diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,21500 +1,1810 @@
 {
-  "best_metric": 0.9729424715042114,
-  "best_model_checkpoint": "runs/deepseek_lora_20240421-183352/checkpoint-25000",
-  "epoch": 0.694577089374707,
+  "best_metric": 0.986103355884552,
+  "best_model_checkpoint": "runs/deepseek_lora_20240422-095359/checkpoint-2500",
+  "epoch": 0.05788142411455891,
   "eval_steps": 500,
-  "global_step": 30000,
+  "global_step": 2500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.0,
-      "grad_norm": 8.893424987792969,
+      "grad_norm": 11.34415054321289,
       "learning_rate": 4.0000000000000003e-07,
-      "loss": 2.1665,
+      "loss": 1.2669,
       "step": 10
     },
     {
       "epoch": 0.0,
-      "grad_norm": 7.90573787689209,
+      "grad_norm": 7.290144443511963,
       "learning_rate": 8.000000000000001e-07,
-      "loss": 2.2972,
+      "loss": 1.3878,
       "step": 20
     },
     {
       "epoch": 0.0,
-      "grad_norm": 4.421755313873291,
+      "grad_norm": 5.349151134490967,
       "learning_rate": 1.2000000000000002e-06,
-      "loss": 2.0517,
+      "loss": 1.1855,
       "step": 30
     },
     {
       "epoch": 0.0,
-      "grad_norm": 4.259636402130127,
+      "grad_norm": 14.503132820129395,
       "learning_rate": 1.6000000000000001e-06,
-      "loss": 2.1163,
+      "loss": 1.0839,
       "step": 40
     },
     {
       "epoch": 0.0,
-      "grad_norm": 4.228359222412109,
+      "grad_norm": 10.504097938537598,
       "learning_rate": 2.0000000000000003e-06,
-      "loss": 2.2697,
+      "loss": 1.3778,
       "step": 50
     },
     {
       "epoch": 0.0,
-      "grad_norm": 8.439579963684082,
+      "grad_norm": 10.759912490844727,
       "learning_rate": 2.4000000000000003e-06,
-      "loss": 2.1731,
+      "loss": 1.2921,
       "step": 60
     },
     {
       "epoch": 0.0,
-      "grad_norm": 2.426865339279175,
+      "grad_norm": 10.052595138549805,
       "learning_rate": 2.8000000000000003e-06,
-      "loss": 2.2052,
+      "loss": 1.2835,
       "step": 70
     },
     {
       "epoch": 0.0,
-      "grad_norm": 4.410788059234619,
+      "grad_norm": 4.416482925415039,
       "learning_rate": 3.2000000000000003e-06,
-      "loss": 2.0863,
+      "loss": 1.1876,
       "step": 80
     },
     {
       "epoch": 0.0,
-      "grad_norm": 5.553213596343994,
+      "grad_norm": 8.467272758483887,
       "learning_rate": 3.6000000000000003e-06,
-      "loss": 1.9664,
+      "loss": 1.0513,
       "step": 90
     },
     {
       "epoch": 0.0,
-      "grad_norm": 6.600622177124023,
+      "grad_norm": 7.2040510177612305,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 2.0656,
+      "loss": 1.1429,
       "step": 100
     },
     {
       "epoch": 0.0,
-      "grad_norm": 4.6402153968811035,
+      "grad_norm": 22.72906494140625,
       "learning_rate": 4.4e-06,
-      "loss": 2.0541,
+      "loss": 1.344,
       "step": 110
     },
     {
       "epoch": 0.0,
-      "grad_norm": 3.2395169734954834,
+      "grad_norm": 2.6358766555786133,
       "learning_rate": 4.800000000000001e-06,
-      "loss": 1.8353,
+      "loss": 1.1027,
       "step": 120
     },
     {
       "epoch": 0.0,
-      "grad_norm": 2.8458945751190186,
+      "grad_norm": 2.7037551403045654,
       "learning_rate": 5.2e-06,
-      "loss": 1.8408,
+      "loss": 1.1483,
       "step": 130
     },
     {
       "epoch": 0.0,
-      "grad_norm": 5.223249912261963,
+      "grad_norm": 7.809167861938477,
       "learning_rate": 5.600000000000001e-06,
-      "loss": 1.905,
+      "loss": 1.2556,
       "step": 140
     },
     {
       "epoch": 0.0,
-      "grad_norm": 3.8996775150299072,
+      "grad_norm": 10.522435188293457,
       "learning_rate": 6e-06,
-      "loss": 1.9277,
+      "loss": 1.3272,
       "step": 150
     },
     {
       "epoch": 0.0,
-      "grad_norm": 3.8909013271331787,
+      "grad_norm": 12.723409652709961,
       "learning_rate": 6.4000000000000006e-06,
-      "loss": 1.777,
+      "loss": 1.1342,
       "step": 160
     },
     {
       "epoch": 0.0,
-      "grad_norm": 2.76326060295105,
+      "grad_norm": 9.65882682800293,
       "learning_rate": 6.800000000000001e-06,
-      "loss": 1.7483,
+      "loss": 1.1509,
       "step": 170
     },
     {
       "epoch": 0.0,
-      "grad_norm": 3.6061465740203857,
+      "grad_norm": 7.01952600479126,
       "learning_rate": 7.2000000000000005e-06,
-      "loss": 1.79,
+      "loss": 1.3069,
       "step": 180
     },
     {
       "epoch": 0.0,
-      "grad_norm": 4.046599388122559,
+      "grad_norm": 3.8367834091186523,
       "learning_rate": 7.600000000000001e-06,
-      "loss": 1.7282,
+      "loss": 1.1551,
       "step": 190
     },
     {
       "epoch": 0.0,
-      "grad_norm": 3.0414559841156006,
+      "grad_norm": 6.172680854797363,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 1.8936,
+      "loss": 1.4156,
       "step": 200
     },
     {
       "epoch": 0.0,
-      "grad_norm": 6.000649452209473,
+      "grad_norm": 13.904732704162598,
       "learning_rate": 8.400000000000001e-06,
-      "loss": 1.7291,
+      "loss": 1.2325,
       "step": 210
     },
     {
       "epoch": 0.01,
-      "grad_norm": 5.2610554695129395,
+      "grad_norm": 7.113356590270996,
       "learning_rate": 8.8e-06,
-      "loss": 1.5919,
+      "loss": 1.142,
       "step": 220
     },
     {
       "epoch": 0.01,
-      "grad_norm": 4.346323490142822,
+      "grad_norm": 14.426301956176758,
       "learning_rate": 9.200000000000002e-06,
-      "loss": 1.7002,
+      "loss": 1.2648,
       "step": 230
     },
     {
       "epoch": 0.01,
-      "grad_norm": 6.625606060028076,
+      "grad_norm": 11.00722885131836,
       "learning_rate": 9.600000000000001e-06,
-      "loss": 1.7204,
+      "loss": 1.3083,
       "step": 240
     },
     {
       "epoch": 0.01,
-      "grad_norm": 5.263577938079834,
+      "grad_norm": 5.72845983505249,
       "learning_rate": 1e-05,
-      "loss": 1.6544,
+      "loss": 1.1742,
       "step": 250
     },
     {
       "epoch": 0.01,
-      "grad_norm": 5.526005268096924,
+      "grad_norm": 2.7809460163116455,
       "learning_rate": 1.04e-05,
-      "loss": 1.6343,
+      "loss": 1.2009,
       "step": 260
     },
     {
       "epoch": 0.01,
-      "grad_norm": 6.9661431312561035,
+      "grad_norm": 11.464003562927246,
       "learning_rate": 1.0800000000000002e-05,
-      "loss": 1.6069,
+      "loss": 1.1236,
       "step": 270
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.983661651611328,
+      "grad_norm": 4.139829158782959,
       "learning_rate": 1.1200000000000001e-05,
-      "loss": 1.718,
+      "loss": 1.412,
       "step": 280
     },
     {
       "epoch": 0.01,
-      "grad_norm": 7.731842517852783,
+      "grad_norm": 11.483891487121582,
       "learning_rate": 1.16e-05,
-      "loss": 1.6998,
+      "loss": 1.2858,
       "step": 290
     },
     {
       "epoch": 0.01,
-      "grad_norm": 6.424485683441162,
+      "grad_norm": 5.432833671569824,
       "learning_rate": 1.2e-05,
-      "loss": 1.7128,
+      "loss": 1.2636,
       "step": 300
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.6413464546203613,
+      "grad_norm": 8.610489845275879,
       "learning_rate": 1.2400000000000002e-05,
-      "loss": 1.7114,
+      "loss": 1.3628,
       "step": 310
     },
     {
       "epoch": 0.01,
-      "grad_norm": 6.612671375274658,
+      "grad_norm": 5.597244739532471,
       "learning_rate": 1.2800000000000001e-05,
-      "loss": 1.717,
+      "loss": 1.3585,
       "step": 320
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.804788827896118,
+      "grad_norm": 11.667290687561035,
       "learning_rate": 1.3200000000000002e-05,
-      "loss": 1.7916,
+      "loss": 1.4112,
       "step": 330
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.907038688659668,
+      "grad_norm": 4.102590084075928,
       "learning_rate": 1.3600000000000002e-05,
-      "loss": 1.6078,
+      "loss": 1.2255,
       "step": 340
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.003390312194824,
+      "grad_norm": 15.342700958251953,
       "learning_rate": 1.4e-05,
-      "loss": 1.7339,
+      "loss": 1.2915,
       "step": 350
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.2724103927612305,
+      "grad_norm": 13.78890323638916,
       "learning_rate": 1.4400000000000001e-05,
-      "loss": 1.8393,
+      "loss": 1.5371,
       "step": 360
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.7622690200805664,
+      "grad_norm": 13.456836700439453,
       "learning_rate": 1.48e-05,
-      "loss": 1.5091,
+      "loss": 1.0586,
       "step": 370
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.663651943206787,
+      "grad_norm": 2.98771071434021,
       "learning_rate": 1.5200000000000002e-05,
-      "loss": 1.639,
+      "loss": 1.3475,
       "step": 380
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.8204166889190674,
+      "grad_norm": 6.61796236038208,
       "learning_rate": 1.5600000000000003e-05,
-      "loss": 1.4148,
+      "loss": 1.1562,
       "step": 390
     },
     {
       "epoch": 0.01,
-      "grad_norm": 4.282427787780762,
+      "grad_norm": 5.871325492858887,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.7065,
+      "loss": 1.3575,
       "step": 400
     },
     {
       "epoch": 0.01,
-      "grad_norm": 9.151147842407227,
+      "grad_norm": 6.282336235046387,
       "learning_rate": 1.64e-05,
-      "loss": 1.5874,
+      "loss": 1.3261,
       "step": 410
     },
     {
       "epoch": 0.01,
-      "grad_norm": 4.359065532684326,
+      "grad_norm": 5.444363594055176,
       "learning_rate": 1.6800000000000002e-05,
-      "loss": 1.4316,
+      "loss": 1.0091,
       "step": 420
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.6506259441375732,
+      "grad_norm": 1.7336666584014893,
       "learning_rate": 1.72e-05,
-      "loss": 1.5907,
+      "loss": 1.3492,
       "step": 430
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.2175769805908203,
+      "grad_norm": 5.654507637023926,
       "learning_rate": 1.76e-05,
-      "loss": 1.6386,
+      "loss": 1.2701,
       "step": 440
     },
     {
       "epoch": 0.01,
-      "grad_norm": 6.790574550628662,
+      "grad_norm": 3.83292555809021,
       "learning_rate": 1.8e-05,
-      "loss": 1.4903,
+      "loss": 1.1073,
       "step": 450
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.407963514328003,
+      "grad_norm": 5.379516124725342,
       "learning_rate": 1.8400000000000003e-05,
-      "loss": 1.5074,
+      "loss": 1.2015,
       "step": 460
     },
     {
       "epoch": 0.01,
-      "grad_norm": 4.586065769195557,
+      "grad_norm": 3.365577220916748,
       "learning_rate": 1.88e-05,
-      "loss": 1.4186,
+      "loss": 1.1174,
       "step": 470
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.2595417499542236,
+      "grad_norm": 11.620206832885742,
       "learning_rate": 1.9200000000000003e-05,
-      "loss": 1.5489,
+      "loss": 1.3091,
       "step": 480
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.8598222732543945,
+      "grad_norm": 10.400999069213867,
       "learning_rate": 1.9600000000000002e-05,
-      "loss": 1.314,
+      "loss": 0.98,
       "step": 490
     },
     {
       "epoch": 0.01,
-      "grad_norm": 5.427089214324951,
+      "grad_norm": 5.732550144195557,
       "learning_rate": 2e-05,
-      "loss": 1.7116,
+      "loss": 1.3638,
       "step": 500
     },
     {
       "epoch": 0.01,
-      "eval_loss": 1.4248766899108887,
-      "eval_runtime": 67.1505,
-      "eval_samples_per_second": 14.892,
-      "eval_steps_per_second": 14.892,
+      "eval_loss": 1.0330229997634888,
+      "eval_runtime": 66.9668,
+      "eval_samples_per_second": 14.933,
+      "eval_steps_per_second": 14.933,
       "step": 500
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.126067876815796,
-      "learning_rate": 1.999322033898305e-05,
-      "loss": 1.6259,
+      "grad_norm": 3.0168514251708984,
+      "learning_rate": 1.9955555555555557e-05,
+      "loss": 1.2762,
       "step": 510
     },
     {
       "epoch": 0.01,
-      "grad_norm": 4.224739074707031,
-      "learning_rate": 1.9986440677966104e-05,
-      "loss": 1.6205,
+      "grad_norm": 3.4872841835021973,
+      "learning_rate": 1.9911111111111112e-05,
+      "loss": 1.3287,
       "step": 520
     },
     {
       "epoch": 0.01,
-      "grad_norm": 7.176899433135986,
-      "learning_rate": 1.9979661016949154e-05,
-      "loss": 1.5216,
+      "grad_norm": 5.1740336418151855,
+      "learning_rate": 1.9866666666666667e-05,
+      "loss": 1.2577,
       "step": 530
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.984959840774536,
-      "learning_rate": 1.9972881355932204e-05,
-      "loss": 1.3944,
+      "grad_norm": 2.3551251888275146,
+      "learning_rate": 1.9822222222222226e-05,
+      "loss": 1.1244,
       "step": 540
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.6579527854919434,
-      "learning_rate": 1.9966101694915257e-05,
-      "loss": 1.5515,
+      "grad_norm": 2.230109691619873,
+      "learning_rate": 1.977777777777778e-05,
+      "loss": 1.2021,
       "step": 550
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.6600215435028076,
-      "learning_rate": 1.9959322033898307e-05,
-      "loss": 1.5275,
+      "grad_norm": 2.284055709838867,
+      "learning_rate": 1.9733333333333336e-05,
+      "loss": 1.2919,
       "step": 560
     },
     {
       "epoch": 0.01,
-      "grad_norm": 5.994769096374512,
-      "learning_rate": 1.995254237288136e-05,
-      "loss": 1.6558,
+      "grad_norm": 2.5889816284179688,
+      "learning_rate": 1.968888888888889e-05,
+      "loss": 1.3648,
       "step": 570
     },
     {
       "epoch": 0.01,
-      "grad_norm": 5.50462532043457,
-      "learning_rate": 1.994576271186441e-05,
-      "loss": 1.641,
+      "grad_norm": 4.799360752105713,
+      "learning_rate": 1.9644444444444447e-05,
+      "loss": 1.396,
       "step": 580
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.41587233543396,
-      "learning_rate": 1.993898305084746e-05,
-      "loss": 1.461,
+      "grad_norm": 2.2582507133483887,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.2263,
       "step": 590
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.1005308628082275,
-      "learning_rate": 1.9932203389830512e-05,
-      "loss": 1.4009,
+      "grad_norm": 9.519213676452637,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 1.1475,
       "step": 600
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3.533121347427368,
-      "learning_rate": 1.992542372881356e-05,
-      "loss": 1.7164,
+      "grad_norm": 1.0547032356262207,
+      "learning_rate": 1.9511111111111113e-05,
+      "loss": 1.4278,
       "step": 610
     },
     {
       "epoch": 0.01,
-      "grad_norm": 4.897855758666992,
-      "learning_rate": 1.991864406779661e-05,
-      "loss": 1.5889,
+      "grad_norm": 9.448994636535645,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 1.3394,
       "step": 620
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.890028953552246,
-      "learning_rate": 1.991186440677966e-05,
-      "loss": 1.4549,
+      "grad_norm": 6.21259069442749,
+      "learning_rate": 1.9422222222222223e-05,
+      "loss": 1.206,
       "step": 630
     },
     {
       "epoch": 0.01,
-      "grad_norm": 5.191209316253662,
-      "learning_rate": 1.990508474576271e-05,
-      "loss": 1.4711,
+      "grad_norm": 3.4361753463745117,
+      "learning_rate": 1.9377777777777778e-05,
+      "loss": 1.2472,
       "step": 640
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.568540334701538,
-      "learning_rate": 1.9898305084745764e-05,
-      "loss": 1.3507,
+      "grad_norm": 2.0469722747802734,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 1.1283,
       "step": 650
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.9555094242095947,
-      "learning_rate": 1.9891525423728814e-05,
-      "loss": 1.4569,
+      "grad_norm": 3.071639060974121,
+      "learning_rate": 1.928888888888889e-05,
+      "loss": 1.2191,
       "step": 660
     },
     {
       "epoch": 0.02,
-      "grad_norm": 7.103305339813232,
-      "learning_rate": 1.9884745762711867e-05,
-      "loss": 1.6679,
+      "grad_norm": 3.428431272506714,
+      "learning_rate": 1.9244444444444444e-05,
+      "loss": 1.3863,
       "step": 670
     },
     {
       "epoch": 0.02,
-      "grad_norm": 5.208203315734863,
-      "learning_rate": 1.9877966101694917e-05,
-      "loss": 1.3169,
+      "grad_norm": 7.4005656242370605,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.0794,
       "step": 680
     },
     {
       "epoch": 0.02,
-      "grad_norm": 7.401998996734619,
-      "learning_rate": 1.9871186440677966e-05,
-      "loss": 1.642,
+      "grad_norm": 2.636923313140869,
+      "learning_rate": 1.9155555555555558e-05,
+      "loss": 1.3761,
       "step": 690
     },
     {
       "epoch": 0.02,
-      "grad_norm": 6.582605361938477,
-      "learning_rate": 1.986440677966102e-05,
-      "loss": 1.5996,
+      "grad_norm": 5.060346603393555,
+      "learning_rate": 1.9111111111111113e-05,
+      "loss": 1.4087,
       "step": 700
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.554516315460205,
-      "learning_rate": 1.985762711864407e-05,
-      "loss": 1.5291,
+      "grad_norm": 6.77576208114624,
+      "learning_rate": 1.9066666666666668e-05,
+      "loss": 1.2442,
       "step": 710
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.563401460647583,
-      "learning_rate": 1.985084745762712e-05,
-      "loss": 1.4608,
+      "grad_norm": 3.3169615268707275,
+      "learning_rate": 1.9022222222222223e-05,
+      "loss": 1.2498,
       "step": 720
     },
     {
       "epoch": 0.02,
-      "grad_norm": 5.061882972717285,
-      "learning_rate": 1.9844067796610172e-05,
-      "loss": 1.6631,
+      "grad_norm": 3.9390623569488525,
+      "learning_rate": 1.897777777777778e-05,
+      "loss": 1.4098,
       "step": 730
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.9863511323928833,
-      "learning_rate": 1.9837288135593222e-05,
-      "loss": 1.5619,
+      "grad_norm": 5.928336143493652,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 1.235,
       "step": 740
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.3460112810134888,
-      "learning_rate": 1.9830508474576275e-05,
-      "loss": 1.3738,
+      "grad_norm": 9.329615592956543,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 1.2166,
       "step": 750
     },
     {
       "epoch": 0.02,
-      "grad_norm": 5.177835464477539,
-      "learning_rate": 1.9823728813559324e-05,
-      "loss": 1.6424,
+      "grad_norm": 6.106197357177734,
+      "learning_rate": 1.8844444444444444e-05,
+      "loss": 1.3728,
       "step": 760
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.877689242362976,
-      "learning_rate": 1.9816949152542374e-05,
-      "loss": 1.4095,
+      "grad_norm": 4.729337215423584,
+      "learning_rate": 1.88e-05,
+      "loss": 1.2131,
       "step": 770
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.7550783157348633,
-      "learning_rate": 1.9810169491525427e-05,
-      "loss": 1.5935,
+      "grad_norm": 3.116116762161255,
+      "learning_rate": 1.8755555555555558e-05,
+      "loss": 1.4169,
       "step": 780
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.4884017705917358,
-      "learning_rate": 1.9803389830508477e-05,
-      "loss": 1.4442,
+      "grad_norm": 8.869202613830566,
+      "learning_rate": 1.8711111111111113e-05,
+      "loss": 1.2377,
       "step": 790
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.309656858444214,
-      "learning_rate": 1.9796610169491527e-05,
-      "loss": 1.5698,
+      "grad_norm": 4.858852863311768,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.3863,
       "step": 800
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.5358872413635254,
-      "learning_rate": 1.978983050847458e-05,
-      "loss": 1.2542,
+      "grad_norm": 10.197395324707031,
+      "learning_rate": 1.8622222222222224e-05,
+      "loss": 1.1053,
       "step": 810
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.623551368713379,
-      "learning_rate": 1.9783050847457626e-05,
-      "loss": 1.663,
+      "grad_norm": 2.7740931510925293,
+      "learning_rate": 1.857777777777778e-05,
+      "loss": 1.41,
       "step": 820
     },
     {
       "epoch": 0.02,
-      "grad_norm": 4.512385845184326,
-      "learning_rate": 1.977627118644068e-05,
-      "loss": 1.5636,
+      "grad_norm": 8.306866645812988,
+      "learning_rate": 1.8533333333333334e-05,
+      "loss": 1.2661,
       "step": 830
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.330457925796509,
-      "learning_rate": 1.976949152542373e-05,
-      "loss": 1.3357,
+      "grad_norm": 5.462616920471191,
+      "learning_rate": 1.848888888888889e-05,
+      "loss": 1.1217,
       "step": 840
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.6951920986175537,
-      "learning_rate": 1.9762711864406782e-05,
-      "loss": 1.5766,
+      "grad_norm": 1.6351518630981445,
+      "learning_rate": 1.8444444444444448e-05,
+      "loss": 1.2512,
       "step": 850
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.8462748527526855,
-      "learning_rate": 1.9755932203389832e-05,
-      "loss": 1.5554,
+      "grad_norm": 4.930731773376465,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.3136,
       "step": 860
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.8278895616531372,
-      "learning_rate": 1.974915254237288e-05,
-      "loss": 1.3479,
+      "grad_norm": 6.907737731933594,
+      "learning_rate": 1.835555555555556e-05,
+      "loss": 1.1728,
       "step": 870
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.1370766162872314,
-      "learning_rate": 1.9742372881355935e-05,
-      "loss": 1.5584,
+      "grad_norm": 2.2834455966949463,
+      "learning_rate": 1.8311111111111114e-05,
+      "loss": 1.3566,
       "step": 880
     },
     {
       "epoch": 0.02,
-      "grad_norm": 8.17900276184082,
-      "learning_rate": 1.9735593220338984e-05,
-      "loss": 1.5172,
+      "grad_norm": 6.938192367553711,
+      "learning_rate": 1.826666666666667e-05,
+      "loss": 1.2879,
       "step": 890
     },
     {
       "epoch": 0.02,
-      "grad_norm": 7.5510640144348145,
-      "learning_rate": 1.9728813559322034e-05,
-      "loss": 1.4574,
+      "grad_norm": 3.4376509189605713,
+      "learning_rate": 1.8222222222222224e-05,
+      "loss": 1.1971,
       "step": 900
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.811005115509033,
-      "learning_rate": 1.9722033898305087e-05,
-      "loss": 1.576,
+      "grad_norm": 4.437848091125488,
+      "learning_rate": 1.817777777777778e-05,
+      "loss": 1.3583,
       "step": 910
     },
     {
       "epoch": 0.02,
-      "grad_norm": 4.8449907302856445,
-      "learning_rate": 1.9715254237288137e-05,
-      "loss": 1.4886,
+      "grad_norm": 6.843893051147461,
+      "learning_rate": 1.8133333333333335e-05,
+      "loss": 1.2323,
       "step": 920
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.2932374477386475,
-      "learning_rate": 1.970847457627119e-05,
-      "loss": 1.4845,
+      "grad_norm": 3.2527034282684326,
+      "learning_rate": 1.808888888888889e-05,
+      "loss": 1.2691,
       "step": 930
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.189581871032715,
-      "learning_rate": 1.970169491525424e-05,
-      "loss": 1.4143,
+      "grad_norm": 2.6528022289276123,
+      "learning_rate": 1.8044444444444445e-05,
+      "loss": 1.2525,
       "step": 940
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.5806180238723755,
-      "learning_rate": 1.969491525423729e-05,
-      "loss": 1.4859,
+      "grad_norm": 2.8804101943969727,
+      "learning_rate": 1.8e-05,
+      "loss": 1.3387,
       "step": 950
     },
     {
       "epoch": 0.02,
-      "grad_norm": 5.293037414550781,
-      "learning_rate": 1.9688135593220342e-05,
-      "loss": 1.4887,
+      "grad_norm": 9.754573822021484,
+      "learning_rate": 1.7955555555555556e-05,
+      "loss": 1.265,
       "step": 960
     },
     {
       "epoch": 0.02,
-      "grad_norm": 5.328525543212891,
-      "learning_rate": 1.9681355932203392e-05,
-      "loss": 1.2715,
+      "grad_norm": 2.54309344291687,
+      "learning_rate": 1.791111111111111e-05,
+      "loss": 1.0998,
       "step": 970
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.8140487670898438,
-      "learning_rate": 1.9674576271186442e-05,
-      "loss": 1.4281,
+      "grad_norm": 9.926447868347168,
+      "learning_rate": 1.7866666666666666e-05,
+      "loss": 1.1927,
       "step": 980
     },
     {
       "epoch": 0.02,
-      "grad_norm": 5.055774688720703,
-      "learning_rate": 1.9667796610169495e-05,
-      "loss": 1.2676,
+      "grad_norm": 6.4870805740356445,
+      "learning_rate": 1.782222222222222e-05,
+      "loss": 1.0565,
       "step": 990
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.9871106147766113,
-      "learning_rate": 1.9661016949152545e-05,
-      "loss": 1.4242,
+      "grad_norm": 3.2183382511138916,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 1.2572,
       "step": 1000
     },
     {
       "epoch": 0.02,
-      "eval_loss": 1.2329587936401367,
-      "eval_runtime": 67.1625,
-      "eval_samples_per_second": 14.889,
-      "eval_steps_per_second": 14.889,
+      "eval_loss": 1.019097089767456,
+      "eval_runtime": 66.8787,
+      "eval_samples_per_second": 14.952,
+      "eval_steps_per_second": 14.952,
       "step": 1000
     },
     {
       "epoch": 0.02,
-      "grad_norm": 9.538952827453613,
-      "learning_rate": 1.9654237288135594e-05,
-      "loss": 1.664,
+      "grad_norm": 9.61516284942627,
+      "learning_rate": 1.7733333333333335e-05,
+      "loss": 1.4306,
       "step": 1010
     },
     {
       "epoch": 0.02,
-      "grad_norm": 7.894416809082031,
-      "learning_rate": 1.9647457627118644e-05,
-      "loss": 1.5421,
+      "grad_norm": 10.755435943603516,
+      "learning_rate": 1.768888888888889e-05,
+      "loss": 1.3042,
       "step": 1020
     },
     {
       "epoch": 0.02,
-      "grad_norm": 15.61988353729248,
-      "learning_rate": 1.9640677966101697e-05,
-      "loss": 1.5037,
+      "grad_norm": 11.427221298217773,
+      "learning_rate": 1.7644444444444446e-05,
+      "loss": 1.3201,
       "step": 1030
     },
     {
       "epoch": 0.02,
-      "grad_norm": 1.5587459802627563,
-      "learning_rate": 1.9633898305084747e-05,
-      "loss": 1.4768,
+      "grad_norm": 9.656847953796387,
+      "learning_rate": 1.76e-05,
+      "loss": 1.2579,
       "step": 1040
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.0465331077575684,
-      "learning_rate": 1.9627118644067796e-05,
-      "loss": 1.6342,
+      "grad_norm": 4.030166149139404,
+      "learning_rate": 1.7555555555555556e-05,
+      "loss": 1.4462,
       "step": 1050
     },
     {
       "epoch": 0.02,
-      "grad_norm": 2.6867032051086426,
-      "learning_rate": 1.962033898305085e-05,
-      "loss": 1.6395,
+      "grad_norm": 6.84343957901001,
+      "learning_rate": 1.751111111111111e-05,
+      "loss": 1.4213,
       "step": 1060
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.0958399772644043,
-      "learning_rate": 1.96135593220339e-05,
-      "loss": 1.3742,
+      "grad_norm": 2.4579355716705322,
+      "learning_rate": 1.7466666666666667e-05,
+      "loss": 1.2673,
       "step": 1070
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.064396858215332,
-      "learning_rate": 1.960677966101695e-05,
-      "loss": 1.258,
+      "grad_norm": 6.692587375640869,
+      "learning_rate": 1.7422222222222222e-05,
+      "loss": 1.1579,
       "step": 1080
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.9529526233673096,
-      "learning_rate": 1.9600000000000002e-05,
-      "loss": 1.6082,
+      "grad_norm": 4.579308032989502,
+      "learning_rate": 1.737777777777778e-05,
+      "loss": 1.3278,
       "step": 1090
     },
     {
       "epoch": 0.03,
-      "grad_norm": 8.01111125946045,
-      "learning_rate": 1.9593220338983052e-05,
-      "loss": 1.5338,
+      "grad_norm": 7.755845069885254,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 1.3471,
       "step": 1100
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.7759506702423096,
-      "learning_rate": 1.95864406779661e-05,
-      "loss": 1.4728,
+      "grad_norm": 2.065462112426758,
+      "learning_rate": 1.728888888888889e-05,
+      "loss": 1.2811,
       "step": 1110
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.339688777923584,
-      "learning_rate": 1.9579661016949155e-05,
-      "loss": 1.5947,
+      "grad_norm": 1.6691714525222778,
+      "learning_rate": 1.7244444444444446e-05,
+      "loss": 1.4433,
       "step": 1120
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.9827146530151367,
-      "learning_rate": 1.9572881355932204e-05,
-      "loss": 1.4338,
+      "grad_norm": 6.87007474899292,
+      "learning_rate": 1.72e-05,
+      "loss": 1.1916,
       "step": 1130
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.902629852294922,
-      "learning_rate": 1.9566101694915257e-05,
-      "loss": 1.5609,
+      "grad_norm": 1.663476586341858,
+      "learning_rate": 1.7155555555555557e-05,
+      "loss": 1.3006,
       "step": 1140
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.317248582839966,
-      "learning_rate": 1.9559322033898307e-05,
-      "loss": 1.5503,
+      "grad_norm": 2.493046522140503,
+      "learning_rate": 1.7111111111111112e-05,
+      "loss": 1.3099,
       "step": 1150
     },
     {
       "epoch": 0.03,
-      "grad_norm": 1.522884726524353,
-      "learning_rate": 1.9552542372881357e-05,
-      "loss": 1.3661,
+      "grad_norm": 3.2659084796905518,
+      "learning_rate": 1.706666666666667e-05,
+      "loss": 1.1553,
       "step": 1160
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.1419677734375,
-      "learning_rate": 1.954576271186441e-05,
-      "loss": 1.4839,
+      "grad_norm": 3.07832670211792,
+      "learning_rate": 1.7022222222222226e-05,
+      "loss": 1.2604,
       "step": 1170
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.2150516510009766,
-      "learning_rate": 1.953898305084746e-05,
-      "loss": 1.1754,
+      "grad_norm": 2.0068790912628174,
+      "learning_rate": 1.697777777777778e-05,
+      "loss": 0.9679,
       "step": 1180
     },
     {
       "epoch": 0.03,
-      "grad_norm": 4.25466775894165,
-      "learning_rate": 1.953220338983051e-05,
-      "loss": 1.4118,
+      "grad_norm": 14.066789627075195,
+      "learning_rate": 1.6933333333333336e-05,
+      "loss": 1.2291,
       "step": 1190
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.261260986328125,
-      "learning_rate": 1.9525423728813562e-05,
-      "loss": 1.4779,
+      "grad_norm": 5.058719158172607,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 1.3224,
       "step": 1200
     },
     {
       "epoch": 0.03,
-      "grad_norm": 4.751726150512695,
-      "learning_rate": 1.9518644067796612e-05,
-      "loss": 1.3143,
+      "grad_norm": 6.309176445007324,
+      "learning_rate": 1.6844444444444447e-05,
+      "loss": 1.1184,
       "step": 1210
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.0561487674713135,
-      "learning_rate": 1.9511864406779665e-05,
-      "loss": 1.5102,
+      "grad_norm": 9.05258846282959,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.345,
       "step": 1220
     },
     {
       "epoch": 0.03,
-      "grad_norm": 4.0936279296875,
-      "learning_rate": 1.950508474576271e-05,
-      "loss": 1.4763,
+      "grad_norm": 4.345346927642822,
+      "learning_rate": 1.6755555555555557e-05,
+      "loss": 1.2663,
       "step": 1230
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.0847556591033936,
-      "learning_rate": 1.9498305084745765e-05,
-      "loss": 1.3039,
+      "grad_norm": 4.119368076324463,
+      "learning_rate": 1.6711111111111112e-05,
+      "loss": 1.1441,
       "step": 1240
     },
     {
       "epoch": 0.03,
-      "grad_norm": 4.303075313568115,
-      "learning_rate": 1.9491525423728814e-05,
-      "loss": 1.4491,
+      "grad_norm": 5.0422563552856445,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.2199,
       "step": 1250
     },
     {
       "epoch": 0.03,
-      "grad_norm": 1.7520992755889893,
-      "learning_rate": 1.9484745762711864e-05,
-      "loss": 1.3379,
+      "grad_norm": 11.27535343170166,
+      "learning_rate": 1.6622222222222223e-05,
+      "loss": 1.1876,
       "step": 1260
     },
     {
       "epoch": 0.03,
-      "grad_norm": 6.233397006988525,
-      "learning_rate": 1.9477966101694917e-05,
-      "loss": 1.5582,
+      "grad_norm": 6.767408847808838,
+      "learning_rate": 1.6577777777777778e-05,
+      "loss": 1.3928,
       "step": 1270
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.988858461380005,
-      "learning_rate": 1.9471186440677967e-05,
-      "loss": 1.4801,
+      "grad_norm": 6.1706862449646,
+      "learning_rate": 1.6533333333333333e-05,
+      "loss": 1.2268,
       "step": 1280
     },
     {
       "epoch": 0.03,
-      "grad_norm": 5.172889232635498,
-      "learning_rate": 1.9464406779661017e-05,
-      "loss": 1.5481,
+      "grad_norm": 6.185644149780273,
+      "learning_rate": 1.648888888888889e-05,
+      "loss": 1.3086,
       "step": 1290
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.5870141983032227,
-      "learning_rate": 1.945762711864407e-05,
-      "loss": 1.4008,
+      "grad_norm": 9.57487678527832,
+      "learning_rate": 1.6444444444444444e-05,
+      "loss": 1.2079,
       "step": 1300
     },
     {
       "epoch": 0.03,
-      "grad_norm": 4.538797378540039,
-      "learning_rate": 1.945084745762712e-05,
-      "loss": 1.4145,
+      "grad_norm": 4.104099273681641,
+      "learning_rate": 1.64e-05,
+      "loss": 1.2012,
       "step": 1310
     },
     {
       "epoch": 0.03,
-      "grad_norm": 4.275447368621826,
-      "learning_rate": 1.9444067796610172e-05,
-      "loss": 1.4524,
+      "grad_norm": 6.703191757202148,
+      "learning_rate": 1.6355555555555557e-05,
+      "loss": 1.2567,
       "step": 1320
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.357238531112671,
-      "learning_rate": 1.9437288135593222e-05,
-      "loss": 1.487,
+      "grad_norm": 4.194169521331787,
+      "learning_rate": 1.6311111111111113e-05,
+      "loss": 1.3687,
       "step": 1330
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.1907289028167725,
-      "learning_rate": 1.9430508474576272e-05,
-      "loss": 1.3633,
+      "grad_norm": 4.2333269119262695,
+      "learning_rate": 1.6266666666666668e-05,
+      "loss": 1.1547,
       "step": 1340
     },
     {
       "epoch": 0.03,
-      "grad_norm": 1.7265887260437012,
-      "learning_rate": 1.9423728813559325e-05,
-      "loss": 1.3923,
+      "grad_norm": 3.007706880569458,
+      "learning_rate": 1.6222222222222223e-05,
+      "loss": 1.2031,
       "step": 1350
     },
     {
       "epoch": 0.03,
-      "grad_norm": 10.415843963623047,
-      "learning_rate": 1.9416949152542375e-05,
-      "loss": 1.3302,
+      "grad_norm": 22.12256622314453,
+      "learning_rate": 1.617777777777778e-05,
+      "loss": 1.1117,
       "step": 1360
     },
     {
       "epoch": 0.03,
-      "grad_norm": 6.544619560241699,
-      "learning_rate": 1.9410169491525424e-05,
-      "loss": 1.2906,
+      "grad_norm": 5.559662818908691,
+      "learning_rate": 1.6133333333333334e-05,
+      "loss": 1.0796,
       "step": 1370
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.8395652770996094,
-      "learning_rate": 1.9403389830508477e-05,
-      "loss": 1.4636,
+      "grad_norm": 4.618852615356445,
+      "learning_rate": 1.608888888888889e-05,
+      "loss": 1.3308,
       "step": 1380
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.81257963180542,
-      "learning_rate": 1.9396610169491527e-05,
-      "loss": 1.408,
+      "grad_norm": 3.6262142658233643,
+      "learning_rate": 1.6044444444444444e-05,
+      "loss": 1.1913,
       "step": 1390
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.445089817047119,
-      "learning_rate": 1.938983050847458e-05,
-      "loss": 1.5095,
+      "grad_norm": 8.0167875289917,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.341,
       "step": 1400
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.918297290802002,
-      "learning_rate": 1.938305084745763e-05,
-      "loss": 1.154,
+      "grad_norm": 5.142825603485107,
+      "learning_rate": 1.5955555555555558e-05,
+      "loss": 1.03,
       "step": 1410
     },
     {
       "epoch": 0.03,
-      "grad_norm": 1.7477134466171265,
-      "learning_rate": 1.937627118644068e-05,
-      "loss": 1.4866,
+      "grad_norm": 3.186448335647583,
+      "learning_rate": 1.5911111111111113e-05,
+      "loss": 1.2788,
       "step": 1420
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.768603801727295,
-      "learning_rate": 1.9369491525423733e-05,
-      "loss": 1.5642,
+      "grad_norm": 8.11291217803955,
+      "learning_rate": 1.586666666666667e-05,
+      "loss": 1.4419,
       "step": 1430
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.055950164794922,
-      "learning_rate": 1.936271186440678e-05,
-      "loss": 1.4019,
+      "grad_norm": 6.920192718505859,
+      "learning_rate": 1.5822222222222224e-05,
+      "loss": 1.3127,
       "step": 1440
     },
     {
       "epoch": 0.03,
-      "grad_norm": 9.180340766906738,
-      "learning_rate": 1.9355932203389832e-05,
-      "loss": 1.509,
+      "grad_norm": 4.357940196990967,
+      "learning_rate": 1.577777777777778e-05,
+      "loss": 1.3184,
       "step": 1450
     },
     {
       "epoch": 0.03,
-      "grad_norm": 4.526889801025391,
-      "learning_rate": 1.9349152542372882e-05,
-      "loss": 1.5197,
+      "grad_norm": 8.83785343170166,
+      "learning_rate": 1.5733333333333334e-05,
+      "loss": 1.382,
       "step": 1460
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.020207166671753,
-      "learning_rate": 1.934237288135593e-05,
-      "loss": 1.6292,
+      "grad_norm": 2.565369129180908,
+      "learning_rate": 1.5688888888888893e-05,
+      "loss": 1.4148,
       "step": 1470
     },
     {
       "epoch": 0.03,
-      "grad_norm": 3.9965012073516846,
-      "learning_rate": 1.9335593220338985e-05,
-      "loss": 1.6703,
+      "grad_norm": 3.819150686264038,
+      "learning_rate": 1.5644444444444448e-05,
+      "loss": 1.4428,
       "step": 1480
     },
     {
       "epoch": 0.03,
-      "grad_norm": 0.9574942588806152,
-      "learning_rate": 1.9328813559322034e-05,
-      "loss": 1.3525,
+      "grad_norm": 7.107534408569336,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 1.2188,
       "step": 1490
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.8527791500091553,
-      "learning_rate": 1.9322033898305087e-05,
-      "loss": 1.5142,
+      "grad_norm": 9.566027641296387,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 1.2406,
       "step": 1500
     },
     {
       "epoch": 0.03,
-      "eval_loss": 1.1832261085510254,
-      "eval_runtime": 67.0686,
-      "eval_samples_per_second": 14.91,
-      "eval_steps_per_second": 14.91,
+      "eval_loss": 1.0252078771591187,
+      "eval_runtime": 66.9753,
+      "eval_samples_per_second": 14.931,
+      "eval_steps_per_second": 14.931,
       "step": 1500
     },
     {
       "epoch": 0.03,
-      "grad_norm": 5.690788745880127,
-      "learning_rate": 1.9315254237288137e-05,
-      "loss": 1.4003,
+      "grad_norm": 8.054425239562988,
+      "learning_rate": 1.5511111111111114e-05,
+      "loss": 1.2767,
       "step": 1510
     },
     {
       "epoch": 0.04,
-      "grad_norm": 7.6338605880737305,
-      "learning_rate": 1.9308474576271187e-05,
-      "loss": 1.4324,
+      "grad_norm": 2.8494319915771484,
+      "learning_rate": 1.546666666666667e-05,
+      "loss": 1.1632,
       "step": 1520
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.098454713821411,
-      "learning_rate": 1.930169491525424e-05,
-      "loss": 1.5234,
+      "grad_norm": 2.9305810928344727,
+      "learning_rate": 1.5422222222222224e-05,
+      "loss": 1.2075,
       "step": 1530
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.322094440460205,
-      "learning_rate": 1.929491525423729e-05,
-      "loss": 1.2201,
+      "grad_norm": 11.442655563354492,
+      "learning_rate": 1.537777777777778e-05,
+      "loss": 1.0231,
       "step": 1540
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.089012622833252,
-      "learning_rate": 1.928813559322034e-05,
-      "loss": 1.4444,
+      "grad_norm": 1.735372543334961,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 1.282,
       "step": 1550
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.2624104022979736,
-      "learning_rate": 1.9281355932203392e-05,
-      "loss": 1.4144,
+      "grad_norm": 11.126168251037598,
+      "learning_rate": 1.528888888888889e-05,
+      "loss": 1.2106,
       "step": 1560
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.4324264526367188,
-      "learning_rate": 1.9274576271186442e-05,
-      "loss": 1.3715,
+      "grad_norm": 5.569930076599121,
+      "learning_rate": 1.5244444444444447e-05,
+      "loss": 1.2016,
       "step": 1570
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.596635341644287,
-      "learning_rate": 1.9267796610169492e-05,
-      "loss": 1.366,
+      "grad_norm": 4.002272605895996,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.206,
       "step": 1580
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.6570637226104736,
-      "learning_rate": 1.9261016949152545e-05,
-      "loss": 1.5426,
+      "grad_norm": 11.26425838470459,
+      "learning_rate": 1.5155555555555557e-05,
+      "loss": 1.3145,
       "step": 1590
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.935263156890869,
-      "learning_rate": 1.9254237288135595e-05,
-      "loss": 1.2968,
+      "grad_norm": 6.265772819519043,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 1.105,
       "step": 1600
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.188492298126221,
-      "learning_rate": 1.9247457627118648e-05,
-      "loss": 1.396,
+      "grad_norm": 6.139275550842285,
+      "learning_rate": 1.5066666666666668e-05,
+      "loss": 1.2016,
       "step": 1610
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.421194553375244,
-      "learning_rate": 1.9240677966101698e-05,
-      "loss": 1.4905,
+      "grad_norm": 4.753066539764404,
+      "learning_rate": 1.5022222222222223e-05,
+      "loss": 1.3433,
       "step": 1620
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.735612154006958,
-      "learning_rate": 1.9233898305084747e-05,
-      "loss": 1.4508,
+      "grad_norm": 8.761942863464355,
+      "learning_rate": 1.497777777777778e-05,
+      "loss": 1.3153,
       "step": 1630
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.989076018333435,
-      "learning_rate": 1.92271186440678e-05,
-      "loss": 1.4195,
+      "grad_norm": 3.3448381423950195,
+      "learning_rate": 1.4933333333333335e-05,
+      "loss": 1.24,
       "step": 1640
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.4910849332809448,
-      "learning_rate": 1.9220338983050847e-05,
-      "loss": 1.374,
+      "grad_norm": 2.818711519241333,
+      "learning_rate": 1.488888888888889e-05,
+      "loss": 1.2067,
       "step": 1650
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.7809735536575317,
-      "learning_rate": 1.92135593220339e-05,
-      "loss": 1.5008,
+      "grad_norm": 11.795276641845703,
+      "learning_rate": 1.4844444444444445e-05,
+      "loss": 1.3926,
       "step": 1660
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.1804401874542236,
-      "learning_rate": 1.920677966101695e-05,
-      "loss": 1.7274,
+      "grad_norm": 4.47786283493042,
+      "learning_rate": 1.48e-05,
+      "loss": 1.5349,
       "step": 1670
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.3297713994979858,
-      "learning_rate": 1.9200000000000003e-05,
-      "loss": 1.3982,
+      "grad_norm": 4.560647010803223,
+      "learning_rate": 1.4755555555555556e-05,
+      "loss": 1.2318,
       "step": 1680
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.2256460189819336,
-      "learning_rate": 1.9193220338983052e-05,
-      "loss": 1.3758,
+      "grad_norm": 5.564818382263184,
+      "learning_rate": 1.4711111111111111e-05,
+      "loss": 1.2044,
       "step": 1690
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.3911705017089844,
-      "learning_rate": 1.9186440677966102e-05,
-      "loss": 1.5054,
+      "grad_norm": 1.4382556676864624,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 1.3367,
       "step": 1700
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.7374446392059326,
-      "learning_rate": 1.9179661016949155e-05,
-      "loss": 1.3796,
+      "grad_norm": 2.7976467609405518,
+      "learning_rate": 1.4622222222222225e-05,
+      "loss": 1.2304,
       "step": 1710
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.8097071647644043,
-      "learning_rate": 1.9172881355932205e-05,
-      "loss": 1.2966,
+      "grad_norm": 3.021933078765869,
+      "learning_rate": 1.457777777777778e-05,
+      "loss": 1.1339,
       "step": 1720
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.590428352355957,
-      "learning_rate": 1.9166101694915254e-05,
-      "loss": 1.493,
+      "grad_norm": 1.9026000499725342,
+      "learning_rate": 1.4533333333333335e-05,
+      "loss": 1.3733,
       "step": 1730
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.458738327026367,
-      "learning_rate": 1.9159322033898308e-05,
-      "loss": 1.3179,
+      "grad_norm": 4.237542152404785,
+      "learning_rate": 1.448888888888889e-05,
+      "loss": 1.0956,
       "step": 1740
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.689072132110596,
-      "learning_rate": 1.9152542372881357e-05,
-      "loss": 1.2673,
+      "grad_norm": 5.288679599761963,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 1.0542,
       "step": 1750
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.915776252746582,
-      "learning_rate": 1.9145762711864407e-05,
-      "loss": 1.5096,
+      "grad_norm": 4.956302642822266,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.326,
       "step": 1760
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.790291786193848,
-      "learning_rate": 1.913898305084746e-05,
-      "loss": 1.2601,
+      "grad_norm": 4.096738338470459,
+      "learning_rate": 1.4355555555555556e-05,
+      "loss": 1.1285,
       "step": 1770
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.455094337463379,
-      "learning_rate": 1.913220338983051e-05,
-      "loss": 1.5662,
+      "grad_norm": 7.757137298583984,
+      "learning_rate": 1.4311111111111111e-05,
+      "loss": 1.3607,
       "step": 1780
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.6922054290771484,
-      "learning_rate": 1.9125423728813563e-05,
-      "loss": 1.1819,
+      "grad_norm": 5.415702819824219,
+      "learning_rate": 1.4266666666666668e-05,
+      "loss": 1.0,
       "step": 1790
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.807013511657715,
-      "learning_rate": 1.9118644067796613e-05,
-      "loss": 1.4572,
+      "grad_norm": 9.445133209228516,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 1.2187,
       "step": 1800
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.976742267608643,
-      "learning_rate": 1.9111864406779662e-05,
-      "loss": 1.4976,
+      "grad_norm": 5.406456470489502,
+      "learning_rate": 1.4177777777777779e-05,
+      "loss": 1.3459,
       "step": 1810
     },
     {
       "epoch": 0.04,
-      "grad_norm": 5.053407669067383,
-      "learning_rate": 1.9105084745762715e-05,
-      "loss": 1.4944,
+      "grad_norm": 1.619770884513855,
+      "learning_rate": 1.4133333333333334e-05,
+      "loss": 1.2368,
       "step": 1820
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.595452308654785,
-      "learning_rate": 1.9098305084745765e-05,
-      "loss": 1.3392,
+      "grad_norm": 4.307522296905518,
+      "learning_rate": 1.408888888888889e-05,
+      "loss": 1.0995,
       "step": 1830
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.3358185291290283,
-      "learning_rate": 1.9091525423728815e-05,
-      "loss": 1.3567,
+      "grad_norm": 6.3472185134887695,
+      "learning_rate": 1.4044444444444445e-05,
+      "loss": 1.2563,
       "step": 1840
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.6956140995025635,
-      "learning_rate": 1.9084745762711868e-05,
-      "loss": 1.4298,
+      "grad_norm": 2.04168701171875,
+      "learning_rate": 1.4e-05,
+      "loss": 1.2858,
       "step": 1850
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.6025402545928955,
-      "learning_rate": 1.9077966101694914e-05,
-      "loss": 1.3674,
+      "grad_norm": 2.180267810821533,
+      "learning_rate": 1.3955555555555558e-05,
+      "loss": 1.123,
       "step": 1860
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.4478222131729126,
-      "learning_rate": 1.9071186440677967e-05,
-      "loss": 1.5232,
+      "grad_norm": 2.560042381286621,
+      "learning_rate": 1.3911111111111114e-05,
+      "loss": 1.3962,
       "step": 1870
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.7879269123077393,
-      "learning_rate": 1.9064406779661017e-05,
-      "loss": 1.2088,
+      "grad_norm": 5.683982849121094,
+      "learning_rate": 1.3866666666666669e-05,
+      "loss": 1.0984,
       "step": 1880
     },
     {
       "epoch": 0.04,
-      "grad_norm": 4.009925365447998,
-      "learning_rate": 1.905762711864407e-05,
-      "loss": 1.306,
+      "grad_norm": 3.1994190216064453,
+      "learning_rate": 1.3822222222222224e-05,
+      "loss": 1.1746,
       "step": 1890
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.8789501190185547,
-      "learning_rate": 1.905084745762712e-05,
-      "loss": 1.6352,
+      "grad_norm": 8.851926803588867,
+      "learning_rate": 1.377777777777778e-05,
+      "loss": 1.4551,
       "step": 1900
     },
     {
       "epoch": 0.04,
-      "grad_norm": 1.9612170457839966,
-      "learning_rate": 1.904406779661017e-05,
-      "loss": 1.4313,
+      "grad_norm": 2.670786142349243,
+      "learning_rate": 1.3733333333333335e-05,
+      "loss": 1.2446,
       "step": 1910
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.5845580101013184,
-      "learning_rate": 1.9037288135593223e-05,
-      "loss": 1.464,
+      "grad_norm": 5.134795188903809,
+      "learning_rate": 1.368888888888889e-05,
+      "loss": 1.2128,
       "step": 1920
     },
     {
       "epoch": 0.04,
-      "grad_norm": 5.5218095779418945,
-      "learning_rate": 1.9030508474576272e-05,
-      "loss": 1.3589,
+      "grad_norm": 4.486630916595459,
+      "learning_rate": 1.3644444444444445e-05,
+      "loss": 1.1951,
       "step": 1930
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.3145925998687744,
-      "learning_rate": 1.9023728813559322e-05,
-      "loss": 1.5245,
+      "grad_norm": 4.025615215301514,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.3628,
       "step": 1940
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.0829366445541382,
-      "learning_rate": 1.9016949152542375e-05,
-      "loss": 1.5265,
+      "grad_norm": 4.232900619506836,
+      "learning_rate": 1.3555555555555557e-05,
+      "loss": 1.3367,
       "step": 1950
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.3161795139312744,
-      "learning_rate": 1.9010169491525425e-05,
-      "loss": 1.3759,
+      "grad_norm": 1.6469148397445679,
+      "learning_rate": 1.3511111111111112e-05,
+      "loss": 1.2144,
       "step": 1960
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.037906169891357,
-      "learning_rate": 1.9003389830508478e-05,
-      "loss": 1.576,
+      "grad_norm": 8.560945510864258,
+      "learning_rate": 1.3466666666666668e-05,
+      "loss": 1.4125,
       "step": 1970
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.638181447982788,
-      "learning_rate": 1.8996610169491528e-05,
-      "loss": 1.3606,
+      "grad_norm": 3.06697416305542,
+      "learning_rate": 1.3422222222222223e-05,
+      "loss": 1.1769,
       "step": 1980
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.6348967552185059,
-      "learning_rate": 1.8989830508474577e-05,
-      "loss": 1.4664,
+      "grad_norm": 2.721186399459839,
+      "learning_rate": 1.3377777777777778e-05,
+      "loss": 1.3266,
       "step": 1990
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.778348684310913,
-      "learning_rate": 1.898305084745763e-05,
-      "loss": 1.308,
+      "grad_norm": 4.427524566650391,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.1697,
       "step": 2000
     },
     {
       "epoch": 0.05,
-      "eval_loss": 1.1971772909164429,
-      "eval_runtime": 67.1893,
-      "eval_samples_per_second": 14.883,
-      "eval_steps_per_second": 14.883,
+      "eval_loss": 1.042366623878479,
+      "eval_runtime": 67.0143,
+      "eval_samples_per_second": 14.922,
+      "eval_steps_per_second": 14.922,
       "step": 2000
     },
     {
       "epoch": 0.05,
-      "grad_norm": 3.655930757522583,
-      "learning_rate": 1.897627118644068e-05,
-      "loss": 1.5353,
+      "grad_norm": 1.322222113609314,
+      "learning_rate": 1.3288888888888889e-05,
+      "loss": 1.3173,
       "step": 2010
     },
     {
       "epoch": 0.05,
-      "grad_norm": 7.265784740447998,
-      "learning_rate": 1.896949152542373e-05,
-      "loss": 1.4901,
+      "grad_norm": 5.731060028076172,
+      "learning_rate": 1.3244444444444447e-05,
+      "loss": 1.2578,
       "step": 2020
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.902195930480957,
-      "learning_rate": 1.8962711864406783e-05,
-      "loss": 1.3619,
+      "grad_norm": 0.8411041498184204,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 1.1796,
       "step": 2030
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.685209035873413,
-      "learning_rate": 1.8955932203389833e-05,
-      "loss": 1.4934,
+      "grad_norm": 5.4170026779174805,
+      "learning_rate": 1.3155555555555558e-05,
+      "loss": 1.3608,
       "step": 2040
     },
     {
       "epoch": 0.05,
-      "grad_norm": 6.6489458084106445,
-      "learning_rate": 1.8949152542372882e-05,
-      "loss": 1.4807,
+      "grad_norm": 4.58616304397583,
+      "learning_rate": 1.3111111111111113e-05,
+      "loss": 1.2726,
       "step": 2050
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.4234724044799805,
-      "learning_rate": 1.8942372881355932e-05,
-      "loss": 1.4953,
+      "grad_norm": 3.936751365661621,
+      "learning_rate": 1.3066666666666668e-05,
+      "loss": 1.3327,
       "step": 2060
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.339203119277954,
-      "learning_rate": 1.8935593220338985e-05,
-      "loss": 1.4583,
+      "grad_norm": 4.1074042320251465,
+      "learning_rate": 1.3022222222222223e-05,
+      "loss": 1.2446,
       "step": 2070
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.0084259510040283,
-      "learning_rate": 1.8928813559322035e-05,
-      "loss": 1.1904,
+      "grad_norm": 2.657953977584839,
+      "learning_rate": 1.2977777777777779e-05,
+      "loss": 0.9989,
       "step": 2080
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.666440963745117,
-      "learning_rate": 1.8922033898305085e-05,
-      "loss": 1.3025,
+      "grad_norm": 5.181591987609863,
+      "learning_rate": 1.2933333333333334e-05,
+      "loss": 1.1732,
       "step": 2090
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.9041775465011597,
-      "learning_rate": 1.8915254237288138e-05,
-      "loss": 1.3873,
+      "grad_norm": 5.981390953063965,
+      "learning_rate": 1.288888888888889e-05,
+      "loss": 1.1463,
       "step": 2100
     },
     {
       "epoch": 0.05,
-      "grad_norm": 3.1416661739349365,
-      "learning_rate": 1.8908474576271187e-05,
-      "loss": 1.3741,
+      "grad_norm": 6.05696964263916,
+      "learning_rate": 1.2844444444444446e-05,
+      "loss": 1.2061,
       "step": 2110
     },
     {
       "epoch": 0.05,
-      "grad_norm": 3.8044958114624023,
-      "learning_rate": 1.8901694915254237e-05,
-      "loss": 1.5686,
+      "grad_norm": 6.697308540344238,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.3052,
       "step": 2120
     },
     {
       "epoch": 0.05,
-      "grad_norm": 3.8283112049102783,
-      "learning_rate": 1.889491525423729e-05,
-      "loss": 1.3731,
+      "grad_norm": 14.833478927612305,
+      "learning_rate": 1.2755555555555556e-05,
+      "loss": 1.1987,
       "step": 2130
     },
     {
       "epoch": 0.05,
-      "grad_norm": 3.2208967208862305,
-      "learning_rate": 1.888813559322034e-05,
-      "loss": 1.5184,
+      "grad_norm": 4.479732990264893,
+      "learning_rate": 1.2711111111111112e-05,
+      "loss": 1.337,
       "step": 2140
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.5146480798721313,
-      "learning_rate": 1.8881355932203393e-05,
-      "loss": 1.3791,
+      "grad_norm": 3.746943235397339,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 1.3082,
       "step": 2150
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.624804973602295,
-      "learning_rate": 1.8874576271186443e-05,
-      "loss": 1.3399,
+      "grad_norm": 4.7201828956604,
+      "learning_rate": 1.2622222222222222e-05,
+      "loss": 1.1393,
       "step": 2160
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.085561752319336,
-      "learning_rate": 1.8867796610169492e-05,
-      "loss": 1.5298,
+      "grad_norm": 5.082671642303467,
+      "learning_rate": 1.257777777777778e-05,
+      "loss": 1.3047,
       "step": 2170
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.3589959144592285,
-      "learning_rate": 1.8861016949152545e-05,
-      "loss": 1.2609,
+      "grad_norm": 3.5268049240112305,
+      "learning_rate": 1.2533333333333336e-05,
+      "loss": 1.1255,
       "step": 2180
     },
     {
       "epoch": 0.05,
-      "grad_norm": 5.165348529815674,
-      "learning_rate": 1.8854237288135595e-05,
-      "loss": 1.5708,
+      "grad_norm": 3.8856544494628906,
+      "learning_rate": 1.2488888888888891e-05,
+      "loss": 1.4022,
       "step": 2190
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.508254051208496,
-      "learning_rate": 1.8847457627118645e-05,
-      "loss": 1.5046,
+      "grad_norm": 5.552175521850586,
+      "learning_rate": 1.2444444444444446e-05,
+      "loss": 1.3444,
       "step": 2200
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.811107635498047,
-      "learning_rate": 1.8840677966101698e-05,
-      "loss": 1.5798,
+      "grad_norm": 2.772660255432129,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 1.4274,
       "step": 2210
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.915175914764404,
-      "learning_rate": 1.8833898305084748e-05,
-      "loss": 1.3742,
+      "grad_norm": 6.9137773513793945,
+      "learning_rate": 1.2355555555555557e-05,
+      "loss": 1.1356,
       "step": 2220
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.212732791900635,
-      "learning_rate": 1.8827118644067797e-05,
-      "loss": 1.4748,
+      "grad_norm": 2.5068724155426025,
+      "learning_rate": 1.2311111111111112e-05,
+      "loss": 1.25,
       "step": 2230
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.4692771434783936,
-      "learning_rate": 1.882033898305085e-05,
-      "loss": 1.1476,
+      "grad_norm": 0.8587338924407959,
+      "learning_rate": 1.2266666666666667e-05,
+      "loss": 0.9182,
       "step": 2240
     },
     {
       "epoch": 0.05,
-      "grad_norm": 3.057826042175293,
-      "learning_rate": 1.88135593220339e-05,
-      "loss": 1.4235,
+      "grad_norm": 5.356514930725098,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 1.2882,
       "step": 2250
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.5246427059173584,
-      "learning_rate": 1.8806779661016953e-05,
-      "loss": 1.4705,
+      "grad_norm": 8.728981018066406,
+      "learning_rate": 1.217777777777778e-05,
+      "loss": 1.2974,
       "step": 2260
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.8235859870910645,
-      "learning_rate": 1.88e-05,
-      "loss": 1.4017,
+      "grad_norm": 2.5229454040527344,
+      "learning_rate": 1.2133333333333335e-05,
+      "loss": 1.2368,
       "step": 2270
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.3737335205078125,
-      "learning_rate": 1.8793220338983053e-05,
-      "loss": 1.2918,
+      "grad_norm": 2.760233163833618,
+      "learning_rate": 1.208888888888889e-05,
+      "loss": 1.0504,
       "step": 2280
     },
     {
       "epoch": 0.05,
-      "grad_norm": 1.6644233465194702,
-      "learning_rate": 1.8786440677966102e-05,
-      "loss": 1.5225,
+      "grad_norm": 1.7355753183364868,
+      "learning_rate": 1.2044444444444445e-05,
+      "loss": 1.3547,
       "step": 2290
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.259829044342041,
-      "learning_rate": 1.8779661016949152e-05,
-      "loss": 1.428,
+      "grad_norm": 2.2375967502593994,
+      "learning_rate": 1.2e-05,
+      "loss": 1.3058,
       "step": 2300
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.356344223022461,
-      "learning_rate": 1.8772881355932205e-05,
-      "loss": 1.4652,
+      "grad_norm": 4.5727386474609375,
+      "learning_rate": 1.1955555555555556e-05,
+      "loss": 1.3097,
       "step": 2310
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.862521171569824,
-      "learning_rate": 1.8766101694915255e-05,
-      "loss": 1.3431,
+      "grad_norm": 7.387458801269531,
+      "learning_rate": 1.191111111111111e-05,
+      "loss": 1.2058,
       "step": 2320
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.1847054958343506,
-      "learning_rate": 1.8759322033898305e-05,
-      "loss": 1.3383,
+      "grad_norm": 3.593174934387207,
+      "learning_rate": 1.186666666666667e-05,
+      "loss": 1.1711,
       "step": 2330
     },
     {
       "epoch": 0.05,
-      "grad_norm": 5.384062767028809,
-      "learning_rate": 1.8752542372881358e-05,
-      "loss": 1.3222,
+      "grad_norm": 3.3077690601348877,
+      "learning_rate": 1.1822222222222225e-05,
+      "loss": 1.1175,
       "step": 2340
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.504513740539551,
-      "learning_rate": 1.8745762711864407e-05,
-      "loss": 1.4033,
+      "grad_norm": 4.785140514373779,
+      "learning_rate": 1.177777777777778e-05,
+      "loss": 1.2618,
       "step": 2350
     },
     {
       "epoch": 0.05,
-      "grad_norm": 2.035881757736206,
-      "learning_rate": 1.873898305084746e-05,
-      "loss": 1.4297,
+      "grad_norm": 1.8957252502441406,
+      "learning_rate": 1.1733333333333335e-05,
+      "loss": 1.1641,
       "step": 2360
     },
     {
       "epoch": 0.05,
-      "grad_norm": 3.0843729972839355,
-      "learning_rate": 1.873220338983051e-05,
-      "loss": 1.5548,
+      "grad_norm": 7.033701419830322,
+      "learning_rate": 1.168888888888889e-05,
+      "loss": 1.3563,
       "step": 2370
     },
     {
       "epoch": 0.06,
-      "grad_norm": 4.0240912437438965,
-      "learning_rate": 1.872542372881356e-05,
-      "loss": 1.5772,
+      "grad_norm": 2.6566953659057617,
+      "learning_rate": 1.1644444444444446e-05,
+      "loss": 1.3673,
       "step": 2380
     },
     {
       "epoch": 0.06,
-      "grad_norm": 5.927667140960693,
-      "learning_rate": 1.8718644067796613e-05,
-      "loss": 1.471,
+      "grad_norm": 10.329083442687988,
+      "learning_rate": 1.16e-05,
+      "loss": 1.2953,
       "step": 2390
     },
     {
       "epoch": 0.06,
-      "grad_norm": 4.345820903778076,
-      "learning_rate": 1.8711864406779663e-05,
-      "loss": 1.5337,
+      "grad_norm": 3.5499980449676514,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 1.3136,
       "step": 2400
     },
     {
       "epoch": 0.06,
-      "grad_norm": 4.0480122566223145,
-      "learning_rate": 1.8705084745762712e-05,
-      "loss": 1.5327,
+      "grad_norm": 2.051661968231201,
+      "learning_rate": 1.1511111111111113e-05,
+      "loss": 1.3235,
       "step": 2410
     },
     {
       "epoch": 0.06,
-      "grad_norm": 2.2904655933380127,
-      "learning_rate": 1.8698305084745765e-05,
-      "loss": 1.3821,
+      "grad_norm": 2.9961462020874023,
+      "learning_rate": 1.1466666666666668e-05,
+      "loss": 1.1637,
       "step": 2420
     },
     {
       "epoch": 0.06,
-      "grad_norm": 5.9938225746154785,
-      "learning_rate": 1.8691525423728815e-05,
-      "loss": 1.2987,
+      "grad_norm": 8.53834056854248,
+      "learning_rate": 1.1422222222222223e-05,
+      "loss": 1.1066,
       "step": 2430
     },
     {
       "epoch": 0.06,
-      "grad_norm": 2.0605130195617676,
-      "learning_rate": 1.8684745762711868e-05,
-      "loss": 1.7052,
+      "grad_norm": 5.684910297393799,
+      "learning_rate": 1.1377777777777779e-05,
+      "loss": 1.4476,
       "step": 2440
     },
     {
       "epoch": 0.06,
-      "grad_norm": 1.4659353494644165,
-      "learning_rate": 1.8677966101694918e-05,
-      "loss": 1.5719,
+      "grad_norm": 16.79145622253418,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 1.3978,
       "step": 2450
     },
     {
       "epoch": 0.06,
-      "grad_norm": 1.8421217203140259,
-      "learning_rate": 1.8671186440677968e-05,
-      "loss": 1.3397,
+      "grad_norm": 3.3043346405029297,
+      "learning_rate": 1.1288888888888889e-05,
+      "loss": 1.1598,
       "step": 2460
     },
     {
       "epoch": 0.06,
-      "grad_norm": 1.0998902320861816,
-      "learning_rate": 1.866440677966102e-05,
-      "loss": 1.4649,
+      "grad_norm": 3.7516391277313232,
+      "learning_rate": 1.1244444444444444e-05,
+      "loss": 1.3534,
       "step": 2470
     },
     {
       "epoch": 0.06,
-      "grad_norm": 3.088279962539673,
-      "learning_rate": 1.8657627118644067e-05,
-      "loss": 1.4717,
+      "grad_norm": 2.0643539428710938,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.2924,
       "step": 2480
     },
     {
       "epoch": 0.06,
-      "grad_norm": 8.840229034423828,
-      "learning_rate": 1.865084745762712e-05,
-      "loss": 1.4184,
+      "grad_norm": 4.207937717437744,
+      "learning_rate": 1.1155555555555556e-05,
+      "loss": 1.2699,
       "step": 2490
     },
     {
       "epoch": 0.06,
-      "grad_norm": 2.5428450107574463,
-      "learning_rate": 1.864406779661017e-05,
-      "loss": 1.5584,
+      "grad_norm": 3.3195173740386963,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 1.379,
       "step": 2500
     },
     {
       "epoch": 0.06,
-      "eval_loss": 1.1482951641082764,
-      "eval_runtime": 67.0978,
-      "eval_samples_per_second": 14.904,
-      "eval_steps_per_second": 14.904,
+      "eval_loss": 0.986103355884552,
+      "eval_runtime": 67.0933,
+      "eval_samples_per_second": 14.905,
+      "eval_steps_per_second": 14.905,
       "step": 2500
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 5.455661773681641,
-      "learning_rate": 1.863728813559322e-05,
-      "loss": 1.5017,
-      "step": 2510
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 4.0317816734313965,
-      "learning_rate": 1.8630508474576273e-05,
-      "loss": 1.5774,
-      "step": 2520
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.15238618850708,
-      "learning_rate": 1.8623728813559322e-05,
-      "loss": 1.3854,
-      "step": 2530
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 2.9443018436431885,
-      "learning_rate": 1.8616949152542376e-05,
-      "loss": 1.5316,
-      "step": 2540
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.2570126056671143,
-      "learning_rate": 1.8610169491525425e-05,
-      "loss": 1.4141,
-      "step": 2550
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.6843924522399902,
-      "learning_rate": 1.8603389830508475e-05,
-      "loss": 1.4986,
-      "step": 2560
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.697768211364746,
-      "learning_rate": 1.8596610169491528e-05,
-      "loss": 1.3707,
-      "step": 2570
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 5.394255638122559,
-      "learning_rate": 1.8589830508474578e-05,
-      "loss": 1.4716,
-      "step": 2580
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 5.255263805389404,
-      "learning_rate": 1.8583050847457627e-05,
-      "loss": 1.5351,
-      "step": 2590
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 4.587831974029541,
-      "learning_rate": 1.857627118644068e-05,
-      "loss": 1.3261,
-      "step": 2600
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 8.405659675598145,
-      "learning_rate": 1.856949152542373e-05,
-      "loss": 1.4404,
-      "step": 2610
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 4.0395121574401855,
-      "learning_rate": 1.856271186440678e-05,
-      "loss": 1.368,
-      "step": 2620
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 6.0600972175598145,
-      "learning_rate": 1.8555932203389833e-05,
-      "loss": 1.4306,
-      "step": 2630
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.951476573944092,
-      "learning_rate": 1.8549152542372883e-05,
-      "loss": 1.4079,
-      "step": 2640
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 2.6316113471984863,
-      "learning_rate": 1.8542372881355936e-05,
-      "loss": 1.4675,
-      "step": 2650
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 2.2290899753570557,
-      "learning_rate": 1.8535593220338986e-05,
-      "loss": 1.462,
-      "step": 2660
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.8435947895050049,
-      "learning_rate": 1.8528813559322035e-05,
-      "loss": 1.0381,
-      "step": 2670
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 2.033623456954956,
-      "learning_rate": 1.852203389830509e-05,
-      "loss": 1.5768,
-      "step": 2680
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 2.9724245071411133,
-      "learning_rate": 1.8515254237288135e-05,
-      "loss": 1.2659,
-      "step": 2690
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.107775926589966,
-      "learning_rate": 1.8508474576271188e-05,
-      "loss": 1.1751,
-      "step": 2700
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 4.55193567276001,
-      "learning_rate": 1.8501694915254237e-05,
-      "loss": 1.5424,
-      "step": 2710
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.6639865636825562,
-      "learning_rate": 1.849491525423729e-05,
-      "loss": 1.391,
-      "step": 2720
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.7521861791610718,
-      "learning_rate": 1.848813559322034e-05,
-      "loss": 1.5969,
-      "step": 2730
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.5757704973220825,
-      "learning_rate": 1.848135593220339e-05,
-      "loss": 1.4257,
-      "step": 2740
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.4172585010528564,
-      "learning_rate": 1.8474576271186443e-05,
-      "loss": 1.4596,
-      "step": 2750
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.5575098991394043,
-      "learning_rate": 1.8467796610169493e-05,
-      "loss": 1.5946,
-      "step": 2760
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.385815382003784,
-      "learning_rate": 1.8461016949152542e-05,
-      "loss": 1.3725,
-      "step": 2770
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 3.108947992324829,
-      "learning_rate": 1.8454237288135596e-05,
-      "loss": 1.3995,
-      "step": 2780
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.180427074432373,
-      "learning_rate": 1.8447457627118645e-05,
-      "loss": 1.7337,
-      "step": 2790
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 1.628450632095337,
-      "learning_rate": 1.8440677966101695e-05,
-      "loss": 1.3329,
-      "step": 2800
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.4316961765289307,
-      "learning_rate": 1.8433898305084748e-05,
-      "loss": 1.5717,
-      "step": 2810
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.428252100944519,
-      "learning_rate": 1.8427118644067798e-05,
-      "loss": 1.3912,
-      "step": 2820
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.6684863567352295,
-      "learning_rate": 1.842033898305085e-05,
-      "loss": 1.4674,
-      "step": 2830
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.7176380157470703,
-      "learning_rate": 1.84135593220339e-05,
-      "loss": 1.2351,
-      "step": 2840
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.1771388053894043,
-      "learning_rate": 1.840677966101695e-05,
-      "loss": 1.6447,
-      "step": 2850
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 4.724181652069092,
-      "learning_rate": 1.8400000000000003e-05,
-      "loss": 1.4167,
-      "step": 2860
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.6035029888153076,
-      "learning_rate": 1.8393220338983053e-05,
-      "loss": 1.3391,
-      "step": 2870
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.2763707637786865,
-      "learning_rate": 1.8386440677966103e-05,
-      "loss": 1.5107,
-      "step": 2880
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.8805983066558838,
-      "learning_rate": 1.8379661016949153e-05,
-      "loss": 1.6482,
-      "step": 2890
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.5126351118087769,
-      "learning_rate": 1.8372881355932202e-05,
-      "loss": 1.4343,
-      "step": 2900
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 7.735413074493408,
-      "learning_rate": 1.8366101694915255e-05,
-      "loss": 1.2867,
-      "step": 2910
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 3.7856109142303467,
-      "learning_rate": 1.8359322033898305e-05,
-      "loss": 1.3506,
-      "step": 2920
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.2521424293518066,
-      "learning_rate": 1.8352542372881358e-05,
-      "loss": 1.3935,
-      "step": 2930
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.2285022735595703,
-      "learning_rate": 1.8345762711864408e-05,
-      "loss": 1.5142,
-      "step": 2940
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.304668664932251,
-      "learning_rate": 1.8338983050847458e-05,
-      "loss": 1.5006,
-      "step": 2950
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.3826477527618408,
-      "learning_rate": 1.833220338983051e-05,
-      "loss": 1.1512,
-      "step": 2960
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.6900218725204468,
-      "learning_rate": 1.832542372881356e-05,
-      "loss": 1.4017,
-      "step": 2970
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 3.0282979011535645,
-      "learning_rate": 1.831864406779661e-05,
-      "loss": 1.3722,
-      "step": 2980
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 4.85763692855835,
-      "learning_rate": 1.8311864406779663e-05,
-      "loss": 1.3328,
-      "step": 2990
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.2919251918792725,
-      "learning_rate": 1.8305084745762713e-05,
-      "loss": 1.4177,
-      "step": 3000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 1.2010964155197144,
-      "eval_runtime": 67.1519,
-      "eval_samples_per_second": 14.892,
-      "eval_steps_per_second": 14.892,
-      "step": 3000
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.3012161254882812,
-      "learning_rate": 1.8298305084745766e-05,
-      "loss": 1.2741,
-      "step": 3010
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.728154182434082,
-      "learning_rate": 1.8291525423728816e-05,
-      "loss": 1.495,
-      "step": 3020
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.277280569076538,
-      "learning_rate": 1.8284745762711865e-05,
-      "loss": 1.4877,
-      "step": 3030
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.6222431659698486,
-      "learning_rate": 1.827796610169492e-05,
-      "loss": 1.3664,
-      "step": 3040
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.7156178951263428,
-      "learning_rate": 1.8271186440677968e-05,
-      "loss": 1.3912,
-      "step": 3050
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 0.9766186475753784,
-      "learning_rate": 1.8264406779661018e-05,
-      "loss": 1.3334,
-      "step": 3060
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 11.627836227416992,
-      "learning_rate": 1.825762711864407e-05,
-      "loss": 1.6583,
-      "step": 3070
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.1804165840148926,
-      "learning_rate": 1.825084745762712e-05,
-      "loss": 1.2931,
-      "step": 3080
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 12.77298641204834,
-      "learning_rate": 1.824406779661017e-05,
-      "loss": 1.3071,
-      "step": 3090
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 3.189802646636963,
-      "learning_rate": 1.823728813559322e-05,
-      "loss": 1.508,
-      "step": 3100
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.3131775856018066,
-      "learning_rate": 1.8230508474576273e-05,
-      "loss": 1.4604,
-      "step": 3110
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 4.224339962005615,
-      "learning_rate": 1.8223728813559323e-05,
-      "loss": 1.2131,
-      "step": 3120
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 4.142259120941162,
-      "learning_rate": 1.8216949152542373e-05,
-      "loss": 1.409,
-      "step": 3130
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.409524917602539,
-      "learning_rate": 1.8210169491525426e-05,
-      "loss": 1.5591,
-      "step": 3140
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.3798465728759766,
-      "learning_rate": 1.8203389830508475e-05,
-      "loss": 1.4358,
-      "step": 3150
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 0.9663469195365906,
-      "learning_rate": 1.8196610169491525e-05,
-      "loss": 1.4432,
-      "step": 3160
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.4938082695007324,
-      "learning_rate": 1.8189830508474578e-05,
-      "loss": 1.2837,
-      "step": 3170
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.094170093536377,
-      "learning_rate": 1.8183050847457628e-05,
-      "loss": 1.3925,
-      "step": 3180
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 4.124843120574951,
-      "learning_rate": 1.817627118644068e-05,
-      "loss": 1.5111,
-      "step": 3190
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 1.090320348739624,
-      "learning_rate": 1.816949152542373e-05,
-      "loss": 1.4229,
-      "step": 3200
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 2.482816696166992,
-      "learning_rate": 1.816271186440678e-05,
-      "loss": 1.3102,
-      "step": 3210
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 3.5619702339172363,
-      "learning_rate": 1.8155932203389833e-05,
-      "loss": 1.2051,
-      "step": 3220
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 7.419604301452637,
-      "learning_rate": 1.8149152542372883e-05,
-      "loss": 1.3915,
-      "step": 3230
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 4.109569072723389,
-      "learning_rate": 1.8142372881355933e-05,
-      "loss": 1.4158,
-      "step": 3240
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.4256246089935303,
-      "learning_rate": 1.8135593220338986e-05,
-      "loss": 1.4494,
-      "step": 3250
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.4669333696365356,
-      "learning_rate": 1.8128813559322036e-05,
-      "loss": 1.6379,
-      "step": 3260
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.800262689590454,
-      "learning_rate": 1.8122033898305085e-05,
-      "loss": 1.5004,
-      "step": 3270
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.3720695972442627,
-      "learning_rate": 1.811525423728814e-05,
-      "loss": 1.425,
-      "step": 3280
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.8338054418563843,
-      "learning_rate": 1.8108474576271188e-05,
-      "loss": 1.4227,
-      "step": 3290
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 4.969144821166992,
-      "learning_rate": 1.810169491525424e-05,
-      "loss": 1.3806,
-      "step": 3300
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.558366537094116,
-      "learning_rate": 1.8094915254237288e-05,
-      "loss": 1.4254,
-      "step": 3310
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.669677972793579,
-      "learning_rate": 1.808813559322034e-05,
-      "loss": 1.4065,
-      "step": 3320
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.3998602628707886,
-      "learning_rate": 1.808135593220339e-05,
-      "loss": 1.7619,
-      "step": 3330
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.9056684970855713,
-      "learning_rate": 1.807457627118644e-05,
-      "loss": 1.4177,
-      "step": 3340
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.942318916320801,
-      "learning_rate": 1.8067796610169493e-05,
-      "loss": 1.4779,
-      "step": 3350
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.1195051670074463,
-      "learning_rate": 1.8061016949152543e-05,
-      "loss": 1.4865,
-      "step": 3360
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 4.495617866516113,
-      "learning_rate": 1.8054237288135593e-05,
-      "loss": 1.3127,
-      "step": 3370
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.5328311920166016,
-      "learning_rate": 1.8047457627118646e-05,
-      "loss": 1.4624,
-      "step": 3380
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.0216922760009766,
-      "learning_rate": 1.8040677966101695e-05,
-      "loss": 1.4197,
-      "step": 3390
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.578646183013916,
-      "learning_rate": 1.803389830508475e-05,
-      "loss": 1.2317,
-      "step": 3400
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.8747899532318115,
-      "learning_rate": 1.8027118644067798e-05,
-      "loss": 1.5859,
-      "step": 3410
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 4.799411773681641,
-      "learning_rate": 1.8020338983050848e-05,
-      "loss": 1.5663,
-      "step": 3420
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.006511688232422,
-      "learning_rate": 1.80135593220339e-05,
-      "loss": 1.4501,
-      "step": 3430
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.5541688203811646,
-      "learning_rate": 1.800677966101695e-05,
-      "loss": 1.6072,
-      "step": 3440
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.2471415996551514,
-      "learning_rate": 1.8e-05,
-      "loss": 1.4034,
-      "step": 3450
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.43544602394104,
-      "learning_rate": 1.7993220338983054e-05,
-      "loss": 1.3181,
-      "step": 3460
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 6.56860876083374,
-      "learning_rate": 1.7986440677966103e-05,
-      "loss": 1.308,
-      "step": 3470
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.6072529554367065,
-      "learning_rate": 1.7979661016949156e-05,
-      "loss": 1.3989,
-      "step": 3480
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.9302940368652344,
-      "learning_rate": 1.7972881355932206e-05,
-      "loss": 1.5623,
-      "step": 3490
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 5.61619758605957,
-      "learning_rate": 1.7966101694915256e-05,
-      "loss": 1.4135,
-      "step": 3500
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 1.1386640071868896,
-      "eval_runtime": 67.1064,
-      "eval_samples_per_second": 14.902,
-      "eval_steps_per_second": 14.902,
-      "step": 3500
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.315586566925049,
-      "learning_rate": 1.795932203389831e-05,
-      "loss": 1.5134,
-      "step": 3510
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 6.156167507171631,
-      "learning_rate": 1.7952542372881355e-05,
-      "loss": 1.4425,
-      "step": 3520
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.9159029722213745,
-      "learning_rate": 1.7945762711864408e-05,
-      "loss": 1.3159,
-      "step": 3530
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.3781180381774902,
-      "learning_rate": 1.7938983050847458e-05,
-      "loss": 1.3683,
-      "step": 3540
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.8861943483352661,
-      "learning_rate": 1.7932203389830508e-05,
-      "loss": 1.486,
-      "step": 3550
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.9574244022369385,
-      "learning_rate": 1.792542372881356e-05,
-      "loss": 1.4858,
-      "step": 3560
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 5.585728168487549,
-      "learning_rate": 1.791864406779661e-05,
-      "loss": 1.4388,
-      "step": 3570
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 6.258697986602783,
-      "learning_rate": 1.7911864406779664e-05,
-      "loss": 1.5302,
-      "step": 3580
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 4.292418479919434,
-      "learning_rate": 1.7905084745762713e-05,
-      "loss": 1.3228,
-      "step": 3590
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 4.342001438140869,
-      "learning_rate": 1.7898305084745763e-05,
-      "loss": 1.1876,
-      "step": 3600
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 7.243982315063477,
-      "learning_rate": 1.7891525423728816e-05,
-      "loss": 1.4241,
-      "step": 3610
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 8.561788558959961,
-      "learning_rate": 1.7884745762711866e-05,
-      "loss": 1.4567,
-      "step": 3620
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 2.6268651485443115,
-      "learning_rate": 1.7877966101694916e-05,
-      "loss": 1.2928,
-      "step": 3630
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.3656156063079834,
-      "learning_rate": 1.787118644067797e-05,
-      "loss": 1.2128,
-      "step": 3640
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 3.3748042583465576,
-      "learning_rate": 1.7864406779661018e-05,
-      "loss": 1.472,
-      "step": 3650
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 4.713076591491699,
-      "learning_rate": 1.785762711864407e-05,
-      "loss": 1.4151,
-      "step": 3660
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 1.744603157043457,
-      "learning_rate": 1.785084745762712e-05,
-      "loss": 1.3556,
-      "step": 3670
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 5.440206050872803,
-      "learning_rate": 1.784406779661017e-05,
-      "loss": 1.5536,
-      "step": 3680
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.1328643560409546,
-      "learning_rate": 1.7837288135593224e-05,
-      "loss": 1.3039,
-      "step": 3690
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.336589813232422,
-      "learning_rate": 1.7830508474576274e-05,
-      "loss": 1.4308,
-      "step": 3700
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.2741591930389404,
-      "learning_rate": 1.7823728813559323e-05,
-      "loss": 1.3202,
-      "step": 3710
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 4.327836990356445,
-      "learning_rate": 1.7816949152542376e-05,
-      "loss": 1.5221,
-      "step": 3720
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.930452823638916,
-      "learning_rate": 1.7810169491525423e-05,
-      "loss": 1.3796,
-      "step": 3730
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 4.003491401672363,
-      "learning_rate": 1.7803389830508476e-05,
-      "loss": 1.5096,
-      "step": 3740
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.0040380954742432,
-      "learning_rate": 1.7796610169491526e-05,
-      "loss": 1.5853,
-      "step": 3750
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.6716127395629883,
-      "learning_rate": 1.778983050847458e-05,
-      "loss": 1.4564,
-      "step": 3760
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 5.1395158767700195,
-      "learning_rate": 1.778305084745763e-05,
-      "loss": 1.3919,
-      "step": 3770
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.6375240087509155,
-      "learning_rate": 1.7776271186440678e-05,
-      "loss": 1.2678,
-      "step": 3780
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 11.120715141296387,
-      "learning_rate": 1.776949152542373e-05,
-      "loss": 1.5301,
-      "step": 3790
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.9304801225662231,
-      "learning_rate": 1.776271186440678e-05,
-      "loss": 1.3043,
-      "step": 3800
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 3.231403112411499,
-      "learning_rate": 1.775593220338983e-05,
-      "loss": 1.3634,
-      "step": 3810
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.287160873413086,
-      "learning_rate": 1.7749152542372884e-05,
-      "loss": 1.306,
-      "step": 3820
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 4.768348217010498,
-      "learning_rate": 1.7742372881355933e-05,
-      "loss": 1.479,
-      "step": 3830
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 3.244331121444702,
-      "learning_rate": 1.7735593220338983e-05,
-      "loss": 1.2446,
-      "step": 3840
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 3.9224987030029297,
-      "learning_rate": 1.7728813559322036e-05,
-      "loss": 1.4043,
-      "step": 3850
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 3.9674978256225586,
-      "learning_rate": 1.7722033898305086e-05,
-      "loss": 1.3901,
-      "step": 3860
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 7.087489604949951,
-      "learning_rate": 1.771525423728814e-05,
-      "loss": 1.4567,
-      "step": 3870
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.8183858394622803,
-      "learning_rate": 1.770847457627119e-05,
-      "loss": 1.6078,
-      "step": 3880
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.5410759449005127,
-      "learning_rate": 1.770169491525424e-05,
-      "loss": 1.3961,
-      "step": 3890
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.4553983211517334,
-      "learning_rate": 1.769491525423729e-05,
-      "loss": 1.405,
-      "step": 3900
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 5.181529998779297,
-      "learning_rate": 1.768813559322034e-05,
-      "loss": 1.4944,
-      "step": 3910
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.054699420928955,
-      "learning_rate": 1.768135593220339e-05,
-      "loss": 1.5309,
-      "step": 3920
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 3.4803507328033447,
-      "learning_rate": 1.767457627118644e-05,
-      "loss": 1.2931,
-      "step": 3930
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.9554686546325684,
-      "learning_rate": 1.7667796610169494e-05,
-      "loss": 1.3651,
-      "step": 3940
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.0756142139434814,
-      "learning_rate": 1.7661016949152543e-05,
-      "loss": 1.4476,
-      "step": 3950
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.552859306335449,
-      "learning_rate": 1.7654237288135593e-05,
-      "loss": 1.4417,
-      "step": 3960
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.7441787719726562,
-      "learning_rate": 1.7647457627118646e-05,
-      "loss": 1.3769,
-      "step": 3970
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.9613826274871826,
-      "learning_rate": 1.7640677966101696e-05,
-      "loss": 1.4381,
-      "step": 3980
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 4.84116268157959,
-      "learning_rate": 1.7633898305084746e-05,
-      "loss": 1.3539,
-      "step": 3990
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.5078232288360596,
-      "learning_rate": 1.76271186440678e-05,
-      "loss": 1.4277,
-      "step": 4000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 1.158501386642456,
-      "eval_runtime": 67.1002,
-      "eval_samples_per_second": 14.903,
-      "eval_steps_per_second": 14.903,
-      "step": 4000
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.6108415126800537,
-      "learning_rate": 1.762033898305085e-05,
-      "loss": 1.5502,
-      "step": 4010
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 4.075037479400635,
-      "learning_rate": 1.7613559322033898e-05,
-      "loss": 1.3346,
-      "step": 4020
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.969217300415039,
-      "learning_rate": 1.760677966101695e-05,
-      "loss": 1.3709,
-      "step": 4030
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 3.885554313659668,
-      "learning_rate": 1.76e-05,
-      "loss": 1.4961,
-      "step": 4040
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.5712426900863647,
-      "learning_rate": 1.7593220338983054e-05,
-      "loss": 1.3091,
-      "step": 4050
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.020577907562256,
-      "learning_rate": 1.7586440677966104e-05,
-      "loss": 1.3308,
-      "step": 4060
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 4.234301567077637,
-      "learning_rate": 1.7579661016949153e-05,
-      "loss": 1.5322,
-      "step": 4070
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.667493224143982,
-      "learning_rate": 1.7572881355932206e-05,
-      "loss": 1.3986,
-      "step": 4080
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 1.5255711078643799,
-      "learning_rate": 1.7566101694915256e-05,
-      "loss": 1.1095,
-      "step": 4090
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 2.342149257659912,
-      "learning_rate": 1.7559322033898306e-05,
-      "loss": 1.468,
-      "step": 4100
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.444767951965332,
-      "learning_rate": 1.755254237288136e-05,
-      "loss": 1.3548,
-      "step": 4110
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.0781197547912598,
-      "learning_rate": 1.754576271186441e-05,
-      "loss": 1.4868,
-      "step": 4120
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.5227110385894775,
-      "learning_rate": 1.753898305084746e-05,
-      "loss": 1.2477,
-      "step": 4130
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.230503797531128,
-      "learning_rate": 1.7532203389830508e-05,
-      "loss": 1.5486,
-      "step": 4140
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.484776496887207,
-      "learning_rate": 1.752542372881356e-05,
-      "loss": 1.4854,
-      "step": 4150
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.312047004699707,
-      "learning_rate": 1.751864406779661e-05,
-      "loss": 1.45,
-      "step": 4160
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 4.45994758605957,
-      "learning_rate": 1.751186440677966e-05,
-      "loss": 1.2472,
-      "step": 4170
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 1.4740896224975586,
-      "learning_rate": 1.7505084745762714e-05,
-      "loss": 1.1978,
-      "step": 4180
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.2780067920684814,
-      "learning_rate": 1.7498305084745763e-05,
-      "loss": 1.4335,
-      "step": 4190
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.937509298324585,
-      "learning_rate": 1.7491525423728813e-05,
-      "loss": 1.4036,
-      "step": 4200
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 5.987236022949219,
-      "learning_rate": 1.7484745762711866e-05,
-      "loss": 1.3751,
-      "step": 4210
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 1.929296851158142,
-      "learning_rate": 1.7477966101694916e-05,
-      "loss": 1.294,
-      "step": 4220
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.1039040088653564,
-      "learning_rate": 1.747118644067797e-05,
-      "loss": 1.3019,
-      "step": 4230
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 6.1424946784973145,
-      "learning_rate": 1.746440677966102e-05,
-      "loss": 1.2913,
-      "step": 4240
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 0.9339830279350281,
-      "learning_rate": 1.745762711864407e-05,
-      "loss": 1.3142,
-      "step": 4250
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.47251296043396,
-      "learning_rate": 1.745084745762712e-05,
-      "loss": 1.4085,
-      "step": 4260
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 1.0523934364318848,
-      "learning_rate": 1.744406779661017e-05,
-      "loss": 1.3264,
-      "step": 4270
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.680063009262085,
-      "learning_rate": 1.743728813559322e-05,
-      "loss": 1.5524,
-      "step": 4280
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 4.143898010253906,
-      "learning_rate": 1.7430508474576274e-05,
-      "loss": 1.5795,
-      "step": 4290
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.3456549644470215,
-      "learning_rate": 1.7423728813559324e-05,
-      "loss": 1.3907,
-      "step": 4300
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.776109457015991,
-      "learning_rate": 1.7416949152542373e-05,
-      "loss": 1.2968,
-      "step": 4310
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 4.241484642028809,
-      "learning_rate": 1.7410169491525427e-05,
-      "loss": 1.305,
-      "step": 4320
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.8671717643737793,
-      "learning_rate": 1.7403389830508476e-05,
-      "loss": 1.5132,
-      "step": 4330
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.0062782764434814,
-      "learning_rate": 1.739661016949153e-05,
-      "loss": 1.2456,
-      "step": 4340
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 4.044645309448242,
-      "learning_rate": 1.7389830508474576e-05,
-      "loss": 1.4838,
-      "step": 4350
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.291123867034912,
-      "learning_rate": 1.738305084745763e-05,
-      "loss": 1.4495,
-      "step": 4360
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.096938133239746,
-      "learning_rate": 1.737627118644068e-05,
-      "loss": 1.2715,
-      "step": 4370
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.5189833641052246,
-      "learning_rate": 1.7369491525423728e-05,
-      "loss": 1.1358,
-      "step": 4380
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.956369638442993,
-      "learning_rate": 1.736271186440678e-05,
-      "loss": 1.1939,
-      "step": 4390
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 8.631299018859863,
-      "learning_rate": 1.735593220338983e-05,
-      "loss": 1.5566,
-      "step": 4400
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 1.6200116872787476,
-      "learning_rate": 1.734915254237288e-05,
-      "loss": 1.5092,
-      "step": 4410
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 4.132010459899902,
-      "learning_rate": 1.7342372881355934e-05,
-      "loss": 1.4241,
-      "step": 4420
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.7074265480041504,
-      "learning_rate": 1.7335593220338983e-05,
-      "loss": 1.4407,
-      "step": 4430
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.0132627487182617,
-      "learning_rate": 1.7328813559322037e-05,
-      "loss": 1.3255,
-      "step": 4440
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.750293254852295,
-      "learning_rate": 1.7322033898305086e-05,
-      "loss": 1.5557,
-      "step": 4450
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 1.0939888954162598,
-      "learning_rate": 1.7315254237288136e-05,
-      "loss": 1.3632,
-      "step": 4460
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 2.476027727127075,
-      "learning_rate": 1.730847457627119e-05,
-      "loss": 1.3655,
-      "step": 4470
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.4436848163604736,
-      "learning_rate": 1.730169491525424e-05,
-      "loss": 1.3026,
-      "step": 4480
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 5.511134147644043,
-      "learning_rate": 1.729491525423729e-05,
-      "loss": 1.5307,
-      "step": 4490
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 1.5898433923721313,
-      "learning_rate": 1.728813559322034e-05,
-      "loss": 1.4483,
-      "step": 4500
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 1.1334296464920044,
-      "eval_runtime": 67.0278,
-      "eval_samples_per_second": 14.919,
-      "eval_steps_per_second": 14.919,
-      "step": 4500
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 1.7626523971557617,
-      "learning_rate": 1.728135593220339e-05,
-      "loss": 1.3683,
-      "step": 4510
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.1556930541992188,
-      "learning_rate": 1.7274576271186444e-05,
-      "loss": 1.4022,
-      "step": 4520
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 3.2186577320098877,
-      "learning_rate": 1.7267796610169494e-05,
-      "loss": 1.41,
-      "step": 4530
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 6.2501068115234375,
-      "learning_rate": 1.7261016949152544e-05,
-      "loss": 1.4536,
-      "step": 4540
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 4.240054130554199,
-      "learning_rate": 1.7254237288135597e-05,
-      "loss": 1.4073,
-      "step": 4550
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.387373208999634,
-      "learning_rate": 1.7247457627118643e-05,
-      "loss": 1.613,
-      "step": 4560
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 4.530575752258301,
-      "learning_rate": 1.7240677966101696e-05,
-      "loss": 1.2231,
-      "step": 4570
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.739079236984253,
-      "learning_rate": 1.7233898305084746e-05,
-      "loss": 1.4286,
-      "step": 4580
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 4.380158424377441,
-      "learning_rate": 1.7227118644067796e-05,
-      "loss": 1.3608,
-      "step": 4590
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 5.204037189483643,
-      "learning_rate": 1.722033898305085e-05,
-      "loss": 1.2801,
-      "step": 4600
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.6450468301773071,
-      "learning_rate": 1.72135593220339e-05,
-      "loss": 1.29,
-      "step": 4610
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.5391757488250732,
-      "learning_rate": 1.720677966101695e-05,
-      "loss": 1.6189,
-      "step": 4620
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.872121810913086,
-      "learning_rate": 1.72e-05,
-      "loss": 1.2915,
-      "step": 4630
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.8644561767578125,
-      "learning_rate": 1.719322033898305e-05,
-      "loss": 1.5074,
-      "step": 4640
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.774881601333618,
-      "learning_rate": 1.7186440677966104e-05,
-      "loss": 1.3268,
-      "step": 4650
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.5650038719177246,
-      "learning_rate": 1.7179661016949154e-05,
-      "loss": 1.3892,
-      "step": 4660
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.5609863996505737,
-      "learning_rate": 1.7172881355932204e-05,
-      "loss": 1.3675,
-      "step": 4670
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 8.063512802124023,
-      "learning_rate": 1.7166101694915257e-05,
-      "loss": 1.3671,
-      "step": 4680
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.6324501037597656,
-      "learning_rate": 1.7159322033898306e-05,
-      "loss": 1.3284,
-      "step": 4690
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.6894187927246094,
-      "learning_rate": 1.715254237288136e-05,
-      "loss": 1.3032,
-      "step": 4700
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.1557717323303223,
-      "learning_rate": 1.714576271186441e-05,
-      "loss": 1.366,
-      "step": 4710
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.8717806339263916,
-      "learning_rate": 1.713898305084746e-05,
-      "loss": 1.5634,
-      "step": 4720
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.2488980293273926,
-      "learning_rate": 1.7132203389830512e-05,
-      "loss": 1.4814,
-      "step": 4730
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.677203893661499,
-      "learning_rate": 1.712542372881356e-05,
-      "loss": 1.2079,
-      "step": 4740
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.8588718175888062,
-      "learning_rate": 1.711864406779661e-05,
-      "loss": 1.3525,
-      "step": 4750
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.2451605796813965,
-      "learning_rate": 1.711186440677966e-05,
-      "loss": 1.4148,
-      "step": 4760
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.5242979526519775,
-      "learning_rate": 1.710508474576271e-05,
-      "loss": 1.2413,
-      "step": 4770
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.1980478763580322,
-      "learning_rate": 1.7098305084745764e-05,
-      "loss": 1.4558,
-      "step": 4780
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 6.478370666503906,
-      "learning_rate": 1.7091525423728814e-05,
-      "loss": 1.3639,
-      "step": 4790
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.1923305988311768,
-      "learning_rate": 1.7084745762711867e-05,
-      "loss": 1.5741,
-      "step": 4800
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.293412923812866,
-      "learning_rate": 1.7077966101694916e-05,
-      "loss": 1.2903,
-      "step": 4810
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.4575860500335693,
-      "learning_rate": 1.7071186440677966e-05,
-      "loss": 1.4585,
-      "step": 4820
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.9733614921569824,
-      "learning_rate": 1.706440677966102e-05,
-      "loss": 1.1767,
-      "step": 4830
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.9043350219726562,
-      "learning_rate": 1.705762711864407e-05,
-      "loss": 1.4218,
-      "step": 4840
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.5711846351623535,
-      "learning_rate": 1.705084745762712e-05,
-      "loss": 1.2186,
-      "step": 4850
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.8665926456451416,
-      "learning_rate": 1.704406779661017e-05,
-      "loss": 1.3747,
-      "step": 4860
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 3.0901689529418945,
-      "learning_rate": 1.703728813559322e-05,
-      "loss": 1.3356,
-      "step": 4870
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.0309970378875732,
-      "learning_rate": 1.703050847457627e-05,
-      "loss": 1.4572,
-      "step": 4880
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.9482085704803467,
-      "learning_rate": 1.7023728813559324e-05,
-      "loss": 1.3365,
-      "step": 4890
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.5880916118621826,
-      "learning_rate": 1.7016949152542374e-05,
-      "loss": 1.4305,
-      "step": 4900
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 5.12321138381958,
-      "learning_rate": 1.7010169491525427e-05,
-      "loss": 1.5018,
-      "step": 4910
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.508700370788574,
-      "learning_rate": 1.7003389830508477e-05,
-      "loss": 1.237,
-      "step": 4920
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.8924113512039185,
-      "learning_rate": 1.6996610169491526e-05,
-      "loss": 1.3583,
-      "step": 4930
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.551520824432373,
-      "learning_rate": 1.698983050847458e-05,
-      "loss": 1.2687,
-      "step": 4940
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 2.142587184906006,
-      "learning_rate": 1.698305084745763e-05,
-      "loss": 1.4377,
-      "step": 4950
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 1.7592850923538208,
-      "learning_rate": 1.697627118644068e-05,
-      "loss": 1.5202,
-      "step": 4960
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.212038040161133,
-      "learning_rate": 1.696949152542373e-05,
-      "loss": 1.3286,
-      "step": 4970
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.211546897888184,
-      "learning_rate": 1.6962711864406782e-05,
-      "loss": 1.5126,
-      "step": 4980
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 6.877481460571289,
-      "learning_rate": 1.695593220338983e-05,
-      "loss": 1.4129,
-      "step": 4990
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.20015811920166,
-      "learning_rate": 1.694915254237288e-05,
-      "loss": 1.2721,
-      "step": 5000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 1.1188498735427856,
-      "eval_runtime": 67.1151,
-      "eval_samples_per_second": 14.9,
-      "eval_steps_per_second": 14.9,
-      "step": 5000
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.303971290588379,
-      "learning_rate": 1.6942372881355934e-05,
-      "loss": 1.3757,
-      "step": 5010
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.5342512130737305,
-      "learning_rate": 1.6935593220338984e-05,
-      "loss": 1.2597,
-      "step": 5020
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.022207260131836,
-      "learning_rate": 1.6928813559322034e-05,
-      "loss": 1.4939,
-      "step": 5030
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.4725759029388428,
-      "learning_rate": 1.6922033898305087e-05,
-      "loss": 1.2841,
-      "step": 5040
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.944225788116455,
-      "learning_rate": 1.6915254237288136e-05,
-      "loss": 1.3808,
-      "step": 5050
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 1.0974010229110718,
-      "learning_rate": 1.6908474576271186e-05,
-      "loss": 1.2744,
-      "step": 5060
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.6920721530914307,
-      "learning_rate": 1.690169491525424e-05,
-      "loss": 1.5257,
-      "step": 5070
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.715402126312256,
-      "learning_rate": 1.689491525423729e-05,
-      "loss": 1.5197,
-      "step": 5080
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.5713796615600586,
-      "learning_rate": 1.6888135593220342e-05,
-      "loss": 1.3383,
-      "step": 5090
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.9672067165374756,
-      "learning_rate": 1.6881355932203392e-05,
-      "loss": 1.2505,
-      "step": 5100
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.107706069946289,
-      "learning_rate": 1.687457627118644e-05,
-      "loss": 1.3978,
-      "step": 5110
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.694401979446411,
-      "learning_rate": 1.6867796610169495e-05,
-      "loss": 1.5616,
-      "step": 5120
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 1.1937147378921509,
-      "learning_rate": 1.6861016949152544e-05,
-      "loss": 1.4247,
-      "step": 5130
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 1.7874356508255005,
-      "learning_rate": 1.6854237288135594e-05,
-      "loss": 1.1953,
-      "step": 5140
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.081228256225586,
-      "learning_rate": 1.6847457627118647e-05,
-      "loss": 1.3952,
-      "step": 5150
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.792527198791504,
-      "learning_rate": 1.6840677966101697e-05,
-      "loss": 1.3015,
-      "step": 5160
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.482060194015503,
-      "learning_rate": 1.683389830508475e-05,
-      "loss": 1.1953,
-      "step": 5170
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.224519729614258,
-      "learning_rate": 1.6827118644067796e-05,
-      "loss": 1.3729,
-      "step": 5180
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.609736680984497,
-      "learning_rate": 1.682033898305085e-05,
-      "loss": 1.4468,
-      "step": 5190
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.786303997039795,
-      "learning_rate": 1.68135593220339e-05,
-      "loss": 1.4248,
-      "step": 5200
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.584510326385498,
-      "learning_rate": 1.680677966101695e-05,
-      "loss": 1.4896,
-      "step": 5210
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 7.068190574645996,
-      "learning_rate": 1.6800000000000002e-05,
-      "loss": 1.3613,
-      "step": 5220
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.357131242752075,
-      "learning_rate": 1.679322033898305e-05,
-      "loss": 1.1566,
-      "step": 5230
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.547120571136475,
-      "learning_rate": 1.67864406779661e-05,
-      "loss": 1.3428,
-      "step": 5240
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.5816683769226074,
-      "learning_rate": 1.6779661016949154e-05,
-      "loss": 1.3112,
-      "step": 5250
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 4.817299842834473,
-      "learning_rate": 1.6772881355932204e-05,
-      "loss": 1.4183,
-      "step": 5260
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.906155586242676,
-      "learning_rate": 1.6766101694915257e-05,
-      "loss": 1.4456,
-      "step": 5270
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 8.34229850769043,
-      "learning_rate": 1.6759322033898307e-05,
-      "loss": 1.3624,
-      "step": 5280
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.7926371097564697,
-      "learning_rate": 1.6752542372881357e-05,
-      "loss": 1.4581,
-      "step": 5290
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.89165997505188,
-      "learning_rate": 1.674576271186441e-05,
-      "loss": 1.5302,
-      "step": 5300
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 1.7644487619400024,
-      "learning_rate": 1.673898305084746e-05,
-      "loss": 1.2788,
-      "step": 5310
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.700753927230835,
-      "learning_rate": 1.673220338983051e-05,
-      "loss": 1.4187,
-      "step": 5320
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.538086175918579,
-      "learning_rate": 1.6725423728813562e-05,
-      "loss": 1.4627,
-      "step": 5330
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.7824482917785645,
-      "learning_rate": 1.6718644067796612e-05,
-      "loss": 1.3147,
-      "step": 5340
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 3.426088571548462,
-      "learning_rate": 1.671186440677966e-05,
-      "loss": 1.4254,
-      "step": 5350
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.100320339202881,
-      "learning_rate": 1.6705084745762715e-05,
-      "loss": 1.2516,
-      "step": 5360
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.2726573944091797,
-      "learning_rate": 1.6698305084745764e-05,
-      "loss": 1.2826,
-      "step": 5370
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 5.182670593261719,
-      "learning_rate": 1.6691525423728817e-05,
-      "loss": 1.468,
-      "step": 5380
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 2.9416656494140625,
-      "learning_rate": 1.6684745762711864e-05,
-      "loss": 1.4455,
-      "step": 5390
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.5433812141418457,
-      "learning_rate": 1.6677966101694917e-05,
-      "loss": 1.4634,
-      "step": 5400
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 4.191746234893799,
-      "learning_rate": 1.6671186440677967e-05,
-      "loss": 1.3533,
-      "step": 5410
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.96868097782135,
-      "learning_rate": 1.6664406779661016e-05,
-      "loss": 1.3821,
-      "step": 5420
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 5.689413070678711,
-      "learning_rate": 1.665762711864407e-05,
-      "loss": 1.4411,
-      "step": 5430
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.58811354637146,
-      "learning_rate": 1.665084745762712e-05,
-      "loss": 1.3856,
-      "step": 5440
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.642380952835083,
-      "learning_rate": 1.6644067796610172e-05,
-      "loss": 1.4348,
-      "step": 5450
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.2928035259246826,
-      "learning_rate": 1.6637288135593222e-05,
-      "loss": 1.5305,
-      "step": 5460
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 5.261194705963135,
-      "learning_rate": 1.663050847457627e-05,
-      "loss": 1.2432,
-      "step": 5470
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.4704355001449585,
-      "learning_rate": 1.6623728813559325e-05,
-      "loss": 1.5543,
-      "step": 5480
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.8535382747650146,
-      "learning_rate": 1.6616949152542374e-05,
-      "loss": 1.3296,
-      "step": 5490
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.3717048168182373,
-      "learning_rate": 1.6610169491525424e-05,
-      "loss": 1.5835,
-      "step": 5500
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 1.0649362802505493,
-      "eval_runtime": 66.9134,
-      "eval_samples_per_second": 14.945,
-      "eval_steps_per_second": 14.945,
-      "step": 5500
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.3180487155914307,
-      "learning_rate": 1.6603389830508477e-05,
-      "loss": 1.3602,
-      "step": 5510
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.428536891937256,
-      "learning_rate": 1.6596610169491527e-05,
-      "loss": 1.4902,
-      "step": 5520
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.3657978773117065,
-      "learning_rate": 1.6589830508474577e-05,
-      "loss": 1.3019,
-      "step": 5530
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.775023341178894,
-      "learning_rate": 1.658305084745763e-05,
-      "loss": 1.309,
-      "step": 5540
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.047219753265381,
-      "learning_rate": 1.657627118644068e-05,
-      "loss": 1.463,
-      "step": 5550
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.6937634944915771,
-      "learning_rate": 1.6569491525423732e-05,
-      "loss": 1.4777,
-      "step": 5560
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.0562212467193604,
-      "learning_rate": 1.6562711864406782e-05,
-      "loss": 1.4766,
-      "step": 5570
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.577533483505249,
-      "learning_rate": 1.6555932203389832e-05,
-      "loss": 1.3865,
-      "step": 5580
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.4926679134368896,
-      "learning_rate": 1.654915254237288e-05,
-      "loss": 1.3764,
-      "step": 5590
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.904751777648926,
-      "learning_rate": 1.654237288135593e-05,
-      "loss": 1.2022,
-      "step": 5600
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 4.692943096160889,
-      "learning_rate": 1.6535593220338984e-05,
-      "loss": 1.2504,
-      "step": 5610
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.7578885555267334,
-      "learning_rate": 1.6528813559322034e-05,
-      "loss": 1.3805,
-      "step": 5620
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 6.6114630699157715,
-      "learning_rate": 1.6522033898305084e-05,
-      "loss": 1.3391,
-      "step": 5630
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.5723536014556885,
-      "learning_rate": 1.6515254237288137e-05,
-      "loss": 1.2655,
-      "step": 5640
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.141366720199585,
-      "learning_rate": 1.6508474576271187e-05,
-      "loss": 1.4299,
-      "step": 5650
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.970369577407837,
-      "learning_rate": 1.650169491525424e-05,
-      "loss": 1.4145,
-      "step": 5660
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 4.518489837646484,
-      "learning_rate": 1.649491525423729e-05,
-      "loss": 1.419,
-      "step": 5670
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.4335029125213623,
-      "learning_rate": 1.648813559322034e-05,
-      "loss": 1.3781,
-      "step": 5680
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.025974988937378,
-      "learning_rate": 1.6481355932203392e-05,
-      "loss": 1.456,
-      "step": 5690
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.4364434480667114,
-      "learning_rate": 1.6474576271186442e-05,
-      "loss": 1.4295,
-      "step": 5700
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.17189359664917,
-      "learning_rate": 1.646779661016949e-05,
-      "loss": 1.3155,
-      "step": 5710
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 4.905094146728516,
-      "learning_rate": 1.6461016949152545e-05,
-      "loss": 1.3161,
-      "step": 5720
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 2.6364047527313232,
-      "learning_rate": 1.6454237288135594e-05,
-      "loss": 1.378,
-      "step": 5730
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.8336093425750732,
-      "learning_rate": 1.6447457627118648e-05,
-      "loss": 1.2784,
-      "step": 5740
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 4.631802558898926,
-      "learning_rate": 1.6440677966101697e-05,
-      "loss": 1.2603,
-      "step": 5750
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 9.11829948425293,
-      "learning_rate": 1.6433898305084747e-05,
-      "loss": 1.3563,
-      "step": 5760
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.024040460586548,
-      "learning_rate": 1.64271186440678e-05,
-      "loss": 1.3491,
-      "step": 5770
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.558547019958496,
-      "learning_rate": 1.642033898305085e-05,
-      "loss": 1.3992,
-      "step": 5780
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.277186870574951,
-      "learning_rate": 1.64135593220339e-05,
-      "loss": 1.3636,
-      "step": 5790
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 3.6262595653533936,
-      "learning_rate": 1.640677966101695e-05,
-      "loss": 1.2908,
-      "step": 5800
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.5076159238815308,
-      "learning_rate": 1.64e-05,
-      "loss": 1.4525,
-      "step": 5810
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.863429069519043,
-      "learning_rate": 1.6393220338983052e-05,
-      "loss": 1.5479,
-      "step": 5820
-    },
-    {
-      "epoch": 0.13,
-      "grad_norm": 1.3503921031951904,
-      "learning_rate": 1.63864406779661e-05,
-      "loss": 1.2465,
-      "step": 5830
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.307507276535034,
-      "learning_rate": 1.6379661016949155e-05,
-      "loss": 1.3346,
-      "step": 5840
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.301724910736084,
-      "learning_rate": 1.6372881355932204e-05,
-      "loss": 1.4078,
-      "step": 5850
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.821153163909912,
-      "learning_rate": 1.6366101694915254e-05,
-      "loss": 1.4262,
-      "step": 5860
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 1.4718036651611328,
-      "learning_rate": 1.6359322033898307e-05,
-      "loss": 1.5426,
-      "step": 5870
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.044488430023193,
-      "learning_rate": 1.6352542372881357e-05,
-      "loss": 1.3703,
-      "step": 5880
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.0791540145874023,
-      "learning_rate": 1.6345762711864407e-05,
-      "loss": 1.3436,
-      "step": 5890
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.2117247581481934,
-      "learning_rate": 1.633898305084746e-05,
-      "loss": 1.428,
-      "step": 5900
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.8183112144470215,
-      "learning_rate": 1.633220338983051e-05,
-      "loss": 1.6166,
-      "step": 5910
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.566955089569092,
-      "learning_rate": 1.6325423728813563e-05,
-      "loss": 1.4127,
-      "step": 5920
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.455868721008301,
-      "learning_rate": 1.6318644067796612e-05,
-      "loss": 1.2684,
-      "step": 5930
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 1.872295618057251,
-      "learning_rate": 1.6311864406779662e-05,
-      "loss": 1.1753,
-      "step": 5940
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.5006000995635986,
-      "learning_rate": 1.6305084745762715e-05,
-      "loss": 1.1916,
-      "step": 5950
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.872570514678955,
-      "learning_rate": 1.6298305084745765e-05,
-      "loss": 1.1692,
-      "step": 5960
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.5257492065429688,
-      "learning_rate": 1.6291525423728814e-05,
-      "loss": 1.3611,
-      "step": 5970
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.483407497406006,
-      "learning_rate": 1.6284745762711868e-05,
-      "loss": 1.4315,
-      "step": 5980
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 1.8382426500320435,
-      "learning_rate": 1.6277966101694917e-05,
-      "loss": 1.4698,
-      "step": 5990
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.766387939453125,
-      "learning_rate": 1.6271186440677967e-05,
-      "loss": 1.361,
-      "step": 6000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 1.1337964534759521,
-      "eval_runtime": 67.0656,
-      "eval_samples_per_second": 14.911,
-      "eval_steps_per_second": 14.911,
-      "step": 6000
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.5609889030456543,
-      "learning_rate": 1.6264406779661017e-05,
-      "loss": 1.4736,
-      "step": 6010
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.4243054389953613,
-      "learning_rate": 1.625762711864407e-05,
-      "loss": 1.3073,
-      "step": 6020
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 1.1757571697235107,
-      "learning_rate": 1.625084745762712e-05,
-      "loss": 1.5282,
-      "step": 6030
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.4322659969329834,
-      "learning_rate": 1.624406779661017e-05,
-      "loss": 1.534,
-      "step": 6040
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.9377753734588623,
-      "learning_rate": 1.6237288135593222e-05,
-      "loss": 1.354,
-      "step": 6050
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.534852981567383,
-      "learning_rate": 1.6230508474576272e-05,
-      "loss": 1.2613,
-      "step": 6060
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 1.6617101430892944,
-      "learning_rate": 1.6223728813559322e-05,
-      "loss": 1.4414,
-      "step": 6070
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.800033092498779,
-      "learning_rate": 1.6216949152542375e-05,
-      "loss": 1.3286,
-      "step": 6080
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.045539379119873,
-      "learning_rate": 1.6210169491525424e-05,
-      "loss": 1.4305,
-      "step": 6090
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.9078855514526367,
-      "learning_rate": 1.6203389830508474e-05,
-      "loss": 1.3735,
-      "step": 6100
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.0546069145202637,
-      "learning_rate": 1.6196610169491527e-05,
-      "loss": 1.3062,
-      "step": 6110
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.7809054851531982,
-      "learning_rate": 1.6189830508474577e-05,
-      "loss": 1.2676,
-      "step": 6120
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.5306100845336914,
-      "learning_rate": 1.618305084745763e-05,
-      "loss": 1.4725,
-      "step": 6130
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.504068374633789,
-      "learning_rate": 1.617627118644068e-05,
-      "loss": 1.3162,
-      "step": 6140
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 5.180143356323242,
-      "learning_rate": 1.616949152542373e-05,
-      "loss": 1.3681,
-      "step": 6150
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.692683696746826,
-      "learning_rate": 1.6162711864406783e-05,
-      "loss": 1.2732,
-      "step": 6160
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 1.6309053897857666,
-      "learning_rate": 1.6155932203389832e-05,
-      "loss": 1.212,
-      "step": 6170
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.247219562530518,
-      "learning_rate": 1.6149152542372882e-05,
-      "loss": 1.3263,
-      "step": 6180
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 1.420581579208374,
-      "learning_rate": 1.6142372881355935e-05,
-      "loss": 1.2392,
-      "step": 6190
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 3.2501585483551025,
-      "learning_rate": 1.6135593220338985e-05,
-      "loss": 1.3958,
-      "step": 6200
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 5.111384868621826,
-      "learning_rate": 1.6128813559322038e-05,
-      "loss": 1.4578,
-      "step": 6210
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.0279738903045654,
-      "learning_rate": 1.6122033898305084e-05,
-      "loss": 1.4083,
-      "step": 6220
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 0.9996086955070496,
-      "learning_rate": 1.6115254237288137e-05,
-      "loss": 1.2599,
-      "step": 6230
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 2.6425893306732178,
-      "learning_rate": 1.6108474576271187e-05,
-      "loss": 1.3794,
-      "step": 6240
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 4.84988260269165,
-      "learning_rate": 1.6101694915254237e-05,
-      "loss": 1.3801,
-      "step": 6250
-    },
-    {
-      "epoch": 0.14,
-      "grad_norm": 7.806826591491699,
-      "learning_rate": 1.609491525423729e-05,
-      "loss": 1.3006,
-      "step": 6260
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 6.427366733551025,
-      "learning_rate": 1.608813559322034e-05,
-      "loss": 1.4588,
-      "step": 6270
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.51039719581604,
-      "learning_rate": 1.608135593220339e-05,
-      "loss": 1.4209,
-      "step": 6280
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.638364553451538,
-      "learning_rate": 1.6074576271186442e-05,
-      "loss": 1.3952,
-      "step": 6290
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.9240260124206543,
-      "learning_rate": 1.6067796610169492e-05,
-      "loss": 1.3194,
-      "step": 6300
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.798541784286499,
-      "learning_rate": 1.6061016949152545e-05,
-      "loss": 1.2272,
-      "step": 6310
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.5487680435180664,
-      "learning_rate": 1.6054237288135595e-05,
-      "loss": 1.4795,
-      "step": 6320
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.1405465602874756,
-      "learning_rate": 1.6047457627118645e-05,
-      "loss": 1.3574,
-      "step": 6330
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.4945203065872192,
-      "learning_rate": 1.6040677966101698e-05,
-      "loss": 1.3875,
-      "step": 6340
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.4741933345794678,
-      "learning_rate": 1.6033898305084747e-05,
-      "loss": 1.3757,
-      "step": 6350
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.7438557147979736,
-      "learning_rate": 1.6027118644067797e-05,
-      "loss": 1.4486,
-      "step": 6360
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.901851177215576,
-      "learning_rate": 1.602033898305085e-05,
-      "loss": 1.3954,
-      "step": 6370
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.6728180646896362,
-      "learning_rate": 1.60135593220339e-05,
-      "loss": 1.311,
-      "step": 6380
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.970463752746582,
-      "learning_rate": 1.600677966101695e-05,
-      "loss": 1.4011,
-      "step": 6390
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 12.278678894042969,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.4062,
-      "step": 6400
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.6090126037597656,
-      "learning_rate": 1.5993220338983052e-05,
-      "loss": 1.3565,
-      "step": 6410
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 6.103749752044678,
-      "learning_rate": 1.5986440677966105e-05,
-      "loss": 1.4804,
-      "step": 6420
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.822219133377075,
-      "learning_rate": 1.5979661016949152e-05,
-      "loss": 1.4567,
-      "step": 6430
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 4.869916915893555,
-      "learning_rate": 1.5972881355932205e-05,
-      "loss": 1.3093,
-      "step": 6440
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.4998152256011963,
-      "learning_rate": 1.5966101694915255e-05,
-      "loss": 1.5148,
-      "step": 6450
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.6436197757720947,
-      "learning_rate": 1.5959322033898304e-05,
-      "loss": 1.519,
-      "step": 6460
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.9514076709747314,
-      "learning_rate": 1.5952542372881357e-05,
-      "loss": 1.2673,
-      "step": 6470
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.19160795211792,
-      "learning_rate": 1.5945762711864407e-05,
-      "loss": 1.2964,
-      "step": 6480
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.302166700363159,
-      "learning_rate": 1.593898305084746e-05,
-      "loss": 1.5311,
-      "step": 6490
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.648465156555176,
-      "learning_rate": 1.593220338983051e-05,
-      "loss": 1.5642,
-      "step": 6500
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 1.1224900484085083,
-      "eval_runtime": 67.0535,
-      "eval_samples_per_second": 14.913,
-      "eval_steps_per_second": 14.913,
-      "step": 6500
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 4.6730732917785645,
-      "learning_rate": 1.592542372881356e-05,
-      "loss": 1.3842,
-      "step": 6510
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.7324059009552,
-      "learning_rate": 1.5918644067796613e-05,
-      "loss": 1.4838,
-      "step": 6520
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.5643513202667236,
-      "learning_rate": 1.5911864406779662e-05,
-      "loss": 1.326,
-      "step": 6530
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 5.981233596801758,
-      "learning_rate": 1.5905084745762712e-05,
-      "loss": 1.1575,
-      "step": 6540
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.4779703617095947,
-      "learning_rate": 1.5898305084745765e-05,
-      "loss": 1.3118,
-      "step": 6550
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 5.831774711608887,
-      "learning_rate": 1.5891525423728815e-05,
-      "loss": 1.3681,
-      "step": 6560
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 4.515669345855713,
-      "learning_rate": 1.5884745762711865e-05,
-      "loss": 1.2631,
-      "step": 6570
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.7888094186782837,
-      "learning_rate": 1.5877966101694918e-05,
-      "loss": 1.4389,
-      "step": 6580
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.9246420860290527,
-      "learning_rate": 1.5871186440677967e-05,
-      "loss": 1.2706,
-      "step": 6590
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.629147529602051,
-      "learning_rate": 1.586440677966102e-05,
-      "loss": 1.4099,
-      "step": 6600
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.5322067737579346,
-      "learning_rate": 1.585762711864407e-05,
-      "loss": 1.3107,
-      "step": 6610
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 1.1020938158035278,
-      "learning_rate": 1.585084745762712e-05,
-      "loss": 1.3415,
-      "step": 6620
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 5.216763496398926,
-      "learning_rate": 1.584406779661017e-05,
-      "loss": 1.3664,
-      "step": 6630
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.0716798305511475,
-      "learning_rate": 1.583728813559322e-05,
-      "loss": 1.1287,
-      "step": 6640
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.659613609313965,
-      "learning_rate": 1.5830508474576272e-05,
-      "loss": 1.4241,
-      "step": 6650
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 7.536604881286621,
-      "learning_rate": 1.5823728813559322e-05,
-      "loss": 1.426,
-      "step": 6660
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.213017702102661,
-      "learning_rate": 1.5816949152542372e-05,
-      "loss": 1.3564,
-      "step": 6670
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 2.929147481918335,
-      "learning_rate": 1.5810169491525425e-05,
-      "loss": 1.5176,
-      "step": 6680
-    },
-    {
-      "epoch": 0.15,
-      "grad_norm": 3.5247678756713867,
-      "learning_rate": 1.5803389830508475e-05,
-      "loss": 1.4608,
-      "step": 6690
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.313474178314209,
-      "learning_rate": 1.5796610169491528e-05,
-      "loss": 1.411,
-      "step": 6700
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.516223907470703,
-      "learning_rate": 1.5789830508474577e-05,
-      "loss": 1.2478,
-      "step": 6710
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 3.387552499771118,
-      "learning_rate": 1.5783050847457627e-05,
-      "loss": 1.3225,
-      "step": 6720
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.9858660697937012,
-      "learning_rate": 1.577627118644068e-05,
-      "loss": 1.3289,
-      "step": 6730
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.3014678955078125,
-      "learning_rate": 1.576949152542373e-05,
-      "loss": 1.457,
-      "step": 6740
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 10.245163917541504,
-      "learning_rate": 1.576271186440678e-05,
-      "loss": 1.4311,
-      "step": 6750
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.805487871170044,
-      "learning_rate": 1.5755932203389833e-05,
-      "loss": 1.3857,
-      "step": 6760
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.2822433710098267,
-      "learning_rate": 1.5749152542372882e-05,
-      "loss": 1.4489,
-      "step": 6770
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.254396915435791,
-      "learning_rate": 1.5742372881355936e-05,
-      "loss": 1.3914,
-      "step": 6780
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.3311165571212769,
-      "learning_rate": 1.5735593220338985e-05,
-      "loss": 1.243,
-      "step": 6790
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.7225605249404907,
-      "learning_rate": 1.5728813559322035e-05,
-      "loss": 1.425,
-      "step": 6800
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 3.585598945617676,
-      "learning_rate": 1.5722033898305088e-05,
-      "loss": 1.414,
-      "step": 6810
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.3213350772857666,
-      "learning_rate": 1.5715254237288138e-05,
-      "loss": 1.3233,
-      "step": 6820
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.5480399131774902,
-      "learning_rate": 1.5708474576271187e-05,
-      "loss": 1.3096,
-      "step": 6830
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.9461381435394287,
-      "learning_rate": 1.5701694915254237e-05,
-      "loss": 1.2209,
-      "step": 6840
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.8784675598144531,
-      "learning_rate": 1.5694915254237287e-05,
-      "loss": 1.2154,
-      "step": 6850
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.964296817779541,
-      "learning_rate": 1.568813559322034e-05,
-      "loss": 1.3913,
-      "step": 6860
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.013964891433716,
-      "learning_rate": 1.568135593220339e-05,
-      "loss": 1.4206,
-      "step": 6870
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.359316110610962,
-      "learning_rate": 1.5674576271186443e-05,
-      "loss": 1.3806,
-      "step": 6880
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 5.8272175788879395,
-      "learning_rate": 1.5667796610169492e-05,
-      "loss": 1.2084,
-      "step": 6890
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.1895947456359863,
-      "learning_rate": 1.5661016949152542e-05,
-      "loss": 1.5326,
-      "step": 6900
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 3.105100631713867,
-      "learning_rate": 1.5654237288135595e-05,
-      "loss": 1.3586,
-      "step": 6910
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.031907081604004,
-      "learning_rate": 1.5647457627118645e-05,
-      "loss": 1.0995,
-      "step": 6920
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 15.673041343688965,
-      "learning_rate": 1.5640677966101695e-05,
-      "loss": 1.5304,
-      "step": 6930
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 6.542800426483154,
-      "learning_rate": 1.5633898305084748e-05,
-      "loss": 1.4612,
-      "step": 6940
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 4.4466352462768555,
-      "learning_rate": 1.5627118644067798e-05,
-      "loss": 1.4641,
-      "step": 6950
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.0358104705810547,
-      "learning_rate": 1.562033898305085e-05,
-      "loss": 1.5169,
-      "step": 6960
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 3.3914594650268555,
-      "learning_rate": 1.56135593220339e-05,
-      "loss": 1.3483,
-      "step": 6970
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.626600503921509,
-      "learning_rate": 1.560677966101695e-05,
-      "loss": 1.3114,
-      "step": 6980
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 5.424988746643066,
-      "learning_rate": 1.5600000000000003e-05,
-      "loss": 1.2768,
-      "step": 6990
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 7.747529029846191,
-      "learning_rate": 1.5593220338983053e-05,
-      "loss": 1.1494,
-      "step": 7000
-    },
-    {
-      "epoch": 0.16,
-      "eval_loss": 1.0665773153305054,
-      "eval_runtime": 67.1469,
-      "eval_samples_per_second": 14.893,
-      "eval_steps_per_second": 14.893,
-      "step": 7000
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 0.9942423701286316,
-      "learning_rate": 1.5586440677966103e-05,
-      "loss": 1.3625,
-      "step": 7010
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.6136345863342285,
-      "learning_rate": 1.5579661016949156e-05,
-      "loss": 1.2317,
-      "step": 7020
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.3277318477630615,
-      "learning_rate": 1.5572881355932205e-05,
-      "loss": 1.5961,
-      "step": 7030
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 4.660468578338623,
-      "learning_rate": 1.5566101694915255e-05,
-      "loss": 1.3753,
-      "step": 7040
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.3660237789154053,
-      "learning_rate": 1.5559322033898305e-05,
-      "loss": 1.311,
-      "step": 7050
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 5.831724643707275,
-      "learning_rate": 1.5552542372881358e-05,
-      "loss": 1.3895,
-      "step": 7060
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 4.309049606323242,
-      "learning_rate": 1.5545762711864408e-05,
-      "loss": 1.325,
-      "step": 7070
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.5130770206451416,
-      "learning_rate": 1.5538983050847457e-05,
-      "loss": 1.3108,
-      "step": 7080
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 11.385069847106934,
-      "learning_rate": 1.553220338983051e-05,
-      "loss": 1.3483,
-      "step": 7090
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 2.305030107498169,
-      "learning_rate": 1.552542372881356e-05,
-      "loss": 1.12,
-      "step": 7100
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 1.7694568634033203,
-      "learning_rate": 1.551864406779661e-05,
-      "loss": 1.4892,
-      "step": 7110
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 3.4342284202575684,
-      "learning_rate": 1.5511864406779663e-05,
-      "loss": 1.3841,
-      "step": 7120
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 8.617453575134277,
-      "learning_rate": 1.5505084745762713e-05,
-      "loss": 1.2856,
-      "step": 7130
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 4.346908092498779,
-      "learning_rate": 1.5498305084745762e-05,
-      "loss": 1.2906,
-      "step": 7140
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 3.7418603897094727,
-      "learning_rate": 1.5491525423728815e-05,
-      "loss": 1.5261,
-      "step": 7150
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 1.1197525262832642,
-      "learning_rate": 1.5484745762711865e-05,
-      "loss": 1.2972,
-      "step": 7160
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 3.5833466053009033,
-      "learning_rate": 1.5477966101694918e-05,
-      "loss": 1.3393,
-      "step": 7170
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 3.046830892562866,
-      "learning_rate": 1.5471186440677968e-05,
-      "loss": 1.4726,
-      "step": 7180
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 10.847610473632812,
-      "learning_rate": 1.5464406779661018e-05,
-      "loss": 1.4872,
-      "step": 7190
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 1.379472255706787,
-      "learning_rate": 1.545762711864407e-05,
-      "loss": 1.3873,
-      "step": 7200
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.1769838333129883,
-      "learning_rate": 1.545084745762712e-05,
-      "loss": 1.432,
-      "step": 7210
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.547250747680664,
-      "learning_rate": 1.544406779661017e-05,
-      "loss": 1.5696,
-      "step": 7220
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 8.801671981811523,
-      "learning_rate": 1.5437288135593223e-05,
-      "loss": 1.4942,
-      "step": 7230
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.674645185470581,
-      "learning_rate": 1.5430508474576273e-05,
-      "loss": 1.2567,
-      "step": 7240
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 3.047281503677368,
-      "learning_rate": 1.5423728813559326e-05,
-      "loss": 1.1286,
-      "step": 7250
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 1.422094464302063,
-      "learning_rate": 1.5416949152542372e-05,
-      "loss": 1.4765,
-      "step": 7260
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 3.8606622219085693,
-      "learning_rate": 1.5410169491525425e-05,
-      "loss": 1.2492,
-      "step": 7270
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.5654449462890625,
-      "learning_rate": 1.5403389830508475e-05,
-      "loss": 1.2715,
-      "step": 7280
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.5264976024627686,
-      "learning_rate": 1.5396610169491525e-05,
-      "loss": 1.4896,
-      "step": 7290
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 5.585784435272217,
-      "learning_rate": 1.5389830508474578e-05,
-      "loss": 1.3929,
-      "step": 7300
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 1.7968121767044067,
-      "learning_rate": 1.5383050847457628e-05,
-      "loss": 1.3018,
-      "step": 7310
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.1171720027923584,
-      "learning_rate": 1.5376271186440677e-05,
-      "loss": 1.4814,
-      "step": 7320
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 5.140902519226074,
-      "learning_rate": 1.536949152542373e-05,
-      "loss": 1.454,
-      "step": 7330
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 8.208806037902832,
-      "learning_rate": 1.536271186440678e-05,
-      "loss": 1.3176,
-      "step": 7340
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 4.932653903961182,
-      "learning_rate": 1.5355932203389833e-05,
-      "loss": 1.3674,
-      "step": 7350
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 0.9559627175331116,
-      "learning_rate": 1.5349152542372883e-05,
-      "loss": 1.3701,
-      "step": 7360
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 4.199779033660889,
-      "learning_rate": 1.5342372881355933e-05,
-      "loss": 1.2293,
-      "step": 7370
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 4.100563049316406,
-      "learning_rate": 1.5335593220338986e-05,
-      "loss": 1.3448,
-      "step": 7380
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.8607914447784424,
-      "learning_rate": 1.5328813559322035e-05,
-      "loss": 1.4457,
-      "step": 7390
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 6.619777202606201,
-      "learning_rate": 1.5322033898305085e-05,
-      "loss": 1.5492,
-      "step": 7400
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 5.614234447479248,
-      "learning_rate": 1.5315254237288138e-05,
-      "loss": 1.2983,
-      "step": 7410
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.3404347896575928,
-      "learning_rate": 1.5308474576271188e-05,
-      "loss": 1.3885,
-      "step": 7420
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.3918747901916504,
-      "learning_rate": 1.530169491525424e-05,
-      "loss": 1.2908,
-      "step": 7430
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.5815324783325195,
-      "learning_rate": 1.529491525423729e-05,
-      "loss": 1.2632,
-      "step": 7440
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 5.126127243041992,
-      "learning_rate": 1.528813559322034e-05,
-      "loss": 1.6003,
-      "step": 7450
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 4.3783955574035645,
-      "learning_rate": 1.528135593220339e-05,
-      "loss": 1.3901,
-      "step": 7460
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.654667615890503,
-      "learning_rate": 1.527457627118644e-05,
-      "loss": 1.2822,
-      "step": 7470
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 7.821676731109619,
-      "learning_rate": 1.5267796610169493e-05,
-      "loss": 1.396,
-      "step": 7480
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.0116024017333984,
-      "learning_rate": 1.5261016949152543e-05,
-      "loss": 1.4361,
-      "step": 7490
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.1018917560577393,
-      "learning_rate": 1.5254237288135594e-05,
-      "loss": 1.3302,
-      "step": 7500
-    },
-    {
-      "epoch": 0.17,
-      "eval_loss": 1.1425343751907349,
-      "eval_runtime": 67.2314,
-      "eval_samples_per_second": 14.874,
-      "eval_steps_per_second": 14.874,
-      "step": 7500
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 4.509432315826416,
-      "learning_rate": 1.5247457627118645e-05,
-      "loss": 1.3979,
-      "step": 7510
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 2.7782204151153564,
-      "learning_rate": 1.5240677966101695e-05,
-      "loss": 1.4353,
-      "step": 7520
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 1.9378540515899658,
-      "learning_rate": 1.5233898305084747e-05,
-      "loss": 1.2803,
-      "step": 7530
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 3.195469379425049,
-      "learning_rate": 1.5227118644067798e-05,
-      "loss": 1.3325,
-      "step": 7540
-    },
-    {
-      "epoch": 0.17,
-      "grad_norm": 3.075035810470581,
-      "learning_rate": 1.522033898305085e-05,
-      "loss": 1.2988,
-      "step": 7550
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.2794687747955322,
-      "learning_rate": 1.5213559322033899e-05,
-      "loss": 1.3588,
-      "step": 7560
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.5682082176208496,
-      "learning_rate": 1.520677966101695e-05,
-      "loss": 1.2074,
-      "step": 7570
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.2680002450942993,
-      "learning_rate": 1.5200000000000002e-05,
-      "loss": 1.3791,
-      "step": 7580
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 4.0036725997924805,
-      "learning_rate": 1.5193220338983052e-05,
-      "loss": 1.4726,
-      "step": 7590
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.5019006729125977,
-      "learning_rate": 1.5186440677966103e-05,
-      "loss": 1.418,
-      "step": 7600
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.3398947715759277,
-      "learning_rate": 1.5179661016949154e-05,
-      "loss": 1.2933,
-      "step": 7610
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.528242826461792,
-      "learning_rate": 1.5172881355932206e-05,
-      "loss": 1.1812,
-      "step": 7620
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.616392970085144,
-      "learning_rate": 1.5166101694915255e-05,
-      "loss": 1.5172,
-      "step": 7630
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 3.7177088260650635,
-      "learning_rate": 1.5159322033898307e-05,
-      "loss": 1.251,
-      "step": 7640
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.7318824529647827,
-      "learning_rate": 1.5152542372881358e-05,
-      "loss": 1.4652,
-      "step": 7650
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 9.107917785644531,
-      "learning_rate": 1.514576271186441e-05,
-      "loss": 1.2738,
-      "step": 7660
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.2745494842529297,
-      "learning_rate": 1.5138983050847458e-05,
-      "loss": 1.3997,
-      "step": 7670
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.564849376678467,
-      "learning_rate": 1.5132203389830509e-05,
-      "loss": 1.3054,
-      "step": 7680
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 5.596649169921875,
-      "learning_rate": 1.512542372881356e-05,
-      "loss": 1.2581,
-      "step": 7690
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.4178537130355835,
-      "learning_rate": 1.511864406779661e-05,
-      "loss": 1.3319,
-      "step": 7700
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 3.1207597255706787,
-      "learning_rate": 1.5111864406779662e-05,
-      "loss": 1.3919,
-      "step": 7710
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.1499109268188477,
-      "learning_rate": 1.5105084745762713e-05,
-      "loss": 1.2123,
-      "step": 7720
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 7.321339130401611,
-      "learning_rate": 1.5098305084745763e-05,
-      "loss": 1.3294,
-      "step": 7730
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 4.1394243240356445,
-      "learning_rate": 1.5091525423728814e-05,
-      "loss": 1.4105,
-      "step": 7740
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 4.41034460067749,
-      "learning_rate": 1.5084745762711865e-05,
-      "loss": 1.4582,
-      "step": 7750
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 5.8325324058532715,
-      "learning_rate": 1.5077966101694917e-05,
-      "loss": 1.1756,
-      "step": 7760
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 3.059936761856079,
-      "learning_rate": 1.5071186440677967e-05,
-      "loss": 1.4276,
-      "step": 7770
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.1993743181228638,
-      "learning_rate": 1.5064406779661018e-05,
-      "loss": 1.1909,
-      "step": 7780
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 6.689935207366943,
-      "learning_rate": 1.505762711864407e-05,
-      "loss": 1.3381,
-      "step": 7790
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 4.1810150146484375,
-      "learning_rate": 1.505084745762712e-05,
-      "loss": 1.4885,
-      "step": 7800
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 4.501943588256836,
-      "learning_rate": 1.504406779661017e-05,
-      "loss": 1.3556,
-      "step": 7810
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 7.109510898590088,
-      "learning_rate": 1.5037288135593222e-05,
-      "loss": 1.3156,
-      "step": 7820
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 4.330848693847656,
-      "learning_rate": 1.5030508474576273e-05,
-      "loss": 1.508,
-      "step": 7830
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 3.391294240951538,
-      "learning_rate": 1.5023728813559325e-05,
-      "loss": 1.3651,
-      "step": 7840
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.8204346895217896,
-      "learning_rate": 1.5016949152542374e-05,
-      "loss": 1.3925,
-      "step": 7850
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.8383281230926514,
-      "learning_rate": 1.5010169491525426e-05,
-      "loss": 1.4486,
-      "step": 7860
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 7.470884323120117,
-      "learning_rate": 1.5003389830508477e-05,
-      "loss": 1.4738,
-      "step": 7870
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 5.963416576385498,
-      "learning_rate": 1.4996610169491525e-05,
-      "loss": 1.3832,
-      "step": 7880
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 5.30115270614624,
-      "learning_rate": 1.4989830508474577e-05,
-      "loss": 1.4412,
-      "step": 7890
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.7669206857681274,
-      "learning_rate": 1.4983050847457628e-05,
-      "loss": 1.3016,
-      "step": 7900
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.1460033655166626,
-      "learning_rate": 1.4976271186440678e-05,
-      "loss": 1.3938,
-      "step": 7910
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 3.047388792037964,
-      "learning_rate": 1.4969491525423729e-05,
-      "loss": 1.352,
-      "step": 7920
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.9169957637786865,
-      "learning_rate": 1.496271186440678e-05,
-      "loss": 1.274,
-      "step": 7930
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.4173505306243896,
-      "learning_rate": 1.4955932203389832e-05,
-      "loss": 1.5062,
-      "step": 7940
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 2.8410110473632812,
-      "learning_rate": 1.4949152542372882e-05,
-      "loss": 1.3504,
-      "step": 7950
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 4.969642162322998,
-      "learning_rate": 1.4942372881355933e-05,
-      "loss": 1.4739,
-      "step": 7960
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 3.8513638973236084,
-      "learning_rate": 1.4935593220338984e-05,
-      "loss": 1.4283,
-      "step": 7970
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.2723006010055542,
-      "learning_rate": 1.4928813559322036e-05,
-      "loss": 1.5301,
-      "step": 7980
-    },
-    {
-      "epoch": 0.18,
-      "grad_norm": 1.1805778741836548,
-      "learning_rate": 1.4922033898305086e-05,
-      "loss": 1.3213,
-      "step": 7990
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.0252439975738525,
-      "learning_rate": 1.4915254237288137e-05,
-      "loss": 1.2692,
-      "step": 8000
-    },
-    {
-      "epoch": 0.19,
-      "eval_loss": 1.098508358001709,
-      "eval_runtime": 67.0444,
-      "eval_samples_per_second": 14.915,
-      "eval_steps_per_second": 14.915,
-      "step": 8000
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.2070083618164062,
-      "learning_rate": 1.4908474576271188e-05,
-      "loss": 1.3754,
-      "step": 8010
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.910412549972534,
-      "learning_rate": 1.490169491525424e-05,
-      "loss": 1.5372,
-      "step": 8020
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.161473274230957,
-      "learning_rate": 1.489491525423729e-05,
-      "loss": 1.4059,
-      "step": 8030
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.312554121017456,
-      "learning_rate": 1.4888135593220341e-05,
-      "loss": 1.3982,
-      "step": 8040
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.2093117237091064,
-      "learning_rate": 1.4881355932203392e-05,
-      "loss": 1.386,
-      "step": 8050
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.6620709896087646,
-      "learning_rate": 1.4874576271186442e-05,
-      "loss": 1.2786,
-      "step": 8060
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.2062294483184814,
-      "learning_rate": 1.4867796610169493e-05,
-      "loss": 1.419,
-      "step": 8070
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 4.173048973083496,
-      "learning_rate": 1.4861016949152545e-05,
-      "loss": 1.3723,
-      "step": 8080
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 4.707441329956055,
-      "learning_rate": 1.4854237288135593e-05,
-      "loss": 1.1348,
-      "step": 8090
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.876202344894409,
-      "learning_rate": 1.4847457627118644e-05,
-      "loss": 1.1759,
-      "step": 8100
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.0642389059066772,
-      "learning_rate": 1.4840677966101696e-05,
-      "loss": 1.2566,
-      "step": 8110
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.49448823928833,
-      "learning_rate": 1.4833898305084747e-05,
-      "loss": 1.4826,
-      "step": 8120
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.88454008102417,
-      "learning_rate": 1.4827118644067797e-05,
-      "loss": 1.3342,
-      "step": 8130
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 14.040616035461426,
-      "learning_rate": 1.4820338983050848e-05,
-      "loss": 1.293,
-      "step": 8140
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.384743094444275,
-      "learning_rate": 1.48135593220339e-05,
-      "loss": 1.5205,
-      "step": 8150
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.8649702072143555,
-      "learning_rate": 1.4806779661016951e-05,
-      "loss": 1.4878,
-      "step": 8160
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.677607774734497,
-      "learning_rate": 1.48e-05,
-      "loss": 1.3551,
-      "step": 8170
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.7539191246032715,
-      "learning_rate": 1.4793220338983052e-05,
-      "loss": 1.6182,
-      "step": 8180
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.3938488960266113,
-      "learning_rate": 1.4786440677966103e-05,
-      "loss": 1.3727,
-      "step": 8190
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.8780012130737305,
-      "learning_rate": 1.4779661016949153e-05,
-      "loss": 1.3845,
-      "step": 8200
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.8616294860839844,
-      "learning_rate": 1.4772881355932205e-05,
-      "loss": 1.4509,
-      "step": 8210
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.9469797611236572,
-      "learning_rate": 1.4766101694915256e-05,
-      "loss": 1.2653,
-      "step": 8220
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.5003085136413574,
-      "learning_rate": 1.4759322033898307e-05,
-      "loss": 1.4115,
-      "step": 8230
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 5.820629596710205,
-      "learning_rate": 1.4752542372881357e-05,
-      "loss": 1.4058,
-      "step": 8240
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 8.11052417755127,
-      "learning_rate": 1.4745762711864408e-05,
-      "loss": 1.3686,
-      "step": 8250
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 5.473447799682617,
-      "learning_rate": 1.473898305084746e-05,
-      "loss": 1.2394,
-      "step": 8260
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 8.85197925567627,
-      "learning_rate": 1.4732203389830511e-05,
-      "loss": 1.3407,
-      "step": 8270
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.4516103267669678,
-      "learning_rate": 1.4725423728813561e-05,
-      "loss": 1.5529,
-      "step": 8280
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.373441219329834,
-      "learning_rate": 1.4718644067796612e-05,
-      "loss": 1.3036,
-      "step": 8290
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 4.667367458343506,
-      "learning_rate": 1.4711864406779662e-05,
-      "loss": 1.219,
-      "step": 8300
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.8208861351013184,
-      "learning_rate": 1.4705084745762712e-05,
-      "loss": 1.2244,
-      "step": 8310
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 9.953441619873047,
-      "learning_rate": 1.4698305084745763e-05,
-      "loss": 1.2222,
-      "step": 8320
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 2.542130708694458,
-      "learning_rate": 1.4691525423728815e-05,
-      "loss": 1.4311,
-      "step": 8330
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 4.040639400482178,
-      "learning_rate": 1.4684745762711864e-05,
-      "loss": 1.3049,
-      "step": 8340
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.6556200981140137,
-      "learning_rate": 1.4677966101694916e-05,
-      "loss": 1.3456,
-      "step": 8350
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.1826813220977783,
-      "learning_rate": 1.4671186440677967e-05,
-      "loss": 1.3355,
-      "step": 8360
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.7379783391952515,
-      "learning_rate": 1.4664406779661018e-05,
-      "loss": 1.2498,
-      "step": 8370
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 4.671259880065918,
-      "learning_rate": 1.4657627118644068e-05,
-      "loss": 1.4722,
-      "step": 8380
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.5242054462432861,
-      "learning_rate": 1.465084745762712e-05,
-      "loss": 1.4148,
-      "step": 8390
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 1.3360844850540161,
-      "learning_rate": 1.4644067796610171e-05,
-      "loss": 1.2182,
-      "step": 8400
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.401080369949341,
-      "learning_rate": 1.4637288135593222e-05,
-      "loss": 1.2632,
-      "step": 8410
-    },
-    {
-      "epoch": 0.19,
-      "grad_norm": 3.884335517883301,
-      "learning_rate": 1.4630508474576272e-05,
-      "loss": 1.2798,
-      "step": 8420
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.928900718688965,
-      "learning_rate": 1.4623728813559323e-05,
-      "loss": 1.3342,
-      "step": 8430
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.478438377380371,
-      "learning_rate": 1.4616949152542375e-05,
-      "loss": 1.4387,
-      "step": 8440
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.839172124862671,
-      "learning_rate": 1.4610169491525426e-05,
-      "loss": 1.4165,
-      "step": 8450
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 8.968944549560547,
-      "learning_rate": 1.4603389830508476e-05,
-      "loss": 1.3139,
-      "step": 8460
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.869877338409424,
-      "learning_rate": 1.4596610169491527e-05,
-      "loss": 1.4249,
-      "step": 8470
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.0457489490509033,
-      "learning_rate": 1.4589830508474579e-05,
-      "loss": 1.2988,
-      "step": 8480
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 5.064341068267822,
-      "learning_rate": 1.458305084745763e-05,
-      "loss": 1.337,
-      "step": 8490
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.481555700302124,
-      "learning_rate": 1.4576271186440678e-05,
-      "loss": 1.2976,
-      "step": 8500
-    },
-    {
-      "epoch": 0.2,
-      "eval_loss": 1.0718435049057007,
-      "eval_runtime": 67.009,
-      "eval_samples_per_second": 14.923,
-      "eval_steps_per_second": 14.923,
-      "step": 8500
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 1.9016735553741455,
-      "learning_rate": 1.456949152542373e-05,
-      "loss": 1.3663,
-      "step": 8510
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 1.9207221269607544,
-      "learning_rate": 1.456271186440678e-05,
-      "loss": 1.2997,
-      "step": 8520
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.0371081829071045,
-      "learning_rate": 1.455593220338983e-05,
-      "loss": 1.3997,
-      "step": 8530
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.698136806488037,
-      "learning_rate": 1.4549152542372882e-05,
-      "loss": 1.4582,
-      "step": 8540
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 5.272952556610107,
-      "learning_rate": 1.4542372881355933e-05,
-      "loss": 1.2139,
-      "step": 8550
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.3560380935668945,
-      "learning_rate": 1.4535593220338983e-05,
-      "loss": 1.5006,
-      "step": 8560
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 4.281118869781494,
-      "learning_rate": 1.4528813559322035e-05,
-      "loss": 1.3091,
-      "step": 8570
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 5.719073295593262,
-      "learning_rate": 1.4522033898305086e-05,
-      "loss": 1.5326,
-      "step": 8580
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.063105821609497,
-      "learning_rate": 1.4515254237288137e-05,
-      "loss": 1.3689,
-      "step": 8590
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 1.946047306060791,
-      "learning_rate": 1.4508474576271187e-05,
-      "loss": 1.337,
-      "step": 8600
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 0.704504668712616,
-      "learning_rate": 1.4501694915254239e-05,
-      "loss": 1.2556,
-      "step": 8610
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 4.635929584503174,
-      "learning_rate": 1.449491525423729e-05,
-      "loss": 1.2115,
-      "step": 8620
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 1.6211448907852173,
-      "learning_rate": 1.4488135593220341e-05,
-      "loss": 1.4003,
-      "step": 8630
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.5351459980010986,
-      "learning_rate": 1.4481355932203391e-05,
-      "loss": 1.3165,
-      "step": 8640
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.730940580368042,
-      "learning_rate": 1.4474576271186442e-05,
-      "loss": 1.1764,
-      "step": 8650
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.209855794906616,
-      "learning_rate": 1.4467796610169494e-05,
-      "loss": 1.5537,
-      "step": 8660
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 5.097927570343018,
-      "learning_rate": 1.4461016949152544e-05,
-      "loss": 1.2082,
-      "step": 8670
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.625061273574829,
-      "learning_rate": 1.4454237288135595e-05,
-      "loss": 1.4377,
-      "step": 8680
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.1008729934692383,
-      "learning_rate": 1.4447457627118646e-05,
-      "loss": 1.3205,
-      "step": 8690
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.726665735244751,
-      "learning_rate": 1.4440677966101698e-05,
-      "loss": 1.5016,
-      "step": 8700
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.084998369216919,
-      "learning_rate": 1.4433898305084746e-05,
-      "loss": 1.3151,
-      "step": 8710
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 4.455216884613037,
-      "learning_rate": 1.4427118644067797e-05,
-      "loss": 1.2526,
-      "step": 8720
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.0726282596588135,
-      "learning_rate": 1.4420338983050849e-05,
-      "loss": 1.3079,
-      "step": 8730
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 6.472900867462158,
-      "learning_rate": 1.4413559322033898e-05,
-      "loss": 1.4239,
-      "step": 8740
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.418191909790039,
-      "learning_rate": 1.440677966101695e-05,
-      "loss": 1.1407,
-      "step": 8750
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 6.4082489013671875,
-      "learning_rate": 1.4400000000000001e-05,
-      "loss": 1.2607,
-      "step": 8760
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.1980984210968018,
-      "learning_rate": 1.4393220338983052e-05,
-      "loss": 1.4132,
-      "step": 8770
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 6.397686958312988,
-      "learning_rate": 1.4386440677966102e-05,
-      "loss": 1.4096,
-      "step": 8780
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 1.1709277629852295,
-      "learning_rate": 1.4379661016949154e-05,
-      "loss": 1.3932,
-      "step": 8790
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 1.7701913118362427,
-      "learning_rate": 1.4372881355932205e-05,
-      "loss": 1.4644,
-      "step": 8800
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 4.566888809204102,
-      "learning_rate": 1.4366101694915255e-05,
-      "loss": 1.4883,
-      "step": 8810
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 5.88577127456665,
-      "learning_rate": 1.4359322033898306e-05,
-      "loss": 1.4677,
-      "step": 8820
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 3.0660550594329834,
-      "learning_rate": 1.4352542372881357e-05,
-      "loss": 1.3425,
-      "step": 8830
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 5.181571006774902,
-      "learning_rate": 1.4345762711864409e-05,
-      "loss": 1.4117,
-      "step": 8840
-    },
-    {
-      "epoch": 0.2,
-      "grad_norm": 2.3248679637908936,
-      "learning_rate": 1.4338983050847459e-05,
-      "loss": 1.3787,
-      "step": 8850
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.677826404571533,
-      "learning_rate": 1.433220338983051e-05,
-      "loss": 1.3205,
-      "step": 8860
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.778449296951294,
-      "learning_rate": 1.4325423728813561e-05,
-      "loss": 1.183,
-      "step": 8870
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.0881848335266113,
-      "learning_rate": 1.4318644067796613e-05,
-      "loss": 1.4995,
-      "step": 8880
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 10.81136703491211,
-      "learning_rate": 1.4311864406779662e-05,
-      "loss": 1.3487,
-      "step": 8890
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 8.108224868774414,
-      "learning_rate": 1.4305084745762714e-05,
-      "loss": 1.492,
-      "step": 8900
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.3358426094055176,
-      "learning_rate": 1.4298305084745765e-05,
-      "loss": 1.4723,
-      "step": 8910
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.7977956533432007,
-      "learning_rate": 1.4291525423728813e-05,
-      "loss": 1.4662,
-      "step": 8920
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 5.2564921379089355,
-      "learning_rate": 1.4284745762711865e-05,
-      "loss": 1.2359,
-      "step": 8930
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 6.835478782653809,
-      "learning_rate": 1.4277966101694916e-05,
-      "loss": 1.5774,
-      "step": 8940
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 10.185200691223145,
-      "learning_rate": 1.4271186440677966e-05,
-      "loss": 1.5078,
-      "step": 8950
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.049248218536377,
-      "learning_rate": 1.4264406779661017e-05,
-      "loss": 1.3978,
-      "step": 8960
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.1511282920837402,
-      "learning_rate": 1.4257627118644069e-05,
-      "loss": 1.4465,
-      "step": 8970
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 4.345871925354004,
-      "learning_rate": 1.425084745762712e-05,
-      "loss": 1.401,
-      "step": 8980
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 4.231874942779541,
-      "learning_rate": 1.424406779661017e-05,
-      "loss": 1.2993,
-      "step": 8990
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.336160182952881,
-      "learning_rate": 1.4237288135593221e-05,
-      "loss": 1.3631,
-      "step": 9000
-    },
-    {
-      "epoch": 0.21,
-      "eval_loss": 1.0496437549591064,
-      "eval_runtime": 67.0488,
-      "eval_samples_per_second": 14.915,
-      "eval_steps_per_second": 14.915,
-      "step": 9000
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.6596415042877197,
-      "learning_rate": 1.4230508474576273e-05,
-      "loss": 1.3761,
-      "step": 9010
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.467890501022339,
-      "learning_rate": 1.4223728813559324e-05,
-      "loss": 1.459,
-      "step": 9020
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.3155555725097656,
-      "learning_rate": 1.4216949152542374e-05,
-      "loss": 1.3093,
-      "step": 9030
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.9326798915863037,
-      "learning_rate": 1.4210169491525425e-05,
-      "loss": 1.3302,
-      "step": 9040
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.875458002090454,
-      "learning_rate": 1.4203389830508476e-05,
-      "loss": 1.2721,
-      "step": 9050
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.327162742614746,
-      "learning_rate": 1.4196610169491528e-05,
-      "loss": 1.3188,
-      "step": 9060
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 6.240946292877197,
-      "learning_rate": 1.4189830508474578e-05,
-      "loss": 1.2654,
-      "step": 9070
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.747529983520508,
-      "learning_rate": 1.4183050847457629e-05,
-      "loss": 1.3903,
-      "step": 9080
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.1977028846740723,
-      "learning_rate": 1.417627118644068e-05,
-      "loss": 1.484,
-      "step": 9090
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.0423030853271484,
-      "learning_rate": 1.416949152542373e-05,
-      "loss": 1.2111,
-      "step": 9100
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.0299501419067383,
-      "learning_rate": 1.4162711864406781e-05,
-      "loss": 1.4146,
-      "step": 9110
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.8823587894439697,
-      "learning_rate": 1.4155932203389833e-05,
-      "loss": 1.3893,
-      "step": 9120
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.748755693435669,
-      "learning_rate": 1.414915254237288e-05,
-      "loss": 1.2844,
-      "step": 9130
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 7.603834629058838,
-      "learning_rate": 1.4142372881355932e-05,
-      "loss": 1.4416,
-      "step": 9140
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 4.06264591217041,
-      "learning_rate": 1.4135593220338984e-05,
-      "loss": 1.2758,
-      "step": 9150
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.5511293411254883,
-      "learning_rate": 1.4128813559322035e-05,
-      "loss": 1.3112,
-      "step": 9160
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.303243637084961,
-      "learning_rate": 1.4122033898305085e-05,
-      "loss": 1.2676,
-      "step": 9170
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 6.051310062408447,
-      "learning_rate": 1.4115254237288136e-05,
-      "loss": 1.2643,
-      "step": 9180
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 7.211860179901123,
-      "learning_rate": 1.4108474576271188e-05,
-      "loss": 1.3993,
-      "step": 9190
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.2717708349227905,
-      "learning_rate": 1.4101694915254239e-05,
-      "loss": 1.21,
-      "step": 9200
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 4.1961989402771,
-      "learning_rate": 1.4094915254237289e-05,
-      "loss": 1.1805,
-      "step": 9210
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.593219041824341,
-      "learning_rate": 1.408813559322034e-05,
-      "loss": 1.3931,
-      "step": 9220
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.4971346855163574,
-      "learning_rate": 1.4081355932203391e-05,
-      "loss": 1.4354,
-      "step": 9230
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 2.813567876815796,
-      "learning_rate": 1.4074576271186441e-05,
-      "loss": 1.1847,
-      "step": 9240
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.8391880989074707,
-      "learning_rate": 1.4067796610169493e-05,
-      "loss": 1.4731,
-      "step": 9250
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 6.610534191131592,
-      "learning_rate": 1.4061016949152544e-05,
-      "loss": 1.4042,
-      "step": 9260
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 1.301313042640686,
-      "learning_rate": 1.4054237288135595e-05,
-      "loss": 1.4343,
-      "step": 9270
-    },
-    {
-      "epoch": 0.21,
-      "grad_norm": 3.5929203033447266,
-      "learning_rate": 1.4047457627118645e-05,
-      "loss": 1.3178,
-      "step": 9280
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.077080249786377,
-      "learning_rate": 1.4040677966101696e-05,
-      "loss": 1.2344,
-      "step": 9290
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.020028591156006,
-      "learning_rate": 1.4033898305084748e-05,
-      "loss": 1.3188,
-      "step": 9300
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 1.7268164157867432,
-      "learning_rate": 1.40271186440678e-05,
-      "loss": 1.3322,
-      "step": 9310
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.74446177482605,
-      "learning_rate": 1.4020338983050849e-05,
-      "loss": 1.3143,
-      "step": 9320
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 8.375783920288086,
-      "learning_rate": 1.4013559322033899e-05,
-      "loss": 1.3217,
-      "step": 9330
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 1.9759814739227295,
-      "learning_rate": 1.400677966101695e-05,
-      "loss": 1.464,
-      "step": 9340
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 7.948102951049805,
-      "learning_rate": 1.4e-05,
-      "loss": 1.5633,
-      "step": 9350
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 1.1538234949111938,
-      "learning_rate": 1.3993220338983051e-05,
-      "loss": 1.2872,
-      "step": 9360
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.251887798309326,
-      "learning_rate": 1.3986440677966103e-05,
-      "loss": 1.3319,
-      "step": 9370
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 5.3007283210754395,
-      "learning_rate": 1.3979661016949152e-05,
-      "loss": 1.4477,
-      "step": 9380
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.742395877838135,
-      "learning_rate": 1.3972881355932204e-05,
-      "loss": 1.3017,
-      "step": 9390
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.886890411376953,
-      "learning_rate": 1.3966101694915255e-05,
-      "loss": 1.4781,
-      "step": 9400
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.440849781036377,
-      "learning_rate": 1.3959322033898306e-05,
-      "loss": 1.5151,
-      "step": 9410
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 1.388344168663025,
-      "learning_rate": 1.3952542372881356e-05,
-      "loss": 1.1664,
-      "step": 9420
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.4817326068878174,
-      "learning_rate": 1.3945762711864408e-05,
-      "loss": 1.3492,
-      "step": 9430
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.128695011138916,
-      "learning_rate": 1.3938983050847459e-05,
-      "loss": 1.3101,
-      "step": 9440
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 5.121028900146484,
-      "learning_rate": 1.393220338983051e-05,
-      "loss": 1.4188,
-      "step": 9450
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.124645471572876,
-      "learning_rate": 1.392542372881356e-05,
-      "loss": 1.1935,
-      "step": 9460
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.997901439666748,
-      "learning_rate": 1.3918644067796612e-05,
-      "loss": 1.2564,
-      "step": 9470
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.793147087097168,
-      "learning_rate": 1.3911864406779663e-05,
-      "loss": 1.402,
-      "step": 9480
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.9552319049835205,
-      "learning_rate": 1.3905084745762714e-05,
-      "loss": 1.5437,
-      "step": 9490
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 5.294174671173096,
-      "learning_rate": 1.3898305084745764e-05,
-      "loss": 1.4665,
-      "step": 9500
-    },
-    {
-      "epoch": 0.22,
-      "eval_loss": 1.102941632270813,
-      "eval_runtime": 67.0366,
-      "eval_samples_per_second": 14.917,
-      "eval_steps_per_second": 14.917,
-      "step": 9500
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 0.7909783124923706,
-      "learning_rate": 1.3891525423728815e-05,
-      "loss": 1.3756,
-      "step": 9510
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.4454197883605957,
-      "learning_rate": 1.3884745762711867e-05,
-      "loss": 1.3005,
-      "step": 9520
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.054439067840576,
-      "learning_rate": 1.3877966101694918e-05,
-      "loss": 1.4208,
-      "step": 9530
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 8.244361877441406,
-      "learning_rate": 1.3871186440677966e-05,
-      "loss": 1.3597,
-      "step": 9540
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.499704360961914,
-      "learning_rate": 1.3864406779661018e-05,
-      "loss": 1.3397,
-      "step": 9550
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.15438985824585,
-      "learning_rate": 1.3857627118644067e-05,
-      "loss": 1.3366,
-      "step": 9560
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 1.504563808441162,
-      "learning_rate": 1.3850847457627119e-05,
-      "loss": 1.4157,
-      "step": 9570
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 5.045544147491455,
-      "learning_rate": 1.384406779661017e-05,
-      "loss": 1.4529,
-      "step": 9580
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.8790090084075928,
-      "learning_rate": 1.3837288135593222e-05,
-      "loss": 1.2581,
-      "step": 9590
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 4.442860126495361,
-      "learning_rate": 1.3830508474576271e-05,
-      "loss": 1.3185,
-      "step": 9600
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.301625967025757,
-      "learning_rate": 1.3823728813559323e-05,
-      "loss": 1.519,
-      "step": 9610
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.222405195236206,
-      "learning_rate": 1.3816949152542374e-05,
-      "loss": 1.4471,
-      "step": 9620
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 6.543327808380127,
-      "learning_rate": 1.3810169491525425e-05,
-      "loss": 1.3623,
-      "step": 9630
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 7.1420392990112305,
-      "learning_rate": 1.3803389830508475e-05,
-      "loss": 1.1868,
-      "step": 9640
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 6.8577094078063965,
-      "learning_rate": 1.3796610169491527e-05,
-      "loss": 1.3238,
-      "step": 9650
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 3.2370662689208984,
-      "learning_rate": 1.3789830508474578e-05,
-      "loss": 1.2522,
-      "step": 9660
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.8609619140625,
-      "learning_rate": 1.378305084745763e-05,
-      "loss": 1.3918,
-      "step": 9670
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.962815523147583,
-      "learning_rate": 1.3776271186440679e-05,
-      "loss": 1.3291,
-      "step": 9680
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 1.649778962135315,
-      "learning_rate": 1.376949152542373e-05,
-      "loss": 1.4385,
-      "step": 9690
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.7426583766937256,
-      "learning_rate": 1.3762711864406782e-05,
-      "loss": 1.2947,
-      "step": 9700
-    },
-    {
-      "epoch": 0.22,
-      "grad_norm": 2.934476613998413,
-      "learning_rate": 1.3755932203389832e-05,
-      "loss": 1.4168,
-      "step": 9710
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 5.4620747566223145,
-      "learning_rate": 1.3749152542372883e-05,
-      "loss": 1.3825,
-      "step": 9720
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.6670947074890137,
-      "learning_rate": 1.3742372881355934e-05,
-      "loss": 1.1969,
-      "step": 9730
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 5.469496726989746,
-      "learning_rate": 1.3735593220338986e-05,
-      "loss": 1.0666,
-      "step": 9740
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.448007583618164,
-      "learning_rate": 1.3728813559322034e-05,
-      "loss": 1.4138,
-      "step": 9750
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.5976903438568115,
-      "learning_rate": 1.3722033898305085e-05,
-      "loss": 1.4822,
-      "step": 9760
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.9508200883865356,
-      "learning_rate": 1.3715254237288137e-05,
-      "loss": 1.2946,
-      "step": 9770
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.993997812271118,
-      "learning_rate": 1.3708474576271186e-05,
-      "loss": 1.4221,
-      "step": 9780
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.5066683292388916,
-      "learning_rate": 1.3701694915254238e-05,
-      "loss": 1.4399,
-      "step": 9790
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.8072290420532227,
-      "learning_rate": 1.3694915254237289e-05,
-      "loss": 1.2055,
-      "step": 9800
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 5.655754566192627,
-      "learning_rate": 1.368813559322034e-05,
-      "loss": 1.3866,
-      "step": 9810
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 15.120744705200195,
-      "learning_rate": 1.368135593220339e-05,
-      "loss": 1.1179,
-      "step": 9820
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.2080466747283936,
-      "learning_rate": 1.3674576271186442e-05,
-      "loss": 1.412,
-      "step": 9830
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.3344991207122803,
-      "learning_rate": 1.3667796610169493e-05,
-      "loss": 1.3185,
-      "step": 9840
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.8925857543945312,
-      "learning_rate": 1.3661016949152543e-05,
-      "loss": 1.3284,
-      "step": 9850
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 6.113729476928711,
-      "learning_rate": 1.3654237288135594e-05,
-      "loss": 1.3287,
-      "step": 9860
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 8.377769470214844,
-      "learning_rate": 1.3647457627118646e-05,
-      "loss": 1.4654,
-      "step": 9870
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 3.898707866668701,
-      "learning_rate": 1.3640677966101697e-05,
-      "loss": 1.3557,
-      "step": 9880
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 3.4954209327697754,
-      "learning_rate": 1.3633898305084747e-05,
-      "loss": 1.2589,
-      "step": 9890
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.97265362739563,
-      "learning_rate": 1.3627118644067798e-05,
-      "loss": 1.4468,
-      "step": 9900
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 4.927505970001221,
-      "learning_rate": 1.362033898305085e-05,
-      "loss": 1.3386,
-      "step": 9910
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 5.02236795425415,
-      "learning_rate": 1.36135593220339e-05,
-      "loss": 1.6745,
-      "step": 9920
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 3.2489023208618164,
-      "learning_rate": 1.360677966101695e-05,
-      "loss": 1.3722,
-      "step": 9930
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 4.043548583984375,
-      "learning_rate": 1.3600000000000002e-05,
-      "loss": 1.3996,
-      "step": 9940
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.7545967102050781,
-      "learning_rate": 1.3593220338983053e-05,
-      "loss": 1.3781,
-      "step": 9950
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.4097864627838135,
-      "learning_rate": 1.3586440677966101e-05,
-      "loss": 1.3649,
-      "step": 9960
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.1898608207702637,
-      "learning_rate": 1.3579661016949153e-05,
-      "loss": 1.4705,
-      "step": 9970
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 5.296695232391357,
-      "learning_rate": 1.3572881355932204e-05,
-      "loss": 1.2701,
-      "step": 9980
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.6690765619277954,
-      "learning_rate": 1.3566101694915254e-05,
-      "loss": 1.2846,
-      "step": 9990
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.7802351713180542,
-      "learning_rate": 1.3559322033898305e-05,
-      "loss": 1.3004,
-      "step": 10000
-    },
-    {
-      "epoch": 0.23,
-      "eval_loss": 1.0849580764770508,
-      "eval_runtime": 67.0014,
-      "eval_samples_per_second": 14.925,
-      "eval_steps_per_second": 14.925,
-      "step": 10000
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.575228214263916,
-      "learning_rate": 1.3552542372881357e-05,
-      "loss": 1.2472,
-      "step": 10010
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 3.6964104175567627,
-      "learning_rate": 1.3545762711864408e-05,
-      "loss": 1.376,
-      "step": 10020
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.1188055276870728,
-      "learning_rate": 1.3538983050847458e-05,
-      "loss": 1.1798,
-      "step": 10030
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 6.469828128814697,
-      "learning_rate": 1.353220338983051e-05,
-      "loss": 1.3331,
-      "step": 10040
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 6.409282207489014,
-      "learning_rate": 1.352542372881356e-05,
-      "loss": 1.3283,
-      "step": 10050
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 3.001148223876953,
-      "learning_rate": 1.3518644067796612e-05,
-      "loss": 1.4138,
-      "step": 10060
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.1178245544433594,
-      "learning_rate": 1.3511864406779662e-05,
-      "loss": 1.0088,
-      "step": 10070
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.2544403076171875,
-      "learning_rate": 1.3505084745762713e-05,
-      "loss": 1.2236,
-      "step": 10080
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 4.784693241119385,
-      "learning_rate": 1.3498305084745764e-05,
-      "loss": 1.3831,
-      "step": 10090
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 2.8957157135009766,
-      "learning_rate": 1.3491525423728816e-05,
-      "loss": 1.4073,
-      "step": 10100
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.957875370979309,
-      "learning_rate": 1.3484745762711866e-05,
-      "loss": 1.4168,
-      "step": 10110
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 3.314802885055542,
-      "learning_rate": 1.3477966101694917e-05,
-      "loss": 1.2289,
-      "step": 10120
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 3.6212656497955322,
-      "learning_rate": 1.3471186440677968e-05,
-      "loss": 1.5181,
-      "step": 10130
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 1.864043116569519,
-      "learning_rate": 1.346440677966102e-05,
-      "loss": 1.4954,
-      "step": 10140
-    },
-    {
-      "epoch": 0.23,
-      "grad_norm": 4.504016876220703,
-      "learning_rate": 1.345762711864407e-05,
-      "loss": 1.2792,
-      "step": 10150
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 4.482726573944092,
-      "learning_rate": 1.3450847457627121e-05,
-      "loss": 1.0474,
-      "step": 10160
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.6446313858032227,
-      "learning_rate": 1.3444067796610169e-05,
-      "loss": 1.4683,
-      "step": 10170
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 1.929124116897583,
-      "learning_rate": 1.343728813559322e-05,
-      "loss": 1.4199,
-      "step": 10180
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.283059597015381,
-      "learning_rate": 1.3430508474576272e-05,
-      "loss": 1.2784,
-      "step": 10190
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.409111738204956,
-      "learning_rate": 1.3423728813559323e-05,
-      "loss": 1.393,
-      "step": 10200
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 6.90030574798584,
-      "learning_rate": 1.3416949152542373e-05,
-      "loss": 1.2889,
-      "step": 10210
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.19612455368042,
-      "learning_rate": 1.3410169491525424e-05,
-      "loss": 1.646,
-      "step": 10220
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 7.8987650871276855,
-      "learning_rate": 1.3403389830508476e-05,
-      "loss": 1.2646,
-      "step": 10230
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.225478172302246,
-      "learning_rate": 1.3396610169491527e-05,
-      "loss": 1.2869,
-      "step": 10240
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 3.281136989593506,
-      "learning_rate": 1.3389830508474577e-05,
-      "loss": 1.3977,
-      "step": 10250
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.4851903915405273,
-      "learning_rate": 1.3383050847457628e-05,
-      "loss": 1.0693,
-      "step": 10260
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 3.3886020183563232,
-      "learning_rate": 1.337627118644068e-05,
-      "loss": 1.3299,
-      "step": 10270
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 5.062967300415039,
-      "learning_rate": 1.3369491525423731e-05,
-      "loss": 1.4647,
-      "step": 10280
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 5.308380126953125,
-      "learning_rate": 1.336271186440678e-05,
-      "loss": 1.3072,
-      "step": 10290
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.1562750339508057,
-      "learning_rate": 1.3355932203389832e-05,
-      "loss": 1.4808,
-      "step": 10300
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 0.7641376852989197,
-      "learning_rate": 1.3349152542372883e-05,
-      "loss": 1.2411,
-      "step": 10310
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 4.490609169006348,
-      "learning_rate": 1.3342372881355933e-05,
-      "loss": 1.4156,
-      "step": 10320
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 1.7860184907913208,
-      "learning_rate": 1.3335593220338985e-05,
-      "loss": 1.35,
-      "step": 10330
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 5.641195297241211,
-      "learning_rate": 1.3328813559322036e-05,
-      "loss": 1.3174,
-      "step": 10340
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.737027406692505,
-      "learning_rate": 1.3322033898305087e-05,
-      "loss": 1.2823,
-      "step": 10350
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 5.195732116699219,
-      "learning_rate": 1.3315254237288137e-05,
-      "loss": 1.2259,
-      "step": 10360
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.1460018157958984,
-      "learning_rate": 1.3308474576271187e-05,
-      "loss": 1.0539,
-      "step": 10370
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 4.896897315979004,
-      "learning_rate": 1.3301694915254238e-05,
-      "loss": 1.2379,
-      "step": 10380
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 4.282651901245117,
-      "learning_rate": 1.3294915254237288e-05,
-      "loss": 1.4351,
-      "step": 10390
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.8991894721984863,
-      "learning_rate": 1.328813559322034e-05,
-      "loss": 1.3508,
-      "step": 10400
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 4.306196212768555,
-      "learning_rate": 1.328135593220339e-05,
-      "loss": 1.3804,
-      "step": 10410
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.230379104614258,
-      "learning_rate": 1.3274576271186442e-05,
-      "loss": 1.2072,
-      "step": 10420
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 1.0265578031539917,
-      "learning_rate": 1.3267796610169492e-05,
-      "loss": 1.3691,
-      "step": 10430
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 1.1750489473342896,
-      "learning_rate": 1.3261016949152543e-05,
-      "loss": 1.2741,
-      "step": 10440
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 6.429545879364014,
-      "learning_rate": 1.3254237288135595e-05,
-      "loss": 1.4234,
-      "step": 10450
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 8.749063491821289,
-      "learning_rate": 1.3247457627118644e-05,
-      "loss": 1.3652,
-      "step": 10460
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 3.783092498779297,
-      "learning_rate": 1.3240677966101696e-05,
-      "loss": 1.2662,
-      "step": 10470
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 7.148987293243408,
-      "learning_rate": 1.3233898305084747e-05,
-      "loss": 1.2553,
-      "step": 10480
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 4.071252822875977,
-      "learning_rate": 1.3227118644067798e-05,
-      "loss": 1.1767,
-      "step": 10490
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 4.911563396453857,
-      "learning_rate": 1.3220338983050848e-05,
-      "loss": 1.4,
-      "step": 10500
-    },
-    {
-      "epoch": 0.24,
-      "eval_loss": 1.056018590927124,
-      "eval_runtime": 67.0592,
-      "eval_samples_per_second": 14.912,
-      "eval_steps_per_second": 14.912,
-      "step": 10500
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.0508599281311035,
-      "learning_rate": 1.32135593220339e-05,
-      "loss": 1.3227,
-      "step": 10510
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 2.7030413150787354,
-      "learning_rate": 1.3206779661016951e-05,
-      "loss": 1.4168,
-      "step": 10520
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 8.267769813537598,
-      "learning_rate": 1.3200000000000002e-05,
-      "loss": 1.4354,
-      "step": 10530
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 3.2510886192321777,
-      "learning_rate": 1.3193220338983052e-05,
-      "loss": 1.3922,
-      "step": 10540
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 10.665681838989258,
-      "learning_rate": 1.3186440677966103e-05,
-      "loss": 1.3282,
-      "step": 10550
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 1.7952899932861328,
-      "learning_rate": 1.3179661016949155e-05,
-      "loss": 1.2216,
-      "step": 10560
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 3.8615293502807617,
-      "learning_rate": 1.3172881355932206e-05,
-      "loss": 1.3105,
-      "step": 10570
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 5.162840366363525,
-      "learning_rate": 1.3166101694915254e-05,
-      "loss": 1.3709,
-      "step": 10580
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 2.7145211696624756,
-      "learning_rate": 1.3159322033898306e-05,
-      "loss": 1.3356,
-      "step": 10590
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 2.6916394233703613,
-      "learning_rate": 1.3152542372881355e-05,
-      "loss": 1.4327,
-      "step": 10600
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.5745512247085571,
-      "learning_rate": 1.3145762711864407e-05,
-      "loss": 1.4339,
-      "step": 10610
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.7034305334091187,
-      "learning_rate": 1.3138983050847458e-05,
-      "loss": 1.3206,
-      "step": 10620
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.4381961822509766,
-      "learning_rate": 1.313220338983051e-05,
-      "loss": 1.3502,
-      "step": 10630
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.2712607383728027,
-      "learning_rate": 1.312542372881356e-05,
-      "loss": 1.3906,
-      "step": 10640
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 4.805257320404053,
-      "learning_rate": 1.311864406779661e-05,
-      "loss": 1.407,
-      "step": 10650
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.6435729265213013,
-      "learning_rate": 1.3111864406779662e-05,
-      "loss": 1.4471,
-      "step": 10660
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.9431121349334717,
-      "learning_rate": 1.3105084745762714e-05,
-      "loss": 1.2493,
-      "step": 10670
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.9565777778625488,
-      "learning_rate": 1.3098305084745763e-05,
-      "loss": 1.4396,
-      "step": 10680
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 5.754490852355957,
-      "learning_rate": 1.3091525423728815e-05,
-      "loss": 1.3714,
-      "step": 10690
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.628016233444214,
-      "learning_rate": 1.3084745762711866e-05,
-      "loss": 1.2535,
-      "step": 10700
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.5335326194763184,
-      "learning_rate": 1.3077966101694917e-05,
-      "loss": 1.4952,
-      "step": 10710
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 4.725300312042236,
-      "learning_rate": 1.3071186440677967e-05,
-      "loss": 1.2693,
-      "step": 10720
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.5443134307861328,
-      "learning_rate": 1.3064406779661019e-05,
-      "loss": 1.2728,
-      "step": 10730
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.495453119277954,
-      "learning_rate": 1.305762711864407e-05,
-      "loss": 1.4771,
-      "step": 10740
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.3322789669036865,
-      "learning_rate": 1.305084745762712e-05,
-      "loss": 1.4256,
-      "step": 10750
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 7.982933521270752,
-      "learning_rate": 1.3044067796610171e-05,
-      "loss": 1.2098,
-      "step": 10760
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.701568603515625,
-      "learning_rate": 1.3037288135593222e-05,
-      "loss": 1.0726,
-      "step": 10770
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.6933153867721558,
-      "learning_rate": 1.3030508474576274e-05,
-      "loss": 1.5174,
-      "step": 10780
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.9348666667938232,
-      "learning_rate": 1.3023728813559322e-05,
-      "loss": 1.3662,
-      "step": 10790
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 7.25183629989624,
-      "learning_rate": 1.3016949152542373e-05,
-      "loss": 1.2264,
-      "step": 10800
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 4.727850914001465,
-      "learning_rate": 1.3010169491525425e-05,
-      "loss": 1.2664,
-      "step": 10810
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.9714324474334717,
-      "learning_rate": 1.3003389830508474e-05,
-      "loss": 1.316,
-      "step": 10820
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.8657814264297485,
-      "learning_rate": 1.2996610169491526e-05,
-      "loss": 1.5116,
-      "step": 10830
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 4.508814811706543,
-      "learning_rate": 1.2989830508474577e-05,
-      "loss": 1.2307,
-      "step": 10840
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.6838784217834473,
-      "learning_rate": 1.2983050847457629e-05,
-      "loss": 1.1877,
-      "step": 10850
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 4.849525451660156,
-      "learning_rate": 1.2976271186440678e-05,
-      "loss": 1.2414,
-      "step": 10860
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.3128011226654053,
-      "learning_rate": 1.296949152542373e-05,
-      "loss": 1.3896,
-      "step": 10870
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 2.422858238220215,
-      "learning_rate": 1.2962711864406781e-05,
-      "loss": 1.2159,
-      "step": 10880
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 4.698125839233398,
-      "learning_rate": 1.295593220338983e-05,
-      "loss": 1.276,
-      "step": 10890
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 5.102774143218994,
-      "learning_rate": 1.2949152542372882e-05,
-      "loss": 1.2338,
-      "step": 10900
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 2.2019145488739014,
-      "learning_rate": 1.2942372881355934e-05,
-      "loss": 1.2739,
-      "step": 10910
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 9.765484809875488,
-      "learning_rate": 1.2935593220338985e-05,
-      "loss": 1.3403,
-      "step": 10920
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 3.763749837875366,
-      "learning_rate": 1.2928813559322035e-05,
-      "loss": 1.3235,
-      "step": 10930
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 2.33292818069458,
-      "learning_rate": 1.2922033898305086e-05,
-      "loss": 1.4525,
-      "step": 10940
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 5.402053356170654,
-      "learning_rate": 1.2915254237288137e-05,
-      "loss": 1.2736,
-      "step": 10950
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 8.462284088134766,
-      "learning_rate": 1.2908474576271189e-05,
-      "loss": 1.302,
-      "step": 10960
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 2.865405321121216,
-      "learning_rate": 1.2901694915254239e-05,
-      "loss": 1.3355,
-      "step": 10970
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.8566632270812988,
-      "learning_rate": 1.289491525423729e-05,
-      "loss": 1.0694,
-      "step": 10980
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 4.951323986053467,
-      "learning_rate": 1.2888135593220341e-05,
-      "loss": 1.3189,
-      "step": 10990
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 2.248414993286133,
-      "learning_rate": 1.288135593220339e-05,
-      "loss": 1.3578,
-      "step": 11000
-    },
-    {
-      "epoch": 0.25,
-      "eval_loss": 1.1179373264312744,
-      "eval_runtime": 67.0257,
-      "eval_samples_per_second": 14.92,
-      "eval_steps_per_second": 14.92,
-      "step": 11000
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 1.7238454818725586,
-      "learning_rate": 1.287457627118644e-05,
-      "loss": 1.2867,
-      "step": 11010
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.1348774433135986,
-      "learning_rate": 1.2867796610169492e-05,
-      "loss": 1.3616,
-      "step": 11020
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 5.413891315460205,
-      "learning_rate": 1.2861016949152542e-05,
-      "loss": 1.358,
-      "step": 11030
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.0301661491394043,
-      "learning_rate": 1.2854237288135593e-05,
-      "loss": 1.4766,
-      "step": 11040
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.3625426292419434,
-      "learning_rate": 1.2847457627118645e-05,
-      "loss": 1.2889,
-      "step": 11050
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 6.010834217071533,
-      "learning_rate": 1.2840677966101696e-05,
-      "loss": 1.3095,
-      "step": 11060
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 1.5633188486099243,
-      "learning_rate": 1.2833898305084746e-05,
-      "loss": 1.3397,
-      "step": 11070
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.8885929584503174,
-      "learning_rate": 1.2827118644067797e-05,
-      "loss": 1.3388,
-      "step": 11080
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 4.556473255157471,
-      "learning_rate": 1.2820338983050849e-05,
-      "loss": 1.2801,
-      "step": 11090
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.5699386596679688,
-      "learning_rate": 1.28135593220339e-05,
-      "loss": 1.2587,
-      "step": 11100
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.842324733734131,
-      "learning_rate": 1.280677966101695e-05,
-      "loss": 1.2892,
-      "step": 11110
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 1.6393495798110962,
-      "learning_rate": 1.2800000000000001e-05,
-      "loss": 1.2888,
-      "step": 11120
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 4.156044960021973,
-      "learning_rate": 1.2793220338983053e-05,
-      "loss": 1.4945,
-      "step": 11130
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 5.133569717407227,
-      "learning_rate": 1.2786440677966104e-05,
-      "loss": 1.2537,
-      "step": 11140
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 6.207439422607422,
-      "learning_rate": 1.2779661016949154e-05,
-      "loss": 1.2773,
-      "step": 11150
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 7.413145542144775,
-      "learning_rate": 1.2772881355932205e-05,
-      "loss": 1.3265,
-      "step": 11160
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.2502923011779785,
-      "learning_rate": 1.2766101694915256e-05,
-      "loss": 1.1874,
-      "step": 11170
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.847926139831543,
-      "learning_rate": 1.2759322033898308e-05,
-      "loss": 1.2997,
-      "step": 11180
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 5.950617790222168,
-      "learning_rate": 1.2752542372881358e-05,
-      "loss": 1.2431,
-      "step": 11190
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.7104105949401855,
-      "learning_rate": 1.2745762711864407e-05,
-      "loss": 1.2631,
-      "step": 11200
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.444988965988159,
-      "learning_rate": 1.2738983050847457e-05,
-      "loss": 1.3001,
-      "step": 11210
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.375910520553589,
-      "learning_rate": 1.2732203389830508e-05,
-      "loss": 1.4155,
-      "step": 11220
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 6.275600433349609,
-      "learning_rate": 1.272542372881356e-05,
-      "loss": 1.2824,
-      "step": 11230
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 5.417289733886719,
-      "learning_rate": 1.2718644067796611e-05,
-      "loss": 1.4936,
-      "step": 11240
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.149083375930786,
-      "learning_rate": 1.2711864406779661e-05,
-      "loss": 1.4582,
-      "step": 11250
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 1.4166737794876099,
-      "learning_rate": 1.2705084745762712e-05,
-      "loss": 1.274,
-      "step": 11260
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 4.054012775421143,
-      "learning_rate": 1.2698305084745764e-05,
-      "loss": 1.4788,
-      "step": 11270
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.4033005237579346,
-      "learning_rate": 1.2691525423728815e-05,
-      "loss": 1.2238,
-      "step": 11280
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.9126157760620117,
-      "learning_rate": 1.2684745762711865e-05,
-      "loss": 1.283,
-      "step": 11290
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 4.059444427490234,
-      "learning_rate": 1.2677966101694916e-05,
-      "loss": 1.3338,
-      "step": 11300
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 9.896828651428223,
-      "learning_rate": 1.2671186440677968e-05,
-      "loss": 1.5215,
-      "step": 11310
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.37363862991333,
-      "learning_rate": 1.2664406779661019e-05,
-      "loss": 1.4575,
-      "step": 11320
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 4.463009834289551,
-      "learning_rate": 1.2657627118644069e-05,
-      "loss": 1.3001,
-      "step": 11330
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 5.098323345184326,
-      "learning_rate": 1.265084745762712e-05,
-      "loss": 1.3776,
-      "step": 11340
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 8.601290702819824,
-      "learning_rate": 1.2644067796610171e-05,
-      "loss": 1.3683,
-      "step": 11350
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 7.4985880851745605,
-      "learning_rate": 1.2637288135593221e-05,
-      "loss": 1.3786,
-      "step": 11360
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.4713215827941895,
-      "learning_rate": 1.2630508474576273e-05,
-      "loss": 1.4187,
-      "step": 11370
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.570042610168457,
-      "learning_rate": 1.2623728813559324e-05,
-      "loss": 1.3496,
-      "step": 11380
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 2.476733446121216,
-      "learning_rate": 1.2616949152542375e-05,
-      "loss": 1.3347,
-      "step": 11390
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 4.002094745635986,
-      "learning_rate": 1.2610169491525425e-05,
-      "loss": 1.4025,
-      "step": 11400
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 7.079920768737793,
-      "learning_rate": 1.2603389830508475e-05,
-      "loss": 1.4474,
-      "step": 11410
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 3.0040886402130127,
-      "learning_rate": 1.2596610169491526e-05,
-      "loss": 1.3848,
-      "step": 11420
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 5.5370612144470215,
-      "learning_rate": 1.2589830508474576e-05,
-      "loss": 1.3444,
-      "step": 11430
-    },
-    {
-      "epoch": 0.26,
-      "grad_norm": 1.4202508926391602,
-      "learning_rate": 1.2583050847457627e-05,
-      "loss": 1.2744,
-      "step": 11440
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.522201061248779,
-      "learning_rate": 1.2576271186440679e-05,
-      "loss": 1.1479,
-      "step": 11450
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 4.666895866394043,
-      "learning_rate": 1.256949152542373e-05,
-      "loss": 1.3711,
-      "step": 11460
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 4.038332462310791,
-      "learning_rate": 1.256271186440678e-05,
-      "loss": 1.0849,
-      "step": 11470
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 1.2697900533676147,
-      "learning_rate": 1.2555932203389831e-05,
-      "loss": 1.2352,
-      "step": 11480
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 1.8700276613235474,
-      "learning_rate": 1.2549152542372883e-05,
-      "loss": 1.2237,
-      "step": 11490
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 4.803176403045654,
-      "learning_rate": 1.2542372881355932e-05,
-      "loss": 1.3036,
-      "step": 11500
-    },
-    {
-      "epoch": 0.27,
-      "eval_loss": 1.0648910999298096,
-      "eval_runtime": 66.9985,
-      "eval_samples_per_second": 14.926,
-      "eval_steps_per_second": 14.926,
-      "step": 11500
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.106827974319458,
-      "learning_rate": 1.2535593220338984e-05,
-      "loss": 1.3212,
-      "step": 11510
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.4768242835998535,
-      "learning_rate": 1.2528813559322035e-05,
-      "loss": 1.4306,
-      "step": 11520
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 17.52560043334961,
-      "learning_rate": 1.2522033898305087e-05,
-      "loss": 1.5017,
-      "step": 11530
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.7506866455078125,
-      "learning_rate": 1.2515254237288136e-05,
-      "loss": 1.2379,
-      "step": 11540
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.74631929397583,
-      "learning_rate": 1.2508474576271188e-05,
-      "loss": 1.3201,
-      "step": 11550
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 1.5436533689498901,
-      "learning_rate": 1.2501694915254239e-05,
-      "loss": 1.3991,
-      "step": 11560
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.3251969814300537,
-      "learning_rate": 1.249491525423729e-05,
-      "loss": 1.5298,
-      "step": 11570
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.3390953540802,
-      "learning_rate": 1.248813559322034e-05,
-      "loss": 1.272,
-      "step": 11580
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.70845627784729,
-      "learning_rate": 1.2481355932203392e-05,
-      "loss": 1.235,
-      "step": 11590
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 6.170680999755859,
-      "learning_rate": 1.2474576271186443e-05,
-      "loss": 1.3834,
-      "step": 11600
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 1.4609490633010864,
-      "learning_rate": 1.2467796610169494e-05,
-      "loss": 1.2213,
-      "step": 11610
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.1652746200561523,
-      "learning_rate": 1.2461016949152542e-05,
-      "loss": 1.4787,
-      "step": 11620
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 7.743651866912842,
-      "learning_rate": 1.2454237288135594e-05,
-      "loss": 1.476,
-      "step": 11630
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.615380048751831,
-      "learning_rate": 1.2447457627118643e-05,
-      "loss": 1.2527,
-      "step": 11640
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.593022346496582,
-      "learning_rate": 1.2440677966101695e-05,
-      "loss": 1.3468,
-      "step": 11650
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 1.5045115947723389,
-      "learning_rate": 1.2433898305084746e-05,
-      "loss": 1.4336,
-      "step": 11660
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 6.107186317443848,
-      "learning_rate": 1.2427118644067798e-05,
-      "loss": 1.4119,
-      "step": 11670
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.435987710952759,
-      "learning_rate": 1.2420338983050847e-05,
-      "loss": 1.1495,
-      "step": 11680
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.024228096008301,
-      "learning_rate": 1.2413559322033899e-05,
-      "loss": 1.3461,
-      "step": 11690
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.844212293624878,
-      "learning_rate": 1.240677966101695e-05,
-      "loss": 1.2638,
-      "step": 11700
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.6562612056732178,
-      "learning_rate": 1.2400000000000002e-05,
-      "loss": 1.1978,
-      "step": 11710
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.006361484527588,
-      "learning_rate": 1.2393220338983051e-05,
-      "loss": 1.3342,
-      "step": 11720
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.448744297027588,
-      "learning_rate": 1.2386440677966103e-05,
-      "loss": 1.3128,
-      "step": 11730
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.1291112899780273,
-      "learning_rate": 1.2379661016949154e-05,
-      "loss": 1.1585,
-      "step": 11740
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 1.8178492784500122,
-      "learning_rate": 1.2372881355932205e-05,
-      "loss": 1.4148,
-      "step": 11750
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 9.099708557128906,
-      "learning_rate": 1.2366101694915255e-05,
-      "loss": 1.275,
-      "step": 11760
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 4.819830894470215,
-      "learning_rate": 1.2359322033898307e-05,
-      "loss": 1.2382,
-      "step": 11770
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.805961847305298,
-      "learning_rate": 1.2352542372881358e-05,
-      "loss": 1.2934,
-      "step": 11780
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.781748294830322,
-      "learning_rate": 1.234576271186441e-05,
-      "loss": 1.2673,
-      "step": 11790
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.577550411224365,
-      "learning_rate": 1.2338983050847459e-05,
-      "loss": 1.4464,
-      "step": 11800
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.494887351989746,
-      "learning_rate": 1.233220338983051e-05,
-      "loss": 1.5913,
-      "step": 11810
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.071074962615967,
-      "learning_rate": 1.2325423728813562e-05,
-      "loss": 1.1268,
-      "step": 11820
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 2.8962059020996094,
-      "learning_rate": 1.231864406779661e-05,
-      "loss": 1.2039,
-      "step": 11830
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 5.128442287445068,
-      "learning_rate": 1.2311864406779661e-05,
-      "loss": 1.3777,
-      "step": 11840
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.5109148025512695,
-      "learning_rate": 1.2305084745762713e-05,
-      "loss": 1.3644,
-      "step": 11850
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 3.0351719856262207,
-      "learning_rate": 1.2298305084745762e-05,
-      "loss": 1.3146,
-      "step": 11860
-    },
-    {
-      "epoch": 0.27,
-      "grad_norm": 4.8273749351501465,
-      "learning_rate": 1.2291525423728814e-05,
-      "loss": 1.278,
-      "step": 11870
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.8249599933624268,
-      "learning_rate": 1.2284745762711865e-05,
-      "loss": 1.2473,
-      "step": 11880
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.006662607192993,
-      "learning_rate": 1.2277966101694917e-05,
-      "loss": 1.2185,
-      "step": 11890
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 5.592995643615723,
-      "learning_rate": 1.2271186440677966e-05,
-      "loss": 1.3326,
-      "step": 11900
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.6979246139526367,
-      "learning_rate": 1.2264406779661018e-05,
-      "loss": 1.2077,
-      "step": 11910
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 7.2336812019348145,
-      "learning_rate": 1.2257627118644069e-05,
-      "loss": 1.3997,
-      "step": 11920
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 11.921634674072266,
-      "learning_rate": 1.225084745762712e-05,
-      "loss": 1.3697,
-      "step": 11930
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.800546646118164,
-      "learning_rate": 1.224406779661017e-05,
-      "loss": 1.3643,
-      "step": 11940
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 8.160839080810547,
-      "learning_rate": 1.2237288135593222e-05,
-      "loss": 1.3598,
-      "step": 11950
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.9122064113616943,
-      "learning_rate": 1.2230508474576273e-05,
-      "loss": 1.4034,
-      "step": 11960
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.0154528617858887,
-      "learning_rate": 1.2223728813559323e-05,
-      "loss": 1.4547,
-      "step": 11970
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.0988502502441406,
-      "learning_rate": 1.2216949152542374e-05,
-      "loss": 1.2566,
-      "step": 11980
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 4.406126976013184,
-      "learning_rate": 1.2210169491525426e-05,
-      "loss": 1.2769,
-      "step": 11990
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.9631588459014893,
-      "learning_rate": 1.2203389830508477e-05,
-      "loss": 1.2008,
-      "step": 12000
-    },
-    {
-      "epoch": 0.28,
-      "eval_loss": 1.0481234788894653,
-      "eval_runtime": 67.0601,
-      "eval_samples_per_second": 14.912,
-      "eval_steps_per_second": 14.912,
-      "step": 12000
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 1.6879490613937378,
-      "learning_rate": 1.2196610169491527e-05,
-      "loss": 1.318,
-      "step": 12010
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 1.431481957435608,
-      "learning_rate": 1.2189830508474578e-05,
-      "loss": 1.1515,
-      "step": 12020
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.4427268505096436,
-      "learning_rate": 1.2183050847457628e-05,
-      "loss": 1.3034,
-      "step": 12030
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.564251184463501,
-      "learning_rate": 1.2176271186440677e-05,
-      "loss": 1.4187,
-      "step": 12040
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 4.561717510223389,
-      "learning_rate": 1.2169491525423729e-05,
-      "loss": 1.2476,
-      "step": 12050
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.120867967605591,
-      "learning_rate": 1.216271186440678e-05,
-      "loss": 1.3806,
-      "step": 12060
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.209594964981079,
-      "learning_rate": 1.2155932203389832e-05,
-      "loss": 1.384,
-      "step": 12070
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.614448070526123,
-      "learning_rate": 1.2149152542372881e-05,
-      "loss": 1.3525,
-      "step": 12080
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 5.756226062774658,
-      "learning_rate": 1.2142372881355933e-05,
-      "loss": 1.2727,
-      "step": 12090
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.3497910499572754,
-      "learning_rate": 1.2135593220338984e-05,
-      "loss": 1.2477,
-      "step": 12100
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 4.628206729888916,
-      "learning_rate": 1.2128813559322034e-05,
-      "loss": 1.2596,
-      "step": 12110
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.101430654525757,
-      "learning_rate": 1.2122033898305085e-05,
-      "loss": 1.4617,
-      "step": 12120
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 6.928171157836914,
-      "learning_rate": 1.2115254237288137e-05,
-      "loss": 1.1119,
-      "step": 12130
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 8.110835075378418,
-      "learning_rate": 1.2108474576271188e-05,
-      "loss": 1.365,
-      "step": 12140
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.325059652328491,
-      "learning_rate": 1.2101694915254238e-05,
-      "loss": 1.1705,
-      "step": 12150
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 6.017369747161865,
-      "learning_rate": 1.209491525423729e-05,
-      "loss": 1.1434,
-      "step": 12160
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.4774413108825684,
-      "learning_rate": 1.208813559322034e-05,
-      "loss": 1.4273,
-      "step": 12170
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.168473482131958,
-      "learning_rate": 1.2081355932203392e-05,
-      "loss": 1.2839,
-      "step": 12180
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 7.533470153808594,
-      "learning_rate": 1.2074576271186442e-05,
-      "loss": 1.3452,
-      "step": 12190
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 5.115807056427002,
-      "learning_rate": 1.2067796610169493e-05,
-      "loss": 1.3136,
-      "step": 12200
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 6.920198917388916,
-      "learning_rate": 1.2061016949152544e-05,
-      "loss": 1.3263,
-      "step": 12210
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 4.812027454376221,
-      "learning_rate": 1.2054237288135596e-05,
-      "loss": 1.1978,
-      "step": 12220
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 4.1168107986450195,
-      "learning_rate": 1.2047457627118646e-05,
-      "loss": 1.3051,
-      "step": 12230
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.500709056854248,
-      "learning_rate": 1.2040677966101695e-05,
-      "loss": 1.1895,
-      "step": 12240
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.5548219680786133,
-      "learning_rate": 1.2033898305084745e-05,
-      "loss": 1.4415,
-      "step": 12250
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 4.0832295417785645,
-      "learning_rate": 1.2027118644067796e-05,
-      "loss": 1.2931,
-      "step": 12260
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.2954070568084717,
-      "learning_rate": 1.2020338983050848e-05,
-      "loss": 1.052,
-      "step": 12270
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 2.6456809043884277,
-      "learning_rate": 1.20135593220339e-05,
-      "loss": 1.3214,
-      "step": 12280
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 3.9517974853515625,
-      "learning_rate": 1.2006779661016949e-05,
-      "loss": 1.3797,
-      "step": 12290
-    },
-    {
-      "epoch": 0.28,
-      "grad_norm": 6.248586654663086,
-      "learning_rate": 1.2e-05,
-      "loss": 1.3371,
-      "step": 12300
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 3.673175811767578,
-      "learning_rate": 1.1993220338983052e-05,
-      "loss": 1.3891,
-      "step": 12310
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.9075310230255127,
-      "learning_rate": 1.1986440677966103e-05,
-      "loss": 1.3943,
-      "step": 12320
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 6.2103705406188965,
-      "learning_rate": 1.1979661016949153e-05,
-      "loss": 1.3813,
-      "step": 12330
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 3.8567051887512207,
-      "learning_rate": 1.1972881355932204e-05,
-      "loss": 1.4039,
-      "step": 12340
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.756994724273682,
-      "learning_rate": 1.1966101694915256e-05,
-      "loss": 1.2371,
-      "step": 12350
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.874370574951172,
-      "learning_rate": 1.1959322033898307e-05,
-      "loss": 1.3543,
-      "step": 12360
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.5366311073303223,
-      "learning_rate": 1.1952542372881357e-05,
-      "loss": 1.2105,
-      "step": 12370
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 1.0801111459732056,
-      "learning_rate": 1.1945762711864408e-05,
-      "loss": 1.46,
-      "step": 12380
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.35442590713501,
-      "learning_rate": 1.193898305084746e-05,
-      "loss": 1.3463,
-      "step": 12390
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 1.817285180091858,
-      "learning_rate": 1.1932203389830511e-05,
-      "loss": 1.1985,
-      "step": 12400
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 1.607805848121643,
-      "learning_rate": 1.192542372881356e-05,
-      "loss": 1.2979,
-      "step": 12410
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 0.8578438758850098,
-      "learning_rate": 1.1918644067796612e-05,
-      "loss": 1.1953,
-      "step": 12420
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.037782669067383,
-      "learning_rate": 1.1911864406779663e-05,
-      "loss": 1.4013,
-      "step": 12430
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.66656231880188,
-      "learning_rate": 1.1905084745762713e-05,
-      "loss": 1.1451,
-      "step": 12440
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 6.166327953338623,
-      "learning_rate": 1.1898305084745763e-05,
-      "loss": 1.2794,
-      "step": 12450
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 6.4938483238220215,
-      "learning_rate": 1.1891525423728814e-05,
-      "loss": 1.2848,
-      "step": 12460
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.013676166534424,
-      "learning_rate": 1.1884745762711864e-05,
-      "loss": 1.3793,
-      "step": 12470
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 1.673925757408142,
-      "learning_rate": 1.1877966101694915e-05,
-      "loss": 1.1985,
-      "step": 12480
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 3.0690205097198486,
-      "learning_rate": 1.1871186440677967e-05,
-      "loss": 1.3797,
-      "step": 12490
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.671661853790283,
-      "learning_rate": 1.1864406779661018e-05,
-      "loss": 1.3697,
-      "step": 12500
-    },
-    {
-      "epoch": 0.29,
-      "eval_loss": 1.0424882173538208,
-      "eval_runtime": 67.1094,
-      "eval_samples_per_second": 14.901,
-      "eval_steps_per_second": 14.901,
-      "step": 12500
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.892620086669922,
-      "learning_rate": 1.1857627118644068e-05,
-      "loss": 1.3287,
-      "step": 12510
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 7.4180145263671875,
-      "learning_rate": 1.185084745762712e-05,
-      "loss": 1.4554,
-      "step": 12520
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 6.7615966796875,
-      "learning_rate": 1.184406779661017e-05,
-      "loss": 1.3294,
-      "step": 12530
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 5.470348358154297,
-      "learning_rate": 1.183728813559322e-05,
-      "loss": 1.4311,
-      "step": 12540
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 3.9983062744140625,
-      "learning_rate": 1.1830508474576272e-05,
-      "loss": 1.313,
-      "step": 12550
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 7.445504188537598,
-      "learning_rate": 1.1823728813559323e-05,
-      "loss": 1.2363,
-      "step": 12560
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.365860462188721,
-      "learning_rate": 1.1816949152542375e-05,
-      "loss": 1.3424,
-      "step": 12570
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.8447556495666504,
-      "learning_rate": 1.1810169491525424e-05,
-      "loss": 1.3638,
-      "step": 12580
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.913205623626709,
-      "learning_rate": 1.1803389830508476e-05,
-      "loss": 1.3654,
-      "step": 12590
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 6.282583713531494,
-      "learning_rate": 1.1796610169491527e-05,
-      "loss": 1.386,
-      "step": 12600
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.249077320098877,
-      "learning_rate": 1.1789830508474578e-05,
-      "loss": 1.2501,
-      "step": 12610
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 5.649593353271484,
-      "learning_rate": 1.1783050847457628e-05,
-      "loss": 1.3802,
-      "step": 12620
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.310429096221924,
-      "learning_rate": 1.177627118644068e-05,
-      "loss": 1.3718,
-      "step": 12630
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 8.635117530822754,
-      "learning_rate": 1.1769491525423731e-05,
-      "loss": 1.1369,
-      "step": 12640
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.3350889682769775,
-      "learning_rate": 1.1762711864406782e-05,
-      "loss": 1.5591,
-      "step": 12650
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 3.3547542095184326,
-      "learning_rate": 1.175593220338983e-05,
-      "loss": 1.0917,
-      "step": 12660
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 6.9904937744140625,
-      "learning_rate": 1.1749152542372882e-05,
-      "loss": 1.3379,
-      "step": 12670
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 3.06915020942688,
-      "learning_rate": 1.1742372881355931e-05,
-      "loss": 1.3451,
-      "step": 12680
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 6.47864294052124,
-      "learning_rate": 1.1735593220338983e-05,
-      "loss": 1.2393,
-      "step": 12690
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 4.751437664031982,
-      "learning_rate": 1.1728813559322034e-05,
-      "loss": 1.2097,
-      "step": 12700
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 5.026716232299805,
-      "learning_rate": 1.1722033898305086e-05,
-      "loss": 1.3033,
-      "step": 12710
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.250887393951416,
-      "learning_rate": 1.1715254237288135e-05,
-      "loss": 1.3686,
-      "step": 12720
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 2.050124406814575,
-      "learning_rate": 1.1708474576271187e-05,
-      "loss": 1.3823,
-      "step": 12730
-    },
-    {
-      "epoch": 0.29,
-      "grad_norm": 3.651585817337036,
-      "learning_rate": 1.1701694915254238e-05,
-      "loss": 1.3815,
-      "step": 12740
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 1.1647934913635254,
-      "learning_rate": 1.169491525423729e-05,
-      "loss": 1.346,
-      "step": 12750
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.0802316665649414,
-      "learning_rate": 1.168813559322034e-05,
-      "loss": 1.2188,
-      "step": 12760
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 6.935558319091797,
-      "learning_rate": 1.168135593220339e-05,
-      "loss": 1.2733,
-      "step": 12770
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.3272249698638916,
-      "learning_rate": 1.1674576271186442e-05,
-      "loss": 1.4055,
-      "step": 12780
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 7.0462517738342285,
-      "learning_rate": 1.1667796610169494e-05,
-      "loss": 1.3269,
-      "step": 12790
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.325636148452759,
-      "learning_rate": 1.1661016949152543e-05,
-      "loss": 1.4648,
-      "step": 12800
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 1.3131448030471802,
-      "learning_rate": 1.1654237288135595e-05,
-      "loss": 1.3613,
-      "step": 12810
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 5.341305732727051,
-      "learning_rate": 1.1647457627118646e-05,
-      "loss": 1.2867,
-      "step": 12820
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 5.281243324279785,
-      "learning_rate": 1.1640677966101697e-05,
-      "loss": 1.2868,
-      "step": 12830
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.2200121879577637,
-      "learning_rate": 1.1633898305084747e-05,
-      "loss": 1.5082,
-      "step": 12840
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 7.734021186828613,
-      "learning_rate": 1.1627118644067799e-05,
-      "loss": 1.1832,
-      "step": 12850
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.2634496688842773,
-      "learning_rate": 1.162033898305085e-05,
-      "loss": 1.3862,
-      "step": 12860
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.978215456008911,
-      "learning_rate": 1.1613559322033898e-05,
-      "loss": 1.19,
-      "step": 12870
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 1.9776360988616943,
-      "learning_rate": 1.160677966101695e-05,
-      "loss": 1.3434,
-      "step": 12880
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 9.741606712341309,
-      "learning_rate": 1.16e-05,
-      "loss": 1.5127,
-      "step": 12890
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 1.8154492378234863,
-      "learning_rate": 1.159322033898305e-05,
-      "loss": 1.1796,
-      "step": 12900
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.0180463790893555,
-      "learning_rate": 1.1586440677966102e-05,
-      "loss": 1.3409,
-      "step": 12910
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.2689106464385986,
-      "learning_rate": 1.1579661016949153e-05,
-      "loss": 1.279,
-      "step": 12920
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.992006778717041,
-      "learning_rate": 1.1572881355932205e-05,
-      "loss": 1.0891,
-      "step": 12930
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 7.110446453094482,
-      "learning_rate": 1.1566101694915254e-05,
-      "loss": 1.2186,
-      "step": 12940
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.776698350906372,
-      "learning_rate": 1.1559322033898306e-05,
-      "loss": 1.505,
-      "step": 12950
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 4.205793857574463,
-      "learning_rate": 1.1552542372881357e-05,
-      "loss": 1.4795,
-      "step": 12960
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 5.062299728393555,
-      "learning_rate": 1.1545762711864409e-05,
-      "loss": 1.3921,
-      "step": 12970
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 14.015942573547363,
-      "learning_rate": 1.1538983050847458e-05,
-      "loss": 1.3975,
-      "step": 12980
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 1.4116904735565186,
-      "learning_rate": 1.153220338983051e-05,
-      "loss": 1.1743,
-      "step": 12990
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.194725513458252,
-      "learning_rate": 1.1525423728813561e-05,
-      "loss": 1.4538,
-      "step": 13000
-    },
-    {
-      "epoch": 0.3,
-      "eval_loss": 1.071112036705017,
-      "eval_runtime": 67.075,
-      "eval_samples_per_second": 14.909,
-      "eval_steps_per_second": 14.909,
-      "step": 13000
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.108156204223633,
-      "learning_rate": 1.151864406779661e-05,
-      "loss": 1.4626,
-      "step": 13010
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 0.9883029460906982,
-      "learning_rate": 1.1511864406779662e-05,
-      "loss": 1.2526,
-      "step": 13020
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.394761085510254,
-      "learning_rate": 1.1505084745762714e-05,
-      "loss": 1.35,
-      "step": 13030
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.971503973007202,
-      "learning_rate": 1.1498305084745765e-05,
-      "loss": 1.3228,
-      "step": 13040
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.9637510776519775,
-      "learning_rate": 1.1491525423728815e-05,
-      "loss": 1.3027,
-      "step": 13050
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.263624906539917,
-      "learning_rate": 1.1484745762711866e-05,
-      "loss": 1.2493,
-      "step": 13060
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.722733497619629,
-      "learning_rate": 1.1477966101694916e-05,
-      "loss": 1.2733,
-      "step": 13070
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.7181248664855957,
-      "learning_rate": 1.1471186440677965e-05,
-      "loss": 1.1724,
-      "step": 13080
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.551340341567993,
-      "learning_rate": 1.1464406779661017e-05,
-      "loss": 1.1791,
-      "step": 13090
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 9.841154098510742,
-      "learning_rate": 1.1457627118644068e-05,
-      "loss": 1.2801,
-      "step": 13100
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 5.1563191413879395,
-      "learning_rate": 1.145084745762712e-05,
-      "loss": 1.4498,
-      "step": 13110
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 5.396986961364746,
-      "learning_rate": 1.144406779661017e-05,
-      "loss": 1.3006,
-      "step": 13120
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 1.691427230834961,
-      "learning_rate": 1.143728813559322e-05,
-      "loss": 1.3966,
-      "step": 13130
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 3.0465362071990967,
-      "learning_rate": 1.1430508474576272e-05,
-      "loss": 1.4104,
-      "step": 13140
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 5.615993976593018,
-      "learning_rate": 1.1423728813559322e-05,
-      "loss": 1.231,
-      "step": 13150
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 2.800468683242798,
-      "learning_rate": 1.1416949152542373e-05,
-      "loss": 1.1893,
-      "step": 13160
-    },
-    {
-      "epoch": 0.3,
-      "grad_norm": 5.876188278198242,
-      "learning_rate": 1.1410169491525425e-05,
-      "loss": 1.1381,
-      "step": 13170
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.2849178314208984,
-      "learning_rate": 1.1403389830508476e-05,
-      "loss": 1.2292,
-      "step": 13180
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 5.406613826751709,
-      "learning_rate": 1.1396610169491526e-05,
-      "loss": 1.1515,
-      "step": 13190
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 5.904646873474121,
-      "learning_rate": 1.1389830508474577e-05,
-      "loss": 1.5528,
-      "step": 13200
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.9938790798187256,
-      "learning_rate": 1.1383050847457629e-05,
-      "loss": 1.1436,
-      "step": 13210
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.544445037841797,
-      "learning_rate": 1.137627118644068e-05,
-      "loss": 1.3652,
-      "step": 13220
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 5.070495128631592,
-      "learning_rate": 1.136949152542373e-05,
-      "loss": 1.2352,
-      "step": 13230
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.612300395965576,
-      "learning_rate": 1.1362711864406781e-05,
-      "loss": 1.3513,
-      "step": 13240
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.960466146469116,
-      "learning_rate": 1.1355932203389833e-05,
-      "loss": 1.0844,
-      "step": 13250
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.9677114486694336,
-      "learning_rate": 1.1349152542372884e-05,
-      "loss": 1.4191,
-      "step": 13260
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.220777988433838,
-      "learning_rate": 1.1342372881355934e-05,
-      "loss": 1.3462,
-      "step": 13270
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 4.4845452308654785,
-      "learning_rate": 1.1335593220338983e-05,
-      "loss": 1.1878,
-      "step": 13280
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.976458787918091,
-      "learning_rate": 1.1328813559322033e-05,
-      "loss": 1.1003,
-      "step": 13290
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.404944658279419,
-      "learning_rate": 1.1322033898305084e-05,
-      "loss": 1.4279,
-      "step": 13300
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 5.025093078613281,
-      "learning_rate": 1.1315254237288136e-05,
-      "loss": 1.2617,
-      "step": 13310
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 1.922884225845337,
-      "learning_rate": 1.1308474576271187e-05,
-      "loss": 1.4879,
-      "step": 13320
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 5.648707866668701,
-      "learning_rate": 1.1301694915254237e-05,
-      "loss": 1.4279,
-      "step": 13330
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.952711820602417,
-      "learning_rate": 1.1294915254237288e-05,
-      "loss": 1.294,
-      "step": 13340
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.5280747413635254,
-      "learning_rate": 1.128813559322034e-05,
-      "loss": 1.3706,
-      "step": 13350
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.169260025024414,
-      "learning_rate": 1.1281355932203391e-05,
-      "loss": 1.3188,
-      "step": 13360
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.602459192276001,
-      "learning_rate": 1.1274576271186441e-05,
-      "loss": 1.1604,
-      "step": 13370
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.6066842079162598,
-      "learning_rate": 1.1267796610169492e-05,
-      "loss": 1.3923,
-      "step": 13380
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.369154930114746,
-      "learning_rate": 1.1261016949152544e-05,
-      "loss": 1.3249,
-      "step": 13390
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 1.7909421920776367,
-      "learning_rate": 1.1254237288135595e-05,
-      "loss": 1.3519,
-      "step": 13400
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.983210802078247,
-      "learning_rate": 1.1247457627118645e-05,
-      "loss": 1.2492,
-      "step": 13410
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.9838898181915283,
-      "learning_rate": 1.1240677966101696e-05,
-      "loss": 1.4542,
-      "step": 13420
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 6.088125705718994,
-      "learning_rate": 1.1233898305084748e-05,
-      "loss": 1.2546,
-      "step": 13430
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.840747833251953,
-      "learning_rate": 1.1227118644067799e-05,
-      "loss": 1.2426,
-      "step": 13440
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 4.945519924163818,
-      "learning_rate": 1.1220338983050849e-05,
-      "loss": 1.3064,
-      "step": 13450
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 4.030840873718262,
-      "learning_rate": 1.12135593220339e-05,
-      "loss": 1.1405,
-      "step": 13460
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 4.2843194007873535,
-      "learning_rate": 1.1206779661016951e-05,
-      "loss": 1.3749,
-      "step": 13470
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 8.288750648498535,
-      "learning_rate": 1.1200000000000001e-05,
-      "loss": 1.37,
-      "step": 13480
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.233605146408081,
-      "learning_rate": 1.1193220338983051e-05,
-      "loss": 1.1979,
-      "step": 13490
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.9739513397216797,
-      "learning_rate": 1.1186440677966102e-05,
-      "loss": 1.0539,
-      "step": 13500
-    },
-    {
-      "epoch": 0.31,
-      "eval_loss": 1.068673849105835,
-      "eval_runtime": 67.158,
-      "eval_samples_per_second": 14.89,
-      "eval_steps_per_second": 14.89,
-      "step": 13500
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.8984007835388184,
-      "learning_rate": 1.1179661016949152e-05,
-      "loss": 1.1189,
-      "step": 13510
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 5.570311069488525,
-      "learning_rate": 1.1172881355932203e-05,
-      "loss": 1.2069,
-      "step": 13520
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.0752551555633545,
-      "learning_rate": 1.1166101694915255e-05,
-      "loss": 1.304,
-      "step": 13530
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 4.809720039367676,
-      "learning_rate": 1.1159322033898306e-05,
-      "loss": 1.3403,
-      "step": 13540
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 6.9752655029296875,
-      "learning_rate": 1.1152542372881356e-05,
-      "loss": 1.4062,
-      "step": 13550
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.2357780933380127,
-      "learning_rate": 1.1145762711864407e-05,
-      "loss": 1.409,
-      "step": 13560
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.0999038219451904,
-      "learning_rate": 1.1138983050847459e-05,
-      "loss": 1.3877,
-      "step": 13570
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 5.735909461975098,
-      "learning_rate": 1.113220338983051e-05,
-      "loss": 1.3124,
-      "step": 13580
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 2.736616849899292,
-      "learning_rate": 1.112542372881356e-05,
-      "loss": 1.3948,
-      "step": 13590
-    },
-    {
-      "epoch": 0.31,
-      "grad_norm": 3.9741382598876953,
-      "learning_rate": 1.1118644067796611e-05,
-      "loss": 1.3462,
-      "step": 13600
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.8991127014160156,
-      "learning_rate": 1.1111864406779663e-05,
-      "loss": 1.3715,
-      "step": 13610
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 1.8136656284332275,
-      "learning_rate": 1.1105084745762712e-05,
-      "loss": 1.1655,
-      "step": 13620
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 5.556140899658203,
-      "learning_rate": 1.1098305084745764e-05,
-      "loss": 1.1872,
-      "step": 13630
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 4.87309455871582,
-      "learning_rate": 1.1091525423728815e-05,
-      "loss": 1.2085,
-      "step": 13640
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 4.111629486083984,
-      "learning_rate": 1.1084745762711867e-05,
-      "loss": 1.2432,
-      "step": 13650
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.4309825897216797,
-      "learning_rate": 1.1077966101694916e-05,
-      "loss": 1.291,
-      "step": 13660
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 4.081984043121338,
-      "learning_rate": 1.1071186440677968e-05,
-      "loss": 1.2944,
-      "step": 13670
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.980055093765259,
-      "learning_rate": 1.1064406779661019e-05,
-      "loss": 1.4403,
-      "step": 13680
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 6.602412223815918,
-      "learning_rate": 1.105762711864407e-05,
-      "loss": 1.2835,
-      "step": 13690
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 6.045549392700195,
-      "learning_rate": 1.1050847457627118e-05,
-      "loss": 1.2792,
-      "step": 13700
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 8.093133926391602,
-      "learning_rate": 1.104406779661017e-05,
-      "loss": 1.3566,
-      "step": 13710
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.0017004013061523,
-      "learning_rate": 1.1037288135593221e-05,
-      "loss": 1.0925,
-      "step": 13720
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 4.844403266906738,
-      "learning_rate": 1.1030508474576271e-05,
-      "loss": 1.3142,
-      "step": 13730
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 4.788667678833008,
-      "learning_rate": 1.1023728813559322e-05,
-      "loss": 1.4304,
-      "step": 13740
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 12.615195274353027,
-      "learning_rate": 1.1016949152542374e-05,
-      "loss": 1.4411,
-      "step": 13750
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 5.6674323081970215,
-      "learning_rate": 1.1010169491525423e-05,
-      "loss": 1.2066,
-      "step": 13760
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 10.92747688293457,
-      "learning_rate": 1.1003389830508475e-05,
-      "loss": 1.4857,
-      "step": 13770
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.42288875579834,
-      "learning_rate": 1.0996610169491526e-05,
-      "loss": 1.2415,
-      "step": 13780
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 9.594301223754883,
-      "learning_rate": 1.0989830508474578e-05,
-      "loss": 1.3813,
-      "step": 13790
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.701327323913574,
-      "learning_rate": 1.0983050847457627e-05,
-      "loss": 1.1592,
-      "step": 13800
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.5337729454040527,
-      "learning_rate": 1.0976271186440679e-05,
-      "loss": 1.3177,
-      "step": 13810
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 14.992568016052246,
-      "learning_rate": 1.096949152542373e-05,
-      "loss": 1.1037,
-      "step": 13820
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 12.353404998779297,
-      "learning_rate": 1.0962711864406782e-05,
-      "loss": 1.3257,
-      "step": 13830
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.713555335998535,
-      "learning_rate": 1.0955932203389831e-05,
-      "loss": 1.2935,
-      "step": 13840
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 5.5155348777771,
-      "learning_rate": 1.0949152542372883e-05,
-      "loss": 1.4259,
-      "step": 13850
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.6537179946899414,
-      "learning_rate": 1.0942372881355934e-05,
-      "loss": 1.333,
-      "step": 13860
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.3852736949920654,
-      "learning_rate": 1.0935593220338985e-05,
-      "loss": 1.2526,
-      "step": 13870
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.6579887866973877,
-      "learning_rate": 1.0928813559322035e-05,
-      "loss": 1.4214,
-      "step": 13880
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.4810287952423096,
-      "learning_rate": 1.0922033898305087e-05,
-      "loss": 1.4154,
-      "step": 13890
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 4.413066387176514,
-      "learning_rate": 1.0915254237288135e-05,
-      "loss": 1.4967,
-      "step": 13900
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.1642134189605713,
-      "learning_rate": 1.0908474576271186e-05,
-      "loss": 1.2179,
-      "step": 13910
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 1.1288875341415405,
-      "learning_rate": 1.0901694915254237e-05,
-      "loss": 1.2011,
-      "step": 13920
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.4285356998443604,
-      "learning_rate": 1.0894915254237289e-05,
-      "loss": 1.3319,
-      "step": 13930
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.1330859661102295,
-      "learning_rate": 1.0888135593220339e-05,
-      "loss": 1.2524,
-      "step": 13940
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.8436052799224854,
-      "learning_rate": 1.088135593220339e-05,
-      "loss": 1.2021,
-      "step": 13950
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 1.7708642482757568,
-      "learning_rate": 1.0874576271186441e-05,
-      "loss": 1.1705,
-      "step": 13960
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 5.702758312225342,
-      "learning_rate": 1.0867796610169493e-05,
-      "loss": 1.3999,
-      "step": 13970
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 1.627833604812622,
-      "learning_rate": 1.0861016949152542e-05,
-      "loss": 1.497,
-      "step": 13980
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 2.4980685710906982,
-      "learning_rate": 1.0854237288135594e-05,
-      "loss": 1.2449,
-      "step": 13990
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.2824714183807373,
-      "learning_rate": 1.0847457627118645e-05,
-      "loss": 1.1941,
-      "step": 14000
-    },
-    {
-      "epoch": 0.32,
-      "eval_loss": 1.097359299659729,
-      "eval_runtime": 67.0566,
-      "eval_samples_per_second": 14.913,
-      "eval_steps_per_second": 14.913,
-      "step": 14000
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.3165223598480225,
-      "learning_rate": 1.0840677966101697e-05,
-      "loss": 1.2049,
-      "step": 14010
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 3.123730182647705,
-      "learning_rate": 1.0833898305084746e-05,
-      "loss": 1.2982,
-      "step": 14020
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 1.46307373046875,
-      "learning_rate": 1.0827118644067798e-05,
-      "loss": 1.175,
-      "step": 14030
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.6726436614990234,
-      "learning_rate": 1.0820338983050849e-05,
-      "loss": 1.1683,
-      "step": 14040
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.585521936416626,
-      "learning_rate": 1.08135593220339e-05,
-      "loss": 1.4859,
-      "step": 14050
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.8152098655700684,
-      "learning_rate": 1.080677966101695e-05,
-      "loss": 1.2218,
-      "step": 14060
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 5.557183742523193,
-      "learning_rate": 1.0800000000000002e-05,
-      "loss": 1.1304,
-      "step": 14070
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 5.458992958068848,
-      "learning_rate": 1.0793220338983053e-05,
-      "loss": 1.2276,
-      "step": 14080
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.8788580894470215,
-      "learning_rate": 1.0786440677966103e-05,
-      "loss": 1.3762,
-      "step": 14090
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 7.669591426849365,
-      "learning_rate": 1.0779661016949154e-05,
-      "loss": 1.2418,
-      "step": 14100
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 5.8121442794799805,
-      "learning_rate": 1.0772881355932204e-05,
-      "loss": 1.2151,
-      "step": 14110
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 4.32166051864624,
-      "learning_rate": 1.0766101694915254e-05,
-      "loss": 1.3599,
-      "step": 14120
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.139389991760254,
-      "learning_rate": 1.0759322033898305e-05,
-      "loss": 1.1879,
-      "step": 14130
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 1.7812507152557373,
-      "learning_rate": 1.0752542372881356e-05,
-      "loss": 1.3147,
-      "step": 14140
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.0108046531677246,
-      "learning_rate": 1.0745762711864408e-05,
-      "loss": 1.1607,
-      "step": 14150
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 6.003909587860107,
-      "learning_rate": 1.0738983050847457e-05,
-      "loss": 1.3212,
-      "step": 14160
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 6.925515174865723,
-      "learning_rate": 1.0732203389830509e-05,
-      "loss": 1.342,
-      "step": 14170
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 5.450997352600098,
-      "learning_rate": 1.072542372881356e-05,
-      "loss": 1.327,
-      "step": 14180
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 4.291378974914551,
-      "learning_rate": 1.0718644067796612e-05,
-      "loss": 1.2563,
-      "step": 14190
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.7784173488616943,
-      "learning_rate": 1.0711864406779661e-05,
-      "loss": 1.5323,
-      "step": 14200
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.37923526763916,
-      "learning_rate": 1.0705084745762713e-05,
-      "loss": 1.3377,
-      "step": 14210
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 1.3921860456466675,
-      "learning_rate": 1.0698305084745764e-05,
-      "loss": 1.2814,
-      "step": 14220
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 4.000369548797607,
-      "learning_rate": 1.0691525423728814e-05,
-      "loss": 1.1886,
-      "step": 14230
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.731414318084717,
-      "learning_rate": 1.0684745762711865e-05,
-      "loss": 1.5797,
-      "step": 14240
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.6613595485687256,
-      "learning_rate": 1.0677966101694917e-05,
-      "loss": 1.2735,
-      "step": 14250
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 1.467409610748291,
-      "learning_rate": 1.0671186440677968e-05,
-      "loss": 1.4635,
-      "step": 14260
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 1.406770944595337,
-      "learning_rate": 1.0664406779661018e-05,
-      "loss": 1.1051,
-      "step": 14270
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 1.697908878326416,
-      "learning_rate": 1.065762711864407e-05,
-      "loss": 1.2916,
-      "step": 14280
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.9356634616851807,
-      "learning_rate": 1.065084745762712e-05,
-      "loss": 1.2845,
-      "step": 14290
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 4.0369553565979,
-      "learning_rate": 1.0644067796610172e-05,
-      "loss": 1.1785,
-      "step": 14300
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 7.373144149780273,
-      "learning_rate": 1.0637288135593222e-05,
-      "loss": 1.2085,
-      "step": 14310
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.261367082595825,
-      "learning_rate": 1.0630508474576271e-05,
-      "loss": 1.2291,
-      "step": 14320
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.6595752239227295,
-      "learning_rate": 1.0623728813559323e-05,
-      "loss": 1.4075,
-      "step": 14330
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.082679510116577,
-      "learning_rate": 1.0616949152542373e-05,
-      "loss": 1.2423,
-      "step": 14340
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.0270230770111084,
-      "learning_rate": 1.0610169491525424e-05,
-      "loss": 1.3866,
-      "step": 14350
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.6601712703704834,
-      "learning_rate": 1.0603389830508475e-05,
-      "loss": 1.5351,
-      "step": 14360
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.8816442489624023,
-      "learning_rate": 1.0596610169491525e-05,
-      "loss": 1.4206,
-      "step": 14370
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 2.6769402027130127,
-      "learning_rate": 1.0589830508474576e-05,
-      "loss": 1.3384,
-      "step": 14380
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 7.779047966003418,
-      "learning_rate": 1.0583050847457628e-05,
-      "loss": 1.2507,
-      "step": 14390
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 1.7086968421936035,
-      "learning_rate": 1.057627118644068e-05,
-      "loss": 1.4237,
-      "step": 14400
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 6.948585033416748,
-      "learning_rate": 1.0569491525423729e-05,
-      "loss": 1.4241,
-      "step": 14410
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 4.5062103271484375,
-      "learning_rate": 1.056271186440678e-05,
-      "loss": 1.1971,
-      "step": 14420
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 3.1329903602600098,
-      "learning_rate": 1.0555932203389832e-05,
-      "loss": 1.2855,
-      "step": 14430
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 4.693443298339844,
-      "learning_rate": 1.0549152542372883e-05,
-      "loss": 1.4162,
-      "step": 14440
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 6.092784404754639,
-      "learning_rate": 1.0542372881355933e-05,
-      "loss": 1.2929,
-      "step": 14450
-    },
-    {
-      "epoch": 0.33,
-      "grad_norm": 5.795193195343018,
-      "learning_rate": 1.0535593220338984e-05,
-      "loss": 1.1753,
-      "step": 14460
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.002300500869751,
-      "learning_rate": 1.0528813559322036e-05,
-      "loss": 1.2696,
-      "step": 14470
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 8.367115020751953,
-      "learning_rate": 1.0522033898305087e-05,
-      "loss": 1.4971,
-      "step": 14480
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 7.178134918212891,
-      "learning_rate": 1.0515254237288137e-05,
-      "loss": 1.2025,
-      "step": 14490
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 6.813628673553467,
-      "learning_rate": 1.0508474576271188e-05,
-      "loss": 1.2135,
-      "step": 14500
-    },
-    {
-      "epoch": 0.34,
-      "eval_loss": 1.056099772453308,
-      "eval_runtime": 67.1376,
-      "eval_samples_per_second": 14.895,
-      "eval_steps_per_second": 14.895,
-      "step": 14500
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 6.232717990875244,
-      "learning_rate": 1.050169491525424e-05,
-      "loss": 1.2396,
-      "step": 14510
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 6.143510341644287,
-      "learning_rate": 1.049491525423729e-05,
-      "loss": 1.503,
-      "step": 14520
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.227885723114014,
-      "learning_rate": 1.0488135593220339e-05,
-      "loss": 1.2606,
-      "step": 14530
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.4114978313446045,
-      "learning_rate": 1.048135593220339e-05,
-      "loss": 1.3879,
-      "step": 14540
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.572566509246826,
-      "learning_rate": 1.047457627118644e-05,
-      "loss": 1.3413,
-      "step": 14550
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.934830188751221,
-      "learning_rate": 1.0467796610169491e-05,
-      "loss": 1.2421,
-      "step": 14560
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.879978656768799,
-      "learning_rate": 1.0461016949152543e-05,
-      "loss": 1.2392,
-      "step": 14570
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 2.335127353668213,
-      "learning_rate": 1.0454237288135594e-05,
-      "loss": 1.1258,
-      "step": 14580
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 9.824913024902344,
-      "learning_rate": 1.0447457627118644e-05,
-      "loss": 1.2003,
-      "step": 14590
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 2.6796653270721436,
-      "learning_rate": 1.0440677966101695e-05,
-      "loss": 1.3512,
-      "step": 14600
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 2.872577667236328,
-      "learning_rate": 1.0433898305084747e-05,
-      "loss": 1.1608,
-      "step": 14610
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.358550071716309,
-      "learning_rate": 1.0427118644067798e-05,
-      "loss": 1.3207,
-      "step": 14620
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.672595500946045,
-      "learning_rate": 1.0420338983050848e-05,
-      "loss": 1.5645,
-      "step": 14630
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 6.134851455688477,
-      "learning_rate": 1.04135593220339e-05,
-      "loss": 1.4636,
-      "step": 14640
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.763167381286621,
-      "learning_rate": 1.040677966101695e-05,
-      "loss": 1.3488,
-      "step": 14650
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.057564735412598,
-      "learning_rate": 1.04e-05,
-      "loss": 1.1286,
-      "step": 14660
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.70720100402832,
-      "learning_rate": 1.0393220338983052e-05,
-      "loss": 1.2291,
-      "step": 14670
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.895830154418945,
-      "learning_rate": 1.0386440677966103e-05,
-      "loss": 1.3372,
-      "step": 14680
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.813084125518799,
-      "learning_rate": 1.0379661016949155e-05,
-      "loss": 1.388,
-      "step": 14690
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.9301507472991943,
-      "learning_rate": 1.0372881355932204e-05,
-      "loss": 1.3411,
-      "step": 14700
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 1.1061105728149414,
-      "learning_rate": 1.0366101694915256e-05,
-      "loss": 1.1152,
-      "step": 14710
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 2.1552724838256836,
-      "learning_rate": 1.0359322033898307e-05,
-      "loss": 1.3387,
-      "step": 14720
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.779531002044678,
-      "learning_rate": 1.0352542372881358e-05,
-      "loss": 1.2235,
-      "step": 14730
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.468095779418945,
-      "learning_rate": 1.0345762711864406e-05,
-      "loss": 1.458,
-      "step": 14740
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.9823062419891357,
-      "learning_rate": 1.0338983050847458e-05,
-      "loss": 1.2354,
-      "step": 14750
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.236638069152832,
-      "learning_rate": 1.033220338983051e-05,
-      "loss": 1.2741,
-      "step": 14760
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.5503621101379395,
-      "learning_rate": 1.0325423728813559e-05,
-      "loss": 1.1736,
-      "step": 14770
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 8.381211280822754,
-      "learning_rate": 1.031864406779661e-05,
-      "loss": 1.1965,
-      "step": 14780
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 6.995142459869385,
-      "learning_rate": 1.0311864406779662e-05,
-      "loss": 1.2316,
-      "step": 14790
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 8.22109317779541,
-      "learning_rate": 1.0305084745762712e-05,
-      "loss": 1.3987,
-      "step": 14800
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 1.7374868392944336,
-      "learning_rate": 1.0298305084745763e-05,
-      "loss": 1.2053,
-      "step": 14810
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.8697991371154785,
-      "learning_rate": 1.0291525423728814e-05,
-      "loss": 1.1784,
-      "step": 14820
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 9.522156715393066,
-      "learning_rate": 1.0284745762711866e-05,
-      "loss": 1.4226,
-      "step": 14830
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.162941932678223,
-      "learning_rate": 1.0277966101694915e-05,
-      "loss": 1.3565,
-      "step": 14840
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 4.80945348739624,
-      "learning_rate": 1.0271186440677967e-05,
-      "loss": 1.3013,
-      "step": 14850
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 3.0665078163146973,
-      "learning_rate": 1.0264406779661018e-05,
-      "loss": 1.4093,
-      "step": 14860
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 2.34722900390625,
-      "learning_rate": 1.025762711864407e-05,
-      "loss": 1.3063,
-      "step": 14870
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 5.438226699829102,
-      "learning_rate": 1.025084745762712e-05,
-      "loss": 1.1209,
-      "step": 14880
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 2.0597119331359863,
-      "learning_rate": 1.024406779661017e-05,
-      "loss": 1.3328,
-      "step": 14890
-    },
-    {
-      "epoch": 0.34,
-      "grad_norm": 1.7713210582733154,
-      "learning_rate": 1.0237288135593222e-05,
-      "loss": 1.2971,
-      "step": 14900
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 1.6047179698944092,
-      "learning_rate": 1.0230508474576274e-05,
-      "loss": 1.226,
-      "step": 14910
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.7487828731536865,
-      "learning_rate": 1.0223728813559323e-05,
-      "loss": 1.0162,
-      "step": 14920
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.8491158485412598,
-      "learning_rate": 1.0216949152542375e-05,
-      "loss": 1.3324,
-      "step": 14930
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.6983745098114014,
-      "learning_rate": 1.0210169491525423e-05,
-      "loss": 1.3416,
-      "step": 14940
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 7.2397589683532715,
-      "learning_rate": 1.0203389830508474e-05,
-      "loss": 1.4135,
-      "step": 14950
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 6.375013828277588,
-      "learning_rate": 1.0196610169491525e-05,
-      "loss": 1.2063,
-      "step": 14960
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 7.71971321105957,
-      "learning_rate": 1.0189830508474577e-05,
-      "loss": 1.2969,
-      "step": 14970
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 5.973591327667236,
-      "learning_rate": 1.0183050847457627e-05,
-      "loss": 1.179,
-      "step": 14980
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 6.778249263763428,
-      "learning_rate": 1.0176271186440678e-05,
-      "loss": 1.2906,
-      "step": 14990
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.3302714824676514,
-      "learning_rate": 1.016949152542373e-05,
-      "loss": 1.2764,
-      "step": 15000
-    },
-    {
-      "epoch": 0.35,
-      "eval_loss": 1.0367546081542969,
-      "eval_runtime": 67.0312,
-      "eval_samples_per_second": 14.918,
-      "eval_steps_per_second": 14.918,
-      "step": 15000
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.099465847015381,
-      "learning_rate": 1.016271186440678e-05,
-      "loss": 1.2667,
-      "step": 15010
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.9707233905792236,
-      "learning_rate": 1.015593220338983e-05,
-      "loss": 1.1102,
-      "step": 15020
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.409958362579346,
-      "learning_rate": 1.0149152542372882e-05,
-      "loss": 1.406,
-      "step": 15030
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 6.402803421020508,
-      "learning_rate": 1.0142372881355933e-05,
-      "loss": 1.3003,
-      "step": 15040
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 1.084704041481018,
-      "learning_rate": 1.0135593220338985e-05,
-      "loss": 1.2643,
-      "step": 15050
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 7.1039910316467285,
-      "learning_rate": 1.0128813559322034e-05,
-      "loss": 1.3562,
-      "step": 15060
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.218326568603516,
-      "learning_rate": 1.0122033898305086e-05,
-      "loss": 1.3977,
-      "step": 15070
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 5.44041109085083,
-      "learning_rate": 1.0115254237288137e-05,
-      "loss": 1.3783,
-      "step": 15080
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.026660442352295,
-      "learning_rate": 1.0108474576271189e-05,
-      "loss": 1.0689,
-      "step": 15090
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.522537708282471,
-      "learning_rate": 1.0101694915254238e-05,
-      "loss": 1.2961,
-      "step": 15100
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 2.490846872329712,
-      "learning_rate": 1.009491525423729e-05,
-      "loss": 1.4791,
-      "step": 15110
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 6.137222766876221,
-      "learning_rate": 1.0088135593220341e-05,
-      "loss": 1.4384,
-      "step": 15120
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 2.501014232635498,
-      "learning_rate": 1.008135593220339e-05,
-      "loss": 1.4224,
-      "step": 15130
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.077551364898682,
-      "learning_rate": 1.0074576271186442e-05,
-      "loss": 1.4564,
-      "step": 15140
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.0064074993133545,
-      "learning_rate": 1.0067796610169492e-05,
-      "loss": 1.3707,
-      "step": 15150
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 6.761713027954102,
-      "learning_rate": 1.0061016949152542e-05,
-      "loss": 1.2507,
-      "step": 15160
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 2.3495090007781982,
-      "learning_rate": 1.0054237288135593e-05,
-      "loss": 1.3328,
-      "step": 15170
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 2.105854034423828,
-      "learning_rate": 1.0047457627118644e-05,
-      "loss": 1.3227,
-      "step": 15180
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 2.0211234092712402,
-      "learning_rate": 1.0040677966101696e-05,
-      "loss": 1.3816,
-      "step": 15190
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.667546272277832,
-      "learning_rate": 1.0033898305084746e-05,
-      "loss": 1.1994,
-      "step": 15200
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 3.69437837600708,
-      "learning_rate": 1.0027118644067797e-05,
-      "loss": 1.3234,
-      "step": 15210
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 6.547877311706543,
-      "learning_rate": 1.0020338983050848e-05,
-      "loss": 1.4976,
-      "step": 15220
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.0074992179870605,
-      "learning_rate": 1.00135593220339e-05,
-      "loss": 1.141,
-      "step": 15230
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 6.511413097381592,
-      "learning_rate": 1.000677966101695e-05,
-      "loss": 1.447,
-      "step": 15240
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 5.6213603019714355,
-      "learning_rate": 1e-05,
-      "loss": 1.3133,
-      "step": 15250
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 5.2933831214904785,
-      "learning_rate": 9.993220338983052e-06,
-      "loss": 1.5486,
-      "step": 15260
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 7.773339748382568,
-      "learning_rate": 9.986440677966102e-06,
-      "loss": 1.192,
-      "step": 15270
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 2.686492443084717,
-      "learning_rate": 9.979661016949153e-06,
-      "loss": 1.3922,
-      "step": 15280
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.0056610107421875,
-      "learning_rate": 9.972881355932205e-06,
-      "loss": 1.3592,
-      "step": 15290
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 4.592884063720703,
-      "learning_rate": 9.966101694915256e-06,
-      "loss": 1.3173,
-      "step": 15300
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 5.68196964263916,
-      "learning_rate": 9.959322033898306e-06,
-      "loss": 1.3826,
-      "step": 15310
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 8.197639465332031,
-      "learning_rate": 9.952542372881356e-06,
-      "loss": 1.1825,
-      "step": 15320
-    },
-    {
-      "epoch": 0.35,
-      "grad_norm": 9.864594459533691,
-      "learning_rate": 9.945762711864407e-06,
-      "loss": 1.1862,
-      "step": 15330
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 7.495543003082275,
-      "learning_rate": 9.938983050847458e-06,
-      "loss": 1.2566,
-      "step": 15340
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 6.1277899742126465,
-      "learning_rate": 9.93220338983051e-06,
-      "loss": 1.3178,
-      "step": 15350
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 8.173718452453613,
-      "learning_rate": 9.92542372881356e-06,
-      "loss": 1.3092,
-      "step": 15360
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 3.474973678588867,
-      "learning_rate": 9.918644067796611e-06,
-      "loss": 1.3152,
-      "step": 15370
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 3.8019461631774902,
-      "learning_rate": 9.911864406779662e-06,
-      "loss": 1.4228,
-      "step": 15380
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 3.221346139907837,
-      "learning_rate": 9.905084745762714e-06,
-      "loss": 1.4204,
-      "step": 15390
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.06070613861084,
-      "learning_rate": 9.898305084745763e-06,
-      "loss": 1.3482,
-      "step": 15400
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.18265962600708,
-      "learning_rate": 9.891525423728813e-06,
-      "loss": 1.1693,
-      "step": 15410
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 1.873125433921814,
-      "learning_rate": 9.884745762711864e-06,
-      "loss": 1.3536,
-      "step": 15420
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.7039427757263184,
-      "learning_rate": 9.877966101694916e-06,
-      "loss": 1.3127,
-      "step": 15430
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 1.9308912754058838,
-      "learning_rate": 9.871186440677967e-06,
-      "loss": 1.2807,
-      "step": 15440
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.275124549865723,
-      "learning_rate": 9.864406779661017e-06,
-      "loss": 1.2304,
-      "step": 15450
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 15.040656089782715,
-      "learning_rate": 9.857627118644068e-06,
-      "loss": 1.1825,
-      "step": 15460
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 6.073838710784912,
-      "learning_rate": 9.85084745762712e-06,
-      "loss": 1.1309,
-      "step": 15470
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 6.772380828857422,
-      "learning_rate": 9.844067796610171e-06,
-      "loss": 1.4216,
-      "step": 15480
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.349153995513916,
-      "learning_rate": 9.837288135593221e-06,
-      "loss": 1.2706,
-      "step": 15490
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.798342227935791,
-      "learning_rate": 9.830508474576272e-06,
-      "loss": 1.3675,
-      "step": 15500
-    },
-    {
-      "epoch": 0.36,
-      "eval_loss": 1.0479532480239868,
-      "eval_runtime": 66.9728,
-      "eval_samples_per_second": 14.931,
-      "eval_steps_per_second": 14.931,
-      "step": 15500
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 3.6128780841827393,
-      "learning_rate": 9.823728813559322e-06,
-      "loss": 1.2832,
-      "step": 15510
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 3.2983522415161133,
-      "learning_rate": 9.816949152542373e-06,
-      "loss": 1.4409,
-      "step": 15520
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.769399881362915,
-      "learning_rate": 9.810169491525425e-06,
-      "loss": 1.2814,
-      "step": 15530
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 1.404224157333374,
-      "learning_rate": 9.803389830508474e-06,
-      "loss": 1.5151,
-      "step": 15540
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 1.0403872728347778,
-      "learning_rate": 9.796610169491526e-06,
-      "loss": 1.1822,
-      "step": 15550
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 8.834808349609375,
-      "learning_rate": 9.789830508474577e-06,
-      "loss": 1.2262,
-      "step": 15560
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.028998851776123,
-      "learning_rate": 9.783050847457629e-06,
-      "loss": 1.1775,
-      "step": 15570
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 6.36124849319458,
-      "learning_rate": 9.776271186440678e-06,
-      "loss": 1.2234,
-      "step": 15580
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 6.729506969451904,
-      "learning_rate": 9.76949152542373e-06,
-      "loss": 1.5841,
-      "step": 15590
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 1.4375039339065552,
-      "learning_rate": 9.762711864406781e-06,
-      "loss": 1.2468,
-      "step": 15600
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 5.384178161621094,
-      "learning_rate": 9.755932203389833e-06,
-      "loss": 1.572,
-      "step": 15610
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 6.532320499420166,
-      "learning_rate": 9.749152542372882e-06,
-      "loss": 1.2631,
-      "step": 15620
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 6.689794540405273,
-      "learning_rate": 9.742372881355932e-06,
-      "loss": 1.3595,
-      "step": 15630
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 3.815664768218994,
-      "learning_rate": 9.735593220338983e-06,
-      "loss": 1.3499,
-      "step": 15640
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 1.9858266115188599,
-      "learning_rate": 9.728813559322035e-06,
-      "loss": 1.2567,
-      "step": 15650
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.2085723876953125,
-      "learning_rate": 9.722033898305086e-06,
-      "loss": 1.3345,
-      "step": 15660
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.255419731140137,
-      "learning_rate": 9.715254237288136e-06,
-      "loss": 1.3081,
-      "step": 15670
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.9408528804779053,
-      "learning_rate": 9.708474576271187e-06,
-      "loss": 1.3258,
-      "step": 15680
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 7.384323596954346,
-      "learning_rate": 9.701694915254239e-06,
-      "loss": 1.2384,
-      "step": 15690
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.5653786659240723,
-      "learning_rate": 9.69491525423729e-06,
-      "loss": 1.3809,
-      "step": 15700
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 5.427888870239258,
-      "learning_rate": 9.68813559322034e-06,
-      "loss": 1.1215,
-      "step": 15710
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 5.703902244567871,
-      "learning_rate": 9.68135593220339e-06,
-      "loss": 1.3512,
-      "step": 15720
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.905301570892334,
-      "learning_rate": 9.674576271186441e-06,
-      "loss": 1.3938,
-      "step": 15730
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.5882115364074707,
-      "learning_rate": 9.667796610169492e-06,
-      "loss": 1.2503,
-      "step": 15740
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 2.825324535369873,
-      "learning_rate": 9.661016949152544e-06,
-      "loss": 1.2018,
-      "step": 15750
-    },
-    {
-      "epoch": 0.36,
-      "grad_norm": 4.857571601867676,
-      "learning_rate": 9.654237288135593e-06,
-      "loss": 1.5004,
-      "step": 15760
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.466376543045044,
-      "learning_rate": 9.647457627118645e-06,
-      "loss": 1.3539,
-      "step": 15770
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 6.780067443847656,
-      "learning_rate": 9.640677966101696e-06,
-      "loss": 1.2378,
-      "step": 15780
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 6.265717506408691,
-      "learning_rate": 9.633898305084746e-06,
-      "loss": 1.3072,
-      "step": 15790
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 2.7451460361480713,
-      "learning_rate": 9.627118644067797e-06,
-      "loss": 1.1819,
-      "step": 15800
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 4.156853675842285,
-      "learning_rate": 9.620338983050849e-06,
-      "loss": 1.3481,
-      "step": 15810
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.0684568881988525,
-      "learning_rate": 9.6135593220339e-06,
-      "loss": 1.2366,
-      "step": 15820
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 8.240663528442383,
-      "learning_rate": 9.60677966101695e-06,
-      "loss": 1.101,
-      "step": 15830
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 1.4690364599227905,
-      "learning_rate": 9.600000000000001e-06,
-      "loss": 1.2979,
-      "step": 15840
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 5.775753021240234,
-      "learning_rate": 9.593220338983051e-06,
-      "loss": 1.2432,
-      "step": 15850
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 6.069706439971924,
-      "learning_rate": 9.586440677966102e-06,
-      "loss": 1.0817,
-      "step": 15860
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.6312060356140137,
-      "learning_rate": 9.579661016949154e-06,
-      "loss": 1.3273,
-      "step": 15870
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 2.591268539428711,
-      "learning_rate": 9.572881355932203e-06,
-      "loss": 1.3013,
-      "step": 15880
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 9.177107810974121,
-      "learning_rate": 9.566101694915255e-06,
-      "loss": 1.4208,
-      "step": 15890
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 4.9722490310668945,
-      "learning_rate": 9.559322033898306e-06,
-      "loss": 1.3375,
-      "step": 15900
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 5.919930458068848,
-      "learning_rate": 9.552542372881358e-06,
-      "loss": 1.3232,
-      "step": 15910
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.5664663314819336,
-      "learning_rate": 9.545762711864407e-06,
-      "loss": 1.3428,
-      "step": 15920
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 2.6872286796569824,
-      "learning_rate": 9.538983050847457e-06,
-      "loss": 1.4764,
-      "step": 15930
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 1.3224186897277832,
-      "learning_rate": 9.532203389830508e-06,
-      "loss": 1.2986,
-      "step": 15940
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.2773120403289795,
-      "learning_rate": 9.52542372881356e-06,
-      "loss": 1.2231,
-      "step": 15950
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.5411736965179443,
-      "learning_rate": 9.518644067796611e-06,
-      "loss": 1.3613,
-      "step": 15960
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 2.742453098297119,
-      "learning_rate": 9.511864406779661e-06,
-      "loss": 1.3298,
-      "step": 15970
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 5.865776538848877,
-      "learning_rate": 9.505084745762712e-06,
-      "loss": 1.3194,
-      "step": 15980
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 4.6556715965271,
-      "learning_rate": 9.498305084745764e-06,
-      "loss": 1.33,
-      "step": 15990
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 8.305102348327637,
-      "learning_rate": 9.491525423728815e-06,
-      "loss": 1.3663,
-      "step": 16000
-    },
-    {
-      "epoch": 0.37,
-      "eval_loss": 1.01907479763031,
-      "eval_runtime": 67.0887,
-      "eval_samples_per_second": 14.906,
-      "eval_steps_per_second": 14.906,
-      "step": 16000
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 5.403275012969971,
-      "learning_rate": 9.484745762711865e-06,
-      "loss": 1.3266,
-      "step": 16010
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 7.849989891052246,
-      "learning_rate": 9.477966101694916e-06,
-      "loss": 1.2395,
-      "step": 16020
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 10.609501838684082,
-      "learning_rate": 9.471186440677966e-06,
-      "loss": 1.3162,
-      "step": 16030
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 4.230147361755371,
-      "learning_rate": 9.464406779661017e-06,
-      "loss": 1.4996,
-      "step": 16040
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 2.444842576980591,
-      "learning_rate": 9.457627118644069e-06,
-      "loss": 1.3524,
-      "step": 16050
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.158411741256714,
-      "learning_rate": 9.450847457627119e-06,
-      "loss": 1.4404,
-      "step": 16060
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 4.415703773498535,
-      "learning_rate": 9.44406779661017e-06,
-      "loss": 1.2824,
-      "step": 16070
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.985328435897827,
-      "learning_rate": 9.437288135593221e-06,
-      "loss": 1.1726,
-      "step": 16080
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 6.863582134246826,
-      "learning_rate": 9.430508474576273e-06,
-      "loss": 1.2735,
-      "step": 16090
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 4.635952472686768,
-      "learning_rate": 9.423728813559322e-06,
-      "loss": 1.1791,
-      "step": 16100
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 0.7141504287719727,
-      "learning_rate": 9.416949152542374e-06,
-      "loss": 1.2523,
-      "step": 16110
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.634265661239624,
-      "learning_rate": 9.410169491525425e-06,
-      "loss": 1.2288,
-      "step": 16120
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 2.960615634918213,
-      "learning_rate": 9.403389830508477e-06,
-      "loss": 1.1966,
-      "step": 16130
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 4.205376148223877,
-      "learning_rate": 9.396610169491526e-06,
-      "loss": 1.2102,
-      "step": 16140
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.364990472793579,
-      "learning_rate": 9.389830508474576e-06,
-      "loss": 1.2442,
-      "step": 16150
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 6.015055179595947,
-      "learning_rate": 9.383050847457627e-06,
-      "loss": 1.3283,
-      "step": 16160
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 3.9518237113952637,
-      "learning_rate": 9.376271186440679e-06,
-      "loss": 1.4116,
-      "step": 16170
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 7.87483024597168,
-      "learning_rate": 9.36949152542373e-06,
-      "loss": 1.4171,
-      "step": 16180
-    },
-    {
-      "epoch": 0.37,
-      "grad_norm": 1.998564600944519,
-      "learning_rate": 9.36271186440678e-06,
-      "loss": 1.275,
-      "step": 16190
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 1.7648671865463257,
-      "learning_rate": 9.355932203389831e-06,
-      "loss": 1.155,
-      "step": 16200
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 3.3635053634643555,
-      "learning_rate": 9.349152542372883e-06,
-      "loss": 1.245,
-      "step": 16210
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 5.173689842224121,
-      "learning_rate": 9.342372881355934e-06,
-      "loss": 1.1604,
-      "step": 16220
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 7.293972492218018,
-      "learning_rate": 9.335593220338984e-06,
-      "loss": 1.3044,
-      "step": 16230
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.6143198013305664,
-      "learning_rate": 9.328813559322034e-06,
-      "loss": 1.0515,
-      "step": 16240
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 10.650688171386719,
-      "learning_rate": 9.322033898305085e-06,
-      "loss": 1.0784,
-      "step": 16250
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 3.8562324047088623,
-      "learning_rate": 9.315254237288136e-06,
-      "loss": 1.1731,
-      "step": 16260
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 7.339954853057861,
-      "learning_rate": 9.308474576271188e-06,
-      "loss": 1.3072,
-      "step": 16270
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 11.781778335571289,
-      "learning_rate": 9.301694915254237e-06,
-      "loss": 1.369,
-      "step": 16280
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.4666757583618164,
-      "learning_rate": 9.294915254237289e-06,
-      "loss": 1.3628,
-      "step": 16290
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 3.2646923065185547,
-      "learning_rate": 9.28813559322034e-06,
-      "loss": 1.33,
-      "step": 16300
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.234771251678467,
-      "learning_rate": 9.28135593220339e-06,
-      "loss": 1.2017,
-      "step": 16310
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 10.070908546447754,
-      "learning_rate": 9.274576271186441e-06,
-      "loss": 1.019,
-      "step": 16320
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.7779605388641357,
-      "learning_rate": 9.267796610169493e-06,
-      "loss": 1.3126,
-      "step": 16330
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 0.898559033870697,
-      "learning_rate": 9.261016949152544e-06,
-      "loss": 1.1969,
-      "step": 16340
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 6.065027236938477,
-      "learning_rate": 9.254237288135594e-06,
-      "loss": 1.3629,
-      "step": 16350
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 18.8370361328125,
-      "learning_rate": 9.247457627118645e-06,
-      "loss": 1.4472,
-      "step": 16360
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 4.5132598876953125,
-      "learning_rate": 9.240677966101695e-06,
-      "loss": 1.219,
-      "step": 16370
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 5.263208389282227,
-      "learning_rate": 9.233898305084746e-06,
-      "loss": 1.2699,
-      "step": 16380
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.9755587577819824,
-      "learning_rate": 9.227118644067798e-06,
-      "loss": 1.3508,
-      "step": 16390
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.9797961711883545,
-      "learning_rate": 9.220338983050847e-06,
-      "loss": 1.0866,
-      "step": 16400
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 5.462897777557373,
-      "learning_rate": 9.213559322033899e-06,
-      "loss": 1.1406,
-      "step": 16410
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 7.175684928894043,
-      "learning_rate": 9.20677966101695e-06,
-      "loss": 1.1563,
-      "step": 16420
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 5.566606044769287,
-      "learning_rate": 9.200000000000002e-06,
-      "loss": 1.4053,
-      "step": 16430
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 5.2373762130737305,
-      "learning_rate": 9.193220338983051e-06,
-      "loss": 1.3666,
-      "step": 16440
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.5914816856384277,
-      "learning_rate": 9.186440677966101e-06,
-      "loss": 1.3011,
-      "step": 16450
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 7.714232921600342,
-      "learning_rate": 9.179661016949153e-06,
-      "loss": 1.3538,
-      "step": 16460
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 6.387141227722168,
-      "learning_rate": 9.172881355932204e-06,
-      "loss": 1.2126,
-      "step": 16470
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 12.002847671508789,
-      "learning_rate": 9.166101694915255e-06,
-      "loss": 1.3061,
-      "step": 16480
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 4.265731334686279,
-      "learning_rate": 9.159322033898305e-06,
-      "loss": 1.3956,
-      "step": 16490
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 7.408833980560303,
-      "learning_rate": 9.152542372881356e-06,
-      "loss": 1.2394,
-      "step": 16500
-    },
-    {
-      "epoch": 0.38,
-      "eval_loss": 1.063765048980713,
-      "eval_runtime": 65.9613,
-      "eval_samples_per_second": 15.16,
-      "eval_steps_per_second": 15.16,
-      "step": 16500
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 1.7360392808914185,
-      "learning_rate": 9.145762711864408e-06,
-      "loss": 1.2752,
-      "step": 16510
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 4.30725622177124,
-      "learning_rate": 9.13898305084746e-06,
-      "loss": 1.4356,
-      "step": 16520
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 3.4266469478607178,
-      "learning_rate": 9.132203389830509e-06,
-      "loss": 1.2648,
-      "step": 16530
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 4.342678070068359,
-      "learning_rate": 9.12542372881356e-06,
-      "loss": 1.452,
-      "step": 16540
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 3.7193539142608643,
-      "learning_rate": 9.11864406779661e-06,
-      "loss": 1.2244,
-      "step": 16550
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 4.744646072387695,
-      "learning_rate": 9.111864406779661e-06,
-      "loss": 1.1743,
-      "step": 16560
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 6.784905910491943,
-      "learning_rate": 9.105084745762713e-06,
-      "loss": 1.3432,
-      "step": 16570
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 3.6726789474487305,
-      "learning_rate": 9.098305084745763e-06,
-      "loss": 1.4602,
-      "step": 16580
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.33778715133667,
-      "learning_rate": 9.091525423728814e-06,
-      "loss": 1.4069,
-      "step": 16590
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 9.637640953063965,
-      "learning_rate": 9.084745762711865e-06,
-      "loss": 1.408,
-      "step": 16600
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 2.0369181632995605,
-      "learning_rate": 9.077966101694917e-06,
-      "loss": 1.3524,
-      "step": 16610
-    },
-    {
-      "epoch": 0.38,
-      "grad_norm": 4.292532920837402,
-      "learning_rate": 9.071186440677966e-06,
-      "loss": 1.1809,
-      "step": 16620
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 7.0022735595703125,
-      "learning_rate": 9.064406779661018e-06,
-      "loss": 1.2977,
-      "step": 16630
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 10.691664695739746,
-      "learning_rate": 9.05762711864407e-06,
-      "loss": 1.2714,
-      "step": 16640
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 11.661876678466797,
-      "learning_rate": 9.05084745762712e-06,
-      "loss": 1.3014,
-      "step": 16650
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 11.593371391296387,
-      "learning_rate": 9.04406779661017e-06,
-      "loss": 1.2261,
-      "step": 16660
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.483781337738037,
-      "learning_rate": 9.03728813559322e-06,
-      "loss": 1.3501,
-      "step": 16670
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.500730037689209,
-      "learning_rate": 9.030508474576271e-06,
-      "loss": 1.2256,
-      "step": 16680
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.4255197048187256,
-      "learning_rate": 9.023728813559323e-06,
-      "loss": 1.2245,
-      "step": 16690
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 5.569160461425781,
-      "learning_rate": 9.016949152542374e-06,
-      "loss": 1.4078,
-      "step": 16700
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 1.7080422639846802,
-      "learning_rate": 9.010169491525424e-06,
-      "loss": 1.4236,
-      "step": 16710
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.914741516113281,
-      "learning_rate": 9.003389830508475e-06,
-      "loss": 1.2597,
-      "step": 16720
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.742468357086182,
-      "learning_rate": 8.996610169491527e-06,
-      "loss": 1.1909,
-      "step": 16730
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 1.7503118515014648,
-      "learning_rate": 8.989830508474578e-06,
-      "loss": 1.3589,
-      "step": 16740
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.5486040115356445,
-      "learning_rate": 8.983050847457628e-06,
-      "loss": 1.3431,
-      "step": 16750
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 1.0202205181121826,
-      "learning_rate": 8.976271186440678e-06,
-      "loss": 1.3625,
-      "step": 16760
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 7.154101371765137,
-      "learning_rate": 8.969491525423729e-06,
-      "loss": 1.1777,
-      "step": 16770
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 2.6144134998321533,
-      "learning_rate": 8.96271186440678e-06,
-      "loss": 1.2143,
-      "step": 16780
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.3913540840148926,
-      "learning_rate": 8.955932203389832e-06,
-      "loss": 1.2127,
-      "step": 16790
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.7216756343841553,
-      "learning_rate": 8.949152542372881e-06,
-      "loss": 1.1532,
-      "step": 16800
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 8.174360275268555,
-      "learning_rate": 8.942372881355933e-06,
-      "loss": 1.2403,
-      "step": 16810
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.247581958770752,
-      "learning_rate": 8.935593220338984e-06,
-      "loss": 1.2231,
-      "step": 16820
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 2.393038749694824,
-      "learning_rate": 8.928813559322036e-06,
-      "loss": 1.3235,
-      "step": 16830
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 1.777172327041626,
-      "learning_rate": 8.922033898305085e-06,
-      "loss": 1.2097,
-      "step": 16840
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.403905868530273,
-      "learning_rate": 8.915254237288137e-06,
-      "loss": 1.1161,
-      "step": 16850
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.6407907009124756,
-      "learning_rate": 8.908474576271188e-06,
-      "loss": 1.3792,
-      "step": 16860
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.673400402069092,
-      "learning_rate": 8.901694915254238e-06,
-      "loss": 1.2941,
-      "step": 16870
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 5.5751848220825195,
-      "learning_rate": 8.89491525423729e-06,
-      "loss": 1.2355,
-      "step": 16880
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.737181186676025,
-      "learning_rate": 8.888135593220339e-06,
-      "loss": 1.1943,
-      "step": 16890
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.271194934844971,
-      "learning_rate": 8.88135593220339e-06,
-      "loss": 1.2646,
-      "step": 16900
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 1.9672961235046387,
-      "learning_rate": 8.874576271186442e-06,
-      "loss": 1.3767,
-      "step": 16910
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.734816074371338,
-      "learning_rate": 8.867796610169492e-06,
-      "loss": 1.2552,
-      "step": 16920
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.603008270263672,
-      "learning_rate": 8.861016949152543e-06,
-      "loss": 1.428,
-      "step": 16930
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 9.842779159545898,
-      "learning_rate": 8.854237288135594e-06,
-      "loss": 1.2608,
-      "step": 16940
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.2394957542419434,
-      "learning_rate": 8.847457627118646e-06,
-      "loss": 1.1189,
-      "step": 16950
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.726215362548828,
-      "learning_rate": 8.840677966101695e-06,
-      "loss": 1.3895,
-      "step": 16960
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.315942287445068,
-      "learning_rate": 8.833898305084747e-06,
-      "loss": 1.2961,
-      "step": 16970
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 2.8540353775024414,
-      "learning_rate": 8.827118644067797e-06,
-      "loss": 1.2502,
-      "step": 16980
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 5.723995685577393,
-      "learning_rate": 8.820338983050848e-06,
-      "loss": 1.3241,
-      "step": 16990
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 4.132941246032715,
-      "learning_rate": 8.8135593220339e-06,
-      "loss": 1.2777,
-      "step": 17000
-    },
-    {
-      "epoch": 0.39,
-      "eval_loss": 1.0359538793563843,
-      "eval_runtime": 66.4953,
-      "eval_samples_per_second": 15.039,
-      "eval_steps_per_second": 15.039,
-      "step": 17000
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.6101505756378174,
-      "learning_rate": 8.806779661016949e-06,
-      "loss": 1.302,
-      "step": 17010
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 2.8257572650909424,
-      "learning_rate": 8.8e-06,
-      "loss": 1.3581,
-      "step": 17020
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 2.660344123840332,
-      "learning_rate": 8.793220338983052e-06,
-      "loss": 1.2804,
-      "step": 17030
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 9.050156593322754,
-      "learning_rate": 8.786440677966103e-06,
-      "loss": 1.2465,
-      "step": 17040
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 3.4612698554992676,
-      "learning_rate": 8.779661016949153e-06,
-      "loss": 1.4898,
-      "step": 17050
-    },
-    {
-      "epoch": 0.39,
-      "grad_norm": 10.567662239074707,
-      "learning_rate": 8.772881355932204e-06,
-      "loss": 1.1548,
-      "step": 17060
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 5.1014275550842285,
-      "learning_rate": 8.766101694915254e-06,
-      "loss": 1.3642,
-      "step": 17070
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 7.804174423217773,
-      "learning_rate": 8.759322033898305e-06,
-      "loss": 1.2271,
-      "step": 17080
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 1.9340236186981201,
-      "learning_rate": 8.752542372881357e-06,
-      "loss": 1.3143,
-      "step": 17090
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 7.017419815063477,
-      "learning_rate": 8.745762711864407e-06,
-      "loss": 1.189,
-      "step": 17100
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 7.2211995124816895,
-      "learning_rate": 8.738983050847458e-06,
-      "loss": 1.3399,
-      "step": 17110
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.3273162841796875,
-      "learning_rate": 8.73220338983051e-06,
-      "loss": 1.3636,
-      "step": 17120
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.173155784606934,
-      "learning_rate": 8.72542372881356e-06,
-      "loss": 1.2755,
-      "step": 17130
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.535107374191284,
-      "learning_rate": 8.71864406779661e-06,
-      "loss": 1.2036,
-      "step": 17140
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 5.383171081542969,
-      "learning_rate": 8.711864406779662e-06,
-      "loss": 1.2716,
-      "step": 17150
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.0449705123901367,
-      "learning_rate": 8.705084745762713e-06,
-      "loss": 1.287,
-      "step": 17160
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 6.901495933532715,
-      "learning_rate": 8.698305084745765e-06,
-      "loss": 1.4108,
-      "step": 17170
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.3814964294433594,
-      "learning_rate": 8.691525423728814e-06,
-      "loss": 1.3023,
-      "step": 17180
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 2.271169424057007,
-      "learning_rate": 8.684745762711864e-06,
-      "loss": 1.2989,
-      "step": 17190
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.1464428901672363,
-      "learning_rate": 8.677966101694915e-06,
-      "loss": 1.2671,
-      "step": 17200
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.007338047027588,
-      "learning_rate": 8.671186440677967e-06,
-      "loss": 1.3452,
-      "step": 17210
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 5.046366214752197,
-      "learning_rate": 8.664406779661018e-06,
-      "loss": 1.3612,
-      "step": 17220
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 2.2831978797912598,
-      "learning_rate": 8.657627118644068e-06,
-      "loss": 1.2544,
-      "step": 17230
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 2.5876657962799072,
-      "learning_rate": 8.65084745762712e-06,
-      "loss": 1.1922,
-      "step": 17240
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.5324642658233643,
-      "learning_rate": 8.64406779661017e-06,
-      "loss": 1.2893,
-      "step": 17250
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.1740715503692627,
-      "learning_rate": 8.637288135593222e-06,
-      "loss": 1.3322,
-      "step": 17260
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 6.979104995727539,
-      "learning_rate": 8.630508474576272e-06,
-      "loss": 1.0987,
-      "step": 17270
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 2.1378564834594727,
-      "learning_rate": 8.623728813559322e-06,
-      "loss": 1.2972,
-      "step": 17280
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.754072666168213,
-      "learning_rate": 8.616949152542373e-06,
-      "loss": 1.1845,
-      "step": 17290
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 5.770691394805908,
-      "learning_rate": 8.610169491525424e-06,
-      "loss": 1.2964,
-      "step": 17300
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 1.7626854181289673,
-      "learning_rate": 8.603389830508476e-06,
-      "loss": 1.2309,
-      "step": 17310
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.450747966766357,
-      "learning_rate": 8.596610169491526e-06,
-      "loss": 1.3676,
-      "step": 17320
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.775254249572754,
-      "learning_rate": 8.589830508474577e-06,
-      "loss": 1.3344,
-      "step": 17330
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 8.103263854980469,
-      "learning_rate": 8.583050847457628e-06,
-      "loss": 1.3521,
-      "step": 17340
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 6.5187201499938965,
-      "learning_rate": 8.57627118644068e-06,
-      "loss": 1.323,
-      "step": 17350
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.5734620094299316,
-      "learning_rate": 8.56949152542373e-06,
-      "loss": 1.3845,
-      "step": 17360
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 6.281393527984619,
-      "learning_rate": 8.56271186440678e-06,
-      "loss": 1.2092,
-      "step": 17370
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.1913156509399414,
-      "learning_rate": 8.55593220338983e-06,
-      "loss": 1.2908,
-      "step": 17380
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 12.576632499694824,
-      "learning_rate": 8.549152542372882e-06,
-      "loss": 1.2864,
-      "step": 17390
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.168999195098877,
-      "learning_rate": 8.542372881355933e-06,
-      "loss": 1.3254,
-      "step": 17400
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.167811870574951,
-      "learning_rate": 8.535593220338983e-06,
-      "loss": 1.2052,
-      "step": 17410
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 6.4759135246276855,
-      "learning_rate": 8.528813559322034e-06,
-      "loss": 1.2101,
-      "step": 17420
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.099415302276611,
-      "learning_rate": 8.522033898305086e-06,
-      "loss": 1.2088,
-      "step": 17430
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 5.991481781005859,
-      "learning_rate": 8.515254237288136e-06,
-      "loss": 1.3606,
-      "step": 17440
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 5.722553253173828,
-      "learning_rate": 8.508474576271187e-06,
-      "loss": 1.2074,
-      "step": 17450
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 5.729592800140381,
-      "learning_rate": 8.501694915254238e-06,
-      "loss": 1.2838,
-      "step": 17460
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 3.647024393081665,
-      "learning_rate": 8.49491525423729e-06,
-      "loss": 1.2778,
-      "step": 17470
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 4.270313262939453,
-      "learning_rate": 8.48813559322034e-06,
-      "loss": 1.3161,
-      "step": 17480
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 8.540990829467773,
-      "learning_rate": 8.481355932203391e-06,
-      "loss": 1.1557,
-      "step": 17490
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.2297253608703613,
-      "learning_rate": 8.47457627118644e-06,
-      "loss": 1.4842,
-      "step": 17500
-    },
-    {
-      "epoch": 0.41,
-      "eval_loss": 1.0164830684661865,
-      "eval_runtime": 67.3932,
-      "eval_samples_per_second": 14.838,
-      "eval_steps_per_second": 14.838,
-      "step": 17500
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 1.271462082862854,
-      "learning_rate": 8.467796610169492e-06,
-      "loss": 1.3579,
-      "step": 17510
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.716254711151123,
-      "learning_rate": 8.461016949152543e-06,
-      "loss": 1.3783,
-      "step": 17520
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 4.856709957122803,
-      "learning_rate": 8.454237288135593e-06,
-      "loss": 1.2537,
-      "step": 17530
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 11.454696655273438,
-      "learning_rate": 8.447457627118644e-06,
-      "loss": 1.1546,
-      "step": 17540
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.388411045074463,
-      "learning_rate": 8.440677966101696e-06,
-      "loss": 1.3001,
-      "step": 17550
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 1.902894139289856,
-      "learning_rate": 8.433898305084747e-06,
-      "loss": 1.1925,
-      "step": 17560
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 2.1532018184661865,
-      "learning_rate": 8.427118644067797e-06,
-      "loss": 1.3065,
-      "step": 17570
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.9372098445892334,
-      "learning_rate": 8.420338983050848e-06,
-      "loss": 1.3958,
-      "step": 17580
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 7.873089790344238,
-      "learning_rate": 8.413559322033898e-06,
-      "loss": 1.1727,
-      "step": 17590
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.4659693241119385,
-      "learning_rate": 8.40677966101695e-06,
-      "loss": 1.355,
-      "step": 17600
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 18.613733291625977,
-      "learning_rate": 8.400000000000001e-06,
-      "loss": 1.3742,
-      "step": 17610
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.247617721557617,
-      "learning_rate": 8.39322033898305e-06,
-      "loss": 1.1606,
-      "step": 17620
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 2.1550047397613525,
-      "learning_rate": 8.386440677966102e-06,
-      "loss": 1.3164,
-      "step": 17630
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 6.266103744506836,
-      "learning_rate": 8.379661016949153e-06,
-      "loss": 1.3522,
-      "step": 17640
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 2.7962443828582764,
-      "learning_rate": 8.372881355932205e-06,
-      "loss": 1.4491,
-      "step": 17650
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 5.119645118713379,
-      "learning_rate": 8.366101694915255e-06,
-      "loss": 1.447,
-      "step": 17660
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 6.110286712646484,
-      "learning_rate": 8.359322033898306e-06,
-      "loss": 1.2631,
-      "step": 17670
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 9.99429702758789,
-      "learning_rate": 8.352542372881357e-06,
-      "loss": 1.3755,
-      "step": 17680
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 2.3583624362945557,
-      "learning_rate": 8.345762711864409e-06,
-      "loss": 1.4084,
-      "step": 17690
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.867703437805176,
-      "learning_rate": 8.338983050847458e-06,
-      "loss": 1.3138,
-      "step": 17700
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 8.489445686340332,
-      "learning_rate": 8.332203389830508e-06,
-      "loss": 1.2486,
-      "step": 17710
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 5.573801040649414,
-      "learning_rate": 8.32542372881356e-06,
-      "loss": 1.2657,
-      "step": 17720
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 5.751363754272461,
-      "learning_rate": 8.318644067796611e-06,
-      "loss": 1.2009,
-      "step": 17730
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.761434316635132,
-      "learning_rate": 8.311864406779662e-06,
-      "loss": 1.3223,
-      "step": 17740
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.0040767192840576,
-      "learning_rate": 8.305084745762712e-06,
-      "loss": 1.2842,
-      "step": 17750
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 4.172005653381348,
-      "learning_rate": 8.298305084745763e-06,
-      "loss": 1.2445,
-      "step": 17760
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 6.599450588226318,
-      "learning_rate": 8.291525423728815e-06,
-      "loss": 1.0738,
-      "step": 17770
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 7.292091369628906,
-      "learning_rate": 8.284745762711866e-06,
-      "loss": 1.3037,
-      "step": 17780
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 2.842538595199585,
-      "learning_rate": 8.277966101694916e-06,
-      "loss": 1.3071,
-      "step": 17790
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 4.334558010101318,
-      "learning_rate": 8.271186440677966e-06,
-      "loss": 1.5332,
-      "step": 17800
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.017799139022827,
-      "learning_rate": 8.264406779661017e-06,
-      "loss": 1.3143,
-      "step": 17810
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 8.827996253967285,
-      "learning_rate": 8.257627118644068e-06,
-      "loss": 1.0828,
-      "step": 17820
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 10.552117347717285,
-      "learning_rate": 8.25084745762712e-06,
-      "loss": 1.2198,
-      "step": 17830
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 4.006974697113037,
-      "learning_rate": 8.24406779661017e-06,
-      "loss": 1.2621,
-      "step": 17840
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 4.397966384887695,
-      "learning_rate": 8.237288135593221e-06,
-      "loss": 1.1562,
-      "step": 17850
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 7.467845916748047,
-      "learning_rate": 8.230508474576272e-06,
-      "loss": 1.3113,
-      "step": 17860
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.8428890705108643,
-      "learning_rate": 8.223728813559324e-06,
-      "loss": 0.9949,
-      "step": 17870
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 3.2165286540985107,
-      "learning_rate": 8.216949152542373e-06,
-      "loss": 1.3019,
-      "step": 17880
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 9.67989444732666,
-      "learning_rate": 8.210169491525425e-06,
-      "loss": 1.2628,
-      "step": 17890
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 0.9366636276245117,
-      "learning_rate": 8.203389830508475e-06,
-      "loss": 1.0139,
-      "step": 17900
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 6.279294967651367,
-      "learning_rate": 8.196610169491526e-06,
-      "loss": 1.2977,
-      "step": 17910
-    },
-    {
-      "epoch": 0.41,
-      "grad_norm": 5.070333003997803,
-      "learning_rate": 8.189830508474577e-06,
-      "loss": 1.3068,
-      "step": 17920
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.6785717010498047,
-      "learning_rate": 8.183050847457627e-06,
-      "loss": 1.2096,
-      "step": 17930
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 2.7473363876342773,
-      "learning_rate": 8.176271186440678e-06,
-      "loss": 1.2947,
-      "step": 17940
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 8.139301300048828,
-      "learning_rate": 8.16949152542373e-06,
-      "loss": 1.3725,
-      "step": 17950
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 5.076933860778809,
-      "learning_rate": 8.162711864406781e-06,
-      "loss": 1.3476,
-      "step": 17960
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 35.75548553466797,
-      "learning_rate": 8.155932203389831e-06,
-      "loss": 1.2893,
-      "step": 17970
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 5.612383842468262,
-      "learning_rate": 8.149152542372882e-06,
-      "loss": 1.278,
-      "step": 17980
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 4.277181148529053,
-      "learning_rate": 8.142372881355934e-06,
-      "loss": 1.2526,
-      "step": 17990
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.2299575805664062,
-      "learning_rate": 8.135593220338983e-06,
-      "loss": 1.2866,
-      "step": 18000
-    },
-    {
-      "epoch": 0.42,
-      "eval_loss": 1.0518407821655273,
-      "eval_runtime": 67.5281,
-      "eval_samples_per_second": 14.809,
-      "eval_steps_per_second": 14.809,
-      "step": 18000
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.5516388416290283,
-      "learning_rate": 8.128813559322035e-06,
-      "loss": 1.3295,
-      "step": 18010
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 5.150609970092773,
-      "learning_rate": 8.122033898305085e-06,
-      "loss": 1.1718,
-      "step": 18020
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 9.092333793640137,
-      "learning_rate": 8.115254237288136e-06,
-      "loss": 1.2794,
-      "step": 18030
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 4.030599117279053,
-      "learning_rate": 8.108474576271187e-06,
-      "loss": 1.3244,
-      "step": 18040
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 2.527414083480835,
-      "learning_rate": 8.101694915254237e-06,
-      "loss": 1.3997,
-      "step": 18050
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.05289626121521,
-      "learning_rate": 8.094915254237289e-06,
-      "loss": 1.4377,
-      "step": 18060
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 1.4701480865478516,
-      "learning_rate": 8.08813559322034e-06,
-      "loss": 1.2932,
-      "step": 18070
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 7.444494247436523,
-      "learning_rate": 8.081355932203391e-06,
-      "loss": 1.171,
-      "step": 18080
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.2663989067077637,
-      "learning_rate": 8.074576271186441e-06,
-      "loss": 1.4609,
-      "step": 18090
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.3425662517547607,
-      "learning_rate": 8.067796610169492e-06,
-      "loss": 1.2233,
-      "step": 18100
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 4.549169063568115,
-      "learning_rate": 8.061016949152542e-06,
-      "loss": 1.3096,
-      "step": 18110
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 7.133522033691406,
-      "learning_rate": 8.054237288135594e-06,
-      "loss": 1.3431,
-      "step": 18120
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.107785224914551,
-      "learning_rate": 8.047457627118645e-06,
-      "loss": 1.2448,
-      "step": 18130
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.182917356491089,
-      "learning_rate": 8.040677966101695e-06,
-      "loss": 1.2348,
-      "step": 18140
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 10.907735824584961,
-      "learning_rate": 8.033898305084746e-06,
-      "loss": 1.102,
-      "step": 18150
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 4.007683753967285,
-      "learning_rate": 8.027118644067797e-06,
-      "loss": 1.3091,
-      "step": 18160
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 7.087812900543213,
-      "learning_rate": 8.020338983050849e-06,
-      "loss": 1.3511,
-      "step": 18170
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.0910608768463135,
-      "learning_rate": 8.013559322033899e-06,
-      "loss": 1.2003,
-      "step": 18180
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 8.238131523132324,
-      "learning_rate": 8.00677966101695e-06,
-      "loss": 1.0435,
-      "step": 18190
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 1.3896546363830566,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 1.3478,
-      "step": 18200
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 2.614786148071289,
-      "learning_rate": 7.993220338983053e-06,
-      "loss": 1.4047,
-      "step": 18210
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 5.2004714012146,
-      "learning_rate": 7.986440677966102e-06,
-      "loss": 1.2459,
-      "step": 18220
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 8.22817325592041,
-      "learning_rate": 7.979661016949152e-06,
-      "loss": 1.3499,
-      "step": 18230
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.287627696990967,
-      "learning_rate": 7.972881355932204e-06,
-      "loss": 1.1953,
-      "step": 18240
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 7.237671852111816,
-      "learning_rate": 7.966101694915255e-06,
-      "loss": 1.254,
-      "step": 18250
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 2.2025198936462402,
-      "learning_rate": 7.959322033898306e-06,
-      "loss": 1.2611,
-      "step": 18260
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 3.426156520843506,
-      "learning_rate": 7.952542372881356e-06,
-      "loss": 1.2885,
-      "step": 18270
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 1.2702093124389648,
-      "learning_rate": 7.945762711864407e-06,
-      "loss": 1.1662,
-      "step": 18280
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 4.970961570739746,
-      "learning_rate": 7.938983050847459e-06,
-      "loss": 1.3333,
-      "step": 18290
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 11.826562881469727,
-      "learning_rate": 7.93220338983051e-06,
-      "loss": 1.2809,
-      "step": 18300
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 5.391317367553711,
-      "learning_rate": 7.92542372881356e-06,
-      "loss": 1.2781,
-      "step": 18310
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 4.817717552185059,
-      "learning_rate": 7.91864406779661e-06,
-      "loss": 1.2691,
-      "step": 18320
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 7.846217632293701,
-      "learning_rate": 7.911864406779661e-06,
-      "loss": 1.3397,
-      "step": 18330
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 2.251600742340088,
-      "learning_rate": 7.905084745762712e-06,
-      "loss": 1.1359,
-      "step": 18340
-    },
-    {
-      "epoch": 0.42,
-      "grad_norm": 2.6260218620300293,
-      "learning_rate": 7.898305084745764e-06,
-      "loss": 1.2603,
-      "step": 18350
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 1.5109496116638184,
-      "learning_rate": 7.891525423728814e-06,
-      "loss": 1.1711,
-      "step": 18360
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 4.166934013366699,
-      "learning_rate": 7.884745762711865e-06,
-      "loss": 1.2943,
-      "step": 18370
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.4080865383148193,
-      "learning_rate": 7.877966101694916e-06,
-      "loss": 1.3551,
-      "step": 18380
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.820498466491699,
-      "learning_rate": 7.871186440677968e-06,
-      "loss": 1.2382,
-      "step": 18390
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 7.695006847381592,
-      "learning_rate": 7.864406779661017e-06,
-      "loss": 1.295,
-      "step": 18400
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 9.21014404296875,
-      "learning_rate": 7.857627118644069e-06,
-      "loss": 1.0621,
-      "step": 18410
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.656609535217285,
-      "learning_rate": 7.850847457627119e-06,
-      "loss": 1.2838,
-      "step": 18420
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 7.399263381958008,
-      "learning_rate": 7.84406779661017e-06,
-      "loss": 1.2595,
-      "step": 18430
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.139892101287842,
-      "learning_rate": 7.837288135593221e-06,
-      "loss": 1.2488,
-      "step": 18440
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.8797223567962646,
-      "learning_rate": 7.830508474576271e-06,
-      "loss": 1.1638,
-      "step": 18450
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.705308198928833,
-      "learning_rate": 7.823728813559322e-06,
-      "loss": 1.22,
-      "step": 18460
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 7.386727333068848,
-      "learning_rate": 7.816949152542374e-06,
-      "loss": 1.3104,
-      "step": 18470
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.4528744220733643,
-      "learning_rate": 7.810169491525425e-06,
-      "loss": 1.3548,
-      "step": 18480
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 5.502497673034668,
-      "learning_rate": 7.803389830508475e-06,
-      "loss": 1.1125,
-      "step": 18490
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 6.493344783782959,
-      "learning_rate": 7.796610169491526e-06,
-      "loss": 1.5306,
-      "step": 18500
-    },
-    {
-      "epoch": 0.43,
-      "eval_loss": 1.0274684429168701,
-      "eval_runtime": 68.6334,
-      "eval_samples_per_second": 14.57,
-      "eval_steps_per_second": 14.57,
-      "step": 18500
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 8.007162094116211,
-      "learning_rate": 7.789830508474578e-06,
-      "loss": 1.4575,
-      "step": 18510
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 4.510526657104492,
-      "learning_rate": 7.783050847457628e-06,
-      "loss": 1.3597,
-      "step": 18520
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 6.399461269378662,
-      "learning_rate": 7.776271186440679e-06,
-      "loss": 1.249,
-      "step": 18530
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.6798839569091797,
-      "learning_rate": 7.769491525423729e-06,
-      "loss": 1.1667,
-      "step": 18540
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.981032133102417,
-      "learning_rate": 7.76271186440678e-06,
-      "loss": 1.4405,
-      "step": 18550
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.538968086242676,
-      "learning_rate": 7.755932203389831e-06,
-      "loss": 1.2578,
-      "step": 18560
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 1.8838539123535156,
-      "learning_rate": 7.749152542372881e-06,
-      "loss": 1.1573,
-      "step": 18570
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.920963764190674,
-      "learning_rate": 7.742372881355933e-06,
-      "loss": 1.3485,
-      "step": 18580
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 1.1590783596038818,
-      "learning_rate": 7.735593220338984e-06,
-      "loss": 1.28,
-      "step": 18590
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 7.310429573059082,
-      "learning_rate": 7.728813559322035e-06,
-      "loss": 1.5065,
-      "step": 18600
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.0086493492126465,
-      "learning_rate": 7.722033898305085e-06,
-      "loss": 1.4836,
-      "step": 18610
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.2605090141296387,
-      "learning_rate": 7.715254237288136e-06,
-      "loss": 1.3881,
-      "step": 18620
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.652811288833618,
-      "learning_rate": 7.708474576271186e-06,
-      "loss": 1.218,
-      "step": 18630
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 1.251134991645813,
-      "learning_rate": 7.701694915254238e-06,
-      "loss": 1.3565,
-      "step": 18640
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 4.377309322357178,
-      "learning_rate": 7.694915254237289e-06,
-      "loss": 1.1737,
-      "step": 18650
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 6.178280353546143,
-      "learning_rate": 7.688135593220339e-06,
-      "loss": 1.241,
-      "step": 18660
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 1.76272451877594,
-      "learning_rate": 7.68135593220339e-06,
-      "loss": 1.1855,
-      "step": 18670
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 5.491523742675781,
-      "learning_rate": 7.674576271186441e-06,
-      "loss": 1.3062,
-      "step": 18680
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.912921190261841,
-      "learning_rate": 7.667796610169493e-06,
-      "loss": 1.2213,
-      "step": 18690
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.2884957790374756,
-      "learning_rate": 7.661016949152543e-06,
-      "loss": 1.2815,
-      "step": 18700
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 2.6772258281707764,
-      "learning_rate": 7.654237288135594e-06,
-      "loss": 1.109,
-      "step": 18710
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 5.152644157409668,
-      "learning_rate": 7.647457627118645e-06,
-      "loss": 1.2494,
-      "step": 18720
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 6.88754940032959,
-      "learning_rate": 7.640677966101695e-06,
-      "loss": 1.3743,
-      "step": 18730
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 5.835016250610352,
-      "learning_rate": 7.633898305084746e-06,
-      "loss": 1.4158,
-      "step": 18740
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 5.073338508605957,
-      "learning_rate": 7.627118644067797e-06,
-      "loss": 1.1477,
-      "step": 18750
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 3.2228505611419678,
-      "learning_rate": 7.6203389830508476e-06,
-      "loss": 0.979,
-      "step": 18760
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 6.220423698425293,
-      "learning_rate": 7.613559322033899e-06,
-      "loss": 1.2679,
-      "step": 18770
-    },
-    {
-      "epoch": 0.43,
-      "grad_norm": 5.256744861602783,
-      "learning_rate": 7.6067796610169495e-06,
-      "loss": 1.1012,
-      "step": 18780
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 7.843850612640381,
-      "learning_rate": 7.600000000000001e-06,
-      "loss": 1.2904,
-      "step": 18790
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 7.586652755737305,
-      "learning_rate": 7.5932203389830515e-06,
-      "loss": 1.3106,
-      "step": 18800
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 8.580673217773438,
-      "learning_rate": 7.586440677966103e-06,
-      "loss": 1.0683,
-      "step": 18810
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 9.023046493530273,
-      "learning_rate": 7.5796610169491534e-06,
-      "loss": 1.3849,
-      "step": 18820
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 8.47220516204834,
-      "learning_rate": 7.572881355932205e-06,
-      "loss": 1.3014,
-      "step": 18830
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.808041572570801,
-      "learning_rate": 7.5661016949152545e-06,
-      "loss": 1.4718,
-      "step": 18840
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.828659534454346,
-      "learning_rate": 7.559322033898305e-06,
-      "loss": 1.226,
-      "step": 18850
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 3.637918710708618,
-      "learning_rate": 7.5525423728813565e-06,
-      "loss": 1.4644,
-      "step": 18860
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.678231716156006,
-      "learning_rate": 7.545762711864407e-06,
-      "loss": 1.3399,
-      "step": 18870
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 2.2599503993988037,
-      "learning_rate": 7.5389830508474584e-06,
-      "loss": 1.395,
-      "step": 18880
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 1.7738438844680786,
-      "learning_rate": 7.532203389830509e-06,
-      "loss": 1.2668,
-      "step": 18890
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 7.916548252105713,
-      "learning_rate": 7.52542372881356e-06,
-      "loss": 1.3632,
-      "step": 18900
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 9.459602355957031,
-      "learning_rate": 7.518644067796611e-06,
-      "loss": 1.4194,
-      "step": 18910
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 6.116416931152344,
-      "learning_rate": 7.511864406779662e-06,
-      "loss": 1.3118,
-      "step": 18920
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.6954731941223145,
-      "learning_rate": 7.505084745762713e-06,
-      "loss": 1.1006,
-      "step": 18930
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 2.977074384689331,
-      "learning_rate": 7.498305084745763e-06,
-      "loss": 1.0664,
-      "step": 18940
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.856612682342529,
-      "learning_rate": 7.491525423728814e-06,
-      "loss": 1.2166,
-      "step": 18950
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.739044189453125,
-      "learning_rate": 7.4847457627118646e-06,
-      "loss": 1.1675,
-      "step": 18960
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 2.360950231552124,
-      "learning_rate": 7.477966101694916e-06,
-      "loss": 1.3736,
-      "step": 18970
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.020655155181885,
-      "learning_rate": 7.4711864406779665e-06,
-      "loss": 1.4255,
-      "step": 18980
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 10.082642555236816,
-      "learning_rate": 7.464406779661018e-06,
-      "loss": 1.2732,
-      "step": 18990
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.41738748550415,
-      "learning_rate": 7.4576271186440685e-06,
-      "loss": 1.4897,
-      "step": 19000
-    },
-    {
-      "epoch": 0.44,
-      "eval_loss": 1.0319198369979858,
-      "eval_runtime": 66.943,
-      "eval_samples_per_second": 14.938,
-      "eval_steps_per_second": 14.938,
-      "step": 19000
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 3.4302687644958496,
-      "learning_rate": 7.45084745762712e-06,
-      "loss": 1.1379,
-      "step": 19010
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.5467529296875,
-      "learning_rate": 7.4440677966101704e-06,
-      "loss": 1.1605,
-      "step": 19020
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 3.840532064437866,
-      "learning_rate": 7.437288135593221e-06,
-      "loss": 1.318,
-      "step": 19030
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 8.278573989868164,
-      "learning_rate": 7.430508474576272e-06,
-      "loss": 1.4069,
-      "step": 19040
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 8.6283597946167,
-      "learning_rate": 7.423728813559322e-06,
-      "loss": 1.2437,
-      "step": 19050
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.682373523712158,
-      "learning_rate": 7.4169491525423735e-06,
-      "loss": 1.2322,
-      "step": 19060
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.24190616607666,
-      "learning_rate": 7.410169491525424e-06,
-      "loss": 1.137,
-      "step": 19070
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.345545291900635,
-      "learning_rate": 7.4033898305084754e-06,
-      "loss": 1.2763,
-      "step": 19080
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 6.952364444732666,
-      "learning_rate": 7.396610169491526e-06,
-      "loss": 1.3103,
-      "step": 19090
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 4.714378356933594,
-      "learning_rate": 7.3898305084745766e-06,
-      "loss": 1.3069,
-      "step": 19100
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 1.7550067901611328,
-      "learning_rate": 7.383050847457628e-06,
-      "loss": 1.3664,
-      "step": 19110
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.599852561950684,
-      "learning_rate": 7.3762711864406785e-06,
-      "loss": 1.2475,
-      "step": 19120
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 7.171874046325684,
-      "learning_rate": 7.36949152542373e-06,
-      "loss": 1.1256,
-      "step": 19130
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 2.445876359939575,
-      "learning_rate": 7.3627118644067805e-06,
-      "loss": 1.3442,
-      "step": 19140
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 3.083585500717163,
-      "learning_rate": 7.355932203389831e-06,
-      "loss": 1.3258,
-      "step": 19150
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.623142719268799,
-      "learning_rate": 7.3491525423728816e-06,
-      "loss": 1.1306,
-      "step": 19160
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 7.293304920196533,
-      "learning_rate": 7.342372881355932e-06,
-      "loss": 1.2778,
-      "step": 19170
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 2.301374673843384,
-      "learning_rate": 7.3355932203389835e-06,
-      "loss": 1.3221,
-      "step": 19180
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 5.147269248962402,
-      "learning_rate": 7.328813559322034e-06,
-      "loss": 1.1004,
-      "step": 19190
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 3.936070203781128,
-      "learning_rate": 7.3220338983050855e-06,
-      "loss": 1.3795,
-      "step": 19200
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 1.6208925247192383,
-      "learning_rate": 7.315254237288136e-06,
-      "loss": 1.2793,
-      "step": 19210
-    },
-    {
-      "epoch": 0.44,
-      "grad_norm": 6.317408084869385,
-      "learning_rate": 7.3084745762711874e-06,
-      "loss": 1.3759,
-      "step": 19220
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 3.9492716789245605,
-      "learning_rate": 7.301694915254238e-06,
-      "loss": 1.1688,
-      "step": 19230
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.192453384399414,
-      "learning_rate": 7.294915254237289e-06,
-      "loss": 1.2945,
-      "step": 19240
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 3.642240524291992,
-      "learning_rate": 7.288135593220339e-06,
-      "loss": 1.3392,
-      "step": 19250
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 1.0650874376296997,
-      "learning_rate": 7.28135593220339e-06,
-      "loss": 1.202,
-      "step": 19260
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.307705402374268,
-      "learning_rate": 7.274576271186441e-06,
-      "loss": 1.3651,
-      "step": 19270
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 6.180150985717773,
-      "learning_rate": 7.267796610169492e-06,
-      "loss": 1.3021,
-      "step": 19280
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.6004438400268555,
-      "learning_rate": 7.261016949152543e-06,
-      "loss": 1.2091,
-      "step": 19290
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 1.1322782039642334,
-      "learning_rate": 7.2542372881355936e-06,
-      "loss": 1.1944,
-      "step": 19300
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.901897430419922,
-      "learning_rate": 7.247457627118645e-06,
-      "loss": 1.2975,
-      "step": 19310
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.5151190757751465,
-      "learning_rate": 7.2406779661016955e-06,
-      "loss": 1.2506,
-      "step": 19320
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 3.7965264320373535,
-      "learning_rate": 7.233898305084747e-06,
-      "loss": 1.2635,
-      "step": 19330
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 2.9827957153320312,
-      "learning_rate": 7.2271186440677975e-06,
-      "loss": 1.3371,
-      "step": 19340
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.417064666748047,
-      "learning_rate": 7.220338983050849e-06,
-      "loss": 1.2549,
-      "step": 19350
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 6.065445423126221,
-      "learning_rate": 7.2135593220338986e-06,
-      "loss": 1.1856,
-      "step": 19360
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.525112628936768,
-      "learning_rate": 7.206779661016949e-06,
-      "loss": 1.0842,
-      "step": 19370
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 9.134572982788086,
-      "learning_rate": 7.2000000000000005e-06,
-      "loss": 1.1286,
-      "step": 19380
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 1.2406967878341675,
-      "learning_rate": 7.193220338983051e-06,
-      "loss": 1.0884,
-      "step": 19390
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 3.5600786209106445,
-      "learning_rate": 7.1864406779661025e-06,
-      "loss": 1.299,
-      "step": 19400
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 7.953360557556152,
-      "learning_rate": 7.179661016949153e-06,
-      "loss": 1.3167,
-      "step": 19410
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 6.8465986251831055,
-      "learning_rate": 7.1728813559322044e-06,
-      "loss": 1.2951,
-      "step": 19420
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.177648544311523,
-      "learning_rate": 7.166101694915255e-06,
-      "loss": 1.3032,
-      "step": 19430
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 6.590419292449951,
-      "learning_rate": 7.159322033898306e-06,
-      "loss": 1.3326,
-      "step": 19440
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 10.273443222045898,
-      "learning_rate": 7.152542372881357e-06,
-      "loss": 1.1746,
-      "step": 19450
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.114923000335693,
-      "learning_rate": 7.145762711864407e-06,
-      "loss": 1.5382,
-      "step": 19460
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 6.737642765045166,
-      "learning_rate": 7.138983050847458e-06,
-      "loss": 1.3364,
-      "step": 19470
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.186859607696533,
-      "learning_rate": 7.132203389830509e-06,
-      "loss": 1.2651,
-      "step": 19480
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.74080753326416,
-      "learning_rate": 7.12542372881356e-06,
-      "loss": 1.2772,
-      "step": 19490
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.5528883934021,
-      "learning_rate": 7.1186440677966106e-06,
-      "loss": 1.2557,
-      "step": 19500
-    },
-    {
-      "epoch": 0.45,
-      "eval_loss": 1.0282971858978271,
-      "eval_runtime": 67.939,
-      "eval_samples_per_second": 14.719,
-      "eval_steps_per_second": 14.719,
-      "step": 19500
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 3.818643093109131,
-      "learning_rate": 7.111864406779662e-06,
-      "loss": 1.3204,
-      "step": 19510
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.886016845703125,
-      "learning_rate": 7.1050847457627125e-06,
-      "loss": 1.4398,
-      "step": 19520
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 1.2431572675704956,
-      "learning_rate": 7.098305084745764e-06,
-      "loss": 1.1902,
-      "step": 19530
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 11.248309135437012,
-      "learning_rate": 7.0915254237288145e-06,
-      "loss": 1.2207,
-      "step": 19540
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 11.708429336547852,
-      "learning_rate": 7.084745762711865e-06,
-      "loss": 1.3756,
-      "step": 19550
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 2.923466920852661,
-      "learning_rate": 7.077966101694916e-06,
-      "loss": 1.4096,
-      "step": 19560
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 8.958571434020996,
-      "learning_rate": 7.071186440677966e-06,
-      "loss": 1.2582,
-      "step": 19570
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 6.289533615112305,
-      "learning_rate": 7.0644067796610175e-06,
-      "loss": 1.1581,
-      "step": 19580
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 2.9920499324798584,
-      "learning_rate": 7.057627118644068e-06,
-      "loss": 1.2995,
-      "step": 19590
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.625074863433838,
-      "learning_rate": 7.0508474576271195e-06,
-      "loss": 1.211,
-      "step": 19600
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 15.013193130493164,
-      "learning_rate": 7.04406779661017e-06,
-      "loss": 1.302,
-      "step": 19610
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 4.741778373718262,
-      "learning_rate": 7.037288135593221e-06,
-      "loss": 1.3025,
-      "step": 19620
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 9.325716972351074,
-      "learning_rate": 7.030508474576272e-06,
-      "loss": 1.2179,
-      "step": 19630
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.192024230957031,
-      "learning_rate": 7.0237288135593225e-06,
-      "loss": 1.1417,
-      "step": 19640
-    },
-    {
-      "epoch": 0.45,
-      "grad_norm": 5.244385719299316,
-      "learning_rate": 7.016949152542374e-06,
-      "loss": 1.3456,
-      "step": 19650
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 3.3191256523132324,
-      "learning_rate": 7.0101694915254245e-06,
-      "loss": 1.307,
-      "step": 19660
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 6.4485039710998535,
-      "learning_rate": 7.003389830508475e-06,
-      "loss": 1.238,
-      "step": 19670
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 8.261616706848145,
-      "learning_rate": 6.996610169491526e-06,
-      "loss": 1.1738,
-      "step": 19680
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 6.185995101928711,
-      "learning_rate": 6.989830508474576e-06,
-      "loss": 1.3407,
-      "step": 19690
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 9.428878784179688,
-      "learning_rate": 6.9830508474576275e-06,
-      "loss": 1.2947,
-      "step": 19700
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 1.5929450988769531,
-      "learning_rate": 6.976271186440678e-06,
-      "loss": 1.025,
-      "step": 19710
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 4.876119136810303,
-      "learning_rate": 6.9694915254237295e-06,
-      "loss": 1.4673,
-      "step": 19720
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 2.259251356124878,
-      "learning_rate": 6.96271186440678e-06,
-      "loss": 1.5241,
-      "step": 19730
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 4.431079864501953,
-      "learning_rate": 6.9559322033898315e-06,
-      "loss": 1.0948,
-      "step": 19740
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 5.089422225952148,
-      "learning_rate": 6.949152542372882e-06,
-      "loss": 1.2419,
-      "step": 19750
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 3.312706708908081,
-      "learning_rate": 6.942372881355933e-06,
-      "loss": 1.3715,
-      "step": 19760
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 9.536430358886719,
-      "learning_rate": 6.935593220338983e-06,
-      "loss": 1.3488,
-      "step": 19770
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 3.4993770122528076,
-      "learning_rate": 6.928813559322034e-06,
-      "loss": 1.2832,
-      "step": 19780
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 3.5873186588287354,
-      "learning_rate": 6.922033898305085e-06,
-      "loss": 1.2135,
-      "step": 19790
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 15.397970199584961,
-      "learning_rate": 6.915254237288136e-06,
-      "loss": 1.3824,
-      "step": 19800
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 5.0459794998168945,
-      "learning_rate": 6.908474576271187e-06,
-      "loss": 1.298,
-      "step": 19810
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 2.630267858505249,
-      "learning_rate": 6.901694915254238e-06,
-      "loss": 1.3083,
-      "step": 19820
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 7.766806602478027,
-      "learning_rate": 6.894915254237289e-06,
-      "loss": 1.2522,
-      "step": 19830
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 2.8050918579101562,
-      "learning_rate": 6.8881355932203395e-06,
-      "loss": 1.2368,
-      "step": 19840
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 2.825176239013672,
-      "learning_rate": 6.881355932203391e-06,
-      "loss": 1.2529,
-      "step": 19850
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 2.7948811054229736,
-      "learning_rate": 6.8745762711864415e-06,
-      "loss": 1.1087,
-      "step": 19860
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 5.275668144226074,
-      "learning_rate": 6.867796610169493e-06,
-      "loss": 1.3271,
-      "step": 19870
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 4.819080352783203,
-      "learning_rate": 6.861016949152543e-06,
-      "loss": 1.2839,
-      "step": 19880
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 6.651522159576416,
-      "learning_rate": 6.854237288135593e-06,
-      "loss": 1.1781,
-      "step": 19890
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 4.455416202545166,
-      "learning_rate": 6.8474576271186445e-06,
-      "loss": 1.3672,
-      "step": 19900
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 5.660415172576904,
-      "learning_rate": 6.840677966101695e-06,
-      "loss": 1.1597,
-      "step": 19910
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 4.537483215332031,
-      "learning_rate": 6.8338983050847465e-06,
-      "loss": 1.0678,
-      "step": 19920
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 8.044699668884277,
-      "learning_rate": 6.827118644067797e-06,
-      "loss": 1.1595,
-      "step": 19930
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 6.171726226806641,
-      "learning_rate": 6.8203389830508485e-06,
-      "loss": 1.3844,
-      "step": 19940
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 7.469864368438721,
-      "learning_rate": 6.813559322033899e-06,
-      "loss": 1.3641,
-      "step": 19950
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 1.8413645029067993,
-      "learning_rate": 6.80677966101695e-06,
-      "loss": 1.3229,
-      "step": 19960
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 7.607746124267578,
-      "learning_rate": 6.800000000000001e-06,
-      "loss": 1.3169,
-      "step": 19970
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 6.271143436431885,
-      "learning_rate": 6.793220338983051e-06,
-      "loss": 1.1564,
-      "step": 19980
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 6.224188804626465,
-      "learning_rate": 6.786440677966102e-06,
-      "loss": 1.426,
-      "step": 19990
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 5.205558776855469,
-      "learning_rate": 6.779661016949153e-06,
-      "loss": 1.0772,
-      "step": 20000
-    },
-    {
-      "epoch": 0.46,
-      "eval_loss": 0.9983803629875183,
-      "eval_runtime": 67.8281,
-      "eval_samples_per_second": 14.743,
-      "eval_steps_per_second": 14.743,
-      "step": 20000
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 2.1063528060913086,
-      "learning_rate": 6.772881355932204e-06,
-      "loss": 1.2372,
-      "step": 20010
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 8.878252983093262,
-      "learning_rate": 6.766101694915255e-06,
-      "loss": 1.3002,
-      "step": 20020
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 4.35091495513916,
-      "learning_rate": 6.759322033898306e-06,
-      "loss": 1.2132,
-      "step": 20030
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 3.8765366077423096,
-      "learning_rate": 6.7525423728813565e-06,
-      "loss": 1.2695,
-      "step": 20040
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 3.4200284481048584,
-      "learning_rate": 6.745762711864408e-06,
-      "loss": 1.4985,
-      "step": 20050
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 4.551451683044434,
-      "learning_rate": 6.7389830508474585e-06,
-      "loss": 1.303,
-      "step": 20060
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 5.922582149505615,
-      "learning_rate": 6.73220338983051e-06,
-      "loss": 1.2245,
-      "step": 20070
-    },
-    {
-      "epoch": 0.46,
-      "grad_norm": 7.12042760848999,
-      "learning_rate": 6.7254237288135604e-06,
-      "loss": 1.1991,
-      "step": 20080
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 3.4062719345092773,
-      "learning_rate": 6.71864406779661e-06,
-      "loss": 1.236,
-      "step": 20090
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 8.163664817810059,
-      "learning_rate": 6.7118644067796615e-06,
-      "loss": 1.3316,
-      "step": 20100
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 5.4666876792907715,
-      "learning_rate": 6.705084745762712e-06,
-      "loss": 1.2533,
-      "step": 20110
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 2.7013297080993652,
-      "learning_rate": 6.6983050847457635e-06,
-      "loss": 1.3016,
-      "step": 20120
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 5.204336166381836,
-      "learning_rate": 6.691525423728814e-06,
-      "loss": 1.2233,
-      "step": 20130
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 4.580149173736572,
-      "learning_rate": 6.6847457627118655e-06,
-      "loss": 1.1783,
-      "step": 20140
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 6.327539920806885,
-      "learning_rate": 6.677966101694916e-06,
-      "loss": 1.1495,
-      "step": 20150
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 12.739646911621094,
-      "learning_rate": 6.6711864406779666e-06,
-      "loss": 1.4032,
-      "step": 20160
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 2.9852492809295654,
-      "learning_rate": 6.664406779661018e-06,
-      "loss": 1.4179,
-      "step": 20170
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 5.05040979385376,
-      "learning_rate": 6.6576271186440685e-06,
-      "loss": 1.2684,
-      "step": 20180
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 7.304352760314941,
-      "learning_rate": 6.650847457627119e-06,
-      "loss": 1.3162,
-      "step": 20190
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 1.9470860958099365,
-      "learning_rate": 6.64406779661017e-06,
-      "loss": 1.3961,
-      "step": 20200
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 22.234329223632812,
-      "learning_rate": 6.637288135593221e-06,
-      "loss": 1.4659,
-      "step": 20210
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 5.601120948791504,
-      "learning_rate": 6.6305084745762716e-06,
-      "loss": 1.3222,
-      "step": 20220
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 4.113243579864502,
-      "learning_rate": 6.623728813559322e-06,
-      "loss": 1.2246,
-      "step": 20230
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 2.9469447135925293,
-      "learning_rate": 6.6169491525423735e-06,
-      "loss": 1.3428,
-      "step": 20240
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 3.2708497047424316,
-      "learning_rate": 6.610169491525424e-06,
-      "loss": 1.2096,
-      "step": 20250
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 2.1987311840057373,
-      "learning_rate": 6.6033898305084755e-06,
-      "loss": 1.2681,
-      "step": 20260
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 7.16361665725708,
-      "learning_rate": 6.596610169491526e-06,
-      "loss": 1.2586,
-      "step": 20270
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 4.537783622741699,
-      "learning_rate": 6.5898305084745774e-06,
-      "loss": 1.2015,
-      "step": 20280
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 6.448202610015869,
-      "learning_rate": 6.583050847457627e-06,
-      "loss": 1.1919,
-      "step": 20290
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 0.4741743206977844,
-      "learning_rate": 6.576271186440678e-06,
-      "loss": 1.4343,
-      "step": 20300
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 7.066291332244873,
-      "learning_rate": 6.569491525423729e-06,
-      "loss": 1.2109,
-      "step": 20310
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 3.167917490005493,
-      "learning_rate": 6.56271186440678e-06,
-      "loss": 1.2345,
-      "step": 20320
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 4.370973110198975,
-      "learning_rate": 6.555932203389831e-06,
-      "loss": 1.1774,
-      "step": 20330
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 3.6259400844573975,
-      "learning_rate": 6.549152542372882e-06,
-      "loss": 1.338,
-      "step": 20340
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 2.601332664489746,
-      "learning_rate": 6.542372881355933e-06,
-      "loss": 1.3791,
-      "step": 20350
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 9.787617683410645,
-      "learning_rate": 6.5355932203389836e-06,
-      "loss": 1.2684,
-      "step": 20360
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 1.965387225151062,
-      "learning_rate": 6.528813559322035e-06,
-      "loss": 1.2012,
-      "step": 20370
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 7.392675399780273,
-      "learning_rate": 6.5220338983050855e-06,
-      "loss": 1.0166,
-      "step": 20380
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 6.649271011352539,
-      "learning_rate": 6.515254237288137e-06,
-      "loss": 1.1422,
-      "step": 20390
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 6.833590984344482,
-      "learning_rate": 6.508474576271187e-06,
-      "loss": 1.3592,
-      "step": 20400
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 1.1567243337631226,
-      "learning_rate": 6.501694915254237e-06,
-      "loss": 1.4019,
-      "step": 20410
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 1.9526944160461426,
-      "learning_rate": 6.4949152542372886e-06,
-      "loss": 1.3748,
-      "step": 20420
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 9.20057201385498,
-      "learning_rate": 6.488135593220339e-06,
-      "loss": 1.1884,
-      "step": 20430
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 5.133617401123047,
-      "learning_rate": 6.4813559322033905e-06,
-      "loss": 1.2444,
-      "step": 20440
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 4.195832252502441,
-      "learning_rate": 6.474576271186441e-06,
-      "loss": 1.3187,
-      "step": 20450
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 8.69914722442627,
-      "learning_rate": 6.4677966101694925e-06,
-      "loss": 1.2932,
-      "step": 20460
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 3.7510390281677246,
-      "learning_rate": 6.461016949152543e-06,
-      "loss": 1.1252,
-      "step": 20470
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 5.250388145446777,
-      "learning_rate": 6.4542372881355944e-06,
-      "loss": 1.2828,
-      "step": 20480
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 8.589373588562012,
-      "learning_rate": 6.447457627118645e-06,
-      "loss": 1.2028,
-      "step": 20490
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 7.695883274078369,
-      "learning_rate": 6.440677966101695e-06,
-      "loss": 1.3743,
-      "step": 20500
-    },
-    {
-      "epoch": 0.47,
-      "eval_loss": 1.0409547090530396,
-      "eval_runtime": 67.4208,
-      "eval_samples_per_second": 14.832,
-      "eval_steps_per_second": 14.832,
-      "step": 20500
-    },
-    {
-      "epoch": 0.47,
-      "grad_norm": 10.85682487487793,
-      "learning_rate": 6.433898305084746e-06,
-      "loss": 1.4262,
-      "step": 20510
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 6.717245101928711,
-      "learning_rate": 6.427118644067797e-06,
-      "loss": 1.3546,
-      "step": 20520
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 9.486088752746582,
-      "learning_rate": 6.420338983050848e-06,
-      "loss": 1.2189,
-      "step": 20530
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 4.002749919891357,
-      "learning_rate": 6.413559322033899e-06,
-      "loss": 1.3337,
-      "step": 20540
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 8.079071044921875,
-      "learning_rate": 6.40677966101695e-06,
-      "loss": 1.3489,
-      "step": 20550
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 6.364570617675781,
-      "learning_rate": 6.4000000000000006e-06,
-      "loss": 1.0267,
-      "step": 20560
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 4.548829555511475,
-      "learning_rate": 6.393220338983052e-06,
-      "loss": 1.3025,
-      "step": 20570
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 8.967122077941895,
-      "learning_rate": 6.3864406779661025e-06,
-      "loss": 1.2183,
-      "step": 20580
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 8.529352188110352,
-      "learning_rate": 6.379661016949154e-06,
-      "loss": 1.2688,
-      "step": 20590
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 7.445988178253174,
-      "learning_rate": 6.372881355932204e-06,
-      "loss": 1.3147,
-      "step": 20600
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 8.145844459533691,
-      "learning_rate": 6.366101694915254e-06,
-      "loss": 1.3358,
-      "step": 20610
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 3.0077157020568848,
-      "learning_rate": 6.3593220338983056e-06,
-      "loss": 1.4507,
-      "step": 20620
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 7.429611682891846,
-      "learning_rate": 6.352542372881356e-06,
-      "loss": 1.0919,
-      "step": 20630
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 6.513698101043701,
-      "learning_rate": 6.3457627118644075e-06,
-      "loss": 1.392,
-      "step": 20640
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 10.298941612243652,
-      "learning_rate": 6.338983050847458e-06,
-      "loss": 1.1327,
-      "step": 20650
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 3.2288424968719482,
-      "learning_rate": 6.3322033898305095e-06,
-      "loss": 1.1693,
-      "step": 20660
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 4.072999000549316,
-      "learning_rate": 6.32542372881356e-06,
-      "loss": 1.1615,
-      "step": 20670
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 2.6640424728393555,
-      "learning_rate": 6.318644067796611e-06,
-      "loss": 1.1804,
-      "step": 20680
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 7.22818660736084,
-      "learning_rate": 6.311864406779662e-06,
-      "loss": 1.3203,
-      "step": 20690
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 3.37728214263916,
-      "learning_rate": 6.3050847457627125e-06,
-      "loss": 1.1533,
-      "step": 20700
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 2.2169246673583984,
-      "learning_rate": 6.298305084745763e-06,
-      "loss": 1.3152,
-      "step": 20710
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 8.210066795349121,
-      "learning_rate": 6.291525423728814e-06,
-      "loss": 1.292,
-      "step": 20720
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 8.596230506896973,
-      "learning_rate": 6.284745762711865e-06,
-      "loss": 1.2931,
-      "step": 20730
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 4.568326950073242,
-      "learning_rate": 6.277966101694916e-06,
-      "loss": 1.3412,
-      "step": 20740
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 7.391994953155518,
-      "learning_rate": 6.271186440677966e-06,
-      "loss": 1.1354,
-      "step": 20750
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 5.019871711730957,
-      "learning_rate": 6.2644067796610176e-06,
-      "loss": 1.3048,
-      "step": 20760
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 2.394231081008911,
-      "learning_rate": 6.257627118644068e-06,
-      "loss": 1.372,
-      "step": 20770
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 13.933137893676758,
-      "learning_rate": 6.2508474576271195e-06,
-      "loss": 1.2364,
-      "step": 20780
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 0.63595050573349,
-      "learning_rate": 6.24406779661017e-06,
-      "loss": 1.3786,
-      "step": 20790
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 4.854935646057129,
-      "learning_rate": 6.2372881355932215e-06,
-      "loss": 1.2068,
-      "step": 20800
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 8.855067253112793,
-      "learning_rate": 6.230508474576271e-06,
-      "loss": 1.1401,
-      "step": 20810
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 2.5127904415130615,
-      "learning_rate": 6.223728813559322e-06,
-      "loss": 1.3908,
-      "step": 20820
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 4.830283164978027,
-      "learning_rate": 6.216949152542373e-06,
-      "loss": 1.2906,
-      "step": 20830
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 5.127600193023682,
-      "learning_rate": 6.210169491525424e-06,
-      "loss": 1.2737,
-      "step": 20840
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 2.2829642295837402,
-      "learning_rate": 6.203389830508475e-06,
-      "loss": 1.1188,
-      "step": 20850
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 3.748267889022827,
-      "learning_rate": 6.196610169491526e-06,
-      "loss": 1.2071,
-      "step": 20860
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 5.889699935913086,
-      "learning_rate": 6.189830508474577e-06,
-      "loss": 1.2175,
-      "step": 20870
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 2.2203211784362793,
-      "learning_rate": 6.183050847457628e-06,
-      "loss": 1.4178,
-      "step": 20880
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 9.717460632324219,
-      "learning_rate": 6.176271186440679e-06,
-      "loss": 1.2213,
-      "step": 20890
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 4.968123912811279,
-      "learning_rate": 6.1694915254237295e-06,
-      "loss": 1.2122,
-      "step": 20900
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 11.27860164642334,
-      "learning_rate": 6.162711864406781e-06,
-      "loss": 1.2388,
-      "step": 20910
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 7.691869735717773,
-      "learning_rate": 6.155932203389831e-06,
-      "loss": 1.3148,
-      "step": 20920
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 3.1556854248046875,
-      "learning_rate": 6.149152542372881e-06,
-      "loss": 1.0672,
-      "step": 20930
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 11.499942779541016,
-      "learning_rate": 6.142372881355933e-06,
-      "loss": 1.2098,
-      "step": 20940
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.9171266555786133,
-      "learning_rate": 6.135593220338983e-06,
-      "loss": 1.2329,
-      "step": 20950
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 12.616515159606934,
-      "learning_rate": 6.1288135593220346e-06,
-      "loss": 1.1922,
-      "step": 20960
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 4.825931072235107,
-      "learning_rate": 6.122033898305085e-06,
-      "loss": 1.2403,
-      "step": 20970
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 9.974098205566406,
-      "learning_rate": 6.1152542372881365e-06,
-      "loss": 1.4907,
-      "step": 20980
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 3.1674904823303223,
-      "learning_rate": 6.108474576271187e-06,
-      "loss": 1.1076,
-      "step": 20990
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 5.067592144012451,
-      "learning_rate": 6.1016949152542385e-06,
-      "loss": 1.3289,
-      "step": 21000
-    },
-    {
-      "epoch": 0.49,
-      "eval_loss": 1.0621670484542847,
-      "eval_runtime": 67.2686,
-      "eval_samples_per_second": 14.866,
-      "eval_steps_per_second": 14.866,
-      "step": 21000
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.247994899749756,
-      "learning_rate": 6.094915254237289e-06,
-      "loss": 0.9261,
-      "step": 21010
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 8.553903579711914,
-      "learning_rate": 6.088135593220339e-06,
-      "loss": 1.3122,
-      "step": 21020
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 3.4144701957702637,
-      "learning_rate": 6.08135593220339e-06,
-      "loss": 1.5104,
-      "step": 21030
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 3.242537021636963,
-      "learning_rate": 6.074576271186441e-06,
-      "loss": 1.3842,
-      "step": 21040
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.6115124225616455,
-      "learning_rate": 6.067796610169492e-06,
-      "loss": 1.4534,
-      "step": 21050
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 3.6139254570007324,
-      "learning_rate": 6.061016949152543e-06,
-      "loss": 1.1461,
-      "step": 21060
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 8.1109619140625,
-      "learning_rate": 6.054237288135594e-06,
-      "loss": 1.2601,
-      "step": 21070
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.504269123077393,
-      "learning_rate": 6.047457627118645e-06,
-      "loss": 0.9487,
-      "step": 21080
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 3.8124566078186035,
-      "learning_rate": 6.040677966101696e-06,
-      "loss": 1.1965,
-      "step": 21090
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 6.297351837158203,
-      "learning_rate": 6.0338983050847465e-06,
-      "loss": 1.4083,
-      "step": 21100
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.318817138671875,
-      "learning_rate": 6.027118644067798e-06,
-      "loss": 1.3782,
-      "step": 21110
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 3.4005706310272217,
-      "learning_rate": 6.020338983050848e-06,
-      "loss": 1.3072,
-      "step": 21120
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 6.140954971313477,
-      "learning_rate": 6.013559322033898e-06,
-      "loss": 1.3371,
-      "step": 21130
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 9.956124305725098,
-      "learning_rate": 6.00677966101695e-06,
-      "loss": 1.3825,
-      "step": 21140
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 5.303738117218018,
-      "learning_rate": 6e-06,
-      "loss": 1.2292,
-      "step": 21150
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.743120193481445,
-      "learning_rate": 5.9932203389830516e-06,
-      "loss": 1.1453,
-      "step": 21160
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 9.754213333129883,
-      "learning_rate": 5.986440677966102e-06,
-      "loss": 1.2849,
-      "step": 21170
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 4.077347755432129,
-      "learning_rate": 5.9796610169491535e-06,
-      "loss": 1.3983,
-      "step": 21180
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.309597969055176,
-      "learning_rate": 5.972881355932204e-06,
-      "loss": 1.3914,
-      "step": 21190
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 8.950671195983887,
-      "learning_rate": 5.9661016949152555e-06,
-      "loss": 1.3262,
-      "step": 21200
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 4.134212493896484,
-      "learning_rate": 5.959322033898306e-06,
-      "loss": 1.2166,
-      "step": 21210
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.0232081413269043,
-      "learning_rate": 5.9525423728813566e-06,
-      "loss": 1.2959,
-      "step": 21220
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.094144582748413,
-      "learning_rate": 5.945762711864407e-06,
-      "loss": 1.2989,
-      "step": 21230
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.6865925788879395,
-      "learning_rate": 5.938983050847458e-06,
-      "loss": 1.3427,
-      "step": 21240
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.456575870513916,
-      "learning_rate": 5.932203389830509e-06,
-      "loss": 1.1175,
-      "step": 21250
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 5.038335800170898,
-      "learning_rate": 5.92542372881356e-06,
-      "loss": 1.1995,
-      "step": 21260
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 6.489980697631836,
-      "learning_rate": 5.91864406779661e-06,
-      "loss": 1.2752,
-      "step": 21270
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 4.00714111328125,
-      "learning_rate": 5.911864406779662e-06,
-      "loss": 1.3853,
-      "step": 21280
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.6164281368255615,
-      "learning_rate": 5.905084745762712e-06,
-      "loss": 1.0804,
-      "step": 21290
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 3.9665138721466064,
-      "learning_rate": 5.8983050847457635e-06,
-      "loss": 1.1759,
-      "step": 21300
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.67433500289917,
-      "learning_rate": 5.891525423728814e-06,
-      "loss": 1.393,
-      "step": 21310
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.603160381317139,
-      "learning_rate": 5.8847457627118655e-06,
-      "loss": 1.5365,
-      "step": 21320
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.298281192779541,
-      "learning_rate": 5.877966101694915e-06,
-      "loss": 1.3866,
-      "step": 21330
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 2.494446039199829,
-      "learning_rate": 5.871186440677966e-06,
-      "loss": 1.2081,
-      "step": 21340
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 8.670098304748535,
-      "learning_rate": 5.864406779661017e-06,
-      "loss": 1.2346,
-      "step": 21350
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 8.977124214172363,
-      "learning_rate": 5.857627118644068e-06,
-      "loss": 1.0865,
-      "step": 21360
-    },
-    {
-      "epoch": 0.49,
-      "grad_norm": 7.3538384437561035,
-      "learning_rate": 5.850847457627119e-06,
-      "loss": 1.0902,
-      "step": 21370
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 5.1297783851623535,
-      "learning_rate": 5.84406779661017e-06,
-      "loss": 1.532,
-      "step": 21380
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 9.002881050109863,
-      "learning_rate": 5.837288135593221e-06,
-      "loss": 1.2216,
-      "step": 21390
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 7.277677536010742,
-      "learning_rate": 5.830508474576272e-06,
-      "loss": 1.1974,
-      "step": 21400
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 3.383936643600464,
-      "learning_rate": 5.823728813559323e-06,
-      "loss": 1.2097,
-      "step": 21410
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 9.847234725952148,
-      "learning_rate": 5.8169491525423736e-06,
-      "loss": 1.225,
-      "step": 21420
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 9.401296615600586,
-      "learning_rate": 5.810169491525425e-06,
-      "loss": 1.4399,
-      "step": 21430
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 6.085318088531494,
-      "learning_rate": 5.803389830508475e-06,
-      "loss": 1.2921,
-      "step": 21440
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 1.833952784538269,
-      "learning_rate": 5.796610169491525e-06,
-      "loss": 1.3274,
-      "step": 21450
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.7563090324401855,
-      "learning_rate": 5.789830508474577e-06,
-      "loss": 1.2354,
-      "step": 21460
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 2.760796546936035,
-      "learning_rate": 5.783050847457627e-06,
-      "loss": 1.2845,
-      "step": 21470
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 8.02619743347168,
-      "learning_rate": 5.776271186440679e-06,
-      "loss": 1.3089,
-      "step": 21480
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 7.648224830627441,
-      "learning_rate": 5.769491525423729e-06,
-      "loss": 1.3202,
-      "step": 21490
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 3.990171194076538,
-      "learning_rate": 5.7627118644067805e-06,
-      "loss": 1.3624,
-      "step": 21500
-    },
-    {
-      "epoch": 0.5,
-      "eval_loss": 1.033817172050476,
-      "eval_runtime": 68.2554,
-      "eval_samples_per_second": 14.651,
-      "eval_steps_per_second": 14.651,
-      "step": 21500
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 5.392577171325684,
-      "learning_rate": 5.755932203389831e-06,
-      "loss": 1.1602,
-      "step": 21510
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 2.582357883453369,
-      "learning_rate": 5.7491525423728825e-06,
-      "loss": 1.0574,
-      "step": 21520
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.663832187652588,
-      "learning_rate": 5.742372881355933e-06,
-      "loss": 1.1496,
-      "step": 21530
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 8.770710945129395,
-      "learning_rate": 5.735593220338983e-06,
-      "loss": 1.2921,
-      "step": 21540
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 9.296341896057129,
-      "learning_rate": 5.728813559322034e-06,
-      "loss": 1.284,
-      "step": 21550
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 9.320971488952637,
-      "learning_rate": 5.722033898305085e-06,
-      "loss": 1.3229,
-      "step": 21560
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 5.163783073425293,
-      "learning_rate": 5.715254237288136e-06,
-      "loss": 1.3256,
-      "step": 21570
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 7.684600830078125,
-      "learning_rate": 5.708474576271187e-06,
-      "loss": 1.2054,
-      "step": 21580
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 11.25024700164795,
-      "learning_rate": 5.701694915254238e-06,
-      "loss": 1.3583,
-      "step": 21590
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 1.7152315378189087,
-      "learning_rate": 5.694915254237289e-06,
-      "loss": 1.2261,
-      "step": 21600
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.451650619506836,
-      "learning_rate": 5.68813559322034e-06,
-      "loss": 1.2716,
-      "step": 21610
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 1.9379420280456543,
-      "learning_rate": 5.6813559322033906e-06,
-      "loss": 1.2794,
-      "step": 21620
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 2.325803756713867,
-      "learning_rate": 5.674576271186442e-06,
-      "loss": 1.3336,
-      "step": 21630
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 5.5600457191467285,
-      "learning_rate": 5.667796610169492e-06,
-      "loss": 1.2249,
-      "step": 21640
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 6.445628643035889,
-      "learning_rate": 5.661016949152542e-06,
-      "loss": 1.1567,
-      "step": 21650
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.747307300567627,
-      "learning_rate": 5.654237288135594e-06,
-      "loss": 1.1082,
-      "step": 21660
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.780632495880127,
-      "learning_rate": 5.647457627118644e-06,
-      "loss": 1.2751,
-      "step": 21670
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 6.126337051391602,
-      "learning_rate": 5.640677966101696e-06,
-      "loss": 1.2164,
-      "step": 21680
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 7.235857963562012,
-      "learning_rate": 5.633898305084746e-06,
-      "loss": 1.5172,
-      "step": 21690
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 5.676464080810547,
-      "learning_rate": 5.6271186440677975e-06,
-      "loss": 1.1409,
-      "step": 21700
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.40902853012085,
-      "learning_rate": 5.620338983050848e-06,
-      "loss": 1.2341,
-      "step": 21710
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 2.4780306816101074,
-      "learning_rate": 5.6135593220338995e-06,
-      "loss": 1.2982,
-      "step": 21720
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.747418403625488,
-      "learning_rate": 5.60677966101695e-06,
-      "loss": 1.1433,
-      "step": 21730
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 1.319555401802063,
-      "learning_rate": 5.600000000000001e-06,
-      "loss": 1.1173,
-      "step": 21740
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 6.180632591247559,
-      "learning_rate": 5.593220338983051e-06,
-      "loss": 1.3304,
-      "step": 21750
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 17.385282516479492,
-      "learning_rate": 5.586440677966102e-06,
-      "loss": 1.3295,
-      "step": 21760
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 1.6295877695083618,
-      "learning_rate": 5.579661016949153e-06,
-      "loss": 1.329,
-      "step": 21770
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 7.083805561065674,
-      "learning_rate": 5.572881355932204e-06,
-      "loss": 1.337,
-      "step": 21780
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 9.662304878234863,
-      "learning_rate": 5.566101694915255e-06,
-      "loss": 1.2301,
-      "step": 21790
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 7.854396820068359,
-      "learning_rate": 5.559322033898306e-06,
-      "loss": 1.3312,
-      "step": 21800
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.626899242401123,
-      "learning_rate": 5.552542372881356e-06,
-      "loss": 1.1979,
-      "step": 21810
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 7.825695037841797,
-      "learning_rate": 5.5457627118644076e-06,
-      "loss": 1.2997,
-      "step": 21820
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 4.459179401397705,
-      "learning_rate": 5.538983050847458e-06,
-      "loss": 1.4772,
-      "step": 21830
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.008397579193115,
-      "learning_rate": 5.5322033898305095e-06,
-      "loss": 1.2599,
-      "step": 21840
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.558854103088379,
-      "learning_rate": 5.525423728813559e-06,
-      "loss": 1.3258,
-      "step": 21850
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 2.846191883087158,
-      "learning_rate": 5.518644067796611e-06,
-      "loss": 1.344,
-      "step": 21860
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.7226576805114746,
-      "learning_rate": 5.511864406779661e-06,
-      "loss": 1.1642,
-      "step": 21870
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 8.011185646057129,
-      "learning_rate": 5.505084745762712e-06,
-      "loss": 1.1073,
-      "step": 21880
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 8.683090209960938,
-      "learning_rate": 5.498305084745763e-06,
-      "loss": 1.2852,
-      "step": 21890
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.06953763961792,
-      "learning_rate": 5.491525423728814e-06,
-      "loss": 1.2416,
-      "step": 21900
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.721527576446533,
-      "learning_rate": 5.484745762711865e-06,
-      "loss": 1.3548,
-      "step": 21910
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 4.444101810455322,
-      "learning_rate": 5.477966101694916e-06,
-      "loss": 1.3911,
-      "step": 21920
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 9.347046852111816,
-      "learning_rate": 5.471186440677967e-06,
-      "loss": 1.237,
-      "step": 21930
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.472478866577148,
-      "learning_rate": 5.464406779661018e-06,
-      "loss": 1.1721,
-      "step": 21940
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.854987382888794,
-      "learning_rate": 5.457627118644067e-06,
-      "loss": 1.245,
-      "step": 21950
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 8.591747283935547,
-      "learning_rate": 5.450847457627119e-06,
-      "loss": 1.2754,
-      "step": 21960
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 2.2374930381774902,
-      "learning_rate": 5.444067796610169e-06,
-      "loss": 1.248,
-      "step": 21970
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.7026805877685547,
-      "learning_rate": 5.437288135593221e-06,
-      "loss": 1.2337,
-      "step": 21980
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.4988057613372803,
-      "learning_rate": 5.430508474576271e-06,
-      "loss": 1.2272,
-      "step": 21990
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.3364341259002686,
-      "learning_rate": 5.423728813559323e-06,
-      "loss": 1.2301,
-      "step": 22000
-    },
-    {
-      "epoch": 0.51,
-      "eval_loss": 0.9484620690345764,
-      "eval_runtime": 67.8436,
-      "eval_samples_per_second": 14.74,
-      "eval_steps_per_second": 14.74,
-      "step": 22000
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 8.392367362976074,
-      "learning_rate": 5.416949152542373e-06,
-      "loss": 1.3843,
-      "step": 22010
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 7.365686416625977,
-      "learning_rate": 5.4101694915254246e-06,
-      "loss": 1.2589,
-      "step": 22020
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 2.972006320953369,
-      "learning_rate": 5.403389830508475e-06,
-      "loss": 1.428,
-      "step": 22030
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.8066885471343994,
-      "learning_rate": 5.3966101694915265e-06,
-      "loss": 1.2202,
-      "step": 22040
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.83513879776001,
-      "learning_rate": 5.389830508474577e-06,
-      "loss": 1.3555,
-      "step": 22050
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 10.071037292480469,
-      "learning_rate": 5.383050847457627e-06,
-      "loss": 1.4515,
-      "step": 22060
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.605524063110352,
-      "learning_rate": 5.376271186440678e-06,
-      "loss": 1.3172,
-      "step": 22070
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 7.423837184906006,
-      "learning_rate": 5.369491525423729e-06,
-      "loss": 1.2454,
-      "step": 22080
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 7.91333532333374,
-      "learning_rate": 5.36271186440678e-06,
-      "loss": 0.9647,
-      "step": 22090
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.196317672729492,
-      "learning_rate": 5.355932203389831e-06,
-      "loss": 1.3829,
-      "step": 22100
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 1.629678726196289,
-      "learning_rate": 5.349152542372882e-06,
-      "loss": 1.306,
-      "step": 22110
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.215089797973633,
-      "learning_rate": 5.342372881355933e-06,
-      "loss": 1.3034,
-      "step": 22120
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.362987995147705,
-      "learning_rate": 5.335593220338984e-06,
-      "loss": 1.3381,
-      "step": 22130
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.5946292877197266,
-      "learning_rate": 5.328813559322035e-06,
-      "loss": 1.2814,
-      "step": 22140
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.043945789337158,
-      "learning_rate": 5.322033898305086e-06,
-      "loss": 1.1752,
-      "step": 22150
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 6.917520046234131,
-      "learning_rate": 5.315254237288136e-06,
-      "loss": 1.3277,
-      "step": 22160
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 10.332216262817383,
-      "learning_rate": 5.308474576271186e-06,
-      "loss": 1.2168,
-      "step": 22170
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 7.612697124481201,
-      "learning_rate": 5.301694915254238e-06,
-      "loss": 1.2865,
-      "step": 22180
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 7.4925360679626465,
-      "learning_rate": 5.294915254237288e-06,
-      "loss": 1.1815,
-      "step": 22190
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 1.8448129892349243,
-      "learning_rate": 5.28813559322034e-06,
-      "loss": 1.2843,
-      "step": 22200
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 5.239171981811523,
-      "learning_rate": 5.28135593220339e-06,
-      "loss": 1.2266,
-      "step": 22210
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.959794759750366,
-      "learning_rate": 5.2745762711864416e-06,
-      "loss": 1.1932,
-      "step": 22220
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 4.557924270629883,
-      "learning_rate": 5.267796610169492e-06,
-      "loss": 1.386,
-      "step": 22230
-    },
-    {
-      "epoch": 0.51,
-      "grad_norm": 3.9078621864318848,
-      "learning_rate": 5.2610169491525435e-06,
-      "loss": 1.2638,
-      "step": 22240
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 0.628909707069397,
-      "learning_rate": 5.254237288135594e-06,
-      "loss": 1.1575,
-      "step": 22250
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 7.980692386627197,
-      "learning_rate": 5.247457627118645e-06,
-      "loss": 1.2014,
-      "step": 22260
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 13.897843360900879,
-      "learning_rate": 5.240677966101695e-06,
-      "loss": 1.3383,
-      "step": 22270
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 7.637335300445557,
-      "learning_rate": 5.233898305084746e-06,
-      "loss": 1.1604,
-      "step": 22280
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 8.646268844604492,
-      "learning_rate": 5.227118644067797e-06,
-      "loss": 1.3669,
-      "step": 22290
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 10.057101249694824,
-      "learning_rate": 5.220338983050848e-06,
-      "loss": 1.1849,
-      "step": 22300
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.541333198547363,
-      "learning_rate": 5.213559322033899e-06,
-      "loss": 1.116,
-      "step": 22310
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 16.73844337463379,
-      "learning_rate": 5.20677966101695e-06,
-      "loss": 1.3564,
-      "step": 22320
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 13.738208770751953,
-      "learning_rate": 5.2e-06,
-      "loss": 1.365,
-      "step": 22330
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.143723011016846,
-      "learning_rate": 5.193220338983052e-06,
-      "loss": 1.2913,
-      "step": 22340
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 7.556609153747559,
-      "learning_rate": 5.186440677966102e-06,
-      "loss": 1.2513,
-      "step": 22350
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 5.468164920806885,
-      "learning_rate": 5.1796610169491535e-06,
-      "loss": 1.2641,
-      "step": 22360
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 6.362597942352295,
-      "learning_rate": 5.172881355932203e-06,
-      "loss": 1.4793,
-      "step": 22370
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 3.212599992752075,
-      "learning_rate": 5.166101694915255e-06,
-      "loss": 1.4001,
-      "step": 22380
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 7.666508197784424,
-      "learning_rate": 5.159322033898305e-06,
-      "loss": 1.3046,
-      "step": 22390
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 7.109092712402344,
-      "learning_rate": 5.152542372881356e-06,
-      "loss": 1.2948,
-      "step": 22400
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 6.648116588592529,
-      "learning_rate": 5.145762711864407e-06,
-      "loss": 1.1994,
-      "step": 22410
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.890117645263672,
-      "learning_rate": 5.138983050847458e-06,
-      "loss": 1.2836,
-      "step": 22420
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 10.657286643981934,
-      "learning_rate": 5.132203389830509e-06,
-      "loss": 1.0558,
-      "step": 22430
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 2.1854753494262695,
-      "learning_rate": 5.12542372881356e-06,
-      "loss": 1.3231,
-      "step": 22440
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 5.85990571975708,
-      "learning_rate": 5.118644067796611e-06,
-      "loss": 1.0908,
-      "step": 22450
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 6.751573085784912,
-      "learning_rate": 5.111864406779662e-06,
-      "loss": 1.1804,
-      "step": 22460
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 8.723662376403809,
-      "learning_rate": 5.105084745762711e-06,
-      "loss": 1.3736,
-      "step": 22470
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.84275484085083,
-      "learning_rate": 5.098305084745763e-06,
-      "loss": 1.3529,
-      "step": 22480
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 3.9603848457336426,
-      "learning_rate": 5.091525423728813e-06,
-      "loss": 1.1711,
-      "step": 22490
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 1.6602897644042969,
-      "learning_rate": 5.084745762711865e-06,
-      "loss": 1.2808,
-      "step": 22500
-    },
-    {
-      "epoch": 0.52,
-      "eval_loss": 1.0503261089324951,
-      "eval_runtime": 68.1322,
-      "eval_samples_per_second": 14.677,
-      "eval_steps_per_second": 14.677,
-      "step": 22500
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.92212438583374,
-      "learning_rate": 5.077966101694915e-06,
-      "loss": 1.0928,
-      "step": 22510
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 11.511126518249512,
-      "learning_rate": 5.071186440677967e-06,
-      "loss": 1.3047,
-      "step": 22520
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.2978596687316895,
-      "learning_rate": 5.064406779661017e-06,
-      "loss": 1.3928,
-      "step": 22530
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.02852201461792,
-      "learning_rate": 5.057627118644069e-06,
-      "loss": 1.3077,
-      "step": 22540
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.235876560211182,
-      "learning_rate": 5.050847457627119e-06,
-      "loss": 1.1435,
-      "step": 22550
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 3.8855769634246826,
-      "learning_rate": 5.0440677966101705e-06,
-      "loss": 1.2596,
-      "step": 22560
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 10.052867889404297,
-      "learning_rate": 5.037288135593221e-06,
-      "loss": 1.1957,
-      "step": 22570
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 6.440583229064941,
-      "learning_rate": 5.030508474576271e-06,
-      "loss": 1.1874,
-      "step": 22580
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.625791072845459,
-      "learning_rate": 5.023728813559322e-06,
-      "loss": 1.4208,
-      "step": 22590
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 2.446409225463867,
-      "learning_rate": 5.016949152542373e-06,
-      "loss": 1.142,
-      "step": 22600
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 10.698454856872559,
-      "learning_rate": 5.010169491525424e-06,
-      "loss": 1.1436,
-      "step": 22610
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 1.8976472616195679,
-      "learning_rate": 5.003389830508475e-06,
-      "loss": 1.1139,
-      "step": 22620
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.049407958984375,
-      "learning_rate": 4.996610169491526e-06,
-      "loss": 1.2085,
-      "step": 22630
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 14.139172554016113,
-      "learning_rate": 4.989830508474577e-06,
-      "loss": 1.192,
-      "step": 22640
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 12.04554271697998,
-      "learning_rate": 4.983050847457628e-06,
-      "loss": 1.2503,
-      "step": 22650
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 4.248260498046875,
-      "learning_rate": 4.976271186440678e-06,
-      "loss": 0.9741,
-      "step": 22660
-    },
-    {
-      "epoch": 0.52,
-      "grad_norm": 8.9169282913208,
-      "learning_rate": 4.969491525423729e-06,
-      "loss": 1.4536,
-      "step": 22670
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 2.942535877227783,
-      "learning_rate": 4.96271186440678e-06,
-      "loss": 1.1215,
-      "step": 22680
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 3.7272868156433105,
-      "learning_rate": 4.955932203389831e-06,
-      "loss": 1.2022,
-      "step": 22690
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 6.489871501922607,
-      "learning_rate": 4.949152542372882e-06,
-      "loss": 1.1204,
-      "step": 22700
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 5.630454063415527,
-      "learning_rate": 4.942372881355932e-06,
-      "loss": 1.2873,
-      "step": 22710
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 2.8798165321350098,
-      "learning_rate": 4.935593220338984e-06,
-      "loss": 1.3039,
-      "step": 22720
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 5.791500568389893,
-      "learning_rate": 4.928813559322034e-06,
-      "loss": 1.5419,
-      "step": 22730
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 4.488035202026367,
-      "learning_rate": 4.922033898305086e-06,
-      "loss": 1.0108,
-      "step": 22740
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 2.9141738414764404,
-      "learning_rate": 4.915254237288136e-06,
-      "loss": 1.0819,
-      "step": 22750
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 7.7986626625061035,
-      "learning_rate": 4.908474576271187e-06,
-      "loss": 1.1406,
-      "step": 22760
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 5.593525409698486,
-      "learning_rate": 4.901694915254237e-06,
-      "loss": 1.1587,
-      "step": 22770
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 6.104484558105469,
-      "learning_rate": 4.894915254237289e-06,
-      "loss": 1.3049,
-      "step": 22780
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 10.20970630645752,
-      "learning_rate": 4.888135593220339e-06,
-      "loss": 1.1083,
-      "step": 22790
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 6.432941913604736,
-      "learning_rate": 4.881355932203391e-06,
-      "loss": 1.2579,
-      "step": 22800
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 5.444655895233154,
-      "learning_rate": 4.874576271186441e-06,
-      "loss": 1.1385,
-      "step": 22810
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 6.204164028167725,
-      "learning_rate": 4.867796610169492e-06,
-      "loss": 1.2969,
-      "step": 22820
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 2.851778984069824,
-      "learning_rate": 4.861016949152543e-06,
-      "loss": 1.3118,
-      "step": 22830
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 4.721585750579834,
-      "learning_rate": 4.854237288135594e-06,
-      "loss": 1.0812,
-      "step": 22840
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 7.1442646980285645,
-      "learning_rate": 4.847457627118645e-06,
-      "loss": 1.258,
-      "step": 22850
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 3.9416871070861816,
-      "learning_rate": 4.840677966101695e-06,
-      "loss": 1.0249,
-      "step": 22860
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 4.044296741485596,
-      "learning_rate": 4.833898305084746e-06,
-      "loss": 1.2126,
-      "step": 22870
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 2.5377376079559326,
-      "learning_rate": 4.827118644067797e-06,
-      "loss": 1.4611,
-      "step": 22880
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 3.37980055809021,
-      "learning_rate": 4.820338983050848e-06,
-      "loss": 1.2157,
-      "step": 22890
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 7.17388916015625,
-      "learning_rate": 4.813559322033899e-06,
-      "loss": 1.252,
-      "step": 22900
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 5.8747124671936035,
-      "learning_rate": 4.80677966101695e-06,
-      "loss": 1.3771,
-      "step": 22910
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 8.872361183166504,
-      "learning_rate": 4.800000000000001e-06,
-      "loss": 1.3137,
-      "step": 22920
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 6.978908061981201,
-      "learning_rate": 4.793220338983051e-06,
-      "loss": 1.1414,
-      "step": 22930
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 1.8234301805496216,
-      "learning_rate": 4.786440677966102e-06,
-      "loss": 1.2421,
-      "step": 22940
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 9.105183601379395,
-      "learning_rate": 4.779661016949153e-06,
-      "loss": 1.4391,
-      "step": 22950
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 27.357824325561523,
-      "learning_rate": 4.772881355932204e-06,
-      "loss": 1.1089,
-      "step": 22960
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 5.128312110900879,
-      "learning_rate": 4.766101694915254e-06,
-      "loss": 1.0676,
-      "step": 22970
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 4.661288261413574,
-      "learning_rate": 4.759322033898306e-06,
-      "loss": 1.2435,
-      "step": 22980
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 3.2815043926239014,
-      "learning_rate": 4.752542372881356e-06,
-      "loss": 1.1472,
-      "step": 22990
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 5.986001014709473,
-      "learning_rate": 4.745762711864408e-06,
-      "loss": 1.3882,
-      "step": 23000
-    },
-    {
-      "epoch": 0.53,
-      "eval_loss": 0.9571949243545532,
-      "eval_runtime": 68.0527,
-      "eval_samples_per_second": 14.694,
-      "eval_steps_per_second": 14.694,
-      "step": 23000
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 2.8159420490264893,
-      "learning_rate": 4.738983050847458e-06,
-      "loss": 1.193,
-      "step": 23010
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 4.794797897338867,
-      "learning_rate": 4.732203389830509e-06,
-      "loss": 1.3156,
-      "step": 23020
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 3.945427656173706,
-      "learning_rate": 4.725423728813559e-06,
-      "loss": 1.2774,
-      "step": 23030
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 9.099642753601074,
-      "learning_rate": 4.718644067796611e-06,
-      "loss": 1.2536,
-      "step": 23040
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 6.1287455558776855,
-      "learning_rate": 4.711864406779661e-06,
-      "loss": 1.225,
-      "step": 23050
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 4.507009506225586,
-      "learning_rate": 4.705084745762713e-06,
-      "loss": 1.3274,
-      "step": 23060
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 14.352507591247559,
-      "learning_rate": 4.698305084745763e-06,
-      "loss": 1.3935,
-      "step": 23070
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 2.108365297317505,
-      "learning_rate": 4.691525423728814e-06,
-      "loss": 1.294,
-      "step": 23080
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 6.877602577209473,
-      "learning_rate": 4.684745762711865e-06,
-      "loss": 1.3154,
-      "step": 23090
-    },
-    {
-      "epoch": 0.53,
-      "grad_norm": 12.653403282165527,
-      "learning_rate": 4.677966101694916e-06,
-      "loss": 1.1215,
-      "step": 23100
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 6.832087993621826,
-      "learning_rate": 4.671186440677967e-06,
-      "loss": 1.3124,
-      "step": 23110
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 8.196941375732422,
-      "learning_rate": 4.664406779661017e-06,
-      "loss": 1.357,
-      "step": 23120
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 1.573502540588379,
-      "learning_rate": 4.657627118644068e-06,
-      "loss": 1.3047,
-      "step": 23130
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 6.144334316253662,
-      "learning_rate": 4.650847457627119e-06,
-      "loss": 1.2824,
-      "step": 23140
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 2.6499195098876953,
-      "learning_rate": 4.64406779661017e-06,
-      "loss": 1.3157,
-      "step": 23150
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 3.0066230297088623,
-      "learning_rate": 4.637288135593221e-06,
-      "loss": 1.2067,
-      "step": 23160
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 2.056504726409912,
-      "learning_rate": 4.630508474576272e-06,
-      "loss": 1.3069,
-      "step": 23170
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 5.741653919219971,
-      "learning_rate": 4.623728813559323e-06,
-      "loss": 1.2342,
-      "step": 23180
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 14.528912544250488,
-      "learning_rate": 4.616949152542373e-06,
-      "loss": 1.1091,
-      "step": 23190
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 2.9248435497283936,
-      "learning_rate": 4.610169491525424e-06,
-      "loss": 1.0753,
-      "step": 23200
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 7.365577697753906,
-      "learning_rate": 4.603389830508475e-06,
-      "loss": 1.2809,
-      "step": 23210
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 16.456893920898438,
-      "learning_rate": 4.596610169491526e-06,
-      "loss": 1.0776,
-      "step": 23220
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 2.7638049125671387,
-      "learning_rate": 4.589830508474576e-06,
-      "loss": 1.3565,
-      "step": 23230
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 3.463066816329956,
-      "learning_rate": 4.583050847457628e-06,
-      "loss": 1.2449,
-      "step": 23240
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 5.465227127075195,
-      "learning_rate": 4.576271186440678e-06,
-      "loss": 1.2915,
-      "step": 23250
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 4.380557060241699,
-      "learning_rate": 4.56949152542373e-06,
-      "loss": 1.3565,
-      "step": 23260
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 4.977199554443359,
-      "learning_rate": 4.56271186440678e-06,
-      "loss": 1.3312,
-      "step": 23270
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 4.714416027069092,
-      "learning_rate": 4.555932203389831e-06,
-      "loss": 1.318,
-      "step": 23280
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 5.0472092628479,
-      "learning_rate": 4.549152542372881e-06,
-      "loss": 1.3783,
-      "step": 23290
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 10.963289260864258,
-      "learning_rate": 4.542372881355933e-06,
-      "loss": 1.3035,
-      "step": 23300
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 2.592956066131592,
-      "learning_rate": 4.535593220338983e-06,
-      "loss": 1.3272,
-      "step": 23310
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 11.702526092529297,
-      "learning_rate": 4.528813559322035e-06,
-      "loss": 1.1363,
-      "step": 23320
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 4.2335710525512695,
-      "learning_rate": 4.522033898305085e-06,
-      "loss": 1.3878,
-      "step": 23330
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 7.734435558319092,
-      "learning_rate": 4.515254237288136e-06,
-      "loss": 1.3184,
-      "step": 23340
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 3.9147768020629883,
-      "learning_rate": 4.508474576271187e-06,
-      "loss": 1.2077,
-      "step": 23350
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 3.1659061908721924,
-      "learning_rate": 4.501694915254238e-06,
-      "loss": 1.3058,
-      "step": 23360
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 3.9177017211914062,
-      "learning_rate": 4.494915254237289e-06,
-      "loss": 1.1948,
-      "step": 23370
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 6.0111870765686035,
-      "learning_rate": 4.488135593220339e-06,
-      "loss": 1.1927,
-      "step": 23380
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 5.929579257965088,
-      "learning_rate": 4.48135593220339e-06,
-      "loss": 1.299,
-      "step": 23390
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 10.30470085144043,
-      "learning_rate": 4.474576271186441e-06,
-      "loss": 1.3172,
-      "step": 23400
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 2.541947841644287,
-      "learning_rate": 4.467796610169492e-06,
-      "loss": 1.1529,
-      "step": 23410
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 9.664104461669922,
-      "learning_rate": 4.461016949152543e-06,
-      "loss": 1.008,
-      "step": 23420
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 3.3850762844085693,
-      "learning_rate": 4.454237288135594e-06,
-      "loss": 1.2801,
-      "step": 23430
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 4.204078197479248,
-      "learning_rate": 4.447457627118645e-06,
-      "loss": 1.3791,
-      "step": 23440
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 1.8552970886230469,
-      "learning_rate": 4.440677966101695e-06,
-      "loss": 1.1793,
-      "step": 23450
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 9.347552299499512,
-      "learning_rate": 4.433898305084746e-06,
-      "loss": 1.2106,
-      "step": 23460
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 3.1846823692321777,
-      "learning_rate": 4.427118644067797e-06,
-      "loss": 1.5138,
-      "step": 23470
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 7.087679862976074,
-      "learning_rate": 4.420338983050848e-06,
-      "loss": 1.1433,
-      "step": 23480
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 6.835491180419922,
-      "learning_rate": 4.413559322033898e-06,
-      "loss": 1.221,
-      "step": 23490
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 5.99943733215332,
-      "learning_rate": 4.40677966101695e-06,
-      "loss": 1.4262,
-      "step": 23500
-    },
-    {
-      "epoch": 0.54,
-      "eval_loss": 1.024583339691162,
-      "eval_runtime": 67.2037,
-      "eval_samples_per_second": 14.88,
-      "eval_steps_per_second": 14.88,
-      "step": 23500
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 11.216558456420898,
-      "learning_rate": 4.4e-06,
-      "loss": 1.2844,
-      "step": 23510
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 8.619065284729004,
-      "learning_rate": 4.393220338983052e-06,
-      "loss": 1.4356,
-      "step": 23520
-    },
-    {
-      "epoch": 0.54,
-      "grad_norm": 11.311079978942871,
-      "learning_rate": 4.386440677966102e-06,
-      "loss": 1.1927,
-      "step": 23530
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 4.765082836151123,
-      "learning_rate": 4.379661016949153e-06,
-      "loss": 1.3196,
-      "step": 23540
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.4813079833984375,
-      "learning_rate": 4.372881355932203e-06,
-      "loss": 1.3019,
-      "step": 23550
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 2.4694595336914062,
-      "learning_rate": 4.366101694915255e-06,
-      "loss": 1.0386,
-      "step": 23560
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 5.6850481033325195,
-      "learning_rate": 4.359322033898305e-06,
-      "loss": 1.3539,
-      "step": 23570
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 4.677857398986816,
-      "learning_rate": 4.352542372881357e-06,
-      "loss": 1.3064,
-      "step": 23580
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 7.017549514770508,
-      "learning_rate": 4.345762711864407e-06,
-      "loss": 1.2506,
-      "step": 23590
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 7.754286766052246,
-      "learning_rate": 4.338983050847458e-06,
-      "loss": 1.2139,
-      "step": 23600
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 8.794230461120605,
-      "learning_rate": 4.332203389830509e-06,
-      "loss": 1.2354,
-      "step": 23610
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 8.946222305297852,
-      "learning_rate": 4.32542372881356e-06,
-      "loss": 1.1265,
-      "step": 23620
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 6.819628715515137,
-      "learning_rate": 4.318644067796611e-06,
-      "loss": 1.2775,
-      "step": 23630
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 5.190433502197266,
-      "learning_rate": 4.311864406779661e-06,
-      "loss": 1.3512,
-      "step": 23640
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.929642677307129,
-      "learning_rate": 4.305084745762712e-06,
-      "loss": 1.2403,
-      "step": 23650
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 8.023259162902832,
-      "learning_rate": 4.298305084745763e-06,
-      "loss": 1.336,
-      "step": 23660
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 5.015733242034912,
-      "learning_rate": 4.291525423728814e-06,
-      "loss": 1.2565,
-      "step": 23670
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 7.108937740325928,
-      "learning_rate": 4.284745762711865e-06,
-      "loss": 1.2121,
-      "step": 23680
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 10.912981033325195,
-      "learning_rate": 4.277966101694915e-06,
-      "loss": 1.353,
-      "step": 23690
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 2.0632991790771484,
-      "learning_rate": 4.271186440677967e-06,
-      "loss": 1.2225,
-      "step": 23700
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 9.637491226196289,
-      "learning_rate": 4.264406779661017e-06,
-      "loss": 1.1805,
-      "step": 23710
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 8.959497451782227,
-      "learning_rate": 4.257627118644068e-06,
-      "loss": 1.2289,
-      "step": 23720
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 9.83276081085205,
-      "learning_rate": 4.250847457627119e-06,
-      "loss": 1.0274,
-      "step": 23730
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 9.423148155212402,
-      "learning_rate": 4.24406779661017e-06,
-      "loss": 1.2569,
-      "step": 23740
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 4.011594772338867,
-      "learning_rate": 4.23728813559322e-06,
-      "loss": 1.1647,
-      "step": 23750
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 6.655368328094482,
-      "learning_rate": 4.230508474576272e-06,
-      "loss": 1.4665,
-      "step": 23760
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.5101451873779297,
-      "learning_rate": 4.223728813559322e-06,
-      "loss": 1.1459,
-      "step": 23770
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 20.52288818359375,
-      "learning_rate": 4.216949152542374e-06,
-      "loss": 1.1491,
-      "step": 23780
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 9.20742130279541,
-      "learning_rate": 4.210169491525424e-06,
-      "loss": 1.2733,
-      "step": 23790
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 6.216460704803467,
-      "learning_rate": 4.203389830508475e-06,
-      "loss": 1.2745,
-      "step": 23800
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 2.417329788208008,
-      "learning_rate": 4.196610169491525e-06,
-      "loss": 1.035,
-      "step": 23810
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 10.275775909423828,
-      "learning_rate": 4.189830508474577e-06,
-      "loss": 1.2926,
-      "step": 23820
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.9145314693450928,
-      "learning_rate": 4.183050847457627e-06,
-      "loss": 1.2843,
-      "step": 23830
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 15.721736907958984,
-      "learning_rate": 4.176271186440679e-06,
-      "loss": 1.2988,
-      "step": 23840
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.7596094608306885,
-      "learning_rate": 4.169491525423729e-06,
-      "loss": 1.3231,
-      "step": 23850
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 1.9014712572097778,
-      "learning_rate": 4.16271186440678e-06,
-      "loss": 1.3156,
-      "step": 23860
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.3489229679107666,
-      "learning_rate": 4.155932203389831e-06,
-      "loss": 1.3292,
-      "step": 23870
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 4.741728782653809,
-      "learning_rate": 4.149152542372882e-06,
-      "loss": 1.317,
-      "step": 23880
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 8.91144847869873,
-      "learning_rate": 4.142372881355933e-06,
-      "loss": 1.3414,
-      "step": 23890
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 1.4547756910324097,
-      "learning_rate": 4.135593220338983e-06,
-      "loss": 1.2971,
-      "step": 23900
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 8.662474632263184,
-      "learning_rate": 4.128813559322034e-06,
-      "loss": 1.1533,
-      "step": 23910
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.9637677669525146,
-      "learning_rate": 4.122033898305085e-06,
-      "loss": 1.2564,
-      "step": 23920
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 4.98095703125,
-      "learning_rate": 4.115254237288136e-06,
-      "loss": 1.0988,
-      "step": 23930
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 6.112912654876709,
-      "learning_rate": 4.108474576271187e-06,
-      "loss": 1.058,
-      "step": 23940
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 3.396317720413208,
-      "learning_rate": 4.101694915254237e-06,
-      "loss": 1.0615,
-      "step": 23950
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 2.1811277866363525,
-      "learning_rate": 4.094915254237289e-06,
-      "loss": 1.0981,
-      "step": 23960
-    },
-    {
-      "epoch": 0.55,
-      "grad_norm": 7.639281749725342,
-      "learning_rate": 4.088135593220339e-06,
-      "loss": 1.39,
-      "step": 23970
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 6.836162567138672,
-      "learning_rate": 4.081355932203391e-06,
-      "loss": 1.3098,
-      "step": 23980
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 4.436983108520508,
-      "learning_rate": 4.074576271186441e-06,
-      "loss": 1.2947,
-      "step": 23990
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 2.344386339187622,
-      "learning_rate": 4.067796610169492e-06,
-      "loss": 1.23,
-      "step": 24000
-    },
-    {
-      "epoch": 0.56,
-      "eval_loss": 1.038734793663025,
-      "eval_runtime": 68.6728,
-      "eval_samples_per_second": 14.562,
-      "eval_steps_per_second": 14.562,
-      "step": 24000
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 2.359607219696045,
-      "learning_rate": 4.061016949152542e-06,
-      "loss": 1.4914,
-      "step": 24010
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 3.044067621231079,
-      "learning_rate": 4.054237288135594e-06,
-      "loss": 1.0596,
-      "step": 24020
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 6.635824203491211,
-      "learning_rate": 4.047457627118644e-06,
-      "loss": 1.1495,
-      "step": 24030
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 2.152034044265747,
-      "learning_rate": 4.040677966101696e-06,
-      "loss": 1.2823,
-      "step": 24040
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 6.193742752075195,
-      "learning_rate": 4.033898305084746e-06,
-      "loss": 1.2369,
-      "step": 24050
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 5.2856245040893555,
-      "learning_rate": 4.027118644067797e-06,
-      "loss": 1.3229,
-      "step": 24060
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 4.423139572143555,
-      "learning_rate": 4.020338983050847e-06,
-      "loss": 1.4499,
-      "step": 24070
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 1.6255598068237305,
-      "learning_rate": 4.013559322033899e-06,
-      "loss": 1.1325,
-      "step": 24080
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 1.646109700202942,
-      "learning_rate": 4.006779661016949e-06,
-      "loss": 1.2737,
-      "step": 24090
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 5.978227615356445,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 1.1537,
-      "step": 24100
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 9.052789688110352,
-      "learning_rate": 3.993220338983051e-06,
-      "loss": 1.4989,
-      "step": 24110
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 5.04679012298584,
-      "learning_rate": 3.986440677966102e-06,
-      "loss": 1.3025,
-      "step": 24120
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 4.390671730041504,
-      "learning_rate": 3.979661016949153e-06,
-      "loss": 1.3109,
-      "step": 24130
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 11.515645027160645,
-      "learning_rate": 3.972881355932204e-06,
-      "loss": 1.3073,
-      "step": 24140
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 8.161099433898926,
-      "learning_rate": 3.966101694915255e-06,
-      "loss": 1.0133,
-      "step": 24150
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 5.4128851890563965,
-      "learning_rate": 3.959322033898305e-06,
-      "loss": 1.3091,
-      "step": 24160
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 1.8805668354034424,
-      "learning_rate": 3.952542372881356e-06,
-      "loss": 1.4176,
-      "step": 24170
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 7.201099395751953,
-      "learning_rate": 3.945762711864407e-06,
-      "loss": 1.0447,
-      "step": 24180
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 2.0222816467285156,
-      "learning_rate": 3.938983050847458e-06,
-      "loss": 1.2018,
-      "step": 24190
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 4.226831912994385,
-      "learning_rate": 3.932203389830509e-06,
-      "loss": 1.1199,
-      "step": 24200
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 6.030657768249512,
-      "learning_rate": 3.925423728813559e-06,
-      "loss": 1.2632,
-      "step": 24210
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 2.487889528274536,
-      "learning_rate": 3.918644067796611e-06,
-      "loss": 1.3595,
-      "step": 24220
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 2.413583755493164,
-      "learning_rate": 3.911864406779661e-06,
-      "loss": 1.2903,
-      "step": 24230
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 8.041131019592285,
-      "learning_rate": 3.905084745762713e-06,
-      "loss": 1.1923,
-      "step": 24240
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 7.538460731506348,
-      "learning_rate": 3.898305084745763e-06,
-      "loss": 1.3482,
-      "step": 24250
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 3.833472728729248,
-      "learning_rate": 3.891525423728814e-06,
-      "loss": 1.2592,
-      "step": 24260
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 3.3971173763275146,
-      "learning_rate": 3.884745762711864e-06,
-      "loss": 1.2468,
-      "step": 24270
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 10.157732009887695,
-      "learning_rate": 3.877966101694916e-06,
-      "loss": 1.3728,
-      "step": 24280
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 3.008604049682617,
-      "learning_rate": 3.871186440677966e-06,
-      "loss": 1.0438,
-      "step": 24290
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 9.531476020812988,
-      "learning_rate": 3.864406779661018e-06,
-      "loss": 1.3266,
-      "step": 24300
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 6.8482818603515625,
-      "learning_rate": 3.857627118644068e-06,
-      "loss": 1.3832,
-      "step": 24310
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 2.2738068103790283,
-      "learning_rate": 3.850847457627119e-06,
-      "loss": 1.3126,
-      "step": 24320
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 9.141778945922852,
-      "learning_rate": 3.844067796610169e-06,
-      "loss": 1.1799,
-      "step": 24330
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 4.968504905700684,
-      "learning_rate": 3.837288135593221e-06,
-      "loss": 1.2729,
-      "step": 24340
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 1.8136253356933594,
-      "learning_rate": 3.830508474576271e-06,
-      "loss": 1.2373,
-      "step": 24350
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 10.53274917602539,
-      "learning_rate": 3.823728813559323e-06,
-      "loss": 1.1203,
-      "step": 24360
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 4.648149013519287,
-      "learning_rate": 3.816949152542373e-06,
-      "loss": 1.1483,
-      "step": 24370
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 8.629014015197754,
-      "learning_rate": 3.8101694915254238e-06,
-      "loss": 1.3082,
-      "step": 24380
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 6.339970588684082,
-      "learning_rate": 3.8033898305084748e-06,
-      "loss": 1.4333,
-      "step": 24390
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 5.55668830871582,
-      "learning_rate": 3.7966101694915257e-06,
-      "loss": 1.4178,
-      "step": 24400
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 6.1273884773254395,
-      "learning_rate": 3.7898305084745767e-06,
-      "loss": 1.1529,
-      "step": 24410
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.123335361480713,
-      "learning_rate": 3.7830508474576273e-06,
-      "loss": 1.1294,
-      "step": 24420
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 4.093799591064453,
-      "learning_rate": 3.7762711864406782e-06,
-      "loss": 1.263,
-      "step": 24430
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 7.780729293823242,
-      "learning_rate": 3.7694915254237292e-06,
-      "loss": 1.1628,
-      "step": 24440
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 5.5196614265441895,
-      "learning_rate": 3.76271186440678e-06,
-      "loss": 1.204,
-      "step": 24450
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 4.023863315582275,
-      "learning_rate": 3.755932203389831e-06,
-      "loss": 1.2161,
-      "step": 24460
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 12.464215278625488,
-      "learning_rate": 3.7491525423728813e-06,
-      "loss": 1.1825,
-      "step": 24470
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.583104372024536,
-      "learning_rate": 3.7423728813559323e-06,
-      "loss": 1.2815,
-      "step": 24480
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 2.9160475730895996,
-      "learning_rate": 3.7355932203389833e-06,
-      "loss": 1.2902,
-      "step": 24490
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 5.577888488769531,
-      "learning_rate": 3.7288135593220342e-06,
-      "loss": 1.2927,
-      "step": 24500
-    },
-    {
-      "epoch": 0.57,
-      "eval_loss": 1.001274585723877,
-      "eval_runtime": 69.0094,
-      "eval_samples_per_second": 14.491,
-      "eval_steps_per_second": 14.491,
-      "step": 24500
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 26.491436004638672,
-      "learning_rate": 3.7220338983050852e-06,
-      "loss": 1.3748,
-      "step": 24510
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.675736904144287,
-      "learning_rate": 3.715254237288136e-06,
-      "loss": 1.3274,
-      "step": 24520
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 0.7855976223945618,
-      "learning_rate": 3.7084745762711867e-06,
-      "loss": 1.1811,
-      "step": 24530
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 16.270326614379883,
-      "learning_rate": 3.7016949152542377e-06,
-      "loss": 1.2293,
-      "step": 24540
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.6461169719696045,
-      "learning_rate": 3.6949152542372883e-06,
-      "loss": 1.2937,
-      "step": 24550
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 5.9607954025268555,
-      "learning_rate": 3.6881355932203393e-06,
-      "loss": 1.2017,
-      "step": 24560
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 8.822928428649902,
-      "learning_rate": 3.6813559322033902e-06,
-      "loss": 1.2073,
-      "step": 24570
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.9838669300079346,
-      "learning_rate": 3.6745762711864408e-06,
-      "loss": 1.0297,
-      "step": 24580
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.180518865585327,
-      "learning_rate": 3.6677966101694918e-06,
-      "loss": 1.2039,
-      "step": 24590
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.557624340057373,
-      "learning_rate": 3.6610169491525427e-06,
-      "loss": 1.2882,
-      "step": 24600
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 2.1503684520721436,
-      "learning_rate": 3.6542372881355937e-06,
-      "loss": 1.2697,
-      "step": 24610
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 4.290503978729248,
-      "learning_rate": 3.6474576271186447e-06,
-      "loss": 1.3283,
-      "step": 24620
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 12.416634559631348,
-      "learning_rate": 3.640677966101695e-06,
-      "loss": 1.2059,
-      "step": 24630
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.7786366939544678,
-      "learning_rate": 3.633898305084746e-06,
-      "loss": 1.1957,
-      "step": 24640
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 6.75616455078125,
-      "learning_rate": 3.6271186440677968e-06,
-      "loss": 1.3586,
-      "step": 24650
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 8.630069732666016,
-      "learning_rate": 3.6203389830508478e-06,
-      "loss": 1.1611,
-      "step": 24660
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.7713470458984375,
-      "learning_rate": 3.6135593220338987e-06,
-      "loss": 1.1613,
-      "step": 24670
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 8.666251182556152,
-      "learning_rate": 3.6067796610169493e-06,
-      "loss": 1.3801,
-      "step": 24680
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 13.79195785522461,
-      "learning_rate": 3.6000000000000003e-06,
-      "loss": 1.3715,
-      "step": 24690
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 1.1425895690917969,
-      "learning_rate": 3.5932203389830512e-06,
-      "loss": 1.2213,
-      "step": 24700
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 4.308348655700684,
-      "learning_rate": 3.5864406779661022e-06,
-      "loss": 1.2354,
-      "step": 24710
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 7.868862152099609,
-      "learning_rate": 3.579661016949153e-06,
-      "loss": 1.2111,
-      "step": 24720
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 8.803450584411621,
-      "learning_rate": 3.5728813559322033e-06,
-      "loss": 1.2107,
-      "step": 24730
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 5.428405284881592,
-      "learning_rate": 3.5661016949152543e-06,
-      "loss": 1.3327,
-      "step": 24740
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 4.115501880645752,
-      "learning_rate": 3.5593220338983053e-06,
-      "loss": 1.167,
-      "step": 24750
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 19.842674255371094,
-      "learning_rate": 3.5525423728813563e-06,
-      "loss": 1.2822,
-      "step": 24760
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 17.216901779174805,
-      "learning_rate": 3.5457627118644072e-06,
-      "loss": 1.1838,
-      "step": 24770
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.8147220611572266,
-      "learning_rate": 3.538983050847458e-06,
-      "loss": 1.3505,
-      "step": 24780
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 6.949516773223877,
-      "learning_rate": 3.5322033898305088e-06,
-      "loss": 1.3103,
-      "step": 24790
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 3.2688727378845215,
-      "learning_rate": 3.5254237288135597e-06,
-      "loss": 1.1356,
-      "step": 24800
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 9.062920570373535,
-      "learning_rate": 3.5186440677966103e-06,
-      "loss": 1.182,
-      "step": 24810
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 5.235373497009277,
-      "learning_rate": 3.5118644067796613e-06,
-      "loss": 1.3181,
-      "step": 24820
-    },
-    {
-      "epoch": 0.57,
-      "grad_norm": 6.707292079925537,
-      "learning_rate": 3.5050847457627122e-06,
-      "loss": 1.204,
-      "step": 24830
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 3.9347891807556152,
-      "learning_rate": 3.498305084745763e-06,
-      "loss": 1.2906,
-      "step": 24840
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 4.399641513824463,
-      "learning_rate": 3.4915254237288138e-06,
-      "loss": 1.3014,
-      "step": 24850
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 2.8803553581237793,
-      "learning_rate": 3.4847457627118648e-06,
-      "loss": 1.2676,
-      "step": 24860
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 8.522972106933594,
-      "learning_rate": 3.4779661016949157e-06,
-      "loss": 1.194,
-      "step": 24870
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 6.1468825340271,
-      "learning_rate": 3.4711864406779667e-06,
-      "loss": 1.172,
-      "step": 24880
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 4.32474422454834,
-      "learning_rate": 3.464406779661017e-06,
-      "loss": 1.2484,
-      "step": 24890
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 11.452080726623535,
-      "learning_rate": 3.457627118644068e-06,
-      "loss": 1.291,
-      "step": 24900
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 1.0818170309066772,
-      "learning_rate": 3.450847457627119e-06,
-      "loss": 1.1327,
-      "step": 24910
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.037398815155029,
-      "learning_rate": 3.4440677966101698e-06,
-      "loss": 1.3494,
-      "step": 24920
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 3.8455469608306885,
-      "learning_rate": 3.4372881355932207e-06,
-      "loss": 1.2283,
-      "step": 24930
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.2556681632995605,
-      "learning_rate": 3.4305084745762713e-06,
-      "loss": 1.1897,
-      "step": 24940
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 8.240093231201172,
-      "learning_rate": 3.4237288135593223e-06,
-      "loss": 1.3777,
-      "step": 24950
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 2.0133845806121826,
-      "learning_rate": 3.4169491525423733e-06,
-      "loss": 1.2694,
-      "step": 24960
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 1.9996819496154785,
-      "learning_rate": 3.4101694915254242e-06,
-      "loss": 1.2311,
-      "step": 24970
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 3.0491392612457275,
-      "learning_rate": 3.403389830508475e-06,
-      "loss": 1.2316,
-      "step": 24980
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 21.063587188720703,
-      "learning_rate": 3.3966101694915253e-06,
-      "loss": 1.2148,
-      "step": 24990
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 13.847090721130371,
-      "learning_rate": 3.3898305084745763e-06,
-      "loss": 1.2155,
-      "step": 25000
-    },
-    {
-      "epoch": 0.58,
-      "eval_loss": 0.9729424715042114,
-      "eval_runtime": 68.3419,
-      "eval_samples_per_second": 14.632,
-      "eval_steps_per_second": 14.632,
-      "step": 25000
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.260663032531738,
-      "learning_rate": 3.3830508474576273e-06,
-      "loss": 1.2702,
-      "step": 25010
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.188079357147217,
-      "learning_rate": 3.3762711864406783e-06,
-      "loss": 1.2995,
-      "step": 25020
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 7.982115745544434,
-      "learning_rate": 3.3694915254237292e-06,
-      "loss": 1.2687,
-      "step": 25030
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 3.1299421787261963,
-      "learning_rate": 3.3627118644067802e-06,
-      "loss": 1.3372,
-      "step": 25040
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.5412726402282715,
-      "learning_rate": 3.3559322033898308e-06,
-      "loss": 1.2002,
-      "step": 25050
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 6.219215393066406,
-      "learning_rate": 3.3491525423728817e-06,
-      "loss": 1.1117,
-      "step": 25060
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 3.6740598678588867,
-      "learning_rate": 3.3423728813559327e-06,
-      "loss": 1.2222,
-      "step": 25070
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.168256759643555,
-      "learning_rate": 3.3355932203389833e-06,
-      "loss": 1.2059,
-      "step": 25080
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 4.997997283935547,
-      "learning_rate": 3.3288135593220343e-06,
-      "loss": 1.3613,
-      "step": 25090
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 10.958406448364258,
-      "learning_rate": 3.322033898305085e-06,
-      "loss": 1.2508,
-      "step": 25100
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 7.793730735778809,
-      "learning_rate": 3.3152542372881358e-06,
-      "loss": 1.1418,
-      "step": 25110
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 6.706384181976318,
-      "learning_rate": 3.3084745762711868e-06,
-      "loss": 1.1661,
-      "step": 25120
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 9.17857551574707,
-      "learning_rate": 3.3016949152542377e-06,
-      "loss": 1.1569,
-      "step": 25130
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 1.5717544555664062,
-      "learning_rate": 3.2949152542372887e-06,
-      "loss": 1.0838,
-      "step": 25140
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 7.292511940002441,
-      "learning_rate": 3.288135593220339e-06,
-      "loss": 1.3231,
-      "step": 25150
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 6.030792236328125,
-      "learning_rate": 3.28135593220339e-06,
-      "loss": 1.1247,
-      "step": 25160
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 2.3071470260620117,
-      "learning_rate": 3.274576271186441e-06,
-      "loss": 1.113,
-      "step": 25170
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.088674545288086,
-      "learning_rate": 3.2677966101694918e-06,
-      "loss": 1.0984,
-      "step": 25180
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 13.216058731079102,
-      "learning_rate": 3.2610169491525428e-06,
-      "loss": 1.3596,
-      "step": 25190
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.061989784240723,
-      "learning_rate": 3.2542372881355933e-06,
-      "loss": 1.1028,
-      "step": 25200
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.740258693695068,
-      "learning_rate": 3.2474576271186443e-06,
-      "loss": 1.14,
-      "step": 25210
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 5.7218098640441895,
-      "learning_rate": 3.2406779661016953e-06,
-      "loss": 1.3268,
-      "step": 25220
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 10.579949378967285,
-      "learning_rate": 3.2338983050847462e-06,
-      "loss": 1.1409,
-      "step": 25230
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 10.984354019165039,
-      "learning_rate": 3.2271186440677972e-06,
-      "loss": 1.3797,
-      "step": 25240
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 8.028718948364258,
-      "learning_rate": 3.2203389830508473e-06,
-      "loss": 1.2275,
-      "step": 25250
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 2.102750301361084,
-      "learning_rate": 3.2135593220338983e-06,
-      "loss": 1.3638,
-      "step": 25260
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 12.67302417755127,
-      "learning_rate": 3.2067796610169493e-06,
-      "loss": 1.3955,
-      "step": 25270
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 9.690311431884766,
-      "learning_rate": 3.2000000000000003e-06,
-      "loss": 1.178,
-      "step": 25280
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 9.31050968170166,
-      "learning_rate": 3.1932203389830513e-06,
-      "loss": 1.3644,
-      "step": 25290
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.5038933753967285,
-      "learning_rate": 3.186440677966102e-06,
-      "loss": 1.4397,
-      "step": 25300
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 4.691766262054443,
-      "learning_rate": 3.1796610169491528e-06,
-      "loss": 1.3062,
-      "step": 25310
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 9.108213424682617,
-      "learning_rate": 3.1728813559322038e-06,
-      "loss": 1.1835,
-      "step": 25320
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.10239315032959,
-      "learning_rate": 3.1661016949152547e-06,
-      "loss": 1.3249,
-      "step": 25330
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 3.791691780090332,
-      "learning_rate": 3.1593220338983053e-06,
-      "loss": 1.2023,
-      "step": 25340
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 5.800852298736572,
-      "learning_rate": 3.1525423728813563e-06,
-      "loss": 1.352,
-      "step": 25350
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.795603275299072,
-      "learning_rate": 3.145762711864407e-06,
-      "loss": 1.3013,
-      "step": 25360
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 8.434121131896973,
-      "learning_rate": 3.138983050847458e-06,
-      "loss": 1.1872,
-      "step": 25370
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 4.576048851013184,
-      "learning_rate": 3.1322033898305088e-06,
-      "loss": 1.2377,
-      "step": 25380
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 11.319803237915039,
-      "learning_rate": 3.1254237288135598e-06,
-      "loss": 1.1306,
-      "step": 25390
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 11.858842849731445,
-      "learning_rate": 3.1186440677966107e-06,
-      "loss": 1.3123,
-      "step": 25400
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 10.17744255065918,
-      "learning_rate": 3.111864406779661e-06,
-      "loss": 1.2951,
-      "step": 25410
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 5.539567947387695,
-      "learning_rate": 3.105084745762712e-06,
-      "loss": 1.369,
-      "step": 25420
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.66847038269043,
-      "learning_rate": 3.098305084745763e-06,
-      "loss": 1.1254,
-      "step": 25430
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 2.0820562839508057,
-      "learning_rate": 3.091525423728814e-06,
-      "loss": 1.0718,
-      "step": 25440
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 1.1296765804290771,
-      "learning_rate": 3.0847457627118648e-06,
-      "loss": 1.3201,
-      "step": 25450
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.302907943725586,
-      "learning_rate": 3.0779661016949153e-06,
-      "loss": 1.3956,
-      "step": 25460
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 7.594703674316406,
-      "learning_rate": 3.0711864406779663e-06,
-      "loss": 1.109,
-      "step": 25470
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 4.60818338394165,
-      "learning_rate": 3.0644067796610173e-06,
-      "loss": 1.194,
-      "step": 25480
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 9.426482200622559,
-      "learning_rate": 3.0576271186440683e-06,
-      "loss": 1.1175,
-      "step": 25490
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 1.6626375913619995,
-      "learning_rate": 3.0508474576271192e-06,
-      "loss": 1.241,
-      "step": 25500
-    },
-    {
-      "epoch": 0.59,
-      "eval_loss": 0.998432993888855,
-      "eval_runtime": 66.8613,
-      "eval_samples_per_second": 14.956,
-      "eval_steps_per_second": 14.956,
-      "step": 25500
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 5.7481255531311035,
-      "learning_rate": 3.0440677966101694e-06,
-      "loss": 1.0914,
-      "step": 25510
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.295405864715576,
-      "learning_rate": 3.0372881355932203e-06,
-      "loss": 1.2265,
-      "step": 25520
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 7.302733898162842,
-      "learning_rate": 3.0305084745762713e-06,
-      "loss": 1.2384,
-      "step": 25530
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.079695224761963,
-      "learning_rate": 3.0237288135593223e-06,
-      "loss": 1.3285,
-      "step": 25540
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 3.6032302379608154,
-      "learning_rate": 3.0169491525423733e-06,
-      "loss": 1.3156,
-      "step": 25550
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 4.823906421661377,
-      "learning_rate": 3.010169491525424e-06,
-      "loss": 0.8886,
-      "step": 25560
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 3.6529293060302734,
-      "learning_rate": 3.003389830508475e-06,
-      "loss": 1.2771,
-      "step": 25570
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 8.288748741149902,
-      "learning_rate": 2.9966101694915258e-06,
-      "loss": 1.1765,
-      "step": 25580
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 17.39524269104004,
-      "learning_rate": 2.9898305084745768e-06,
-      "loss": 1.3095,
-      "step": 25590
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 6.149013042449951,
-      "learning_rate": 2.9830508474576277e-06,
-      "loss": 0.9663,
-      "step": 25600
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 9.429570198059082,
-      "learning_rate": 2.9762711864406783e-06,
-      "loss": 1.1291,
-      "step": 25610
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 7.571442127227783,
-      "learning_rate": 2.969491525423729e-06,
-      "loss": 1.0891,
-      "step": 25620
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 8.435641288757324,
-      "learning_rate": 2.96271186440678e-06,
-      "loss": 1.207,
-      "step": 25630
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 5.694164276123047,
-      "learning_rate": 2.955932203389831e-06,
-      "loss": 1.2288,
-      "step": 25640
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 7.887094974517822,
-      "learning_rate": 2.9491525423728818e-06,
-      "loss": 1.2683,
-      "step": 25650
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 1.4776153564453125,
-      "learning_rate": 2.9423728813559327e-06,
-      "loss": 1.2045,
-      "step": 25660
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 22.860342025756836,
-      "learning_rate": 2.935593220338983e-06,
-      "loss": 1.1633,
-      "step": 25670
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 5.612551689147949,
-      "learning_rate": 2.928813559322034e-06,
-      "loss": 1.2065,
-      "step": 25680
-    },
-    {
-      "epoch": 0.59,
-      "grad_norm": 4.532423973083496,
-      "learning_rate": 2.922033898305085e-06,
-      "loss": 1.2672,
-      "step": 25690
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 8.843605041503906,
-      "learning_rate": 2.915254237288136e-06,
-      "loss": 1.2333,
-      "step": 25700
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 7.5249481201171875,
-      "learning_rate": 2.9084745762711868e-06,
-      "loss": 1.2351,
-      "step": 25710
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 4.021324634552002,
-      "learning_rate": 2.9016949152542373e-06,
-      "loss": 1.302,
-      "step": 25720
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 12.174166679382324,
-      "learning_rate": 2.8949152542372883e-06,
-      "loss": 1.1858,
-      "step": 25730
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.587007522583008,
-      "learning_rate": 2.8881355932203393e-06,
-      "loss": 1.2249,
-      "step": 25740
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.73293399810791,
-      "learning_rate": 2.8813559322033903e-06,
-      "loss": 1.1867,
-      "step": 25750
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 17.562885284423828,
-      "learning_rate": 2.8745762711864412e-06,
-      "loss": 1.2162,
-      "step": 25760
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 2.855794668197632,
-      "learning_rate": 2.8677966101694914e-06,
-      "loss": 1.2218,
-      "step": 25770
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 8.89600658416748,
-      "learning_rate": 2.8610169491525424e-06,
-      "loss": 1.288,
-      "step": 25780
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.732436180114746,
-      "learning_rate": 2.8542372881355933e-06,
-      "loss": 1.4694,
-      "step": 25790
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 7.497152805328369,
-      "learning_rate": 2.8474576271186443e-06,
-      "loss": 1.1525,
-      "step": 25800
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 1.6418439149856567,
-      "learning_rate": 2.8406779661016953e-06,
-      "loss": 1.1873,
-      "step": 25810
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.75865650177002,
-      "learning_rate": 2.833898305084746e-06,
-      "loss": 1.3353,
-      "step": 25820
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.915677070617676,
-      "learning_rate": 2.827118644067797e-06,
-      "loss": 1.2384,
-      "step": 25830
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 11.455119132995605,
-      "learning_rate": 2.820338983050848e-06,
-      "loss": 1.0484,
-      "step": 25840
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 4.330808162689209,
-      "learning_rate": 2.8135593220338988e-06,
-      "loss": 1.1905,
-      "step": 25850
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 7.7616376876831055,
-      "learning_rate": 2.8067796610169497e-06,
-      "loss": 1.2721,
-      "step": 25860
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.885458946228027,
-      "learning_rate": 2.8000000000000003e-06,
-      "loss": 0.9887,
-      "step": 25870
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 7.448554039001465,
-      "learning_rate": 2.793220338983051e-06,
-      "loss": 1.1294,
-      "step": 25880
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 3.008898973464966,
-      "learning_rate": 2.786440677966102e-06,
-      "loss": 1.2489,
-      "step": 25890
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 8.211142539978027,
-      "learning_rate": 2.779661016949153e-06,
-      "loss": 1.2185,
-      "step": 25900
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 7.966844081878662,
-      "learning_rate": 2.7728813559322038e-06,
-      "loss": 1.3666,
-      "step": 25910
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 7.830656051635742,
-      "learning_rate": 2.7661016949152548e-06,
-      "loss": 1.3537,
-      "step": 25920
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 4.714284420013428,
-      "learning_rate": 2.7593220338983053e-06,
-      "loss": 1.1768,
-      "step": 25930
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 11.090384483337402,
-      "learning_rate": 2.752542372881356e-06,
-      "loss": 1.3252,
-      "step": 25940
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 7.201011657714844,
-      "learning_rate": 2.745762711864407e-06,
-      "loss": 1.3252,
-      "step": 25950
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 3.7255921363830566,
-      "learning_rate": 2.738983050847458e-06,
-      "loss": 1.2302,
-      "step": 25960
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.323339462280273,
-      "learning_rate": 2.732203389830509e-06,
-      "loss": 1.2951,
-      "step": 25970
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 24.529966354370117,
-      "learning_rate": 2.7254237288135593e-06,
-      "loss": 1.2912,
-      "step": 25980
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 4.739007472991943,
-      "learning_rate": 2.7186440677966103e-06,
-      "loss": 0.9912,
-      "step": 25990
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 8.324442863464355,
-      "learning_rate": 2.7118644067796613e-06,
-      "loss": 1.1363,
-      "step": 26000
-    },
-    {
-      "epoch": 0.6,
-      "eval_loss": 0.9869400858879089,
-      "eval_runtime": 66.847,
-      "eval_samples_per_second": 14.96,
-      "eval_steps_per_second": 14.96,
-      "step": 26000
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 8.513010025024414,
-      "learning_rate": 2.7050847457627123e-06,
-      "loss": 1.1538,
-      "step": 26010
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 3.9168806076049805,
-      "learning_rate": 2.6983050847457633e-06,
-      "loss": 1.0728,
-      "step": 26020
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 5.090580463409424,
-      "learning_rate": 2.6915254237288134e-06,
-      "loss": 1.2442,
-      "step": 26030
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 5.569304466247559,
-      "learning_rate": 2.6847457627118644e-06,
-      "loss": 0.9942,
-      "step": 26040
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 6.856686115264893,
-      "learning_rate": 2.6779661016949153e-06,
-      "loss": 1.1926,
-      "step": 26050
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 4.85548210144043,
-      "learning_rate": 2.6711864406779663e-06,
-      "loss": 1.1911,
-      "step": 26060
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 5.3112335205078125,
-      "learning_rate": 2.6644067796610173e-06,
-      "loss": 1.1381,
-      "step": 26070
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 5.764985084533691,
-      "learning_rate": 2.657627118644068e-06,
-      "loss": 1.1531,
-      "step": 26080
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 6.8243184089660645,
-      "learning_rate": 2.650847457627119e-06,
-      "loss": 1.1892,
-      "step": 26090
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 6.5991692543029785,
-      "learning_rate": 2.64406779661017e-06,
-      "loss": 1.0998,
-      "step": 26100
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 9.651345252990723,
-      "learning_rate": 2.6372881355932208e-06,
-      "loss": 1.2483,
-      "step": 26110
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 3.886197328567505,
-      "learning_rate": 2.6305084745762718e-06,
-      "loss": 1.1655,
-      "step": 26120
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 4.277401924133301,
-      "learning_rate": 2.6237288135593223e-06,
-      "loss": 1.2908,
-      "step": 26130
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 4.072823524475098,
-      "learning_rate": 2.616949152542373e-06,
-      "loss": 1.0785,
-      "step": 26140
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 4.426355838775635,
-      "learning_rate": 2.610169491525424e-06,
-      "loss": 1.157,
-      "step": 26150
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 1.601050853729248,
-      "learning_rate": 2.603389830508475e-06,
-      "loss": 1.3465,
-      "step": 26160
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 22.35835075378418,
-      "learning_rate": 2.596610169491526e-06,
-      "loss": 1.3287,
-      "step": 26170
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 4.574925422668457,
-      "learning_rate": 2.5898305084745768e-06,
-      "loss": 1.1298,
-      "step": 26180
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 6.630609512329102,
-      "learning_rate": 2.5830508474576273e-06,
-      "loss": 1.1522,
-      "step": 26190
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 5.038731098175049,
-      "learning_rate": 2.576271186440678e-06,
-      "loss": 1.1709,
-      "step": 26200
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 9.418252944946289,
-      "learning_rate": 2.569491525423729e-06,
-      "loss": 1.2216,
-      "step": 26210
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 16.634244918823242,
-      "learning_rate": 2.56271186440678e-06,
-      "loss": 1.2476,
-      "step": 26220
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 6.329607009887695,
-      "learning_rate": 2.555932203389831e-06,
-      "loss": 1.3251,
-      "step": 26230
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 7.770134449005127,
-      "learning_rate": 2.5491525423728814e-06,
-      "loss": 1.3357,
-      "step": 26240
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 8.234681129455566,
-      "learning_rate": 2.5423728813559323e-06,
-      "loss": 1.3181,
-      "step": 26250
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 4.855250358581543,
-      "learning_rate": 2.5355932203389833e-06,
-      "loss": 1.3914,
-      "step": 26260
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 9.435198783874512,
-      "learning_rate": 2.5288135593220343e-06,
-      "loss": 1.225,
-      "step": 26270
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 2.1608779430389404,
-      "learning_rate": 2.5220338983050853e-06,
-      "loss": 1.4329,
-      "step": 26280
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 5.846624851226807,
-      "learning_rate": 2.5152542372881354e-06,
-      "loss": 1.2809,
-      "step": 26290
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 2.737983465194702,
-      "learning_rate": 2.5084745762711864e-06,
-      "loss": 1.1873,
-      "step": 26300
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 2.908550977706909,
-      "learning_rate": 2.5016949152542374e-06,
-      "loss": 1.1736,
-      "step": 26310
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 36.61166000366211,
-      "learning_rate": 2.4949152542372883e-06,
-      "loss": 1.1875,
-      "step": 26320
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 20.149581909179688,
-      "learning_rate": 2.488135593220339e-06,
-      "loss": 1.3924,
-      "step": 26330
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 8.97331428527832,
-      "learning_rate": 2.48135593220339e-06,
-      "loss": 1.0913,
-      "step": 26340
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 6.3798441886901855,
-      "learning_rate": 2.474576271186441e-06,
-      "loss": 1.0683,
-      "step": 26350
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 13.193538665771484,
-      "learning_rate": 2.467796610169492e-06,
-      "loss": 1.0535,
-      "step": 26360
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 14.773574829101562,
-      "learning_rate": 2.461016949152543e-06,
-      "loss": 1.0665,
-      "step": 26370
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 3.1088144779205322,
-      "learning_rate": 2.4542372881355933e-06,
-      "loss": 1.3309,
-      "step": 26380
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 3.1904797554016113,
-      "learning_rate": 2.4474576271186443e-06,
-      "loss": 1.2429,
-      "step": 26390
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 3.5016331672668457,
-      "learning_rate": 2.4406779661016953e-06,
-      "loss": 1.2189,
-      "step": 26400
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 6.315566539764404,
-      "learning_rate": 2.433898305084746e-06,
-      "loss": 1.2291,
-      "step": 26410
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 14.041393280029297,
-      "learning_rate": 2.427118644067797e-06,
-      "loss": 1.1644,
-      "step": 26420
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 20.225662231445312,
-      "learning_rate": 2.4203389830508474e-06,
-      "loss": 1.2319,
-      "step": 26430
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 3.7186009883880615,
-      "learning_rate": 2.4135593220338984e-06,
-      "loss": 1.2453,
-      "step": 26440
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 2.4279704093933105,
-      "learning_rate": 2.4067796610169493e-06,
-      "loss": 1.0585,
-      "step": 26450
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 4.853421211242676,
-      "learning_rate": 2.4000000000000003e-06,
-      "loss": 1.2464,
-      "step": 26460
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 4.436877250671387,
-      "learning_rate": 2.393220338983051e-06,
-      "loss": 1.3915,
-      "step": 26470
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 3.7520599365234375,
-      "learning_rate": 2.386440677966102e-06,
-      "loss": 1.3439,
-      "step": 26480
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 1.7208768129348755,
-      "learning_rate": 2.379661016949153e-06,
-      "loss": 1.2509,
-      "step": 26490
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 2.465243101119995,
-      "learning_rate": 2.372881355932204e-06,
-      "loss": 1.2785,
-      "step": 26500
-    },
-    {
-      "epoch": 0.61,
-      "eval_loss": 0.9529283046722412,
-      "eval_runtime": 66.4596,
-      "eval_samples_per_second": 15.047,
-      "eval_steps_per_second": 15.047,
-      "step": 26500
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 16.25847816467285,
-      "learning_rate": 2.3661016949152544e-06,
-      "loss": 1.2935,
-      "step": 26510
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 5.584811687469482,
-      "learning_rate": 2.3593220338983053e-06,
-      "loss": 1.1677,
-      "step": 26520
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 5.604238033294678,
-      "learning_rate": 2.3525423728813563e-06,
-      "loss": 1.1095,
-      "step": 26530
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 7.715309143066406,
-      "learning_rate": 2.345762711864407e-06,
-      "loss": 1.3044,
-      "step": 26540
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 1.3220744132995605,
-      "learning_rate": 2.338983050847458e-06,
-      "loss": 1.1298,
-      "step": 26550
-    },
-    {
-      "epoch": 0.61,
-      "grad_norm": 2.8409392833709717,
-      "learning_rate": 2.3322033898305084e-06,
-      "loss": 1.2565,
-      "step": 26560
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 3.3840484619140625,
-      "learning_rate": 2.3254237288135594e-06,
-      "loss": 1.2529,
-      "step": 26570
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 2.952516555786133,
-      "learning_rate": 2.3186440677966103e-06,
-      "loss": 1.3135,
-      "step": 26580
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 6.298173904418945,
-      "learning_rate": 2.3118644067796613e-06,
-      "loss": 1.2517,
-      "step": 26590
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 17.91318702697754,
-      "learning_rate": 2.305084745762712e-06,
-      "loss": 1.1782,
-      "step": 26600
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 9.400506973266602,
-      "learning_rate": 2.298305084745763e-06,
-      "loss": 1.3047,
-      "step": 26610
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 8.402453422546387,
-      "learning_rate": 2.291525423728814e-06,
-      "loss": 1.3617,
-      "step": 26620
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 6.365287780761719,
-      "learning_rate": 2.284745762711865e-06,
-      "loss": 1.2257,
-      "step": 26630
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 2.8725926876068115,
-      "learning_rate": 2.2779661016949154e-06,
-      "loss": 1.0295,
-      "step": 26640
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 13.41898250579834,
-      "learning_rate": 2.2711864406779663e-06,
-      "loss": 1.1358,
-      "step": 26650
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 3.4462170600891113,
-      "learning_rate": 2.2644067796610173e-06,
-      "loss": 1.2295,
-      "step": 26660
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 1.85061514377594,
-      "learning_rate": 2.257627118644068e-06,
-      "loss": 1.2474,
-      "step": 26670
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 4.921010971069336,
-      "learning_rate": 2.250847457627119e-06,
-      "loss": 1.4324,
-      "step": 26680
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 5.649210453033447,
-      "learning_rate": 2.2440677966101694e-06,
-      "loss": 1.2167,
-      "step": 26690
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 13.460455894470215,
-      "learning_rate": 2.2372881355932204e-06,
-      "loss": 1.1124,
-      "step": 26700
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 4.873124599456787,
-      "learning_rate": 2.2305084745762714e-06,
-      "loss": 1.2533,
-      "step": 26710
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 4.260035514831543,
-      "learning_rate": 2.2237288135593223e-06,
-      "loss": 1.1214,
-      "step": 26720
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 9.646221160888672,
-      "learning_rate": 2.216949152542373e-06,
-      "loss": 1.3115,
-      "step": 26730
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 2.029663324356079,
-      "learning_rate": 2.210169491525424e-06,
-      "loss": 1.1277,
-      "step": 26740
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 2.2031261920928955,
-      "learning_rate": 2.203389830508475e-06,
-      "loss": 1.275,
-      "step": 26750
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 7.809500217437744,
-      "learning_rate": 2.196610169491526e-06,
-      "loss": 1.4312,
-      "step": 26760
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 3.175096273422241,
-      "learning_rate": 2.1898305084745764e-06,
-      "loss": 1.3533,
-      "step": 26770
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 7.439695835113525,
-      "learning_rate": 2.1830508474576273e-06,
-      "loss": 1.2436,
-      "step": 26780
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 6.698858737945557,
-      "learning_rate": 2.1762711864406783e-06,
-      "loss": 1.3145,
-      "step": 26790
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 4.412804126739502,
-      "learning_rate": 2.169491525423729e-06,
-      "loss": 1.185,
-      "step": 26800
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 9.170161247253418,
-      "learning_rate": 2.16271186440678e-06,
-      "loss": 1.1868,
-      "step": 26810
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 21.11345672607422,
-      "learning_rate": 2.1559322033898304e-06,
-      "loss": 1.2738,
-      "step": 26820
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 52.31527328491211,
-      "learning_rate": 2.1491525423728814e-06,
-      "loss": 1.1694,
-      "step": 26830
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 4.136544704437256,
-      "learning_rate": 2.1423728813559324e-06,
-      "loss": 1.0897,
-      "step": 26840
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 3.128509759902954,
-      "learning_rate": 2.1355932203389833e-06,
-      "loss": 1.1818,
-      "step": 26850
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 12.374292373657227,
-      "learning_rate": 2.128813559322034e-06,
-      "loss": 1.3163,
-      "step": 26860
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 8.347363471984863,
-      "learning_rate": 2.122033898305085e-06,
-      "loss": 1.1177,
-      "step": 26870
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 17.513408660888672,
-      "learning_rate": 2.115254237288136e-06,
-      "loss": 1.2128,
-      "step": 26880
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 5.053252220153809,
-      "learning_rate": 2.108474576271187e-06,
-      "loss": 1.4111,
-      "step": 26890
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 2.5421042442321777,
-      "learning_rate": 2.1016949152542374e-06,
-      "loss": 1.3909,
-      "step": 26900
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 14.27386474609375,
-      "learning_rate": 2.0949152542372883e-06,
-      "loss": 1.3056,
-      "step": 26910
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 3.5711681842803955,
-      "learning_rate": 2.0881355932203393e-06,
-      "loss": 1.1021,
-      "step": 26920
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 7.276656627655029,
-      "learning_rate": 2.08135593220339e-06,
-      "loss": 1.0432,
-      "step": 26930
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 2.4546101093292236,
-      "learning_rate": 2.074576271186441e-06,
-      "loss": 1.3391,
-      "step": 26940
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 6.01200008392334,
-      "learning_rate": 2.0677966101694914e-06,
-      "loss": 1.1382,
-      "step": 26950
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 6.168879508972168,
-      "learning_rate": 2.0610169491525424e-06,
-      "loss": 1.1858,
-      "step": 26960
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 8.381854057312012,
-      "learning_rate": 2.0542372881355934e-06,
-      "loss": 1.2046,
-      "step": 26970
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 3.6406259536743164,
-      "learning_rate": 2.0474576271186443e-06,
-      "loss": 1.2098,
-      "step": 26980
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 7.411496162414551,
-      "learning_rate": 2.0406779661016953e-06,
-      "loss": 1.2822,
-      "step": 26990
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 4.259427547454834,
-      "learning_rate": 2.033898305084746e-06,
-      "loss": 1.0367,
-      "step": 27000
-    },
-    {
-      "epoch": 0.63,
-      "eval_loss": 1.0261023044586182,
-      "eval_runtime": 66.6287,
-      "eval_samples_per_second": 15.009,
-      "eval_steps_per_second": 15.009,
-      "step": 27000
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 1.8957105875015259,
-      "learning_rate": 2.027118644067797e-06,
-      "loss": 1.1367,
-      "step": 27010
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 8.313997268676758,
-      "learning_rate": 2.020338983050848e-06,
-      "loss": 1.1379,
-      "step": 27020
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.363570690155029,
-      "learning_rate": 2.0135593220338984e-06,
-      "loss": 1.1295,
-      "step": 27030
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 15.122271537780762,
-      "learning_rate": 2.0067796610169494e-06,
-      "loss": 1.1451,
-      "step": 27040
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 4.8829755783081055,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 1.411,
-      "step": 27050
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 9.953533172607422,
-      "learning_rate": 1.993220338983051e-06,
-      "loss": 1.0443,
-      "step": 27060
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 6.28991174697876,
-      "learning_rate": 1.986440677966102e-06,
-      "loss": 1.0607,
-      "step": 27070
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 8.279200553894043,
-      "learning_rate": 1.9796610169491524e-06,
-      "loss": 1.3226,
-      "step": 27080
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 6.716332912445068,
-      "learning_rate": 1.9728813559322034e-06,
-      "loss": 1.1337,
-      "step": 27090
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 8.114790916442871,
-      "learning_rate": 1.9661016949152544e-06,
-      "loss": 1.1927,
-      "step": 27100
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 8.28928279876709,
-      "learning_rate": 1.9593220338983053e-06,
-      "loss": 1.2793,
-      "step": 27110
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.299505710601807,
-      "learning_rate": 1.9525423728813563e-06,
-      "loss": 1.1913,
-      "step": 27120
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.9343061447143555,
-      "learning_rate": 1.945762711864407e-06,
-      "loss": 1.162,
-      "step": 27130
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.81511926651001,
-      "learning_rate": 1.938983050847458e-06,
-      "loss": 1.3911,
-      "step": 27140
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 2.0106678009033203,
-      "learning_rate": 1.932203389830509e-06,
-      "loss": 1.3768,
-      "step": 27150
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.970021724700928,
-      "learning_rate": 1.9254237288135594e-06,
-      "loss": 1.2548,
-      "step": 27160
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 12.892632484436035,
-      "learning_rate": 1.9186440677966104e-06,
-      "loss": 1.3527,
-      "step": 27170
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 12.3954496383667,
-      "learning_rate": 1.9118644067796613e-06,
-      "loss": 1.2633,
-      "step": 27180
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 9.317601203918457,
-      "learning_rate": 1.9050847457627119e-06,
-      "loss": 1.1907,
-      "step": 27190
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 2.2695984840393066,
-      "learning_rate": 1.8983050847457629e-06,
-      "loss": 1.2513,
-      "step": 27200
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.8285908699035645,
-      "learning_rate": 1.8915254237288136e-06,
-      "loss": 1.2707,
-      "step": 27210
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 1.861101508140564,
-      "learning_rate": 1.8847457627118646e-06,
-      "loss": 0.9353,
-      "step": 27220
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.153651714324951,
-      "learning_rate": 1.8779661016949156e-06,
-      "loss": 1.2232,
-      "step": 27230
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 4.55015230178833,
-      "learning_rate": 1.8711864406779661e-06,
-      "loss": 1.2731,
-      "step": 27240
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 2.4303572177886963,
-      "learning_rate": 1.8644067796610171e-06,
-      "loss": 1.3679,
-      "step": 27250
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 7.1049017906188965,
-      "learning_rate": 1.857627118644068e-06,
-      "loss": 1.2542,
-      "step": 27260
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 7.2281036376953125,
-      "learning_rate": 1.8508474576271189e-06,
-      "loss": 1.197,
-      "step": 27270
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 7.088107109069824,
-      "learning_rate": 1.8440677966101696e-06,
-      "loss": 1.3486,
-      "step": 27280
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.903868675231934,
-      "learning_rate": 1.8372881355932204e-06,
-      "loss": 1.2879,
-      "step": 27290
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 1.4684103727340698,
-      "learning_rate": 1.8305084745762714e-06,
-      "loss": 1.2303,
-      "step": 27300
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 10.257400512695312,
-      "learning_rate": 1.8237288135593223e-06,
-      "loss": 1.2068,
-      "step": 27310
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 4.519632816314697,
-      "learning_rate": 1.816949152542373e-06,
-      "loss": 1.1828,
-      "step": 27320
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 2.953141689300537,
-      "learning_rate": 1.8101694915254239e-06,
-      "loss": 1.2411,
-      "step": 27330
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 3.631997585296631,
-      "learning_rate": 1.8033898305084746e-06,
-      "loss": 1.267,
-      "step": 27340
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 4.540164947509766,
-      "learning_rate": 1.7966101694915256e-06,
-      "loss": 1.1731,
-      "step": 27350
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 5.641757965087891,
-      "learning_rate": 1.7898305084745766e-06,
-      "loss": 1.1819,
-      "step": 27360
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 10.886664390563965,
-      "learning_rate": 1.7830508474576271e-06,
-      "loss": 1.1085,
-      "step": 27370
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 8.04320240020752,
-      "learning_rate": 1.7762711864406781e-06,
-      "loss": 1.2514,
-      "step": 27380
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 8.10466480255127,
-      "learning_rate": 1.769491525423729e-06,
-      "loss": 1.3978,
-      "step": 27390
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 8.447578430175781,
-      "learning_rate": 1.7627118644067799e-06,
-      "loss": 1.1422,
-      "step": 27400
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 15.72360610961914,
-      "learning_rate": 1.7559322033898306e-06,
-      "loss": 1.2554,
-      "step": 27410
-    },
-    {
-      "epoch": 0.63,
-      "grad_norm": 4.728666305541992,
-      "learning_rate": 1.7491525423728814e-06,
-      "loss": 1.0925,
-      "step": 27420
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 6.482224464416504,
-      "learning_rate": 1.7423728813559324e-06,
-      "loss": 1.2841,
-      "step": 27430
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 17.51344871520996,
-      "learning_rate": 1.7355932203389834e-06,
-      "loss": 1.2075,
-      "step": 27440
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 4.776230335235596,
-      "learning_rate": 1.728813559322034e-06,
-      "loss": 1.2672,
-      "step": 27450
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 11.152587890625,
-      "learning_rate": 1.7220338983050849e-06,
-      "loss": 1.3695,
-      "step": 27460
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 4.942652702331543,
-      "learning_rate": 1.7152542372881356e-06,
-      "loss": 1.4412,
-      "step": 27470
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 7.248074054718018,
-      "learning_rate": 1.7084745762711866e-06,
-      "loss": 1.2882,
-      "step": 27480
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 3.3746461868286133,
-      "learning_rate": 1.7016949152542376e-06,
-      "loss": 1.2192,
-      "step": 27490
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 0.75871741771698,
-      "learning_rate": 1.6949152542372882e-06,
-      "loss": 1.1612,
-      "step": 27500
-    },
-    {
-      "epoch": 0.64,
-      "eval_loss": 0.9824415445327759,
-      "eval_runtime": 66.6076,
-      "eval_samples_per_second": 15.013,
-      "eval_steps_per_second": 15.013,
-      "step": 27500
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 8.995759963989258,
-      "learning_rate": 1.6881355932203391e-06,
-      "loss": 1.1534,
-      "step": 27510
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 4.333745002746582,
-      "learning_rate": 1.6813559322033901e-06,
-      "loss": 1.3447,
-      "step": 27520
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 9.009721755981445,
-      "learning_rate": 1.6745762711864409e-06,
-      "loss": 1.102,
-      "step": 27530
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 1.8660717010498047,
-      "learning_rate": 1.6677966101694916e-06,
-      "loss": 1.3257,
-      "step": 27540
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 8.743804931640625,
-      "learning_rate": 1.6610169491525424e-06,
-      "loss": 1.3621,
-      "step": 27550
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 4.872500419616699,
-      "learning_rate": 1.6542372881355934e-06,
-      "loss": 1.348,
-      "step": 27560
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 6.630684852600098,
-      "learning_rate": 1.6474576271186444e-06,
-      "loss": 1.1863,
-      "step": 27570
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 3.588062286376953,
-      "learning_rate": 1.640677966101695e-06,
-      "loss": 1.1248,
-      "step": 27580
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 9.038145065307617,
-      "learning_rate": 1.6338983050847459e-06,
-      "loss": 1.2381,
-      "step": 27590
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 2.703775644302368,
-      "learning_rate": 1.6271186440677967e-06,
-      "loss": 1.2388,
-      "step": 27600
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 6.438141345977783,
-      "learning_rate": 1.6203389830508476e-06,
-      "loss": 1.4026,
-      "step": 27610
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 23.630895614624023,
-      "learning_rate": 1.6135593220338986e-06,
-      "loss": 1.2498,
-      "step": 27620
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 3.4313204288482666,
-      "learning_rate": 1.6067796610169492e-06,
-      "loss": 1.275,
-      "step": 27630
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 9.49851131439209,
-      "learning_rate": 1.6000000000000001e-06,
-      "loss": 1.2096,
-      "step": 27640
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 3.8425252437591553,
-      "learning_rate": 1.593220338983051e-06,
-      "loss": 1.225,
-      "step": 27650
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 4.10614538192749,
-      "learning_rate": 1.5864406779661019e-06,
-      "loss": 1.182,
-      "step": 27660
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 6.613473892211914,
-      "learning_rate": 1.5796610169491526e-06,
-      "loss": 1.316,
-      "step": 27670
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 3.006357431411743,
-      "learning_rate": 1.5728813559322034e-06,
-      "loss": 0.991,
-      "step": 27680
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 6.98958158493042,
-      "learning_rate": 1.5661016949152544e-06,
-      "loss": 1.2518,
-      "step": 27690
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 18.53824806213379,
-      "learning_rate": 1.5593220338983054e-06,
-      "loss": 1.2037,
-      "step": 27700
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 9.595442771911621,
-      "learning_rate": 1.552542372881356e-06,
-      "loss": 1.1781,
-      "step": 27710
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 8.525481224060059,
-      "learning_rate": 1.545762711864407e-06,
-      "loss": 1.0589,
-      "step": 27720
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 7.7480010986328125,
-      "learning_rate": 1.5389830508474577e-06,
-      "loss": 1.2321,
-      "step": 27730
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 4.403713703155518,
-      "learning_rate": 1.5322033898305086e-06,
-      "loss": 1.2394,
-      "step": 27740
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 8.966909408569336,
-      "learning_rate": 1.5254237288135596e-06,
-      "loss": 1.1507,
-      "step": 27750
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 3.3880155086517334,
-      "learning_rate": 1.5186440677966102e-06,
-      "loss": 1.3728,
-      "step": 27760
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 5.387197971343994,
-      "learning_rate": 1.5118644067796611e-06,
-      "loss": 1.3463,
-      "step": 27770
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 9.160344123840332,
-      "learning_rate": 1.505084745762712e-06,
-      "loss": 1.2807,
-      "step": 27780
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 9.067780494689941,
-      "learning_rate": 1.4983050847457629e-06,
-      "loss": 1.2772,
-      "step": 27790
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 11.345379829406738,
-      "learning_rate": 1.4915254237288139e-06,
-      "loss": 1.2251,
-      "step": 27800
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 1.177361249923706,
-      "learning_rate": 1.4847457627118644e-06,
-      "loss": 1.2393,
-      "step": 27810
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 10.291409492492676,
-      "learning_rate": 1.4779661016949154e-06,
-      "loss": 1.1844,
-      "step": 27820
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 3.2554609775543213,
-      "learning_rate": 1.4711864406779664e-06,
-      "loss": 1.4098,
-      "step": 27830
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 10.152715682983398,
-      "learning_rate": 1.464406779661017e-06,
-      "loss": 1.1902,
-      "step": 27840
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 7.060266494750977,
-      "learning_rate": 1.457627118644068e-06,
-      "loss": 1.255,
-      "step": 27850
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 1.495127558708191,
-      "learning_rate": 1.4508474576271187e-06,
-      "loss": 1.2665,
-      "step": 27860
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 4.548370838165283,
-      "learning_rate": 1.4440677966101696e-06,
-      "loss": 1.301,
-      "step": 27870
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 7.4621734619140625,
-      "learning_rate": 1.4372881355932206e-06,
-      "loss": 1.2617,
-      "step": 27880
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 7.851536273956299,
-      "learning_rate": 1.4305084745762712e-06,
-      "loss": 1.2416,
-      "step": 27890
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 8.992350578308105,
-      "learning_rate": 1.4237288135593222e-06,
-      "loss": 1.1459,
-      "step": 27900
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 4.336245059967041,
-      "learning_rate": 1.416949152542373e-06,
-      "loss": 1.4332,
-      "step": 27910
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 11.527148246765137,
-      "learning_rate": 1.410169491525424e-06,
-      "loss": 1.1412,
-      "step": 27920
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 3.6670548915863037,
-      "learning_rate": 1.4033898305084749e-06,
-      "loss": 1.3735,
-      "step": 27930
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 10.903753280639648,
-      "learning_rate": 1.3966101694915254e-06,
-      "loss": 1.2655,
-      "step": 27940
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 17.927316665649414,
-      "learning_rate": 1.3898305084745764e-06,
-      "loss": 1.2136,
-      "step": 27950
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 8.938633918762207,
-      "learning_rate": 1.3830508474576274e-06,
-      "loss": 1.4995,
-      "step": 27960
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 8.094979286193848,
-      "learning_rate": 1.376271186440678e-06,
-      "loss": 1.2659,
-      "step": 27970
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 9.596714973449707,
-      "learning_rate": 1.369491525423729e-06,
-      "loss": 1.1559,
-      "step": 27980
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 10.698532104492188,
-      "learning_rate": 1.3627118644067797e-06,
-      "loss": 1.3288,
-      "step": 27990
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 2.6723885536193848,
-      "learning_rate": 1.3559322033898307e-06,
-      "loss": 1.2338,
-      "step": 28000
-    },
-    {
-      "epoch": 0.65,
-      "eval_loss": 1.0027884244918823,
-      "eval_runtime": 67.3345,
-      "eval_samples_per_second": 14.851,
-      "eval_steps_per_second": 14.851,
-      "step": 28000
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 7.515227794647217,
-      "learning_rate": 1.3491525423728816e-06,
-      "loss": 1.365,
-      "step": 28010
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 3.3241326808929443,
-      "learning_rate": 1.3423728813559322e-06,
-      "loss": 1.2401,
-      "step": 28020
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 8.155590057373047,
-      "learning_rate": 1.3355932203389832e-06,
-      "loss": 1.3003,
-      "step": 28030
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 2.6986613273620605,
-      "learning_rate": 1.328813559322034e-06,
-      "loss": 1.3645,
-      "step": 28040
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 3.558117151260376,
-      "learning_rate": 1.322033898305085e-06,
-      "loss": 1.2466,
-      "step": 28050
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 2.6389589309692383,
-      "learning_rate": 1.3152542372881359e-06,
-      "loss": 1.4448,
-      "step": 28060
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 4.357456684112549,
-      "learning_rate": 1.3084745762711864e-06,
-      "loss": 1.0606,
-      "step": 28070
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 3.8627960681915283,
-      "learning_rate": 1.3016949152542374e-06,
-      "loss": 1.2686,
-      "step": 28080
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 6.192730903625488,
-      "learning_rate": 1.2949152542372884e-06,
-      "loss": 1.4787,
-      "step": 28090
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 4.1315131187438965,
-      "learning_rate": 1.288135593220339e-06,
-      "loss": 1.2422,
-      "step": 28100
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 13.30528450012207,
-      "learning_rate": 1.28135593220339e-06,
-      "loss": 1.2519,
-      "step": 28110
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 6.925759315490723,
-      "learning_rate": 1.2745762711864407e-06,
-      "loss": 1.388,
-      "step": 28120
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 5.491391181945801,
-      "learning_rate": 1.2677966101694917e-06,
-      "loss": 1.2848,
-      "step": 28130
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 6.054429054260254,
-      "learning_rate": 1.2610169491525426e-06,
-      "loss": 1.1429,
-      "step": 28140
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 13.854817390441895,
-      "learning_rate": 1.2542372881355932e-06,
-      "loss": 1.207,
-      "step": 28150
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 10.493091583251953,
-      "learning_rate": 1.2474576271186442e-06,
-      "loss": 1.0886,
-      "step": 28160
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 2.1718876361846924,
-      "learning_rate": 1.240677966101695e-06,
-      "loss": 1.2488,
-      "step": 28170
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 2.642390727996826,
-      "learning_rate": 1.233898305084746e-06,
-      "loss": 1.1109,
-      "step": 28180
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 7.8057332038879395,
-      "learning_rate": 1.2271186440677967e-06,
-      "loss": 1.121,
-      "step": 28190
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 4.232807636260986,
-      "learning_rate": 1.2203389830508477e-06,
-      "loss": 1.193,
-      "step": 28200
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 7.1039862632751465,
-      "learning_rate": 1.2135593220338984e-06,
-      "loss": 1.142,
-      "step": 28210
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 4.6590752601623535,
-      "learning_rate": 1.2067796610169492e-06,
-      "loss": 1.2522,
-      "step": 28220
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 6.117308139801025,
-      "learning_rate": 1.2000000000000002e-06,
-      "loss": 1.3453,
-      "step": 28230
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 6.451774597167969,
-      "learning_rate": 1.193220338983051e-06,
-      "loss": 1.297,
-      "step": 28240
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 4.05746603012085,
-      "learning_rate": 1.186440677966102e-06,
-      "loss": 1.2008,
-      "step": 28250
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 12.00547981262207,
-      "learning_rate": 1.1796610169491527e-06,
-      "loss": 1.1584,
-      "step": 28260
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 5.004361152648926,
-      "learning_rate": 1.1728813559322034e-06,
-      "loss": 1.1504,
-      "step": 28270
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 5.276272296905518,
-      "learning_rate": 1.1661016949152542e-06,
-      "loss": 1.1542,
-      "step": 28280
-    },
-    {
-      "epoch": 0.65,
-      "grad_norm": 6.736107349395752,
-      "learning_rate": 1.1593220338983052e-06,
-      "loss": 1.2236,
-      "step": 28290
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 3.8124780654907227,
-      "learning_rate": 1.152542372881356e-06,
-      "loss": 1.2236,
-      "step": 28300
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 7.842329025268555,
-      "learning_rate": 1.145762711864407e-06,
-      "loss": 1.3622,
-      "step": 28310
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 6.599586486816406,
-      "learning_rate": 1.1389830508474577e-06,
-      "loss": 1.1218,
-      "step": 28320
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 4.077427387237549,
-      "learning_rate": 1.1322033898305087e-06,
-      "loss": 1.2038,
-      "step": 28330
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 8.3840970993042,
-      "learning_rate": 1.1254237288135594e-06,
-      "loss": 1.3661,
-      "step": 28340
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 8.212132453918457,
-      "learning_rate": 1.1186440677966102e-06,
-      "loss": 1.1966,
-      "step": 28350
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 7.084817886352539,
-      "learning_rate": 1.1118644067796612e-06,
-      "loss": 1.2721,
-      "step": 28360
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 9.409709930419922,
-      "learning_rate": 1.105084745762712e-06,
-      "loss": 1.2941,
-      "step": 28370
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 4.141493320465088,
-      "learning_rate": 1.098305084745763e-06,
-      "loss": 1.4355,
-      "step": 28380
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 8.694832801818848,
-      "learning_rate": 1.0915254237288137e-06,
-      "loss": 1.2153,
-      "step": 28390
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 12.472442626953125,
-      "learning_rate": 1.0847457627118644e-06,
-      "loss": 1.1019,
-      "step": 28400
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 6.5215163230896,
-      "learning_rate": 1.0779661016949152e-06,
-      "loss": 1.0885,
-      "step": 28410
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 5.509222030639648,
-      "learning_rate": 1.0711864406779662e-06,
-      "loss": 1.2676,
-      "step": 28420
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 9.21153736114502,
-      "learning_rate": 1.064406779661017e-06,
-      "loss": 1.3133,
-      "step": 28430
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 5.743494987487793,
-      "learning_rate": 1.057627118644068e-06,
-      "loss": 1.292,
-      "step": 28440
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 8.526168823242188,
-      "learning_rate": 1.0508474576271187e-06,
-      "loss": 1.0585,
-      "step": 28450
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 3.5986156463623047,
-      "learning_rate": 1.0440677966101697e-06,
-      "loss": 1.1234,
-      "step": 28460
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 13.26816463470459,
-      "learning_rate": 1.0372881355932204e-06,
-      "loss": 1.1792,
-      "step": 28470
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 3.407280445098877,
-      "learning_rate": 1.0305084745762712e-06,
-      "loss": 1.1652,
-      "step": 28480
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 8.368290901184082,
-      "learning_rate": 1.0237288135593222e-06,
-      "loss": 1.3459,
-      "step": 28490
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 8.060779571533203,
-      "learning_rate": 1.016949152542373e-06,
-      "loss": 1.3085,
-      "step": 28500
-    },
-    {
-      "epoch": 0.66,
-      "eval_loss": 1.0383493900299072,
-      "eval_runtime": 68.471,
-      "eval_samples_per_second": 14.605,
-      "eval_steps_per_second": 14.605,
-      "step": 28500
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 2.879181385040283,
-      "learning_rate": 1.010169491525424e-06,
-      "loss": 1.2482,
-      "step": 28510
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 10.214177131652832,
-      "learning_rate": 1.0033898305084747e-06,
-      "loss": 1.2781,
-      "step": 28520
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 17.13050079345703,
-      "learning_rate": 9.966101694915254e-07,
-      "loss": 1.0837,
-      "step": 28530
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 15.841142654418945,
-      "learning_rate": 9.898305084745762e-07,
-      "loss": 1.2962,
-      "step": 28540
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 7.079030990600586,
-      "learning_rate": 9.830508474576272e-07,
-      "loss": 0.9777,
-      "step": 28550
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 3.9758317470550537,
-      "learning_rate": 9.762711864406782e-07,
-      "loss": 1.1584,
-      "step": 28560
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 4.3860392570495605,
-      "learning_rate": 9.69491525423729e-07,
-      "loss": 1.1247,
-      "step": 28570
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 3.759638786315918,
-      "learning_rate": 9.627118644067797e-07,
-      "loss": 1.2665,
-      "step": 28580
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 16.677642822265625,
-      "learning_rate": 9.559322033898307e-07,
-      "loss": 1.4559,
-      "step": 28590
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 7.994900703430176,
-      "learning_rate": 9.491525423728814e-07,
-      "loss": 1.0231,
-      "step": 28600
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 5.5230841636657715,
-      "learning_rate": 9.423728813559323e-07,
-      "loss": 1.1959,
-      "step": 28610
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 1.7889764308929443,
-      "learning_rate": 9.355932203389831e-07,
-      "loss": 1.3042,
-      "step": 28620
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 1.4875215291976929,
-      "learning_rate": 9.28813559322034e-07,
-      "loss": 1.1124,
-      "step": 28630
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 4.865635395050049,
-      "learning_rate": 9.220338983050848e-07,
-      "loss": 1.2639,
-      "step": 28640
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 3.1794333457946777,
-      "learning_rate": 9.152542372881357e-07,
-      "loss": 1.2244,
-      "step": 28650
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 9.40485954284668,
-      "learning_rate": 9.084745762711864e-07,
-      "loss": 1.244,
-      "step": 28660
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 8.51891803741455,
-      "learning_rate": 9.016949152542373e-07,
-      "loss": 1.2456,
-      "step": 28670
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 1.864579200744629,
-      "learning_rate": 8.949152542372883e-07,
-      "loss": 1.1113,
-      "step": 28680
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 12.539873123168945,
-      "learning_rate": 8.881355932203391e-07,
-      "loss": 1.1996,
-      "step": 28690
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 6.042193412780762,
-      "learning_rate": 8.813559322033899e-07,
-      "loss": 1.2614,
-      "step": 28700
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 5.701644420623779,
-      "learning_rate": 8.745762711864407e-07,
-      "loss": 1.1356,
-      "step": 28710
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 1.5281639099121094,
-      "learning_rate": 8.677966101694917e-07,
-      "loss": 1.244,
-      "step": 28720
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 11.545940399169922,
-      "learning_rate": 8.610169491525424e-07,
-      "loss": 1.2447,
-      "step": 28730
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 3.2043371200561523,
-      "learning_rate": 8.542372881355933e-07,
-      "loss": 1.2653,
-      "step": 28740
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 9.815803527832031,
-      "learning_rate": 8.474576271186441e-07,
-      "loss": 1.2609,
-      "step": 28750
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 2.5789248943328857,
-      "learning_rate": 8.406779661016951e-07,
-      "loss": 1.3839,
-      "step": 28760
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 6.396823883056641,
-      "learning_rate": 8.338983050847458e-07,
-      "loss": 1.2267,
-      "step": 28770
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 4.092123985290527,
-      "learning_rate": 8.271186440677967e-07,
-      "loss": 1.16,
-      "step": 28780
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 1.5519795417785645,
-      "learning_rate": 8.203389830508475e-07,
-      "loss": 1.1852,
-      "step": 28790
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 10.254398345947266,
-      "learning_rate": 8.135593220338983e-07,
-      "loss": 1.3109,
-      "step": 28800
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.28373384475708,
-      "learning_rate": 8.067796610169493e-07,
-      "loss": 1.2634,
-      "step": 28810
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.046334266662598,
-      "learning_rate": 8.000000000000001e-07,
-      "loss": 1.3029,
-      "step": 28820
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 7.71566104888916,
-      "learning_rate": 7.932203389830509e-07,
-      "loss": 1.1061,
-      "step": 28830
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 2.412494659423828,
-      "learning_rate": 7.864406779661017e-07,
-      "loss": 1.4086,
-      "step": 28840
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 6.458393096923828,
-      "learning_rate": 7.796610169491527e-07,
-      "loss": 1.238,
-      "step": 28850
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 4.167256832122803,
-      "learning_rate": 7.728813559322034e-07,
-      "loss": 1.24,
-      "step": 28860
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 7.2174177169799805,
-      "learning_rate": 7.661016949152543e-07,
-      "loss": 0.9788,
-      "step": 28870
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 8.519877433776855,
-      "learning_rate": 7.593220338983051e-07,
-      "loss": 1.3748,
-      "step": 28880
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.170273303985596,
-      "learning_rate": 7.52542372881356e-07,
-      "loss": 1.2798,
-      "step": 28890
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 9.433271408081055,
-      "learning_rate": 7.457627118644069e-07,
-      "loss": 1.2496,
-      "step": 28900
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 3.0811305046081543,
-      "learning_rate": 7.389830508474577e-07,
-      "loss": 1.3467,
-      "step": 28910
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 3.777050495147705,
-      "learning_rate": 7.322033898305085e-07,
-      "loss": 1.3679,
-      "step": 28920
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 2.2685747146606445,
-      "learning_rate": 7.254237288135593e-07,
-      "loss": 1.3076,
-      "step": 28930
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 6.493512153625488,
-      "learning_rate": 7.186440677966103e-07,
-      "loss": 1.1433,
-      "step": 28940
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 3.230098009109497,
-      "learning_rate": 7.118644067796611e-07,
-      "loss": 1.3191,
-      "step": 28950
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 1.8938180208206177,
-      "learning_rate": 7.05084745762712e-07,
-      "loss": 1.138,
-      "step": 28960
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 7.385470867156982,
-      "learning_rate": 6.983050847457627e-07,
-      "loss": 1.1831,
-      "step": 28970
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 4.667689800262451,
-      "learning_rate": 6.915254237288137e-07,
-      "loss": 1.3282,
-      "step": 28980
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 7.002562999725342,
-      "learning_rate": 6.847457627118645e-07,
-      "loss": 1.3384,
-      "step": 28990
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 7.729028701782227,
-      "learning_rate": 6.779661016949153e-07,
-      "loss": 1.2527,
-      "step": 29000
-    },
-    {
-      "epoch": 0.67,
-      "eval_loss": 1.036794900894165,
-      "eval_runtime": 67.0053,
-      "eval_samples_per_second": 14.924,
-      "eval_steps_per_second": 14.924,
-      "step": 29000
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.627978801727295,
-      "learning_rate": 6.711864406779661e-07,
-      "loss": 1.2501,
-      "step": 29010
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 18.444440841674805,
-      "learning_rate": 6.64406779661017e-07,
-      "loss": 1.0848,
-      "step": 29020
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.111781120300293,
-      "learning_rate": 6.576271186440679e-07,
-      "loss": 1.1205,
-      "step": 29030
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 10.785350799560547,
-      "learning_rate": 6.508474576271187e-07,
-      "loss": 1.0386,
-      "step": 29040
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.114529609680176,
-      "learning_rate": 6.440677966101695e-07,
-      "loss": 1.2195,
-      "step": 29050
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 9.179306983947754,
-      "learning_rate": 6.372881355932203e-07,
-      "loss": 1.2742,
-      "step": 29060
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 3.526697874069214,
-      "learning_rate": 6.305084745762713e-07,
-      "loss": 1.2442,
-      "step": 29070
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 2.499582052230835,
-      "learning_rate": 6.237288135593221e-07,
-      "loss": 1.1093,
-      "step": 29080
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 7.2823638916015625,
-      "learning_rate": 6.16949152542373e-07,
-      "loss": 1.3122,
-      "step": 29090
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 4.526163578033447,
-      "learning_rate": 6.101694915254238e-07,
-      "loss": 1.3623,
-      "step": 29100
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 3.685267686843872,
-      "learning_rate": 6.033898305084746e-07,
-      "loss": 1.0836,
-      "step": 29110
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.807446479797363,
-      "learning_rate": 5.966101694915255e-07,
-      "loss": 1.4172,
-      "step": 29120
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 7.100170135498047,
-      "learning_rate": 5.898305084745763e-07,
-      "loss": 1.424,
-      "step": 29130
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 2.293673276901245,
-      "learning_rate": 5.830508474576271e-07,
-      "loss": 1.2277,
-      "step": 29140
-    },
-    {
-      "epoch": 0.67,
-      "grad_norm": 5.1463165283203125,
-      "learning_rate": 5.76271186440678e-07,
-      "loss": 1.0603,
-      "step": 29150
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 2.3667409420013428,
-      "learning_rate": 5.694915254237288e-07,
-      "loss": 1.235,
-      "step": 29160
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 12.005767822265625,
-      "learning_rate": 5.627118644067797e-07,
-      "loss": 1.2028,
-      "step": 29170
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 6.225417613983154,
-      "learning_rate": 5.559322033898306e-07,
-      "loss": 1.0662,
-      "step": 29180
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.9254140853881836,
-      "learning_rate": 5.491525423728815e-07,
-      "loss": 1.209,
-      "step": 29190
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 2.4858005046844482,
-      "learning_rate": 5.423728813559322e-07,
-      "loss": 1.1661,
-      "step": 29200
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 5.9373884201049805,
-      "learning_rate": 5.355932203389831e-07,
-      "loss": 1.1282,
-      "step": 29210
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 6.60464334487915,
-      "learning_rate": 5.28813559322034e-07,
-      "loss": 1.3383,
-      "step": 29220
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 4.6242828369140625,
-      "learning_rate": 5.220338983050848e-07,
-      "loss": 1.3385,
-      "step": 29230
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 7.414794445037842,
-      "learning_rate": 5.152542372881356e-07,
-      "loss": 1.2059,
-      "step": 29240
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.5053508281707764,
-      "learning_rate": 5.084745762711865e-07,
-      "loss": 1.2981,
-      "step": 29250
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 5.528652191162109,
-      "learning_rate": 5.016949152542373e-07,
-      "loss": 1.2069,
-      "step": 29260
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 7.826475143432617,
-      "learning_rate": 4.949152542372881e-07,
-      "loss": 1.286,
-      "step": 29270
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 2.1104512214660645,
-      "learning_rate": 4.881355932203391e-07,
-      "loss": 1.0453,
-      "step": 29280
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.8154537677764893,
-      "learning_rate": 4.813559322033898e-07,
-      "loss": 1.265,
-      "step": 29290
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 8.516002655029297,
-      "learning_rate": 4.745762711864407e-07,
-      "loss": 1.2142,
-      "step": 29300
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.6356046199798584,
-      "learning_rate": 4.6779661016949154e-07,
-      "loss": 1.3932,
-      "step": 29310
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 5.065585613250732,
-      "learning_rate": 4.610169491525424e-07,
-      "loss": 1.1825,
-      "step": 29320
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.2396178245544434,
-      "learning_rate": 4.542372881355932e-07,
-      "loss": 1.269,
-      "step": 29330
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.066288471221924,
-      "learning_rate": 4.4745762711864415e-07,
-      "loss": 1.1986,
-      "step": 29340
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 8.957977294921875,
-      "learning_rate": 4.4067796610169497e-07,
-      "loss": 1.2721,
-      "step": 29350
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 6.242004871368408,
-      "learning_rate": 4.3389830508474584e-07,
-      "loss": 1.1724,
-      "step": 29360
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 10.986865997314453,
-      "learning_rate": 4.2711864406779666e-07,
-      "loss": 1.3121,
-      "step": 29370
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 7.176580905914307,
-      "learning_rate": 4.2033898305084753e-07,
-      "loss": 1.0941,
-      "step": 29380
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 8.999109268188477,
-      "learning_rate": 4.1355932203389835e-07,
-      "loss": 1.2917,
-      "step": 29390
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.9739575386047363,
-      "learning_rate": 4.0677966101694916e-07,
-      "loss": 1.1372,
-      "step": 29400
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 7.752053737640381,
-      "learning_rate": 4.0000000000000003e-07,
-      "loss": 1.229,
-      "step": 29410
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 15.617894172668457,
-      "learning_rate": 3.9322033898305085e-07,
-      "loss": 1.2178,
-      "step": 29420
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 5.103687763214111,
-      "learning_rate": 3.864406779661017e-07,
-      "loss": 1.1923,
-      "step": 29430
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 13.210460662841797,
-      "learning_rate": 3.7966101694915254e-07,
-      "loss": 1.3562,
-      "step": 29440
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 14.239701271057129,
-      "learning_rate": 3.7288135593220347e-07,
-      "loss": 1.2422,
-      "step": 29450
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 15.124138832092285,
-      "learning_rate": 3.6610169491525423e-07,
-      "loss": 1.3159,
-      "step": 29460
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 10.07345199584961,
-      "learning_rate": 3.5932203389830516e-07,
-      "loss": 1.0358,
-      "step": 29470
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 8.566622734069824,
-      "learning_rate": 3.52542372881356e-07,
-      "loss": 1.0562,
-      "step": 29480
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 13.473173141479492,
-      "learning_rate": 3.4576271186440684e-07,
-      "loss": 1.1538,
-      "step": 29490
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.2546353340148926,
-      "learning_rate": 3.3898305084745766e-07,
-      "loss": 1.3963,
-      "step": 29500
-    },
-    {
-      "epoch": 0.68,
-      "eval_loss": 0.9776638150215149,
-      "eval_runtime": 67.0335,
-      "eval_samples_per_second": 14.918,
-      "eval_steps_per_second": 14.918,
-      "step": 29500
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 8.819803237915039,
-      "learning_rate": 3.322033898305085e-07,
-      "loss": 1.3515,
-      "step": 29510
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 26.38185691833496,
-      "learning_rate": 3.2542372881355935e-07,
-      "loss": 1.1068,
-      "step": 29520
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 5.794632434844971,
-      "learning_rate": 3.1864406779661017e-07,
-      "loss": 1.2475,
-      "step": 29530
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 9.786145210266113,
-      "learning_rate": 3.1186440677966104e-07,
-      "loss": 1.2613,
-      "step": 29540
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 8.388969421386719,
-      "learning_rate": 3.050847457627119e-07,
-      "loss": 1.4074,
-      "step": 29550
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 3.0043601989746094,
-      "learning_rate": 2.9830508474576273e-07,
-      "loss": 1.4106,
-      "step": 29560
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 12.958208084106445,
-      "learning_rate": 2.9152542372881355e-07,
-      "loss": 1.2264,
-      "step": 29570
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 6.8235063552856445,
-      "learning_rate": 2.847457627118644e-07,
-      "loss": 1.1727,
-      "step": 29580
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 4.284834384918213,
-      "learning_rate": 2.779661016949153e-07,
-      "loss": 1.2897,
-      "step": 29590
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 3.338212251663208,
-      "learning_rate": 2.711864406779661e-07,
-      "loss": 1.2449,
-      "step": 29600
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 23.881393432617188,
-      "learning_rate": 2.64406779661017e-07,
-      "loss": 1.2986,
-      "step": 29610
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 9.715893745422363,
-      "learning_rate": 2.576271186440678e-07,
-      "loss": 1.2566,
-      "step": 29620
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 2.027327060699463,
-      "learning_rate": 2.5084745762711867e-07,
-      "loss": 1.2739,
-      "step": 29630
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 6.084362030029297,
-      "learning_rate": 2.4406779661016954e-07,
-      "loss": 1.1787,
-      "step": 29640
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 9.463132858276367,
-      "learning_rate": 2.3728813559322036e-07,
-      "loss": 1.1894,
-      "step": 29650
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 12.049840927124023,
-      "learning_rate": 2.305084745762712e-07,
-      "loss": 1.2903,
-      "step": 29660
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 4.152366638183594,
-      "learning_rate": 2.2372881355932207e-07,
-      "loss": 1.096,
-      "step": 29670
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 9.684791564941406,
-      "learning_rate": 2.1694915254237292e-07,
-      "loss": 1.2017,
-      "step": 29680
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 7.867280960083008,
-      "learning_rate": 2.1016949152542376e-07,
-      "loss": 1.1625,
-      "step": 29690
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 1.4374995231628418,
-      "learning_rate": 2.0338983050847458e-07,
-      "loss": 1.4949,
-      "step": 29700
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 4.6095428466796875,
-      "learning_rate": 1.9661016949152543e-07,
-      "loss": 1.3103,
-      "step": 29710
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 5.289913177490234,
-      "learning_rate": 1.8983050847457627e-07,
-      "loss": 1.3265,
-      "step": 29720
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 9.2072172164917,
-      "learning_rate": 1.8305084745762712e-07,
-      "loss": 1.1148,
-      "step": 29730
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 4.103106498718262,
-      "learning_rate": 1.76271186440678e-07,
-      "loss": 1.3842,
-      "step": 29740
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 15.820337295532227,
-      "learning_rate": 1.6949152542372883e-07,
-      "loss": 1.2755,
-      "step": 29750
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 4.733695030212402,
-      "learning_rate": 1.6271186440677968e-07,
-      "loss": 1.462,
-      "step": 29760
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 8.206982612609863,
-      "learning_rate": 1.5593220338983052e-07,
-      "loss": 1.235,
-      "step": 29770
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 7.3087663650512695,
-      "learning_rate": 1.4915254237288137e-07,
-      "loss": 1.1845,
-      "step": 29780
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 15.122031211853027,
-      "learning_rate": 1.423728813559322e-07,
-      "loss": 1.1609,
-      "step": 29790
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 5.786062717437744,
-      "learning_rate": 1.3559322033898305e-07,
-      "loss": 1.0769,
-      "step": 29800
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 1.0411622524261475,
-      "learning_rate": 1.288135593220339e-07,
-      "loss": 1.2727,
-      "step": 29810
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 3.1188580989837646,
-      "learning_rate": 1.2203389830508477e-07,
-      "loss": 1.2504,
-      "step": 29820
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 1.6318976879119873,
-      "learning_rate": 1.152542372881356e-07,
-      "loss": 1.2839,
-      "step": 29830
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 7.041867733001709,
-      "learning_rate": 1.0847457627118646e-07,
-      "loss": 1.0513,
-      "step": 29840
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 12.029023170471191,
-      "learning_rate": 1.0169491525423729e-07,
-      "loss": 1.132,
-      "step": 29850
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 5.014979839324951,
-      "learning_rate": 9.491525423728814e-08,
-      "loss": 1.2018,
-      "step": 29860
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 2.5008723735809326,
-      "learning_rate": 8.8135593220339e-08,
-      "loss": 1.2404,
-      "step": 29870
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 10.014387130737305,
-      "learning_rate": 8.135593220338984e-08,
-      "loss": 1.2046,
-      "step": 29880
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 12.765832901000977,
-      "learning_rate": 7.457627118644068e-08,
-      "loss": 1.0579,
-      "step": 29890
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 4.352108001708984,
-      "learning_rate": 6.779661016949153e-08,
-      "loss": 1.3139,
-      "step": 29900
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 10.630574226379395,
-      "learning_rate": 6.101694915254239e-08,
-      "loss": 1.3174,
-      "step": 29910
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 11.868256568908691,
-      "learning_rate": 5.423728813559323e-08,
-      "loss": 0.9875,
-      "step": 29920
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 3.4866442680358887,
-      "learning_rate": 4.745762711864407e-08,
-      "loss": 1.3073,
-      "step": 29930
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 9.97262191772461,
-      "learning_rate": 4.067796610169492e-08,
-      "loss": 1.276,
-      "step": 29940
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 3.1585495471954346,
-      "learning_rate": 3.3898305084745764e-08,
-      "loss": 1.3528,
-      "step": 29950
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 14.819723129272461,
-      "learning_rate": 2.7118644067796615e-08,
-      "loss": 1.1333,
-      "step": 29960
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 4.731696128845215,
-      "learning_rate": 2.033898305084746e-08,
-      "loss": 1.5296,
-      "step": 29970
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 9.690890312194824,
-      "learning_rate": 1.3559322033898307e-08,
-      "loss": 1.2169,
-      "step": 29980
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 3.7671804428100586,
-      "learning_rate": 6.779661016949154e-09,
-      "loss": 1.304,
-      "step": 29990
-    },
-    {
-      "epoch": 0.69,
-      "grad_norm": 13.029277801513672,
-      "learning_rate": 0.0,
-      "loss": 1.0771,
-      "step": 30000
-    },
-    {
-      "epoch": 0.69,
-      "eval_loss": 1.0159568786621094,
-      "eval_runtime": 67.116,
-      "eval_samples_per_second": 14.9,
-      "eval_steps_per_second": 14.9,
-      "step": 30000
     }
   ],
   "logging_steps": 10,
-  "max_steps": 30000,
+  "max_steps": 5000,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 2500,
-  "total_flos": 4.8306377981952e+17,
+  "total_flos": 4.025531498496e+16,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null