch": 1.0082217973231358,
+      "grad_norm": 1.94535493850708,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 5273
+    },
+    {
+      "epoch": 1.0084130019120459,
+      "grad_norm": 1.2606523036956787,
+      "learning_rate": 5e-06,
+      "loss": 0.0534,
+      "step": 5274
+    },
+    {
+      "epoch": 1.008604206500956,
+      "grad_norm": 2.3237664699554443,
+      "learning_rate": 5e-06,
+      "loss": 0.2517,
+      "step": 5275
+    },
+    {
+      "epoch": 1.0087954110898663,
+      "grad_norm": 2.197056293487549,
+      "learning_rate": 5e-06,
+      "loss": 0.3001,
+      "step": 5276
+    },
+    {
+      "epoch": 1.0089866156787763,
+      "grad_norm": 1.2944321632385254,
+      "learning_rate": 5e-06,
+      "loss": 0.103,
+      "step": 5277
+    },
+    {
+      "epoch": 1.0091778202676864,
+      "grad_norm": 1.5966204404830933,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 5278
+    },
+    {
+      "epoch": 1.0093690248565965,
+      "grad_norm": 1.0905331373214722,
+      "learning_rate": 5e-06,
+      "loss": 0.0653,
+      "step": 5279
+    },
+    {
+      "epoch": 1.0095602294455066,
+      "grad_norm": 1.1597042083740234,
+      "learning_rate": 5e-06,
+      "loss": 0.0651,
+      "step": 5280
+    },
+    {
+      "epoch": 1.009751434034417,
+      "grad_norm": 2.1614441871643066,
+      "learning_rate": 5e-06,
+      "loss": 0.1937,
+      "step": 5281
+    },
+    {
+      "epoch": 1.009942638623327,
+      "grad_norm": 2.6233420372009277,
+      "learning_rate": 5e-06,
+      "loss": 0.3533,
+      "step": 5282
+    },
+    {
+      "epoch": 1.010133843212237,
+      "grad_norm": 1.4536938667297363,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 5283
+    },
+    {
+      "epoch": 1.0103250478011472,
+      "grad_norm": 0.9678513407707214,
+      "learning_rate": 5e-06,
+      "loss": 0.0769,
+      "step": 5284
+    },
+    {
+      "epoch": 1.0105162523900573,
+      "grad_norm": 1.2624231576919556,
+      "learning_rate": 5e-06,
+      "loss": 0.0625,
+      "step": 5285
+    },
+    {
+      "epoch": 1.0107074569789676,
+      "grad_norm": 1.1822149753570557,
+      "learning_rate": 5e-06,
+      "loss": 0.0367,
+      "step": 5286
+    },
+    {
+      "epoch": 1.0108986615678777,
+      "grad_norm": 1.7562121152877808,
+      "learning_rate": 5e-06,
+      "loss": 0.1614,
+      "step": 5287
+    },
+    {
+      "epoch": 1.0110898661567878,
+      "grad_norm": 2.224581003189087,
+      "learning_rate": 5e-06,
+      "loss": 0.3048,
+      "step": 5288
+    },
+    {
+      "epoch": 1.0112810707456978,
+      "grad_norm": 1.7163106203079224,
+      "learning_rate": 5e-06,
+      "loss": 0.2074,
+      "step": 5289
+    },
+    {
+      "epoch": 1.011472275334608,
+      "grad_norm": 1.9225854873657227,
+      "learning_rate": 5e-06,
+      "loss": 0.2347,
+      "step": 5290
+    },
+    {
+      "epoch": 1.0116634799235182,
+      "grad_norm": 0.8340242505073547,
+      "learning_rate": 5e-06,
+      "loss": 0.0415,
+      "step": 5291
+    },
+    {
+      "epoch": 1.0118546845124283,
+      "grad_norm": 0.9325177073478699,
+      "learning_rate": 5e-06,
+      "loss": 0.0336,
+      "step": 5292
+    },
+    {
+      "epoch": 1.0120458891013384,
+      "grad_norm": 1.3594424724578857,
+      "learning_rate": 5e-06,
+      "loss": 0.0575,
+      "step": 5293
+    },
+    {
+      "epoch": 1.0122370936902485,
+      "grad_norm": 2.631110429763794,
+      "learning_rate": 5e-06,
+      "loss": 0.4276,
+      "step": 5294
+    },
+    {
+      "epoch": 1.0124282982791586,
+      "grad_norm": 1.8750988245010376,
+      "learning_rate": 5e-06,
+      "loss": 0.3076,
+      "step": 5295
+    },
+    {
+      "epoch": 1.012619502868069,
+      "grad_norm": 1.2204387187957764,
+      "learning_rate": 5e-06,
+      "loss": 0.074,
+      "step": 5296
+    },
+    {
+      "epoch": 1.012810707456979,
+      "grad_norm": 1.7058926820755005,
+      "learning_rate": 5e-06,
+      "loss": 0.1998,
+      "step": 5297
+    },
+    {
+      "epoch": 1.013001912045889,
+      "grad_norm": 1.6197351217269897,
+      "learning_rate": 5e-06,
+      "loss": 0.071,
+      "step": 5298
+    },
+    {
+      "epoch": 1.0131931166347992,
+      "grad_norm": 1.8550605773925781,
+      "learning_rate": 5e-06,
+      "loss": 0.0786,
+      "step": 5299
+    },
+    {
+      "epoch": 1.0133843212237095,
+      "grad_norm": 1.1241230964660645,
+      "learning_rate": 5e-06,
+      "loss": 0.0807,
+      "step": 5300
+    },
+    {
+      "epoch": 1.0135755258126196,
+      "grad_norm": 2.6108241081237793,
+      "learning_rate": 5e-06,
+      "loss": 0.2384,
+      "step": 5301
+    },
+    {
+      "epoch": 1.0137667304015296,
+      "grad_norm": 1.967355728149414,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 5302
+    },
+    {
+      "epoch": 1.0139579349904397,
+      "grad_norm": 1.6688017845153809,
+      "learning_rate": 5e-06,
+      "loss": 0.0525,
+      "step": 5303
+    },
+    {
+      "epoch": 1.0141491395793498,
+      "grad_norm": 1.1315393447875977,
+      "learning_rate": 5e-06,
+      "loss": 0.0706,
+      "step": 5304
+    },
+    {
+      "epoch": 1.0143403441682601,
+      "grad_norm": 1.7984042167663574,
+      "learning_rate": 5e-06,
+      "loss": 0.078,
+      "step": 5305
+    },
+    {
+      "epoch": 1.0145315487571702,
+      "grad_norm": 1.8164480924606323,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 5306
+    },
+    {
+      "epoch": 1.0147227533460803,
+      "grad_norm": 1.3991438150405884,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 5307
+    },
+    {
+      "epoch": 1.0149139579349904,
+      "grad_norm": 1.8388389348983765,
+      "learning_rate": 5e-06,
+      "loss": 0.0591,
+      "step": 5308
+    },
+    {
+      "epoch": 1.0151051625239005,
+      "grad_norm": 1.4206583499908447,
+      "learning_rate": 5e-06,
+      "loss": 0.0782,
+      "step": 5309
+    },
+    {
+      "epoch": 1.0152963671128108,
+      "grad_norm": 0.5835526585578918,
+      "learning_rate": 5e-06,
+      "loss": 0.0423,
+      "step": 5310
+    },
+    {
+      "epoch": 1.0154875717017209,
+      "grad_norm": 1.2603877782821655,
+      "learning_rate": 5e-06,
+      "loss": 0.0581,
+      "step": 5311
+    },
+    {
+      "epoch": 1.015678776290631,
+      "grad_norm": 2.069904327392578,
+      "learning_rate": 5e-06,
+      "loss": 0.2226,
+      "step": 5312
+    },
+    {
+      "epoch": 1.015869980879541,
+      "grad_norm": 1.6898385286331177,
+      "learning_rate": 5e-06,
+      "loss": 0.1273,
+      "step": 5313
+    },
+    {
+      "epoch": 1.0160611854684511,
+      "grad_norm": 2.38932466506958,
+      "learning_rate": 5e-06,
+      "loss": 0.1309,
+      "step": 5314
+    },
+    {
+      "epoch": 1.0162523900573615,
+      "grad_norm": 1.2808679342269897,
+      "learning_rate": 5e-06,
+      "loss": 0.0763,
+      "step": 5315
+    },
+    {
+      "epoch": 1.0164435946462715,
+      "grad_norm": 1.2353378534317017,
+      "learning_rate": 5e-06,
+      "loss": 0.0659,
+      "step": 5316
+    },
+    {
+      "epoch": 1.0166347992351816,
+      "grad_norm": 1.1403536796569824,
+      "learning_rate": 5e-06,
+      "loss": 0.059,
+      "step": 5317
+    },
+    {
+      "epoch": 1.0168260038240917,
+      "grad_norm": 1.8925150632858276,
+      "learning_rate": 5e-06,
+      "loss": 0.1917,
+      "step": 5318
+    },
+    {
+      "epoch": 1.0170172084130018,
+      "grad_norm": 1.5492744445800781,
+      "learning_rate": 5e-06,
+      "loss": 0.1937,
+      "step": 5319
+    },
+    {
+      "epoch": 1.0172084130019121,
+      "grad_norm": 2.6364314556121826,
+      "learning_rate": 5e-06,
+      "loss": 0.2993,
+      "step": 5320
+    },
+    {
+      "epoch": 1.0173996175908222,
+      "grad_norm": 1.9068900346755981,
+      "learning_rate": 5e-06,
+      "loss": 0.2033,
+      "step": 5321
+    },
+    {
+      "epoch": 1.0175908221797323,
+      "grad_norm": 1.244210958480835,
+      "learning_rate": 5e-06,
+      "loss": 0.084,
+      "step": 5322
+    },
+    {
+      "epoch": 1.0177820267686424,
+      "grad_norm": 1.4095242023468018,
+      "learning_rate": 5e-06,
+      "loss": 0.0952,
+      "step": 5323
+    },
+    {
+      "epoch": 1.0179732313575527,
+      "grad_norm": 1.3021609783172607,
+      "learning_rate": 5e-06,
+      "loss": 0.0732,
+      "step": 5324
+    },
+    {
+      "epoch": 1.0181644359464628,
+      "grad_norm": 2.593829870223999,
+      "learning_rate": 5e-06,
+      "loss": 0.3372,
+      "step": 5325
+    },
+    {
+      "epoch": 1.0183556405353729,
+      "grad_norm": 0.7937406897544861,
+      "learning_rate": 5e-06,
+      "loss": 0.0694,
+      "step": 5326
+    },
+    {
+      "epoch": 1.018546845124283,
+      "grad_norm": 1.4523568153381348,
+      "learning_rate": 5e-06,
+      "loss": 0.0889,
+      "step": 5327
+    },
+    {
+      "epoch": 1.018738049713193,
+      "grad_norm": 2.000311851501465,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 5328
+    },
+    {
+      "epoch": 1.0189292543021033,
+      "grad_norm": 1.0171622037887573,
+      "learning_rate": 5e-06,
+      "loss": 0.0692,
+      "step": 5329
+    },
+    {
+      "epoch": 1.0191204588910134,
+      "grad_norm": 0.9312133193016052,
+      "learning_rate": 5e-06,
+      "loss": 0.0239,
+      "step": 5330
+    },
+    {
+      "epoch": 1.0193116634799235,
+      "grad_norm": 2.9251790046691895,
+      "learning_rate": 5e-06,
+      "loss": 0.1988,
+      "step": 5331
+    },
+    {
+      "epoch": 1.0195028680688336,
+      "grad_norm": 1.3772673606872559,
+      "learning_rate": 5e-06,
+      "loss": 0.0688,
+      "step": 5332
+    },
+    {
+      "epoch": 1.0196940726577437,
+      "grad_norm": 1.559841275215149,
+      "learning_rate": 5e-06,
+      "loss": 0.2109,
+      "step": 5333
+    },
+    {
+      "epoch": 1.019885277246654,
+      "grad_norm": 1.385416865348816,
+      "learning_rate": 5e-06,
+      "loss": 0.0531,
+      "step": 5334
+    },
+    {
+      "epoch": 1.020076481835564,
+      "grad_norm": 1.2152140140533447,
+      "learning_rate": 5e-06,
+      "loss": 0.0514,
+      "step": 5335
+    },
+    {
+      "epoch": 1.0202676864244742,
+      "grad_norm": 1.0661227703094482,
+      "learning_rate": 5e-06,
+      "loss": 0.0496,
+      "step": 5336
+    },
+    {
+      "epoch": 1.0204588910133843,
+      "grad_norm": 2.223980188369751,
+      "learning_rate": 5e-06,
+      "loss": 0.2001,
+      "step": 5337
+    },
+    {
+      "epoch": 1.0206500956022944,
+      "grad_norm": 2.8563883304595947,
+      "learning_rate": 5e-06,
+      "loss": 0.3093,
+      "step": 5338
+    },
+    {
+      "epoch": 1.0208413001912047,
+      "grad_norm": 1.7817109823226929,
+      "learning_rate": 5e-06,
+      "loss": 0.0834,
+      "step": 5339
+    },
+    {
+      "epoch": 1.0210325047801148,
+      "grad_norm": 1.39713716506958,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 5340
+    },
+    {
+      "epoch": 1.0212237093690248,
+      "grad_norm": 0.9216711521148682,
+      "learning_rate": 5e-06,
+      "loss": 0.046,
+      "step": 5341
+    },
+    {
+      "epoch": 1.021414913957935,
+      "grad_norm": 3.6287028789520264,
+      "learning_rate": 5e-06,
+      "loss": 0.066,
+      "step": 5342
+    },
+    {
+      "epoch": 1.021606118546845,
+      "grad_norm": 1.79201078414917,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 5343
+    },
+    {
+      "epoch": 1.0217973231357553,
+      "grad_norm": 1.3624271154403687,
+      "learning_rate": 5e-06,
+      "loss": 0.0847,
+      "step": 5344
+    },
+    {
+      "epoch": 1.0219885277246654,
+      "grad_norm": 1.809230089187622,
+      "learning_rate": 5e-06,
+      "loss": 0.0799,
+      "step": 5345
+    },
+    {
+      "epoch": 1.0221797323135755,
+      "grad_norm": 1.3011666536331177,
+      "learning_rate": 5e-06,
+      "loss": 0.0849,
+      "step": 5346
+    },
+    {
+      "epoch": 1.0223709369024856,
+      "grad_norm": 0.8237725496292114,
+      "learning_rate": 5e-06,
+      "loss": 0.0524,
+      "step": 5347
+    },
+    {
+      "epoch": 1.0225621414913957,
+      "grad_norm": 1.9484095573425293,
+      "learning_rate": 5e-06,
+      "loss": 0.0832,
+      "step": 5348
+    },
+    {
+      "epoch": 1.022753346080306,
+      "grad_norm": 1.2306699752807617,
+      "learning_rate": 5e-06,
+      "loss": 0.0433,
+      "step": 5349
+    },
+    {
+      "epoch": 1.022944550669216,
+      "grad_norm": 2.3627052307128906,
+      "learning_rate": 5e-06,
+      "loss": 0.2586,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0231357552581262,
+      "grad_norm": 1.5180178880691528,
+      "learning_rate": 5e-06,
+      "loss": 0.1013,
+      "step": 5351
+    },
+    {
+      "epoch": 1.0233269598470363,
+      "grad_norm": 1.259223222732544,
+      "learning_rate": 5e-06,
+      "loss": 0.1714,
+      "step": 5352
+    },
+    {
+      "epoch": 1.0235181644359466,
+      "grad_norm": 1.2246848344802856,
+      "learning_rate": 5e-06,
+      "loss": 0.0907,
+      "step": 5353
+    },
+    {
+      "epoch": 1.0237093690248567,
+      "grad_norm": 2.1600708961486816,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 5354
+    },
+    {
+      "epoch": 1.0239005736137667,
+      "grad_norm": 1.1809910535812378,
+      "learning_rate": 5e-06,
+      "loss": 0.0521,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0240917782026768,
+      "grad_norm": 2.0541279315948486,
+      "learning_rate": 5e-06,
+      "loss": 0.1937,
+      "step": 5356
+    },
+    {
+      "epoch": 1.024282982791587,
+      "grad_norm": 1.9363659620285034,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 5357
+    },
+    {
+      "epoch": 1.0244741873804972,
+      "grad_norm": 2.4680871963500977,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 5358
+    },
+    {
+      "epoch": 1.0246653919694073,
+      "grad_norm": 1.0142306089401245,
+      "learning_rate": 5e-06,
+      "loss": 0.0611,
+      "step": 5359
+    },
+    {
+      "epoch": 1.0248565965583174,
+      "grad_norm": 1.288448452949524,
+      "learning_rate": 5e-06,
+      "loss": 0.0642,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0250478011472275,
+      "grad_norm": 1.4450470209121704,
+      "learning_rate": 5e-06,
+      "loss": 0.0648,
+      "step": 5361
+    },
+    {
+      "epoch": 1.0252390057361376,
+      "grad_norm": 1.5391541719436646,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 5362
+    },
+    {
+      "epoch": 1.0254302103250479,
+      "grad_norm": 2.144197940826416,
+      "learning_rate": 5e-06,
+      "loss": 0.3029,
+      "step": 5363
+    },
+    {
+      "epoch": 1.025621414913958,
+      "grad_norm": 1.9811673164367676,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 5364
+    },
+    {
+      "epoch": 1.025812619502868,
+      "grad_norm": 0.9573968648910522,
+      "learning_rate": 5e-06,
+      "loss": 0.0651,
+      "step": 5365
+    },
+    {
+      "epoch": 1.0260038240917781,
+      "grad_norm": 1.0169429779052734,
+      "learning_rate": 5e-06,
+      "loss": 0.0364,
+      "step": 5366
+    },
+    {
+      "epoch": 1.0261950286806882,
+      "grad_norm": 0.7411577701568604,
+      "learning_rate": 5e-06,
+      "loss": 0.0185,
+      "step": 5367
+    },
+    {
+      "epoch": 1.0263862332695985,
+      "grad_norm": 1.0757877826690674,
+      "learning_rate": 5e-06,
+      "loss": 0.082,
+      "step": 5368
+    },
+    {
+      "epoch": 1.0265774378585086,
+      "grad_norm": 2.224330425262451,
+      "learning_rate": 5e-06,
+      "loss": 0.2543,
+      "step": 5369
+    },
+    {
+      "epoch": 1.0267686424474187,
+      "grad_norm": 1.7419105768203735,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0269598470363288,
+      "grad_norm": 1.4337900876998901,
+      "learning_rate": 5e-06,
+      "loss": 0.0853,
+      "step": 5371
+    },
+    {
+      "epoch": 1.027151051625239,
+      "grad_norm": 1.0431238412857056,
+      "learning_rate": 5e-06,
+      "loss": 0.0712,
+      "step": 5372
+    },
+    {
+      "epoch": 1.0273422562141492,
+      "grad_norm": 1.8076082468032837,
+      "learning_rate": 5e-06,
+      "loss": 0.0804,
+      "step": 5373
+    },
+    {
+      "epoch": 1.0275334608030593,
+      "grad_norm": 1.998808741569519,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 5374
+    },
+    {
+      "epoch": 1.0277246653919694,
+      "grad_norm": 1.9513105154037476,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 5375
+    },
+    {
+      "epoch": 1.0279158699808795,
+      "grad_norm": 0.6905821561813354,
+      "learning_rate": 5e-06,
+      "loss": 0.0502,
+      "step": 5376
+    },
+    {
+      "epoch": 1.0281070745697898,
+      "grad_norm": 1.0172103643417358,
+      "learning_rate": 5e-06,
+      "loss": 0.0641,
+      "step": 5377
+    },
+    {
+      "epoch": 1.0282982791586999,
+      "grad_norm": 1.166339635848999,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 5378
+    },
+    {
+      "epoch": 1.02848948374761,
+      "grad_norm": 1.3585234880447388,
+      "learning_rate": 5e-06,
+      "loss": 0.0646,
+      "step": 5379
+    },
+    {
+      "epoch": 1.02868068833652,
+      "grad_norm": 1.65544593334198,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0288718929254301,
+      "grad_norm": 2.576430082321167,
+      "learning_rate": 5e-06,
+      "loss": 0.2197,
+      "step": 5381
+    },
+    {
+      "epoch": 1.0290630975143404,
+      "grad_norm": 2.2190842628479004,
+      "learning_rate": 5e-06,
+      "loss": 0.1957,
+      "step": 5382
+    },
+    {
+      "epoch": 1.0292543021032505,
+      "grad_norm": 1.092215895652771,
+      "learning_rate": 5e-06,
+      "loss": 0.0609,
+      "step": 5383
+    },
+    {
+      "epoch": 1.0294455066921606,
+      "grad_norm": 1.6553555727005005,
+      "learning_rate": 5e-06,
+      "loss": 0.085,
+      "step": 5384
+    },
+    {
+      "epoch": 1.0296367112810707,
+      "grad_norm": 1.6299556493759155,
+      "learning_rate": 5e-06,
+      "loss": 0.0454,
+      "step": 5385
+    },
+    {
+      "epoch": 1.0298279158699808,
+      "grad_norm": 1.2839386463165283,
+      "learning_rate": 5e-06,
+      "loss": 0.068,
+      "step": 5386
+    },
+    {
+      "epoch": 1.030019120458891,
+      "grad_norm": 1.5672410726547241,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 5387
+    },
+    {
+      "epoch": 1.0302103250478012,
+      "grad_norm": 1.3381305932998657,
+      "learning_rate": 5e-06,
+      "loss": 0.0716,
+      "step": 5388
+    },
+    {
+      "epoch": 1.0304015296367113,
+      "grad_norm": 0.9112756252288818,
+      "learning_rate": 5e-06,
+      "loss": 0.0466,
+      "step": 5389
+    },
+    {
+      "epoch": 1.0305927342256214,
+      "grad_norm": 0.6958617568016052,
+      "learning_rate": 5e-06,
+      "loss": 0.0528,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0307839388145315,
+      "grad_norm": 1.5371332168579102,
+      "learning_rate": 5e-06,
+      "loss": 0.0767,
+      "step": 5391
+    },
+    {
+      "epoch": 1.0309751434034418,
+      "grad_norm": 2.49212646484375,
+      "learning_rate": 5e-06,
+      "loss": 0.1944,
+      "step": 5392
+    },
+    {
+      "epoch": 1.0311663479923519,
+      "grad_norm": 1.2017213106155396,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 5393
+    },
+    {
+      "epoch": 1.031357552581262,
+      "grad_norm": 1.103096604347229,
+      "learning_rate": 5e-06,
+      "loss": 0.0552,
+      "step": 5394
+    },
+    {
+      "epoch": 1.031548757170172,
+      "grad_norm": 2.4898500442504883,
+      "learning_rate": 5e-06,
+      "loss": 0.1824,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0317399617590821,
+      "grad_norm": 1.9159975051879883,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 5396
+    },
+    {
+      "epoch": 1.0319311663479924,
+      "grad_norm": 1.6219099760055542,
+      "learning_rate": 5e-06,
+      "loss": 0.0913,
+      "step": 5397
+    },
+    {
+      "epoch": 1.0321223709369025,
+      "grad_norm": 1.076381802558899,
+      "learning_rate": 5e-06,
+      "loss": 0.0744,
+      "step": 5398
+    },
+    {
+      "epoch": 1.0323135755258126,
+      "grad_norm": 1.1917436122894287,
+      "learning_rate": 5e-06,
+      "loss": 0.0657,
+      "step": 5399
+    },
+    {
+      "epoch": 1.0325047801147227,
+      "grad_norm": 1.8527624607086182,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0326959847036328,
+      "grad_norm": 2.29830002784729,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 5401
+    },
+    {
+      "epoch": 1.032887189292543,
+      "grad_norm": 2.4546074867248535,
+      "learning_rate": 5e-06,
+      "loss": 0.2168,
+      "step": 5402
+    },
+    {
+      "epoch": 1.0330783938814532,
+      "grad_norm": 0.7161755561828613,
+      "learning_rate": 5e-06,
+      "loss": 0.0753,
+      "step": 5403
+    },
+    {
+      "epoch": 1.0332695984703633,
+      "grad_norm": 1.9103176593780518,
+      "learning_rate": 5e-06,
+      "loss": 0.0824,
+      "step": 5404
+    },
+    {
+      "epoch": 1.0334608030592733,
+      "grad_norm": 2.567988634109497,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0336520076481837,
+      "grad_norm": 2.4080469608306885,
+      "learning_rate": 5e-06,
+      "loss": 0.3055,
+      "step": 5406
+    },
+    {
+      "epoch": 1.0338432122370937,
+      "grad_norm": 2.5664243698120117,
+      "learning_rate": 5e-06,
+      "loss": 0.3235,
+      "step": 5407
+    },
+    {
+      "epoch": 1.0340344168260038,
+      "grad_norm": 0.7243027091026306,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 5408
+    },
+    {
+      "epoch": 1.034225621414914,
+      "grad_norm": 2.28975510597229,
+      "learning_rate": 5e-06,
+      "loss": 0.089,
+      "step": 5409
+    },
+    {
+      "epoch": 1.034416826003824,
+      "grad_norm": 1.4908415079116821,
+      "learning_rate": 5e-06,
+      "loss": 0.0705,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0346080305927343,
+      "grad_norm": 0.9343728423118591,
+      "learning_rate": 5e-06,
+      "loss": 0.0416,
+      "step": 5411
+    },
+    {
+      "epoch": 1.0347992351816444,
+      "grad_norm": 2.183262586593628,
+      "learning_rate": 5e-06,
+      "loss": 0.1471,
+      "step": 5412
+    },
+    {
+      "epoch": 1.0349904397705545,
+      "grad_norm": 1.1275417804718018,
+      "learning_rate": 5e-06,
+      "loss": 0.0581,
+      "step": 5413
+    },
+    {
+      "epoch": 1.0351816443594646,
+      "grad_norm": 3.412479877471924,
+      "learning_rate": 5e-06,
+      "loss": 0.4229,
+      "step": 5414
+    },
+    {
+      "epoch": 1.0353728489483747,
+      "grad_norm": 0.81615149974823,
+      "learning_rate": 5e-06,
+      "loss": 0.0412,
+      "step": 5415
+    },
+    {
+      "epoch": 1.035564053537285,
+      "grad_norm": 1.7186781167984009,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 5416
+    },
+    {
+      "epoch": 1.035755258126195,
+      "grad_norm": 3.903059482574463,
+      "learning_rate": 5e-06,
+      "loss": 0.0223,
+      "step": 5417
+    },
+    {
+      "epoch": 1.0359464627151052,
+      "grad_norm": 1.3878508806228638,
+      "learning_rate": 5e-06,
+      "loss": 0.069,
+      "step": 5418
+    },
+    {
+      "epoch": 1.0361376673040152,
+      "grad_norm": 2.2635209560394287,
+      "learning_rate": 5e-06,
+      "loss": 0.1977,
+      "step": 5419
+    },
+    {
+      "epoch": 1.0363288718929253,
+      "grad_norm": 1.8340933322906494,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0365200764818356,
+      "grad_norm": 1.2530016899108887,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 5421
+    },
+    {
+      "epoch": 1.0367112810707457,
+      "grad_norm": 0.9224182367324829,
+      "learning_rate": 5e-06,
+      "loss": 0.0661,
+      "step": 5422
+    },
+    {
+      "epoch": 1.0369024856596558,
+      "grad_norm": 1.306754231452942,
+      "learning_rate": 5e-06,
+      "loss": 0.0599,
+      "step": 5423
+    },
+    {
+      "epoch": 1.037093690248566,
+      "grad_norm": 1.311036467552185,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 5424
+    },
+    {
+      "epoch": 1.0372848948374762,
+      "grad_norm": 1.1835675239562988,
+      "learning_rate": 5e-06,
+      "loss": 0.0955,
+      "step": 5425
+    },
+    {
+      "epoch": 1.0374760994263863,
+      "grad_norm": 0.7402036190032959,
+      "learning_rate": 5e-06,
+      "loss": 0.0427,
+      "step": 5426
+    },
+    {
+      "epoch": 1.0376673040152964,
+      "grad_norm": 2.02817964553833,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 5427
+    },
+    {
+      "epoch": 1.0378585086042065,
+      "grad_norm": 2.148780107498169,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 5428
+    },
+    {
+      "epoch": 1.0380497131931166,
+      "grad_norm": 1.8620522022247314,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 5429
+    },
+    {
+      "epoch": 1.0382409177820269,
+      "grad_norm": 0.7214341759681702,
+      "learning_rate": 5e-06,
+      "loss": 0.0191,
+      "step": 5430
+    },
+    {
+      "epoch": 1.038432122370937,
+      "grad_norm": 1.828162670135498,
+      "learning_rate": 5e-06,
+      "loss": 0.1798,
+      "step": 5431
+    },
+    {
+      "epoch": 1.038623326959847,
+      "grad_norm": 2.594289541244507,
+      "learning_rate": 5e-06,
+      "loss": 0.2456,
+      "step": 5432
+    },
+    {
+      "epoch": 1.0388145315487571,
+      "grad_norm": 0.8717902302742004,
+      "learning_rate": 5e-06,
+      "loss": 0.0624,
+      "step": 5433
+    },
+    {
+      "epoch": 1.0390057361376672,
+      "grad_norm": 2.1278491020202637,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 5434
+    },
+    {
+      "epoch": 1.0391969407265775,
+      "grad_norm": 1.1939607858657837,
+      "learning_rate": 5e-06,
+      "loss": 0.0389,
+      "step": 5435
+    },
+    {
+      "epoch": 1.0393881453154876,
+      "grad_norm": 2.5707457065582275,
+      "learning_rate": 5e-06,
+      "loss": 0.2028,
+      "step": 5436
+    },
+    {
+      "epoch": 1.0395793499043977,
+      "grad_norm": 1.869125247001648,
+      "learning_rate": 5e-06,
+      "loss": 0.2081,
+      "step": 5437
+    },
+    {
+      "epoch": 1.0397705544933078,
+      "grad_norm": 3.4773709774017334,
+      "learning_rate": 5e-06,
+      "loss": 0.4955,
+      "step": 5438
+    },
+    {
+      "epoch": 1.0399617590822179,
+      "grad_norm": 1.6218843460083008,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 5439
+    },
+    {
+      "epoch": 1.0401529636711282,
+      "grad_norm": 2.7059316635131836,
+      "learning_rate": 5e-06,
+      "loss": 0.3133,
+      "step": 5440
+    },
+    {
+      "epoch": 1.0403441682600383,
+      "grad_norm": 1.0128536224365234,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 5441
+    },
+    {
+      "epoch": 1.0405353728489484,
+      "grad_norm": 6.157577037811279,
+      "learning_rate": 5e-06,
+      "loss": 0.2846,
+      "step": 5442
+    },
+    {
+      "epoch": 1.0407265774378585,
+      "grad_norm": 1.409836769104004,
+      "learning_rate": 5e-06,
+      "loss": 0.0739,
+      "step": 5443
+    },
+    {
+      "epoch": 1.0409177820267685,
+      "grad_norm": 2.545416831970215,
+      "learning_rate": 5e-06,
+      "loss": 0.3916,
+      "step": 5444
+    },
+    {
+      "epoch": 1.0411089866156789,
+      "grad_norm": 1.2188429832458496,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 5445
+    },
+    {
+      "epoch": 1.041300191204589,
+      "grad_norm": 3.023515224456787,
+      "learning_rate": 5e-06,
+      "loss": 0.4011,
+      "step": 5446
+    },
+    {
+      "epoch": 1.041491395793499,
+      "grad_norm": 2.3562259674072266,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 5447
+    },
+    {
+      "epoch": 1.0416826003824091,
+      "grad_norm": 1.0878733396530151,
+      "learning_rate": 5e-06,
+      "loss": 0.0712,
+      "step": 5448
+    },
+    {
+      "epoch": 1.0418738049713192,
+      "grad_norm": 2.7172553539276123,
+      "learning_rate": 5e-06,
+      "loss": 0.1926,
+      "step": 5449
+    },
+    {
+      "epoch": 1.0420650095602295,
+      "grad_norm": 2.0290565490722656,
+      "learning_rate": 5e-06,
+      "loss": 0.2468,
+      "step": 5450
+    },
+    {
+      "epoch": 1.0422562141491396,
+      "grad_norm": 0.8205150961875916,
+      "learning_rate": 5e-06,
+      "loss": 0.0483,
+      "step": 5451
+    },
+    {
+      "epoch": 1.0424474187380497,
+      "grad_norm": 1.924567699432373,
+      "learning_rate": 5e-06,
+      "loss": 0.2065,
+      "step": 5452
+    },
+    {
+      "epoch": 1.0426386233269598,
+      "grad_norm": 0.8093776702880859,
+      "learning_rate": 5e-06,
+      "loss": 0.0619,
+      "step": 5453
+    },
+    {
+      "epoch": 1.0428298279158699,
+      "grad_norm": 1.8981060981750488,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 5454
+    },
+    {
+      "epoch": 1.0430210325047802,
+      "grad_norm": 0.724181056022644,
+      "learning_rate": 5e-06,
+      "loss": 0.0284,
+      "step": 5455
+    },
+    {
+      "epoch": 1.0432122370936903,
+      "grad_norm": 1.9111442565917969,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 5456
+    },
+    {
+      "epoch": 1.0434034416826004,
+      "grad_norm": 1.8289096355438232,
+      "learning_rate": 5e-06,
+      "loss": 0.1876,
+      "step": 5457
+    },
+    {
+      "epoch": 1.0435946462715104,
+      "grad_norm": 2.535822629928589,
+      "learning_rate": 5e-06,
+      "loss": 0.2191,
+      "step": 5458
+    },
+    {
+      "epoch": 1.0437858508604207,
+      "grad_norm": 1.2734262943267822,
+      "learning_rate": 5e-06,
+      "loss": 0.0492,
+      "step": 5459
+    },
+    {
+      "epoch": 1.0439770554493308,
+      "grad_norm": 0.6029118895530701,
+      "learning_rate": 5e-06,
+      "loss": 0.0392,
+      "step": 5460
+    },
+    {
+      "epoch": 1.044168260038241,
+      "grad_norm": 1.2994762659072876,
+      "learning_rate": 5e-06,
+      "loss": 0.058,
+      "step": 5461
+    },
+    {
+      "epoch": 1.044359464627151,
+      "grad_norm": 1.6217557191848755,
+      "learning_rate": 5e-06,
+      "loss": 0.1581,
+      "step": 5462
+    },
+    {
+      "epoch": 1.044550669216061,
+      "grad_norm": 1.316620111465454,
+      "learning_rate": 5e-06,
+      "loss": 0.0736,
+      "step": 5463
+    },
+    {
+      "epoch": 1.0447418738049714,
+      "grad_norm": 1.5537855625152588,
+      "learning_rate": 5e-06,
+      "loss": 0.1094,
+      "step": 5464
+    },
+    {
+      "epoch": 1.0449330783938815,
+      "grad_norm": 0.6778096556663513,
+      "learning_rate": 5e-06,
+      "loss": 0.0312,
+      "step": 5465
+    },
+    {
+      "epoch": 1.0451242829827916,
+      "grad_norm": 1.1639436483383179,
+      "learning_rate": 5e-06,
+      "loss": 0.0931,
+      "step": 5466
+    },
+    {
+      "epoch": 1.0453154875717017,
+      "grad_norm": 1.2169827222824097,
+      "learning_rate": 5e-06,
+      "loss": 0.035,
+      "step": 5467
+    },
+    {
+      "epoch": 1.0455066921606118,
+      "grad_norm": 3.1080920696258545,
+      "learning_rate": 5e-06,
+      "loss": 0.2154,
+      "step": 5468
+    },
+    {
+      "epoch": 1.045697896749522,
+      "grad_norm": 1.4423012733459473,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 5469
+    },
+    {
+      "epoch": 1.0458891013384322,
+      "grad_norm": 1.3293002843856812,
+      "learning_rate": 5e-06,
+      "loss": 0.112,
+      "step": 5470
+    },
+    {
+      "epoch": 1.0460803059273422,
+      "grad_norm": 4.677468299865723,
+      "learning_rate": 5e-06,
+      "loss": 0.1826,
+      "step": 5471
+    },
+    {
+      "epoch": 1.0462715105162523,
+      "grad_norm": 1.4991474151611328,
+      "learning_rate": 5e-06,
+      "loss": 0.1735,
+      "step": 5472
+    },
+    {
+      "epoch": 1.0464627151051624,
+      "grad_norm": 0.8098369836807251,
+      "learning_rate": 5e-06,
+      "loss": 0.0329,
+      "step": 5473
+    },
+    {
+      "epoch": 1.0466539196940727,
+      "grad_norm": 1.0252405405044556,
+      "learning_rate": 5e-06,
+      "loss": 0.0512,
+      "step": 5474
+    },
+    {
+      "epoch": 1.0468451242829828,
+      "grad_norm": 2.2641191482543945,
+      "learning_rate": 5e-06,
+      "loss": 0.263,
+      "step": 5475
+    },
+    {
+      "epoch": 1.047036328871893,
+      "grad_norm": 2.4479317665100098,
+      "learning_rate": 5e-06,
+      "loss": 0.1826,
+      "step": 5476
+    },
+    {
+      "epoch": 1.047227533460803,
+      "grad_norm": 1.0218003988265991,
+      "learning_rate": 5e-06,
+      "loss": 0.0605,
+      "step": 5477
+    },
+    {
+      "epoch": 1.0474187380497133,
+      "grad_norm": 1.1280349493026733,
+      "learning_rate": 5e-06,
+      "loss": 0.0742,
+      "step": 5478
+    },
+    {
+      "epoch": 1.0476099426386234,
+      "grad_norm": 2.0644752979278564,
+      "learning_rate": 5e-06,
+      "loss": 0.1082,
+      "step": 5479
+    },
+    {
+      "epoch": 1.0478011472275335,
+      "grad_norm": 1.3084979057312012,
+      "learning_rate": 5e-06,
+      "loss": 0.0601,
+      "step": 5480
+    },
+    {
+      "epoch": 1.0479923518164436,
+      "grad_norm": 1.2070064544677734,
+      "learning_rate": 5e-06,
+      "loss": 0.0784,
+      "step": 5481
+    },
+    {
+      "epoch": 1.0481835564053537,
+      "grad_norm": 1.7970086336135864,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 5482
+    },
+    {
+      "epoch": 1.048374760994264,
+      "grad_norm": 2.143139362335205,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 5483
+    },
+    {
+      "epoch": 1.048565965583174,
+      "grad_norm": 1.9227676391601562,
+      "learning_rate": 5e-06,
+      "loss": 0.1499,
+      "step": 5484
+    },
+    {
+      "epoch": 1.0487571701720841,
+      "grad_norm": 1.6543242931365967,
+      "learning_rate": 5e-06,
+      "loss": 0.0493,
+      "step": 5485
+    },
+    {
+      "epoch": 1.0489483747609942,
+      "grad_norm": 0.5971057415008545,
+      "learning_rate": 5e-06,
+      "loss": 0.0124,
+      "step": 5486
+    },
+    {
+      "epoch": 1.0491395793499043,
+      "grad_norm": 2.2405505180358887,
+      "learning_rate": 5e-06,
+      "loss": 0.2293,
+      "step": 5487
+    },
+    {
+      "epoch": 1.0493307839388146,
+      "grad_norm": 2.6099843978881836,
+      "learning_rate": 5e-06,
+      "loss": 0.2646,
+      "step": 5488
+    },
+    {
+      "epoch": 1.0495219885277247,
+      "grad_norm": 1.8190488815307617,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 5489
+    },
+    {
+      "epoch": 1.0497131931166348,
+      "grad_norm": 2.8577523231506348,
+      "learning_rate": 5e-06,
+      "loss": 0.2463,
+      "step": 5490
+    },
+    {
+      "epoch": 1.049904397705545,
+      "grad_norm": 1.3947685956954956,
+      "learning_rate": 5e-06,
+      "loss": 0.115,
+      "step": 5491
+    },
+    {
+      "epoch": 1.050095602294455,
+      "grad_norm": 1.2213077545166016,
+      "learning_rate": 5e-06,
+      "loss": 0.0604,
+      "step": 5492
+    },
+    {
+      "epoch": 1.0502868068833653,
+      "grad_norm": 1.0234959125518799,
+      "learning_rate": 5e-06,
+      "loss": 0.0541,
+      "step": 5493
+    },
+    {
+      "epoch": 1.0504780114722754,
+      "grad_norm": 2.8558273315429688,
+      "learning_rate": 5e-06,
+      "loss": 0.266,
+      "step": 5494
+    },
+    {
+      "epoch": 1.0506692160611855,
+      "grad_norm": 1.4367798566818237,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 5495
+    },
+    {
+      "epoch": 1.0508604206500956,
+      "grad_norm": 1.1142265796661377,
+      "learning_rate": 5e-06,
+      "loss": 0.0389,
+      "step": 5496
+    },
+    {
+      "epoch": 1.0510516252390056,
+      "grad_norm": 2.527897596359253,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 5497
+    },
+    {
+      "epoch": 1.051242829827916,
+      "grad_norm": 1.8113329410552979,
+      "learning_rate": 5e-06,
+      "loss": 0.1046,
+      "step": 5498
+    },
+    {
+      "epoch": 1.051434034416826,
+      "grad_norm": 1.4811005592346191,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 5499
+    },
+    {
+      "epoch": 1.0516252390057361,
+      "grad_norm": 1.7655647993087769,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 5500
+    },
+    {
+      "epoch": 1.0516252390057361,
+      "eval_runtime": 834.4534,
+      "eval_samples_per_second": 1.838,
+      "eval_steps_per_second": 0.23,
+      "step": 5500
+    },
+    {
+      "epoch": 1.0518164435946462,
+      "grad_norm": 1.7707321643829346,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 5501
+    },
+    {
+      "epoch": 1.0520076481835563,
+      "grad_norm": 1.482149362564087,
+      "learning_rate": 5e-06,
+      "loss": 0.0581,
+      "step": 5502
+    },
+    {
+      "epoch": 1.0521988527724666,
+      "grad_norm": 0.7621336579322815,
+      "learning_rate": 5e-06,
+      "loss": 0.057,
+      "step": 5503
+    },
+    {
+      "epoch": 1.0523900573613767,
+      "grad_norm": 1.2675679922103882,
+      "learning_rate": 5e-06,
+      "loss": 0.0503,
+      "step": 5504
+    },
+    {
+      "epoch": 1.0525812619502868,
+      "grad_norm": 1.968985676765442,
+      "learning_rate": 5e-06,
+      "loss": 0.0581,
+      "step": 5505
+    },
+    {
+      "epoch": 1.0527724665391969,
+      "grad_norm": 2.8859829902648926,
+      "learning_rate": 5e-06,
+      "loss": 0.2671,
+      "step": 5506
+    },
+    {
+      "epoch": 1.0529636711281072,
+      "grad_norm": 0.8013049960136414,
+      "learning_rate": 5e-06,
+      "loss": 0.045,
+      "step": 5507
+    },
+    {
+      "epoch": 1.0531548757170173,
+      "grad_norm": 1.266425609588623,
+      "learning_rate": 5e-06,
+      "loss": 0.0874,
+      "step": 5508
+    },
+    {
+      "epoch": 1.0533460803059274,
+      "grad_norm": 2.2374284267425537,
+      "learning_rate": 5e-06,
+      "loss": 0.0577,
+      "step": 5509
+    },
+    {
+      "epoch": 1.0535372848948374,
+      "grad_norm": 1.1194673776626587,
+      "learning_rate": 5e-06,
+      "loss": 0.0354,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0537284894837475,
+      "grad_norm": 1.0510578155517578,
+      "learning_rate": 5e-06,
+      "loss": 0.03,
+      "step": 5511
+    },
+    {
+      "epoch": 1.0539196940726578,
+      "grad_norm": 1.7931699752807617,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 5512
+    },
+    {
+      "epoch": 1.054110898661568,
+      "grad_norm": 1.9386789798736572,
+      "learning_rate": 5e-06,
+      "loss": 0.0997,
+      "step": 5513
+    },
+    {
+      "epoch": 1.054302103250478,
+      "grad_norm": 0.9754608869552612,
+      "learning_rate": 5e-06,
+      "loss": 0.0676,
+      "step": 5514
+    },
+    {
+      "epoch": 1.054493307839388,
+      "grad_norm": 1.582482099533081,
+      "learning_rate": 5e-06,
+      "loss": 0.0941,
+      "step": 5515
+    },
+    {
+      "epoch": 1.0546845124282982,
+      "grad_norm": 0.8954532742500305,
+      "learning_rate": 5e-06,
+      "loss": 0.0351,
+      "step": 5516
+    },
+    {
+      "epoch": 1.0548757170172085,
+      "grad_norm": 1.9438811540603638,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 5517
+    },
+    {
+      "epoch": 1.0550669216061186,
+      "grad_norm": 2.3192455768585205,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 5518
+    },
+    {
+      "epoch": 1.0552581261950287,
+      "grad_norm": 2.396188974380493,
+      "learning_rate": 5e-06,
+      "loss": 0.2317,
+      "step": 5519
+    },
+    {
+      "epoch": 1.0554493307839388,
+      "grad_norm": 1.4901149272918701,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 5520
+    },
+    {
+      "epoch": 1.0556405353728489,
+      "grad_norm": 1.5792475938796997,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 5521
+    },
+    {
+      "epoch": 1.0558317399617592,
+      "grad_norm": 1.3586615324020386,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 5522
+    },
+    {
+      "epoch": 1.0560229445506693,
+      "grad_norm": 1.5009024143218994,
+      "learning_rate": 5e-06,
+      "loss": 0.0418,
+      "step": 5523
+    },
+    {
+      "epoch": 1.0562141491395793,
+      "grad_norm": 1.5317716598510742,
+      "learning_rate": 5e-06,
+      "loss": 0.0772,
+      "step": 5524
+    },
+    {
+      "epoch": 1.0564053537284894,
+      "grad_norm": 1.119912028312683,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 5525
+    },
+    {
+      "epoch": 1.0565965583173995,
+      "grad_norm": 1.4832195043563843,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 5526
+    },
+    {
+      "epoch": 1.0567877629063098,
+      "grad_norm": 1.1946576833724976,
+      "learning_rate": 5e-06,
+      "loss": 0.0627,
+      "step": 5527
+    },
+    {
+      "epoch": 1.05697896749522,
+      "grad_norm": 1.3606984615325928,
+      "learning_rate": 5e-06,
+      "loss": 0.0645,
+      "step": 5528
+    },
+    {
+      "epoch": 1.05717017208413,
+      "grad_norm": 0.7441090941429138,
+      "learning_rate": 5e-06,
+      "loss": 0.0384,
+      "step": 5529
+    },
+    {
+      "epoch": 1.05736137667304,
+      "grad_norm": 1.1623265743255615,
+      "learning_rate": 5e-06,
+      "loss": 0.0436,
+      "step": 5530
+    },
+    {
+      "epoch": 1.0575525812619504,
+      "grad_norm": 1.458216667175293,
+      "learning_rate": 5e-06,
+      "loss": 0.1655,
+      "step": 5531
+    },
+    {
+      "epoch": 1.0577437858508605,
+      "grad_norm": 1.6108118295669556,
+      "learning_rate": 5e-06,
+      "loss": 0.1881,
+      "step": 5532
+    },
+    {
+      "epoch": 1.0579349904397706,
+      "grad_norm": 1.7551403045654297,
+      "learning_rate": 5e-06,
+      "loss": 0.1642,
+      "step": 5533
+    },
+    {
+      "epoch": 1.0581261950286807,
+      "grad_norm": 1.0185822248458862,
+      "learning_rate": 5e-06,
+      "loss": 0.055,
+      "step": 5534
+    },
+    {
+      "epoch": 1.0583173996175907,
+      "grad_norm": 0.6421904563903809,
+      "learning_rate": 5e-06,
+      "loss": 0.0271,
+      "step": 5535
+    },
+    {
+      "epoch": 1.058508604206501,
+      "grad_norm": 1.085301160812378,
+      "learning_rate": 5e-06,
+      "loss": 0.0458,
+      "step": 5536
+    },
+    {
+      "epoch": 1.0586998087954111,
+      "grad_norm": 1.3988016843795776,
+      "learning_rate": 5e-06,
+      "loss": 0.0921,
+      "step": 5537
+    },
+    {
+      "epoch": 1.0588910133843212,
+      "grad_norm": 3.1126415729522705,
+      "learning_rate": 5e-06,
+      "loss": 0.2017,
+      "step": 5538
+    },
+    {
+      "epoch": 1.0590822179732313,
+      "grad_norm": 1.5114021301269531,
+      "learning_rate": 5e-06,
+      "loss": 0.0989,
+      "step": 5539
+    },
+    {
+      "epoch": 1.0592734225621414,
+      "grad_norm": 1.0168532133102417,
+      "learning_rate": 5e-06,
+      "loss": 0.0519,
+      "step": 5540
+    },
+    {
+      "epoch": 1.0594646271510517,
+      "grad_norm": 0.8101119995117188,
+      "learning_rate": 5e-06,
+      "loss": 0.0505,
+      "step": 5541
+    },
+    {
+      "epoch": 1.0596558317399618,
+      "grad_norm": 1.7537641525268555,
+      "learning_rate": 5e-06,
+      "loss": 0.0264,
+      "step": 5542
+    },
+    {
+      "epoch": 1.059847036328872,
+      "grad_norm": 1.9035402536392212,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 5543
+    },
+    {
+      "epoch": 1.060038240917782,
+      "grad_norm": 2.1393039226531982,
+      "learning_rate": 5e-06,
+      "loss": 0.2744,
+      "step": 5544
+    },
+    {
+      "epoch": 1.060229445506692,
+      "grad_norm": 1.2142505645751953,
+      "learning_rate": 5e-06,
+      "loss": 0.0952,
+      "step": 5545
+    },
+    {
+      "epoch": 1.0604206500956024,
+      "grad_norm": 1.868961215019226,
+      "learning_rate": 5e-06,
+      "loss": 0.0702,
+      "step": 5546
+    },
+    {
+      "epoch": 1.0606118546845125,
+      "grad_norm": 1.7776259183883667,
+      "learning_rate": 5e-06,
+      "loss": 0.198,
+      "step": 5547
+    },
+    {
+      "epoch": 1.0608030592734226,
+      "grad_norm": 0.7262550592422485,
+      "learning_rate": 5e-06,
+      "loss": 0.0306,
+      "step": 5548
+    },
+    {
+      "epoch": 1.0609942638623326,
+      "grad_norm": 1.554850697517395,
+      "learning_rate": 5e-06,
+      "loss": 0.0685,
+      "step": 5549
+    },
+    {
+      "epoch": 1.0611854684512427,
+      "grad_norm": 2.132505178451538,
+      "learning_rate": 5e-06,
+      "loss": 0.3059,
+      "step": 5550
+    },
+    {
+      "epoch": 1.061376673040153,
+      "grad_norm": 0.8302657008171082,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 5551
+    },
+    {
+      "epoch": 1.0615678776290631,
+      "grad_norm": 1.9379445314407349,
+      "learning_rate": 5e-06,
+      "loss": 0.1882,
+      "step": 5552
+    },
+    {
+      "epoch": 1.0617590822179732,
+      "grad_norm": 1.474511742591858,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 5553
+    },
+    {
+      "epoch": 1.0619502868068833,
+      "grad_norm": 2.032550573348999,
+      "learning_rate": 5e-06,
+      "loss": 0.1122,
+      "step": 5554
+    },
+    {
+      "epoch": 1.0621414913957934,
+      "grad_norm": 1.345967411994934,
+      "learning_rate": 5e-06,
+      "loss": 0.0447,
+      "step": 5555
+    },
+    {
+      "epoch": 1.0623326959847037,
+      "grad_norm": 2.1029436588287354,
+      "learning_rate": 5e-06,
+      "loss": 0.2457,
+      "step": 5556
+    },
+    {
+      "epoch": 1.0625239005736138,
+      "grad_norm": 1.188673734664917,
+      "learning_rate": 5e-06,
+      "loss": 0.0634,
+      "step": 5557
+    },
+    {
+      "epoch": 1.0627151051625239,
+      "grad_norm": 0.6115692257881165,
+      "learning_rate": 5e-06,
+      "loss": 0.0683,
+      "step": 5558
+    },
+    {
+      "epoch": 1.062906309751434,
+      "grad_norm": 1.1898462772369385,
+      "learning_rate": 5e-06,
+      "loss": 0.042,
+      "step": 5559
+    },
+    {
+      "epoch": 1.063097514340344,
+      "grad_norm": 2.9459636211395264,
+      "learning_rate": 5e-06,
+      "loss": 0.1648,
+      "step": 5560
+    },
+    {
+      "epoch": 1.0632887189292544,
+      "grad_norm": 1.1826893091201782,
+      "learning_rate": 5e-06,
+      "loss": 0.0368,
+      "step": 5561
+    },
+    {
+      "epoch": 1.0634799235181644,
+      "grad_norm": 2.073742389678955,
+      "learning_rate": 5e-06,
+      "loss": 0.2756,
+      "step": 5562
+    },
+    {
+      "epoch": 1.0636711281070745,
+      "grad_norm": 1.3164788484573364,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 5563
+    },
+    {
+      "epoch": 1.0638623326959846,
+      "grad_norm": 1.7094725370407104,
+      "learning_rate": 5e-06,
+      "loss": 0.2081,
+      "step": 5564
+    },
+    {
+      "epoch": 1.064053537284895,
+      "grad_norm": 1.2266261577606201,
+      "learning_rate": 5e-06,
+      "loss": 0.0791,
+      "step": 5565
+    },
+    {
+      "epoch": 1.064244741873805,
+      "grad_norm": 0.5611611008644104,
+      "learning_rate": 5e-06,
+      "loss": 0.0304,
+      "step": 5566
+    },
+    {
+      "epoch": 1.064435946462715,
+      "grad_norm": 1.0190863609313965,
+      "learning_rate": 5e-06,
+      "loss": 0.0428,
+      "step": 5567
+    },
+    {
+      "epoch": 1.0646271510516252,
+      "grad_norm": 1.5948188304901123,
+      "learning_rate": 5e-06,
+      "loss": 0.1038,
+      "step": 5568
+    },
+    {
+      "epoch": 1.0648183556405353,
+      "grad_norm": 1.694000005722046,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 5569
+    },
+    {
+      "epoch": 1.0650095602294456,
+      "grad_norm": 2.386840581893921,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 5570
+    },
+    {
+      "epoch": 1.0652007648183557,
+      "grad_norm": 2.1206300258636475,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 5571
+    },
+    {
+      "epoch": 1.0653919694072658,
+      "grad_norm": 1.195825457572937,
+      "learning_rate": 5e-06,
+      "loss": 0.0482,
+      "step": 5572
+    },
+    {
+      "epoch": 1.0655831739961759,
+      "grad_norm": 0.9524969458580017,
+      "learning_rate": 5e-06,
+      "loss": 0.0435,
+      "step": 5573
+    },
+    {
+      "epoch": 1.065774378585086,
+      "grad_norm": 1.50506591796875,
+      "learning_rate": 5e-06,
+      "loss": 0.1009,
+      "step": 5574
+    },
+    {
+      "epoch": 1.0659655831739963,
+      "grad_norm": 2.291487216949463,
+      "learning_rate": 5e-06,
+      "loss": 0.2958,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0661567877629063,
+      "grad_norm": 1.8258750438690186,
+      "learning_rate": 5e-06,
+      "loss": 0.2111,
+      "step": 5576
+    },
+    {
+      "epoch": 1.0663479923518164,
+      "grad_norm": 1.87051260471344,
+      "learning_rate": 5e-06,
+      "loss": 0.1976,
+      "step": 5577
+    },
+    {
+      "epoch": 1.0665391969407265,
+      "grad_norm": 2.5715811252593994,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 5578
+    },
+    {
+      "epoch": 1.0667304015296368,
+      "grad_norm": 0.7487394213676453,
+      "learning_rate": 5e-06,
+      "loss": 0.0376,
+      "step": 5579
+    },
+    {
+      "epoch": 1.066921606118547,
+      "grad_norm": 2.3859152793884277,
+      "learning_rate": 5e-06,
+      "loss": 0.0333,
+      "step": 5580
+    },
+    {
+      "epoch": 1.067112810707457,
+      "grad_norm": 1.7255477905273438,
+      "learning_rate": 5e-06,
+      "loss": 0.072,
+      "step": 5581
+    },
+    {
+      "epoch": 1.067304015296367,
+      "grad_norm": 2.3104193210601807,
+      "learning_rate": 5e-06,
+      "loss": 0.3214,
+      "step": 5582
+    },
+    {
+      "epoch": 1.0674952198852772,
+      "grad_norm": 1.6485390663146973,
+      "learning_rate": 5e-06,
+      "loss": 0.1561,
+      "step": 5583
+    },
+    {
+      "epoch": 1.0676864244741875,
+      "grad_norm": 0.9040722846984863,
+      "learning_rate": 5e-06,
+      "loss": 0.0759,
+      "step": 5584
+    },
+    {
+      "epoch": 1.0678776290630976,
+      "grad_norm": 0.904682993888855,
+      "learning_rate": 5e-06,
+      "loss": 0.0422,
+      "step": 5585
+    },
+    {
+      "epoch": 1.0680688336520077,
+      "grad_norm": 1.8263367414474487,
+      "learning_rate": 5e-06,
+      "loss": 0.0747,
+      "step": 5586
+    },
+    {
+      "epoch": 1.0682600382409178,
+      "grad_norm": 2.2079832553863525,
+      "learning_rate": 5e-06,
+      "loss": 0.3012,
+      "step": 5587
+    },
+    {
+      "epoch": 1.0684512428298278,
+      "grad_norm": 1.0028715133666992,
+      "learning_rate": 5e-06,
+      "loss": 0.0642,
+      "step": 5588
+    },
+    {
+      "epoch": 1.0686424474187382,
+      "grad_norm": 2.1110336780548096,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 5589
+    },
+    {
+      "epoch": 1.0688336520076482,
+      "grad_norm": 1.1553751230239868,
+      "learning_rate": 5e-06,
+      "loss": 0.0648,
+      "step": 5590
+    },
+    {
+      "epoch": 1.0690248565965583,
+      "grad_norm": 1.4524354934692383,
+      "learning_rate": 5e-06,
+      "loss": 0.0671,
+      "step": 5591
+    },
+    {
+      "epoch": 1.0692160611854684,
+      "grad_norm": 1.1088978052139282,
+      "learning_rate": 5e-06,
+      "loss": 0.0488,
+      "step": 5592
+    },
+    {
+      "epoch": 1.0694072657743785,
+      "grad_norm": 1.0848307609558105,
+      "learning_rate": 5e-06,
+      "loss": 0.0353,
+      "step": 5593
+    },
+    {
+      "epoch": 1.0695984703632888,
+      "grad_norm": 1.9532465934753418,
+      "learning_rate": 5e-06,
+      "loss": 0.2058,
+      "step": 5594
+    },
+    {
+      "epoch": 1.069789674952199,
+      "grad_norm": 2.134166955947876,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 5595
+    },
+    {
+      "epoch": 1.069980879541109,
+      "grad_norm": 0.9889646172523499,
+      "learning_rate": 5e-06,
+      "loss": 0.076,
+      "step": 5596
+    },
+    {
+      "epoch": 1.070172084130019,
+      "grad_norm": 4.0119452476501465,
+      "learning_rate": 5e-06,
+      "loss": 0.2381,
+      "step": 5597
+    },
+    {
+      "epoch": 1.0703632887189292,
+      "grad_norm": 1.1719545125961304,
+      "learning_rate": 5e-06,
+      "loss": 0.0532,
+      "step": 5598
+    },
+    {
+      "epoch": 1.0705544933078395,
+      "grad_norm": 0.7966266870498657,
+      "learning_rate": 5e-06,
+      "loss": 0.0488,
+      "step": 5599
+    },
+    {
+      "epoch": 1.0707456978967496,
+      "grad_norm": 2.092799186706543,
+      "learning_rate": 5e-06,
+      "loss": 0.2781,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0709369024856596,
+      "grad_norm": 1.770919680595398,
+      "learning_rate": 5e-06,
+      "loss": 0.2382,
+      "step": 5601
+    },
+    {
+      "epoch": 1.0711281070745697,
+      "grad_norm": 1.2337734699249268,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 5602
+    },
+    {
+      "epoch": 1.0713193116634798,
+      "grad_norm": 1.2285529375076294,
+      "learning_rate": 5e-06,
+      "loss": 0.0923,
+      "step": 5603
+    },
+    {
+      "epoch": 1.0715105162523901,
+      "grad_norm": 0.8008958101272583,
+      "learning_rate": 5e-06,
+      "loss": 0.0347,
+      "step": 5604
+    },
+    {
+      "epoch": 1.0717017208413002,
+      "grad_norm": 1.2167035341262817,
+      "learning_rate": 5e-06,
+      "loss": 0.0495,
+      "step": 5605
+    },
+    {
+      "epoch": 1.0718929254302103,
+      "grad_norm": 2.4168663024902344,
+      "learning_rate": 5e-06,
+      "loss": 0.3622,
+      "step": 5606
+    },
+    {
+      "epoch": 1.0720841300191204,
+      "grad_norm": 2.050773859024048,
+      "learning_rate": 5e-06,
+      "loss": 0.3046,
+      "step": 5607
+    },
+    {
+      "epoch": 1.0722753346080305,
+      "grad_norm": 2.666419267654419,
+      "learning_rate": 5e-06,
+      "loss": 0.1967,
+      "step": 5608
+    },
+    {
+      "epoch": 1.0724665391969408,
+      "grad_norm": 0.6062523722648621,
+      "learning_rate": 5e-06,
+      "loss": 0.0457,
+      "step": 5609
+    },
+    {
+      "epoch": 1.0726577437858509,
+      "grad_norm": 1.7348066568374634,
+      "learning_rate": 5e-06,
+      "loss": 0.0553,
+      "step": 5610
+    },
+    {
+      "epoch": 1.072848948374761,
+      "grad_norm": 2.2407853603363037,
+      "learning_rate": 5e-06,
+      "loss": 0.0718,
+      "step": 5611
+    },
+    {
+      "epoch": 1.073040152963671,
+      "grad_norm": 2.270958662033081,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 5612
+    },
+    {
+      "epoch": 1.0732313575525811,
+      "grad_norm": 2.6858391761779785,
+      "learning_rate": 5e-06,
+      "loss": 0.3042,
+      "step": 5613
+    },
+    {
+      "epoch": 1.0734225621414915,
+      "grad_norm": 2.1360089778900146,
+      "learning_rate": 5e-06,
+      "loss": 0.2963,
+      "step": 5614
+    },
+    {
+      "epoch": 1.0736137667304015,
+      "grad_norm": 1.7626913785934448,
+      "learning_rate": 5e-06,
+      "loss": 0.1656,
+      "step": 5615
+    },
+    {
+      "epoch": 1.0738049713193116,
+      "grad_norm": 0.7246912717819214,
+      "learning_rate": 5e-06,
+      "loss": 0.0466,
+      "step": 5616
+    },
+    {
+      "epoch": 1.0739961759082217,
+      "grad_norm": 1.5797908306121826,
+      "learning_rate": 5e-06,
+      "loss": 0.0513,
+      "step": 5617
+    },
+    {
+      "epoch": 1.074187380497132,
+      "grad_norm": 1.0937782526016235,
+      "learning_rate": 5e-06,
+      "loss": 0.066,
+      "step": 5618
+    },
+    {
+      "epoch": 1.0743785850860421,
+      "grad_norm": 1.6355293989181519,
+      "learning_rate": 5e-06,
+      "loss": 0.2058,
+      "step": 5619
+    },
+    {
+      "epoch": 1.0745697896749522,
+      "grad_norm": 1.8476656675338745,
+      "learning_rate": 5e-06,
+      "loss": 0.2098,
+      "step": 5620
+    },
+    {
+      "epoch": 1.0747609942638623,
+      "grad_norm": 1.877700686454773,
+      "learning_rate": 5e-06,
+      "loss": 0.0347,
+      "step": 5621
+    },
+    {
+      "epoch": 1.0749521988527724,
+      "grad_norm": 1.5795397758483887,
+      "learning_rate": 5e-06,
+      "loss": 0.092,
+      "step": 5622
+    },
+    {
+      "epoch": 1.0751434034416827,
+      "grad_norm": 1.5501025915145874,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 5623
+    },
+    {
+      "epoch": 1.0753346080305928,
+      "grad_norm": 1.7347816228866577,
+      "learning_rate": 5e-06,
+      "loss": 0.0974,
+      "step": 5624
+    },
+    {
+      "epoch": 1.0755258126195029,
+      "grad_norm": 2.559631109237671,
+      "learning_rate": 5e-06,
+      "loss": 0.3785,
+      "step": 5625
+    },
+    {
+      "epoch": 1.075717017208413,
+      "grad_norm": 1.182547688484192,
+      "learning_rate": 5e-06,
+      "loss": 0.0788,
+      "step": 5626
+    },
+    {
+      "epoch": 1.075908221797323,
+      "grad_norm": 1.3121072053909302,
+      "learning_rate": 5e-06,
+      "loss": 0.0665,
+      "step": 5627
+    },
+    {
+      "epoch": 1.0760994263862333,
+      "grad_norm": 1.6444857120513916,
+      "learning_rate": 5e-06,
+      "loss": 0.0654,
+      "step": 5628
+    },
+    {
+      "epoch": 1.0762906309751434,
+      "grad_norm": 1.5076884031295776,
+      "learning_rate": 5e-06,
+      "loss": 0.0763,
+      "step": 5629
+    },
+    {
+      "epoch": 1.0764818355640535,
+      "grad_norm": 1.1843534708023071,
+      "learning_rate": 5e-06,
+      "loss": 0.037,
+      "step": 5630
+    },
+    {
+      "epoch": 1.0766730401529636,
+      "grad_norm": 1.5834217071533203,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 5631
+    },
+    {
+      "epoch": 1.076864244741874,
+      "grad_norm": 2.579793691635132,
+      "learning_rate": 5e-06,
+      "loss": 0.2568,
+      "step": 5632
+    },
+    {
+      "epoch": 1.077055449330784,
+      "grad_norm": 1.1158941984176636,
+      "learning_rate": 5e-06,
+      "loss": 0.0999,
+      "step": 5633
+    },
+    {
+      "epoch": 1.077246653919694,
+      "grad_norm": 1.8821864128112793,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 5634
+    },
+    {
+      "epoch": 1.0774378585086042,
+      "grad_norm": 1.1883410215377808,
+      "learning_rate": 5e-06,
+      "loss": 0.0513,
+      "step": 5635
+    },
+    {
+      "epoch": 1.0776290630975143,
+      "grad_norm": 2.1233952045440674,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 5636
+    },
+    {
+      "epoch": 1.0778202676864246,
+      "grad_norm": 2.0192222595214844,
+      "learning_rate": 5e-06,
+      "loss": 0.2886,
+      "step": 5637
+    },
+    {
+      "epoch": 1.0780114722753347,
+      "grad_norm": 1.116824746131897,
+      "learning_rate": 5e-06,
+      "loss": 0.081,
+      "step": 5638
+    },
+    {
+      "epoch": 1.0782026768642448,
+      "grad_norm": 1.6397048234939575,
+      "learning_rate": 5e-06,
+      "loss": 0.0968,
+      "step": 5639
+    },
+    {
+      "epoch": 1.0783938814531548,
+      "grad_norm": 0.804574191570282,
+      "learning_rate": 5e-06,
+      "loss": 0.0296,
+      "step": 5640
+    },
+    {
+      "epoch": 1.078585086042065,
+      "grad_norm": 0.6931762099266052,
+      "learning_rate": 5e-06,
+      "loss": 0.0159,
+      "step": 5641
+    },
+    {
+      "epoch": 1.0787762906309752,
+      "grad_norm": 2.571634531021118,
+      "learning_rate": 5e-06,
+      "loss": 0.1144,
+      "step": 5642
+    },
+    {
+      "epoch": 1.0789674952198853,
+      "grad_norm": 1.3164654970169067,
+      "learning_rate": 5e-06,
+      "loss": 0.0838,
+      "step": 5643
+    },
+    {
+      "epoch": 1.0791586998087954,
+      "grad_norm": 1.9440586566925049,
+      "learning_rate": 5e-06,
+      "loss": 0.2427,
+      "step": 5644
+    },
+    {
+      "epoch": 1.0793499043977055,
+      "grad_norm": 1.7131543159484863,
+      "learning_rate": 5e-06,
+      "loss": 0.1543,
+      "step": 5645
+    },
+    {
+      "epoch": 1.0795411089866156,
+      "grad_norm": 0.7819735407829285,
+      "learning_rate": 5e-06,
+      "loss": 0.062,
+      "step": 5646
+    },
+    {
+      "epoch": 1.079732313575526,
+      "grad_norm": 1.6629984378814697,
+      "learning_rate": 5e-06,
+      "loss": 0.0666,
+      "step": 5647
+    },
+    {
+      "epoch": 1.079923518164436,
+      "grad_norm": 1.143560528755188,
+      "learning_rate": 5e-06,
+      "loss": 0.0335,
+      "step": 5648
+    },
+    {
+      "epoch": 1.080114722753346,
+      "grad_norm": 1.9708220958709717,
+      "learning_rate": 5e-06,
+      "loss": 0.1641,
+      "step": 5649
+    },
+    {
+      "epoch": 1.0803059273422562,
+      "grad_norm": 1.5629487037658691,
+      "learning_rate": 5e-06,
+      "loss": 0.2152,
+      "step": 5650
+    },
+    {
+      "epoch": 1.0804971319311663,
+      "grad_norm": 1.2981854677200317,
+      "learning_rate": 5e-06,
+      "loss": 0.0818,
+      "step": 5651
+    },
+    {
+      "epoch": 1.0806883365200766,
+      "grad_norm": 2.244119644165039,
+      "learning_rate": 5e-06,
+      "loss": 0.2363,
+      "step": 5652
+    },
+    {
+      "epoch": 1.0808795411089867,
+      "grad_norm": 0.7663077116012573,
+      "learning_rate": 5e-06,
+      "loss": 0.0666,
+      "step": 5653
+    },
+    {
+      "epoch": 1.0810707456978967,
+      "grad_norm": 0.9932133555412292,
+      "learning_rate": 5e-06,
+      "loss": 0.0601,
+      "step": 5654
+    },
+    {
+      "epoch": 1.0812619502868068,
+      "grad_norm": 0.8187389969825745,
+      "learning_rate": 5e-06,
+      "loss": 0.0233,
+      "step": 5655
+    },
+    {
+      "epoch": 1.081453154875717,
+      "grad_norm": 1.5282580852508545,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 5656
+    },
+    {
+      "epoch": 1.0816443594646272,
+      "grad_norm": 1.7210948467254639,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 5657
+    },
+    {
+      "epoch": 1.0818355640535373,
+      "grad_norm": 0.6434665322303772,
+      "learning_rate": 5e-06,
+      "loss": 0.0408,
+      "step": 5658
+    },
+    {
+      "epoch": 1.0820267686424474,
+      "grad_norm": 1.558206558227539,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 5659
+    },
+    {
+      "epoch": 1.0822179732313575,
+      "grad_norm": 1.6600244045257568,
+      "learning_rate": 5e-06,
+      "loss": 0.0921,
+      "step": 5660
+    },
+    {
+      "epoch": 1.0824091778202676,
+      "grad_norm": 1.1153446435928345,
+      "learning_rate": 5e-06,
+      "loss": 0.024,
+      "step": 5661
+    },
+    {
+      "epoch": 1.0826003824091779,
+      "grad_norm": 2.0310747623443604,
+      "learning_rate": 5e-06,
+      "loss": 0.1647,
+      "step": 5662
+    },
+    {
+      "epoch": 1.082791586998088,
+      "grad_norm": 1.8215316534042358,
+      "learning_rate": 5e-06,
+      "loss": 0.1737,
+      "step": 5663
+    },
+    {
+      "epoch": 1.082982791586998,
+      "grad_norm": 1.404730200767517,
+      "learning_rate": 5e-06,
+      "loss": 0.0599,
+      "step": 5664
+    },
+    {
+      "epoch": 1.0831739961759081,
+      "grad_norm": 5.3708176612854,
+      "learning_rate": 5e-06,
+      "loss": 0.0709,
+      "step": 5665
+    },
+    {
+      "epoch": 1.0833652007648185,
+      "grad_norm": 2.5323567390441895,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 5666
+    },
+    {
+      "epoch": 1.0835564053537285,
+      "grad_norm": 0.777966320514679,
+      "learning_rate": 5e-06,
+      "loss": 0.0265,
+      "step": 5667
+    },
+    {
+      "epoch": 1.0837476099426386,
+      "grad_norm": 1.7438549995422363,
+      "learning_rate": 5e-06,
+      "loss": 0.0811,
+      "step": 5668
+    },
+    {
+      "epoch": 1.0839388145315487,
+      "grad_norm": 1.896105408668518,
+      "learning_rate": 5e-06,
+      "loss": 0.0971,
+      "step": 5669
+    },
+    {
+      "epoch": 1.0841300191204588,
+      "grad_norm": 1.2311056852340698,
+      "learning_rate": 5e-06,
+      "loss": 0.0739,
+      "step": 5670
+    },
+    {
+      "epoch": 1.0843212237093691,
+      "grad_norm": 1.2870060205459595,
+      "learning_rate": 5e-06,
+      "loss": 0.0791,
+      "step": 5671
+    },
+    {
+      "epoch": 1.0845124282982792,
+      "grad_norm": 1.5939801931381226,
+      "learning_rate": 5e-06,
+      "loss": 0.035,
+      "step": 5672
+    },
+    {
+      "epoch": 1.0847036328871893,
+      "grad_norm": 1.4833265542984009,
+      "learning_rate": 5e-06,
+      "loss": 0.0654,
+      "step": 5673
+    },
+    {
+      "epoch": 1.0848948374760994,
+      "grad_norm": 2.980525255203247,
+      "learning_rate": 5e-06,
+      "loss": 0.3089,
+      "step": 5674
+    },
+    {
+      "epoch": 1.0850860420650095,
+      "grad_norm": 2.04789137840271,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 5675
+    },
+    {
+      "epoch": 1.0852772466539198,
+      "grad_norm": 1.1447137594223022,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 5676
+    },
+    {
+      "epoch": 1.0854684512428299,
+      "grad_norm": 2.3989462852478027,
+      "learning_rate": 5e-06,
+      "loss": 0.2504,
+      "step": 5677
+    },
+    {
+      "epoch": 1.08565965583174,
+      "grad_norm": 1.2400035858154297,
+      "learning_rate": 5e-06,
+      "loss": 0.0553,
+      "step": 5678
+    },
+    {
+      "epoch": 1.08585086042065,
+      "grad_norm": 0.6254027485847473,
+      "learning_rate": 5e-06,
+      "loss": 0.0353,
+      "step": 5679
+    },
+    {
+      "epoch": 1.0860420650095601,
+      "grad_norm": 0.8050343990325928,
+      "learning_rate": 5e-06,
+      "loss": 0.0261,
+      "step": 5680
+    },
+    {
+      "epoch": 1.0862332695984704,
+      "grad_norm": 2.053995132446289,
+      "learning_rate": 5e-06,
+      "loss": 0.3031,
+      "step": 5681
+    },
+    {
+      "epoch": 1.0864244741873805,
+      "grad_norm": 2.055222272872925,
+      "learning_rate": 5e-06,
+      "loss": 0.0618,
+      "step": 5682
+    },
+    {
+      "epoch": 1.0866156787762906,
+      "grad_norm": 1.6199257373809814,
+      "learning_rate": 5e-06,
+      "loss": 0.1672,
+      "step": 5683
+    },
+    {
+      "epoch": 1.0868068833652007,
+      "grad_norm": 1.113048791885376,
+      "learning_rate": 5e-06,
+      "loss": 0.0616,
+      "step": 5684
+    },
+    {
+      "epoch": 1.086998087954111,
+      "grad_norm": 0.8362628817558289,
+      "learning_rate": 5e-06,
+      "loss": 0.0341,
+      "step": 5685
+    },
+    {
+      "epoch": 1.087189292543021,
+      "grad_norm": 5.398261070251465,
+      "learning_rate": 5e-06,
+      "loss": 0.1019,
+      "step": 5686
+    },
+    {
+      "epoch": 1.0873804971319312,
+      "grad_norm": 1.4154852628707886,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 5687
+    },
+    {
+      "epoch": 1.0875717017208413,
+      "grad_norm": 1.3424711227416992,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 5688
+    },
+    {
+      "epoch": 1.0877629063097514,
+      "grad_norm": 1.525865912437439,
+      "learning_rate": 5e-06,
+      "loss": 0.0945,
+      "step": 5689
+    },
+    {
+      "epoch": 1.0879541108986617,
+      "grad_norm": 1.0463107824325562,
+      "learning_rate": 5e-06,
+      "loss": 0.0452,
+      "step": 5690
+    },
+    {
+      "epoch": 1.0881453154875718,
+      "grad_norm": 2.542323589324951,
+      "learning_rate": 5e-06,
+      "loss": 0.2381,
+      "step": 5691
+    },
+    {
+      "epoch": 1.0883365200764819,
+      "grad_norm": 1.3651113510131836,
+      "learning_rate": 5e-06,
+      "loss": 0.0745,
+      "step": 5692
+    },
+    {
+      "epoch": 1.088527724665392,
+      "grad_norm": 1.5302486419677734,
+      "learning_rate": 5e-06,
+      "loss": 0.1264,
+      "step": 5693
+    },
+    {
+      "epoch": 1.088718929254302,
+      "grad_norm": 1.7708359956741333,
+      "learning_rate": 5e-06,
+      "loss": 0.1359,
+      "step": 5694
+    },
+    {
+      "epoch": 1.0889101338432123,
+      "grad_norm": 1.8701834678649902,
+      "learning_rate": 5e-06,
+      "loss": 0.2137,
+      "step": 5695
+    },
+    {
+      "epoch": 1.0891013384321224,
+      "grad_norm": 2.5269713401794434,
+      "learning_rate": 5e-06,
+      "loss": 0.0765,
+      "step": 5696
+    },
+    {
+      "epoch": 1.0892925430210325,
+      "grad_norm": 1.7278193235397339,
+      "learning_rate": 5e-06,
+      "loss": 0.0885,
+      "step": 5697
+    },
+    {
+      "epoch": 1.0894837476099426,
+      "grad_norm": 1.453526496887207,
+      "learning_rate": 5e-06,
+      "loss": 0.0511,
+      "step": 5698
+    },
+    {
+      "epoch": 1.0896749521988527,
+      "grad_norm": 5.5870819091796875,
+      "learning_rate": 5e-06,
+      "loss": 0.2771,
+      "step": 5699
+    },
+    {
+      "epoch": 1.089866156787763,
+      "grad_norm": 1.4349433183670044,
+      "learning_rate": 5e-06,
+      "loss": 0.1705,
+      "step": 5700
+    },
+    {
+      "epoch": 1.090057361376673,
+      "grad_norm": 1.717117190361023,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 5701
+    },
+    {
+      "epoch": 1.0902485659655832,
+      "grad_norm": 1.0715373754501343,
+      "learning_rate": 5e-06,
+      "loss": 0.0745,
+      "step": 5702
+    },
+    {
+      "epoch": 1.0904397705544933,
+      "grad_norm": 1.4231574535369873,
+      "learning_rate": 5e-06,
+      "loss": 0.0805,
+      "step": 5703
+    },
+    {
+      "epoch": 1.0906309751434033,
+      "grad_norm": 2.196079730987549,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 5704
+    },
+    {
+      "epoch": 1.0908221797323137,
+      "grad_norm": 1.016573429107666,
+      "learning_rate": 5e-06,
+      "loss": 0.0582,
+      "step": 5705
+    },
+    {
+      "epoch": 1.0910133843212237,
+      "grad_norm": 1.7154649496078491,
+      "learning_rate": 5e-06,
+      "loss": 0.2169,
+      "step": 5706
+    },
+    {
+      "epoch": 1.0912045889101338,
+      "grad_norm": 1.9732598066329956,
+      "learning_rate": 5e-06,
+      "loss": 0.1264,
+      "step": 5707
+    },
+    {
+      "epoch": 1.091395793499044,
+      "grad_norm": 1.975980281829834,
+      "learning_rate": 5e-06,
+      "loss": 0.1932,
+      "step": 5708
+    },
+    {
+      "epoch": 1.091586998087954,
+      "grad_norm": 1.5534021854400635,
+      "learning_rate": 5e-06,
+      "loss": 0.0715,
+      "step": 5709
+    },
+    {
+      "epoch": 1.0917782026768643,
+      "grad_norm": 1.1966547966003418,
+      "learning_rate": 5e-06,
+      "loss": 0.0663,
+      "step": 5710
+    },
+    {
+      "epoch": 1.0919694072657744,
+      "grad_norm": 1.5349960327148438,
+      "learning_rate": 5e-06,
+      "loss": 0.0343,
+      "step": 5711
+    },
+    {
+      "epoch": 1.0921606118546845,
+      "grad_norm": 1.9171141386032104,
+      "learning_rate": 5e-06,
+      "loss": 0.0983,
+      "step": 5712
+    },
+    {
+      "epoch": 1.0923518164435946,
+      "grad_norm": 2.5431854724884033,
+      "learning_rate": 5e-06,
+      "loss": 0.4502,
+      "step": 5713
+    },
+    {
+      "epoch": 1.0925430210325047,
+      "grad_norm": 1.5429447889328003,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 5714
+    },
+    {
+      "epoch": 1.092734225621415,
+      "grad_norm": 2.018566131591797,
+      "learning_rate": 5e-06,
+      "loss": 0.2262,
+      "step": 5715
+    },
+    {
+      "epoch": 1.092925430210325,
+      "grad_norm": 1.1679967641830444,
+      "learning_rate": 5e-06,
+      "loss": 0.0731,
+      "step": 5716
+    },
+    {
+      "epoch": 1.0931166347992352,
+      "grad_norm": 1.0205624103546143,
+      "learning_rate": 5e-06,
+      "loss": 0.0543,
+      "step": 5717
+    },
+    {
+      "epoch": 1.0933078393881452,
+      "grad_norm": 1.0302752256393433,
+      "learning_rate": 5e-06,
+      "loss": 0.0479,
+      "step": 5718
+    },
+    {
+      "epoch": 1.0934990439770556,
+      "grad_norm": 2.270318031311035,
+      "learning_rate": 5e-06,
+      "loss": 0.251,
+      "step": 5719
+    },
+    {
+      "epoch": 1.0936902485659656,
+      "grad_norm": 1.8095113039016724,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 5720
+    },
+    {
+      "epoch": 1.0938814531548757,
+      "grad_norm": 0.7488591074943542,
+      "learning_rate": 5e-06,
+      "loss": 0.0514,
+      "step": 5721
+    },
+    {
+      "epoch": 1.0940726577437858,
+      "grad_norm": 1.4656505584716797,
+      "learning_rate": 5e-06,
+      "loss": 0.0985,
+      "step": 5722
+    },
+    {
+      "epoch": 1.094263862332696,
+      "grad_norm": 0.7714589238166809,
+      "learning_rate": 5e-06,
+      "loss": 0.0423,
+      "step": 5723
+    },
+    {
+      "epoch": 1.0944550669216062,
+      "grad_norm": 2.106182813644409,
+      "learning_rate": 5e-06,
+      "loss": 0.2728,
+      "step": 5724
+    },
+    {
+      "epoch": 1.0946462715105163,
+      "grad_norm": 2.305906057357788,
+      "learning_rate": 5e-06,
+      "loss": 0.2627,
+      "step": 5725
+    },
+    {
+      "epoch": 1.0948374760994264,
+      "grad_norm": 0.6955013275146484,
+      "learning_rate": 5e-06,
+      "loss": 0.0651,
+      "step": 5726
+    },
+    {
+      "epoch": 1.0950286806883365,
+      "grad_norm": 1.7944916486740112,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 5727
+    },
+    {
+      "epoch": 1.0952198852772466,
+      "grad_norm": 2.637842893600464,
+      "learning_rate": 5e-06,
+      "loss": 0.1017,
+      "step": 5728
+    },
+    {
+      "epoch": 1.0954110898661569,
+      "grad_norm": 0.9952343106269836,
+      "learning_rate": 5e-06,
+      "loss": 0.0618,
+      "step": 5729
+    },
+    {
+      "epoch": 1.095602294455067,
+      "grad_norm": 1.2145094871520996,
+      "learning_rate": 5e-06,
+      "loss": 0.0363,
+      "step": 5730
+    },
+    {
+      "epoch": 1.095793499043977,
+      "grad_norm": 2.777398109436035,
+      "learning_rate": 5e-06,
+      "loss": 0.1873,
+      "step": 5731
+    },
+    {
+      "epoch": 1.0959847036328871,
+      "grad_norm": 0.9324513673782349,
+      "learning_rate": 5e-06,
+      "loss": 0.069,
+      "step": 5732
+    },
+    {
+      "epoch": 1.0961759082217972,
+      "grad_norm": 1.6317458152770996,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 5733
+    },
+    {
+      "epoch": 1.0963671128107075,
+      "grad_norm": 1.5389405488967896,
+      "learning_rate": 5e-06,
+      "loss": 0.0722,
+      "step": 5734
+    },
+    {
+      "epoch": 1.0965583173996176,
+      "grad_norm": 3.062220335006714,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 5735
+    },
+    {
+      "epoch": 1.0967495219885277,
+      "grad_norm": 1.457374930381775,
+      "learning_rate": 5e-06,
+      "loss": 0.0802,
+      "step": 5736
+    },
+    {
+      "epoch": 1.0969407265774378,
+      "grad_norm": 2.036764144897461,
+      "learning_rate": 5e-06,
+      "loss": 0.1513,
+      "step": 5737
+    },
+    {
+      "epoch": 1.097131931166348,
+      "grad_norm": 2.1664798259735107,
+      "learning_rate": 5e-06,
+      "loss": 0.2764,
+      "step": 5738
+    },
+    {
+      "epoch": 1.0973231357552582,
+      "grad_norm": 1.5670701265335083,
+      "learning_rate": 5e-06,
+      "loss": 0.0703,
+      "step": 5739
+    },
+    {
+      "epoch": 1.0975143403441683,
+      "grad_norm": 0.9358064532279968,
+      "learning_rate": 5e-06,
+      "loss": 0.0345,
+      "step": 5740
+    },
+    {
+      "epoch": 1.0977055449330784,
+      "grad_norm": 1.4006162881851196,
+      "learning_rate": 5e-06,
+      "loss": 0.0829,
+      "step": 5741
+    },
+    {
+      "epoch": 1.0978967495219885,
+      "grad_norm": 1.403589129447937,
+      "learning_rate": 5e-06,
+      "loss": 0.061,
+      "step": 5742
+    },
+    {
+      "epoch": 1.0980879541108988,
+      "grad_norm": 1.3776750564575195,
+      "learning_rate": 5e-06,
+      "loss": 0.0641,
+      "step": 5743
+    },
+    {
+      "epoch": 1.0982791586998089,
+      "grad_norm": 1.898293375968933,
+      "learning_rate": 5e-06,
+      "loss": 0.1924,
+      "step": 5744
+    },
+    {
+      "epoch": 1.098470363288719,
+      "grad_norm": 1.39720618724823,
+      "learning_rate": 5e-06,
+      "loss": 0.0802,
+      "step": 5745
+    },
+    {
+      "epoch": 1.098661567877629,
+      "grad_norm": 1.3132373094558716,
+      "learning_rate": 5e-06,
+      "loss": 0.063,
+      "step": 5746
+    },
+    {
+      "epoch": 1.0988527724665391,
+      "grad_norm": 1.5557122230529785,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 5747
+    },
+    {
+      "epoch": 1.0990439770554494,
+      "grad_norm": 2.059504508972168,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 5748
+    },
+    {
+      "epoch": 1.0992351816443595,
+      "grad_norm": 1.1423454284667969,
+      "learning_rate": 5e-06,
+      "loss": 0.0325,
+      "step": 5749
+    },
+    {
+      "epoch": 1.0994263862332696,
+      "grad_norm": 2.8969271183013916,
+      "learning_rate": 5e-06,
+      "loss": 0.4175,
+      "step": 5750
+    },
+    {
+      "epoch": 1.0996175908221797,
+      "grad_norm": 1.9927480220794678,
+      "learning_rate": 5e-06,
+      "loss": 0.0938,
+      "step": 5751
+    },
+    {
+      "epoch": 1.0998087954110898,
+      "grad_norm": 1.3723220825195312,
+      "learning_rate": 5e-06,
+      "loss": 0.1006,
+      "step": 5752
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7870752811431885,
+      "learning_rate": 5e-06,
+      "loss": 0.0486,
+      "step": 5753
+    },
+    {
+      "epoch": 1.1001912045889102,
+      "grad_norm": 0.7222342491149902,
+      "learning_rate": 5e-06,
+      "loss": 0.0125,
+      "step": 5754
+    },
+    {
+      "epoch": 1.1003824091778203,
+      "grad_norm": 1.8501266241073608,
+      "learning_rate": 5e-06,
+      "loss": 0.0527,
+      "step": 5755
+    },
+    {
+      "epoch": 1.1005736137667304,
+      "grad_norm": 2.527247905731201,
+      "learning_rate": 5e-06,
+      "loss": 0.2819,
+      "step": 5756
+    },
+    {
+      "epoch": 1.1007648183556404,
+      "grad_norm": 2.091769218444824,
+      "learning_rate": 5e-06,
+      "loss": 0.2612,
+      "step": 5757
+    },
+    {
+      "epoch": 1.1009560229445507,
+      "grad_norm": 1.6816740036010742,
+      "learning_rate": 5e-06,
+      "loss": 0.1683,
+      "step": 5758
+    },
+    {
+      "epoch": 1.1011472275334608,
+      "grad_norm": 1.4403516054153442,
+      "learning_rate": 5e-06,
+      "loss": 0.0814,
+      "step": 5759
+    },
+    {
+      "epoch": 1.101338432122371,
+      "grad_norm": 0.9857239723205566,
+      "learning_rate": 5e-06,
+      "loss": 0.0488,
+      "step": 5760
+    },
+    {
+      "epoch": 1.101529636711281,
+      "grad_norm": 2.1389448642730713,
+      "learning_rate": 5e-06,
+      "loss": 0.0507,
+      "step": 5761
+    },
+    {
+      "epoch": 1.101720841300191,
+      "grad_norm": 1.8587839603424072,
+      "learning_rate": 5e-06,
+      "loss": 0.1574,
+      "step": 5762
+    },
+    {
+      "epoch": 1.1019120458891014,
+      "grad_norm": 1.851593017578125,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 5763
+    },
+    {
+      "epoch": 1.1021032504780115,
+      "grad_norm": 1.861345648765564,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 5764
+    },
+    {
+      "epoch": 1.1022944550669216,
+      "grad_norm": 1.249029278755188,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 5765
+    },
+    {
+      "epoch": 1.1024856596558317,
+      "grad_norm": 2.417471408843994,
+      "learning_rate": 5e-06,
+      "loss": 0.2049,
+      "step": 5766
+    },
+    {
+      "epoch": 1.1026768642447418,
+      "grad_norm": 1.392234206199646,
+      "learning_rate": 5e-06,
+      "loss": 0.0759,
+      "step": 5767
+    },
+    {
+      "epoch": 1.102868068833652,
+      "grad_norm": 1.4488508701324463,
+      "learning_rate": 5e-06,
+      "loss": 0.0945,
+      "step": 5768
+    },
+    {
+      "epoch": 1.1030592734225622,
+      "grad_norm": 1.731885552406311,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 5769
+    },
+    {
+      "epoch": 1.1032504780114722,
+      "grad_norm": 1.2528032064437866,
+      "learning_rate": 5e-06,
+      "loss": 0.0849,
+      "step": 5770
+    },
+    {
+      "epoch": 1.1034416826003823,
+      "grad_norm": 1.366706132888794,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 5771
+    },
+    {
+      "epoch": 1.1036328871892926,
+      "grad_norm": 1.9706212282180786,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 5772
+    },
+    {
+      "epoch": 1.1038240917782027,
+      "grad_norm": 6.170831680297852,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 5773
+    },
+    {
+      "epoch": 1.1040152963671128,
+      "grad_norm": 1.9874886274337769,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 5774
+    },
+    {
+      "epoch": 1.104206500956023,
+      "grad_norm": 2.058870553970337,
+      "learning_rate": 5e-06,
+      "loss": 0.2049,
+      "step": 5775
+    },
+    {
+      "epoch": 1.104397705544933,
+      "grad_norm": 1.8878477811813354,
+      "learning_rate": 5e-06,
+      "loss": 0.2064,
+      "step": 5776
+    },
+    {
+      "epoch": 1.1045889101338433,
+      "grad_norm": 0.7636244893074036,
+      "learning_rate": 5e-06,
+      "loss": 0.0408,
+      "step": 5777
+    },
+    {
+      "epoch": 1.1047801147227534,
+      "grad_norm": 1.490784764289856,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 5778
+    },
+    {
+      "epoch": 1.1049713193116635,
+      "grad_norm": 0.795039713382721,
+      "learning_rate": 5e-06,
+      "loss": 0.0424,
+      "step": 5779
+    },
+    {
+      "epoch": 1.1051625239005736,
+      "grad_norm": 1.5659568309783936,
+      "learning_rate": 5e-06,
+      "loss": 0.0788,
+      "step": 5780
+    },
+    {
+      "epoch": 1.1053537284894837,
+      "grad_norm": 0.8034716844558716,
+      "learning_rate": 5e-06,
+      "loss": 0.0485,
+      "step": 5781
+    },
+    {
+      "epoch": 1.105544933078394,
+      "grad_norm": 1.0298188924789429,
+      "learning_rate": 5e-06,
+      "loss": 0.0677,
+      "step": 5782
+    },
+    {
+      "epoch": 1.105736137667304,
+      "grad_norm": 1.8984624147415161,
+      "learning_rate": 5e-06,
+      "loss": 0.1402,
+      "step": 5783
+    },
+    {
+      "epoch": 1.1059273422562141,
+      "grad_norm": 1.9967960119247437,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 5784
+    },
+    {
+      "epoch": 1.1061185468451242,
+      "grad_norm": 1.45951247215271,
+      "learning_rate": 5e-06,
+      "loss": 0.0652,
+      "step": 5785
+    },
+    {
+      "epoch": 1.1063097514340343,
+      "grad_norm": 1.4071576595306396,
+      "learning_rate": 5e-06,
+      "loss": 0.0426,
+      "step": 5786
+    },
+    {
+      "epoch": 1.1065009560229446,
+      "grad_norm": 3.093616247177124,
+      "learning_rate": 5e-06,
+      "loss": 0.2499,
+      "step": 5787
+    },
+    {
+      "epoch": 1.1066921606118547,
+      "grad_norm": 1.2860229015350342,
+      "learning_rate": 5e-06,
+      "loss": 0.0862,
+      "step": 5788
+    },
+    {
+      "epoch": 1.1068833652007648,
+      "grad_norm": 1.9504213333129883,
+      "learning_rate": 5e-06,
+      "loss": 0.2004,
+      "step": 5789
+    },
+    {
+      "epoch": 1.107074569789675,
+      "grad_norm": 2.4501025676727295,
+      "learning_rate": 5e-06,
+      "loss": 0.1979,
+      "step": 5790
+    },
+    {
+      "epoch": 1.1072657743785852,
+      "grad_norm": 3.6061577796936035,
+      "learning_rate": 5e-06,
+      "loss": 0.1549,
+      "step": 5791
+    },
+    {
+      "epoch": 1.1074569789674953,
+      "grad_norm": 0.46892136335372925,
+      "learning_rate": 5e-06,
+      "loss": 0.0101,
+      "step": 5792
+    },
+    {
+      "epoch": 1.1076481835564054,
+      "grad_norm": 1.5113166570663452,
+      "learning_rate": 5e-06,
+      "loss": 0.0745,
+      "step": 5793
+    },
+    {
+      "epoch": 1.1078393881453155,
+      "grad_norm": 2.696958303451538,
+      "learning_rate": 5e-06,
+      "loss": 0.3514,
+      "step": 5794
+    },
+    {
+      "epoch": 1.1080305927342256,
+      "grad_norm": 2.0832767486572266,
+      "learning_rate": 5e-06,
+      "loss": 0.1931,
+      "step": 5795
+    },
+    {
+      "epoch": 1.1082217973231359,
+      "grad_norm": 1.0346436500549316,
+      "learning_rate": 5e-06,
+      "loss": 0.0442,
+      "step": 5796
+    },
+    {
+      "epoch": 1.108413001912046,
+      "grad_norm": 2.7505767345428467,
+      "learning_rate": 5e-06,
+      "loss": 0.134,
+      "step": 5797
+    },
+    {
+      "epoch": 1.108604206500956,
+      "grad_norm": 1.7536537647247314,
+      "learning_rate": 5e-06,
+      "loss": 0.1032,
+      "step": 5798
+    },
+    {
+      "epoch": 1.1087954110898661,
+      "grad_norm": 0.7006078362464905,
+      "learning_rate": 5e-06,
+      "loss": 0.0423,
+      "step": 5799
+    },
+    {
+      "epoch": 1.1089866156787762,
+      "grad_norm": 1.4535229206085205,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 5800
+    },
+    {
+      "epoch": 1.1091778202676865,
+      "grad_norm": 2.1496787071228027,
+      "learning_rate": 5e-06,
+      "loss": 0.2779,
+      "step": 5801
+    },
+    {
+      "epoch": 1.1093690248565966,
+      "grad_norm": 2.135427474975586,
+      "learning_rate": 5e-06,
+      "loss": 0.2009,
+      "step": 5802
+    },
+    {
+      "epoch": 1.1095602294455067,
+      "grad_norm": 2.1016106605529785,
+      "learning_rate": 5e-06,
+      "loss": 0.111,
+      "step": 5803
+    },
+    {
+      "epoch": 1.1097514340344168,
+      "grad_norm": 0.7748586535453796,
+      "learning_rate": 5e-06,
+      "loss": 0.064,
+      "step": 5804
+    },
+    {
+      "epoch": 1.1099426386233269,
+      "grad_norm": 1.6331055164337158,
+      "learning_rate": 5e-06,
+      "loss": 0.0694,
+      "step": 5805
+    },
+    {
+      "epoch": 1.1101338432122372,
+      "grad_norm": 1.2938159704208374,
+      "learning_rate": 5e-06,
+      "loss": 0.0924,
+      "step": 5806
+    },
+    {
+      "epoch": 1.1103250478011473,
+      "grad_norm": 1.352454662322998,
+      "learning_rate": 5e-06,
+      "loss": 0.074,
+      "step": 5807
+    },
+    {
+      "epoch": 1.1105162523900574,
+      "grad_norm": 1.8348606824874878,
+      "learning_rate": 5e-06,
+      "loss": 0.2089,
+      "step": 5808
+    },
+    {
+      "epoch": 1.1107074569789674,
+      "grad_norm": 1.0210434198379517,
+      "learning_rate": 5e-06,
+      "loss": 0.0382,
+      "step": 5809
+    },
+    {
+      "epoch": 1.1108986615678775,
+      "grad_norm": 2.106792449951172,
+      "learning_rate": 5e-06,
+      "loss": 0.1903,
+      "step": 5810
+    },
+    {
+      "epoch": 1.1110898661567878,
+      "grad_norm": 1.2394472360610962,
+      "learning_rate": 5e-06,
+      "loss": 0.0562,
+      "step": 5811
+    },
+    {
+      "epoch": 1.111281070745698,
+      "grad_norm": 1.395771861076355,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 5812
+    },
+    {
+      "epoch": 1.111472275334608,
+      "grad_norm": 3.2685039043426514,
+      "learning_rate": 5e-06,
+      "loss": 0.1339,
+      "step": 5813
+    },
+    {
+      "epoch": 1.111663479923518,
+      "grad_norm": 1.9594659805297852,
+      "learning_rate": 5e-06,
+      "loss": 0.2694,
+      "step": 5814
+    },
+    {
+      "epoch": 1.1118546845124282,
+      "grad_norm": 1.7652355432510376,
+      "learning_rate": 5e-06,
+      "loss": 0.0963,
+      "step": 5815
+    },
+    {
+      "epoch": 1.1120458891013385,
+      "grad_norm": 2.3200769424438477,
+      "learning_rate": 5e-06,
+      "loss": 0.2094,
+      "step": 5816
+    },
+    {
+      "epoch": 1.1122370936902486,
+      "grad_norm": 1.5320895910263062,
+      "learning_rate": 5e-06,
+      "loss": 0.0662,
+      "step": 5817
+    },
+    {
+      "epoch": 1.1124282982791587,
+      "grad_norm": 0.7863326668739319,
+      "learning_rate": 5e-06,
+      "loss": 0.044,
+      "step": 5818
+    },
+    {
+      "epoch": 1.1126195028680688,
+      "grad_norm": 1.5986088514328003,
+      "learning_rate": 5e-06,
+      "loss": 0.2212,
+      "step": 5819
+    },
+    {
+      "epoch": 1.1128107074569789,
+      "grad_norm": 0.8004074096679688,
+      "learning_rate": 5e-06,
+      "loss": 0.0729,
+      "step": 5820
+    },
+    {
+      "epoch": 1.1130019120458892,
+      "grad_norm": 1.8215408325195312,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 5821
+    },
+    {
+      "epoch": 1.1131931166347993,
+      "grad_norm": 0.8063507080078125,
+      "learning_rate": 5e-06,
+      "loss": 0.0511,
+      "step": 5822
+    },
+    {
+      "epoch": 1.1133843212237093,
+      "grad_norm": 0.7384288311004639,
+      "learning_rate": 5e-06,
+      "loss": 0.0304,
+      "step": 5823
+    },
+    {
+      "epoch": 1.1135755258126194,
+      "grad_norm": 0.910090446472168,
+      "learning_rate": 5e-06,
+      "loss": 0.0327,
+      "step": 5824
+    },
+    {
+      "epoch": 1.1137667304015297,
+      "grad_norm": 2.1890387535095215,
+      "learning_rate": 5e-06,
+      "loss": 0.2288,
+      "step": 5825
+    },
+    {
+      "epoch": 1.1139579349904398,
+      "grad_norm": 0.9365002512931824,
+      "learning_rate": 5e-06,
+      "loss": 0.074,
+      "step": 5826
+    },
+    {
+      "epoch": 1.11414913957935,
+      "grad_norm": 1.6376757621765137,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 5827
+    },
+    {
+      "epoch": 1.11434034416826,
+      "grad_norm": 0.6365956664085388,
+      "learning_rate": 5e-06,
+      "loss": 0.0318,
+      "step": 5828
+    },
+    {
+      "epoch": 1.11453154875717,
+      "grad_norm": 2.4572699069976807,
+      "learning_rate": 5e-06,
+      "loss": 0.1096,
+      "step": 5829
+    },
+    {
+      "epoch": 1.1147227533460804,
+      "grad_norm": 0.6337371468544006,
+      "learning_rate": 5e-06,
+      "loss": 0.0204,
+      "step": 5830
+    },
+    {
+      "epoch": 1.1149139579349905,
+      "grad_norm": 1.9503843784332275,
+      "learning_rate": 5e-06,
+      "loss": 0.2206,
+      "step": 5831
+    },
+    {
+      "epoch": 1.1151051625239006,
+      "grad_norm": 2.674628973007202,
+      "learning_rate": 5e-06,
+      "loss": 0.2115,
+      "step": 5832
+    },
+    {
+      "epoch": 1.1152963671128107,
+      "grad_norm": 1.2188706398010254,
+      "learning_rate": 5e-06,
+      "loss": 0.0726,
+      "step": 5833
+    },
+    {
+      "epoch": 1.1154875717017207,
+      "grad_norm": 4.067685127258301,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 5834
+    },
+    {
+      "epoch": 1.115678776290631,
+      "grad_norm": 1.1010500192642212,
+      "learning_rate": 5e-06,
+      "loss": 0.0777,
+      "step": 5835
+    },
+    {
+      "epoch": 1.1158699808795411,
+      "grad_norm": 1.6503252983093262,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 5836
+    },
+    {
+      "epoch": 1.1160611854684512,
+      "grad_norm": 1.9464383125305176,
+      "learning_rate": 5e-06,
+      "loss": 0.2607,
+      "step": 5837
+    },
+    {
+      "epoch": 1.1162523900573613,
+      "grad_norm": 2.022468090057373,
+      "learning_rate": 5e-06,
+      "loss": 0.2771,
+      "step": 5838
+    },
+    {
+      "epoch": 1.1164435946462714,
+      "grad_norm": 1.6697824001312256,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 5839
+    },
+    {
+      "epoch": 1.1166347992351817,
+      "grad_norm": 2.806473731994629,
+      "learning_rate": 5e-06,
+      "loss": 0.282,
+      "step": 5840
+    },
+    {
+      "epoch": 1.1168260038240918,
+      "grad_norm": 1.8859554529190063,
+      "learning_rate": 5e-06,
+      "loss": 0.1404,
+      "step": 5841
+    },
+    {
+      "epoch": 1.117017208413002,
+      "grad_norm": 2.372626543045044,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 5842
+    },
+    {
+      "epoch": 1.117208413001912,
+      "grad_norm": 1.7864800691604614,
+      "learning_rate": 5e-06,
+      "loss": 0.093,
+      "step": 5843
+    },
+    {
+      "epoch": 1.1173996175908223,
+      "grad_norm": 2.1923136711120605,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 5844
+    },
+    {
+      "epoch": 1.1175908221797324,
+      "grad_norm": 1.0703550577163696,
+      "learning_rate": 5e-06,
+      "loss": 0.0707,
+      "step": 5845
+    },
+    {
+      "epoch": 1.1177820267686425,
+      "grad_norm": 0.8329720497131348,
+      "learning_rate": 5e-06,
+      "loss": 0.073,
+      "step": 5846
+    },
+    {
+      "epoch": 1.1179732313575526,
+      "grad_norm": 1.4551035165786743,
+      "learning_rate": 5e-06,
+      "loss": 0.0402,
+      "step": 5847
+    },
+    {
+      "epoch": 1.1181644359464626,
+      "grad_norm": 1.0404424667358398,
+      "learning_rate": 5e-06,
+      "loss": 0.0406,
+      "step": 5848
+    },
+    {
+      "epoch": 1.118355640535373,
+      "grad_norm": 1.1678119897842407,
+      "learning_rate": 5e-06,
+      "loss": 0.0683,
+      "step": 5849
+    },
+    {
+      "epoch": 1.118546845124283,
+      "grad_norm": 3.122912883758545,
+      "learning_rate": 5e-06,
+      "loss": 0.496,
+      "step": 5850
+    },
+    {
+      "epoch": 1.1187380497131931,
+      "grad_norm": 1.2659235000610352,
+      "learning_rate": 5e-06,
+      "loss": 0.0904,
+      "step": 5851
+    },
+    {
+      "epoch": 1.1189292543021032,
+      "grad_norm": 2.43969988822937,
+      "learning_rate": 5e-06,
+      "loss": 0.308,
+      "step": 5852
+    },
+    {
+      "epoch": 1.1191204588910133,
+      "grad_norm": 1.8691391944885254,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 5853
+    },
+    {
+      "epoch": 1.1193116634799236,
+      "grad_norm": 2.168302059173584,
+      "learning_rate": 5e-06,
+      "loss": 0.0874,
+      "step": 5854
+    },
+    {
+      "epoch": 1.1195028680688337,
+      "grad_norm": 1.7108036279678345,
+      "learning_rate": 5e-06,
+      "loss": 0.0426,
+      "step": 5855
+    },
+    {
+      "epoch": 1.1196940726577438,
+      "grad_norm": 2.807419538497925,
+      "learning_rate": 5e-06,
+      "loss": 0.3353,
+      "step": 5856
+    },
+    {
+      "epoch": 1.1198852772466539,
+      "grad_norm": 1.1391634941101074,
+      "learning_rate": 5e-06,
+      "loss": 0.0533,
+      "step": 5857
+    },
+    {
+      "epoch": 1.120076481835564,
+      "grad_norm": 0.9980553984642029,
+      "learning_rate": 5e-06,
+      "loss": 0.0557,
+      "step": 5858
+    },
+    {
+      "epoch": 1.1202676864244743,
+      "grad_norm": 1.4949246644973755,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 5859
+    },
+    {
+      "epoch": 1.1204588910133844,
+      "grad_norm": 0.8271896839141846,
+      "learning_rate": 5e-06,
+      "loss": 0.0383,
+      "step": 5860
+    },
+    {
+      "epoch": 1.1206500956022944,
+      "grad_norm": 1.2064502239227295,
+      "learning_rate": 5e-06,
+      "loss": 0.0586,
+      "step": 5861
+    },
+    {
+      "epoch": 1.1208413001912045,
+      "grad_norm": 1.5888252258300781,
+      "learning_rate": 5e-06,
+      "loss": 0.1418,
+      "step": 5862
+    },
+    {
+      "epoch": 1.1210325047801146,
+      "grad_norm": 1.441205382347107,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 5863
+    },
+    {
+      "epoch": 1.121223709369025,
+      "grad_norm": 1.0657639503479004,
+      "learning_rate": 5e-06,
+      "loss": 0.0368,
+      "step": 5864
+    },
+    {
+      "epoch": 1.121414913957935,
+      "grad_norm": 1.1127941608428955,
+      "learning_rate": 5e-06,
+      "loss": 0.0692,
+      "step": 5865
+    },
+    {
+      "epoch": 1.121606118546845,
+      "grad_norm": 1.5667933225631714,
+      "learning_rate": 5e-06,
+      "loss": 0.0617,
+      "step": 5866
+    },
+    {
+      "epoch": 1.1217973231357552,
+      "grad_norm": 1.4817070960998535,
+      "learning_rate": 5e-06,
+      "loss": 0.0875,
+      "step": 5867
+    },
+    {
+      "epoch": 1.1219885277246653,
+      "grad_norm": 2.7602312564849854,
+      "learning_rate": 5e-06,
+      "loss": 0.2847,
+      "step": 5868
+    },
+    {
+      "epoch": 1.1221797323135756,
+      "grad_norm": 2.3214738368988037,
+      "learning_rate": 5e-06,
+      "loss": 0.3105,
+      "step": 5869
+    },
+    {
+      "epoch": 1.1223709369024857,
+      "grad_norm": 1.2657291889190674,
+      "learning_rate": 5e-06,
+      "loss": 0.0898,
+      "step": 5870
+    },
+    {
+      "epoch": 1.1225621414913958,
+      "grad_norm": 1.5459098815917969,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 5871
+    },
+    {
+      "epoch": 1.1227533460803059,
+      "grad_norm": 1.9266481399536133,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 5872
+    },
+    {
+      "epoch": 1.122944550669216,
+      "grad_norm": 1.1586101055145264,
+      "learning_rate": 5e-06,
+      "loss": 0.0829,
+      "step": 5873
+    },
+    {
+      "epoch": 1.1231357552581263,
+      "grad_norm": 1.7313776016235352,
+      "learning_rate": 5e-06,
+      "loss": 0.0654,
+      "step": 5874
+    },
+    {
+      "epoch": 1.1233269598470363,
+      "grad_norm": 2.0912880897521973,
+      "learning_rate": 5e-06,
+      "loss": 0.1978,
+      "step": 5875
+    },
+    {
+      "epoch": 1.1235181644359464,
+      "grad_norm": 2.500584840774536,
+      "learning_rate": 5e-06,
+      "loss": 0.2266,
+      "step": 5876
+    },
+    {
+      "epoch": 1.1237093690248565,
+      "grad_norm": 1.7160296440124512,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 5877
+    },
+    {
+      "epoch": 1.1239005736137668,
+      "grad_norm": 0.9689812660217285,
+      "learning_rate": 5e-06,
+      "loss": 0.0418,
+      "step": 5878
+    },
+    {
+      "epoch": 1.124091778202677,
+      "grad_norm": 1.4203402996063232,
+      "learning_rate": 5e-06,
+      "loss": 0.0807,
+      "step": 5879
+    },
+    {
+      "epoch": 1.124282982791587,
+      "grad_norm": 3.106449842453003,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 5880
+    },
+    {
+      "epoch": 1.124474187380497,
+      "grad_norm": 2.113834857940674,
+      "learning_rate": 5e-06,
+      "loss": 0.2381,
+      "step": 5881
+    },
+    {
+      "epoch": 1.1246653919694072,
+      "grad_norm": 1.5136511325836182,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 5882
+    },
+    {
+      "epoch": 1.1248565965583175,
+      "grad_norm": 2.4438841342926025,
+      "learning_rate": 5e-06,
+      "loss": 0.1835,
+      "step": 5883
+    },
+    {
+      "epoch": 1.1250478011472276,
+      "grad_norm": 1.4762717485427856,
+      "learning_rate": 5e-06,
+      "loss": 0.0685,
+      "step": 5884
+    },
+    {
+      "epoch": 1.1252390057361377,
+      "grad_norm": 0.3290434181690216,
+      "learning_rate": 5e-06,
+      "loss": 0.013,
+      "step": 5885
+    },
+    {
+      "epoch": 1.1254302103250478,
+      "grad_norm": 1.6333457231521606,
+      "learning_rate": 5e-06,
+      "loss": 0.048,
+      "step": 5886
+    },
+    {
+      "epoch": 1.1256214149139578,
+      "grad_norm": 2.1950032711029053,
+      "learning_rate": 5e-06,
+      "loss": 0.192,
+      "step": 5887
+    },
+    {
+      "epoch": 1.1258126195028682,
+      "grad_norm": 0.8317578434944153,
+      "learning_rate": 5e-06,
+      "loss": 0.0635,
+      "step": 5888
+    },
+    {
+      "epoch": 1.1260038240917782,
+      "grad_norm": 1.20002281665802,
+      "learning_rate": 5e-06,
+      "loss": 0.0849,
+      "step": 5889
+    },
+    {
+      "epoch": 1.1261950286806883,
+      "grad_norm": 1.6884593963623047,
+      "learning_rate": 5e-06,
+      "loss": 0.0486,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1263862332695984,
+      "grad_norm": 2.1091883182525635,
+      "learning_rate": 5e-06,
+      "loss": 0.1591,
+      "step": 5891
+    },
+    {
+      "epoch": 1.1265774378585087,
+      "grad_norm": 1.4711389541625977,
+      "learning_rate": 5e-06,
+      "loss": 0.0747,
+      "step": 5892
+    },
+    {
+      "epoch": 1.1267686424474188,
+      "grad_norm": 2.467928409576416,
+      "learning_rate": 5e-06,
+      "loss": 0.3312,
+      "step": 5893
+    },
+    {
+      "epoch": 1.126959847036329,
+      "grad_norm": 1.4449323415756226,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 5894
+    },
+    {
+      "epoch": 1.127151051625239,
+      "grad_norm": 2.1099445819854736,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 5895
+    },
+    {
+      "epoch": 1.127342256214149,
+      "grad_norm": 2.1566996574401855,
+      "learning_rate": 5e-06,
+      "loss": 0.1596,
+      "step": 5896
+    },
+    {
+      "epoch": 1.1275334608030594,
+      "grad_norm": 1.25776207447052,
+      "learning_rate": 5e-06,
+      "loss": 0.0662,
+      "step": 5897
+    },
+    {
+      "epoch": 1.1277246653919695,
+      "grad_norm": 1.7440266609191895,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 5898
+    },
+    {
+      "epoch": 1.1279158699808796,
+      "grad_norm": 1.6530354022979736,
+      "learning_rate": 5e-06,
+      "loss": 0.0837,
+      "step": 5899
+    },
+    {
+      "epoch": 1.1281070745697896,
+      "grad_norm": 2.073582649230957,
+      "learning_rate": 5e-06,
+      "loss": 0.209,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1282982791586997,
+      "grad_norm": 1.2013795375823975,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 5901
+    },
+    {
+      "epoch": 1.12848948374761,
+      "grad_norm": 0.7417505979537964,
+      "learning_rate": 5e-06,
+      "loss": 0.061,
+      "step": 5902
+    },
+    {
+      "epoch": 1.1286806883365201,
+      "grad_norm": 0.7994497418403625,
+      "learning_rate": 5e-06,
+      "loss": 0.0461,
+      "step": 5903
+    },
+    {
+      "epoch": 1.1288718929254302,
+      "grad_norm": 0.9240248203277588,
+      "learning_rate": 5e-06,
+      "loss": 0.0244,
+      "step": 5904
+    },
+    {
+      "epoch": 1.1290630975143403,
+      "grad_norm": 1.514775276184082,
+      "learning_rate": 5e-06,
+      "loss": 0.0728,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1292543021032504,
+      "grad_norm": 2.3339710235595703,
+      "learning_rate": 5e-06,
+      "loss": 0.2318,
+      "step": 5906
+    },
+    {
+      "epoch": 1.1294455066921607,
+      "grad_norm": 1.739105463027954,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 5907
+    },
+    {
+      "epoch": 1.1296367112810708,
+      "grad_norm": 1.3210444450378418,
+      "learning_rate": 5e-06,
+      "loss": 0.1018,
+      "step": 5908
+    },
+    {
+      "epoch": 1.1298279158699809,
+      "grad_norm": 1.95649254322052,
+      "learning_rate": 5e-06,
+      "loss": 0.127,
+      "step": 5909
+    },
+    {
+      "epoch": 1.130019120458891,
+      "grad_norm": 1.2586040496826172,
+      "learning_rate": 5e-06,
+      "loss": 0.0602,
+      "step": 5910
+    },
+    {
+      "epoch": 1.130210325047801,
+      "grad_norm": 0.6684005260467529,
+      "learning_rate": 5e-06,
+      "loss": 0.0093,
+      "step": 5911
+    },
+    {
+      "epoch": 1.1304015296367114,
+      "grad_norm": 1.8613158464431763,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 5912
+    },
+    {
+      "epoch": 1.1305927342256215,
+      "grad_norm": 1.138454794883728,
+      "learning_rate": 5e-06,
+      "loss": 0.0658,
+      "step": 5913
+    },
+    {
+      "epoch": 1.1307839388145315,
+      "grad_norm": 1.3337939977645874,
+      "learning_rate": 5e-06,
+      "loss": 0.1279,
+      "step": 5914
+    },
+    {
+      "epoch": 1.1309751434034416,
+      "grad_norm": 1.7277913093566895,
+      "learning_rate": 5e-06,
+      "loss": 0.0672,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1311663479923517,
+      "grad_norm": 1.699109673500061,
+      "learning_rate": 5e-06,
+      "loss": 0.0633,
+      "step": 5916
+    },
+    {
+      "epoch": 1.131357552581262,
+      "grad_norm": 2.2800168991088867,
+      "learning_rate": 5e-06,
+      "loss": 0.1858,
+      "step": 5917
+    },
+    {
+      "epoch": 1.1315487571701721,
+      "grad_norm": 2.4426426887512207,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 5918
+    },
+    {
+      "epoch": 1.1317399617590822,
+      "grad_norm": 1.9887288808822632,
+      "learning_rate": 5e-06,
+      "loss": 0.2119,
+      "step": 5919
+    },
+    {
+      "epoch": 1.1319311663479923,
+      "grad_norm": 2.3811888694763184,
+      "learning_rate": 5e-06,
+      "loss": 0.2509,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1321223709369024,
+      "grad_norm": 2.353497266769409,
+      "learning_rate": 5e-06,
+      "loss": 0.2354,
+      "step": 5921
+    },
+    {
+      "epoch": 1.1323135755258127,
+      "grad_norm": 0.9172454476356506,
+      "learning_rate": 5e-06,
+      "loss": 0.0662,
+      "step": 5922
+    },
+    {
+      "epoch": 1.1325047801147228,
+      "grad_norm": 2.1203384399414062,
+      "learning_rate": 5e-06,
+      "loss": 0.085,
+      "step": 5923
+    },
+    {
+      "epoch": 1.1326959847036329,
+      "grad_norm": 2.3185367584228516,
+      "learning_rate": 5e-06,
+      "loss": 0.0868,
+      "step": 5924
+    },
+    {
+      "epoch": 1.132887189292543,
+      "grad_norm": 2.7123560905456543,
+      "learning_rate": 5e-06,
+      "loss": 0.3535,
+      "step": 5925
+    },
+    {
+      "epoch": 1.133078393881453,
+      "grad_norm": 2.161722183227539,
+      "learning_rate": 5e-06,
+      "loss": 0.2252,
+      "step": 5926
+    },
+    {
+      "epoch": 1.1332695984703633,
+      "grad_norm": 1.7932413816452026,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 5927
+    },
+    {
+      "epoch": 1.1334608030592734,
+      "grad_norm": 1.5227441787719727,
+      "learning_rate": 5e-06,
+      "loss": 0.0524,
+      "step": 5928
+    },
+    {
+      "epoch": 1.1336520076481835,
+      "grad_norm": 1.1363903284072876,
+      "learning_rate": 5e-06,
+      "loss": 0.065,
+      "step": 5929
+    },
+    {
+      "epoch": 1.1338432122370936,
+      "grad_norm": 2.1632423400878906,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1340344168260037,
+      "grad_norm": 0.6506508588790894,
+      "learning_rate": 5e-06,
+      "loss": 0.0442,
+      "step": 5931
+    },
+    {
+      "epoch": 1.134225621414914,
+      "grad_norm": 3.3030409812927246,
+      "learning_rate": 5e-06,
+      "loss": 0.3152,
+      "step": 5932
+    },
+    {
+      "epoch": 1.134416826003824,
+      "grad_norm": 1.9436111450195312,
+      "learning_rate": 5e-06,
+      "loss": 0.0738,
+      "step": 5933
+    },
+    {
+      "epoch": 1.1346080305927342,
+      "grad_norm": 3.1385626792907715,
+      "learning_rate": 5e-06,
+      "loss": 0.3247,
+      "step": 5934
+    },
+    {
+      "epoch": 1.1347992351816443,
+      "grad_norm": 3.9077584743499756,
+      "learning_rate": 5e-06,
+      "loss": 0.0565,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1349904397705546,
+      "grad_norm": 1.5972917079925537,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 5936
+    },
+    {
+      "epoch": 1.1351816443594647,
+      "grad_norm": 1.5931931734085083,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 5937
+    },
+    {
+      "epoch": 1.1353728489483748,
+      "grad_norm": 1.6486834287643433,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 5938
+    },
+    {
+      "epoch": 1.1355640535372848,
+      "grad_norm": 2.0563056468963623,
+      "learning_rate": 5e-06,
+      "loss": 0.2244,
+      "step": 5939
+    },
+    {
+      "epoch": 1.135755258126195,
+      "grad_norm": 1.0756323337554932,
+      "learning_rate": 5e-06,
+      "loss": 0.0842,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1359464627151052,
+      "grad_norm": 1.615752935409546,
+      "learning_rate": 5e-06,
+      "loss": 0.1011,
+      "step": 5941
+    },
+    {
+      "epoch": 1.1361376673040153,
+      "grad_norm": 1.9002844095230103,
+      "learning_rate": 5e-06,
+      "loss": 0.0837,
+      "step": 5942
+    },
+    {
+      "epoch": 1.1363288718929254,
+      "grad_norm": 1.3055564165115356,
+      "learning_rate": 5e-06,
+      "loss": 0.0799,
+      "step": 5943
+    },
+    {
+      "epoch": 1.1365200764818355,
+      "grad_norm": 2.0098092555999756,
+      "learning_rate": 5e-06,
+      "loss": 0.2034,
+      "step": 5944
+    },
+    {
+      "epoch": 1.1367112810707458,
+      "grad_norm": 1.3458549976348877,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 5945
+    },
+    {
+      "epoch": 1.136902485659656,
+      "grad_norm": 1.6178194284439087,
+      "learning_rate": 5e-06,
+      "loss": 0.1039,
+      "step": 5946
+    },
+    {
+      "epoch": 1.137093690248566,
+      "grad_norm": 2.4097132682800293,
+      "learning_rate": 5e-06,
+      "loss": 0.0646,
+      "step": 5947
+    },
+    {
+      "epoch": 1.137284894837476,
+      "grad_norm": 1.505638599395752,
+      "learning_rate": 5e-06,
+      "loss": 0.0537,
+      "step": 5948
+    },
+    {
+      "epoch": 1.1374760994263862,
+      "grad_norm": 1.2082053422927856,
+      "learning_rate": 5e-06,
+      "loss": 0.047,
+      "step": 5949
+    },
+    {
+      "epoch": 1.1376673040152965,
+      "grad_norm": 1.2429389953613281,
+      "learning_rate": 5e-06,
+      "loss": 0.0812,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1378585086042066,
+      "grad_norm": 2.585109233856201,
+      "learning_rate": 5e-06,
+      "loss": 0.069,
+      "step": 5951
+    },
+    {
+      "epoch": 1.1380497131931167,
+      "grad_norm": 2.0243451595306396,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 5952
+    },
+    {
+      "epoch": 1.1382409177820267,
+      "grad_norm": 1.217773199081421,
+      "learning_rate": 5e-06,
+      "loss": 0.0684,
+      "step": 5953
+    },
+    {
+      "epoch": 1.1384321223709368,
+      "grad_norm": 0.9870648384094238,
+      "learning_rate": 5e-06,
+      "loss": 0.0429,
+      "step": 5954
+    },
+    {
+      "epoch": 1.1386233269598471,
+      "grad_norm": 1.2159374952316284,
+      "learning_rate": 5e-06,
+      "loss": 0.0458,
+      "step": 5955
+    },
+    {
+      "epoch": 1.1388145315487572,
+      "grad_norm": 1.7573827505111694,
+      "learning_rate": 5e-06,
+      "loss": 0.2002,
+      "step": 5956
+    },
+    {
+      "epoch": 1.1390057361376673,
+      "grad_norm": 1.0106160640716553,
+      "learning_rate": 5e-06,
+      "loss": 0.1065,
+      "step": 5957
+    },
+    {
+      "epoch": 1.1391969407265774,
+      "grad_norm": 1.294363260269165,
+      "learning_rate": 5e-06,
+      "loss": 0.0982,
+      "step": 5958
+    },
+    {
+      "epoch": 1.1393881453154875,
+      "grad_norm": 1.239507794380188,
+      "learning_rate": 5e-06,
+      "loss": 0.1389,
+      "step": 5959
+    },
+    {
+      "epoch": 1.1395793499043978,
+      "grad_norm": 1.3383917808532715,
+      "learning_rate": 5e-06,
+      "loss": 0.0797,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1397705544933079,
+      "grad_norm": 1.403836965560913,
+      "learning_rate": 5e-06,
+      "loss": 0.0702,
+      "step": 5961
+    },
+    {
+      "epoch": 1.139961759082218,
+      "grad_norm": 2.712651491165161,
+      "learning_rate": 5e-06,
+      "loss": 0.0612,
+      "step": 5962
+    },
+    {
+      "epoch": 1.140152963671128,
+      "grad_norm": 2.9551162719726562,
+      "learning_rate": 5e-06,
+      "loss": 0.2241,
+      "step": 5963
+    },
+    {
+      "epoch": 1.1403441682600381,
+      "grad_norm": 0.7462462186813354,
+      "learning_rate": 5e-06,
+      "loss": 0.0658,
+      "step": 5964
+    },
+    {
+      "epoch": 1.1405353728489485,
+      "grad_norm": 0.8238846659660339,
+      "learning_rate": 5e-06,
+      "loss": 0.0566,
+      "step": 5965
+    },
+    {
+      "epoch": 1.1407265774378585,
+      "grad_norm": 1.3812743425369263,
+      "learning_rate": 5e-06,
+      "loss": 0.0468,
+      "step": 5966
+    },
+    {
+      "epoch": 1.1409177820267686,
+      "grad_norm": 1.1476503610610962,
+      "learning_rate": 5e-06,
+      "loss": 0.0558,
+      "step": 5967
+    },
+    {
+      "epoch": 1.1411089866156787,
+      "grad_norm": 1.7411783933639526,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 5968
+    },
+    {
+      "epoch": 1.1413001912045888,
+      "grad_norm": 2.1010475158691406,
+      "learning_rate": 5e-06,
+      "loss": 0.176,
+      "step": 5969
+    },
+    {
+      "epoch": 1.1414913957934991,
+      "grad_norm": 0.9700941443443298,
+      "learning_rate": 5e-06,
+      "loss": 0.0863,
+      "step": 5970
+    },
+    {
+      "epoch": 1.1416826003824092,
+      "grad_norm": 1.139196753501892,
+      "learning_rate": 5e-06,
+      "loss": 0.0632,
+      "step": 5971
+    },
+    {
+      "epoch": 1.1418738049713193,
+      "grad_norm": 1.9784907102584839,
+      "learning_rate": 5e-06,
+      "loss": 0.1047,
+      "step": 5972
+    },
+    {
+      "epoch": 1.1420650095602294,
+      "grad_norm": 1.6605274677276611,
+      "learning_rate": 5e-06,
+      "loss": 0.1016,
+      "step": 5973
+    },
+    {
+      "epoch": 1.1422562141491395,
+      "grad_norm": 2.2134573459625244,
+      "learning_rate": 5e-06,
+      "loss": 0.1446,
+      "step": 5974
+    },
+    {
+      "epoch": 1.1424474187380498,
+      "grad_norm": 2.1722609996795654,
+      "learning_rate": 5e-06,
+      "loss": 0.3741,
+      "step": 5975
+    },
+    {
+      "epoch": 1.1426386233269599,
+      "grad_norm": 2.2831599712371826,
+      "learning_rate": 5e-06,
+      "loss": 0.3514,
+      "step": 5976
+    },
+    {
+      "epoch": 1.14282982791587,
+      "grad_norm": 3.3943796157836914,
+      "learning_rate": 5e-06,
+      "loss": 0.4212,
+      "step": 5977
+    },
+    {
+      "epoch": 1.14302103250478,
+      "grad_norm": 1.330841302871704,
+      "learning_rate": 5e-06,
+      "loss": 0.0732,
+      "step": 5978
+    },
+    {
+      "epoch": 1.1432122370936901,
+      "grad_norm": 1.9645999670028687,
+      "learning_rate": 5e-06,
+      "loss": 0.0853,
+      "step": 5979
+    },
+    {
+      "epoch": 1.1434034416826004,
+      "grad_norm": 1.6076546907424927,
+      "learning_rate": 5e-06,
+      "loss": 0.0658,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1435946462715105,
+      "grad_norm": 2.182152509689331,
+      "learning_rate": 5e-06,
+      "loss": 0.1959,
+      "step": 5981
+    },
+    {
+      "epoch": 1.1437858508604206,
+      "grad_norm": 2.1681408882141113,
+      "learning_rate": 5e-06,
+      "loss": 0.2228,
+      "step": 5982
+    },
+    {
+      "epoch": 1.1439770554493307,
+      "grad_norm": 2.4408507347106934,
+      "learning_rate": 5e-06,
+      "loss": 0.2301,
+      "step": 5983
+    },
+    {
+      "epoch": 1.144168260038241,
+      "grad_norm": 1.2057291269302368,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 5984
+    },
+    {
+      "epoch": 1.144359464627151,
+      "grad_norm": 0.7931608557701111,
+      "learning_rate": 5e-06,
+      "loss": 0.0782,
+      "step": 5985
+    },
+    {
+      "epoch": 1.1445506692160612,
+      "grad_norm": 1.4083235263824463,
+      "learning_rate": 5e-06,
+      "loss": 0.0425,
+      "step": 5986
+    },
+    {
+      "epoch": 1.1447418738049713,
+      "grad_norm": 3.754728317260742,
+      "learning_rate": 5e-06,
+      "loss": 0.32,
+      "step": 5987
+    },
+    {
+      "epoch": 1.1449330783938814,
+      "grad_norm": 2.447702407836914,
+      "learning_rate": 5e-06,
+      "loss": 0.2783,
+      "step": 5988
+    },
+    {
+      "epoch": 1.1451242829827917,
+      "grad_norm": 1.0956696271896362,
+      "learning_rate": 5e-06,
+      "loss": 0.1021,
+      "step": 5989
+    },
+    {
+      "epoch": 1.1453154875717018,
+      "grad_norm": 0.831962525844574,
+      "learning_rate": 5e-06,
+      "loss": 0.0399,
+      "step": 5990
+    },
+    {
+      "epoch": 1.1455066921606119,
+      "grad_norm": 1.3386517763137817,
+      "learning_rate": 5e-06,
+      "loss": 0.0901,
+      "step": 5991
+    },
+    {
+      "epoch": 1.145697896749522,
+      "grad_norm": 2.1000680923461914,
+      "learning_rate": 5e-06,
+      "loss": 0.0823,
+      "step": 5992
+    },
+    {
+      "epoch": 1.145889101338432,
+      "grad_norm": 1.4930262565612793,
+      "learning_rate": 5e-06,
+      "loss": 0.0767,
+      "step": 5993
+    },
+    {
+      "epoch": 1.1460803059273423,
+      "grad_norm": 1.3432178497314453,
+      "learning_rate": 5e-06,
+      "loss": 0.1057,
+      "step": 5994
+    },
+    {
+      "epoch": 1.1462715105162524,
+      "grad_norm": 1.8772666454315186,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 5995
+    },
+    {
+      "epoch": 1.1464627151051625,
+      "grad_norm": 0.8968759179115295,
+      "learning_rate": 5e-06,
+      "loss": 0.0477,
+      "step": 5996
+    },
+    {
+      "epoch": 1.1466539196940726,
+      "grad_norm": 1.3898628950119019,
+      "learning_rate": 5e-06,
+      "loss": 0.067,
+      "step": 5997
+    },
+    {
+      "epoch": 1.146845124282983,
+      "grad_norm": 0.6423681974411011,
+      "learning_rate": 5e-06,
+      "loss": 0.0118,
+      "step": 5998
+    },
+    {
+      "epoch": 1.147036328871893,
+      "grad_norm": 1.6495819091796875,
+      "learning_rate": 5e-06,
+      "loss": 0.1907,
+      "step": 5999
+    },
+    {
+      "epoch": 1.147227533460803,
+      "grad_norm": 1.3852123022079468,
+      "learning_rate": 5e-06,
+      "loss": 0.0823,
+      "step": 6000
+    },
+    {
+      "epoch": 1.147227533460803,
+      "eval_runtime": 801.3896,
+      "eval_samples_per_second": 1.914,
+      "eval_steps_per_second": 0.24,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1474187380497132,
+      "grad_norm": 1.957743763923645,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 6001
+    },
+    {
+      "epoch": 1.1476099426386233,
+      "grad_norm": 2.059922218322754,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 6002
+    },
+    {
+      "epoch": 1.1478011472275336,
+      "grad_norm": 1.291016936302185,
+      "learning_rate": 5e-06,
+      "loss": 0.0634,
+      "step": 6003
+    },
+    {
+      "epoch": 1.1479923518164437,
+      "grad_norm": 1.9278820753097534,
+      "learning_rate": 5e-06,
+      "loss": 0.2147,
+      "step": 6004
+    },
+    {
+      "epoch": 1.1481835564053537,
+      "grad_norm": 1.496976613998413,
+      "learning_rate": 5e-06,
+      "loss": 0.0778,
+      "step": 6005
+    },
+    {
+      "epoch": 1.1483747609942638,
+      "grad_norm": 1.579020619392395,
+      "learning_rate": 5e-06,
+      "loss": 0.0946,
+      "step": 6006
+    },
+    {
+      "epoch": 1.148565965583174,
+      "grad_norm": 1.9187579154968262,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 6007
+    },
+    {
+      "epoch": 1.1487571701720842,
+      "grad_norm": 1.770951271057129,
+      "learning_rate": 5e-06,
+      "loss": 0.1904,
+      "step": 6008
+    },
+    {
+      "epoch": 1.1489483747609943,
+      "grad_norm": 1.5869866609573364,
+      "learning_rate": 5e-06,
+      "loss": 0.1424,
+      "step": 6009
+    },
+    {
+      "epoch": 1.1491395793499044,
+      "grad_norm": 2.293226718902588,
+      "learning_rate": 5e-06,
+      "loss": 0.0983,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1493307839388145,
+      "grad_norm": 1.4532525539398193,
+      "learning_rate": 5e-06,
+      "loss": 0.0784,
+      "step": 6011
+    },
+    {
+      "epoch": 1.1495219885277246,
+      "grad_norm": 1.6799381971359253,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 6012
+    },
+    {
+      "epoch": 1.149713193116635,
+      "grad_norm": 1.4979642629623413,
+      "learning_rate": 5e-06,
+      "loss": 0.0636,
+      "step": 6013
+    },
+    {
+      "epoch": 1.149904397705545,
+      "grad_norm": 2.968336582183838,
+      "learning_rate": 5e-06,
+      "loss": 0.3399,
+      "step": 6014
+    },
+    {
+      "epoch": 1.150095602294455,
+      "grad_norm": 2.201011896133423,
+      "learning_rate": 5e-06,
+      "loss": 0.1948,
+      "step": 6015
+    },
+    {
+      "epoch": 1.1502868068833652,
+      "grad_norm": 1.8662936687469482,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 6016
+    },
+    {
+      "epoch": 1.1504780114722752,
+      "grad_norm": 1.6262683868408203,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 6017
+    },
+    {
+      "epoch": 1.1506692160611856,
+      "grad_norm": 1.8292471170425415,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 6018
+    },
+    {
+      "epoch": 1.1508604206500956,
+      "grad_norm": 1.0284324884414673,
+      "learning_rate": 5e-06,
+      "loss": 0.0604,
+      "step": 6019
+    },
+    {
+      "epoch": 1.1510516252390057,
+      "grad_norm": 2.682225465774536,
+      "learning_rate": 5e-06,
+      "loss": 0.2026,
+      "step": 6020
+    },
+    {
+      "epoch": 1.1512428298279158,
+      "grad_norm": 0.996833086013794,
+      "learning_rate": 5e-06,
+      "loss": 0.06,
+      "step": 6021
+    },
+    {
+      "epoch": 1.151434034416826,
+      "grad_norm": 1.5025444030761719,
+      "learning_rate": 5e-06,
+      "loss": 0.0522,
+      "step": 6022
+    },
+    {
+      "epoch": 1.1516252390057362,
+      "grad_norm": 1.4137760400772095,
+      "learning_rate": 5e-06,
+      "loss": 0.079,
+      "step": 6023
+    },
+    {
+      "epoch": 1.1518164435946463,
+      "grad_norm": 2.3708930015563965,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 6024
+    },
+    {
+      "epoch": 1.1520076481835564,
+      "grad_norm": 1.4922761917114258,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 6025
+    },
+    {
+      "epoch": 1.1521988527724665,
+      "grad_norm": 0.9097602367401123,
+      "learning_rate": 5e-06,
+      "loss": 0.0681,
+      "step": 6026
+    },
+    {
+      "epoch": 1.1523900573613766,
+      "grad_norm": 1.8793368339538574,
+      "learning_rate": 5e-06,
+      "loss": 0.1962,
+      "step": 6027
+    },
+    {
+      "epoch": 1.1525812619502869,
+      "grad_norm": 3.2023956775665283,
+      "learning_rate": 5e-06,
+      "loss": 0.0744,
+      "step": 6028
+    },
+    {
+      "epoch": 1.152772466539197,
+      "grad_norm": 1.3651431798934937,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 6029
+    },
+    {
+      "epoch": 1.152963671128107,
+      "grad_norm": 2.276491641998291,
+      "learning_rate": 5e-06,
+      "loss": 0.1804,
+      "step": 6030
+    },
+    {
+      "epoch": 1.1531548757170171,
+      "grad_norm": 1.6140533685684204,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 6031
+    },
+    {
+      "epoch": 1.1533460803059272,
+      "grad_norm": 1.3192780017852783,
+      "learning_rate": 5e-06,
+      "loss": 0.0654,
+      "step": 6032
+    },
+    {
+      "epoch": 1.1535372848948375,
+      "grad_norm": 2.071134567260742,
+      "learning_rate": 5e-06,
+      "loss": 0.2245,
+      "step": 6033
+    },
+    {
+      "epoch": 1.1537284894837476,
+      "grad_norm": 1.0834347009658813,
+      "learning_rate": 5e-06,
+      "loss": 0.0735,
+      "step": 6034
+    },
+    {
+      "epoch": 1.1539196940726577,
+      "grad_norm": 5.82857084274292,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 6035
+    },
+    {
+      "epoch": 1.1541108986615678,
+      "grad_norm": 1.3361073732376099,
+      "learning_rate": 5e-06,
+      "loss": 0.0476,
+      "step": 6036
+    },
+    {
+      "epoch": 1.154302103250478,
+      "grad_norm": 2.2594399452209473,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 6037
+    },
+    {
+      "epoch": 1.1544933078393882,
+      "grad_norm": 1.668912649154663,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 6038
+    },
+    {
+      "epoch": 1.1546845124282983,
+      "grad_norm": 1.6249210834503174,
+      "learning_rate": 5e-06,
+      "loss": 0.0849,
+      "step": 6039
+    },
+    {
+      "epoch": 1.1548757170172084,
+      "grad_norm": 1.301710605621338,
+      "learning_rate": 5e-06,
+      "loss": 0.081,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1550669216061185,
+      "grad_norm": 0.9434700608253479,
+      "learning_rate": 5e-06,
+      "loss": 0.0392,
+      "step": 6041
+    },
+    {
+      "epoch": 1.1552581261950288,
+      "grad_norm": 0.9116209745407104,
+      "learning_rate": 5e-06,
+      "loss": 0.0235,
+      "step": 6042
+    },
+    {
+      "epoch": 1.1554493307839389,
+      "grad_norm": 0.9934965968132019,
+      "learning_rate": 5e-06,
+      "loss": 0.0644,
+      "step": 6043
+    },
+    {
+      "epoch": 1.155640535372849,
+      "grad_norm": 1.8826981782913208,
+      "learning_rate": 5e-06,
+      "loss": 0.2933,
+      "step": 6044
+    },
+    {
+      "epoch": 1.155831739961759,
+      "grad_norm": 1.0682721138000488,
+      "learning_rate": 5e-06,
+      "loss": 0.0483,
+      "step": 6045
+    },
+    {
+      "epoch": 1.1560229445506693,
+      "grad_norm": 1.3939552307128906,
+      "learning_rate": 5e-06,
+      "loss": 0.0893,
+      "step": 6046
+    },
+    {
+      "epoch": 1.1562141491395794,
+      "grad_norm": 0.7219886183738708,
+      "learning_rate": 5e-06,
+      "loss": 0.0433,
+      "step": 6047
+    },
+    {
+      "epoch": 1.1564053537284895,
+      "grad_norm": 0.8082029819488525,
+      "learning_rate": 5e-06,
+      "loss": 0.0271,
+      "step": 6048
+    },
+    {
+      "epoch": 1.1565965583173996,
+      "grad_norm": 0.8083641529083252,
+      "learning_rate": 5e-06,
+      "loss": 0.0503,
+      "step": 6049
+    },
+    {
+      "epoch": 1.1567877629063097,
+      "grad_norm": 1.0870314836502075,
+      "learning_rate": 5e-06,
+      "loss": 0.0769,
+      "step": 6050
+    },
+    {
+      "epoch": 1.15697896749522,
+      "grad_norm": 0.9009522795677185,
+      "learning_rate": 5e-06,
+      "loss": 0.0477,
+      "step": 6051
+    },
+    {
+      "epoch": 1.15717017208413,
+      "grad_norm": 0.9931294322013855,
+      "learning_rate": 5e-06,
+      "loss": 0.0635,
+      "step": 6052
+    },
+    {
+      "epoch": 1.1573613766730402,
+      "grad_norm": 1.5237797498703003,
+      "learning_rate": 5e-06,
+      "loss": 0.0961,
+      "step": 6053
+    },
+    {
+      "epoch": 1.1575525812619503,
+      "grad_norm": 1.4311531782150269,
+      "learning_rate": 5e-06,
+      "loss": 0.0464,
+      "step": 6054
+    },
+    {
+      "epoch": 1.1577437858508604,
+      "grad_norm": 2.5143849849700928,
+      "learning_rate": 5e-06,
+      "loss": 0.2268,
+      "step": 6055
+    },
+    {
+      "epoch": 1.1579349904397707,
+      "grad_norm": 1.216488003730774,
+      "learning_rate": 5e-06,
+      "loss": 0.0839,
+      "step": 6056
+    },
+    {
+      "epoch": 1.1581261950286807,
+      "grad_norm": 2.2803962230682373,
+      "learning_rate": 5e-06,
+      "loss": 0.3013,
+      "step": 6057
+    },
+    {
+      "epoch": 1.1583173996175908,
+      "grad_norm": 0.7981217503547668,
+      "learning_rate": 5e-06,
+      "loss": 0.0436,
+      "step": 6058
+    },
+    {
+      "epoch": 1.158508604206501,
+      "grad_norm": 1.780248761177063,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 6059
+    },
+    {
+      "epoch": 1.158699808795411,
+      "grad_norm": 5.465449810028076,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 6060
+    },
+    {
+      "epoch": 1.1588910133843213,
+      "grad_norm": 1.3702939748764038,
+      "learning_rate": 5e-06,
+      "loss": 0.0291,
+      "step": 6061
+    },
+    {
+      "epoch": 1.1590822179732314,
+      "grad_norm": 2.034011125564575,
+      "learning_rate": 5e-06,
+      "loss": 0.2144,
+      "step": 6062
+    },
+    {
+      "epoch": 1.1592734225621415,
+      "grad_norm": 1.4905356168746948,
+      "learning_rate": 5e-06,
+      "loss": 0.1838,
+      "step": 6063
+    },
+    {
+      "epoch": 1.1594646271510516,
+      "grad_norm": 2.3215534687042236,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 6064
+    },
+    {
+      "epoch": 1.1596558317399617,
+      "grad_norm": 0.8124056458473206,
+      "learning_rate": 5e-06,
+      "loss": 0.0456,
+      "step": 6065
+    },
+    {
+      "epoch": 1.159847036328872,
+      "grad_norm": 1.8607640266418457,
+      "learning_rate": 5e-06,
+      "loss": 0.131,
+      "step": 6066
+    },
+    {
+      "epoch": 1.160038240917782,
+      "grad_norm": 1.4982339143753052,
+      "learning_rate": 5e-06,
+      "loss": 0.0986,
+      "step": 6067
+    },
+    {
+      "epoch": 1.1602294455066922,
+      "grad_norm": 2.7232859134674072,
+      "learning_rate": 5e-06,
+      "loss": 0.2743,
+      "step": 6068
+    },
+    {
+      "epoch": 1.1604206500956022,
+      "grad_norm": 2.740915536880493,
+      "learning_rate": 5e-06,
+      "loss": 0.4415,
+      "step": 6069
+    },
+    {
+      "epoch": 1.1606118546845123,
+      "grad_norm": 1.1084314584732056,
+      "learning_rate": 5e-06,
+      "loss": 0.0626,
+      "step": 6070
+    },
+    {
+      "epoch": 1.1608030592734226,
+      "grad_norm": 3.3199756145477295,
+      "learning_rate": 5e-06,
+      "loss": 0.2035,
+      "step": 6071
+    },
+    {
+      "epoch": 1.1609942638623327,
+      "grad_norm": 0.5316591262817383,
+      "learning_rate": 5e-06,
+      "loss": 0.0395,
+      "step": 6072
+    },
+    {
+      "epoch": 1.1611854684512428,
+      "grad_norm": 0.8929632306098938,
+      "learning_rate": 5e-06,
+      "loss": 0.0379,
+      "step": 6073
+    },
+    {
+      "epoch": 1.161376673040153,
+      "grad_norm": 1.1094083786010742,
+      "learning_rate": 5e-06,
+      "loss": 0.0883,
+      "step": 6074
+    },
+    {
+      "epoch": 1.161567877629063,
+      "grad_norm": 3.199720621109009,
+      "learning_rate": 5e-06,
+      "loss": 0.4122,
+      "step": 6075
+    },
+    {
+      "epoch": 1.1617590822179733,
+      "grad_norm": 2.4385669231414795,
+      "learning_rate": 5e-06,
+      "loss": 0.3021,
+      "step": 6076
+    },
+    {
+      "epoch": 1.1619502868068834,
+      "grad_norm": 1.3131438493728638,
+      "learning_rate": 5e-06,
+      "loss": 0.0734,
+      "step": 6077
+    },
+    {
+      "epoch": 1.1621414913957935,
+      "grad_norm": 2.2846639156341553,
+      "learning_rate": 5e-06,
+      "loss": 0.162,
+      "step": 6078
+    },
+    {
+      "epoch": 1.1623326959847036,
+      "grad_norm": 1.474809169769287,
+      "learning_rate": 5e-06,
+      "loss": 0.0868,
+      "step": 6079
+    },
+    {
+      "epoch": 1.1625239005736137,
+      "grad_norm": 1.7511301040649414,
+      "learning_rate": 5e-06,
+      "loss": 0.0742,
+      "step": 6080
+    },
+    {
+      "epoch": 1.162715105162524,
+      "grad_norm": 1.1560462713241577,
+      "learning_rate": 5e-06,
+      "loss": 0.1053,
+      "step": 6081
+    },
+    {
+      "epoch": 1.162906309751434,
+      "grad_norm": 1.9263542890548706,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 6082
+    },
+    {
+      "epoch": 1.1630975143403441,
+      "grad_norm": 1.997341275215149,
+      "learning_rate": 5e-06,
+      "loss": 0.0731,
+      "step": 6083
+    },
+    {
+      "epoch": 1.1632887189292542,
+      "grad_norm": 2.5195326805114746,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 6084
+    },
+    {
+      "epoch": 1.1634799235181643,
+      "grad_norm": 1.3920161724090576,
+      "learning_rate": 5e-06,
+      "loss": 0.0689,
+      "step": 6085
+    },
+    {
+      "epoch": 1.1636711281070746,
+      "grad_norm": 1.848002552986145,
+      "learning_rate": 5e-06,
+      "loss": 0.0813,
+      "step": 6086
+    },
+    {
+      "epoch": 1.1638623326959847,
+      "grad_norm": 1.6859074831008911,
+      "learning_rate": 5e-06,
+      "loss": 0.2863,
+      "step": 6087
+    },
+    {
+      "epoch": 1.1640535372848948,
+      "grad_norm": 2.795480251312256,
+      "learning_rate": 5e-06,
+      "loss": 0.3935,
+      "step": 6088
+    },
+    {
+      "epoch": 1.164244741873805,
+      "grad_norm": 2.764321804046631,
+      "learning_rate": 5e-06,
+      "loss": 0.268,
+      "step": 6089
+    },
+    {
+      "epoch": 1.1644359464627152,
+      "grad_norm": 0.9207529425621033,
+      "learning_rate": 5e-06,
+      "loss": 0.0395,
+      "step": 6090
+    },
+    {
+      "epoch": 1.1646271510516253,
+      "grad_norm": 2.6257948875427246,
+      "learning_rate": 5e-06,
+      "loss": 0.2537,
+      "step": 6091
+    },
+    {
+      "epoch": 1.1648183556405354,
+      "grad_norm": 1.9078866243362427,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 6092
+    },
+    {
+      "epoch": 1.1650095602294455,
+      "grad_norm": 6.173279285430908,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 6093
+    },
+    {
+      "epoch": 1.1652007648183555,
+      "grad_norm": 2.563831090927124,
+      "learning_rate": 5e-06,
+      "loss": 0.1948,
+      "step": 6094
+    },
+    {
+      "epoch": 1.1653919694072659,
+      "grad_norm": 1.040425181388855,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 6095
+    },
+    {
+      "epoch": 1.165583173996176,
+      "grad_norm": 0.45935672521591187,
+      "learning_rate": 5e-06,
+      "loss": 0.0246,
+      "step": 6096
+    },
+    {
+      "epoch": 1.165774378585086,
+      "grad_norm": 1.1373692750930786,
+      "learning_rate": 5e-06,
+      "loss": 0.0963,
+      "step": 6097
+    },
+    {
+      "epoch": 1.1659655831739961,
+      "grad_norm": 2.903848886489868,
+      "learning_rate": 5e-06,
+      "loss": 0.1996,
+      "step": 6098
+    },
+    {
+      "epoch": 1.1661567877629064,
+      "grad_norm": 1.5372850894927979,
+      "learning_rate": 5e-06,
+      "loss": 0.075,
+      "step": 6099
+    },
+    {
+      "epoch": 1.1663479923518165,
+      "grad_norm": 2.0959360599517822,
+      "learning_rate": 5e-06,
+      "loss": 0.2309,
+      "step": 6100
+    },
+    {
+      "epoch": 1.1665391969407266,
+      "grad_norm": 3.3226559162139893,
+      "learning_rate": 5e-06,
+      "loss": 0.233,
+      "step": 6101
+    },
+    {
+      "epoch": 1.1667304015296367,
+      "grad_norm": 1.1195683479309082,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 6102
+    },
+    {
+      "epoch": 1.1669216061185468,
+      "grad_norm": 1.387656569480896,
+      "learning_rate": 5e-06,
+      "loss": 0.0717,
+      "step": 6103
+    },
+    {
+      "epoch": 1.167112810707457,
+      "grad_norm": 1.3395695686340332,
+      "learning_rate": 5e-06,
+      "loss": 0.0479,
+      "step": 6104
+    },
+    {
+      "epoch": 1.1673040152963672,
+      "grad_norm": 1.6233667135238647,
+      "learning_rate": 5e-06,
+      "loss": 0.0703,
+      "step": 6105
+    },
+    {
+      "epoch": 1.1674952198852773,
+      "grad_norm": 2.596073627471924,
+      "learning_rate": 5e-06,
+      "loss": 0.1554,
+      "step": 6106
+    },
+    {
+      "epoch": 1.1676864244741874,
+      "grad_norm": 1.9220967292785645,
+      "learning_rate": 5e-06,
+      "loss": 0.2205,
+      "step": 6107
+    },
+    {
+      "epoch": 1.1678776290630974,
+      "grad_norm": 1.6835957765579224,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 6108
+    },
+    {
+      "epoch": 1.1680688336520078,
+      "grad_norm": 1.530069351196289,
+      "learning_rate": 5e-06,
+      "loss": 0.0651,
+      "step": 6109
+    },
+    {
+      "epoch": 1.1682600382409178,
+      "grad_norm": 0.737148106098175,
+      "learning_rate": 5e-06,
+      "loss": 0.0299,
+      "step": 6110
+    },
+    {
+      "epoch": 1.168451242829828,
+      "grad_norm": 1.4814426898956299,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 6111
+    },
+    {
+      "epoch": 1.168642447418738,
+      "grad_norm": 0.973702609539032,
+      "learning_rate": 5e-06,
+      "loss": 0.0567,
+      "step": 6112
+    },
+    {
+      "epoch": 1.168833652007648,
+      "grad_norm": 2.017190933227539,
+      "learning_rate": 5e-06,
+      "loss": 0.2373,
+      "step": 6113
+    },
+    {
+      "epoch": 1.1690248565965584,
+      "grad_norm": 1.150895595550537,
+      "learning_rate": 5e-06,
+      "loss": 0.0863,
+      "step": 6114
+    },
+    {
+      "epoch": 1.1692160611854685,
+      "grad_norm": 2.035149335861206,
+      "learning_rate": 5e-06,
+      "loss": 0.0599,
+      "step": 6115
+    },
+    {
+      "epoch": 1.1694072657743786,
+      "grad_norm": 1.2829310894012451,
+      "learning_rate": 5e-06,
+      "loss": 0.0662,
+      "step": 6116
+    },
+    {
+      "epoch": 1.1695984703632887,
+      "grad_norm": 0.9252585172653198,
+      "learning_rate": 5e-06,
+      "loss": 0.0334,
+      "step": 6117
+    },
+    {
+      "epoch": 1.1697896749521988,
+      "grad_norm": 2.3354899883270264,
+      "learning_rate": 5e-06,
+      "loss": 0.1792,
+      "step": 6118
+    },
+    {
+      "epoch": 1.169980879541109,
+      "grad_norm": 1.619714617729187,
+      "learning_rate": 5e-06,
+      "loss": 0.2189,
+      "step": 6119
+    },
+    {
+      "epoch": 1.1701720841300192,
+      "grad_norm": 1.9781533479690552,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 6120
+    },
+    {
+      "epoch": 1.1703632887189293,
+      "grad_norm": 1.8020018339157104,
+      "learning_rate": 5e-06,
+      "loss": 0.2103,
+      "step": 6121
+    },
+    {
+      "epoch": 1.1705544933078393,
+      "grad_norm": 1.7455902099609375,
+      "learning_rate": 5e-06,
+      "loss": 0.1837,
+      "step": 6122
+    },
+    {
+      "epoch": 1.1707456978967494,
+      "grad_norm": 1.8223216533660889,
+      "learning_rate": 5e-06,
+      "loss": 0.0997,
+      "step": 6123
+    },
+    {
+      "epoch": 1.1709369024856597,
+      "grad_norm": 1.6536144018173218,
+      "learning_rate": 5e-06,
+      "loss": 0.0922,
+      "step": 6124
+    },
+    {
+      "epoch": 1.1711281070745698,
+      "grad_norm": 2.5655362606048584,
+      "learning_rate": 5e-06,
+      "loss": 0.3305,
+      "step": 6125
+    },
+    {
+      "epoch": 1.17131931166348,
+      "grad_norm": 1.325499415397644,
+      "learning_rate": 5e-06,
+      "loss": 0.105,
+      "step": 6126
+    },
+    {
+      "epoch": 1.17151051625239,
+      "grad_norm": 1.0477885007858276,
+      "learning_rate": 5e-06,
+      "loss": 0.082,
+      "step": 6127
+    },
+    {
+      "epoch": 1.1717017208413,
+      "grad_norm": 1.0732998847961426,
+      "learning_rate": 5e-06,
+      "loss": 0.0464,
+      "step": 6128
+    },
+    {
+      "epoch": 1.1718929254302104,
+      "grad_norm": 1.3447692394256592,
+      "learning_rate": 5e-06,
+      "loss": 0.0626,
+      "step": 6129
+    },
+    {
+      "epoch": 1.1720841300191205,
+      "grad_norm": 2.3556089401245117,
+      "learning_rate": 5e-06,
+      "loss": 0.0672,
+      "step": 6130
+    },
+    {
+      "epoch": 1.1722753346080306,
+      "grad_norm": 2.144721031188965,
+      "learning_rate": 5e-06,
+      "loss": 0.2276,
+      "step": 6131
+    },
+    {
+      "epoch": 1.1724665391969407,
+      "grad_norm": 1.2007111310958862,
+      "learning_rate": 5e-06,
+      "loss": 0.0624,
+      "step": 6132
+    },
+    {
+      "epoch": 1.1726577437858507,
+      "grad_norm": 1.4135570526123047,
+      "learning_rate": 5e-06,
+      "loss": 0.082,
+      "step": 6133
+    },
+    {
+      "epoch": 1.172848948374761,
+      "grad_norm": 1.207506537437439,
+      "learning_rate": 5e-06,
+      "loss": 0.0531,
+      "step": 6134
+    },
+    {
+      "epoch": 1.1730401529636711,
+      "grad_norm": 0.7473751306533813,
+      "learning_rate": 5e-06,
+      "loss": 0.0336,
+      "step": 6135
+    },
+    {
+      "epoch": 1.1732313575525812,
+      "grad_norm": 1.5870224237442017,
+      "learning_rate": 5e-06,
+      "loss": 0.0666,
+      "step": 6136
+    },
+    {
+      "epoch": 1.1734225621414913,
+      "grad_norm": 2.2782132625579834,
+      "learning_rate": 5e-06,
+      "loss": 0.2869,
+      "step": 6137
+    },
+    {
+      "epoch": 1.1736137667304014,
+      "grad_norm": 2.2009999752044678,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 6138
+    },
+    {
+      "epoch": 1.1738049713193117,
+      "grad_norm": 2.3219454288482666,
+      "learning_rate": 5e-06,
+      "loss": 0.2503,
+      "step": 6139
+    },
+    {
+      "epoch": 1.1739961759082218,
+      "grad_norm": 1.4381366968154907,
+      "learning_rate": 5e-06,
+      "loss": 0.0768,
+      "step": 6140
+    },
+    {
+      "epoch": 1.174187380497132,
+      "grad_norm": 2.0978338718414307,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 6141
+    },
+    {
+      "epoch": 1.174378585086042,
+      "grad_norm": 2.1573426723480225,
+      "learning_rate": 5e-06,
+      "loss": 0.083,
+      "step": 6142
+    },
+    {
+      "epoch": 1.1745697896749523,
+      "grad_norm": 1.2375134229660034,
+      "learning_rate": 5e-06,
+      "loss": 0.0635,
+      "step": 6143
+    },
+    {
+      "epoch": 1.1747609942638624,
+      "grad_norm": 1.8515455722808838,
+      "learning_rate": 5e-06,
+      "loss": 0.2427,
+      "step": 6144
+    },
+    {
+      "epoch": 1.1749521988527725,
+      "grad_norm": 1.246639609336853,
+      "learning_rate": 5e-06,
+      "loss": 0.0717,
+      "step": 6145
+    },
+    {
+      "epoch": 1.1751434034416826,
+      "grad_norm": 1.7270658016204834,
+      "learning_rate": 5e-06,
+      "loss": 0.0917,
+      "step": 6146
+    },
+    {
+      "epoch": 1.1753346080305926,
+      "grad_norm": 1.3657946586608887,
+      "learning_rate": 5e-06,
+      "loss": 0.1012,
+      "step": 6147
+    },
+    {
+      "epoch": 1.175525812619503,
+      "grad_norm": 1.2972500324249268,
+      "learning_rate": 5e-06,
+      "loss": 0.0726,
+      "step": 6148
+    },
+    {
+      "epoch": 1.175717017208413,
+      "grad_norm": 2.3606135845184326,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 6149
+    },
+    {
+      "epoch": 1.1759082217973231,
+      "grad_norm": 1.981109619140625,
+      "learning_rate": 5e-06,
+      "loss": 0.2327,
+      "step": 6150
+    },
+    {
+      "epoch": 1.1760994263862332,
+      "grad_norm": 2.298509120941162,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 6151
+    },
+    {
+      "epoch": 1.1762906309751435,
+      "grad_norm": 2.005854606628418,
+      "learning_rate": 5e-06,
+      "loss": 0.1567,
+      "step": 6152
+    },
+    {
+      "epoch": 1.1764818355640536,
+      "grad_norm": 0.9500100612640381,
+      "learning_rate": 5e-06,
+      "loss": 0.0708,
+      "step": 6153
+    },
+    {
+      "epoch": 1.1766730401529637,
+      "grad_norm": 1.3746157884597778,
+      "learning_rate": 5e-06,
+      "loss": 0.0448,
+      "step": 6154
+    },
+    {
+      "epoch": 1.1768642447418738,
+      "grad_norm": 0.5043768882751465,
+      "learning_rate": 5e-06,
+      "loss": 0.0102,
+      "step": 6155
+    },
+    {
+      "epoch": 1.1770554493307839,
+      "grad_norm": 1.2846918106079102,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 6156
+    },
+    {
+      "epoch": 1.1772466539196942,
+      "grad_norm": 2.024791955947876,
+      "learning_rate": 5e-06,
+      "loss": 0.2269,
+      "step": 6157
+    },
+    {
+      "epoch": 1.1774378585086043,
+      "grad_norm": 0.7490910887718201,
+      "learning_rate": 5e-06,
+      "loss": 0.0589,
+      "step": 6158
+    },
+    {
+      "epoch": 1.1776290630975144,
+      "grad_norm": 1.5239231586456299,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 6159
+    },
+    {
+      "epoch": 1.1778202676864244,
+      "grad_norm": 1.2455819845199585,
+      "learning_rate": 5e-06,
+      "loss": 0.0616,
+      "step": 6160
+    },
+    {
+      "epoch": 1.1780114722753345,
+      "grad_norm": 1.4771140813827515,
+      "learning_rate": 5e-06,
+      "loss": 0.0756,
+      "step": 6161
+    },
+    {
+      "epoch": 1.1782026768642448,
+      "grad_norm": 2.541170120239258,
+      "learning_rate": 5e-06,
+      "loss": 0.1628,
+      "step": 6162
+    },
+    {
+      "epoch": 1.178393881453155,
+      "grad_norm": 2.1921584606170654,
+      "learning_rate": 5e-06,
+      "loss": 0.2259,
+      "step": 6163
+    },
+    {
+      "epoch": 1.178585086042065,
+      "grad_norm": 1.606255292892456,
+      "learning_rate": 5e-06,
+      "loss": 0.0649,
+      "step": 6164
+    },
+    {
+      "epoch": 1.178776290630975,
+      "grad_norm": 3.125506639480591,
+      "learning_rate": 5e-06,
+      "loss": 0.2574,
+      "step": 6165
+    },
+    {
+      "epoch": 1.1789674952198852,
+      "grad_norm": 1.2741800546646118,
+      "learning_rate": 5e-06,
+      "loss": 0.0475,
+      "step": 6166
+    },
+    {
+      "epoch": 1.1791586998087955,
+      "grad_norm": 2.3493738174438477,
+      "learning_rate": 5e-06,
+      "loss": 0.059,
+      "step": 6167
+    },
+    {
+      "epoch": 1.1793499043977056,
+      "grad_norm": 1.9525309801101685,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 6168
+    },
+    {
+      "epoch": 1.1795411089866157,
+      "grad_norm": 2.7668604850769043,
+      "learning_rate": 5e-06,
+      "loss": 0.309,
+      "step": 6169
+    },
+    {
+      "epoch": 1.1797323135755258,
+      "grad_norm": 2.444951057434082,
+      "learning_rate": 5e-06,
+      "loss": 0.2414,
+      "step": 6170
+    },
+    {
+      "epoch": 1.1799235181644359,
+      "grad_norm": 1.980841040611267,
+      "learning_rate": 5e-06,
+      "loss": 0.0647,
+      "step": 6171
+    },
+    {
+      "epoch": 1.1801147227533462,
+      "grad_norm": 1.0069383382797241,
+      "learning_rate": 5e-06,
+      "loss": 0.0172,
+      "step": 6172
+    },
+    {
+      "epoch": 1.1803059273422563,
+      "grad_norm": 1.7199982404708862,
+      "learning_rate": 5e-06,
+      "loss": 0.1625,
+      "step": 6173
+    },
+    {
+      "epoch": 1.1804971319311663,
+      "grad_norm": 1.0968600511550903,
+      "learning_rate": 5e-06,
+      "loss": 0.0659,
+      "step": 6174
+    },
+    {
+      "epoch": 1.1806883365200764,
+      "grad_norm": 2.4999165534973145,
+      "learning_rate": 5e-06,
+      "loss": 0.3111,
+      "step": 6175
+    },
+    {
+      "epoch": 1.1808795411089865,
+      "grad_norm": 1.1034135818481445,
+      "learning_rate": 5e-06,
+      "loss": 0.0808,
+      "step": 6176
+    },
+    {
+      "epoch": 1.1810707456978968,
+      "grad_norm": 2.0255119800567627,
+      "learning_rate": 5e-06,
+      "loss": 0.1643,
+      "step": 6177
+    },
+    {
+      "epoch": 1.181261950286807,
+      "grad_norm": 1.7352789640426636,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 6178
+    },
+    {
+      "epoch": 1.181453154875717,
+      "grad_norm": 0.796521782875061,
+      "learning_rate": 5e-06,
+      "loss": 0.0369,
+      "step": 6179
+    },
+    {
+      "epoch": 1.181644359464627,
+      "grad_norm": 2.0245773792266846,
+      "learning_rate": 5e-06,
+      "loss": 0.0833,
+      "step": 6180
+    },
+    {
+      "epoch": 1.1818355640535372,
+      "grad_norm": 1.772589921951294,
+      "learning_rate": 5e-06,
+      "loss": 0.2376,
+      "step": 6181
+    },
+    {
+      "epoch": 1.1820267686424475,
+      "grad_norm": 2.6077778339385986,
+      "learning_rate": 5e-06,
+      "loss": 0.3178,
+      "step": 6182
+    },
+    {
+      "epoch": 1.1822179732313576,
+      "grad_norm": 1.4697093963623047,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 6183
+    },
+    {
+      "epoch": 1.1824091778202677,
+      "grad_norm": 1.1189794540405273,
+      "learning_rate": 5e-06,
+      "loss": 0.0712,
+      "step": 6184
+    },
+    {
+      "epoch": 1.1826003824091778,
+      "grad_norm": 1.7616618871688843,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 6185
+    },
+    {
+      "epoch": 1.1827915869980878,
+      "grad_norm": 1.4431734085083008,
+      "learning_rate": 5e-06,
+      "loss": 0.0509,
+      "step": 6186
+    },
+    {
+      "epoch": 1.1829827915869982,
+      "grad_norm": 1.4921287298202515,
+      "learning_rate": 5e-06,
+      "loss": 0.0924,
+      "step": 6187
+    },
+    {
+      "epoch": 1.1831739961759082,
+      "grad_norm": 3.1014909744262695,
+      "learning_rate": 5e-06,
+      "loss": 0.3719,
+      "step": 6188
+    },
+    {
+      "epoch": 1.1833652007648183,
+      "grad_norm": 1.158164620399475,
+      "learning_rate": 5e-06,
+      "loss": 0.0296,
+      "step": 6189
+    },
+    {
+      "epoch": 1.1835564053537284,
+      "grad_norm": 1.8710097074508667,
+      "learning_rate": 5e-06,
+      "loss": 0.1908,
+      "step": 6190
+    },
+    {
+      "epoch": 1.1837476099426385,
+      "grad_norm": 1.8682113885879517,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 6191
+    },
+    {
+      "epoch": 1.1839388145315488,
+      "grad_norm": 0.9372738003730774,
+      "learning_rate": 5e-06,
+      "loss": 0.0528,
+      "step": 6192
+    },
+    {
+      "epoch": 1.184130019120459,
+      "grad_norm": 1.8785148859024048,
+      "learning_rate": 5e-06,
+      "loss": 0.3054,
+      "step": 6193
+    },
+    {
+      "epoch": 1.184321223709369,
+      "grad_norm": 1.2328784465789795,
+      "learning_rate": 5e-06,
+      "loss": 0.082,
+      "step": 6194
+    },
+    {
+      "epoch": 1.184512428298279,
+      "grad_norm": 1.1626081466674805,
+      "learning_rate": 5e-06,
+      "loss": 0.0991,
+      "step": 6195
+    },
+    {
+      "epoch": 1.1847036328871894,
+      "grad_norm": 2.7644124031066895,
+      "learning_rate": 5e-06,
+      "loss": 0.2075,
+      "step": 6196
+    },
+    {
+      "epoch": 1.1848948374760995,
+      "grad_norm": 1.14145827293396,
+      "learning_rate": 5e-06,
+      "loss": 0.0526,
+      "step": 6197
+    },
+    {
+      "epoch": 1.1850860420650096,
+      "grad_norm": 1.1169764995574951,
+      "learning_rate": 5e-06,
+      "loss": 0.0239,
+      "step": 6198
+    },
+    {
+      "epoch": 1.1852772466539196,
+      "grad_norm": 2.4677770137786865,
+      "learning_rate": 5e-06,
+      "loss": 0.0712,
+      "step": 6199
+    },
+    {
+      "epoch": 1.1854684512428297,
+      "grad_norm": 2.896726369857788,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 6200
+    },
+    {
+      "epoch": 1.18565965583174,
+      "grad_norm": 1.0080317258834839,
+      "learning_rate": 5e-06,
+      "loss": 0.0729,
+      "step": 6201
+    },
+    {
+      "epoch": 1.1858508604206501,
+      "grad_norm": 1.406341791152954,
+      "learning_rate": 5e-06,
+      "loss": 0.1634,
+      "step": 6202
+    },
+    {
+      "epoch": 1.1860420650095602,
+      "grad_norm": 2.233196258544922,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 6203
+    },
+    {
+      "epoch": 1.1862332695984703,
+      "grad_norm": 0.7236731052398682,
+      "learning_rate": 5e-06,
+      "loss": 0.0436,
+      "step": 6204
+    },
+    {
+      "epoch": 1.1864244741873806,
+      "grad_norm": 1.104647159576416,
+      "learning_rate": 5e-06,
+      "loss": 0.0803,
+      "step": 6205
+    },
+    {
+      "epoch": 1.1866156787762907,
+      "grad_norm": 1.6566107273101807,
+      "learning_rate": 5e-06,
+      "loss": 0.1847,
+      "step": 6206
+    },
+    {
+      "epoch": 1.1868068833652008,
+      "grad_norm": 2.225419282913208,
+      "learning_rate": 5e-06,
+      "loss": 0.1847,
+      "step": 6207
+    },
+    {
+      "epoch": 1.1869980879541109,
+      "grad_norm": 1.3101707696914673,
+      "learning_rate": 5e-06,
+      "loss": 0.0773,
+      "step": 6208
+    },
+    {
+      "epoch": 1.187189292543021,
+      "grad_norm": 1.1030720472335815,
+      "learning_rate": 5e-06,
+      "loss": 0.0684,
+      "step": 6209
+    },
+    {
+      "epoch": 1.1873804971319313,
+      "grad_norm": 1.0501564741134644,
+      "learning_rate": 5e-06,
+      "loss": 0.0591,
+      "step": 6210
+    },
+    {
+      "epoch": 1.1875717017208414,
+      "grad_norm": 1.0577927827835083,
+      "learning_rate": 5e-06,
+      "loss": 0.0496,
+      "step": 6211
+    },
+    {
+      "epoch": 1.1877629063097515,
+      "grad_norm": 2.4910309314727783,
+      "learning_rate": 5e-06,
+      "loss": 0.3094,
+      "step": 6212
+    },
+    {
+      "epoch": 1.1879541108986615,
+      "grad_norm": 1.5930724143981934,
+      "learning_rate": 5e-06,
+      "loss": 0.1574,
+      "step": 6213
+    },
+    {
+      "epoch": 1.1881453154875716,
+      "grad_norm": 1.2082481384277344,
+      "learning_rate": 5e-06,
+      "loss": 0.0612,
+      "step": 6214
+    },
+    {
+      "epoch": 1.188336520076482,
+      "grad_norm": 1.5550799369812012,
+      "learning_rate": 5e-06,
+      "loss": 0.0683,
+      "step": 6215
+    },
+    {
+      "epoch": 1.188527724665392,
+      "grad_norm": 1.548312783241272,
+      "learning_rate": 5e-06,
+      "loss": 0.0548,
+      "step": 6216
+    },
+    {
+      "epoch": 1.1887189292543021,
+      "grad_norm": 1.306344747543335,
+      "learning_rate": 5e-06,
+      "loss": 0.0559,
+      "step": 6217
+    },
+    {
+      "epoch": 1.1889101338432122,
+      "grad_norm": 1.2293124198913574,
+      "learning_rate": 5e-06,
+      "loss": 0.065,
+      "step": 6218
+    },
+    {
+      "epoch": 1.1891013384321223,
+      "grad_norm": 3.0053391456604004,
+      "learning_rate": 5e-06,
+      "loss": 0.4573,
+      "step": 6219
+    },
+    {
+      "epoch": 1.1892925430210326,
+      "grad_norm": 2.9051673412323,
+      "learning_rate": 5e-06,
+      "loss": 0.0862,
+      "step": 6220
+    },
+    {
+      "epoch": 1.1894837476099427,
+      "grad_norm": 1.1124510765075684,
+      "learning_rate": 5e-06,
+      "loss": 0.0373,
+      "step": 6221
+    },
+    {
+      "epoch": 1.1896749521988528,
+      "grad_norm": 6.297783374786377,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 6222
+    },
+    {
+      "epoch": 1.1898661567877629,
+      "grad_norm": 3.238295316696167,
+      "learning_rate": 5e-06,
+      "loss": 0.1066,
+      "step": 6223
+    },
+    {
+      "epoch": 1.190057361376673,
+      "grad_norm": 1.506626844406128,
+      "learning_rate": 5e-06,
+      "loss": 0.0639,
+      "step": 6224
+    },
+    {
+      "epoch": 1.1902485659655833,
+      "grad_norm": 1.8110731840133667,
+      "learning_rate": 5e-06,
+      "loss": 0.1739,
+      "step": 6225
+    },
+    {
+      "epoch": 1.1904397705544933,
+      "grad_norm": 1.8189138174057007,
+      "learning_rate": 5e-06,
+      "loss": 0.2069,
+      "step": 6226
+    },
+    {
+      "epoch": 1.1906309751434034,
+      "grad_norm": 1.4828901290893555,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 6227
+    },
+    {
+      "epoch": 1.1908221797323135,
+      "grad_norm": 1.3421870470046997,
+      "learning_rate": 5e-06,
+      "loss": 0.0718,
+      "step": 6228
+    },
+    {
+      "epoch": 1.1910133843212236,
+      "grad_norm": 1.5782335996627808,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 6229
+    },
+    {
+      "epoch": 1.191204588910134,
+      "grad_norm": 0.5779447555541992,
+      "learning_rate": 5e-06,
+      "loss": 0.0196,
+      "step": 6230
+    },
+    {
+      "epoch": 1.191395793499044,
+      "grad_norm": 3.198089599609375,
+      "learning_rate": 5e-06,
+      "loss": 0.4032,
+      "step": 6231
+    },
+    {
+      "epoch": 1.191586998087954,
+      "grad_norm": 2.1662838459014893,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 6232
+    },
+    {
+      "epoch": 1.1917782026768642,
+      "grad_norm": 0.6701384782791138,
+      "learning_rate": 5e-06,
+      "loss": 0.0796,
+      "step": 6233
+    },
+    {
+      "epoch": 1.1919694072657743,
+      "grad_norm": 3.48876953125,
+      "learning_rate": 5e-06,
+      "loss": 0.3998,
+      "step": 6234
+    },
+    {
+      "epoch": 1.1921606118546846,
+      "grad_norm": 1.7155407667160034,
+      "learning_rate": 5e-06,
+      "loss": 0.0601,
+      "step": 6235
+    },
+    {
+      "epoch": 1.1923518164435947,
+      "grad_norm": 1.6817251443862915,
+      "learning_rate": 5e-06,
+      "loss": 0.0674,
+      "step": 6236
+    },
+    {
+      "epoch": 1.1925430210325048,
+      "grad_norm": 1.244370460510254,
+      "learning_rate": 5e-06,
+      "loss": 0.0725,
+      "step": 6237
+    },
+    {
+      "epoch": 1.1927342256214148,
+      "grad_norm": 0.7383753657341003,
+      "learning_rate": 5e-06,
+      "loss": 0.0387,
+      "step": 6238
+    },
+    {
+      "epoch": 1.192925430210325,
+      "grad_norm": 3.243826150894165,
+      "learning_rate": 5e-06,
+      "loss": 0.2708,
+      "step": 6239
+    },
+    {
+      "epoch": 1.1931166347992352,
+      "grad_norm": 1.0814484357833862,
+      "learning_rate": 5e-06,
+      "loss": 0.0429,
+      "step": 6240
+    },
+    {
+      "epoch": 1.1933078393881453,
+      "grad_norm": 1.800237774848938,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 6241
+    },
+    {
+      "epoch": 1.1934990439770554,
+      "grad_norm": 1.2890852689743042,
+      "learning_rate": 5e-06,
+      "loss": 0.0736,
+      "step": 6242
+    },
+    {
+      "epoch": 1.1936902485659655,
+      "grad_norm": 1.504570722579956,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 6243
+    },
+    {
+      "epoch": 1.1938814531548756,
+      "grad_norm": 3.0813817977905273,
+      "learning_rate": 5e-06,
+      "loss": 0.2378,
+      "step": 6244
+    },
+    {
+      "epoch": 1.194072657743786,
+      "grad_norm": 2.5581462383270264,
+      "learning_rate": 5e-06,
+      "loss": 0.2248,
+      "step": 6245
+    },
+    {
+      "epoch": 1.194263862332696,
+      "grad_norm": 1.9704996347427368,
+      "learning_rate": 5e-06,
+      "loss": 0.1308,
+      "step": 6246
+    },
+    {
+      "epoch": 1.194455066921606,
+      "grad_norm": 1.2662582397460938,
+      "learning_rate": 5e-06,
+      "loss": 0.0757,
+      "step": 6247
+    },
+    {
+      "epoch": 1.1946462715105162,
+      "grad_norm": 1.3485127687454224,
+      "learning_rate": 5e-06,
+      "loss": 0.0613,
+      "step": 6248
+    },
+    {
+      "epoch": 1.1948374760994265,
+      "grad_norm": 1.4152460098266602,
+      "learning_rate": 5e-06,
+      "loss": 0.0603,
+      "step": 6249
+    },
+    {
+      "epoch": 1.1950286806883366,
+      "grad_norm": 2.333681583404541,
+      "learning_rate": 5e-06,
+      "loss": 0.1836,
+      "step": 6250
+    },
+    {
+      "epoch": 1.1952198852772467,
+      "grad_norm": 1.0232439041137695,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 6251
+    },
+    {
+      "epoch": 1.1954110898661567,
+      "grad_norm": 1.8188303709030151,
+      "learning_rate": 5e-06,
+      "loss": 0.088,
+      "step": 6252
+    },
+    {
+      "epoch": 1.1956022944550668,
+      "grad_norm": 2.2055118083953857,
+      "learning_rate": 5e-06,
+      "loss": 0.1996,
+      "step": 6253
+    },
+    {
+      "epoch": 1.1957934990439771,
+      "grad_norm": 1.0319563150405884,
+      "learning_rate": 5e-06,
+      "loss": 0.0685,
+      "step": 6254
+    },
+    {
+      "epoch": 1.1959847036328872,
+      "grad_norm": 1.0295190811157227,
+      "learning_rate": 5e-06,
+      "loss": 0.0491,
+      "step": 6255
+    },
+    {
+      "epoch": 1.1961759082217973,
+      "grad_norm": 1.7119534015655518,
+      "learning_rate": 5e-06,
+      "loss": 0.1958,
+      "step": 6256
+    },
+    {
+      "epoch": 1.1963671128107074,
+      "grad_norm": 2.6020867824554443,
+      "learning_rate": 5e-06,
+      "loss": 0.3227,
+      "step": 6257
+    },
+    {
+      "epoch": 1.1965583173996177,
+      "grad_norm": 0.8010663390159607,
+      "learning_rate": 5e-06,
+      "loss": 0.0648,
+      "step": 6258
+    },
+    {
+      "epoch": 1.1967495219885278,
+      "grad_norm": 1.836167335510254,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 6259
+    },
+    {
+      "epoch": 1.1969407265774379,
+      "grad_norm": 2.1391921043395996,
+      "learning_rate": 5e-06,
+      "loss": 0.2105,
+      "step": 6260
+    },
+    {
+      "epoch": 1.197131931166348,
+      "grad_norm": 1.2611051797866821,
+      "learning_rate": 5e-06,
+      "loss": 0.0711,
+      "step": 6261
+    },
+    {
+      "epoch": 1.197323135755258,
+      "grad_norm": 1.799573540687561,
+      "learning_rate": 5e-06,
+      "loss": 0.1808,
+      "step": 6262
+    },
+    {
+      "epoch": 1.1975143403441684,
+      "grad_norm": 1.2242969274520874,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 6263
+    },
+    {
+      "epoch": 1.1977055449330785,
+      "grad_norm": 1.190840721130371,
+      "learning_rate": 5e-06,
+      "loss": 0.0509,
+      "step": 6264
+    },
+    {
+      "epoch": 1.1978967495219885,
+      "grad_norm": 1.2675566673278809,
+      "learning_rate": 5e-06,
+      "loss": 0.0601,
+      "step": 6265
+    },
+    {
+      "epoch": 1.1980879541108986,
+      "grad_norm": 1.62624990940094,
+      "learning_rate": 5e-06,
+      "loss": 0.0855,
+      "step": 6266
+    },
+    {
+      "epoch": 1.1982791586998087,
+      "grad_norm": 0.9653342366218567,
+      "learning_rate": 5e-06,
+      "loss": 0.0298,
+      "step": 6267
+    },
+    {
+      "epoch": 1.198470363288719,
+      "grad_norm": 2.015363931655884,
+      "learning_rate": 5e-06,
+      "loss": 0.1699,
+      "step": 6268
+    },
+    {
+      "epoch": 1.1986615678776291,
+      "grad_norm": 3.233055591583252,
+      "learning_rate": 5e-06,
+      "loss": 0.3099,
+      "step": 6269
+    },
+    {
+      "epoch": 1.1988527724665392,
+      "grad_norm": 3.2268948554992676,
+      "learning_rate": 5e-06,
+      "loss": 0.4068,
+      "step": 6270
+    },
+    {
+      "epoch": 1.1990439770554493,
+      "grad_norm": 2.4663195610046387,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 6271
+    },
+    {
+      "epoch": 1.1992351816443594,
+      "grad_norm": 1.4187419414520264,
+      "learning_rate": 5e-06,
+      "loss": 0.0471,
+      "step": 6272
+    },
+    {
+      "epoch": 1.1994263862332697,
+      "grad_norm": 0.7417177557945251,
+      "learning_rate": 5e-06,
+      "loss": 0.0204,
+      "step": 6273
+    },
+    {
+      "epoch": 1.1996175908221798,
+      "grad_norm": 1.6655139923095703,
+      "learning_rate": 5e-06,
+      "loss": 0.0895,
+      "step": 6274
+    },
+    {
+      "epoch": 1.1998087954110899,
+      "grad_norm": 2.4011824131011963,
+      "learning_rate": 5e-06,
+      "loss": 0.2601,
+      "step": 6275
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6423674821853638,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 6276
+    },
+    {
+      "epoch": 1.20019120458891,
+      "grad_norm": 1.1955245733261108,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 6277
+    },
+    {
+      "epoch": 1.2003824091778204,
+      "grad_norm": 2.667212724685669,
+      "learning_rate": 5e-06,
+      "loss": 0.2089,
+      "step": 6278
+    },
+    {
+      "epoch": 1.2005736137667304,
+      "grad_norm": 1.371573567390442,
+      "learning_rate": 5e-06,
+      "loss": 0.0467,
+      "step": 6279
+    },
+    {
+      "epoch": 1.2007648183556405,
+      "grad_norm": 0.9084519147872925,
+      "learning_rate": 5e-06,
+      "loss": 0.0358,
+      "step": 6280
+    },
+    {
+      "epoch": 1.2009560229445506,
+      "grad_norm": 2.8557240962982178,
+      "learning_rate": 5e-06,
+      "loss": 0.3219,
+      "step": 6281
+    },
+    {
+      "epoch": 1.2011472275334607,
+      "grad_norm": 3.1929121017456055,
+      "learning_rate": 5e-06,
+      "loss": 0.369,
+      "step": 6282
+    },
+    {
+      "epoch": 1.201338432122371,
+      "grad_norm": 1.8422592878341675,
+      "learning_rate": 5e-06,
+      "loss": 0.1406,
+      "step": 6283
+    },
+    {
+      "epoch": 1.201529636711281,
+      "grad_norm": 1.644729733467102,
+      "learning_rate": 5e-06,
+      "loss": 0.0833,
+      "step": 6284
+    },
+    {
+      "epoch": 1.2017208413001912,
+      "grad_norm": 1.2755012512207031,
+      "learning_rate": 5e-06,
+      "loss": 0.0553,
+      "step": 6285
+    },
+    {
+      "epoch": 1.2019120458891013,
+      "grad_norm": 0.9884834885597229,
+      "learning_rate": 5e-06,
+      "loss": 0.034,
+      "step": 6286
+    },
+    {
+      "epoch": 1.2021032504780114,
+      "grad_norm": 1.7062479257583618,
+      "learning_rate": 5e-06,
+      "loss": 0.1823,
+      "step": 6287
+    },
+    {
+      "epoch": 1.2022944550669217,
+      "grad_norm": 1.1571954488754272,
+      "learning_rate": 5e-06,
+      "loss": 0.0542,
+      "step": 6288
+    },
+    {
+      "epoch": 1.2024856596558318,
+      "grad_norm": 2.0687966346740723,
+      "learning_rate": 5e-06,
+      "loss": 0.2359,
+      "step": 6289
+    },
+    {
+      "epoch": 1.2026768642447419,
+      "grad_norm": 1.4236472845077515,
+      "learning_rate": 5e-06,
+      "loss": 0.0533,
+      "step": 6290
+    },
+    {
+      "epoch": 1.202868068833652,
+      "grad_norm": 1.979692816734314,
+      "learning_rate": 5e-06,
+      "loss": 0.0762,
+      "step": 6291
+    },
+    {
+      "epoch": 1.203059273422562,
+      "grad_norm": 0.6412269473075867,
+      "learning_rate": 5e-06,
+      "loss": 0.0188,
+      "step": 6292
+    },
+    {
+      "epoch": 1.2032504780114723,
+      "grad_norm": 2.0749592781066895,
+      "learning_rate": 5e-06,
+      "loss": 0.1959,
+      "step": 6293
+    },
+    {
+      "epoch": 1.2034416826003824,
+      "grad_norm": 1.9312831163406372,
+      "learning_rate": 5e-06,
+      "loss": 0.1836,
+      "step": 6294
+    },
+    {
+      "epoch": 1.2036328871892925,
+      "grad_norm": 1.3121813535690308,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 6295
+    },
+    {
+      "epoch": 1.2038240917782026,
+      "grad_norm": 1.8166366815567017,
+      "learning_rate": 5e-06,
+      "loss": 0.1763,
+      "step": 6296
+    },
+    {
+      "epoch": 1.2040152963671127,
+      "grad_norm": 2.4472837448120117,
+      "learning_rate": 5e-06,
+      "loss": 0.2342,
+      "step": 6297
+    },
+    {
+      "epoch": 1.204206500956023,
+      "grad_norm": 1.2451283931732178,
+      "learning_rate": 5e-06,
+      "loss": 0.0518,
+      "step": 6298
+    },
+    {
+      "epoch": 1.204397705544933,
+      "grad_norm": 1.7306233644485474,
+      "learning_rate": 5e-06,
+      "loss": 0.0941,
+      "step": 6299
+    },
+    {
+      "epoch": 1.2045889101338432,
+      "grad_norm": 1.8587627410888672,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 6300
+    },
+    {
+      "epoch": 1.2047801147227533,
+      "grad_norm": 1.9993771314620972,
+      "learning_rate": 5e-06,
+      "loss": 0.2328,
+      "step": 6301
+    },
+    {
+      "epoch": 1.2049713193116636,
+      "grad_norm": 1.5691461563110352,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 6302
+    },
+    {
+      "epoch": 1.2051625239005737,
+      "grad_norm": 1.0496647357940674,
+      "learning_rate": 5e-06,
+      "loss": 0.0498,
+      "step": 6303
+    },
+    {
+      "epoch": 1.2053537284894837,
+      "grad_norm": 6.206069469451904,
+      "learning_rate": 5e-06,
+      "loss": 0.0789,
+      "step": 6304
+    },
+    {
+      "epoch": 1.2055449330783938,
+      "grad_norm": 1.0517714023590088,
+      "learning_rate": 5e-06,
+      "loss": 0.0533,
+      "step": 6305
+    },
+    {
+      "epoch": 1.205736137667304,
+      "grad_norm": 1.230154275894165,
+      "learning_rate": 5e-06,
+      "loss": 0.0728,
+      "step": 6306
+    },
+    {
+      "epoch": 1.2059273422562142,
+      "grad_norm": 1.2463685274124146,
+      "learning_rate": 5e-06,
+      "loss": 0.0973,
+      "step": 6307
+    },
+    {
+      "epoch": 1.2061185468451243,
+      "grad_norm": 2.0279324054718018,
+      "learning_rate": 5e-06,
+      "loss": 0.1416,
+      "step": 6308
+    },
+    {
+      "epoch": 1.2063097514340344,
+      "grad_norm": 1.0040415525436401,
+      "learning_rate": 5e-06,
+      "loss": 0.0787,
+      "step": 6309
+    },
+    {
+      "epoch": 1.2065009560229445,
+      "grad_norm": 0.8246625065803528,
+      "learning_rate": 5e-06,
+      "loss": 0.0398,
+      "step": 6310
+    },
+    {
+      "epoch": 1.2066921606118548,
+      "grad_norm": 1.2535613775253296,
+      "learning_rate": 5e-06,
+      "loss": 0.0478,
+      "step": 6311
+    },
+    {
+      "epoch": 1.206883365200765,
+      "grad_norm": 1.6264280080795288,
+      "learning_rate": 5e-06,
+      "loss": 0.0806,
+      "step": 6312
+    },
+    {
+      "epoch": 1.207074569789675,
+      "grad_norm": 2.006242275238037,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 6313
+    },
+    {
+      "epoch": 1.207265774378585,
+      "grad_norm": 1.1853926181793213,
+      "learning_rate": 5e-06,
+      "loss": 0.0983,
+      "step": 6314
+    },
+    {
+      "epoch": 1.2074569789674952,
+      "grad_norm": 1.7839405536651611,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 6315
+    },
+    {
+      "epoch": 1.2076481835564055,
+      "grad_norm": 1.2711913585662842,
+      "learning_rate": 5e-06,
+      "loss": 0.0652,
+      "step": 6316
+    },
+    {
+      "epoch": 1.2078393881453156,
+      "grad_norm": 3.0699965953826904,
+      "learning_rate": 5e-06,
+      "loss": 0.0587,
+      "step": 6317
+    },
+    {
+      "epoch": 1.2080305927342256,
+      "grad_norm": 1.580519199371338,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 6318
+    },
+    {
+      "epoch": 1.2082217973231357,
+      "grad_norm": 1.9091516733169556,
+      "learning_rate": 5e-06,
+      "loss": 0.2536,
+      "step": 6319
+    },
+    {
+      "epoch": 1.2084130019120458,
+      "grad_norm": 1.5716497898101807,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 6320
+    },
+    {
+      "epoch": 1.2086042065009561,
+      "grad_norm": 1.1605068445205688,
+      "learning_rate": 5e-06,
+      "loss": 0.0474,
+      "step": 6321
+    },
+    {
+      "epoch": 1.2087954110898662,
+      "grad_norm": 1.2060127258300781,
+      "learning_rate": 5e-06,
+      "loss": 0.0683,
+      "step": 6322
+    },
+    {
+      "epoch": 1.2089866156787763,
+      "grad_norm": 1.9277080297470093,
+      "learning_rate": 5e-06,
+      "loss": 0.196,
+      "step": 6323
+    },
+    {
+      "epoch": 1.2091778202676864,
+      "grad_norm": 2.57090163230896,
+      "learning_rate": 5e-06,
+      "loss": 0.0813,
+      "step": 6324
+    },
+    {
+      "epoch": 1.2093690248565965,
+      "grad_norm": 2.1226699352264404,
+      "learning_rate": 5e-06,
+      "loss": 0.249,
+      "step": 6325
+    },
+    {
+      "epoch": 1.2095602294455068,
+      "grad_norm": 1.5362803936004639,
+      "learning_rate": 5e-06,
+      "loss": 0.183,
+      "step": 6326
+    },
+    {
+      "epoch": 1.2097514340344169,
+      "grad_norm": 2.6527233123779297,
+      "learning_rate": 5e-06,
+      "loss": 0.2657,
+      "step": 6327
+    },
+    {
+      "epoch": 1.209942638623327,
+      "grad_norm": 1.9042359590530396,
+      "learning_rate": 5e-06,
+      "loss": 0.1758,
+      "step": 6328
+    },
+    {
+      "epoch": 1.210133843212237,
+      "grad_norm": 0.6970901489257812,
+      "learning_rate": 5e-06,
+      "loss": 0.0262,
+      "step": 6329
+    },
+    {
+      "epoch": 1.2103250478011471,
+      "grad_norm": 1.6701407432556152,
+      "learning_rate": 5e-06,
+      "loss": 0.0508,
+      "step": 6330
+    },
+    {
+      "epoch": 1.2105162523900574,
+      "grad_norm": 2.219998836517334,
+      "learning_rate": 5e-06,
+      "loss": 0.376,
+      "step": 6331
+    },
+    {
+      "epoch": 1.2107074569789675,
+      "grad_norm": 1.266753077507019,
+      "learning_rate": 5e-06,
+      "loss": 0.078,
+      "step": 6332
+    },
+    {
+      "epoch": 1.2108986615678776,
+      "grad_norm": 1.639849066734314,
+      "learning_rate": 5e-06,
+      "loss": 0.1245,
+      "step": 6333
+    },
+    {
+      "epoch": 1.2110898661567877,
+      "grad_norm": 1.036638855934143,
+      "learning_rate": 5e-06,
+      "loss": 0.043,
+      "step": 6334
+    },
+    {
+      "epoch": 1.2112810707456978,
+      "grad_norm": 1.1263424158096313,
+      "learning_rate": 5e-06,
+      "loss": 0.0321,
+      "step": 6335
+    },
+    {
+      "epoch": 1.211472275334608,
+      "grad_norm": 1.8749959468841553,
+      "learning_rate": 5e-06,
+      "loss": 0.0971,
+      "step": 6336
+    },
+    {
+      "epoch": 1.2116634799235182,
+      "grad_norm": 0.9465709924697876,
+      "learning_rate": 5e-06,
+      "loss": 0.0425,
+      "step": 6337
+    },
+    {
+      "epoch": 1.2118546845124283,
+      "grad_norm": 2.4710988998413086,
+      "learning_rate": 5e-06,
+      "loss": 0.274,
+      "step": 6338
+    },
+    {
+      "epoch": 1.2120458891013384,
+      "grad_norm": 1.7354429960250854,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 6339
+    },
+    {
+      "epoch": 1.2122370936902485,
+      "grad_norm": 1.3433597087860107,
+      "learning_rate": 5e-06,
+      "loss": 0.0579,
+      "step": 6340
+    },
+    {
+      "epoch": 1.2124282982791588,
+      "grad_norm": 1.5866605043411255,
+      "learning_rate": 5e-06,
+      "loss": 0.0936,
+      "step": 6341
+    },
+    {
+      "epoch": 1.2126195028680689,
+      "grad_norm": 2.422170400619507,
+      "learning_rate": 5e-06,
+      "loss": 0.0584,
+      "step": 6342
+    },
+    {
+      "epoch": 1.212810707456979,
+      "grad_norm": 2.6649551391601562,
+      "learning_rate": 5e-06,
+      "loss": 0.1525,
+      "step": 6343
+    },
+    {
+      "epoch": 1.213001912045889,
+      "grad_norm": 1.7258577346801758,
+      "learning_rate": 5e-06,
+      "loss": 0.1976,
+      "step": 6344
+    },
+    {
+      "epoch": 1.2131931166347991,
+      "grad_norm": 2.841470241546631,
+      "learning_rate": 5e-06,
+      "loss": 0.3982,
+      "step": 6345
+    },
+    {
+      "epoch": 1.2133843212237094,
+      "grad_norm": 1.485338807106018,
+      "learning_rate": 5e-06,
+      "loss": 0.044,
+      "step": 6346
+    },
+    {
+      "epoch": 1.2135755258126195,
+      "grad_norm": 1.383081316947937,
+      "learning_rate": 5e-06,
+      "loss": 0.2076,
+      "step": 6347
+    },
+    {
+      "epoch": 1.2137667304015296,
+      "grad_norm": 0.8841148018836975,
+      "learning_rate": 5e-06,
+      "loss": 0.0334,
+      "step": 6348
+    },
+    {
+      "epoch": 1.2139579349904397,
+      "grad_norm": 1.4105879068374634,
+      "learning_rate": 5e-06,
+      "loss": 0.043,
+      "step": 6349
+    },
+    {
+      "epoch": 1.21414913957935,
+      "grad_norm": 2.799672842025757,
+      "learning_rate": 5e-06,
+      "loss": 0.2811,
+      "step": 6350
+    },
+    {
+      "epoch": 1.21434034416826,
+      "grad_norm": 1.106540322303772,
+      "learning_rate": 5e-06,
+      "loss": 0.0508,
+      "step": 6351
+    },
+    {
+      "epoch": 1.2145315487571702,
+      "grad_norm": 1.1798187494277954,
+      "learning_rate": 5e-06,
+      "loss": 0.071,
+      "step": 6352
+    },
+    {
+      "epoch": 1.2147227533460803,
+      "grad_norm": 1.4219733476638794,
+      "learning_rate": 5e-06,
+      "loss": 0.0609,
+      "step": 6353
+    },
+    {
+      "epoch": 1.2149139579349904,
+      "grad_norm": 0.49215996265411377,
+      "learning_rate": 5e-06,
+      "loss": 0.013,
+      "step": 6354
+    },
+    {
+      "epoch": 1.2151051625239007,
+      "grad_norm": 1.4540727138519287,
+      "learning_rate": 5e-06,
+      "loss": 0.066,
+      "step": 6355
+    },
+    {
+      "epoch": 1.2152963671128107,
+      "grad_norm": 2.5020601749420166,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 6356
+    },
+    {
+      "epoch": 1.2154875717017208,
+      "grad_norm": 1.5029916763305664,
+      "learning_rate": 5e-06,
+      "loss": 0.1123,
+      "step": 6357
+    },
+    {
+      "epoch": 1.215678776290631,
+      "grad_norm": 1.260717749595642,
+      "learning_rate": 5e-06,
+      "loss": 0.0678,
+      "step": 6358
+    },
+    {
+      "epoch": 1.215869980879541,
+      "grad_norm": 1.8185780048370361,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 6359
+    },
+    {
+      "epoch": 1.2160611854684513,
+      "grad_norm": 0.7864851951599121,
+      "learning_rate": 5e-06,
+      "loss": 0.0319,
+      "step": 6360
+    },
+    {
+      "epoch": 1.2162523900573614,
+      "grad_norm": 1.0918238162994385,
+      "learning_rate": 5e-06,
+      "loss": 0.0447,
+      "step": 6361
+    },
+    {
+      "epoch": 1.2164435946462715,
+      "grad_norm": 2.916642427444458,
+      "learning_rate": 5e-06,
+      "loss": 0.2256,
+      "step": 6362
+    },
+    {
+      "epoch": 1.2166347992351816,
+      "grad_norm": 2.603426218032837,
+      "learning_rate": 5e-06,
+      "loss": 0.1929,
+      "step": 6363
+    },
+    {
+      "epoch": 1.216826003824092,
+      "grad_norm": 1.888663411140442,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 6364
+    },
+    {
+      "epoch": 1.217017208413002,
+      "grad_norm": 5.557328701019287,
+      "learning_rate": 5e-06,
+      "loss": 0.1513,
+      "step": 6365
+    },
+    {
+      "epoch": 1.217208413001912,
+      "grad_norm": 0.6387971043586731,
+      "learning_rate": 5e-06,
+      "loss": 0.0625,
+      "step": 6366
+    },
+    {
+      "epoch": 1.2173996175908222,
+      "grad_norm": 0.80973881483078,
+      "learning_rate": 5e-06,
+      "loss": 0.0612,
+      "step": 6367
+    },
+    {
+      "epoch": 1.2175908221797322,
+      "grad_norm": 1.4961767196655273,
+      "learning_rate": 5e-06,
+      "loss": 0.0794,
+      "step": 6368
+    },
+    {
+      "epoch": 1.2177820267686426,
+      "grad_norm": 1.3402931690216064,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 6369
+    },
+    {
+      "epoch": 1.2179732313575526,
+      "grad_norm": 1.615916132926941,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 6370
+    },
+    {
+      "epoch": 1.2181644359464627,
+      "grad_norm": 0.9041286110877991,
+      "learning_rate": 5e-06,
+      "loss": 0.0461,
+      "step": 6371
+    },
+    {
+      "epoch": 1.2183556405353728,
+      "grad_norm": 1.5138685703277588,
+      "learning_rate": 5e-06,
+      "loss": 0.0752,
+      "step": 6372
+    },
+    {
+      "epoch": 1.218546845124283,
+      "grad_norm": 2.4668264389038086,
+      "learning_rate": 5e-06,
+      "loss": 0.1996,
+      "step": 6373
+    },
+    {
+      "epoch": 1.2187380497131932,
+      "grad_norm": 1.754960060119629,
+      "learning_rate": 5e-06,
+      "loss": 0.0677,
+      "step": 6374
+    },
+    {
+      "epoch": 1.2189292543021033,
+      "grad_norm": 1.3340129852294922,
+      "learning_rate": 5e-06,
+      "loss": 0.0818,
+      "step": 6375
+    },
+    {
+      "epoch": 1.2191204588910134,
+      "grad_norm": 2.1007907390594482,
+      "learning_rate": 5e-06,
+      "loss": 0.1689,
+      "step": 6376
+    },
+    {
+      "epoch": 1.2193116634799235,
+      "grad_norm": 2.982419729232788,
+      "learning_rate": 5e-06,
+      "loss": 0.217,
+      "step": 6377
+    },
+    {
+      "epoch": 1.2195028680688336,
+      "grad_norm": 0.9394291043281555,
+      "learning_rate": 5e-06,
+      "loss": 0.079,
+      "step": 6378
+    },
+    {
+      "epoch": 1.2196940726577439,
+      "grad_norm": 0.9420931935310364,
+      "learning_rate": 5e-06,
+      "loss": 0.0518,
+      "step": 6379
+    },
+    {
+      "epoch": 1.219885277246654,
+      "grad_norm": 1.3374733924865723,
+      "learning_rate": 5e-06,
+      "loss": 0.0267,
+      "step": 6380
+    },
+    {
+      "epoch": 1.220076481835564,
+      "grad_norm": 3.1424825191497803,
+      "learning_rate": 5e-06,
+      "loss": 0.2596,
+      "step": 6381
+    },
+    {
+      "epoch": 1.2202676864244741,
+      "grad_norm": 1.0249494314193726,
+      "learning_rate": 5e-06,
+      "loss": 0.0696,
+      "step": 6382
+    },
+    {
+      "epoch": 1.2204588910133842,
+      "grad_norm": 1.8867278099060059,
+      "learning_rate": 5e-06,
+      "loss": 0.0802,
+      "step": 6383
+    },
+    {
+      "epoch": 1.2206500956022945,
+      "grad_norm": 1.559537649154663,
+      "learning_rate": 5e-06,
+      "loss": 0.0624,
+      "step": 6384
+    },
+    {
+      "epoch": 1.2208413001912046,
+      "grad_norm": 0.673812747001648,
+      "learning_rate": 5e-06,
+      "loss": 0.0599,
+      "step": 6385
+    },
+    {
+      "epoch": 1.2210325047801147,
+      "grad_norm": 0.6307937502861023,
+      "learning_rate": 5e-06,
+      "loss": 0.0229,
+      "step": 6386
+    },
+    {
+      "epoch": 1.2212237093690248,
+      "grad_norm": 1.3093122243881226,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 6387
+    },
+    {
+      "epoch": 1.221414913957935,
+      "grad_norm": 2.759263277053833,
+      "learning_rate": 5e-06,
+      "loss": 0.4477,
+      "step": 6388
+    },
+    {
+      "epoch": 1.2216061185468452,
+      "grad_norm": 2.4608709812164307,
+      "learning_rate": 5e-06,
+      "loss": 0.313,
+      "step": 6389
+    },
+    {
+      "epoch": 1.2217973231357553,
+      "grad_norm": 1.8315072059631348,
+      "learning_rate": 5e-06,
+      "loss": 0.089,
+      "step": 6390
+    },
+    {
+      "epoch": 1.2219885277246654,
+      "grad_norm": 0.7214317321777344,
+      "learning_rate": 5e-06,
+      "loss": 0.0515,
+      "step": 6391
+    },
+    {
+      "epoch": 1.2221797323135755,
+      "grad_norm": 0.8006981611251831,
+      "learning_rate": 5e-06,
+      "loss": 0.0421,
+      "step": 6392
+    },
+    {
+      "epoch": 1.2223709369024855,
+      "grad_norm": 1.7554163932800293,
+      "learning_rate": 5e-06,
+      "loss": 0.147,
+      "step": 6393
+    },
+    {
+      "epoch": 1.2225621414913959,
+      "grad_norm": 2.516019344329834,
+      "learning_rate": 5e-06,
+      "loss": 0.3259,
+      "step": 6394
+    },
+    {
+      "epoch": 1.222753346080306,
+      "grad_norm": 1.6270321607589722,
+      "learning_rate": 5e-06,
+      "loss": 0.0723,
+      "step": 6395
+    },
+    {
+      "epoch": 1.222944550669216,
+      "grad_norm": 0.9091405272483826,
+      "learning_rate": 5e-06,
+      "loss": 0.1058,
+      "step": 6396
+    },
+    {
+      "epoch": 1.2231357552581261,
+      "grad_norm": 0.7934021353721619,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 6397
+    },
+    {
+      "epoch": 1.2233269598470362,
+      "grad_norm": 0.9105468988418579,
+      "learning_rate": 5e-06,
+      "loss": 0.0267,
+      "step": 6398
+    },
+    {
+      "epoch": 1.2235181644359465,
+      "grad_norm": 1.7011659145355225,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 6399
+    },
+    {
+      "epoch": 1.2237093690248566,
+      "grad_norm": 1.6316510438919067,
+      "learning_rate": 5e-06,
+      "loss": 0.1647,
+      "step": 6400
+    },
+    {
+      "epoch": 1.2239005736137667,
+      "grad_norm": 1.9868152141571045,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 6401
+    },
+    {
+      "epoch": 1.2240917782026768,
+      "grad_norm": 1.912008285522461,
+      "learning_rate": 5e-06,
+      "loss": 0.1873,
+      "step": 6402
+    },
+    {
+      "epoch": 1.224282982791587,
+      "grad_norm": 1.5919640064239502,
+      "learning_rate": 5e-06,
+      "loss": 0.1538,
+      "step": 6403
+    },
+    {
+      "epoch": 1.2244741873804972,
+      "grad_norm": 1.775577425956726,
+      "learning_rate": 5e-06,
+      "loss": 0.0358,
+      "step": 6404
+    },
+    {
+      "epoch": 1.2246653919694073,
+      "grad_norm": 0.8517948985099792,
+      "learning_rate": 5e-06,
+      "loss": 0.0253,
+      "step": 6405
+    },
+    {
+      "epoch": 1.2248565965583174,
+      "grad_norm": 2.3561553955078125,
+      "learning_rate": 5e-06,
+      "loss": 0.2985,
+      "step": 6406
+    },
+    {
+      "epoch": 1.2250478011472274,
+      "grad_norm": 1.6562501192092896,
+      "learning_rate": 5e-06,
+      "loss": 0.2061,
+      "step": 6407
+    },
+    {
+      "epoch": 1.2252390057361378,
+      "grad_norm": 1.6132574081420898,
+      "learning_rate": 5e-06,
+      "loss": 0.1679,
+      "step": 6408
+    },
+    {
+      "epoch": 1.2254302103250478,
+      "grad_norm": 1.9934953451156616,
+      "learning_rate": 5e-06,
+      "loss": 0.2289,
+      "step": 6409
+    },
+    {
+      "epoch": 1.225621414913958,
+      "grad_norm": 1.7691314220428467,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 6410
+    },
+    {
+      "epoch": 1.225812619502868,
+      "grad_norm": 1.1167519092559814,
+      "learning_rate": 5e-06,
+      "loss": 0.0377,
+      "step": 6411
+    },
+    {
+      "epoch": 1.2260038240917783,
+      "grad_norm": 1.710638165473938,
+      "learning_rate": 5e-06,
+      "loss": 0.2259,
+      "step": 6412
+    },
+    {
+      "epoch": 1.2261950286806884,
+      "grad_norm": 2.857348680496216,
+      "learning_rate": 5e-06,
+      "loss": 0.442,
+      "step": 6413
+    },
+    {
+      "epoch": 1.2263862332695985,
+      "grad_norm": 2.343129873275757,
+      "learning_rate": 5e-06,
+      "loss": 0.2763,
+      "step": 6414
+    },
+    {
+      "epoch": 1.2265774378585086,
+      "grad_norm": 0.8242167830467224,
+      "learning_rate": 5e-06,
+      "loss": 0.0545,
+      "step": 6415
+    },
+    {
+      "epoch": 1.2267686424474187,
+      "grad_norm": 1.5538809299468994,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 6416
+    },
+    {
+      "epoch": 1.226959847036329,
+      "grad_norm": 1.174586534500122,
+      "learning_rate": 5e-06,
+      "loss": 0.0375,
+      "step": 6417
+    },
+    {
+      "epoch": 1.227151051625239,
+      "grad_norm": 2.826963186264038,
+      "learning_rate": 5e-06,
+      "loss": 0.3052,
+      "step": 6418
+    },
+    {
+      "epoch": 1.2273422562141492,
+      "grad_norm": 2.5396435260772705,
+      "learning_rate": 5e-06,
+      "loss": 0.2724,
+      "step": 6419
+    },
+    {
+      "epoch": 1.2275334608030593,
+      "grad_norm": 2.1552212238311768,
+      "learning_rate": 5e-06,
+      "loss": 0.2733,
+      "step": 6420
+    },
+    {
+      "epoch": 1.2277246653919693,
+      "grad_norm": 2.812382936477661,
+      "learning_rate": 5e-06,
+      "loss": 0.2159,
+      "step": 6421
+    },
+    {
+      "epoch": 1.2279158699808796,
+      "grad_norm": 1.1729477643966675,
+      "learning_rate": 5e-06,
+      "loss": 0.1066,
+      "step": 6422
+    },
+    {
+      "epoch": 1.2281070745697897,
+      "grad_norm": 1.4861737489700317,
+      "learning_rate": 5e-06,
+      "loss": 0.1204,
+      "step": 6423
+    },
+    {
+      "epoch": 1.2282982791586998,
+      "grad_norm": 0.9737563729286194,
+      "learning_rate": 5e-06,
+      "loss": 0.0359,
+      "step": 6424
+    },
+    {
+      "epoch": 1.22848948374761,
+      "grad_norm": 2.244095802307129,
+      "learning_rate": 5e-06,
+      "loss": 0.3383,
+      "step": 6425
+    },
+    {
+      "epoch": 1.22868068833652,
+      "grad_norm": 1.4030261039733887,
+      "learning_rate": 5e-06,
+      "loss": 0.0753,
+      "step": 6426
+    },
+    {
+      "epoch": 1.2288718929254303,
+      "grad_norm": 1.439172625541687,
+      "learning_rate": 5e-06,
+      "loss": 0.0642,
+      "step": 6427
+    },
+    {
+      "epoch": 1.2290630975143404,
+      "grad_norm": 0.6380961537361145,
+      "learning_rate": 5e-06,
+      "loss": 0.0633,
+      "step": 6428
+    },
+    {
+      "epoch": 1.2292543021032505,
+      "grad_norm": 0.9303730726242065,
+      "learning_rate": 5e-06,
+      "loss": 0.0676,
+      "step": 6429
+    },
+    {
+      "epoch": 1.2294455066921606,
+      "grad_norm": 1.2828363180160522,
+      "learning_rate": 5e-06,
+      "loss": 0.0468,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2296367112810707,
+      "grad_norm": 2.6689090728759766,
+      "learning_rate": 5e-06,
+      "loss": 0.3648,
+      "step": 6431
+    },
+    {
+      "epoch": 1.229827915869981,
+      "grad_norm": 2.0346622467041016,
+      "learning_rate": 5e-06,
+      "loss": 0.2261,
+      "step": 6432
+    },
+    {
+      "epoch": 1.230019120458891,
+      "grad_norm": 1.939487338066101,
+      "learning_rate": 5e-06,
+      "loss": 0.193,
+      "step": 6433
+    },
+    {
+      "epoch": 1.2302103250478011,
+      "grad_norm": 1.3966894149780273,
+      "learning_rate": 5e-06,
+      "loss": 0.0372,
+      "step": 6434
+    },
+    {
+      "epoch": 1.2304015296367112,
+      "grad_norm": 1.1000654697418213,
+      "learning_rate": 5e-06,
+      "loss": 0.0814,
+      "step": 6435
+    },
+    {
+      "epoch": 1.2305927342256213,
+      "grad_norm": 4.047434329986572,
+      "learning_rate": 5e-06,
+      "loss": 0.1489,
+      "step": 6436
+    },
+    {
+      "epoch": 1.2307839388145316,
+      "grad_norm": 1.382005214691162,
+      "learning_rate": 5e-06,
+      "loss": 0.0809,
+      "step": 6437
+    },
+    {
+      "epoch": 1.2309751434034417,
+      "grad_norm": 1.00590181350708,
+      "learning_rate": 5e-06,
+      "loss": 0.0726,
+      "step": 6438
+    },
+    {
+      "epoch": 1.2311663479923518,
+      "grad_norm": 1.4503659009933472,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 6439
+    },
+    {
+      "epoch": 1.231357552581262,
+      "grad_norm": 1.0048933029174805,
+      "learning_rate": 5e-06,
+      "loss": 0.0531,
+      "step": 6440
+    },
+    {
+      "epoch": 1.231548757170172,
+      "grad_norm": 1.0850677490234375,
+      "learning_rate": 5e-06,
+      "loss": 0.0485,
+      "step": 6441
+    },
+    {
+      "epoch": 1.2317399617590823,
+      "grad_norm": 2.6655921936035156,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 6442
+    },
+    {
+      "epoch": 1.2319311663479924,
+      "grad_norm": 0.8888630867004395,
+      "learning_rate": 5e-06,
+      "loss": 0.0555,
+      "step": 6443
+    },
+    {
+      "epoch": 1.2321223709369025,
+      "grad_norm": 2.1980793476104736,
+      "learning_rate": 5e-06,
+      "loss": 0.3322,
+      "step": 6444
+    },
+    {
+      "epoch": 1.2323135755258126,
+      "grad_norm": 1.1009891033172607,
+      "learning_rate": 5e-06,
+      "loss": 0.0593,
+      "step": 6445
+    },
+    {
+      "epoch": 1.2325047801147226,
+      "grad_norm": 2.154589891433716,
+      "learning_rate": 5e-06,
+      "loss": 0.1893,
+      "step": 6446
+    },
+    {
+      "epoch": 1.232695984703633,
+      "grad_norm": 2.7047064304351807,
+      "learning_rate": 5e-06,
+      "loss": 0.178,
+      "step": 6447
+    },
+    {
+      "epoch": 1.232887189292543,
+      "grad_norm": 1.372484803199768,
+      "learning_rate": 5e-06,
+      "loss": 0.0953,
+      "step": 6448
+    },
+    {
+      "epoch": 1.2330783938814531,
+      "grad_norm": 1.1343607902526855,
+      "learning_rate": 5e-06,
+      "loss": 0.0352,
+      "step": 6449
+    },
+    {
+      "epoch": 1.2332695984703632,
+      "grad_norm": 2.3495070934295654,
+      "learning_rate": 5e-06,
+      "loss": 0.2928,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2334608030592733,
+      "grad_norm": 1.3010920286178589,
+      "learning_rate": 5e-06,
+      "loss": 0.0774,
+      "step": 6451
+    },
+    {
+      "epoch": 1.2336520076481836,
+      "grad_norm": 1.11711585521698,
+      "learning_rate": 5e-06,
+      "loss": 0.0818,
+      "step": 6452
+    },
+    {
+      "epoch": 1.2338432122370937,
+      "grad_norm": 2.2713661193847656,
+      "learning_rate": 5e-06,
+      "loss": 0.1201,
+      "step": 6453
+    },
+    {
+      "epoch": 1.2340344168260038,
+      "grad_norm": 1.1702678203582764,
+      "learning_rate": 5e-06,
+      "loss": 0.0606,
+      "step": 6454
+    },
+    {
+      "epoch": 1.2342256214149139,
+      "grad_norm": 1.4318171739578247,
+      "learning_rate": 5e-06,
+      "loss": 0.0504,
+      "step": 6455
+    },
+    {
+      "epoch": 1.2344168260038242,
+      "grad_norm": 1.7486087083816528,
+      "learning_rate": 5e-06,
+      "loss": 0.209,
+      "step": 6456
+    },
+    {
+      "epoch": 1.2346080305927343,
+      "grad_norm": 2.538674831390381,
+      "learning_rate": 5e-06,
+      "loss": 0.2618,
+      "step": 6457
+    },
+    {
+      "epoch": 1.2347992351816444,
+      "grad_norm": 1.6931205987930298,
+      "learning_rate": 5e-06,
+      "loss": 0.1153,
+      "step": 6458
+    },
+    {
+      "epoch": 1.2349904397705544,
+      "grad_norm": 2.5916154384613037,
+      "learning_rate": 5e-06,
+      "loss": 0.2899,
+      "step": 6459
+    },
+    {
+      "epoch": 1.2351816443594645,
+      "grad_norm": 0.5500595569610596,
+      "learning_rate": 5e-06,
+      "loss": 0.0244,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2353728489483748,
+      "grad_norm": 1.0323785543441772,
+      "learning_rate": 5e-06,
+      "loss": 0.0267,
+      "step": 6461
+    },
+    {
+      "epoch": 1.235564053537285,
+      "grad_norm": 2.1304259300231934,
+      "learning_rate": 5e-06,
+      "loss": 0.2838,
+      "step": 6462
+    },
+    {
+      "epoch": 1.235755258126195,
+      "grad_norm": 1.0863720178604126,
+      "learning_rate": 5e-06,
+      "loss": 0.0669,
+      "step": 6463
+    },
+    {
+      "epoch": 1.235946462715105,
+      "grad_norm": 1.7608684301376343,
+      "learning_rate": 5e-06,
+      "loss": 0.1987,
+      "step": 6464
+    },
+    {
+      "epoch": 1.2361376673040154,
+      "grad_norm": 2.235177516937256,
+      "learning_rate": 5e-06,
+      "loss": 0.2417,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2363288718929255,
+      "grad_norm": 0.9923659563064575,
+      "learning_rate": 5e-06,
+      "loss": 0.0576,
+      "step": 6466
+    },
+    {
+      "epoch": 1.2365200764818356,
+      "grad_norm": 0.47206011414527893,
+      "learning_rate": 5e-06,
+      "loss": 0.0123,
+      "step": 6467
+    },
+    {
+      "epoch": 1.2367112810707457,
+      "grad_norm": 1.7471652030944824,
+      "learning_rate": 5e-06,
+      "loss": 0.1048,
+      "step": 6468
+    },
+    {
+      "epoch": 1.2369024856596558,
+      "grad_norm": 2.4238178730010986,
+      "learning_rate": 5e-06,
+      "loss": 0.3712,
+      "step": 6469
+    },
+    {
+      "epoch": 1.237093690248566,
+      "grad_norm": 1.817383050918579,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2372848948374762,
+      "grad_norm": 0.883569598197937,
+      "learning_rate": 5e-06,
+      "loss": 0.0748,
+      "step": 6471
+    },
+    {
+      "epoch": 1.2374760994263863,
+      "grad_norm": 1.4984726905822754,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 6472
+    },
+    {
+      "epoch": 1.2376673040152963,
+      "grad_norm": 1.6612300872802734,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 6473
+    },
+    {
+      "epoch": 1.2378585086042064,
+      "grad_norm": 1.7457292079925537,
+      "learning_rate": 5e-06,
+      "loss": 0.0619,
+      "step": 6474
+    },
+    {
+      "epoch": 1.2380497131931167,
+      "grad_norm": 2.305072784423828,
+      "learning_rate": 5e-06,
+      "loss": 0.2619,
+      "step": 6475
+    },
+    {
+      "epoch": 1.2382409177820268,
+      "grad_norm": 0.9911337494850159,
+      "learning_rate": 5e-06,
+      "loss": 0.0859,
+      "step": 6476
+    },
+    {
+      "epoch": 1.238432122370937,
+      "grad_norm": 1.6552451848983765,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 6477
+    },
+    {
+      "epoch": 1.238623326959847,
+      "grad_norm": 2.003192901611328,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 6478
+    },
+    {
+      "epoch": 1.238814531548757,
+      "grad_norm": 1.888157844543457,
+      "learning_rate": 5e-06,
+      "loss": 0.0987,
+      "step": 6479
+    },
+    {
+      "epoch": 1.2390057361376674,
+      "grad_norm": 1.2396515607833862,
+      "learning_rate": 5e-06,
+      "loss": 0.0398,
+      "step": 6480
+    },
+    {
+      "epoch": 1.2391969407265775,
+      "grad_norm": 0.947607159614563,
+      "learning_rate": 5e-06,
+      "loss": 0.0804,
+      "step": 6481
+    },
+    {
+      "epoch": 1.2393881453154876,
+      "grad_norm": 2.7749950885772705,
+      "learning_rate": 5e-06,
+      "loss": 0.1755,
+      "step": 6482
+    },
+    {
+      "epoch": 1.2395793499043977,
+      "grad_norm": 2.7286288738250732,
+      "learning_rate": 5e-06,
+      "loss": 0.3329,
+      "step": 6483
+    },
+    {
+      "epoch": 1.2397705544933078,
+      "grad_norm": 2.670870065689087,
+      "learning_rate": 5e-06,
+      "loss": 0.0679,
+      "step": 6484
+    },
+    {
+      "epoch": 1.239961759082218,
+      "grad_norm": 0.9692270159721375,
+      "learning_rate": 5e-06,
+      "loss": 0.0543,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2401529636711282,
+      "grad_norm": 1.525709867477417,
+      "learning_rate": 5e-06,
+      "loss": 0.0621,
+      "step": 6486
+    },
+    {
+      "epoch": 1.2403441682600382,
+      "grad_norm": 2.301687479019165,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 6487
+    },
+    {
+      "epoch": 1.2405353728489483,
+      "grad_norm": 1.8523778915405273,
+      "learning_rate": 5e-06,
+      "loss": 0.1042,
+      "step": 6488
+    },
+    {
+      "epoch": 1.2407265774378584,
+      "grad_norm": 1.0919690132141113,
+      "learning_rate": 5e-06,
+      "loss": 0.0533,
+      "step": 6489
+    },
+    {
+      "epoch": 1.2409177820267687,
+      "grad_norm": 0.7931455373764038,
+      "learning_rate": 5e-06,
+      "loss": 0.0574,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2411089866156788,
+      "grad_norm": 0.8481590747833252,
+      "learning_rate": 5e-06,
+      "loss": 0.0492,
+      "step": 6491
+    },
+    {
+      "epoch": 1.241300191204589,
+      "grad_norm": 1.4083970785140991,
+      "learning_rate": 5e-06,
+      "loss": 0.0413,
+      "step": 6492
+    },
+    {
+      "epoch": 1.241491395793499,
+      "grad_norm": 0.9851438999176025,
+      "learning_rate": 5e-06,
+      "loss": 0.0501,
+      "step": 6493
+    },
+    {
+      "epoch": 1.241682600382409,
+      "grad_norm": 2.721176862716675,
+      "learning_rate": 5e-06,
+      "loss": 0.4178,
+      "step": 6494
+    },
+    {
+      "epoch": 1.2418738049713194,
+      "grad_norm": 1.2428650856018066,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 6495
+    },
+    {
+      "epoch": 1.2420650095602295,
+      "grad_norm": 3.327238082885742,
+      "learning_rate": 5e-06,
+      "loss": 0.3819,
+      "step": 6496
+    },
+    {
+      "epoch": 1.2422562141491396,
+      "grad_norm": 1.2107603549957275,
+      "learning_rate": 5e-06,
+      "loss": 0.0878,
+      "step": 6497
+    },
+    {
+      "epoch": 1.2424474187380496,
+      "grad_norm": 1.4254592657089233,
+      "learning_rate": 5e-06,
+      "loss": 0.0726,
+      "step": 6498
+    },
+    {
+      "epoch": 1.2426386233269597,
+      "grad_norm": 0.8053144216537476,
+      "learning_rate": 5e-06,
+      "loss": 0.0288,
+      "step": 6499
+    },
+    {
+      "epoch": 1.24282982791587,
+      "grad_norm": 2.3608970642089844,
+      "learning_rate": 5e-06,
+      "loss": 0.2719,
+      "step": 6500
+    },
+    {
+      "epoch": 1.24282982791587,
+      "eval_runtime": 801.4109,
+      "eval_samples_per_second": 1.914,
+      "eval_steps_per_second": 0.24,
+      "step": 6500
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 26150,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.592210007442588e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}