diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,758 +1,14647 @@
 {
-  "best_metric": 0.897123396396637,
-  "best_model_checkpoint": "./cocoa_outputs_resnet/checkpoint-980",
-  "epoch": 5.0,
+  "best_metric": 0.26566287875175476,
+  "best_model_checkpoint": "./cocoa_outputs_resnet/checkpoint-6468",
+  "epoch": 100.0,
   "eval_steps": 500,
-  "global_step": 980,
+  "global_step": 19600,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.05102040816326531,
-      "grad_norm": 9.393156051635742,
-      "learning_rate": 1.979591836734694e-05,
-      "loss": 1.8572,
+      "grad_norm": 4.57908821105957,
+      "learning_rate": 1.998979591836735e-05,
+      "loss": 1.7911,
       "step": 10
     },
     {
       "epoch": 0.10204081632653061,
-      "grad_norm": 6.8461503982543945,
-      "learning_rate": 1.9591836734693877e-05,
-      "loss": 1.8224,
+      "grad_norm": 4.909427165985107,
+      "learning_rate": 1.9979591836734697e-05,
+      "loss": 1.7576,
       "step": 20
     },
     {
       "epoch": 0.15306122448979592,
-      "grad_norm": 8.049448013305664,
-      "learning_rate": 1.9387755102040817e-05,
-      "loss": 1.7955,
+      "grad_norm": 4.646119117736816,
+      "learning_rate": 1.9969387755102042e-05,
+      "loss": 1.7171,
       "step": 30
     },
     {
       "epoch": 0.20408163265306123,
-      "grad_norm": 6.181807041168213,
-      "learning_rate": 1.9183673469387756e-05,
-      "loss": 1.7641,
+      "grad_norm": 4.798040866851807,
+      "learning_rate": 1.9959183673469388e-05,
+      "loss": 1.6832,
       "step": 40
     },
     {
       "epoch": 0.25510204081632654,
-      "grad_norm": 6.991772651672363,
-      "learning_rate": 1.8979591836734696e-05,
-      "loss": 1.7421,
+      "grad_norm": 4.766178131103516,
+      "learning_rate": 1.9948979591836737e-05,
+      "loss": 1.615,
       "step": 50
     },
     {
       "epoch": 0.30612244897959184,
-      "grad_norm": 6.233796119689941,
-      "learning_rate": 1.8775510204081636e-05,
-      "loss": 1.7018,
+      "grad_norm": 4.4977617263793945,
+      "learning_rate": 1.9938775510204083e-05,
+      "loss": 1.5784,
       "step": 60
     },
     {
       "epoch": 0.35714285714285715,
-      "grad_norm": 7.064742088317871,
-      "learning_rate": 1.8571428571428575e-05,
-      "loss": 1.6763,
+      "grad_norm": 4.778722286224365,
+      "learning_rate": 1.992857142857143e-05,
+      "loss": 1.5465,
       "step": 70
     },
     {
       "epoch": 0.40816326530612246,
-      "grad_norm": 7.047808647155762,
-      "learning_rate": 1.836734693877551e-05,
-      "loss": 1.654,
+      "grad_norm": 4.473388195037842,
+      "learning_rate": 1.9918367346938775e-05,
+      "loss": 1.506,
       "step": 80
     },
     {
       "epoch": 0.45918367346938777,
-      "grad_norm": 8.950271606445312,
-      "learning_rate": 1.816326530612245e-05,
-      "loss": 1.6461,
+      "grad_norm": 4.812692642211914,
+      "learning_rate": 1.9908163265306124e-05,
+      "loss": 1.4882,
       "step": 90
     },
     {
       "epoch": 0.5102040816326531,
-      "grad_norm": 7.686429023742676,
-      "learning_rate": 1.795918367346939e-05,
-      "loss": 1.6049,
+      "grad_norm": 4.0863494873046875,
+      "learning_rate": 1.9897959183673473e-05,
+      "loss": 1.4078,
       "step": 100
     },
     {
       "epoch": 0.5612244897959183,
-      "grad_norm": 6.434399127960205,
-      "learning_rate": 1.7755102040816327e-05,
-      "loss": 1.5926,
+      "grad_norm": 4.4336323738098145,
+      "learning_rate": 1.988775510204082e-05,
+      "loss": 1.404,
       "step": 110
     },
     {
       "epoch": 0.6122448979591837,
-      "grad_norm": 6.048853397369385,
-      "learning_rate": 1.7551020408163266e-05,
-      "loss": 1.5614,
+      "grad_norm": 4.0491719245910645,
+      "learning_rate": 1.9877551020408165e-05,
+      "loss": 1.34,
       "step": 120
     },
     {
       "epoch": 0.6632653061224489,
-      "grad_norm": 12.146856307983398,
-      "learning_rate": 1.7346938775510206e-05,
-      "loss": 1.5344,
+      "grad_norm": 4.225604057312012,
+      "learning_rate": 1.986734693877551e-05,
+      "loss": 1.2999,
       "step": 130
     },
     {
       "epoch": 0.7142857142857143,
-      "grad_norm": 8.939432144165039,
-      "learning_rate": 1.7142857142857142e-05,
-      "loss": 1.4892,
+      "grad_norm": 4.075715065002441,
+      "learning_rate": 1.985714285714286e-05,
+      "loss": 1.2529,
       "step": 140
     },
     {
       "epoch": 0.7653061224489796,
-      "grad_norm": 16.924053192138672,
-      "learning_rate": 1.6938775510204085e-05,
-      "loss": 1.4829,
+      "grad_norm": 4.563647747039795,
+      "learning_rate": 1.9846938775510205e-05,
+      "loss": 1.2366,
       "step": 150
     },
     {
       "epoch": 0.8163265306122449,
-      "grad_norm": 10.929306983947754,
-      "learning_rate": 1.673469387755102e-05,
-      "loss": 1.5136,
+      "grad_norm": 4.373153209686279,
+      "learning_rate": 1.983673469387755e-05,
+      "loss": 1.1994,
       "step": 160
     },
     {
       "epoch": 0.8673469387755102,
-      "grad_norm": 12.943446159362793,
-      "learning_rate": 1.653061224489796e-05,
-      "loss": 1.4316,
+      "grad_norm": 4.007843971252441,
+      "learning_rate": 1.9826530612244897e-05,
+      "loss": 1.149,
       "step": 170
     },
     {
       "epoch": 0.9183673469387755,
-      "grad_norm": 6.890258312225342,
-      "learning_rate": 1.63265306122449e-05,
-      "loss": 1.4373,
+      "grad_norm": 3.800847291946411,
+      "learning_rate": 1.9816326530612246e-05,
+      "loss": 1.137,
       "step": 180
     },
     {
       "epoch": 0.9693877551020408,
-      "grad_norm": 9.790729522705078,
-      "learning_rate": 1.612244897959184e-05,
-      "loss": 1.4094,
+      "grad_norm": 3.8902652263641357,
+      "learning_rate": 1.9806122448979595e-05,
+      "loss": 1.0627,
       "step": 190
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8592057761732852,
-      "eval_loss": 1.4767718315124512,
-      "eval_runtime": 14.5787,
-      "eval_samples_per_second": 19.0,
-      "eval_steps_per_second": 2.401,
+      "eval_accuracy": 0.5595667870036101,
+      "eval_loss": 1.5222723484039307,
+      "eval_runtime": 0.9755,
+      "eval_samples_per_second": 283.96,
+      "eval_steps_per_second": 35.879,
       "step": 196
     },
     {
       "epoch": 1.0204081632653061,
-      "grad_norm": 20.24932861328125,
-      "learning_rate": 1.5918367346938776e-05,
-      "loss": 1.4063,
+      "grad_norm": 5.571325778961182,
+      "learning_rate": 1.979591836734694e-05,
+      "loss": 1.0586,
       "step": 200
     },
     {
       "epoch": 1.0714285714285714,
-      "grad_norm": 14.811591148376465,
-      "learning_rate": 1.5714285714285715e-05,
-      "loss": 1.3551,
+      "grad_norm": 5.324435710906982,
+      "learning_rate": 1.9785714285714287e-05,
+      "loss": 1.009,
       "step": 210
     },
     {
       "epoch": 1.1224489795918366,
-      "grad_norm": 18.894763946533203,
-      "learning_rate": 1.5510204081632655e-05,
-      "loss": 1.3314,
+      "grad_norm": 4.557438850402832,
+      "learning_rate": 1.9775510204081633e-05,
+      "loss": 0.9619,
       "step": 220
     },
     {
       "epoch": 1.1734693877551021,
-      "grad_norm": 13.06529426574707,
-      "learning_rate": 1.530612244897959e-05,
-      "loss": 1.3112,
+      "grad_norm": 4.46217679977417,
+      "learning_rate": 1.9765306122448982e-05,
+      "loss": 0.9068,
       "step": 230
     },
     {
       "epoch": 1.2244897959183674,
-      "grad_norm": 16.423351287841797,
-      "learning_rate": 1.510204081632653e-05,
-      "loss": 1.2519,
+      "grad_norm": 3.9600579738616943,
+      "learning_rate": 1.9755102040816328e-05,
+      "loss": 0.8059,
       "step": 240
     },
     {
       "epoch": 1.2755102040816326,
-      "grad_norm": 11.016277313232422,
-      "learning_rate": 1.4897959183673472e-05,
-      "loss": 1.3565,
+      "grad_norm": 5.61907958984375,
+      "learning_rate": 1.9744897959183677e-05,
+      "loss": 0.9749,
       "step": 250
     },
     {
       "epoch": 1.3265306122448979,
-      "grad_norm": 7.271716594696045,
-      "learning_rate": 1.469387755102041e-05,
-      "loss": 1.2681,
+      "grad_norm": 3.644207239151001,
+      "learning_rate": 1.9734693877551023e-05,
+      "loss": 0.8702,
       "step": 260
     },
     {
       "epoch": 1.3775510204081631,
-      "grad_norm": 8.437969207763672,
-      "learning_rate": 1.448979591836735e-05,
-      "loss": 1.2675,
+      "grad_norm": 3.3476579189300537,
+      "learning_rate": 1.972448979591837e-05,
+      "loss": 0.8272,
       "step": 270
     },
     {
       "epoch": 1.4285714285714286,
-      "grad_norm": 6.352739334106445,
-      "learning_rate": 1.4285714285714287e-05,
-      "loss": 1.1948,
+      "grad_norm": 3.4033284187316895,
+      "learning_rate": 1.9714285714285718e-05,
+      "loss": 0.7383,
       "step": 280
     },
     {
       "epoch": 1.4795918367346939,
-      "grad_norm": 17.921051025390625,
-      "learning_rate": 1.4081632653061225e-05,
-      "loss": 1.3366,
+      "grad_norm": 5.9955034255981445,
+      "learning_rate": 1.9704081632653063e-05,
+      "loss": 0.9357,
       "step": 290
     },
     {
       "epoch": 1.5306122448979593,
-      "grad_norm": 22.707721710205078,
-      "learning_rate": 1.3877551020408165e-05,
-      "loss": 1.2684,
+      "grad_norm": 5.240922451019287,
+      "learning_rate": 1.969387755102041e-05,
+      "loss": 0.8436,
       "step": 300
     },
     {
       "epoch": 1.5816326530612246,
-      "grad_norm": 8.037663459777832,
-      "learning_rate": 1.3673469387755102e-05,
-      "loss": 1.2322,
+      "grad_norm": 2.973389148712158,
+      "learning_rate": 1.9683673469387755e-05,
+      "loss": 0.7505,
       "step": 310
     },
     {
       "epoch": 1.6326530612244898,
-      "grad_norm": 9.421684265136719,
-      "learning_rate": 1.3469387755102042e-05,
-      "loss": 1.1945,
+      "grad_norm": 4.997092247009277,
+      "learning_rate": 1.9673469387755104e-05,
+      "loss": 0.7463,
       "step": 320
     },
     {
       "epoch": 1.683673469387755,
-      "grad_norm": 27.85601043701172,
-      "learning_rate": 1.326530612244898e-05,
-      "loss": 1.2519,
+      "grad_norm": 5.728514671325684,
+      "learning_rate": 1.966326530612245e-05,
+      "loss": 0.8559,
       "step": 330
     },
     {
       "epoch": 1.7346938775510203,
-      "grad_norm": 15.291147232055664,
-      "learning_rate": 1.3061224489795918e-05,
-      "loss": 1.1934,
+      "grad_norm": 4.56866455078125,
+      "learning_rate": 1.96530612244898e-05,
+      "loss": 0.7695,
       "step": 340
     },
     {
       "epoch": 1.7857142857142856,
-      "grad_norm": 10.985498428344727,
-      "learning_rate": 1.2857142857142859e-05,
-      "loss": 1.123,
+      "grad_norm": 5.454906940460205,
+      "learning_rate": 1.9642857142857145e-05,
+      "loss": 0.6812,
       "step": 350
     },
     {
       "epoch": 1.836734693877551,
-      "grad_norm": 9.827385902404785,
-      "learning_rate": 1.2653061224489798e-05,
-      "loss": 1.1677,
+      "grad_norm": 3.0533580780029297,
+      "learning_rate": 1.963265306122449e-05,
+      "loss": 0.7244,
       "step": 360
     },
     {
       "epoch": 1.8877551020408163,
-      "grad_norm": 16.994632720947266,
-      "learning_rate": 1.2448979591836736e-05,
-      "loss": 1.171,
+      "grad_norm": 5.244307994842529,
+      "learning_rate": 1.962244897959184e-05,
+      "loss": 0.7089,
       "step": 370
     },
     {
       "epoch": 1.9387755102040818,
-      "grad_norm": 16.16167449951172,
-      "learning_rate": 1.2244897959183674e-05,
-      "loss": 1.1488,
+      "grad_norm": 2.4669547080993652,
+      "learning_rate": 1.9612244897959186e-05,
+      "loss": 0.7261,
       "step": 380
     },
     {
       "epoch": 1.989795918367347,
-      "grad_norm": 12.466065406799316,
-      "learning_rate": 1.2040816326530614e-05,
-      "loss": 1.0664,
+      "grad_norm": 3.281259059906006,
+      "learning_rate": 1.960204081632653e-05,
+      "loss": 0.591,
       "step": 390
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8628158844765343,
-      "eval_loss": 1.2089825868606567,
-      "eval_runtime": 17.1021,
-      "eval_samples_per_second": 16.197,
-      "eval_steps_per_second": 2.047,
+      "eval_accuracy": 0.8303249097472925,
+      "eval_loss": 0.8974758386611938,
+      "eval_runtime": 0.9365,
+      "eval_samples_per_second": 295.777,
+      "eval_steps_per_second": 37.373,
       "step": 392
     },
     {
       "epoch": 2.0408163265306123,
-      "grad_norm": 14.415572166442871,
-      "learning_rate": 1.1836734693877552e-05,
-      "loss": 1.1081,
+      "grad_norm": 3.226776361465454,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 0.5875,
       "step": 400
     },
     {
       "epoch": 2.0918367346938775,
-      "grad_norm": 16.120677947998047,
-      "learning_rate": 1.1632653061224491e-05,
-      "loss": 1.1347,
+      "grad_norm": 4.076236724853516,
+      "learning_rate": 1.9581632653061227e-05,
+      "loss": 0.5969,
       "step": 410
     },
     {
       "epoch": 2.142857142857143,
-      "grad_norm": 9.960640907287598,
-      "learning_rate": 1.1428571428571429e-05,
-      "loss": 1.0755,
+      "grad_norm": 6.054863929748535,
+      "learning_rate": 1.9571428571428572e-05,
+      "loss": 0.6501,
       "step": 420
     },
     {
       "epoch": 2.193877551020408,
-      "grad_norm": 14.734106063842773,
-      "learning_rate": 1.1224489795918367e-05,
-      "loss": 1.0976,
+      "grad_norm": 3.8678038120269775,
+      "learning_rate": 1.956122448979592e-05,
+      "loss": 0.602,
       "step": 430
     },
     {
       "epoch": 2.2448979591836733,
-      "grad_norm": 12.98324203491211,
-      "learning_rate": 1.1020408163265306e-05,
-      "loss": 0.9844,
+      "grad_norm": 3.923830032348633,
+      "learning_rate": 1.9551020408163267e-05,
+      "loss": 0.5332,
       "step": 440
     },
     {
       "epoch": 2.295918367346939,
-      "grad_norm": 8.834336280822754,
-      "learning_rate": 1.0816326530612246e-05,
-      "loss": 1.0923,
+      "grad_norm": 2.2238922119140625,
+      "learning_rate": 1.9540816326530613e-05,
+      "loss": 0.6505,
       "step": 450
     },
     {
       "epoch": 2.3469387755102042,
-      "grad_norm": 10.948126792907715,
-      "learning_rate": 1.0612244897959186e-05,
-      "loss": 0.9711,
+      "grad_norm": 3.5019683837890625,
+      "learning_rate": 1.9530612244897962e-05,
+      "loss": 0.5529,
       "step": 460
     },
     {
       "epoch": 2.3979591836734695,
-      "grad_norm": 13.707374572753906,
-      "learning_rate": 1.0408163265306123e-05,
-      "loss": 1.0786,
+      "grad_norm": 3.935518264770508,
+      "learning_rate": 1.9520408163265308e-05,
+      "loss": 0.6241,
       "step": 470
     },
     {
       "epoch": 2.4489795918367347,
-      "grad_norm": 18.081064224243164,
-      "learning_rate": 1.0204081632653063e-05,
-      "loss": 1.0399,
+      "grad_norm": 2.3772923946380615,
+      "learning_rate": 1.9510204081632654e-05,
+      "loss": 0.6224,
       "step": 480
     },
     {
       "epoch": 2.5,
-      "grad_norm": 12.470099449157715,
-      "learning_rate": 1e-05,
-      "loss": 0.9344,
+      "grad_norm": 6.30583381652832,
+      "learning_rate": 1.95e-05,
+      "loss": 0.4854,
       "step": 490
     },
     {
       "epoch": 2.5510204081632653,
-      "grad_norm": 10.96716022491455,
-      "learning_rate": 9.795918367346939e-06,
-      "loss": 1.043,
+      "grad_norm": 2.9505112171173096,
+      "learning_rate": 1.948979591836735e-05,
+      "loss": 0.6642,
       "step": 500
     },
     {
       "epoch": 2.6020408163265305,
-      "grad_norm": 7.244050025939941,
-      "learning_rate": 9.591836734693878e-06,
-      "loss": 0.9058,
+      "grad_norm": 2.7075178623199463,
+      "learning_rate": 1.9479591836734695e-05,
+      "loss": 0.5366,
       "step": 510
     },
     {
       "epoch": 2.6530612244897958,
-      "grad_norm": 10.213088035583496,
-      "learning_rate": 9.387755102040818e-06,
-      "loss": 0.9683,
+      "grad_norm": 2.4478352069854736,
+      "learning_rate": 1.9469387755102044e-05,
+      "loss": 0.52,
       "step": 520
     },
     {
       "epoch": 2.704081632653061,
-      "grad_norm": 30.410015106201172,
-      "learning_rate": 9.183673469387756e-06,
-      "loss": 1.0396,
+      "grad_norm": 6.208697319030762,
+      "learning_rate": 1.945918367346939e-05,
+      "loss": 0.6528,
       "step": 530
     },
     {
       "epoch": 2.7551020408163263,
-      "grad_norm": 10.457310676574707,
-      "learning_rate": 8.979591836734695e-06,
-      "loss": 1.0215,
+      "grad_norm": 5.765097141265869,
+      "learning_rate": 1.9448979591836735e-05,
+      "loss": 0.683,
       "step": 540
     },
     {
       "epoch": 2.806122448979592,
-      "grad_norm": 12.294684410095215,
-      "learning_rate": 8.775510204081633e-06,
-      "loss": 0.9301,
+      "grad_norm": 2.4912049770355225,
+      "learning_rate": 1.9438775510204085e-05,
+      "loss": 0.4997,
       "step": 550
     },
     {
       "epoch": 2.857142857142857,
-      "grad_norm": 10.194703102111816,
-      "learning_rate": 8.571428571428571e-06,
-      "loss": 0.9541,
+      "grad_norm": 3.521888494491577,
+      "learning_rate": 1.942857142857143e-05,
+      "loss": 0.5453,
       "step": 560
     },
     {
       "epoch": 2.9081632653061225,
-      "grad_norm": 17.80396842956543,
-      "learning_rate": 8.36734693877551e-06,
-      "loss": 0.964,
+      "grad_norm": 3.898709535598755,
+      "learning_rate": 1.941836734693878e-05,
+      "loss": 0.5242,
       "step": 570
     },
     {
       "epoch": 2.9591836734693877,
-      "grad_norm": 12.601577758789062,
-      "learning_rate": 8.16326530612245e-06,
-      "loss": 1.0295,
+      "grad_norm": 6.8352437019348145,
+      "learning_rate": 1.9408163265306122e-05,
+      "loss": 0.6623,
       "step": 580
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8628158844765343,
-      "eval_loss": 0.9923866987228394,
-      "eval_runtime": 14.8495,
-      "eval_samples_per_second": 18.654,
-      "eval_steps_per_second": 2.357,
+      "eval_accuracy": 0.8772563176895307,
+      "eval_loss": 0.6563544273376465,
+      "eval_runtime": 0.9345,
+      "eval_samples_per_second": 296.427,
+      "eval_steps_per_second": 37.455,
       "step": 588
     },
     {
       "epoch": 3.010204081632653,
-      "grad_norm": 12.689225196838379,
-      "learning_rate": 7.959183673469388e-06,
-      "loss": 0.9564,
+      "grad_norm": 6.536755084991455,
+      "learning_rate": 1.939795918367347e-05,
+      "loss": 0.5816,
       "step": 590
     },
     {
       "epoch": 3.061224489795918,
-      "grad_norm": 13.568258285522461,
-      "learning_rate": 7.755102040816327e-06,
-      "loss": 1.0031,
+      "grad_norm": 4.247134685516357,
+      "learning_rate": 1.9387755102040817e-05,
+      "loss": 0.6536,
       "step": 600
     },
     {
       "epoch": 3.1122448979591835,
-      "grad_norm": 15.86646842956543,
-      "learning_rate": 7.551020408163265e-06,
-      "loss": 0.9434,
+      "grad_norm": 5.93636417388916,
+      "learning_rate": 1.9377551020408166e-05,
+      "loss": 0.5808,
       "step": 610
     },
     {
       "epoch": 3.163265306122449,
-      "grad_norm": 16.88734245300293,
-      "learning_rate": 7.346938775510205e-06,
-      "loss": 0.9468,
+      "grad_norm": 3.8680365085601807,
+      "learning_rate": 1.9367346938775512e-05,
+      "loss": 0.5401,
       "step": 620
     },
     {
       "epoch": 3.2142857142857144,
-      "grad_norm": 14.894389152526855,
-      "learning_rate": 7.1428571428571436e-06,
-      "loss": 0.9284,
+      "grad_norm": 4.432985305786133,
+      "learning_rate": 1.9357142857142858e-05,
+      "loss": 0.5204,
       "step": 630
     },
     {
       "epoch": 3.2653061224489797,
-      "grad_norm": 13.203142166137695,
-      "learning_rate": 6.938775510204082e-06,
-      "loss": 0.9288,
+      "grad_norm": 6.848782062530518,
+      "learning_rate": 1.9346938775510207e-05,
+      "loss": 0.5037,
       "step": 640
     },
     {
       "epoch": 3.316326530612245,
-      "grad_norm": 13.70313835144043,
-      "learning_rate": 6.734693877551021e-06,
-      "loss": 0.8879,
+      "grad_norm": 3.865110158920288,
+      "learning_rate": 1.9336734693877553e-05,
+      "loss": 0.4718,
       "step": 650
     },
     {
       "epoch": 3.36734693877551,
-      "grad_norm": 22.619674682617188,
-      "learning_rate": 6.530612244897959e-06,
-      "loss": 0.9367,
+      "grad_norm": 7.135827541351318,
+      "learning_rate": 1.9326530612244902e-05,
+      "loss": 0.5719,
       "step": 660
     },
     {
       "epoch": 3.4183673469387754,
-      "grad_norm": 17.59436798095703,
-      "learning_rate": 6.326530612244899e-06,
-      "loss": 0.9176,
+      "grad_norm": 6.565048694610596,
+      "learning_rate": 1.9316326530612248e-05,
+      "loss": 0.5502,
       "step": 670
     },
     {
       "epoch": 3.4693877551020407,
-      "grad_norm": 12.488346099853516,
-      "learning_rate": 6.122448979591837e-06,
-      "loss": 0.9448,
+      "grad_norm": 3.097395420074463,
+      "learning_rate": 1.9306122448979593e-05,
+      "loss": 0.6063,
       "step": 680
     },
     {
       "epoch": 3.520408163265306,
-      "grad_norm": 7.665136337280273,
-      "learning_rate": 5.918367346938776e-06,
-      "loss": 0.8446,
+      "grad_norm": 2.708237886428833,
+      "learning_rate": 1.929591836734694e-05,
+      "loss": 0.4113,
       "step": 690
     },
     {
       "epoch": 3.571428571428571,
-      "grad_norm": 7.547515392303467,
-      "learning_rate": 5.7142857142857145e-06,
-      "loss": 0.8465,
+      "grad_norm": 3.084146738052368,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 0.4986,
       "step": 700
     },
     {
       "epoch": 3.622448979591837,
-      "grad_norm": 5.4680304527282715,
-      "learning_rate": 5.510204081632653e-06,
-      "loss": 0.8302,
+      "grad_norm": 1.5593074560165405,
+      "learning_rate": 1.9275510204081634e-05,
+      "loss": 0.465,
       "step": 710
     },
     {
       "epoch": 3.673469387755102,
-      "grad_norm": 22.36733055114746,
-      "learning_rate": 5.306122448979593e-06,
-      "loss": 0.9601,
+      "grad_norm": 3.1597211360931396,
+      "learning_rate": 1.926530612244898e-05,
+      "loss": 0.611,
       "step": 720
     },
     {
       "epoch": 3.7244897959183674,
-      "grad_norm": 12.001816749572754,
-      "learning_rate": 5.1020408163265315e-06,
-      "loss": 0.8835,
+      "grad_norm": 4.91085147857666,
+      "learning_rate": 1.925510204081633e-05,
+      "loss": 0.5452,
       "step": 730
     },
     {
       "epoch": 3.7755102040816326,
-      "grad_norm": 23.104984283447266,
-      "learning_rate": 4.897959183673469e-06,
-      "loss": 0.8024,
+      "grad_norm": 4.3785505294799805,
+      "learning_rate": 1.9244897959183675e-05,
+      "loss": 0.3833,
       "step": 740
     },
     {
       "epoch": 3.826530612244898,
-      "grad_norm": 13.406716346740723,
-      "learning_rate": 4.693877551020409e-06,
-      "loss": 1.0554,
+      "grad_norm": 1.9421954154968262,
+      "learning_rate": 1.9234693877551024e-05,
+      "loss": 0.7101,
       "step": 750
     },
     {
       "epoch": 3.877551020408163,
-      "grad_norm": 9.763223648071289,
-      "learning_rate": 4.489795918367348e-06,
-      "loss": 0.79,
+      "grad_norm": 2.2657201290130615,
+      "learning_rate": 1.922448979591837e-05,
+      "loss": 0.3817,
       "step": 760
     },
     {
       "epoch": 3.928571428571429,
-      "grad_norm": 10.93140983581543,
-      "learning_rate": 4.2857142857142855e-06,
-      "loss": 0.8532,
+      "grad_norm": 2.7672324180603027,
+      "learning_rate": 1.9214285714285716e-05,
+      "loss": 0.4854,
       "step": 770
     },
     {
       "epoch": 3.979591836734694,
-      "grad_norm": 10.841426849365234,
-      "learning_rate": 4.081632653061225e-06,
-      "loss": 0.8401,
+      "grad_norm": 2.3359146118164062,
+      "learning_rate": 1.920408163265306e-05,
+      "loss": 0.4874,
       "step": 780
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8628158844765343,
-      "eval_loss": 0.9142875075340271,
-      "eval_runtime": 14.5094,
-      "eval_samples_per_second": 19.091,
-      "eval_steps_per_second": 2.412,
+      "eval_accuracy": 0.8339350180505415,
+      "eval_loss": 0.6841569542884827,
+      "eval_runtime": 0.9488,
+      "eval_samples_per_second": 291.94,
+      "eval_steps_per_second": 36.888,
       "step": 784
     },
     {
       "epoch": 4.030612244897959,
-      "grad_norm": 11.179267883300781,
-      "learning_rate": 3.877551020408164e-06,
-      "loss": 0.8061,
+      "grad_norm": 2.9338905811309814,
+      "learning_rate": 1.919387755102041e-05,
+      "loss": 0.4634,
       "step": 790
     },
     {
       "epoch": 4.081632653061225,
-      "grad_norm": 6.5305585861206055,
-      "learning_rate": 3.6734693877551024e-06,
-      "loss": 0.8525,
+      "grad_norm": 1.7526339292526245,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 0.5248,
       "step": 800
     },
     {
       "epoch": 4.13265306122449,
-      "grad_norm": 7.101966857910156,
-      "learning_rate": 3.469387755102041e-06,
-      "loss": 0.7967,
+      "grad_norm": 2.1538479328155518,
+      "learning_rate": 1.9173469387755102e-05,
+      "loss": 0.3662,
       "step": 810
     },
     {
       "epoch": 4.183673469387755,
-      "grad_norm": 9.439040184020996,
-      "learning_rate": 3.2653061224489794e-06,
-      "loss": 0.8487,
+      "grad_norm": 4.566152572631836,
+      "learning_rate": 1.916326530612245e-05,
+      "loss": 0.483,
       "step": 820
     },
     {
       "epoch": 4.23469387755102,
-      "grad_norm": 16.658971786499023,
-      "learning_rate": 3.0612244897959185e-06,
-      "loss": 0.8698,
+      "grad_norm": 5.012964725494385,
+      "learning_rate": 1.9153061224489797e-05,
+      "loss": 0.5225,
       "step": 830
     },
     {
       "epoch": 4.285714285714286,
-      "grad_norm": 14.88980484008789,
-      "learning_rate": 2.8571428571428573e-06,
-      "loss": 0.7877,
+      "grad_norm": 1.5150026082992554,
+      "learning_rate": 1.9142857142857146e-05,
+      "loss": 0.389,
       "step": 840
     },
     {
       "epoch": 4.336734693877551,
-      "grad_norm": 10.878366470336914,
-      "learning_rate": 2.6530612244897964e-06,
-      "loss": 0.7731,
+      "grad_norm": 7.2560296058654785,
+      "learning_rate": 1.9132653061224492e-05,
+      "loss": 0.3931,
       "step": 850
     },
     {
       "epoch": 4.387755102040816,
-      "grad_norm": 11.978877067565918,
-      "learning_rate": 2.4489795918367347e-06,
-      "loss": 0.8627,
+      "grad_norm": 1.9941030740737915,
+      "learning_rate": 1.9122448979591838e-05,
+      "loss": 0.5159,
       "step": 860
     },
     {
       "epoch": 4.438775510204081,
-      "grad_norm": 20.543222427368164,
-      "learning_rate": 2.244897959183674e-06,
-      "loss": 0.8106,
+      "grad_norm": 4.368014335632324,
+      "learning_rate": 1.9112244897959184e-05,
+      "loss": 0.468,
       "step": 870
     },
     {
       "epoch": 4.489795918367347,
-      "grad_norm": 5.99076509475708,
-      "learning_rate": 2.0408163265306125e-06,
-      "loss": 0.8779,
+      "grad_norm": 2.1148953437805176,
+      "learning_rate": 1.9102040816326533e-05,
+      "loss": 0.5074,
       "step": 880
     },
     {
       "epoch": 4.540816326530612,
-      "grad_norm": 13.510481834411621,
-      "learning_rate": 1.8367346938775512e-06,
-      "loss": 0.8535,
+      "grad_norm": 8.519877433776855,
+      "learning_rate": 1.909183673469388e-05,
+      "loss": 0.5314,
       "step": 890
     },
     {
       "epoch": 4.591836734693878,
-      "grad_norm": 14.781784057617188,
-      "learning_rate": 1.6326530612244897e-06,
-      "loss": 0.8972,
+      "grad_norm": 2.9456112384796143,
+      "learning_rate": 1.9081632653061225e-05,
+      "loss": 0.534,
       "step": 900
     },
     {
       "epoch": 4.642857142857143,
-      "grad_norm": 14.63200855255127,
-      "learning_rate": 1.4285714285714286e-06,
-      "loss": 0.9312,
+      "grad_norm": 7.007885932922363,
+      "learning_rate": 1.9071428571428574e-05,
+      "loss": 0.6075,
       "step": 910
     },
     {
       "epoch": 4.6938775510204085,
-      "grad_norm": 19.910818099975586,
-      "learning_rate": 1.2244897959183673e-06,
-      "loss": 0.8134,
+      "grad_norm": 5.156570911407471,
+      "learning_rate": 1.906122448979592e-05,
+      "loss": 0.4616,
       "step": 920
     },
     {
       "epoch": 4.744897959183674,
-      "grad_norm": 16.637697219848633,
-      "learning_rate": 1.0204081632653063e-06,
-      "loss": 0.8341,
+      "grad_norm": 1.7120983600616455,
+      "learning_rate": 1.905102040816327e-05,
+      "loss": 0.4637,
       "step": 930
     },
     {
       "epoch": 4.795918367346939,
-      "grad_norm": 5.32245397567749,
-      "learning_rate": 8.163265306122449e-07,
-      "loss": 0.7553,
+      "grad_norm": 1.9243345260620117,
+      "learning_rate": 1.9040816326530614e-05,
+      "loss": 0.3998,
       "step": 940
     },
     {
       "epoch": 4.846938775510204,
-      "grad_norm": 13.930354118347168,
-      "learning_rate": 6.122448979591837e-07,
-      "loss": 0.8303,
+      "grad_norm": 5.795691013336182,
+      "learning_rate": 1.903061224489796e-05,
+      "loss": 0.4539,
       "step": 950
     },
     {
       "epoch": 4.8979591836734695,
-      "grad_norm": 18.44146728515625,
-      "learning_rate": 4.0816326530612243e-07,
-      "loss": 0.8247,
+      "grad_norm": 1.691705584526062,
+      "learning_rate": 1.9020408163265306e-05,
+      "loss": 0.4351,
       "step": 960
     },
     {
       "epoch": 4.948979591836735,
-      "grad_norm": 15.450093269348145,
-      "learning_rate": 2.0408163265306121e-07,
-      "loss": 0.9838,
+      "grad_norm": 6.1225199699401855,
+      "learning_rate": 1.9010204081632655e-05,
+      "loss": 0.7722,
       "step": 970
     },
     {
       "epoch": 5.0,
-      "grad_norm": 38.064430236816406,
-      "learning_rate": 0.0,
-      "loss": 0.8213,
+      "grad_norm": 8.150745391845703,
+      "learning_rate": 1.9e-05,
+      "loss": 0.4671,
       "step": 980
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8628158844765343,
-      "eval_loss": 0.897123396396637,
-      "eval_runtime": 14.3653,
-      "eval_samples_per_second": 19.283,
-      "eval_steps_per_second": 2.436,
+      "eval_accuracy": 0.8808664259927798,
+      "eval_loss": 0.48941850662231445,
+      "eval_runtime": 0.9622,
+      "eval_samples_per_second": 287.881,
+      "eval_steps_per_second": 36.375,
       "step": 980
     },
     {
-      "epoch": 5.0,
-      "step": 980,
-      "total_flos": 1.6633116935737344e+17,
-      "train_loss": 1.1215976812401596,
-      "train_runtime": 1562.8364,
-      "train_samples_per_second": 5.01,
-      "train_steps_per_second": 0.627
+      "epoch": 5.051020408163265,
+      "grad_norm": 6.351683139801025,
+      "learning_rate": 1.898979591836735e-05,
+      "loss": 0.3814,
+      "step": 990
+    },
+    {
+      "epoch": 5.1020408163265305,
+      "grad_norm": 3.669525384902954,
+      "learning_rate": 1.8979591836734696e-05,
+      "loss": 0.6377,
+      "step": 1000
+    },
+    {
+      "epoch": 5.153061224489796,
+      "grad_norm": 4.838572025299072,
+      "learning_rate": 1.8969387755102042e-05,
+      "loss": 0.283,
+      "step": 1010
+    },
+    {
+      "epoch": 5.204081632653061,
+      "grad_norm": 5.346072196960449,
+      "learning_rate": 1.895918367346939e-05,
+      "loss": 0.3222,
+      "step": 1020
+    },
+    {
+      "epoch": 5.255102040816326,
+      "grad_norm": 7.749392986297607,
+      "learning_rate": 1.8948979591836737e-05,
+      "loss": 0.5852,
+      "step": 1030
+    },
+    {
+      "epoch": 5.3061224489795915,
+      "grad_norm": 5.017237186431885,
+      "learning_rate": 1.8938775510204083e-05,
+      "loss": 0.4062,
+      "step": 1040
+    },
+    {
+      "epoch": 5.357142857142857,
+      "grad_norm": 3.6488564014434814,
+      "learning_rate": 1.892857142857143e-05,
+      "loss": 0.4372,
+      "step": 1050
+    },
+    {
+      "epoch": 5.408163265306122,
+      "grad_norm": 1.9676110744476318,
+      "learning_rate": 1.8918367346938778e-05,
+      "loss": 0.5442,
+      "step": 1060
+    },
+    {
+      "epoch": 5.459183673469388,
+      "grad_norm": 1.5495046377182007,
+      "learning_rate": 1.8908163265306123e-05,
+      "loss": 0.3801,
+      "step": 1070
+    },
+    {
+      "epoch": 5.510204081632653,
+      "grad_norm": 3.1108131408691406,
+      "learning_rate": 1.8897959183673473e-05,
+      "loss": 0.5089,
+      "step": 1080
+    },
+    {
+      "epoch": 5.561224489795919,
+      "grad_norm": 0.8886075615882874,
+      "learning_rate": 1.888775510204082e-05,
+      "loss": 0.6294,
+      "step": 1090
+    },
+    {
+      "epoch": 5.612244897959184,
+      "grad_norm": 2.3292956352233887,
+      "learning_rate": 1.8877551020408164e-05,
+      "loss": 0.7783,
+      "step": 1100
+    },
+    {
+      "epoch": 5.663265306122449,
+      "grad_norm": 6.512946128845215,
+      "learning_rate": 1.8867346938775513e-05,
+      "loss": 0.3822,
+      "step": 1110
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 1.3272405862808228,
+      "learning_rate": 1.885714285714286e-05,
+      "loss": 0.463,
+      "step": 1120
+    },
+    {
+      "epoch": 5.76530612244898,
+      "grad_norm": 5.868877410888672,
+      "learning_rate": 1.8846938775510205e-05,
+      "loss": 0.4467,
+      "step": 1130
+    },
+    {
+      "epoch": 5.816326530612245,
+      "grad_norm": 3.6009857654571533,
+      "learning_rate": 1.883673469387755e-05,
+      "loss": 0.4185,
+      "step": 1140
+    },
+    {
+      "epoch": 5.86734693877551,
+      "grad_norm": 4.642253875732422,
+      "learning_rate": 1.88265306122449e-05,
+      "loss": 0.4854,
+      "step": 1150
+    },
+    {
+      "epoch": 5.918367346938775,
+      "grad_norm": 3.2264819145202637,
+      "learning_rate": 1.8816326530612246e-05,
+      "loss": 0.3584,
+      "step": 1160
+    },
+    {
+      "epoch": 5.969387755102041,
+      "grad_norm": 2.7333457469940186,
+      "learning_rate": 1.8806122448979595e-05,
+      "loss": 0.5623,
+      "step": 1170
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8736462093862816,
+      "eval_loss": 0.41603580117225647,
+      "eval_runtime": 0.9437,
+      "eval_samples_per_second": 293.529,
+      "eval_steps_per_second": 37.088,
+      "step": 1176
+    },
+    {
+      "epoch": 6.020408163265306,
+      "grad_norm": 3.8172926902770996,
+      "learning_rate": 1.879591836734694e-05,
+      "loss": 0.5799,
+      "step": 1180
+    },
+    {
+      "epoch": 6.071428571428571,
+      "grad_norm": 5.174800395965576,
+      "learning_rate": 1.8785714285714286e-05,
+      "loss": 0.3616,
+      "step": 1190
+    },
+    {
+      "epoch": 6.122448979591836,
+      "grad_norm": 4.672287940979004,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 0.362,
+      "step": 1200
+    },
+    {
+      "epoch": 6.173469387755102,
+      "grad_norm": 1.3142999410629272,
+      "learning_rate": 1.876530612244898e-05,
+      "loss": 0.3637,
+      "step": 1210
+    },
+    {
+      "epoch": 6.224489795918367,
+      "grad_norm": 3.4073710441589355,
+      "learning_rate": 1.8755102040816327e-05,
+      "loss": 0.3755,
+      "step": 1220
+    },
+    {
+      "epoch": 6.275510204081632,
+      "grad_norm": 0.8577567338943481,
+      "learning_rate": 1.8744897959183673e-05,
+      "loss": 0.2254,
+      "step": 1230
+    },
+    {
+      "epoch": 6.326530612244898,
+      "grad_norm": 3.5044026374816895,
+      "learning_rate": 1.8734693877551022e-05,
+      "loss": 0.3727,
+      "step": 1240
+    },
+    {
+      "epoch": 6.377551020408164,
+      "grad_norm": 6.610156059265137,
+      "learning_rate": 1.8724489795918368e-05,
+      "loss": 0.8675,
+      "step": 1250
+    },
+    {
+      "epoch": 6.428571428571429,
+      "grad_norm": 2.974456787109375,
+      "learning_rate": 1.8714285714285717e-05,
+      "loss": 0.6604,
+      "step": 1260
+    },
+    {
+      "epoch": 6.479591836734694,
+      "grad_norm": 3.033073902130127,
+      "learning_rate": 1.8704081632653063e-05,
+      "loss": 0.3997,
+      "step": 1270
+    },
+    {
+      "epoch": 6.530612244897959,
+      "grad_norm": 4.493134498596191,
+      "learning_rate": 1.869387755102041e-05,
+      "loss": 0.4472,
+      "step": 1280
+    },
+    {
+      "epoch": 6.581632653061225,
+      "grad_norm": 2.4048378467559814,
+      "learning_rate": 1.8683673469387758e-05,
+      "loss": 0.2935,
+      "step": 1290
+    },
+    {
+      "epoch": 6.63265306122449,
+      "grad_norm": 4.755849838256836,
+      "learning_rate": 1.8673469387755104e-05,
+      "loss": 0.5393,
+      "step": 1300
+    },
+    {
+      "epoch": 6.683673469387755,
+      "grad_norm": 1.5519723892211914,
+      "learning_rate": 1.866326530612245e-05,
+      "loss": 0.4035,
+      "step": 1310
+    },
+    {
+      "epoch": 6.73469387755102,
+      "grad_norm": 2.7238337993621826,
+      "learning_rate": 1.8653061224489795e-05,
+      "loss": 0.367,
+      "step": 1320
+    },
+    {
+      "epoch": 6.785714285714286,
+      "grad_norm": 7.095693588256836,
+      "learning_rate": 1.8642857142857144e-05,
+      "loss": 0.5464,
+      "step": 1330
+    },
+    {
+      "epoch": 6.836734693877551,
+      "grad_norm": 1.5902338027954102,
+      "learning_rate": 1.863265306122449e-05,
+      "loss": 0.5423,
+      "step": 1340
+    },
+    {
+      "epoch": 6.887755102040816,
+      "grad_norm": 3.9222233295440674,
+      "learning_rate": 1.862244897959184e-05,
+      "loss": 0.3587,
+      "step": 1350
+    },
+    {
+      "epoch": 6.938775510204081,
+      "grad_norm": 6.730372905731201,
+      "learning_rate": 1.8612244897959185e-05,
+      "loss": 0.384,
+      "step": 1360
+    },
+    {
+      "epoch": 6.989795918367347,
+      "grad_norm": 3.2234292030334473,
+      "learning_rate": 1.860204081632653e-05,
+      "loss": 0.3917,
+      "step": 1370
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8844765342960289,
+      "eval_loss": 0.40221381187438965,
+      "eval_runtime": 0.994,
+      "eval_samples_per_second": 278.684,
+      "eval_steps_per_second": 35.213,
+      "step": 1372
+    },
+    {
+      "epoch": 7.040816326530612,
+      "grad_norm": 7.335348606109619,
+      "learning_rate": 1.859183673469388e-05,
+      "loss": 0.6036,
+      "step": 1380
+    },
+    {
+      "epoch": 7.091836734693878,
+      "grad_norm": 1.62227201461792,
+      "learning_rate": 1.8581632653061226e-05,
+      "loss": 0.1956,
+      "step": 1390
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 3.523392915725708,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 0.3493,
+      "step": 1400
+    },
+    {
+      "epoch": 7.1938775510204085,
+      "grad_norm": 9.984464645385742,
+      "learning_rate": 1.856122448979592e-05,
+      "loss": 0.6654,
+      "step": 1410
+    },
+    {
+      "epoch": 7.244897959183674,
+      "grad_norm": 3.143925189971924,
+      "learning_rate": 1.8551020408163267e-05,
+      "loss": 0.3959,
+      "step": 1420
+    },
+    {
+      "epoch": 7.295918367346939,
+      "grad_norm": 6.098402976989746,
+      "learning_rate": 1.8540816326530613e-05,
+      "loss": 0.6455,
+      "step": 1430
+    },
+    {
+      "epoch": 7.346938775510204,
+      "grad_norm": 3.7786874771118164,
+      "learning_rate": 1.853061224489796e-05,
+      "loss": 0.2466,
+      "step": 1440
+    },
+    {
+      "epoch": 7.3979591836734695,
+      "grad_norm": 5.968966484069824,
+      "learning_rate": 1.8520408163265307e-05,
+      "loss": 0.1962,
+      "step": 1450
+    },
+    {
+      "epoch": 7.448979591836735,
+      "grad_norm": 3.820188045501709,
+      "learning_rate": 1.8510204081632653e-05,
+      "loss": 0.5675,
+      "step": 1460
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 7.745204448699951,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 0.3934,
+      "step": 1470
+    },
+    {
+      "epoch": 7.551020408163265,
+      "grad_norm": 4.091207981109619,
+      "learning_rate": 1.8489795918367348e-05,
+      "loss": 0.4264,
+      "step": 1480
+    },
+    {
+      "epoch": 7.6020408163265305,
+      "grad_norm": 3.334214448928833,
+      "learning_rate": 1.8479591836734697e-05,
+      "loss": 0.3577,
+      "step": 1490
+    },
+    {
+      "epoch": 7.653061224489796,
+      "grad_norm": 1.8498873710632324,
+      "learning_rate": 1.8469387755102043e-05,
+      "loss": 0.6491,
+      "step": 1500
+    },
+    {
+      "epoch": 7.704081632653061,
+      "grad_norm": 2.124959945678711,
+      "learning_rate": 1.845918367346939e-05,
+      "loss": 0.3526,
+      "step": 1510
+    },
+    {
+      "epoch": 7.755102040816326,
+      "grad_norm": 5.348793983459473,
+      "learning_rate": 1.8448979591836735e-05,
+      "loss": 0.5161,
+      "step": 1520
+    },
+    {
+      "epoch": 7.8061224489795915,
+      "grad_norm": 5.554693222045898,
+      "learning_rate": 1.8438775510204084e-05,
+      "loss": 0.3795,
+      "step": 1530
+    },
+    {
+      "epoch": 7.857142857142857,
+      "grad_norm": 4.88253927230835,
+      "learning_rate": 1.842857142857143e-05,
+      "loss": 0.5021,
+      "step": 1540
+    },
+    {
+      "epoch": 7.908163265306122,
+      "grad_norm": 3.2645864486694336,
+      "learning_rate": 1.8418367346938776e-05,
+      "loss": 0.3677,
+      "step": 1550
+    },
+    {
+      "epoch": 7.959183673469388,
+      "grad_norm": 6.154125690460205,
+      "learning_rate": 1.8408163265306125e-05,
+      "loss": 0.3153,
+      "step": 1560
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8411552346570397,
+      "eval_loss": 0.4939042627811432,
+      "eval_runtime": 0.9307,
+      "eval_samples_per_second": 297.613,
+      "eval_steps_per_second": 37.605,
+      "step": 1568
+    },
+    {
+      "epoch": 8.010204081632653,
+      "grad_norm": 5.937544822692871,
+      "learning_rate": 1.839795918367347e-05,
+      "loss": 0.3933,
+      "step": 1570
+    },
+    {
+      "epoch": 8.061224489795919,
+      "grad_norm": 6.466835021972656,
+      "learning_rate": 1.838775510204082e-05,
+      "loss": 0.4626,
+      "step": 1580
+    },
+    {
+      "epoch": 8.112244897959183,
+      "grad_norm": 2.7466700077056885,
+      "learning_rate": 1.8377551020408165e-05,
+      "loss": 0.5429,
+      "step": 1590
+    },
+    {
+      "epoch": 8.16326530612245,
+      "grad_norm": 0.5636491179466248,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 0.4122,
+      "step": 1600
+    },
+    {
+      "epoch": 8.214285714285714,
+      "grad_norm": 2.0550692081451416,
+      "learning_rate": 1.8357142857142857e-05,
+      "loss": 0.3833,
+      "step": 1610
+    },
+    {
+      "epoch": 8.26530612244898,
+      "grad_norm": 1.9433763027191162,
+      "learning_rate": 1.8346938775510206e-05,
+      "loss": 0.3483,
+      "step": 1620
+    },
+    {
+      "epoch": 8.316326530612244,
+      "grad_norm": 1.361655831336975,
+      "learning_rate": 1.8336734693877552e-05,
+      "loss": 0.5256,
+      "step": 1630
+    },
+    {
+      "epoch": 8.36734693877551,
+      "grad_norm": 0.9611966013908386,
+      "learning_rate": 1.8326530612244898e-05,
+      "loss": 0.4249,
+      "step": 1640
+    },
+    {
+      "epoch": 8.418367346938776,
+      "grad_norm": 1.3870189189910889,
+      "learning_rate": 1.8316326530612247e-05,
+      "loss": 0.2141,
+      "step": 1650
+    },
+    {
+      "epoch": 8.46938775510204,
+      "grad_norm": 1.577960729598999,
+      "learning_rate": 1.8306122448979593e-05,
+      "loss": 0.5005,
+      "step": 1660
+    },
+    {
+      "epoch": 8.520408163265307,
+      "grad_norm": 3.716379404067993,
+      "learning_rate": 1.8295918367346942e-05,
+      "loss": 0.3459,
+      "step": 1670
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 5.005441188812256,
+      "learning_rate": 1.8285714285714288e-05,
+      "loss": 0.3062,
+      "step": 1680
+    },
+    {
+      "epoch": 8.622448979591837,
+      "grad_norm": 4.600987434387207,
+      "learning_rate": 1.8275510204081634e-05,
+      "loss": 0.3241,
+      "step": 1690
+    },
+    {
+      "epoch": 8.673469387755102,
+      "grad_norm": 5.137531757354736,
+      "learning_rate": 1.826530612244898e-05,
+      "loss": 0.5951,
+      "step": 1700
+    },
+    {
+      "epoch": 8.724489795918368,
+      "grad_norm": 2.3758299350738525,
+      "learning_rate": 1.825510204081633e-05,
+      "loss": 0.3967,
+      "step": 1710
+    },
+    {
+      "epoch": 8.775510204081632,
+      "grad_norm": 1.4290903806686401,
+      "learning_rate": 1.8244897959183674e-05,
+      "loss": 0.3694,
+      "step": 1720
+    },
+    {
+      "epoch": 8.826530612244898,
+      "grad_norm": 5.8770432472229,
+      "learning_rate": 1.823469387755102e-05,
+      "loss": 0.4959,
+      "step": 1730
+    },
+    {
+      "epoch": 8.877551020408163,
+      "grad_norm": 8.544288635253906,
+      "learning_rate": 1.822448979591837e-05,
+      "loss": 0.416,
+      "step": 1740
+    },
+    {
+      "epoch": 8.928571428571429,
+      "grad_norm": 0.7648389339447021,
+      "learning_rate": 1.8214285714285715e-05,
+      "loss": 0.3586,
+      "step": 1750
+    },
+    {
+      "epoch": 8.979591836734693,
+      "grad_norm": 7.747754096984863,
+      "learning_rate": 1.8204081632653064e-05,
+      "loss": 0.5814,
+      "step": 1760
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.8772563176895307,
+      "eval_loss": 0.35403916239738464,
+      "eval_runtime": 0.9416,
+      "eval_samples_per_second": 294.185,
+      "eval_steps_per_second": 37.171,
+      "step": 1764
+    },
+    {
+      "epoch": 9.03061224489796,
+      "grad_norm": 7.367336750030518,
+      "learning_rate": 1.819387755102041e-05,
+      "loss": 0.39,
+      "step": 1770
+    },
+    {
+      "epoch": 9.081632653061224,
+      "grad_norm": 5.418698310852051,
+      "learning_rate": 1.8183673469387756e-05,
+      "loss": 0.3316,
+      "step": 1780
+    },
+    {
+      "epoch": 9.13265306122449,
+      "grad_norm": 1.490496277809143,
+      "learning_rate": 1.81734693877551e-05,
+      "loss": 0.4636,
+      "step": 1790
+    },
+    {
+      "epoch": 9.183673469387756,
+      "grad_norm": 4.0167646408081055,
+      "learning_rate": 1.816326530612245e-05,
+      "loss": 0.2302,
+      "step": 1800
+    },
+    {
+      "epoch": 9.23469387755102,
+      "grad_norm": 6.253035545349121,
+      "learning_rate": 1.8153061224489797e-05,
+      "loss": 0.4014,
+      "step": 1810
+    },
+    {
+      "epoch": 9.285714285714286,
+      "grad_norm": 9.081564903259277,
+      "learning_rate": 1.8142857142857146e-05,
+      "loss": 0.4289,
+      "step": 1820
+    },
+    {
+      "epoch": 9.33673469387755,
+      "grad_norm": 3.3292837142944336,
+      "learning_rate": 1.813265306122449e-05,
+      "loss": 0.4435,
+      "step": 1830
+    },
+    {
+      "epoch": 9.387755102040817,
+      "grad_norm": 4.081367492675781,
+      "learning_rate": 1.8122448979591837e-05,
+      "loss": 0.6172,
+      "step": 1840
+    },
+    {
+      "epoch": 9.438775510204081,
+      "grad_norm": 4.028370380401611,
+      "learning_rate": 1.8112244897959187e-05,
+      "loss": 0.3676,
+      "step": 1850
+    },
+    {
+      "epoch": 9.489795918367347,
+      "grad_norm": 2.214043378829956,
+      "learning_rate": 1.8102040816326532e-05,
+      "loss": 0.2903,
+      "step": 1860
+    },
+    {
+      "epoch": 9.540816326530612,
+      "grad_norm": 7.913087844848633,
+      "learning_rate": 1.8091836734693878e-05,
+      "loss": 0.5894,
+      "step": 1870
+    },
+    {
+      "epoch": 9.591836734693878,
+      "grad_norm": 5.671935558319092,
+      "learning_rate": 1.8081632653061224e-05,
+      "loss": 0.2988,
+      "step": 1880
+    },
+    {
+      "epoch": 9.642857142857142,
+      "grad_norm": 5.60788631439209,
+      "learning_rate": 1.8071428571428573e-05,
+      "loss": 0.462,
+      "step": 1890
+    },
+    {
+      "epoch": 9.693877551020408,
+      "grad_norm": 2.223788261413574,
+      "learning_rate": 1.806122448979592e-05,
+      "loss": 0.4142,
+      "step": 1900
+    },
+    {
+      "epoch": 9.744897959183673,
+      "grad_norm": 2.553497791290283,
+      "learning_rate": 1.8051020408163268e-05,
+      "loss": 0.3389,
+      "step": 1910
+    },
+    {
+      "epoch": 9.795918367346939,
+      "grad_norm": 2.754664659500122,
+      "learning_rate": 1.8040816326530614e-05,
+      "loss": 0.3824,
+      "step": 1920
+    },
+    {
+      "epoch": 9.846938775510203,
+      "grad_norm": 0.7434467077255249,
+      "learning_rate": 1.803061224489796e-05,
+      "loss": 0.4129,
+      "step": 1930
+    },
+    {
+      "epoch": 9.89795918367347,
+      "grad_norm": 0.7156695127487183,
+      "learning_rate": 1.802040816326531e-05,
+      "loss": 0.3073,
+      "step": 1940
+    },
+    {
+      "epoch": 9.948979591836736,
+      "grad_norm": 2.1877171993255615,
+      "learning_rate": 1.8010204081632655e-05,
+      "loss": 0.3985,
+      "step": 1950
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 4.8080620765686035,
+      "learning_rate": 1.8e-05,
+      "loss": 0.5883,
+      "step": 1960
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.8953068592057761,
+      "eval_loss": 0.349290668964386,
+      "eval_runtime": 0.9968,
+      "eval_samples_per_second": 277.89,
+      "eval_steps_per_second": 35.113,
+      "step": 1960
+    },
+    {
+      "epoch": 10.051020408163266,
+      "grad_norm": 0.7896466255187988,
+      "learning_rate": 1.7989795918367346e-05,
+      "loss": 0.5032,
+      "step": 1970
+    },
+    {
+      "epoch": 10.10204081632653,
+      "grad_norm": 0.5900990962982178,
+      "learning_rate": 1.7979591836734695e-05,
+      "loss": 0.3054,
+      "step": 1980
+    },
+    {
+      "epoch": 10.153061224489797,
+      "grad_norm": 2.6685738563537598,
+      "learning_rate": 1.796938775510204e-05,
+      "loss": 0.4091,
+      "step": 1990
+    },
+    {
+      "epoch": 10.204081632653061,
+      "grad_norm": 1.8895847797393799,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 0.337,
+      "step": 2000
+    },
+    {
+      "epoch": 10.255102040816327,
+      "grad_norm": 5.3162665367126465,
+      "learning_rate": 1.7948979591836736e-05,
+      "loss": 0.3889,
+      "step": 2010
+    },
+    {
+      "epoch": 10.306122448979592,
+      "grad_norm": 4.576247215270996,
+      "learning_rate": 1.7938775510204082e-05,
+      "loss": 0.4455,
+      "step": 2020
+    },
+    {
+      "epoch": 10.357142857142858,
+      "grad_norm": 1.8315520286560059,
+      "learning_rate": 1.792857142857143e-05,
+      "loss": 0.2955,
+      "step": 2030
+    },
+    {
+      "epoch": 10.408163265306122,
+      "grad_norm": 6.811487197875977,
+      "learning_rate": 1.7918367346938777e-05,
+      "loss": 0.4975,
+      "step": 2040
+    },
+    {
+      "epoch": 10.459183673469388,
+      "grad_norm": 6.817623615264893,
+      "learning_rate": 1.7908163265306123e-05,
+      "loss": 0.5212,
+      "step": 2050
+    },
+    {
+      "epoch": 10.510204081632653,
+      "grad_norm": 2.7684004306793213,
+      "learning_rate": 1.789795918367347e-05,
+      "loss": 0.3925,
+      "step": 2060
+    },
+    {
+      "epoch": 10.561224489795919,
+      "grad_norm": 1.4127514362335205,
+      "learning_rate": 1.7887755102040818e-05,
+      "loss": 0.4752,
+      "step": 2070
+    },
+    {
+      "epoch": 10.612244897959183,
+      "grad_norm": 6.725620269775391,
+      "learning_rate": 1.7877551020408164e-05,
+      "loss": 0.3967,
+      "step": 2080
+    },
+    {
+      "epoch": 10.66326530612245,
+      "grad_norm": 3.6214230060577393,
+      "learning_rate": 1.7867346938775513e-05,
+      "loss": 0.6277,
+      "step": 2090
+    },
+    {
+      "epoch": 10.714285714285714,
+      "grad_norm": 1.93522047996521,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.214,
+      "step": 2100
+    },
+    {
+      "epoch": 10.76530612244898,
+      "grad_norm": 5.977639198303223,
+      "learning_rate": 1.7846938775510204e-05,
+      "loss": 0.3622,
+      "step": 2110
+    },
+    {
+      "epoch": 10.816326530612244,
+      "grad_norm": 1.9556264877319336,
+      "learning_rate": 1.7836734693877553e-05,
+      "loss": 0.4324,
+      "step": 2120
+    },
+    {
+      "epoch": 10.86734693877551,
+      "grad_norm": 7.346195697784424,
+      "learning_rate": 1.78265306122449e-05,
+      "loss": 0.2873,
+      "step": 2130
+    },
+    {
+      "epoch": 10.918367346938776,
+      "grad_norm": 6.110283374786377,
+      "learning_rate": 1.781632653061225e-05,
+      "loss": 0.5476,
+      "step": 2140
+    },
+    {
+      "epoch": 10.96938775510204,
+      "grad_norm": 10.262624740600586,
+      "learning_rate": 1.780612244897959e-05,
+      "loss": 0.4616,
+      "step": 2150
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.776173285198556,
+      "eval_loss": 0.7928438186645508,
+      "eval_runtime": 0.959,
+      "eval_samples_per_second": 288.836,
+      "eval_steps_per_second": 36.495,
+      "step": 2156
+    },
+    {
+      "epoch": 11.020408163265307,
+      "grad_norm": 1.9167275428771973,
+      "learning_rate": 1.779591836734694e-05,
+      "loss": 0.3289,
+      "step": 2160
+    },
+    {
+      "epoch": 11.071428571428571,
+      "grad_norm": 1.6276081800460815,
+      "learning_rate": 1.7785714285714286e-05,
+      "loss": 0.3506,
+      "step": 2170
+    },
+    {
+      "epoch": 11.122448979591837,
+      "grad_norm": 6.090387344360352,
+      "learning_rate": 1.7775510204081635e-05,
+      "loss": 0.3149,
+      "step": 2180
+    },
+    {
+      "epoch": 11.173469387755102,
+      "grad_norm": 1.20468270778656,
+      "learning_rate": 1.776530612244898e-05,
+      "loss": 0.2804,
+      "step": 2190
+    },
+    {
+      "epoch": 11.224489795918368,
+      "grad_norm": 9.119256019592285,
+      "learning_rate": 1.7755102040816327e-05,
+      "loss": 0.3767,
+      "step": 2200
+    },
+    {
+      "epoch": 11.275510204081632,
+      "grad_norm": 3.9283931255340576,
+      "learning_rate": 1.7744897959183676e-05,
+      "loss": 0.5942,
+      "step": 2210
+    },
+    {
+      "epoch": 11.326530612244898,
+      "grad_norm": 1.8245526552200317,
+      "learning_rate": 1.773469387755102e-05,
+      "loss": 0.3046,
+      "step": 2220
+    },
+    {
+      "epoch": 11.377551020408163,
+      "grad_norm": 4.436433792114258,
+      "learning_rate": 1.772448979591837e-05,
+      "loss": 0.4403,
+      "step": 2230
+    },
+    {
+      "epoch": 11.428571428571429,
+      "grad_norm": 5.837812423706055,
+      "learning_rate": 1.7714285714285717e-05,
+      "loss": 0.4211,
+      "step": 2240
+    },
+    {
+      "epoch": 11.479591836734693,
+      "grad_norm": 1.7079073190689087,
+      "learning_rate": 1.7704081632653062e-05,
+      "loss": 0.5829,
+      "step": 2250
+    },
+    {
+      "epoch": 11.53061224489796,
+      "grad_norm": 0.8393904566764832,
+      "learning_rate": 1.7693877551020408e-05,
+      "loss": 0.211,
+      "step": 2260
+    },
+    {
+      "epoch": 11.581632653061224,
+      "grad_norm": 2.068840742111206,
+      "learning_rate": 1.7683673469387757e-05,
+      "loss": 0.2437,
+      "step": 2270
+    },
+    {
+      "epoch": 11.63265306122449,
+      "grad_norm": 2.982731819152832,
+      "learning_rate": 1.7673469387755103e-05,
+      "loss": 0.2618,
+      "step": 2280
+    },
+    {
+      "epoch": 11.683673469387756,
+      "grad_norm": 4.23781156539917,
+      "learning_rate": 1.766326530612245e-05,
+      "loss": 0.3905,
+      "step": 2290
+    },
+    {
+      "epoch": 11.73469387755102,
+      "grad_norm": 5.205179214477539,
+      "learning_rate": 1.7653061224489798e-05,
+      "loss": 0.7493,
+      "step": 2300
+    },
+    {
+      "epoch": 11.785714285714286,
+      "grad_norm": 1.835451364517212,
+      "learning_rate": 1.7642857142857144e-05,
+      "loss": 0.3456,
+      "step": 2310
+    },
+    {
+      "epoch": 11.83673469387755,
+      "grad_norm": 4.031843662261963,
+      "learning_rate": 1.7632653061224493e-05,
+      "loss": 0.3092,
+      "step": 2320
+    },
+    {
+      "epoch": 11.887755102040817,
+      "grad_norm": 5.397056579589844,
+      "learning_rate": 1.762244897959184e-05,
+      "loss": 0.3759,
+      "step": 2330
+    },
+    {
+      "epoch": 11.938775510204081,
+      "grad_norm": 2.7502498626708984,
+      "learning_rate": 1.7612244897959185e-05,
+      "loss": 0.2369,
+      "step": 2340
+    },
+    {
+      "epoch": 11.989795918367347,
+      "grad_norm": 4.946924209594727,
+      "learning_rate": 1.760204081632653e-05,
+      "loss": 0.499,
+      "step": 2350
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.296028880866426,
+      "eval_loss": 2.06593918800354,
+      "eval_runtime": 0.9333,
+      "eval_samples_per_second": 296.781,
+      "eval_steps_per_second": 37.499,
+      "step": 2352
+    },
+    {
+      "epoch": 12.040816326530612,
+      "grad_norm": 3.4574103355407715,
+      "learning_rate": 1.759183673469388e-05,
+      "loss": 0.2957,
+      "step": 2360
+    },
+    {
+      "epoch": 12.091836734693878,
+      "grad_norm": 8.112685203552246,
+      "learning_rate": 1.7581632653061225e-05,
+      "loss": 0.3105,
+      "step": 2370
+    },
+    {
+      "epoch": 12.142857142857142,
+      "grad_norm": 4.740105628967285,
+      "learning_rate": 1.757142857142857e-05,
+      "loss": 0.4352,
+      "step": 2380
+    },
+    {
+      "epoch": 12.193877551020408,
+      "grad_norm": 1.1439720392227173,
+      "learning_rate": 1.756122448979592e-05,
+      "loss": 0.397,
+      "step": 2390
+    },
+    {
+      "epoch": 12.244897959183673,
+      "grad_norm": 0.700265645980835,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 0.3186,
+      "step": 2400
+    },
+    {
+      "epoch": 12.295918367346939,
+      "grad_norm": 4.206295013427734,
+      "learning_rate": 1.7540816326530615e-05,
+      "loss": 0.5838,
+      "step": 2410
+    },
+    {
+      "epoch": 12.346938775510203,
+      "grad_norm": 3.2011196613311768,
+      "learning_rate": 1.753061224489796e-05,
+      "loss": 0.4628,
+      "step": 2420
+    },
+    {
+      "epoch": 12.39795918367347,
+      "grad_norm": 6.6271281242370605,
+      "learning_rate": 1.7520408163265307e-05,
+      "loss": 0.3546,
+      "step": 2430
+    },
+    {
+      "epoch": 12.448979591836734,
+      "grad_norm": 2.038639783859253,
+      "learning_rate": 1.7510204081632653e-05,
+      "loss": 0.2915,
+      "step": 2440
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 1.8398466110229492,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.3023,
+      "step": 2450
+    },
+    {
+      "epoch": 12.551020408163264,
+      "grad_norm": 4.673349857330322,
+      "learning_rate": 1.748979591836735e-05,
+      "loss": 0.4503,
+      "step": 2460
+    },
+    {
+      "epoch": 12.60204081632653,
+      "grad_norm": 1.638249397277832,
+      "learning_rate": 1.7479591836734693e-05,
+      "loss": 0.3084,
+      "step": 2470
+    },
+    {
+      "epoch": 12.653061224489797,
+      "grad_norm": 2.2692887783050537,
+      "learning_rate": 1.7469387755102043e-05,
+      "loss": 0.3704,
+      "step": 2480
+    },
+    {
+      "epoch": 12.704081632653061,
+      "grad_norm": 8.47905445098877,
+      "learning_rate": 1.745918367346939e-05,
+      "loss": 0.3022,
+      "step": 2490
+    },
+    {
+      "epoch": 12.755102040816327,
+      "grad_norm": 7.4197492599487305,
+      "learning_rate": 1.7448979591836738e-05,
+      "loss": 0.5395,
+      "step": 2500
+    },
+    {
+      "epoch": 12.806122448979592,
+      "grad_norm": 3.075881242752075,
+      "learning_rate": 1.7438775510204083e-05,
+      "loss": 0.316,
+      "step": 2510
+    },
+    {
+      "epoch": 12.857142857142858,
+      "grad_norm": 3.121246814727783,
+      "learning_rate": 1.742857142857143e-05,
+      "loss": 0.5723,
+      "step": 2520
+    },
+    {
+      "epoch": 12.908163265306122,
+      "grad_norm": 1.1997493505477905,
+      "learning_rate": 1.7418367346938775e-05,
+      "loss": 0.409,
+      "step": 2530
+    },
+    {
+      "epoch": 12.959183673469388,
+      "grad_norm": 3.3804609775543213,
+      "learning_rate": 1.7408163265306124e-05,
+      "loss": 0.2236,
+      "step": 2540
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.851985559566787,
+      "eval_loss": 0.44439464807510376,
+      "eval_runtime": 0.9402,
+      "eval_samples_per_second": 294.63,
+      "eval_steps_per_second": 37.228,
+      "step": 2548
+    },
+    {
+      "epoch": 13.010204081632653,
+      "grad_norm": 5.298696994781494,
+      "learning_rate": 1.7397959183673473e-05,
+      "loss": 0.538,
+      "step": 2550
+    },
+    {
+      "epoch": 13.061224489795919,
+      "grad_norm": 5.091602802276611,
+      "learning_rate": 1.738775510204082e-05,
+      "loss": 0.4475,
+      "step": 2560
+    },
+    {
+      "epoch": 13.112244897959183,
+      "grad_norm": 4.266963481903076,
+      "learning_rate": 1.7377551020408165e-05,
+      "loss": 0.3297,
+      "step": 2570
+    },
+    {
+      "epoch": 13.16326530612245,
+      "grad_norm": 4.778618812561035,
+      "learning_rate": 1.736734693877551e-05,
+      "loss": 0.3443,
+      "step": 2580
+    },
+    {
+      "epoch": 13.214285714285714,
+      "grad_norm": 2.724548101425171,
+      "learning_rate": 1.735714285714286e-05,
+      "loss": 0.4932,
+      "step": 2590
+    },
+    {
+      "epoch": 13.26530612244898,
+      "grad_norm": 2.435530185699463,
+      "learning_rate": 1.7346938775510206e-05,
+      "loss": 0.2496,
+      "step": 2600
+    },
+    {
+      "epoch": 13.316326530612244,
+      "grad_norm": 7.0745062828063965,
+      "learning_rate": 1.733673469387755e-05,
+      "loss": 0.4175,
+      "step": 2610
+    },
+    {
+      "epoch": 13.36734693877551,
+      "grad_norm": 2.8941125869750977,
+      "learning_rate": 1.7326530612244897e-05,
+      "loss": 0.4203,
+      "step": 2620
+    },
+    {
+      "epoch": 13.418367346938776,
+      "grad_norm": 4.122393608093262,
+      "learning_rate": 1.7316326530612246e-05,
+      "loss": 0.3874,
+      "step": 2630
+    },
+    {
+      "epoch": 13.46938775510204,
+      "grad_norm": 2.2892613410949707,
+      "learning_rate": 1.7306122448979596e-05,
+      "loss": 0.4277,
+      "step": 2640
+    },
+    {
+      "epoch": 13.520408163265307,
+      "grad_norm": 10.996116638183594,
+      "learning_rate": 1.729591836734694e-05,
+      "loss": 0.4704,
+      "step": 2650
+    },
+    {
+      "epoch": 13.571428571428571,
+      "grad_norm": 3.800659418106079,
+      "learning_rate": 1.7285714285714287e-05,
+      "loss": 0.3591,
+      "step": 2660
+    },
+    {
+      "epoch": 13.622448979591837,
+      "grad_norm": 1.313789963722229,
+      "learning_rate": 1.7275510204081633e-05,
+      "loss": 0.2531,
+      "step": 2670
+    },
+    {
+      "epoch": 13.673469387755102,
+      "grad_norm": 2.3375041484832764,
+      "learning_rate": 1.7265306122448982e-05,
+      "loss": 0.3973,
+      "step": 2680
+    },
+    {
+      "epoch": 13.724489795918368,
+      "grad_norm": 6.231153964996338,
+      "learning_rate": 1.7255102040816328e-05,
+      "loss": 0.4401,
+      "step": 2690
+    },
+    {
+      "epoch": 13.775510204081632,
+      "grad_norm": 4.105325698852539,
+      "learning_rate": 1.7244897959183674e-05,
+      "loss": 0.3107,
+      "step": 2700
+    },
+    {
+      "epoch": 13.826530612244898,
+      "grad_norm": 4.559372901916504,
+      "learning_rate": 1.723469387755102e-05,
+      "loss": 0.6584,
+      "step": 2710
+    },
+    {
+      "epoch": 13.877551020408163,
+      "grad_norm": 6.511418342590332,
+      "learning_rate": 1.722448979591837e-05,
+      "loss": 0.3436,
+      "step": 2720
+    },
+    {
+      "epoch": 13.928571428571429,
+      "grad_norm": 3.735102415084839,
+      "learning_rate": 1.7214285714285718e-05,
+      "loss": 0.2158,
+      "step": 2730
+    },
+    {
+      "epoch": 13.979591836734693,
+      "grad_norm": 0.3978160321712494,
+      "learning_rate": 1.7204081632653064e-05,
+      "loss": 0.2083,
+      "step": 2740
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.8736462093862816,
+      "eval_loss": 0.46400919556617737,
+      "eval_runtime": 0.9405,
+      "eval_samples_per_second": 294.528,
+      "eval_steps_per_second": 37.215,
+      "step": 2744
+    },
+    {
+      "epoch": 14.03061224489796,
+      "grad_norm": 10.202120780944824,
+      "learning_rate": 1.719387755102041e-05,
+      "loss": 0.5765,
+      "step": 2750
+    },
+    {
+      "epoch": 14.081632653061224,
+      "grad_norm": 0.6840581893920898,
+      "learning_rate": 1.7183673469387755e-05,
+      "loss": 0.2435,
+      "step": 2760
+    },
+    {
+      "epoch": 14.13265306122449,
+      "grad_norm": 2.577873945236206,
+      "learning_rate": 1.7173469387755104e-05,
+      "loss": 0.2914,
+      "step": 2770
+    },
+    {
+      "epoch": 14.183673469387756,
+      "grad_norm": 2.1203761100769043,
+      "learning_rate": 1.716326530612245e-05,
+      "loss": 0.3378,
+      "step": 2780
+    },
+    {
+      "epoch": 14.23469387755102,
+      "grad_norm": 0.5613865852355957,
+      "learning_rate": 1.7153061224489796e-05,
+      "loss": 0.3199,
+      "step": 2790
+    },
+    {
+      "epoch": 14.285714285714286,
+      "grad_norm": 2.023918867111206,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.2684,
+      "step": 2800
+    },
+    {
+      "epoch": 14.33673469387755,
+      "grad_norm": 4.932062149047852,
+      "learning_rate": 1.713265306122449e-05,
+      "loss": 0.3795,
+      "step": 2810
+    },
+    {
+      "epoch": 14.387755102040817,
+      "grad_norm": 4.413599967956543,
+      "learning_rate": 1.712244897959184e-05,
+      "loss": 0.3618,
+      "step": 2820
+    },
+    {
+      "epoch": 14.438775510204081,
+      "grad_norm": 9.810211181640625,
+      "learning_rate": 1.7112244897959186e-05,
+      "loss": 0.4044,
+      "step": 2830
+    },
+    {
+      "epoch": 14.489795918367347,
+      "grad_norm": 2.67509126663208,
+      "learning_rate": 1.7102040816326532e-05,
+      "loss": 0.4036,
+      "step": 2840
+    },
+    {
+      "epoch": 14.540816326530612,
+      "grad_norm": 4.9529852867126465,
+      "learning_rate": 1.7091836734693878e-05,
+      "loss": 0.3714,
+      "step": 2850
+    },
+    {
+      "epoch": 14.591836734693878,
+      "grad_norm": 4.113321304321289,
+      "learning_rate": 1.7081632653061227e-05,
+      "loss": 0.2794,
+      "step": 2860
+    },
+    {
+      "epoch": 14.642857142857142,
+      "grad_norm": 1.3466196060180664,
+      "learning_rate": 1.7071428571428573e-05,
+      "loss": 0.7391,
+      "step": 2870
+    },
+    {
+      "epoch": 14.693877551020408,
+      "grad_norm": 6.835224151611328,
+      "learning_rate": 1.7061224489795922e-05,
+      "loss": 0.564,
+      "step": 2880
+    },
+    {
+      "epoch": 14.744897959183673,
+      "grad_norm": 7.45158576965332,
+      "learning_rate": 1.7051020408163264e-05,
+      "loss": 0.3609,
+      "step": 2890
+    },
+    {
+      "epoch": 14.795918367346939,
+      "grad_norm": 6.361330986022949,
+      "learning_rate": 1.7040816326530613e-05,
+      "loss": 0.445,
+      "step": 2900
+    },
+    {
+      "epoch": 14.846938775510203,
+      "grad_norm": 6.676398277282715,
+      "learning_rate": 1.7030612244897962e-05,
+      "loss": 0.3851,
+      "step": 2910
+    },
+    {
+      "epoch": 14.89795918367347,
+      "grad_norm": 2.9715816974639893,
+      "learning_rate": 1.7020408163265308e-05,
+      "loss": 0.3791,
+      "step": 2920
+    },
+    {
+      "epoch": 14.948979591836736,
+      "grad_norm": 3.722626209259033,
+      "learning_rate": 1.7010204081632654e-05,
+      "loss": 0.5322,
+      "step": 2930
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 4.071798324584961,
+      "learning_rate": 1.7e-05,
+      "loss": 0.3408,
+      "step": 2940
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.8772563176895307,
+      "eval_loss": 0.37753719091415405,
+      "eval_runtime": 0.9331,
+      "eval_samples_per_second": 296.873,
+      "eval_steps_per_second": 37.511,
+      "step": 2940
+    },
+    {
+      "epoch": 15.051020408163266,
+      "grad_norm": 2.8461458683013916,
+      "learning_rate": 1.698979591836735e-05,
+      "loss": 0.3063,
+      "step": 2950
+    },
+    {
+      "epoch": 15.10204081632653,
+      "grad_norm": 2.373788595199585,
+      "learning_rate": 1.6979591836734695e-05,
+      "loss": 0.2771,
+      "step": 2960
+    },
+    {
+      "epoch": 15.153061224489797,
+      "grad_norm": 1.0632851123809814,
+      "learning_rate": 1.6969387755102044e-05,
+      "loss": 0.3648,
+      "step": 2970
+    },
+    {
+      "epoch": 15.204081632653061,
+      "grad_norm": 2.820613145828247,
+      "learning_rate": 1.695918367346939e-05,
+      "loss": 0.4814,
+      "step": 2980
+    },
+    {
+      "epoch": 15.255102040816327,
+      "grad_norm": 1.9779471158981323,
+      "learning_rate": 1.6948979591836736e-05,
+      "loss": 0.2805,
+      "step": 2990
+    },
+    {
+      "epoch": 15.306122448979592,
+      "grad_norm": 1.2560408115386963,
+      "learning_rate": 1.6938775510204085e-05,
+      "loss": 0.3092,
+      "step": 3000
+    },
+    {
+      "epoch": 15.357142857142858,
+      "grad_norm": 2.176464557647705,
+      "learning_rate": 1.692857142857143e-05,
+      "loss": 0.5962,
+      "step": 3010
+    },
+    {
+      "epoch": 15.408163265306122,
+      "grad_norm": 3.4904580116271973,
+      "learning_rate": 1.6918367346938776e-05,
+      "loss": 0.3132,
+      "step": 3020
+    },
+    {
+      "epoch": 15.459183673469388,
+      "grad_norm": 3.03112530708313,
+      "learning_rate": 1.6908163265306122e-05,
+      "loss": 0.2164,
+      "step": 3030
+    },
+    {
+      "epoch": 15.510204081632653,
+      "grad_norm": 1.1290910243988037,
+      "learning_rate": 1.689795918367347e-05,
+      "loss": 0.1778,
+      "step": 3040
+    },
+    {
+      "epoch": 15.561224489795919,
+      "grad_norm": 5.56776237487793,
+      "learning_rate": 1.6887755102040817e-05,
+      "loss": 0.2771,
+      "step": 3050
+    },
+    {
+      "epoch": 15.612244897959183,
+      "grad_norm": 3.38120436668396,
+      "learning_rate": 1.6877551020408166e-05,
+      "loss": 0.4485,
+      "step": 3060
+    },
+    {
+      "epoch": 15.66326530612245,
+      "grad_norm": 1.8314141035079956,
+      "learning_rate": 1.6867346938775512e-05,
+      "loss": 0.3359,
+      "step": 3070
+    },
+    {
+      "epoch": 15.714285714285714,
+      "grad_norm": 6.086407661437988,
+      "learning_rate": 1.6857142857142858e-05,
+      "loss": 0.5121,
+      "step": 3080
+    },
+    {
+      "epoch": 15.76530612244898,
+      "grad_norm": 3.3119635581970215,
+      "learning_rate": 1.6846938775510207e-05,
+      "loss": 0.283,
+      "step": 3090
+    },
+    {
+      "epoch": 15.816326530612244,
+      "grad_norm": 9.957314491271973,
+      "learning_rate": 1.6836734693877553e-05,
+      "loss": 0.7381,
+      "step": 3100
+    },
+    {
+      "epoch": 15.86734693877551,
+      "grad_norm": 6.10607385635376,
+      "learning_rate": 1.68265306122449e-05,
+      "loss": 0.5771,
+      "step": 3110
+    },
+    {
+      "epoch": 15.918367346938776,
+      "grad_norm": 1.159850001335144,
+      "learning_rate": 1.6816326530612244e-05,
+      "loss": 0.4173,
+      "step": 3120
+    },
+    {
+      "epoch": 15.96938775510204,
+      "grad_norm": 1.2680679559707642,
+      "learning_rate": 1.6806122448979594e-05,
+      "loss": 0.3529,
+      "step": 3130
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.8880866425992779,
+      "eval_loss": 0.3519056737422943,
+      "eval_runtime": 0.9395,
+      "eval_samples_per_second": 294.852,
+      "eval_steps_per_second": 37.256,
+      "step": 3136
+    },
+    {
+      "epoch": 16.020408163265305,
+      "grad_norm": 5.024880886077881,
+      "learning_rate": 1.679591836734694e-05,
+      "loss": 0.2527,
+      "step": 3140
+    },
+    {
+      "epoch": 16.071428571428573,
+      "grad_norm": 7.212440490722656,
+      "learning_rate": 1.678571428571429e-05,
+      "loss": 0.4876,
+      "step": 3150
+    },
+    {
+      "epoch": 16.122448979591837,
+      "grad_norm": 2.613367795944214,
+      "learning_rate": 1.6775510204081634e-05,
+      "loss": 0.1999,
+      "step": 3160
+    },
+    {
+      "epoch": 16.1734693877551,
+      "grad_norm": 4.305943965911865,
+      "learning_rate": 1.676530612244898e-05,
+      "loss": 0.1841,
+      "step": 3170
+    },
+    {
+      "epoch": 16.224489795918366,
+      "grad_norm": 4.246516227722168,
+      "learning_rate": 1.675510204081633e-05,
+      "loss": 0.2093,
+      "step": 3180
+    },
+    {
+      "epoch": 16.275510204081634,
+      "grad_norm": 1.0356781482696533,
+      "learning_rate": 1.6744897959183675e-05,
+      "loss": 0.3482,
+      "step": 3190
+    },
+    {
+      "epoch": 16.3265306122449,
+      "grad_norm": 6.927212238311768,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 0.4234,
+      "step": 3200
+    },
+    {
+      "epoch": 16.377551020408163,
+      "grad_norm": 1.6659942865371704,
+      "learning_rate": 1.6724489795918367e-05,
+      "loss": 0.3295,
+      "step": 3210
+    },
+    {
+      "epoch": 16.428571428571427,
+      "grad_norm": 3.484823703765869,
+      "learning_rate": 1.6714285714285716e-05,
+      "loss": 0.3618,
+      "step": 3220
+    },
+    {
+      "epoch": 16.479591836734695,
+      "grad_norm": 9.103923797607422,
+      "learning_rate": 1.6704081632653062e-05,
+      "loss": 0.3733,
+      "step": 3230
+    },
+    {
+      "epoch": 16.53061224489796,
+      "grad_norm": 3.0150811672210693,
+      "learning_rate": 1.669387755102041e-05,
+      "loss": 0.2741,
+      "step": 3240
+    },
+    {
+      "epoch": 16.581632653061224,
+      "grad_norm": 3.678077459335327,
+      "learning_rate": 1.6683673469387757e-05,
+      "loss": 0.2475,
+      "step": 3250
+    },
+    {
+      "epoch": 16.632653061224488,
+      "grad_norm": 4.152568817138672,
+      "learning_rate": 1.6673469387755102e-05,
+      "loss": 0.346,
+      "step": 3260
+    },
+    {
+      "epoch": 16.683673469387756,
+      "grad_norm": 3.0340733528137207,
+      "learning_rate": 1.666326530612245e-05,
+      "loss": 0.2615,
+      "step": 3270
+    },
+    {
+      "epoch": 16.73469387755102,
+      "grad_norm": 6.065981388092041,
+      "learning_rate": 1.6653061224489797e-05,
+      "loss": 0.4197,
+      "step": 3280
+    },
+    {
+      "epoch": 16.785714285714285,
+      "grad_norm": 4.897032260894775,
+      "learning_rate": 1.6642857142857147e-05,
+      "loss": 0.1801,
+      "step": 3290
+    },
+    {
+      "epoch": 16.836734693877553,
+      "grad_norm": 5.482302188873291,
+      "learning_rate": 1.6632653061224492e-05,
+      "loss": 0.6186,
+      "step": 3300
+    },
+    {
+      "epoch": 16.887755102040817,
+      "grad_norm": 3.145723581314087,
+      "learning_rate": 1.6622448979591838e-05,
+      "loss": 0.2651,
+      "step": 3310
+    },
+    {
+      "epoch": 16.93877551020408,
+      "grad_norm": 3.2254679203033447,
+      "learning_rate": 1.6612244897959184e-05,
+      "loss": 0.2739,
+      "step": 3320
+    },
+    {
+      "epoch": 16.989795918367346,
+      "grad_norm": 4.273975372314453,
+      "learning_rate": 1.6602040816326533e-05,
+      "loss": 0.3859,
+      "step": 3330
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.9061371841155235,
+      "eval_loss": 0.33103030920028687,
+      "eval_runtime": 0.9376,
+      "eval_samples_per_second": 295.433,
+      "eval_steps_per_second": 37.329,
+      "step": 3332
+    },
+    {
+      "epoch": 17.040816326530614,
+      "grad_norm": 5.280905246734619,
+      "learning_rate": 1.659183673469388e-05,
+      "loss": 0.3949,
+      "step": 3340
+    },
+    {
+      "epoch": 17.091836734693878,
+      "grad_norm": 3.1650149822235107,
+      "learning_rate": 1.6581632653061225e-05,
+      "loss": 0.272,
+      "step": 3350
+    },
+    {
+      "epoch": 17.142857142857142,
+      "grad_norm": 1.1566836833953857,
+      "learning_rate": 1.6571428571428574e-05,
+      "loss": 0.174,
+      "step": 3360
+    },
+    {
+      "epoch": 17.193877551020407,
+      "grad_norm": 3.5513110160827637,
+      "learning_rate": 1.656122448979592e-05,
+      "loss": 0.4144,
+      "step": 3370
+    },
+    {
+      "epoch": 17.244897959183675,
+      "grad_norm": 9.73366928100586,
+      "learning_rate": 1.655102040816327e-05,
+      "loss": 0.5116,
+      "step": 3380
+    },
+    {
+      "epoch": 17.29591836734694,
+      "grad_norm": 4.954830169677734,
+      "learning_rate": 1.6540816326530615e-05,
+      "loss": 0.1691,
+      "step": 3390
+    },
+    {
+      "epoch": 17.346938775510203,
+      "grad_norm": 0.931755781173706,
+      "learning_rate": 1.653061224489796e-05,
+      "loss": 0.397,
+      "step": 3400
+    },
+    {
+      "epoch": 17.397959183673468,
+      "grad_norm": 2.4209439754486084,
+      "learning_rate": 1.6520408163265306e-05,
+      "loss": 0.3927,
+      "step": 3410
+    },
+    {
+      "epoch": 17.448979591836736,
+      "grad_norm": 0.28612983226776123,
+      "learning_rate": 1.6510204081632655e-05,
+      "loss": 0.3412,
+      "step": 3420
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 0.8477025628089905,
+      "learning_rate": 1.65e-05,
+      "loss": 0.5129,
+      "step": 3430
+    },
+    {
+      "epoch": 17.551020408163264,
+      "grad_norm": 8.41015338897705,
+      "learning_rate": 1.6489795918367347e-05,
+      "loss": 0.337,
+      "step": 3440
+    },
+    {
+      "epoch": 17.602040816326532,
+      "grad_norm": 1.4178277254104614,
+      "learning_rate": 1.6479591836734696e-05,
+      "loss": 0.3139,
+      "step": 3450
+    },
+    {
+      "epoch": 17.653061224489797,
+      "grad_norm": 3.9586143493652344,
+      "learning_rate": 1.6469387755102042e-05,
+      "loss": 0.3928,
+      "step": 3460
+    },
+    {
+      "epoch": 17.70408163265306,
+      "grad_norm": 3.131901264190674,
+      "learning_rate": 1.645918367346939e-05,
+      "loss": 0.3992,
+      "step": 3470
+    },
+    {
+      "epoch": 17.755102040816325,
+      "grad_norm": 2.3614821434020996,
+      "learning_rate": 1.6448979591836737e-05,
+      "loss": 0.5565,
+      "step": 3480
+    },
+    {
+      "epoch": 17.806122448979593,
+      "grad_norm": 0.8451371788978577,
+      "learning_rate": 1.6438775510204083e-05,
+      "loss": 0.6182,
+      "step": 3490
+    },
+    {
+      "epoch": 17.857142857142858,
+      "grad_norm": 8.3782377243042,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 0.2668,
+      "step": 3500
+    },
+    {
+      "epoch": 17.908163265306122,
+      "grad_norm": 3.256161689758301,
+      "learning_rate": 1.6418367346938778e-05,
+      "loss": 0.2882,
+      "step": 3510
+    },
+    {
+      "epoch": 17.959183673469386,
+      "grad_norm": 3.928743839263916,
+      "learning_rate": 1.6408163265306124e-05,
+      "loss": 0.3557,
+      "step": 3520
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.8916967509025271,
+      "eval_loss": 0.34751203656196594,
+      "eval_runtime": 0.9989,
+      "eval_samples_per_second": 277.319,
+      "eval_steps_per_second": 35.04,
+      "step": 3528
+    },
+    {
+      "epoch": 18.010204081632654,
+      "grad_norm": 0.5895347595214844,
+      "learning_rate": 1.639795918367347e-05,
+      "loss": 0.3817,
+      "step": 3530
+    },
+    {
+      "epoch": 18.06122448979592,
+      "grad_norm": 0.8526825904846191,
+      "learning_rate": 1.638775510204082e-05,
+      "loss": 0.4163,
+      "step": 3540
+    },
+    {
+      "epoch": 18.112244897959183,
+      "grad_norm": 7.634555816650391,
+      "learning_rate": 1.6377551020408164e-05,
+      "loss": 0.3727,
+      "step": 3550
+    },
+    {
+      "epoch": 18.163265306122447,
+      "grad_norm": 7.4755682945251465,
+      "learning_rate": 1.6367346938775513e-05,
+      "loss": 0.3624,
+      "step": 3560
+    },
+    {
+      "epoch": 18.214285714285715,
+      "grad_norm": 10.033106803894043,
+      "learning_rate": 1.635714285714286e-05,
+      "loss": 0.2825,
+      "step": 3570
+    },
+    {
+      "epoch": 18.26530612244898,
+      "grad_norm": 7.6165289878845215,
+      "learning_rate": 1.6346938775510205e-05,
+      "loss": 0.441,
+      "step": 3580
+    },
+    {
+      "epoch": 18.316326530612244,
+      "grad_norm": 2.656892776489258,
+      "learning_rate": 1.633673469387755e-05,
+      "loss": 0.3865,
+      "step": 3590
+    },
+    {
+      "epoch": 18.367346938775512,
+      "grad_norm": 3.905784845352173,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 0.4773,
+      "step": 3600
+    },
+    {
+      "epoch": 18.418367346938776,
+      "grad_norm": 2.772597551345825,
+      "learning_rate": 1.6316326530612246e-05,
+      "loss": 0.435,
+      "step": 3610
+    },
+    {
+      "epoch": 18.46938775510204,
+      "grad_norm": 8.21883487701416,
+      "learning_rate": 1.630612244897959e-05,
+      "loss": 0.3305,
+      "step": 3620
+    },
+    {
+      "epoch": 18.520408163265305,
+      "grad_norm": 7.803742408752441,
+      "learning_rate": 1.629591836734694e-05,
+      "loss": 0.2956,
+      "step": 3630
+    },
+    {
+      "epoch": 18.571428571428573,
+      "grad_norm": 1.3410613536834717,
+      "learning_rate": 1.6285714285714287e-05,
+      "loss": 0.3765,
+      "step": 3640
+    },
+    {
+      "epoch": 18.622448979591837,
+      "grad_norm": 8.857250213623047,
+      "learning_rate": 1.6275510204081636e-05,
+      "loss": 0.535,
+      "step": 3650
+    },
+    {
+      "epoch": 18.6734693877551,
+      "grad_norm": 7.803662300109863,
+      "learning_rate": 1.626530612244898e-05,
+      "loss": 0.3534,
+      "step": 3660
+    },
+    {
+      "epoch": 18.724489795918366,
+      "grad_norm": 0.8503233194351196,
+      "learning_rate": 1.6255102040816327e-05,
+      "loss": 0.3612,
+      "step": 3670
+    },
+    {
+      "epoch": 18.775510204081634,
+      "grad_norm": 4.057595252990723,
+      "learning_rate": 1.6244897959183673e-05,
+      "loss": 0.3237,
+      "step": 3680
+    },
+    {
+      "epoch": 18.8265306122449,
+      "grad_norm": 3.173178195953369,
+      "learning_rate": 1.6234693877551022e-05,
+      "loss": 0.1646,
+      "step": 3690
+    },
+    {
+      "epoch": 18.877551020408163,
+      "grad_norm": 3.036897659301758,
+      "learning_rate": 1.6224489795918368e-05,
+      "loss": 0.1965,
+      "step": 3700
+    },
+    {
+      "epoch": 18.928571428571427,
+      "grad_norm": 6.649494647979736,
+      "learning_rate": 1.6214285714285717e-05,
+      "loss": 0.4771,
+      "step": 3710
+    },
+    {
+      "epoch": 18.979591836734695,
+      "grad_norm": 1.3008012771606445,
+      "learning_rate": 1.6204081632653063e-05,
+      "loss": 0.4979,
+      "step": 3720
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.8592057761732852,
+      "eval_loss": 0.38391339778900146,
+      "eval_runtime": 0.9328,
+      "eval_samples_per_second": 296.961,
+      "eval_steps_per_second": 37.522,
+      "step": 3724
+    },
+    {
+      "epoch": 19.03061224489796,
+      "grad_norm": 5.8646464347839355,
+      "learning_rate": 1.619387755102041e-05,
+      "loss": 0.4242,
+      "step": 3730
+    },
+    {
+      "epoch": 19.081632653061224,
+      "grad_norm": 4.661500453948975,
+      "learning_rate": 1.6183673469387758e-05,
+      "loss": 0.2847,
+      "step": 3740
+    },
+    {
+      "epoch": 19.132653061224488,
+      "grad_norm": 2.1249194145202637,
+      "learning_rate": 1.6173469387755104e-05,
+      "loss": 0.4618,
+      "step": 3750
+    },
+    {
+      "epoch": 19.183673469387756,
+      "grad_norm": 6.6412458419799805,
+      "learning_rate": 1.616326530612245e-05,
+      "loss": 0.4795,
+      "step": 3760
+    },
+    {
+      "epoch": 19.23469387755102,
+      "grad_norm": 5.910946369171143,
+      "learning_rate": 1.6153061224489795e-05,
+      "loss": 0.2605,
+      "step": 3770
+    },
+    {
+      "epoch": 19.285714285714285,
+      "grad_norm": 5.426766395568848,
+      "learning_rate": 1.6142857142857145e-05,
+      "loss": 0.3467,
+      "step": 3780
+    },
+    {
+      "epoch": 19.336734693877553,
+      "grad_norm": 2.4188570976257324,
+      "learning_rate": 1.613265306122449e-05,
+      "loss": 0.3199,
+      "step": 3790
+    },
+    {
+      "epoch": 19.387755102040817,
+      "grad_norm": 3.9347238540649414,
+      "learning_rate": 1.612244897959184e-05,
+      "loss": 0.4666,
+      "step": 3800
+    },
+    {
+      "epoch": 19.43877551020408,
+      "grad_norm": 1.762243628501892,
+      "learning_rate": 1.6112244897959185e-05,
+      "loss": 0.1945,
+      "step": 3810
+    },
+    {
+      "epoch": 19.489795918367346,
+      "grad_norm": 1.250897765159607,
+      "learning_rate": 1.610204081632653e-05,
+      "loss": 0.3725,
+      "step": 3820
+    },
+    {
+      "epoch": 19.540816326530614,
+      "grad_norm": 4.747711658477783,
+      "learning_rate": 1.609183673469388e-05,
+      "loss": 0.3012,
+      "step": 3830
+    },
+    {
+      "epoch": 19.591836734693878,
+      "grad_norm": 6.576929092407227,
+      "learning_rate": 1.6081632653061226e-05,
+      "loss": 0.3594,
+      "step": 3840
+    },
+    {
+      "epoch": 19.642857142857142,
+      "grad_norm": 5.363382816314697,
+      "learning_rate": 1.6071428571428572e-05,
+      "loss": 0.32,
+      "step": 3850
+    },
+    {
+      "epoch": 19.693877551020407,
+      "grad_norm": 1.3070534467697144,
+      "learning_rate": 1.6061224489795918e-05,
+      "loss": 0.4294,
+      "step": 3860
+    },
+    {
+      "epoch": 19.744897959183675,
+      "grad_norm": 7.561715126037598,
+      "learning_rate": 1.6051020408163267e-05,
+      "loss": 0.3365,
+      "step": 3870
+    },
+    {
+      "epoch": 19.79591836734694,
+      "grad_norm": 3.3143508434295654,
+      "learning_rate": 1.6040816326530613e-05,
+      "loss": 0.3049,
+      "step": 3880
+    },
+    {
+      "epoch": 19.846938775510203,
+      "grad_norm": 3.5196471214294434,
+      "learning_rate": 1.6030612244897962e-05,
+      "loss": 0.2679,
+      "step": 3890
+    },
+    {
+      "epoch": 19.897959183673468,
+      "grad_norm": 0.6857045292854309,
+      "learning_rate": 1.6020408163265308e-05,
+      "loss": 0.2322,
+      "step": 3900
+    },
+    {
+      "epoch": 19.948979591836736,
+      "grad_norm": 0.5474905371665955,
+      "learning_rate": 1.6010204081632653e-05,
+      "loss": 0.2885,
+      "step": 3910
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 2.948782444000244,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7133,
+      "step": 3920
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.9133574007220217,
+      "eval_loss": 0.3031846880912781,
+      "eval_runtime": 0.9345,
+      "eval_samples_per_second": 296.417,
+      "eval_steps_per_second": 37.453,
+      "step": 3920
+    },
+    {
+      "epoch": 20.051020408163264,
+      "grad_norm": 8.756710052490234,
+      "learning_rate": 1.598979591836735e-05,
+      "loss": 0.4335,
+      "step": 3930
+    },
+    {
+      "epoch": 20.102040816326532,
+      "grad_norm": 7.383957862854004,
+      "learning_rate": 1.5979591836734694e-05,
+      "loss": 0.394,
+      "step": 3940
+    },
+    {
+      "epoch": 20.153061224489797,
+      "grad_norm": 5.428812503814697,
+      "learning_rate": 1.596938775510204e-05,
+      "loss": 0.3772,
+      "step": 3950
+    },
+    {
+      "epoch": 20.20408163265306,
+      "grad_norm": 8.893187522888184,
+      "learning_rate": 1.595918367346939e-05,
+      "loss": 0.3712,
+      "step": 3960
+    },
+    {
+      "epoch": 20.255102040816325,
+      "grad_norm": 2.3695433139801025,
+      "learning_rate": 1.5948979591836735e-05,
+      "loss": 0.3681,
+      "step": 3970
+    },
+    {
+      "epoch": 20.306122448979593,
+      "grad_norm": 5.1946611404418945,
+      "learning_rate": 1.5938775510204084e-05,
+      "loss": 0.4013,
+      "step": 3980
+    },
+    {
+      "epoch": 20.357142857142858,
+      "grad_norm": 4.6183342933654785,
+      "learning_rate": 1.592857142857143e-05,
+      "loss": 0.4489,
+      "step": 3990
+    },
+    {
+      "epoch": 20.408163265306122,
+      "grad_norm": 2.7871506214141846,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 0.5088,
+      "step": 4000
+    },
+    {
+      "epoch": 20.459183673469386,
+      "grad_norm": 4.250699996948242,
+      "learning_rate": 1.5908163265306125e-05,
+      "loss": 0.3401,
+      "step": 4010
+    },
+    {
+      "epoch": 20.510204081632654,
+      "grad_norm": 5.782101631164551,
+      "learning_rate": 1.589795918367347e-05,
+      "loss": 0.2461,
+      "step": 4020
+    },
+    {
+      "epoch": 20.56122448979592,
+      "grad_norm": 5.054870128631592,
+      "learning_rate": 1.588775510204082e-05,
+      "loss": 0.2459,
+      "step": 4030
+    },
+    {
+      "epoch": 20.612244897959183,
+      "grad_norm": 8.228409767150879,
+      "learning_rate": 1.5877551020408162e-05,
+      "loss": 0.223,
+      "step": 4040
+    },
+    {
+      "epoch": 20.663265306122447,
+      "grad_norm": 1.5619908571243286,
+      "learning_rate": 1.586734693877551e-05,
+      "loss": 0.3247,
+      "step": 4050
+    },
+    {
+      "epoch": 20.714285714285715,
+      "grad_norm": 8.718620300292969,
+      "learning_rate": 1.5857142857142857e-05,
+      "loss": 0.2373,
+      "step": 4060
+    },
+    {
+      "epoch": 20.76530612244898,
+      "grad_norm": 0.9933898448944092,
+      "learning_rate": 1.5846938775510206e-05,
+      "loss": 0.196,
+      "step": 4070
+    },
+    {
+      "epoch": 20.816326530612244,
+      "grad_norm": 6.243815898895264,
+      "learning_rate": 1.5836734693877552e-05,
+      "loss": 0.1823,
+      "step": 4080
+    },
+    {
+      "epoch": 20.867346938775512,
+      "grad_norm": 4.635849952697754,
+      "learning_rate": 1.5826530612244898e-05,
+      "loss": 0.1604,
+      "step": 4090
+    },
+    {
+      "epoch": 20.918367346938776,
+      "grad_norm": 4.620690822601318,
+      "learning_rate": 1.5816326530612247e-05,
+      "loss": 0.383,
+      "step": 4100
+    },
+    {
+      "epoch": 20.96938775510204,
+      "grad_norm": 1.3753141164779663,
+      "learning_rate": 1.5806122448979593e-05,
+      "loss": 0.4489,
+      "step": 4110
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.851985559566787,
+      "eval_loss": 0.4246116876602173,
+      "eval_runtime": 0.9757,
+      "eval_samples_per_second": 283.894,
+      "eval_steps_per_second": 35.871,
+      "step": 4116
+    },
+    {
+      "epoch": 21.020408163265305,
+      "grad_norm": 10.26496696472168,
+      "learning_rate": 1.5795918367346942e-05,
+      "loss": 0.6521,
+      "step": 4120
+    },
+    {
+      "epoch": 21.071428571428573,
+      "grad_norm": 12.259634971618652,
+      "learning_rate": 1.5785714285714288e-05,
+      "loss": 0.5814,
+      "step": 4130
+    },
+    {
+      "epoch": 21.122448979591837,
+      "grad_norm": 11.424445152282715,
+      "learning_rate": 1.5775510204081634e-05,
+      "loss": 0.3028,
+      "step": 4140
+    },
+    {
+      "epoch": 21.1734693877551,
+      "grad_norm": 3.9125359058380127,
+      "learning_rate": 1.576530612244898e-05,
+      "loss": 0.2002,
+      "step": 4150
+    },
+    {
+      "epoch": 21.224489795918366,
+      "grad_norm": 5.634211540222168,
+      "learning_rate": 1.575510204081633e-05,
+      "loss": 0.2863,
+      "step": 4160
+    },
+    {
+      "epoch": 21.275510204081634,
+      "grad_norm": 0.6257736682891846,
+      "learning_rate": 1.5744897959183675e-05,
+      "loss": 0.3551,
+      "step": 4170
+    },
+    {
+      "epoch": 21.3265306122449,
+      "grad_norm": 5.202589988708496,
+      "learning_rate": 1.573469387755102e-05,
+      "loss": 0.3546,
+      "step": 4180
+    },
+    {
+      "epoch": 21.377551020408163,
+      "grad_norm": 1.4646656513214111,
+      "learning_rate": 1.572448979591837e-05,
+      "loss": 0.2139,
+      "step": 4190
+    },
+    {
+      "epoch": 21.428571428571427,
+      "grad_norm": 2.697577953338623,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.2515,
+      "step": 4200
+    },
+    {
+      "epoch": 21.479591836734695,
+      "grad_norm": 0.9905484318733215,
+      "learning_rate": 1.5704081632653065e-05,
+      "loss": 0.3543,
+      "step": 4210
+    },
+    {
+      "epoch": 21.53061224489796,
+      "grad_norm": 0.8990330100059509,
+      "learning_rate": 1.569387755102041e-05,
+      "loss": 0.2124,
+      "step": 4220
+    },
+    {
+      "epoch": 21.581632653061224,
+      "grad_norm": 2.288618326187134,
+      "learning_rate": 1.5683673469387756e-05,
+      "loss": 0.3029,
+      "step": 4230
+    },
+    {
+      "epoch": 21.632653061224488,
+      "grad_norm": 5.491371154785156,
+      "learning_rate": 1.5673469387755102e-05,
+      "loss": 0.1861,
+      "step": 4240
+    },
+    {
+      "epoch": 21.683673469387756,
+      "grad_norm": 4.103919982910156,
+      "learning_rate": 1.566326530612245e-05,
+      "loss": 0.2237,
+      "step": 4250
+    },
+    {
+      "epoch": 21.73469387755102,
+      "grad_norm": 5.436093330383301,
+      "learning_rate": 1.5653061224489797e-05,
+      "loss": 0.1873,
+      "step": 4260
+    },
+    {
+      "epoch": 21.785714285714285,
+      "grad_norm": 1.5695730447769165,
+      "learning_rate": 1.5642857142857143e-05,
+      "loss": 0.3701,
+      "step": 4270
+    },
+    {
+      "epoch": 21.836734693877553,
+      "grad_norm": 8.12942886352539,
+      "learning_rate": 1.5632653061224492e-05,
+      "loss": 0.5966,
+      "step": 4280
+    },
+    {
+      "epoch": 21.887755102040817,
+      "grad_norm": 1.9971330165863037,
+      "learning_rate": 1.5622448979591838e-05,
+      "loss": 0.4555,
+      "step": 4290
+    },
+    {
+      "epoch": 21.93877551020408,
+      "grad_norm": 8.867376327514648,
+      "learning_rate": 1.5612244897959187e-05,
+      "loss": 0.3845,
+      "step": 4300
+    },
+    {
+      "epoch": 21.989795918367346,
+      "grad_norm": 3.649428129196167,
+      "learning_rate": 1.5602040816326533e-05,
+      "loss": 0.2605,
+      "step": 4310
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.8989169675090253,
+      "eval_loss": 0.29509276151657104,
+      "eval_runtime": 0.9476,
+      "eval_samples_per_second": 292.329,
+      "eval_steps_per_second": 36.937,
+      "step": 4312
+    },
+    {
+      "epoch": 22.040816326530614,
+      "grad_norm": 8.780238151550293,
+      "learning_rate": 1.559183673469388e-05,
+      "loss": 0.3356,
+      "step": 4320
+    },
+    {
+      "epoch": 22.091836734693878,
+      "grad_norm": 4.019279479980469,
+      "learning_rate": 1.5581632653061224e-05,
+      "loss": 0.3704,
+      "step": 4330
+    },
+    {
+      "epoch": 22.142857142857142,
+      "grad_norm": 6.3763556480407715,
+      "learning_rate": 1.5571428571428573e-05,
+      "loss": 0.3621,
+      "step": 4340
+    },
+    {
+      "epoch": 22.193877551020407,
+      "grad_norm": 10.312544822692871,
+      "learning_rate": 1.556122448979592e-05,
+      "loss": 0.364,
+      "step": 4350
+    },
+    {
+      "epoch": 22.244897959183675,
+      "grad_norm": 7.919386386871338,
+      "learning_rate": 1.5551020408163265e-05,
+      "loss": 0.2917,
+      "step": 4360
+    },
+    {
+      "epoch": 22.29591836734694,
+      "grad_norm": 9.648198127746582,
+      "learning_rate": 1.5540816326530614e-05,
+      "loss": 0.4103,
+      "step": 4370
+    },
+    {
+      "epoch": 22.346938775510203,
+      "grad_norm": 10.067304611206055,
+      "learning_rate": 1.553061224489796e-05,
+      "loss": 0.4598,
+      "step": 4380
+    },
+    {
+      "epoch": 22.397959183673468,
+      "grad_norm": 3.0613834857940674,
+      "learning_rate": 1.552040816326531e-05,
+      "loss": 0.2636,
+      "step": 4390
+    },
+    {
+      "epoch": 22.448979591836736,
+      "grad_norm": 4.73079252243042,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 0.2499,
+      "step": 4400
+    },
+    {
+      "epoch": 22.5,
+      "grad_norm": 11.9879150390625,
+      "learning_rate": 1.55e-05,
+      "loss": 0.486,
+      "step": 4410
+    },
+    {
+      "epoch": 22.551020408163264,
+      "grad_norm": 2.766538619995117,
+      "learning_rate": 1.5489795918367346e-05,
+      "loss": 0.3824,
+      "step": 4420
+    },
+    {
+      "epoch": 22.602040816326532,
+      "grad_norm": 1.4572681188583374,
+      "learning_rate": 1.5479591836734696e-05,
+      "loss": 0.3105,
+      "step": 4430
+    },
+    {
+      "epoch": 22.653061224489797,
+      "grad_norm": 4.498220920562744,
+      "learning_rate": 1.546938775510204e-05,
+      "loss": 0.2207,
+      "step": 4440
+    },
+    {
+      "epoch": 22.70408163265306,
+      "grad_norm": 4.226017475128174,
+      "learning_rate": 1.545918367346939e-05,
+      "loss": 0.2277,
+      "step": 4450
+    },
+    {
+      "epoch": 22.755102040816325,
+      "grad_norm": 3.2783775329589844,
+      "learning_rate": 1.5448979591836736e-05,
+      "loss": 0.207,
+      "step": 4460
+    },
+    {
+      "epoch": 22.806122448979593,
+      "grad_norm": 2.267697811126709,
+      "learning_rate": 1.5438775510204082e-05,
+      "loss": 0.3034,
+      "step": 4470
+    },
+    {
+      "epoch": 22.857142857142858,
+      "grad_norm": 2.044019937515259,
+      "learning_rate": 1.542857142857143e-05,
+      "loss": 0.2496,
+      "step": 4480
+    },
+    {
+      "epoch": 22.908163265306122,
+      "grad_norm": 1.8587689399719238,
+      "learning_rate": 1.5418367346938777e-05,
+      "loss": 0.3425,
+      "step": 4490
+    },
+    {
+      "epoch": 22.959183673469386,
+      "grad_norm": 9.889906883239746,
+      "learning_rate": 1.5408163265306123e-05,
+      "loss": 0.3787,
+      "step": 4500
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.851985559566787,
+      "eval_loss": 0.4356595277786255,
+      "eval_runtime": 0.9334,
+      "eval_samples_per_second": 296.758,
+      "eval_steps_per_second": 37.497,
+      "step": 4508
+    },
+    {
+      "epoch": 23.010204081632654,
+      "grad_norm": 9.485286712646484,
+      "learning_rate": 1.539795918367347e-05,
+      "loss": 0.2368,
+      "step": 4510
+    },
+    {
+      "epoch": 23.06122448979592,
+      "grad_norm": 4.098474025726318,
+      "learning_rate": 1.5387755102040818e-05,
+      "loss": 0.3777,
+      "step": 4520
+    },
+    {
+      "epoch": 23.112244897959183,
+      "grad_norm": 3.54447078704834,
+      "learning_rate": 1.5377551020408164e-05,
+      "loss": 0.2732,
+      "step": 4530
+    },
+    {
+      "epoch": 23.163265306122447,
+      "grad_norm": 12.153559684753418,
+      "learning_rate": 1.5367346938775513e-05,
+      "loss": 0.4394,
+      "step": 4540
+    },
+    {
+      "epoch": 23.214285714285715,
+      "grad_norm": 5.04149055480957,
+      "learning_rate": 1.535714285714286e-05,
+      "loss": 0.282,
+      "step": 4550
+    },
+    {
+      "epoch": 23.26530612244898,
+      "grad_norm": 2.7063755989074707,
+      "learning_rate": 1.5346938775510204e-05,
+      "loss": 0.2617,
+      "step": 4560
+    },
+    {
+      "epoch": 23.316326530612244,
+      "grad_norm": 2.060274124145508,
+      "learning_rate": 1.5336734693877554e-05,
+      "loss": 0.2776,
+      "step": 4570
+    },
+    {
+      "epoch": 23.367346938775512,
+      "grad_norm": 4.662898063659668,
+      "learning_rate": 1.53265306122449e-05,
+      "loss": 0.3443,
+      "step": 4580
+    },
+    {
+      "epoch": 23.418367346938776,
+      "grad_norm": 7.403005123138428,
+      "learning_rate": 1.5316326530612245e-05,
+      "loss": 0.3008,
+      "step": 4590
+    },
+    {
+      "epoch": 23.46938775510204,
+      "grad_norm": 7.6282958984375,
+      "learning_rate": 1.530612244897959e-05,
+      "loss": 0.4093,
+      "step": 4600
+    },
+    {
+      "epoch": 23.520408163265305,
+      "grad_norm": 3.8396637439727783,
+      "learning_rate": 1.529591836734694e-05,
+      "loss": 0.2757,
+      "step": 4610
+    },
+    {
+      "epoch": 23.571428571428573,
+      "grad_norm": 3.220073699951172,
+      "learning_rate": 1.5285714285714286e-05,
+      "loss": 0.2751,
+      "step": 4620
+    },
+    {
+      "epoch": 23.622448979591837,
+      "grad_norm": 5.4397993087768555,
+      "learning_rate": 1.5275510204081635e-05,
+      "loss": 0.3633,
+      "step": 4630
+    },
+    {
+      "epoch": 23.6734693877551,
+      "grad_norm": 1.3962507247924805,
+      "learning_rate": 1.526530612244898e-05,
+      "loss": 0.1703,
+      "step": 4640
+    },
+    {
+      "epoch": 23.724489795918366,
+      "grad_norm": 1.0168291330337524,
+      "learning_rate": 1.5255102040816327e-05,
+      "loss": 0.3073,
+      "step": 4650
+    },
+    {
+      "epoch": 23.775510204081634,
+      "grad_norm": 8.263096809387207,
+      "learning_rate": 1.5244897959183676e-05,
+      "loss": 0.2385,
+      "step": 4660
+    },
+    {
+      "epoch": 23.8265306122449,
+      "grad_norm": 6.277915954589844,
+      "learning_rate": 1.5234693877551022e-05,
+      "loss": 0.374,
+      "step": 4670
+    },
+    {
+      "epoch": 23.877551020408163,
+      "grad_norm": 2.2303617000579834,
+      "learning_rate": 1.522448979591837e-05,
+      "loss": 0.4268,
+      "step": 4680
+    },
+    {
+      "epoch": 23.928571428571427,
+      "grad_norm": 8.869429588317871,
+      "learning_rate": 1.5214285714285715e-05,
+      "loss": 0.2736,
+      "step": 4690
+    },
+    {
+      "epoch": 23.979591836734695,
+      "grad_norm": 0.8636966943740845,
+      "learning_rate": 1.5204081632653063e-05,
+      "loss": 0.3015,
+      "step": 4700
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.8916967509025271,
+      "eval_loss": 0.399036169052124,
+      "eval_runtime": 0.9356,
+      "eval_samples_per_second": 296.067,
+      "eval_steps_per_second": 37.409,
+      "step": 4704
+    },
+    {
+      "epoch": 24.03061224489796,
+      "grad_norm": 8.083073616027832,
+      "learning_rate": 1.5193877551020408e-05,
+      "loss": 0.3505,
+      "step": 4710
+    },
+    {
+      "epoch": 24.081632653061224,
+      "grad_norm": 2.8755526542663574,
+      "learning_rate": 1.5183673469387756e-05,
+      "loss": 0.3789,
+      "step": 4720
+    },
+    {
+      "epoch": 24.132653061224488,
+      "grad_norm": 5.0803542137146,
+      "learning_rate": 1.5173469387755105e-05,
+      "loss": 0.3118,
+      "step": 4730
+    },
+    {
+      "epoch": 24.183673469387756,
+      "grad_norm": 2.7088780403137207,
+      "learning_rate": 1.516326530612245e-05,
+      "loss": 0.5032,
+      "step": 4740
+    },
+    {
+      "epoch": 24.23469387755102,
+      "grad_norm": 1.4047300815582275,
+      "learning_rate": 1.5153061224489798e-05,
+      "loss": 0.4985,
+      "step": 4750
+    },
+    {
+      "epoch": 24.285714285714285,
+      "grad_norm": 1.823101282119751,
+      "learning_rate": 1.5142857142857144e-05,
+      "loss": 0.3021,
+      "step": 4760
+    },
+    {
+      "epoch": 24.336734693877553,
+      "grad_norm": 3.2638440132141113,
+      "learning_rate": 1.5132653061224492e-05,
+      "loss": 0.2204,
+      "step": 4770
+    },
+    {
+      "epoch": 24.387755102040817,
+      "grad_norm": 1.3173271417617798,
+      "learning_rate": 1.5122448979591837e-05,
+      "loss": 0.1328,
+      "step": 4780
+    },
+    {
+      "epoch": 24.43877551020408,
+      "grad_norm": 1.607823133468628,
+      "learning_rate": 1.5112244897959185e-05,
+      "loss": 0.2578,
+      "step": 4790
+    },
+    {
+      "epoch": 24.489795918367346,
+      "grad_norm": 4.061956405639648,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 0.3741,
+      "step": 4800
+    },
+    {
+      "epoch": 24.540816326530614,
+      "grad_norm": 5.889439582824707,
+      "learning_rate": 1.5091836734693878e-05,
+      "loss": 0.4102,
+      "step": 4810
+    },
+    {
+      "epoch": 24.591836734693878,
+      "grad_norm": 10.062840461730957,
+      "learning_rate": 1.5081632653061227e-05,
+      "loss": 0.3837,
+      "step": 4820
+    },
+    {
+      "epoch": 24.642857142857142,
+      "grad_norm": 0.9003174304962158,
+      "learning_rate": 1.5071428571428573e-05,
+      "loss": 0.1315,
+      "step": 4830
+    },
+    {
+      "epoch": 24.693877551020407,
+      "grad_norm": 5.709663391113281,
+      "learning_rate": 1.506122448979592e-05,
+      "loss": 0.4183,
+      "step": 4840
+    },
+    {
+      "epoch": 24.744897959183675,
+      "grad_norm": 8.025142669677734,
+      "learning_rate": 1.5051020408163266e-05,
+      "loss": 0.5562,
+      "step": 4850
+    },
+    {
+      "epoch": 24.79591836734694,
+      "grad_norm": 2.7048165798187256,
+      "learning_rate": 1.5040816326530614e-05,
+      "loss": 0.2714,
+      "step": 4860
+    },
+    {
+      "epoch": 24.846938775510203,
+      "grad_norm": 1.3981176614761353,
+      "learning_rate": 1.503061224489796e-05,
+      "loss": 0.2331,
+      "step": 4870
+    },
+    {
+      "epoch": 24.897959183673468,
+      "grad_norm": 6.711188316345215,
+      "learning_rate": 1.5020408163265307e-05,
+      "loss": 0.2991,
+      "step": 4880
+    },
+    {
+      "epoch": 24.948979591836736,
+      "grad_norm": 1.7855308055877686,
+      "learning_rate": 1.5010204081632653e-05,
+      "loss": 0.4607,
+      "step": 4890
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 2.935310125350952,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.1965,
+      "step": 4900
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.9097472924187726,
+      "eval_loss": 0.35359257459640503,
+      "eval_runtime": 0.9318,
+      "eval_samples_per_second": 297.276,
+      "eval_steps_per_second": 37.562,
+      "step": 4900
+    },
+    {
+      "epoch": 25.051020408163264,
+      "grad_norm": 11.405259132385254,
+      "learning_rate": 1.498979591836735e-05,
+      "loss": 0.3897,
+      "step": 4910
+    },
+    {
+      "epoch": 25.102040816326532,
+      "grad_norm": 8.36063003540039,
+      "learning_rate": 1.4979591836734695e-05,
+      "loss": 0.3899,
+      "step": 4920
+    },
+    {
+      "epoch": 25.153061224489797,
+      "grad_norm": 8.212508201599121,
+      "learning_rate": 1.4969387755102043e-05,
+      "loss": 0.2401,
+      "step": 4930
+    },
+    {
+      "epoch": 25.20408163265306,
+      "grad_norm": 0.3883512318134308,
+      "learning_rate": 1.4959183673469389e-05,
+      "loss": 0.4655,
+      "step": 4940
+    },
+    {
+      "epoch": 25.255102040816325,
+      "grad_norm": 8.68212890625,
+      "learning_rate": 1.4948979591836736e-05,
+      "loss": 0.4715,
+      "step": 4950
+    },
+    {
+      "epoch": 25.306122448979593,
+      "grad_norm": 1.6489917039871216,
+      "learning_rate": 1.4938775510204082e-05,
+      "loss": 0.1642,
+      "step": 4960
+    },
+    {
+      "epoch": 25.357142857142858,
+      "grad_norm": 1.582821011543274,
+      "learning_rate": 1.492857142857143e-05,
+      "loss": 0.3276,
+      "step": 4970
+    },
+    {
+      "epoch": 25.408163265306122,
+      "grad_norm": 0.6229325532913208,
+      "learning_rate": 1.4918367346938775e-05,
+      "loss": 0.3723,
+      "step": 4980
+    },
+    {
+      "epoch": 25.459183673469386,
+      "grad_norm": 2.6093995571136475,
+      "learning_rate": 1.4908163265306124e-05,
+      "loss": 0.1918,
+      "step": 4990
+    },
+    {
+      "epoch": 25.510204081632654,
+      "grad_norm": 5.734492301940918,
+      "learning_rate": 1.4897959183673472e-05,
+      "loss": 0.2327,
+      "step": 5000
+    },
+    {
+      "epoch": 25.56122448979592,
+      "grad_norm": 2.498661518096924,
+      "learning_rate": 1.4887755102040818e-05,
+      "loss": 0.1976,
+      "step": 5010
+    },
+    {
+      "epoch": 25.612244897959183,
+      "grad_norm": 4.310215473175049,
+      "learning_rate": 1.4877551020408165e-05,
+      "loss": 0.3573,
+      "step": 5020
+    },
+    {
+      "epoch": 25.663265306122447,
+      "grad_norm": 1.1879396438598633,
+      "learning_rate": 1.4867346938775511e-05,
+      "loss": 0.2937,
+      "step": 5030
+    },
+    {
+      "epoch": 25.714285714285715,
+      "grad_norm": 3.3825039863586426,
+      "learning_rate": 1.4857142857142858e-05,
+      "loss": 0.2267,
+      "step": 5040
+    },
+    {
+      "epoch": 25.76530612244898,
+      "grad_norm": 3.127350091934204,
+      "learning_rate": 1.4846938775510204e-05,
+      "loss": 0.3636,
+      "step": 5050
+    },
+    {
+      "epoch": 25.816326530612244,
+      "grad_norm": 1.4972935914993286,
+      "learning_rate": 1.4836734693877552e-05,
+      "loss": 0.1758,
+      "step": 5060
+    },
+    {
+      "epoch": 25.867346938775512,
+      "grad_norm": 12.983075141906738,
+      "learning_rate": 1.4826530612244897e-05,
+      "loss": 0.3909,
+      "step": 5070
+    },
+    {
+      "epoch": 25.918367346938776,
+      "grad_norm": 5.06458854675293,
+      "learning_rate": 1.4816326530612247e-05,
+      "loss": 0.38,
+      "step": 5080
+    },
+    {
+      "epoch": 25.96938775510204,
+      "grad_norm": 7.696865081787109,
+      "learning_rate": 1.4806122448979594e-05,
+      "loss": 0.3903,
+      "step": 5090
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.8592057761732852,
+      "eval_loss": 0.4166140854358673,
+      "eval_runtime": 0.9346,
+      "eval_samples_per_second": 296.393,
+      "eval_steps_per_second": 37.45,
+      "step": 5096
+    },
+    {
+      "epoch": 26.020408163265305,
+      "grad_norm": 1.0229040384292603,
+      "learning_rate": 1.479591836734694e-05,
+      "loss": 0.1086,
+      "step": 5100
+    },
+    {
+      "epoch": 26.071428571428573,
+      "grad_norm": 3.7563860416412354,
+      "learning_rate": 1.4785714285714287e-05,
+      "loss": 0.3433,
+      "step": 5110
+    },
+    {
+      "epoch": 26.122448979591837,
+      "grad_norm": 2.8265397548675537,
+      "learning_rate": 1.4775510204081633e-05,
+      "loss": 0.2704,
+      "step": 5120
+    },
+    {
+      "epoch": 26.1734693877551,
+      "grad_norm": 0.6223945617675781,
+      "learning_rate": 1.476530612244898e-05,
+      "loss": 0.2733,
+      "step": 5130
+    },
+    {
+      "epoch": 26.224489795918366,
+      "grad_norm": 7.791995525360107,
+      "learning_rate": 1.4755102040816326e-05,
+      "loss": 0.3287,
+      "step": 5140
+    },
+    {
+      "epoch": 26.275510204081634,
+      "grad_norm": 6.363265514373779,
+      "learning_rate": 1.4744897959183676e-05,
+      "loss": 0.4244,
+      "step": 5150
+    },
+    {
+      "epoch": 26.3265306122449,
+      "grad_norm": 1.1195369958877563,
+      "learning_rate": 1.4734693877551021e-05,
+      "loss": 0.2089,
+      "step": 5160
+    },
+    {
+      "epoch": 26.377551020408163,
+      "grad_norm": 0.9491761922836304,
+      "learning_rate": 1.4724489795918369e-05,
+      "loss": 0.3159,
+      "step": 5170
+    },
+    {
+      "epoch": 26.428571428571427,
+      "grad_norm": 0.22193892300128937,
+      "learning_rate": 1.4714285714285716e-05,
+      "loss": 0.3834,
+      "step": 5180
+    },
+    {
+      "epoch": 26.479591836734695,
+      "grad_norm": 4.889684200286865,
+      "learning_rate": 1.4704081632653062e-05,
+      "loss": 0.3216,
+      "step": 5190
+    },
+    {
+      "epoch": 26.53061224489796,
+      "grad_norm": 4.601675987243652,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 0.3598,
+      "step": 5200
+    },
+    {
+      "epoch": 26.581632653061224,
+      "grad_norm": 7.6923441886901855,
+      "learning_rate": 1.4683673469387756e-05,
+      "loss": 0.2714,
+      "step": 5210
+    },
+    {
+      "epoch": 26.632653061224488,
+      "grad_norm": 8.871493339538574,
+      "learning_rate": 1.4673469387755103e-05,
+      "loss": 0.4931,
+      "step": 5220
+    },
+    {
+      "epoch": 26.683673469387756,
+      "grad_norm": 1.7044825553894043,
+      "learning_rate": 1.4663265306122449e-05,
+      "loss": 0.34,
+      "step": 5230
+    },
+    {
+      "epoch": 26.73469387755102,
+      "grad_norm": 2.5381484031677246,
+      "learning_rate": 1.4653061224489798e-05,
+      "loss": 0.2527,
+      "step": 5240
+    },
+    {
+      "epoch": 26.785714285714285,
+      "grad_norm": 11.087688446044922,
+      "learning_rate": 1.4642857142857144e-05,
+      "loss": 0.3773,
+      "step": 5250
+    },
+    {
+      "epoch": 26.836734693877553,
+      "grad_norm": 1.4621928930282593,
+      "learning_rate": 1.4632653061224491e-05,
+      "loss": 0.2989,
+      "step": 5260
+    },
+    {
+      "epoch": 26.887755102040817,
+      "grad_norm": 1.6922417879104614,
+      "learning_rate": 1.4622448979591839e-05,
+      "loss": 0.2947,
+      "step": 5270
+    },
+    {
+      "epoch": 26.93877551020408,
+      "grad_norm": 1.8192353248596191,
+      "learning_rate": 1.4612244897959185e-05,
+      "loss": 0.2285,
+      "step": 5280
+    },
+    {
+      "epoch": 26.989795918367346,
+      "grad_norm": 1.7619237899780273,
+      "learning_rate": 1.4602040816326532e-05,
+      "loss": 0.1902,
+      "step": 5290
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.851985559566787,
+      "eval_loss": 0.4353993237018585,
+      "eval_runtime": 0.9607,
+      "eval_samples_per_second": 288.322,
+      "eval_steps_per_second": 36.431,
+      "step": 5292
+    },
+    {
+      "epoch": 27.040816326530614,
+      "grad_norm": 3.745148181915283,
+      "learning_rate": 1.4591836734693878e-05,
+      "loss": 0.1512,
+      "step": 5300
+    },
+    {
+      "epoch": 27.091836734693878,
+      "grad_norm": 6.893754482269287,
+      "learning_rate": 1.4581632653061227e-05,
+      "loss": 0.5395,
+      "step": 5310
+    },
+    {
+      "epoch": 27.142857142857142,
+      "grad_norm": 5.727985382080078,
+      "learning_rate": 1.4571428571428573e-05,
+      "loss": 0.3727,
+      "step": 5320
+    },
+    {
+      "epoch": 27.193877551020407,
+      "grad_norm": 1.6175438165664673,
+      "learning_rate": 1.456122448979592e-05,
+      "loss": 0.1489,
+      "step": 5330
+    },
+    {
+      "epoch": 27.244897959183675,
+      "grad_norm": 1.226585030555725,
+      "learning_rate": 1.4551020408163266e-05,
+      "loss": 0.3616,
+      "step": 5340
+    },
+    {
+      "epoch": 27.29591836734694,
+      "grad_norm": 3.1076860427856445,
+      "learning_rate": 1.4540816326530614e-05,
+      "loss": 0.2072,
+      "step": 5350
+    },
+    {
+      "epoch": 27.346938775510203,
+      "grad_norm": 5.883277416229248,
+      "learning_rate": 1.4530612244897961e-05,
+      "loss": 0.1949,
+      "step": 5360
+    },
+    {
+      "epoch": 27.397959183673468,
+      "grad_norm": 1.5578702688217163,
+      "learning_rate": 1.4520408163265307e-05,
+      "loss": 0.5629,
+      "step": 5370
+    },
+    {
+      "epoch": 27.448979591836736,
+      "grad_norm": 0.5179188847541809,
+      "learning_rate": 1.4510204081632654e-05,
+      "loss": 0.2987,
+      "step": 5380
+    },
+    {
+      "epoch": 27.5,
+      "grad_norm": 7.038111686706543,
+      "learning_rate": 1.45e-05,
+      "loss": 0.2434,
+      "step": 5390
+    },
+    {
+      "epoch": 27.551020408163264,
+      "grad_norm": 9.72697925567627,
+      "learning_rate": 1.448979591836735e-05,
+      "loss": 0.3516,
+      "step": 5400
+    },
+    {
+      "epoch": 27.602040816326532,
+      "grad_norm": 1.2628819942474365,
+      "learning_rate": 1.4479591836734695e-05,
+      "loss": 0.2303,
+      "step": 5410
+    },
+    {
+      "epoch": 27.653061224489797,
+      "grad_norm": 8.881223678588867,
+      "learning_rate": 1.4469387755102043e-05,
+      "loss": 0.3717,
+      "step": 5420
+    },
+    {
+      "epoch": 27.70408163265306,
+      "grad_norm": 0.9033036828041077,
+      "learning_rate": 1.4459183673469388e-05,
+      "loss": 0.4453,
+      "step": 5430
+    },
+    {
+      "epoch": 27.755102040816325,
+      "grad_norm": 2.869086503982544,
+      "learning_rate": 1.4448979591836736e-05,
+      "loss": 0.3423,
+      "step": 5440
+    },
+    {
+      "epoch": 27.806122448979593,
+      "grad_norm": 3.165790319442749,
+      "learning_rate": 1.4438775510204083e-05,
+      "loss": 0.3382,
+      "step": 5450
+    },
+    {
+      "epoch": 27.857142857142858,
+      "grad_norm": 1.9456369876861572,
+      "learning_rate": 1.4428571428571429e-05,
+      "loss": 0.4745,
+      "step": 5460
+    },
+    {
+      "epoch": 27.908163265306122,
+      "grad_norm": 10.290801048278809,
+      "learning_rate": 1.4418367346938778e-05,
+      "loss": 0.2637,
+      "step": 5470
+    },
+    {
+      "epoch": 27.959183673469386,
+      "grad_norm": 1.7097864151000977,
+      "learning_rate": 1.4408163265306122e-05,
+      "loss": 0.2089,
+      "step": 5480
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.8592057761732852,
+      "eval_loss": 0.4088672697544098,
+      "eval_runtime": 0.9344,
+      "eval_samples_per_second": 296.462,
+      "eval_steps_per_second": 37.459,
+      "step": 5488
+    },
+    {
+      "epoch": 28.010204081632654,
+      "grad_norm": 10.47132682800293,
+      "learning_rate": 1.4397959183673472e-05,
+      "loss": 0.3704,
+      "step": 5490
+    },
+    {
+      "epoch": 28.06122448979592,
+      "grad_norm": 2.4138362407684326,
+      "learning_rate": 1.4387755102040817e-05,
+      "loss": 0.2158,
+      "step": 5500
+    },
+    {
+      "epoch": 28.112244897959183,
+      "grad_norm": 14.150741577148438,
+      "learning_rate": 1.4377551020408165e-05,
+      "loss": 0.4084,
+      "step": 5510
+    },
+    {
+      "epoch": 28.163265306122447,
+      "grad_norm": 5.0319061279296875,
+      "learning_rate": 1.436734693877551e-05,
+      "loss": 0.2888,
+      "step": 5520
+    },
+    {
+      "epoch": 28.214285714285715,
+      "grad_norm": 8.061519622802734,
+      "learning_rate": 1.4357142857142858e-05,
+      "loss": 0.3848,
+      "step": 5530
+    },
+    {
+      "epoch": 28.26530612244898,
+      "grad_norm": 0.9189940690994263,
+      "learning_rate": 1.4346938775510206e-05,
+      "loss": 0.2323,
+      "step": 5540
+    },
+    {
+      "epoch": 28.316326530612244,
+      "grad_norm": 2.1076314449310303,
+      "learning_rate": 1.4336734693877551e-05,
+      "loss": 0.2321,
+      "step": 5550
+    },
+    {
+      "epoch": 28.367346938775512,
+      "grad_norm": 6.96281623840332,
+      "learning_rate": 1.43265306122449e-05,
+      "loss": 0.1701,
+      "step": 5560
+    },
+    {
+      "epoch": 28.418367346938776,
+      "grad_norm": 6.155792713165283,
+      "learning_rate": 1.4316326530612246e-05,
+      "loss": 0.3143,
+      "step": 5570
+    },
+    {
+      "epoch": 28.46938775510204,
+      "grad_norm": 6.938708782196045,
+      "learning_rate": 1.4306122448979594e-05,
+      "loss": 0.6464,
+      "step": 5580
+    },
+    {
+      "epoch": 28.520408163265305,
+      "grad_norm": 5.764797210693359,
+      "learning_rate": 1.429591836734694e-05,
+      "loss": 0.4086,
+      "step": 5590
+    },
+    {
+      "epoch": 28.571428571428573,
+      "grad_norm": 5.969749450683594,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.324,
+      "step": 5600
+    },
+    {
+      "epoch": 28.622448979591837,
+      "grad_norm": 3.8588509559631348,
+      "learning_rate": 1.4275510204081633e-05,
+      "loss": 0.2058,
+      "step": 5610
+    },
+    {
+      "epoch": 28.6734693877551,
+      "grad_norm": 0.7087581753730774,
+      "learning_rate": 1.426530612244898e-05,
+      "loss": 0.4878,
+      "step": 5620
+    },
+    {
+      "epoch": 28.724489795918366,
+      "grad_norm": 0.6094661355018616,
+      "learning_rate": 1.425510204081633e-05,
+      "loss": 0.2933,
+      "step": 5630
+    },
+    {
+      "epoch": 28.775510204081634,
+      "grad_norm": 8.96648120880127,
+      "learning_rate": 1.4244897959183674e-05,
+      "loss": 0.2122,
+      "step": 5640
+    },
+    {
+      "epoch": 28.8265306122449,
+      "grad_norm": 4.653097152709961,
+      "learning_rate": 1.4234693877551023e-05,
+      "loss": 0.2367,
+      "step": 5650
+    },
+    {
+      "epoch": 28.877551020408163,
+      "grad_norm": 1.0812416076660156,
+      "learning_rate": 1.4224489795918369e-05,
+      "loss": 0.2423,
+      "step": 5660
+    },
+    {
+      "epoch": 28.928571428571427,
+      "grad_norm": 1.8441146612167358,
+      "learning_rate": 1.4214285714285716e-05,
+      "loss": 0.3556,
+      "step": 5670
+    },
+    {
+      "epoch": 28.979591836734695,
+      "grad_norm": 6.369021892547607,
+      "learning_rate": 1.4204081632653062e-05,
+      "loss": 0.3574,
+      "step": 5680
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.8231046931407943,
+      "eval_loss": 0.4787161946296692,
+      "eval_runtime": 0.9366,
+      "eval_samples_per_second": 295.746,
+      "eval_steps_per_second": 37.369,
+      "step": 5684
+    },
+    {
+      "epoch": 29.03061224489796,
+      "grad_norm": 7.944931983947754,
+      "learning_rate": 1.419387755102041e-05,
+      "loss": 0.2558,
+      "step": 5690
+    },
+    {
+      "epoch": 29.081632653061224,
+      "grad_norm": 8.573824882507324,
+      "learning_rate": 1.4183673469387755e-05,
+      "loss": 0.2327,
+      "step": 5700
+    },
+    {
+      "epoch": 29.132653061224488,
+      "grad_norm": 11.918915748596191,
+      "learning_rate": 1.4173469387755103e-05,
+      "loss": 0.5803,
+      "step": 5710
+    },
+    {
+      "epoch": 29.183673469387756,
+      "grad_norm": 1.88558030128479,
+      "learning_rate": 1.4163265306122452e-05,
+      "loss": 0.2855,
+      "step": 5720
+    },
+    {
+      "epoch": 29.23469387755102,
+      "grad_norm": 1.4961941242218018,
+      "learning_rate": 1.4153061224489798e-05,
+      "loss": 0.5397,
+      "step": 5730
+    },
+    {
+      "epoch": 29.285714285714285,
+      "grad_norm": 4.03718376159668,
+      "learning_rate": 1.4142857142857145e-05,
+      "loss": 0.3524,
+      "step": 5740
+    },
+    {
+      "epoch": 29.336734693877553,
+      "grad_norm": 0.6138362884521484,
+      "learning_rate": 1.4132653061224491e-05,
+      "loss": 0.2927,
+      "step": 5750
+    },
+    {
+      "epoch": 29.387755102040817,
+      "grad_norm": 3.0943140983581543,
+      "learning_rate": 1.4122448979591838e-05,
+      "loss": 0.2377,
+      "step": 5760
+    },
+    {
+      "epoch": 29.43877551020408,
+      "grad_norm": 2.0186758041381836,
+      "learning_rate": 1.4112244897959184e-05,
+      "loss": 0.2708,
+      "step": 5770
+    },
+    {
+      "epoch": 29.489795918367346,
+      "grad_norm": 1.613853096961975,
+      "learning_rate": 1.4102040816326532e-05,
+      "loss": 0.2674,
+      "step": 5780
+    },
+    {
+      "epoch": 29.540816326530614,
+      "grad_norm": 5.511127948760986,
+      "learning_rate": 1.4091836734693877e-05,
+      "loss": 0.4286,
+      "step": 5790
+    },
+    {
+      "epoch": 29.591836734693878,
+      "grad_norm": 1.3316036462783813,
+      "learning_rate": 1.4081632653061225e-05,
+      "loss": 0.3182,
+      "step": 5800
+    },
+    {
+      "epoch": 29.642857142857142,
+      "grad_norm": 0.6130739450454712,
+      "learning_rate": 1.4071428571428574e-05,
+      "loss": 0.3473,
+      "step": 5810
+    },
+    {
+      "epoch": 29.693877551020407,
+      "grad_norm": 1.8860011100769043,
+      "learning_rate": 1.406122448979592e-05,
+      "loss": 0.2681,
+      "step": 5820
+    },
+    {
+      "epoch": 29.744897959183675,
+      "grad_norm": 10.419530868530273,
+      "learning_rate": 1.4051020408163267e-05,
+      "loss": 0.2665,
+      "step": 5830
+    },
+    {
+      "epoch": 29.79591836734694,
+      "grad_norm": 4.105865955352783,
+      "learning_rate": 1.4040816326530613e-05,
+      "loss": 0.2498,
+      "step": 5840
+    },
+    {
+      "epoch": 29.846938775510203,
+      "grad_norm": 0.8506119847297668,
+      "learning_rate": 1.403061224489796e-05,
+      "loss": 0.2422,
+      "step": 5850
+    },
+    {
+      "epoch": 29.897959183673468,
+      "grad_norm": 6.38096809387207,
+      "learning_rate": 1.4020408163265307e-05,
+      "loss": 0.1484,
+      "step": 5860
+    },
+    {
+      "epoch": 29.948979591836736,
+      "grad_norm": 1.4350731372833252,
+      "learning_rate": 1.4010204081632654e-05,
+      "loss": 0.3318,
+      "step": 5870
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 2.968691110610962,
+      "learning_rate": 1.4e-05,
+      "loss": 0.3532,
+      "step": 5880
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.9097472924187726,
+      "eval_loss": 0.3165305256843567,
+      "eval_runtime": 0.9457,
+      "eval_samples_per_second": 292.897,
+      "eval_steps_per_second": 37.009,
+      "step": 5880
+    },
+    {
+      "epoch": 30.051020408163264,
+      "grad_norm": 1.0100151300430298,
+      "learning_rate": 1.3989795918367349e-05,
+      "loss": 0.2086,
+      "step": 5890
+    },
+    {
+      "epoch": 30.102040816326532,
+      "grad_norm": 2.064863681793213,
+      "learning_rate": 1.3979591836734696e-05,
+      "loss": 0.2836,
+      "step": 5900
+    },
+    {
+      "epoch": 30.153061224489797,
+      "grad_norm": 11.070706367492676,
+      "learning_rate": 1.3969387755102042e-05,
+      "loss": 0.451,
+      "step": 5910
+    },
+    {
+      "epoch": 30.20408163265306,
+      "grad_norm": 1.8765898942947388,
+      "learning_rate": 1.395918367346939e-05,
+      "loss": 0.2461,
+      "step": 5920
+    },
+    {
+      "epoch": 30.255102040816325,
+      "grad_norm": 3.754838466644287,
+      "learning_rate": 1.3948979591836736e-05,
+      "loss": 0.4792,
+      "step": 5930
+    },
+    {
+      "epoch": 30.306122448979593,
+      "grad_norm": 1.705304503440857,
+      "learning_rate": 1.3938775510204083e-05,
+      "loss": 0.2855,
+      "step": 5940
+    },
+    {
+      "epoch": 30.357142857142858,
+      "grad_norm": 7.264845371246338,
+      "learning_rate": 1.3928571428571429e-05,
+      "loss": 0.3534,
+      "step": 5950
+    },
+    {
+      "epoch": 30.408163265306122,
+      "grad_norm": 4.922964572906494,
+      "learning_rate": 1.3918367346938776e-05,
+      "loss": 0.2287,
+      "step": 5960
+    },
+    {
+      "epoch": 30.459183673469386,
+      "grad_norm": 11.60462474822998,
+      "learning_rate": 1.3908163265306122e-05,
+      "loss": 0.3458,
+      "step": 5970
+    },
+    {
+      "epoch": 30.510204081632654,
+      "grad_norm": 9.114325523376465,
+      "learning_rate": 1.3897959183673471e-05,
+      "loss": 0.3271,
+      "step": 5980
+    },
+    {
+      "epoch": 30.56122448979592,
+      "grad_norm": 2.0705983638763428,
+      "learning_rate": 1.3887755102040819e-05,
+      "loss": 0.3466,
+      "step": 5990
+    },
+    {
+      "epoch": 30.612244897959183,
+      "grad_norm": 3.5480122566223145,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 0.267,
+      "step": 6000
+    },
+    {
+      "epoch": 30.663265306122447,
+      "grad_norm": 2.192810535430908,
+      "learning_rate": 1.3867346938775512e-05,
+      "loss": 0.241,
+      "step": 6010
+    },
+    {
+      "epoch": 30.714285714285715,
+      "grad_norm": 2.5832903385162354,
+      "learning_rate": 1.3857142857142858e-05,
+      "loss": 0.2774,
+      "step": 6020
+    },
+    {
+      "epoch": 30.76530612244898,
+      "grad_norm": 0.5127232074737549,
+      "learning_rate": 1.3846938775510205e-05,
+      "loss": 0.1853,
+      "step": 6030
+    },
+    {
+      "epoch": 30.816326530612244,
+      "grad_norm": 1.949842095375061,
+      "learning_rate": 1.3836734693877551e-05,
+      "loss": 0.2198,
+      "step": 6040
+    },
+    {
+      "epoch": 30.867346938775512,
+      "grad_norm": 5.536325931549072,
+      "learning_rate": 1.38265306122449e-05,
+      "loss": 0.4455,
+      "step": 6050
+    },
+    {
+      "epoch": 30.918367346938776,
+      "grad_norm": 7.894504070281982,
+      "learning_rate": 1.3816326530612244e-05,
+      "loss": 0.2814,
+      "step": 6060
+    },
+    {
+      "epoch": 30.96938775510204,
+      "grad_norm": 8.459007263183594,
+      "learning_rate": 1.3806122448979594e-05,
+      "loss": 0.2967,
+      "step": 6070
+    },
+    {
+      "epoch": 31.0,
+      "eval_accuracy": 0.9133574007220217,
+      "eval_loss": 0.31053850054740906,
+      "eval_runtime": 0.9369,
+      "eval_samples_per_second": 295.665,
+      "eval_steps_per_second": 37.358,
+      "step": 6076
+    },
+    {
+      "epoch": 31.020408163265305,
+      "grad_norm": 0.4988439381122589,
+      "learning_rate": 1.3795918367346941e-05,
+      "loss": 0.2044,
+      "step": 6080
+    },
+    {
+      "epoch": 31.071428571428573,
+      "grad_norm": 5.99605131149292,
+      "learning_rate": 1.3785714285714287e-05,
+      "loss": 0.4048,
+      "step": 6090
+    },
+    {
+      "epoch": 31.122448979591837,
+      "grad_norm": 0.37132564187049866,
+      "learning_rate": 1.3775510204081634e-05,
+      "loss": 0.2627,
+      "step": 6100
+    },
+    {
+      "epoch": 31.1734693877551,
+      "grad_norm": 6.084136486053467,
+      "learning_rate": 1.376530612244898e-05,
+      "loss": 0.2869,
+      "step": 6110
+    },
+    {
+      "epoch": 31.224489795918366,
+      "grad_norm": 11.331133842468262,
+      "learning_rate": 1.3755102040816328e-05,
+      "loss": 0.273,
+      "step": 6120
+    },
+    {
+      "epoch": 31.275510204081634,
+      "grad_norm": 3.7786152362823486,
+      "learning_rate": 1.3744897959183673e-05,
+      "loss": 0.2437,
+      "step": 6130
+    },
+    {
+      "epoch": 31.3265306122449,
+      "grad_norm": 3.998039722442627,
+      "learning_rate": 1.3734693877551023e-05,
+      "loss": 0.1321,
+      "step": 6140
+    },
+    {
+      "epoch": 31.377551020408163,
+      "grad_norm": 10.547959327697754,
+      "learning_rate": 1.3724489795918368e-05,
+      "loss": 0.43,
+      "step": 6150
+    },
+    {
+      "epoch": 31.428571428571427,
+      "grad_norm": 1.094752311706543,
+      "learning_rate": 1.3714285714285716e-05,
+      "loss": 0.3469,
+      "step": 6160
+    },
+    {
+      "epoch": 31.479591836734695,
+      "grad_norm": 8.193305969238281,
+      "learning_rate": 1.3704081632653062e-05,
+      "loss": 0.1532,
+      "step": 6170
+    },
+    {
+      "epoch": 31.53061224489796,
+      "grad_norm": 7.120598793029785,
+      "learning_rate": 1.3693877551020409e-05,
+      "loss": 0.3421,
+      "step": 6180
+    },
+    {
+      "epoch": 31.581632653061224,
+      "grad_norm": 6.944319248199463,
+      "learning_rate": 1.3683673469387757e-05,
+      "loss": 0.3574,
+      "step": 6190
+    },
+    {
+      "epoch": 31.632653061224488,
+      "grad_norm": 7.441619873046875,
+      "learning_rate": 1.3673469387755102e-05,
+      "loss": 0.384,
+      "step": 6200
+    },
+    {
+      "epoch": 31.683673469387756,
+      "grad_norm": 5.745674133300781,
+      "learning_rate": 1.366326530612245e-05,
+      "loss": 0.4554,
+      "step": 6210
+    },
+    {
+      "epoch": 31.73469387755102,
+      "grad_norm": 8.12261962890625,
+      "learning_rate": 1.3653061224489796e-05,
+      "loss": 0.2278,
+      "step": 6220
+    },
+    {
+      "epoch": 31.785714285714285,
+      "grad_norm": 1.7826637029647827,
+      "learning_rate": 1.3642857142857145e-05,
+      "loss": 0.1841,
+      "step": 6230
+    },
+    {
+      "epoch": 31.836734693877553,
+      "grad_norm": 8.810450553894043,
+      "learning_rate": 1.363265306122449e-05,
+      "loss": 0.3919,
+      "step": 6240
+    },
+    {
+      "epoch": 31.887755102040817,
+      "grad_norm": 2.942906618118286,
+      "learning_rate": 1.3622448979591838e-05,
+      "loss": 0.1034,
+      "step": 6250
+    },
+    {
+      "epoch": 31.93877551020408,
+      "grad_norm": 7.177513599395752,
+      "learning_rate": 1.3612244897959184e-05,
+      "loss": 0.2669,
+      "step": 6260
+    },
+    {
+      "epoch": 31.989795918367346,
+      "grad_norm": 0.5618299245834351,
+      "learning_rate": 1.3602040816326531e-05,
+      "loss": 0.2364,
+      "step": 6270
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.9061371841155235,
+      "eval_loss": 0.3559906780719757,
+      "eval_runtime": 0.9792,
+      "eval_samples_per_second": 282.89,
+      "eval_steps_per_second": 35.744,
+      "step": 6272
+    },
+    {
+      "epoch": 32.04081632653061,
+      "grad_norm": 0.4897635877132416,
+      "learning_rate": 1.3591836734693879e-05,
+      "loss": 0.4591,
+      "step": 6280
+    },
+    {
+      "epoch": 32.09183673469388,
+      "grad_norm": 5.982337474822998,
+      "learning_rate": 1.3581632653061225e-05,
+      "loss": 0.4622,
+      "step": 6290
+    },
+    {
+      "epoch": 32.142857142857146,
+      "grad_norm": 14.7687349319458,
+      "learning_rate": 1.3571428571428574e-05,
+      "loss": 0.3969,
+      "step": 6300
+    },
+    {
+      "epoch": 32.19387755102041,
+      "grad_norm": 1.2857763767242432,
+      "learning_rate": 1.356122448979592e-05,
+      "loss": 0.0797,
+      "step": 6310
+    },
+    {
+      "epoch": 32.244897959183675,
+      "grad_norm": 5.122455596923828,
+      "learning_rate": 1.3551020408163267e-05,
+      "loss": 0.2258,
+      "step": 6320
+    },
+    {
+      "epoch": 32.295918367346935,
+      "grad_norm": 3.5429632663726807,
+      "learning_rate": 1.3540816326530613e-05,
+      "loss": 0.3753,
+      "step": 6330
+    },
+    {
+      "epoch": 32.3469387755102,
+      "grad_norm": 12.08903694152832,
+      "learning_rate": 1.353061224489796e-05,
+      "loss": 0.3908,
+      "step": 6340
+    },
+    {
+      "epoch": 32.39795918367347,
+      "grad_norm": 12.813987731933594,
+      "learning_rate": 1.3520408163265306e-05,
+      "loss": 0.2859,
+      "step": 6350
+    },
+    {
+      "epoch": 32.44897959183673,
+      "grad_norm": 1.3361541032791138,
+      "learning_rate": 1.3510204081632654e-05,
+      "loss": 0.2899,
+      "step": 6360
+    },
+    {
+      "epoch": 32.5,
+      "grad_norm": 8.692803382873535,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 0.3218,
+      "step": 6370
+    },
+    {
+      "epoch": 32.55102040816327,
+      "grad_norm": 10.775219917297363,
+      "learning_rate": 1.3489795918367347e-05,
+      "loss": 0.3706,
+      "step": 6380
+    },
+    {
+      "epoch": 32.60204081632653,
+      "grad_norm": 6.802149295806885,
+      "learning_rate": 1.3479591836734696e-05,
+      "loss": 0.2627,
+      "step": 6390
+    },
+    {
+      "epoch": 32.6530612244898,
+      "grad_norm": 1.9009625911712646,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 0.2294,
+      "step": 6400
+    },
+    {
+      "epoch": 32.704081632653065,
+      "grad_norm": 5.793837070465088,
+      "learning_rate": 1.345918367346939e-05,
+      "loss": 0.3042,
+      "step": 6410
+    },
+    {
+      "epoch": 32.755102040816325,
+      "grad_norm": 8.77164077758789,
+      "learning_rate": 1.3448979591836735e-05,
+      "loss": 0.5283,
+      "step": 6420
+    },
+    {
+      "epoch": 32.80612244897959,
+      "grad_norm": 6.6339616775512695,
+      "learning_rate": 1.3438775510204083e-05,
+      "loss": 0.3025,
+      "step": 6430
+    },
+    {
+      "epoch": 32.857142857142854,
+      "grad_norm": 9.5437593460083,
+      "learning_rate": 1.3428571428571429e-05,
+      "loss": 0.1714,
+      "step": 6440
+    },
+    {
+      "epoch": 32.90816326530612,
+      "grad_norm": 3.033761739730835,
+      "learning_rate": 1.3418367346938776e-05,
+      "loss": 0.36,
+      "step": 6450
+    },
+    {
+      "epoch": 32.95918367346939,
+      "grad_norm": 5.7934112548828125,
+      "learning_rate": 1.3408163265306125e-05,
+      "loss": 0.3136,
+      "step": 6460
+    },
+    {
+      "epoch": 33.0,
+      "eval_accuracy": 0.9097472924187726,
+      "eval_loss": 0.26566287875175476,
+      "eval_runtime": 0.9353,
+      "eval_samples_per_second": 296.157,
+      "eval_steps_per_second": 37.421,
+      "step": 6468
+    },
+    {
+      "epoch": 33.01020408163265,
+      "grad_norm": 1.0668991804122925,
+      "learning_rate": 1.3397959183673471e-05,
+      "loss": 0.1893,
+      "step": 6470
+    },
+    {
+      "epoch": 33.06122448979592,
+      "grad_norm": 1.4086322784423828,
+      "learning_rate": 1.3387755102040818e-05,
+      "loss": 0.5995,
+      "step": 6480
+    },
+    {
+      "epoch": 33.11224489795919,
+      "grad_norm": 4.279899597167969,
+      "learning_rate": 1.3377551020408164e-05,
+      "loss": 0.4313,
+      "step": 6490
+    },
+    {
+      "epoch": 33.16326530612245,
+      "grad_norm": 5.1103410720825195,
+      "learning_rate": 1.3367346938775512e-05,
+      "loss": 0.1508,
+      "step": 6500
+    },
+    {
+      "epoch": 33.214285714285715,
+      "grad_norm": 0.4270355999469757,
+      "learning_rate": 1.3357142857142858e-05,
+      "loss": 0.1682,
+      "step": 6510
+    },
+    {
+      "epoch": 33.265306122448976,
+      "grad_norm": 9.594346046447754,
+      "learning_rate": 1.3346938775510205e-05,
+      "loss": 0.2922,
+      "step": 6520
+    },
+    {
+      "epoch": 33.316326530612244,
+      "grad_norm": 0.859392523765564,
+      "learning_rate": 1.333673469387755e-05,
+      "loss": 0.2269,
+      "step": 6530
+    },
+    {
+      "epoch": 33.36734693877551,
+      "grad_norm": 2.0459938049316406,
+      "learning_rate": 1.3326530612244898e-05,
+      "loss": 0.211,
+      "step": 6540
+    },
+    {
+      "epoch": 33.41836734693877,
+      "grad_norm": 9.009632110595703,
+      "learning_rate": 1.3316326530612247e-05,
+      "loss": 0.3548,
+      "step": 6550
+    },
+    {
+      "epoch": 33.46938775510204,
+      "grad_norm": 3.5498781204223633,
+      "learning_rate": 1.3306122448979593e-05,
+      "loss": 0.3769,
+      "step": 6560
+    },
+    {
+      "epoch": 33.52040816326531,
+      "grad_norm": 2.0291507244110107,
+      "learning_rate": 1.329591836734694e-05,
+      "loss": 0.1357,
+      "step": 6570
+    },
+    {
+      "epoch": 33.57142857142857,
+      "grad_norm": 5.882681369781494,
+      "learning_rate": 1.3285714285714287e-05,
+      "loss": 0.5029,
+      "step": 6580
+    },
+    {
+      "epoch": 33.62244897959184,
+      "grad_norm": 8.612032890319824,
+      "learning_rate": 1.3275510204081634e-05,
+      "loss": 0.3223,
+      "step": 6590
+    },
+    {
+      "epoch": 33.673469387755105,
+      "grad_norm": 2.1690680980682373,
+      "learning_rate": 1.326530612244898e-05,
+      "loss": 0.3455,
+      "step": 6600
+    },
+    {
+      "epoch": 33.724489795918366,
+      "grad_norm": 2.9613852500915527,
+      "learning_rate": 1.3255102040816327e-05,
+      "loss": 0.3186,
+      "step": 6610
+    },
+    {
+      "epoch": 33.775510204081634,
+      "grad_norm": 2.842536687850952,
+      "learning_rate": 1.3244897959183673e-05,
+      "loss": 0.2595,
+      "step": 6620
+    },
+    {
+      "epoch": 33.826530612244895,
+      "grad_norm": 6.094658374786377,
+      "learning_rate": 1.323469387755102e-05,
+      "loss": 0.2713,
+      "step": 6630
+    },
+    {
+      "epoch": 33.87755102040816,
+      "grad_norm": 1.204209804534912,
+      "learning_rate": 1.322448979591837e-05,
+      "loss": 0.303,
+      "step": 6640
+    },
+    {
+      "epoch": 33.92857142857143,
+      "grad_norm": 8.503417015075684,
+      "learning_rate": 1.3214285714285716e-05,
+      "loss": 0.3531,
+      "step": 6650
+    },
+    {
+      "epoch": 33.97959183673469,
+      "grad_norm": 13.13253116607666,
+      "learning_rate": 1.3204081632653063e-05,
+      "loss": 0.4061,
+      "step": 6660
+    },
+    {
+      "epoch": 34.0,
+      "eval_accuracy": 0.9133574007220217,
+      "eval_loss": 0.2679595947265625,
+      "eval_runtime": 1.0011,
+      "eval_samples_per_second": 276.693,
+      "eval_steps_per_second": 34.961,
+      "step": 6664
+    },
+    {
+      "epoch": 34.03061224489796,
+      "grad_norm": 10.893752098083496,
+      "learning_rate": 1.3193877551020409e-05,
+      "loss": 0.2804,
+      "step": 6670
+    },
+    {
+      "epoch": 34.08163265306123,
+      "grad_norm": 1.127501368522644,
+      "learning_rate": 1.3183673469387756e-05,
+      "loss": 0.2987,
+      "step": 6680
+    },
+    {
+      "epoch": 34.13265306122449,
+      "grad_norm": 11.394792556762695,
+      "learning_rate": 1.3173469387755102e-05,
+      "loss": 0.3087,
+      "step": 6690
+    },
+    {
+      "epoch": 34.183673469387756,
+      "grad_norm": 3.2792036533355713,
+      "learning_rate": 1.316326530612245e-05,
+      "loss": 0.2087,
+      "step": 6700
+    },
+    {
+      "epoch": 34.234693877551024,
+      "grad_norm": 1.6803452968597412,
+      "learning_rate": 1.3153061224489795e-05,
+      "loss": 0.289,
+      "step": 6710
+    },
+    {
+      "epoch": 34.285714285714285,
+      "grad_norm": 5.636178493499756,
+      "learning_rate": 1.3142857142857145e-05,
+      "loss": 0.2163,
+      "step": 6720
+    },
+    {
+      "epoch": 34.33673469387755,
+      "grad_norm": 2.6255643367767334,
+      "learning_rate": 1.3132653061224492e-05,
+      "loss": 0.2604,
+      "step": 6730
+    },
+    {
+      "epoch": 34.38775510204081,
+      "grad_norm": 0.9665950536727905,
+      "learning_rate": 1.3122448979591838e-05,
+      "loss": 0.3136,
+      "step": 6740
+    },
+    {
+      "epoch": 34.43877551020408,
+      "grad_norm": 5.1290435791015625,
+      "learning_rate": 1.3112244897959185e-05,
+      "loss": 0.2152,
+      "step": 6750
+    },
+    {
+      "epoch": 34.48979591836735,
+      "grad_norm": 4.895906925201416,
+      "learning_rate": 1.3102040816326531e-05,
+      "loss": 0.1662,
+      "step": 6760
+    },
+    {
+      "epoch": 34.54081632653061,
+      "grad_norm": 3.9003512859344482,
+      "learning_rate": 1.3091836734693879e-05,
+      "loss": 0.3426,
+      "step": 6770
+    },
+    {
+      "epoch": 34.59183673469388,
+      "grad_norm": 1.398506760597229,
+      "learning_rate": 1.3081632653061224e-05,
+      "loss": 0.2563,
+      "step": 6780
+    },
+    {
+      "epoch": 34.642857142857146,
+      "grad_norm": 7.401740550994873,
+      "learning_rate": 1.3071428571428572e-05,
+      "loss": 0.4108,
+      "step": 6790
+    },
+    {
+      "epoch": 34.69387755102041,
+      "grad_norm": 16.010316848754883,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 0.3883,
+      "step": 6800
+    },
+    {
+      "epoch": 34.744897959183675,
+      "grad_norm": 6.630160808563232,
+      "learning_rate": 1.3051020408163267e-05,
+      "loss": 0.3897,
+      "step": 6810
+    },
+    {
+      "epoch": 34.795918367346935,
+      "grad_norm": 5.36838960647583,
+      "learning_rate": 1.3040816326530614e-05,
+      "loss": 0.2516,
+      "step": 6820
+    },
+    {
+      "epoch": 34.8469387755102,
+      "grad_norm": 3.5793333053588867,
+      "learning_rate": 1.303061224489796e-05,
+      "loss": 0.2888,
+      "step": 6830
+    },
+    {
+      "epoch": 34.89795918367347,
+      "grad_norm": 0.8172181248664856,
+      "learning_rate": 1.3020408163265308e-05,
+      "loss": 0.3432,
+      "step": 6840
+    },
+    {
+      "epoch": 34.94897959183673,
+      "grad_norm": 9.127881050109863,
+      "learning_rate": 1.3010204081632653e-05,
+      "loss": 0.3537,
+      "step": 6850
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 2.6013247966766357,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.3296,
+      "step": 6860
+    },
+    {
+      "epoch": 35.0,
+      "eval_accuracy": 0.9061371841155235,
+      "eval_loss": 0.3797638416290283,
+      "eval_runtime": 0.9276,
+      "eval_samples_per_second": 298.61,
+      "eval_steps_per_second": 37.73,
+      "step": 6860
+    },
+    {
+      "epoch": 35.05102040816327,
+      "grad_norm": 9.181253433227539,
+      "learning_rate": 1.2989795918367347e-05,
+      "loss": 0.2048,
+      "step": 6870
+    },
+    {
+      "epoch": 35.10204081632653,
+      "grad_norm": 4.2278666496276855,
+      "learning_rate": 1.2979591836734696e-05,
+      "loss": 0.4347,
+      "step": 6880
+    },
+    {
+      "epoch": 35.1530612244898,
+      "grad_norm": 1.9831122159957886,
+      "learning_rate": 1.2969387755102042e-05,
+      "loss": 0.2518,
+      "step": 6890
+    },
+    {
+      "epoch": 35.204081632653065,
+      "grad_norm": 2.629470109939575,
+      "learning_rate": 1.2959183673469389e-05,
+      "loss": 0.1867,
+      "step": 6900
+    },
+    {
+      "epoch": 35.255102040816325,
+      "grad_norm": 2.3086001873016357,
+      "learning_rate": 1.2948979591836737e-05,
+      "loss": 0.1446,
+      "step": 6910
+    },
+    {
+      "epoch": 35.30612244897959,
+      "grad_norm": 0.5559498071670532,
+      "learning_rate": 1.2938775510204082e-05,
+      "loss": 0.4085,
+      "step": 6920
+    },
+    {
+      "epoch": 35.357142857142854,
+      "grad_norm": 2.5674479007720947,
+      "learning_rate": 1.292857142857143e-05,
+      "loss": 0.2453,
+      "step": 6930
+    },
+    {
+      "epoch": 35.40816326530612,
+      "grad_norm": 1.0306987762451172,
+      "learning_rate": 1.2918367346938776e-05,
+      "loss": 0.3145,
+      "step": 6940
+    },
+    {
+      "epoch": 35.45918367346939,
+      "grad_norm": 0.5135687589645386,
+      "learning_rate": 1.2908163265306123e-05,
+      "loss": 0.243,
+      "step": 6950
+    },
+    {
+      "epoch": 35.51020408163265,
+      "grad_norm": 1.1978002786636353,
+      "learning_rate": 1.2897959183673469e-05,
+      "loss": 0.1959,
+      "step": 6960
+    },
+    {
+      "epoch": 35.56122448979592,
+      "grad_norm": 1.2959942817687988,
+      "learning_rate": 1.2887755102040818e-05,
+      "loss": 0.5525,
+      "step": 6970
+    },
+    {
+      "epoch": 35.61224489795919,
+      "grad_norm": 2.489936351776123,
+      "learning_rate": 1.2877551020408164e-05,
+      "loss": 0.2946,
+      "step": 6980
+    },
+    {
+      "epoch": 35.66326530612245,
+      "grad_norm": 3.4419736862182617,
+      "learning_rate": 1.2867346938775511e-05,
+      "loss": 0.3306,
+      "step": 6990
+    },
+    {
+      "epoch": 35.714285714285715,
+      "grad_norm": 0.7286782264709473,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 0.1627,
+      "step": 7000
+    },
+    {
+      "epoch": 35.765306122448976,
+      "grad_norm": 8.017316818237305,
+      "learning_rate": 1.2846938775510205e-05,
+      "loss": 0.4873,
+      "step": 7010
+    },
+    {
+      "epoch": 35.816326530612244,
+      "grad_norm": 3.3010361194610596,
+      "learning_rate": 1.2836734693877552e-05,
+      "loss": 0.4093,
+      "step": 7020
+    },
+    {
+      "epoch": 35.86734693877551,
+      "grad_norm": 0.4932214319705963,
+      "learning_rate": 1.2826530612244898e-05,
+      "loss": 0.1981,
+      "step": 7030
+    },
+    {
+      "epoch": 35.91836734693877,
+      "grad_norm": 7.730387210845947,
+      "learning_rate": 1.2816326530612247e-05,
+      "loss": 0.3207,
+      "step": 7040
+    },
+    {
+      "epoch": 35.96938775510204,
+      "grad_norm": 14.77576732635498,
+      "learning_rate": 1.2806122448979591e-05,
+      "loss": 0.2905,
+      "step": 7050
+    },
+    {
+      "epoch": 36.0,
+      "eval_accuracy": 0.855595667870036,
+      "eval_loss": 0.5098311305046082,
+      "eval_runtime": 0.9311,
+      "eval_samples_per_second": 297.499,
+      "eval_steps_per_second": 37.59,
+      "step": 7056
+    },
+    {
+      "epoch": 36.02040816326531,
+      "grad_norm": 3.1446189880371094,
+      "learning_rate": 1.279591836734694e-05,
+      "loss": 0.2348,
+      "step": 7060
+    },
+    {
+      "epoch": 36.07142857142857,
+      "grad_norm": 0.9582618474960327,
+      "learning_rate": 1.2785714285714286e-05,
+      "loss": 0.339,
+      "step": 7070
+    },
+    {
+      "epoch": 36.12244897959184,
+      "grad_norm": 2.8060154914855957,
+      "learning_rate": 1.2775510204081634e-05,
+      "loss": 0.55,
+      "step": 7080
+    },
+    {
+      "epoch": 36.173469387755105,
+      "grad_norm": 3.702188014984131,
+      "learning_rate": 1.2765306122448981e-05,
+      "loss": 0.2332,
+      "step": 7090
+    },
+    {
+      "epoch": 36.224489795918366,
+      "grad_norm": 3.444847345352173,
+      "learning_rate": 1.2755102040816327e-05,
+      "loss": 0.2069,
+      "step": 7100
+    },
+    {
+      "epoch": 36.275510204081634,
+      "grad_norm": 8.948756217956543,
+      "learning_rate": 1.2744897959183674e-05,
+      "loss": 0.2245,
+      "step": 7110
+    },
+    {
+      "epoch": 36.326530612244895,
+      "grad_norm": 0.4190171957015991,
+      "learning_rate": 1.273469387755102e-05,
+      "loss": 0.2305,
+      "step": 7120
+    },
+    {
+      "epoch": 36.37755102040816,
+      "grad_norm": 1.888331651687622,
+      "learning_rate": 1.272448979591837e-05,
+      "loss": 0.2685,
+      "step": 7130
+    },
+    {
+      "epoch": 36.42857142857143,
+      "grad_norm": 2.040560722351074,
+      "learning_rate": 1.2714285714285715e-05,
+      "loss": 0.274,
+      "step": 7140
+    },
+    {
+      "epoch": 36.47959183673469,
+      "grad_norm": 3.657184362411499,
+      "learning_rate": 1.2704081632653063e-05,
+      "loss": 0.3129,
+      "step": 7150
+    },
+    {
+      "epoch": 36.53061224489796,
+      "grad_norm": 7.266042709350586,
+      "learning_rate": 1.2693877551020409e-05,
+      "loss": 0.2951,
+      "step": 7160
+    },
+    {
+      "epoch": 36.58163265306123,
+      "grad_norm": 5.278898239135742,
+      "learning_rate": 1.2683673469387756e-05,
+      "loss": 0.1573,
+      "step": 7170
+    },
+    {
+      "epoch": 36.63265306122449,
+      "grad_norm": 0.6331238746643066,
+      "learning_rate": 1.2673469387755104e-05,
+      "loss": 0.1682,
+      "step": 7180
+    },
+    {
+      "epoch": 36.683673469387756,
+      "grad_norm": 7.0703558921813965,
+      "learning_rate": 1.266326530612245e-05,
+      "loss": 0.282,
+      "step": 7190
+    },
+    {
+      "epoch": 36.734693877551024,
+      "grad_norm": 7.899667263031006,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 0.6806,
+      "step": 7200
+    },
+    {
+      "epoch": 36.785714285714285,
+      "grad_norm": 1.6253725290298462,
+      "learning_rate": 1.2642857142857143e-05,
+      "loss": 0.1931,
+      "step": 7210
+    },
+    {
+      "epoch": 36.83673469387755,
+      "grad_norm": 15.306461334228516,
+      "learning_rate": 1.2632653061224492e-05,
+      "loss": 0.4,
+      "step": 7220
+    },
+    {
+      "epoch": 36.88775510204081,
+      "grad_norm": 1.1493027210235596,
+      "learning_rate": 1.2622448979591838e-05,
+      "loss": 0.1519,
+      "step": 7230
+    },
+    {
+      "epoch": 36.93877551020408,
+      "grad_norm": 2.6926989555358887,
+      "learning_rate": 1.2612244897959185e-05,
+      "loss": 0.4129,
+      "step": 7240
+    },
+    {
+      "epoch": 36.98979591836735,
+      "grad_norm": 3.21987247467041,
+      "learning_rate": 1.260204081632653e-05,
+      "loss": 0.2763,
+      "step": 7250
+    },
+    {
+      "epoch": 37.0,
+      "eval_accuracy": 0.8808664259927798,
+      "eval_loss": 0.42190420627593994,
+      "eval_runtime": 0.9352,
+      "eval_samples_per_second": 296.199,
+      "eval_steps_per_second": 37.426,
+      "step": 7252
+    },
+    {
+      "epoch": 37.04081632653061,
+      "grad_norm": 0.3253488540649414,
+      "learning_rate": 1.2591836734693878e-05,
+      "loss": 0.0951,
+      "step": 7260
+    },
+    {
+      "epoch": 37.09183673469388,
+      "grad_norm": 0.2564552426338196,
+      "learning_rate": 1.2581632653061226e-05,
+      "loss": 0.3048,
+      "step": 7270
+    },
+    {
+      "epoch": 37.142857142857146,
+      "grad_norm": 4.657108306884766,
+      "learning_rate": 1.2571428571428572e-05,
+      "loss": 0.2368,
+      "step": 7280
+    },
+    {
+      "epoch": 37.19387755102041,
+      "grad_norm": 7.701886177062988,
+      "learning_rate": 1.256122448979592e-05,
+      "loss": 0.4069,
+      "step": 7290
+    },
+    {
+      "epoch": 37.244897959183675,
+      "grad_norm": 1.6006007194519043,
+      "learning_rate": 1.2551020408163267e-05,
+      "loss": 0.2523,
+      "step": 7300
+    },
+    {
+      "epoch": 37.295918367346935,
+      "grad_norm": 4.586967945098877,
+      "learning_rate": 1.2540816326530614e-05,
+      "loss": 0.2706,
+      "step": 7310
+    },
+    {
+      "epoch": 37.3469387755102,
+      "grad_norm": 6.338481903076172,
+      "learning_rate": 1.253061224489796e-05,
+      "loss": 0.2888,
+      "step": 7320
+    },
+    {
+      "epoch": 37.39795918367347,
+      "grad_norm": 8.174271583557129,
+      "learning_rate": 1.2520408163265307e-05,
+      "loss": 0.1741,
+      "step": 7330
+    },
+    {
+      "epoch": 37.44897959183673,
+      "grad_norm": 0.7613666653633118,
+      "learning_rate": 1.2510204081632653e-05,
+      "loss": 0.337,
+      "step": 7340
+    },
+    {
+      "epoch": 37.5,
+      "grad_norm": 9.497109413146973,
+      "learning_rate": 1.25e-05,
+      "loss": 0.4445,
+      "step": 7350
+    },
+    {
+      "epoch": 37.55102040816327,
+      "grad_norm": 6.902894973754883,
+      "learning_rate": 1.248979591836735e-05,
+      "loss": 0.4536,
+      "step": 7360
+    },
+    {
+      "epoch": 37.60204081632653,
+      "grad_norm": 0.6783469319343567,
+      "learning_rate": 1.2479591836734694e-05,
+      "loss": 0.344,
+      "step": 7370
+    },
+    {
+      "epoch": 37.6530612244898,
+      "grad_norm": 2.9475955963134766,
+      "learning_rate": 1.2469387755102043e-05,
+      "loss": 0.1476,
+      "step": 7380
+    },
+    {
+      "epoch": 37.704081632653065,
+      "grad_norm": 3.033745527267456,
+      "learning_rate": 1.2459183673469389e-05,
+      "loss": 0.2898,
+      "step": 7390
+    },
+    {
+      "epoch": 37.755102040816325,
+      "grad_norm": 1.2969194650650024,
+      "learning_rate": 1.2448979591836736e-05,
+      "loss": 0.2384,
+      "step": 7400
+    },
+    {
+      "epoch": 37.80612244897959,
+      "grad_norm": 3.2489914894104004,
+      "learning_rate": 1.2438775510204082e-05,
+      "loss": 0.1458,
+      "step": 7410
+    },
+    {
+      "epoch": 37.857142857142854,
+      "grad_norm": 0.2780137360095978,
+      "learning_rate": 1.242857142857143e-05,
+      "loss": 0.385,
+      "step": 7420
+    },
+    {
+      "epoch": 37.90816326530612,
+      "grad_norm": 1.6038283109664917,
+      "learning_rate": 1.2418367346938775e-05,
+      "loss": 0.1636,
+      "step": 7430
+    },
+    {
+      "epoch": 37.95918367346939,
+      "grad_norm": 6.315403461456299,
+      "learning_rate": 1.2408163265306123e-05,
+      "loss": 0.2454,
+      "step": 7440
+    },
+    {
+      "epoch": 38.0,
+      "eval_accuracy": 0.9133574007220217,
+      "eval_loss": 0.2851579189300537,
+      "eval_runtime": 0.9342,
+      "eval_samples_per_second": 296.517,
+      "eval_steps_per_second": 37.466,
+      "step": 7448
+    },
+    {
+      "epoch": 38.01020408163265,
+      "grad_norm": 4.942686080932617,
+      "learning_rate": 1.2397959183673472e-05,
+      "loss": 0.4006,
+      "step": 7450
+    },
+    {
+      "epoch": 38.06122448979592,
+      "grad_norm": 8.764945030212402,
+      "learning_rate": 1.2387755102040818e-05,
+      "loss": 0.3837,
+      "step": 7460
+    },
+    {
+      "epoch": 38.11224489795919,
+      "grad_norm": 8.624089241027832,
+      "learning_rate": 1.2377551020408165e-05,
+      "loss": 0.2409,
+      "step": 7470
+    },
+    {
+      "epoch": 38.16326530612245,
+      "grad_norm": 4.476467609405518,
+      "learning_rate": 1.2367346938775511e-05,
+      "loss": 0.2747,
+      "step": 7480
+    },
+    {
+      "epoch": 38.214285714285715,
+      "grad_norm": 13.317136764526367,
+      "learning_rate": 1.2357142857142859e-05,
+      "loss": 0.306,
+      "step": 7490
+    },
+    {
+      "epoch": 38.265306122448976,
+      "grad_norm": 10.13068962097168,
+      "learning_rate": 1.2346938775510204e-05,
+      "loss": 0.3725,
+      "step": 7500
+    },
+    {
+      "epoch": 38.316326530612244,
+      "grad_norm": 3.029904842376709,
+      "learning_rate": 1.2336734693877552e-05,
+      "loss": 0.2611,
+      "step": 7510
+    },
+    {
+      "epoch": 38.36734693877551,
+      "grad_norm": 5.582315921783447,
+      "learning_rate": 1.2326530612244898e-05,
+      "loss": 0.1853,
+      "step": 7520
+    },
+    {
+      "epoch": 38.41836734693877,
+      "grad_norm": 2.4034624099731445,
+      "learning_rate": 1.2316326530612245e-05,
+      "loss": 0.3121,
+      "step": 7530
+    },
+    {
+      "epoch": 38.46938775510204,
+      "grad_norm": 6.040859222412109,
+      "learning_rate": 1.2306122448979594e-05,
+      "loss": 0.373,
+      "step": 7540
+    },
+    {
+      "epoch": 38.52040816326531,
+      "grad_norm": 9.853171348571777,
+      "learning_rate": 1.229591836734694e-05,
+      "loss": 0.2384,
+      "step": 7550
+    },
+    {
+      "epoch": 38.57142857142857,
+      "grad_norm": 1.5734626054763794,
+      "learning_rate": 1.2285714285714288e-05,
+      "loss": 0.2368,
+      "step": 7560
+    },
+    {
+      "epoch": 38.62244897959184,
+      "grad_norm": 6.7449116706848145,
+      "learning_rate": 1.2275510204081633e-05,
+      "loss": 0.3006,
+      "step": 7570
+    },
+    {
+      "epoch": 38.673469387755105,
+      "grad_norm": 5.025670051574707,
+      "learning_rate": 1.2265306122448981e-05,
+      "loss": 0.3014,
+      "step": 7580
+    },
+    {
+      "epoch": 38.724489795918366,
+      "grad_norm": 0.6662245392799377,
+      "learning_rate": 1.2255102040816327e-05,
+      "loss": 0.2153,
+      "step": 7590
+    },
+    {
+      "epoch": 38.775510204081634,
+      "grad_norm": 5.983375549316406,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 0.3509,
+      "step": 7600
+    },
+    {
+      "epoch": 38.826530612244895,
+      "grad_norm": 1.9515364170074463,
+      "learning_rate": 1.223469387755102e-05,
+      "loss": 0.1339,
+      "step": 7610
+    },
+    {
+      "epoch": 38.87755102040816,
+      "grad_norm": 2.3098132610321045,
+      "learning_rate": 1.222448979591837e-05,
+      "loss": 0.0899,
+      "step": 7620
+    },
+    {
+      "epoch": 38.92857142857143,
+      "grad_norm": 4.075789451599121,
+      "learning_rate": 1.2214285714285717e-05,
+      "loss": 0.2255,
+      "step": 7630
+    },
+    {
+      "epoch": 38.97959183673469,
+      "grad_norm": 16.4624080657959,
+      "learning_rate": 1.2204081632653062e-05,
+      "loss": 0.6077,
+      "step": 7640
+    },
+    {
+      "epoch": 39.0,
+      "eval_accuracy": 0.8989169675090253,
+      "eval_loss": 0.36030298471450806,
+      "eval_runtime": 0.9376,
+      "eval_samples_per_second": 295.437,
+      "eval_steps_per_second": 37.33,
+      "step": 7644
+    },
+    {
+      "epoch": 39.03061224489796,
+      "grad_norm": 0.3404889702796936,
+      "learning_rate": 1.219387755102041e-05,
+      "loss": 0.1859,
+      "step": 7650
+    },
+    {
+      "epoch": 39.08163265306123,
+      "grad_norm": 10.761521339416504,
+      "learning_rate": 1.2183673469387756e-05,
+      "loss": 0.1371,
+      "step": 7660
+    },
+    {
+      "epoch": 39.13265306122449,
+      "grad_norm": 4.0601959228515625,
+      "learning_rate": 1.2173469387755103e-05,
+      "loss": 0.2587,
+      "step": 7670
+    },
+    {
+      "epoch": 39.183673469387756,
+      "grad_norm": 2.3847389221191406,
+      "learning_rate": 1.2163265306122449e-05,
+      "loss": 0.115,
+      "step": 7680
+    },
+    {
+      "epoch": 39.234693877551024,
+      "grad_norm": 0.9999655485153198,
+      "learning_rate": 1.2153061224489796e-05,
+      "loss": 0.456,
+      "step": 7690
+    },
+    {
+      "epoch": 39.285714285714285,
+      "grad_norm": 3.566209316253662,
+      "learning_rate": 1.2142857142857142e-05,
+      "loss": 0.2196,
+      "step": 7700
+    },
+    {
+      "epoch": 39.33673469387755,
+      "grad_norm": 6.022881031036377,
+      "learning_rate": 1.2132653061224491e-05,
+      "loss": 0.2245,
+      "step": 7710
+    },
+    {
+      "epoch": 39.38775510204081,
+      "grad_norm": 2.5836009979248047,
+      "learning_rate": 1.2122448979591839e-05,
+      "loss": 0.291,
+      "step": 7720
+    },
+    {
+      "epoch": 39.43877551020408,
+      "grad_norm": 0.4639554023742676,
+      "learning_rate": 1.2112244897959185e-05,
+      "loss": 0.3267,
+      "step": 7730
+    },
+    {
+      "epoch": 39.48979591836735,
+      "grad_norm": 4.0531511306762695,
+      "learning_rate": 1.2102040816326532e-05,
+      "loss": 0.2647,
+      "step": 7740
+    },
+    {
+      "epoch": 39.54081632653061,
+      "grad_norm": 10.204673767089844,
+      "learning_rate": 1.2091836734693878e-05,
+      "loss": 0.2902,
+      "step": 7750
+    },
+    {
+      "epoch": 39.59183673469388,
+      "grad_norm": 7.673137664794922,
+      "learning_rate": 1.2081632653061225e-05,
+      "loss": 0.2398,
+      "step": 7760
+    },
+    {
+      "epoch": 39.642857142857146,
+      "grad_norm": 6.424933910369873,
+      "learning_rate": 1.2071428571428571e-05,
+      "loss": 0.2942,
+      "step": 7770
+    },
+    {
+      "epoch": 39.69387755102041,
+      "grad_norm": 0.6079707741737366,
+      "learning_rate": 1.206122448979592e-05,
+      "loss": 0.244,
+      "step": 7780
+    },
+    {
+      "epoch": 39.744897959183675,
+      "grad_norm": 9.169522285461426,
+      "learning_rate": 1.2051020408163265e-05,
+      "loss": 0.2587,
+      "step": 7790
+    },
+    {
+      "epoch": 39.795918367346935,
+      "grad_norm": 8.324515342712402,
+      "learning_rate": 1.2040816326530614e-05,
+      "loss": 0.3697,
+      "step": 7800
+    },
+    {
+      "epoch": 39.8469387755102,
+      "grad_norm": 13.407546043395996,
+      "learning_rate": 1.2030612244897961e-05,
+      "loss": 0.5134,
+      "step": 7810
+    },
+    {
+      "epoch": 39.89795918367347,
+      "grad_norm": 1.663726568222046,
+      "learning_rate": 1.2020408163265307e-05,
+      "loss": 0.3162,
+      "step": 7820
+    },
+    {
+      "epoch": 39.94897959183673,
+      "grad_norm": 2.873859405517578,
+      "learning_rate": 1.2010204081632655e-05,
+      "loss": 0.1389,
+      "step": 7830
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 5.262413024902344,
+      "learning_rate": 1.2e-05,
+      "loss": 0.1966,
+      "step": 7840
+    },
+    {
+      "epoch": 40.0,
+      "eval_accuracy": 0.8736462093862816,
+      "eval_loss": 0.351894736289978,
+      "eval_runtime": 0.9359,
+      "eval_samples_per_second": 295.964,
+      "eval_steps_per_second": 37.396,
+      "step": 7840
+    },
+    {
+      "epoch": 40.05102040816327,
+      "grad_norm": 5.261941909790039,
+      "learning_rate": 1.1989795918367348e-05,
+      "loss": 0.1488,
+      "step": 7850
+    },
+    {
+      "epoch": 40.10204081632653,
+      "grad_norm": 1.4000771045684814,
+      "learning_rate": 1.1979591836734694e-05,
+      "loss": 0.1324,
+      "step": 7860
+    },
+    {
+      "epoch": 40.1530612244898,
+      "grad_norm": 1.8460229635238647,
+      "learning_rate": 1.1969387755102043e-05,
+      "loss": 0.2866,
+      "step": 7870
+    },
+    {
+      "epoch": 40.204081632653065,
+      "grad_norm": 1.513795256614685,
+      "learning_rate": 1.1959183673469389e-05,
+      "loss": 0.3759,
+      "step": 7880
+    },
+    {
+      "epoch": 40.255102040816325,
+      "grad_norm": 0.34185507893562317,
+      "learning_rate": 1.1948979591836736e-05,
+      "loss": 0.1073,
+      "step": 7890
+    },
+    {
+      "epoch": 40.30612244897959,
+      "grad_norm": 9.449399948120117,
+      "learning_rate": 1.1938775510204084e-05,
+      "loss": 0.4452,
+      "step": 7900
+    },
+    {
+      "epoch": 40.357142857142854,
+      "grad_norm": 8.813591003417969,
+      "learning_rate": 1.192857142857143e-05,
+      "loss": 0.3319,
+      "step": 7910
+    },
+    {
+      "epoch": 40.40816326530612,
+      "grad_norm": 0.7910783290863037,
+      "learning_rate": 1.1918367346938777e-05,
+      "loss": 0.222,
+      "step": 7920
+    },
+    {
+      "epoch": 40.45918367346939,
+      "grad_norm": 0.7371392846107483,
+      "learning_rate": 1.1908163265306123e-05,
+      "loss": 0.304,
+      "step": 7930
+    },
+    {
+      "epoch": 40.51020408163265,
+      "grad_norm": 3.1124866008758545,
+      "learning_rate": 1.189795918367347e-05,
+      "loss": 0.2276,
+      "step": 7940
+    },
+    {
+      "epoch": 40.56122448979592,
+      "grad_norm": 8.691803932189941,
+      "learning_rate": 1.1887755102040816e-05,
+      "loss": 0.4638,
+      "step": 7950
+    },
+    {
+      "epoch": 40.61224489795919,
+      "grad_norm": 6.327417850494385,
+      "learning_rate": 1.1877551020408165e-05,
+      "loss": 0.2015,
+      "step": 7960
+    },
+    {
+      "epoch": 40.66326530612245,
+      "grad_norm": 3.4650707244873047,
+      "learning_rate": 1.186734693877551e-05,
+      "loss": 0.2033,
+      "step": 7970
+    },
+    {
+      "epoch": 40.714285714285715,
+      "grad_norm": 4.530669689178467,
+      "learning_rate": 1.1857142857142858e-05,
+      "loss": 0.3811,
+      "step": 7980
+    },
+    {
+      "epoch": 40.765306122448976,
+      "grad_norm": 7.533509254455566,
+      "learning_rate": 1.1846938775510206e-05,
+      "loss": 0.632,
+      "step": 7990
+    },
+    {
+      "epoch": 40.816326530612244,
+      "grad_norm": 1.9913711547851562,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 0.1945,
+      "step": 8000
+    },
+    {
+      "epoch": 40.86734693877551,
+      "grad_norm": 3.851255178451538,
+      "learning_rate": 1.1826530612244899e-05,
+      "loss": 0.1426,
+      "step": 8010
+    },
+    {
+      "epoch": 40.91836734693877,
+      "grad_norm": 12.677950859069824,
+      "learning_rate": 1.1816326530612245e-05,
+      "loss": 0.3442,
+      "step": 8020
+    },
+    {
+      "epoch": 40.96938775510204,
+      "grad_norm": 6.756103992462158,
+      "learning_rate": 1.1806122448979594e-05,
+      "loss": 0.2473,
+      "step": 8030
+    },
+    {
+      "epoch": 41.0,
+      "eval_accuracy": 0.9025270758122743,
+      "eval_loss": 0.3343473970890045,
+      "eval_runtime": 0.9305,
+      "eval_samples_per_second": 297.683,
+      "eval_steps_per_second": 37.613,
+      "step": 8036
+    },
+    {
+      "epoch": 41.02040816326531,
+      "grad_norm": 2.6532843112945557,
+      "learning_rate": 1.179591836734694e-05,
+      "loss": 0.5848,
+      "step": 8040
+    },
+    {
+      "epoch": 41.07142857142857,
+      "grad_norm": 2.1744353771209717,
+      "learning_rate": 1.1785714285714287e-05,
+      "loss": 0.2735,
+      "step": 8050
+    },
+    {
+      "epoch": 41.12244897959184,
+      "grad_norm": 0.46514615416526794,
+      "learning_rate": 1.1775510204081633e-05,
+      "loss": 0.1303,
+      "step": 8060
+    },
+    {
+      "epoch": 41.173469387755105,
+      "grad_norm": 1.9780634641647339,
+      "learning_rate": 1.176530612244898e-05,
+      "loss": 0.2968,
+      "step": 8070
+    },
+    {
+      "epoch": 41.224489795918366,
+      "grad_norm": 2.0282843112945557,
+      "learning_rate": 1.1755102040816328e-05,
+      "loss": 0.2812,
+      "step": 8080
+    },
+    {
+      "epoch": 41.275510204081634,
+      "grad_norm": 9.81309986114502,
+      "learning_rate": 1.1744897959183674e-05,
+      "loss": 0.2277,
+      "step": 8090
+    },
+    {
+      "epoch": 41.326530612244895,
+      "grad_norm": 10.753496170043945,
+      "learning_rate": 1.1734693877551021e-05,
+      "loss": 0.3806,
+      "step": 8100
+    },
+    {
+      "epoch": 41.37755102040816,
+      "grad_norm": 8.6217041015625,
+      "learning_rate": 1.1724489795918367e-05,
+      "loss": 0.2826,
+      "step": 8110
+    },
+    {
+      "epoch": 41.42857142857143,
+      "grad_norm": 0.3755444586277008,
+      "learning_rate": 1.1714285714285716e-05,
+      "loss": 0.3433,
+      "step": 8120
+    },
+    {
+      "epoch": 41.47959183673469,
+      "grad_norm": 6.280050277709961,
+      "learning_rate": 1.1704081632653062e-05,
+      "loss": 0.3115,
+      "step": 8130
+    },
+    {
+      "epoch": 41.53061224489796,
+      "grad_norm": 5.224348068237305,
+      "learning_rate": 1.169387755102041e-05,
+      "loss": 0.181,
+      "step": 8140
+    },
+    {
+      "epoch": 41.58163265306123,
+      "grad_norm": 2.38151478767395,
+      "learning_rate": 1.1683673469387755e-05,
+      "loss": 0.2309,
+      "step": 8150
+    },
+    {
+      "epoch": 41.63265306122449,
+      "grad_norm": 2.4411683082580566,
+      "learning_rate": 1.1673469387755103e-05,
+      "loss": 0.4639,
+      "step": 8160
+    },
+    {
+      "epoch": 41.683673469387756,
+      "grad_norm": 1.6756824254989624,
+      "learning_rate": 1.166326530612245e-05,
+      "loss": 0.3495,
+      "step": 8170
+    },
+    {
+      "epoch": 41.734693877551024,
+      "grad_norm": 7.354377269744873,
+      "learning_rate": 1.1653061224489796e-05,
+      "loss": 0.1211,
+      "step": 8180
+    },
+    {
+      "epoch": 41.785714285714285,
+      "grad_norm": 4.022644519805908,
+      "learning_rate": 1.1642857142857145e-05,
+      "loss": 0.2025,
+      "step": 8190
+    },
+    {
+      "epoch": 41.83673469387755,
+      "grad_norm": 3.0944485664367676,
+      "learning_rate": 1.1632653061224491e-05,
+      "loss": 0.389,
+      "step": 8200
+    },
+    {
+      "epoch": 41.88775510204081,
+      "grad_norm": 0.8570572733879089,
+      "learning_rate": 1.1622448979591839e-05,
+      "loss": 0.1641,
+      "step": 8210
+    },
+    {
+      "epoch": 41.93877551020408,
+      "grad_norm": 6.50732946395874,
+      "learning_rate": 1.1612244897959184e-05,
+      "loss": 0.2132,
+      "step": 8220
+    },
+    {
+      "epoch": 41.98979591836735,
+      "grad_norm": 2.8220293521881104,
+      "learning_rate": 1.1602040816326532e-05,
+      "loss": 0.2795,
+      "step": 8230
+    },
+    {
+      "epoch": 42.0,
+      "eval_accuracy": 0.9169675090252708,
+      "eval_loss": 0.3383941352367401,
+      "eval_runtime": 0.9305,
+      "eval_samples_per_second": 297.686,
+      "eval_steps_per_second": 37.614,
+      "step": 8232
+    },
+    {
+      "epoch": 42.04081632653061,
+      "grad_norm": 5.730966091156006,
+      "learning_rate": 1.1591836734693878e-05,
+      "loss": 0.3407,
+      "step": 8240
+    },
+    {
+      "epoch": 42.09183673469388,
+      "grad_norm": 4.490549564361572,
+      "learning_rate": 1.1581632653061225e-05,
+      "loss": 0.2667,
+      "step": 8250
+    },
+    {
+      "epoch": 42.142857142857146,
+      "grad_norm": 1.790333867073059,
+      "learning_rate": 1.1571428571428573e-05,
+      "loss": 0.1892,
+      "step": 8260
+    },
+    {
+      "epoch": 42.19387755102041,
+      "grad_norm": 0.3163524270057678,
+      "learning_rate": 1.1561224489795918e-05,
+      "loss": 0.3141,
+      "step": 8270
+    },
+    {
+      "epoch": 42.244897959183675,
+      "grad_norm": 8.028840065002441,
+      "learning_rate": 1.1551020408163268e-05,
+      "loss": 0.0883,
+      "step": 8280
+    },
+    {
+      "epoch": 42.295918367346935,
+      "grad_norm": 1.7445170879364014,
+      "learning_rate": 1.1540816326530613e-05,
+      "loss": 0.3833,
+      "step": 8290
+    },
+    {
+      "epoch": 42.3469387755102,
+      "grad_norm": 1.6772736310958862,
+      "learning_rate": 1.1530612244897961e-05,
+      "loss": 0.3524,
+      "step": 8300
+    },
+    {
+      "epoch": 42.39795918367347,
+      "grad_norm": 11.508550643920898,
+      "learning_rate": 1.1520408163265307e-05,
+      "loss": 0.2454,
+      "step": 8310
+    },
+    {
+      "epoch": 42.44897959183673,
+      "grad_norm": 2.1842567920684814,
+      "learning_rate": 1.1510204081632654e-05,
+      "loss": 0.151,
+      "step": 8320
+    },
+    {
+      "epoch": 42.5,
+      "grad_norm": 9.214639663696289,
+      "learning_rate": 1.15e-05,
+      "loss": 0.246,
+      "step": 8330
+    },
+    {
+      "epoch": 42.55102040816327,
+      "grad_norm": 12.712059020996094,
+      "learning_rate": 1.1489795918367347e-05,
+      "loss": 0.4519,
+      "step": 8340
+    },
+    {
+      "epoch": 42.60204081632653,
+      "grad_norm": 10.930113792419434,
+      "learning_rate": 1.1479591836734697e-05,
+      "loss": 0.1964,
+      "step": 8350
+    },
+    {
+      "epoch": 42.6530612244898,
+      "grad_norm": 7.795054912567139,
+      "learning_rate": 1.146938775510204e-05,
+      "loss": 0.1999,
+      "step": 8360
+    },
+    {
+      "epoch": 42.704081632653065,
+      "grad_norm": 13.82280445098877,
+      "learning_rate": 1.145918367346939e-05,
+      "loss": 0.3961,
+      "step": 8370
+    },
+    {
+      "epoch": 42.755102040816325,
+      "grad_norm": 3.0605897903442383,
+      "learning_rate": 1.1448979591836736e-05,
+      "loss": 0.4642,
+      "step": 8380
+    },
+    {
+      "epoch": 42.80612244897959,
+      "grad_norm": 2.3108794689178467,
+      "learning_rate": 1.1438775510204083e-05,
+      "loss": 0.1843,
+      "step": 8390
+    },
+    {
+      "epoch": 42.857142857142854,
+      "grad_norm": 8.32869815826416,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.292,
+      "step": 8400
+    },
+    {
+      "epoch": 42.90816326530612,
+      "grad_norm": 3.965181827545166,
+      "learning_rate": 1.1418367346938777e-05,
+      "loss": 0.3194,
+      "step": 8410
+    },
+    {
+      "epoch": 42.95918367346939,
+      "grad_norm": 6.17874002456665,
+      "learning_rate": 1.1408163265306122e-05,
+      "loss": 0.1249,
+      "step": 8420
+    },
+    {
+      "epoch": 43.0,
+      "eval_accuracy": 0.8772563176895307,
+      "eval_loss": 0.40458711981773376,
+      "eval_runtime": 0.9348,
+      "eval_samples_per_second": 296.317,
+      "eval_steps_per_second": 37.441,
+      "step": 8428
+    },
+    {
+      "epoch": 43.01020408163265,
+      "grad_norm": 14.760079383850098,
+      "learning_rate": 1.139795918367347e-05,
+      "loss": 0.1974,
+      "step": 8430
+    },
+    {
+      "epoch": 43.06122448979592,
+      "grad_norm": 4.747530937194824,
+      "learning_rate": 1.1387755102040819e-05,
+      "loss": 0.237,
+      "step": 8440
+    },
+    {
+      "epoch": 43.11224489795919,
+      "grad_norm": 2.4478297233581543,
+      "learning_rate": 1.1377551020408165e-05,
+      "loss": 0.1684,
+      "step": 8450
+    },
+    {
+      "epoch": 43.16326530612245,
+      "grad_norm": 11.319709777832031,
+      "learning_rate": 1.1367346938775512e-05,
+      "loss": 0.328,
+      "step": 8460
+    },
+    {
+      "epoch": 43.214285714285715,
+      "grad_norm": 4.042837619781494,
+      "learning_rate": 1.1357142857142858e-05,
+      "loss": 0.3286,
+      "step": 8470
+    },
+    {
+      "epoch": 43.265306122448976,
+      "grad_norm": 2.105111598968506,
+      "learning_rate": 1.1346938775510206e-05,
+      "loss": 0.0777,
+      "step": 8480
+    },
+    {
+      "epoch": 43.316326530612244,
+      "grad_norm": 3.559504985809326,
+      "learning_rate": 1.1336734693877551e-05,
+      "loss": 0.2873,
+      "step": 8490
+    },
+    {
+      "epoch": 43.36734693877551,
+      "grad_norm": 14.684622764587402,
+      "learning_rate": 1.1326530612244899e-05,
+      "loss": 0.2682,
+      "step": 8500
+    },
+    {
+      "epoch": 43.41836734693877,
+      "grad_norm": 4.134817123413086,
+      "learning_rate": 1.1316326530612245e-05,
+      "loss": 0.2208,
+      "step": 8510
+    },
+    {
+      "epoch": 43.46938775510204,
+      "grad_norm": 5.838129997253418,
+      "learning_rate": 1.1306122448979592e-05,
+      "loss": 0.2491,
+      "step": 8520
+    },
+    {
+      "epoch": 43.52040816326531,
+      "grad_norm": 3.218062162399292,
+      "learning_rate": 1.1295918367346941e-05,
+      "loss": 0.4019,
+      "step": 8530
+    },
+    {
+      "epoch": 43.57142857142857,
+      "grad_norm": 9.169731140136719,
+      "learning_rate": 1.1285714285714287e-05,
+      "loss": 0.212,
+      "step": 8540
+    },
+    {
+      "epoch": 43.62244897959184,
+      "grad_norm": 2.351957082748413,
+      "learning_rate": 1.1275510204081635e-05,
+      "loss": 0.356,
+      "step": 8550
+    },
+    {
+      "epoch": 43.673469387755105,
+      "grad_norm": 13.376652717590332,
+      "learning_rate": 1.126530612244898e-05,
+      "loss": 0.3859,
+      "step": 8560
+    },
+    {
+      "epoch": 43.724489795918366,
+      "grad_norm": 9.093658447265625,
+      "learning_rate": 1.1255102040816328e-05,
+      "loss": 0.2787,
+      "step": 8570
+    },
+    {
+      "epoch": 43.775510204081634,
+      "grad_norm": 1.1991856098175049,
+      "learning_rate": 1.1244897959183674e-05,
+      "loss": 0.1786,
+      "step": 8580
+    },
+    {
+      "epoch": 43.826530612244895,
+      "grad_norm": 8.349421501159668,
+      "learning_rate": 1.1234693877551021e-05,
+      "loss": 0.2838,
+      "step": 8590
+    },
+    {
+      "epoch": 43.87755102040816,
+      "grad_norm": 1.6250555515289307,
+      "learning_rate": 1.1224489795918367e-05,
+      "loss": 0.1497,
+      "step": 8600
+    },
+    {
+      "epoch": 43.92857142857143,
+      "grad_norm": 11.57792854309082,
+      "learning_rate": 1.1214285714285716e-05,
+      "loss": 0.5656,
+      "step": 8610
+    },
+    {
+      "epoch": 43.97959183673469,
+      "grad_norm": 7.346845626831055,
+      "learning_rate": 1.1204081632653062e-05,
+      "loss": 0.2943,
+      "step": 8620
+    },
+    {
+      "epoch": 44.0,
+      "eval_accuracy": 0.8916967509025271,
+      "eval_loss": 0.39525121450424194,
+      "eval_runtime": 0.9323,
+      "eval_samples_per_second": 297.115,
+      "eval_steps_per_second": 37.542,
+      "step": 8624
+    },
+    {
+      "epoch": 44.03061224489796,
+      "grad_norm": 2.321767807006836,
+      "learning_rate": 1.119387755102041e-05,
+      "loss": 0.295,
+      "step": 8630
+    },
+    {
+      "epoch": 44.08163265306123,
+      "grad_norm": 10.005078315734863,
+      "learning_rate": 1.1183673469387757e-05,
+      "loss": 0.216,
+      "step": 8640
+    },
+    {
+      "epoch": 44.13265306122449,
+      "grad_norm": 10.172088623046875,
+      "learning_rate": 1.1173469387755103e-05,
+      "loss": 0.336,
+      "step": 8650
+    },
+    {
+      "epoch": 44.183673469387756,
+      "grad_norm": 9.204157829284668,
+      "learning_rate": 1.116326530612245e-05,
+      "loss": 0.1377,
+      "step": 8660
+    },
+    {
+      "epoch": 44.234693877551024,
+      "grad_norm": 0.3493322730064392,
+      "learning_rate": 1.1153061224489796e-05,
+      "loss": 0.5222,
+      "step": 8670
+    },
+    {
+      "epoch": 44.285714285714285,
+      "grad_norm": 4.95557975769043,
+      "learning_rate": 1.1142857142857143e-05,
+      "loss": 0.292,
+      "step": 8680
+    },
+    {
+      "epoch": 44.33673469387755,
+      "grad_norm": 1.297506332397461,
+      "learning_rate": 1.113265306122449e-05,
+      "loss": 0.1712,
+      "step": 8690
+    },
+    {
+      "epoch": 44.38775510204081,
+      "grad_norm": 2.6922519207000732,
+      "learning_rate": 1.1122448979591838e-05,
+      "loss": 0.2791,
+      "step": 8700
+    },
+    {
+      "epoch": 44.43877551020408,
+      "grad_norm": 2.3091373443603516,
+      "learning_rate": 1.1112244897959184e-05,
+      "loss": 0.3586,
+      "step": 8710
+    },
+    {
+      "epoch": 44.48979591836735,
+      "grad_norm": 0.26918989419937134,
+      "learning_rate": 1.1102040816326532e-05,
+      "loss": 0.3325,
+      "step": 8720
+    },
+    {
+      "epoch": 44.54081632653061,
+      "grad_norm": 11.546154975891113,
+      "learning_rate": 1.1091836734693879e-05,
+      "loss": 0.2399,
+      "step": 8730
+    },
+    {
+      "epoch": 44.59183673469388,
+      "grad_norm": 4.936864852905273,
+      "learning_rate": 1.1081632653061225e-05,
+      "loss": 0.0771,
+      "step": 8740
+    },
+    {
+      "epoch": 44.642857142857146,
+      "grad_norm": 4.6656813621521,
+      "learning_rate": 1.1071428571428572e-05,
+      "loss": 0.1901,
+      "step": 8750
+    },
+    {
+      "epoch": 44.69387755102041,
+      "grad_norm": 2.98370099067688,
+      "learning_rate": 1.1061224489795918e-05,
+      "loss": 0.2181,
+      "step": 8760
+    },
+    {
+      "epoch": 44.744897959183675,
+      "grad_norm": 11.47069263458252,
+      "learning_rate": 1.1051020408163267e-05,
+      "loss": 0.2669,
+      "step": 8770
+    },
+    {
+      "epoch": 44.795918367346935,
+      "grad_norm": 6.381840229034424,
+      "learning_rate": 1.1040816326530611e-05,
+      "loss": 0.2505,
+      "step": 8780
+    },
+    {
+      "epoch": 44.8469387755102,
+      "grad_norm": 0.29808229207992554,
+      "learning_rate": 1.103061224489796e-05,
+      "loss": 0.3411,
+      "step": 8790
+    },
+    {
+      "epoch": 44.89795918367347,
+      "grad_norm": 6.780035972595215,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 0.215,
+      "step": 8800
+    },
+    {
+      "epoch": 44.94897959183673,
+      "grad_norm": 10.969614028930664,
+      "learning_rate": 1.1010204081632654e-05,
+      "loss": 0.2879,
+      "step": 8810
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 0.764427900314331,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.3002,
+      "step": 8820
+    },
+    {
+      "epoch": 45.0,
+      "eval_accuracy": 0.8592057761732852,
+      "eval_loss": 0.5002943873405457,
+      "eval_runtime": 0.9404,
+      "eval_samples_per_second": 294.558,
+      "eval_steps_per_second": 37.218,
+      "step": 8820
+    },
+    {
+      "epoch": 45.05102040816327,
+      "grad_norm": 5.812986850738525,
+      "learning_rate": 1.0989795918367347e-05,
+      "loss": 0.15,
+      "step": 8830
+    },
+    {
+      "epoch": 45.10204081632653,
+      "grad_norm": 1.0182520151138306,
+      "learning_rate": 1.0979591836734695e-05,
+      "loss": 0.2247,
+      "step": 8840
+    },
+    {
+      "epoch": 45.1530612244898,
+      "grad_norm": 3.7717044353485107,
+      "learning_rate": 1.096938775510204e-05,
+      "loss": 0.1436,
+      "step": 8850
+    },
+    {
+      "epoch": 45.204081632653065,
+      "grad_norm": 17.78550148010254,
+      "learning_rate": 1.095918367346939e-05,
+      "loss": 0.2528,
+      "step": 8860
+    },
+    {
+      "epoch": 45.255102040816325,
+      "grad_norm": 3.487760305404663,
+      "learning_rate": 1.0948979591836735e-05,
+      "loss": 0.4551,
+      "step": 8870
+    },
+    {
+      "epoch": 45.30612244897959,
+      "grad_norm": 2.6087698936462402,
+      "learning_rate": 1.0938775510204083e-05,
+      "loss": 0.3815,
+      "step": 8880
+    },
+    {
+      "epoch": 45.357142857142854,
+      "grad_norm": 13.519532203674316,
+      "learning_rate": 1.0928571428571429e-05,
+      "loss": 0.3009,
+      "step": 8890
+    },
+    {
+      "epoch": 45.40816326530612,
+      "grad_norm": 3.4221670627593994,
+      "learning_rate": 1.0918367346938776e-05,
+      "loss": 0.3341,
+      "step": 8900
+    },
+    {
+      "epoch": 45.45918367346939,
+      "grad_norm": 0.41899165511131287,
+      "learning_rate": 1.0908163265306124e-05,
+      "loss": 0.0961,
+      "step": 8910
+    },
+    {
+      "epoch": 45.51020408163265,
+      "grad_norm": 1.055353045463562,
+      "learning_rate": 1.089795918367347e-05,
+      "loss": 0.4265,
+      "step": 8920
+    },
+    {
+      "epoch": 45.56122448979592,
+      "grad_norm": 1.2329767942428589,
+      "learning_rate": 1.0887755102040819e-05,
+      "loss": 0.3215,
+      "step": 8930
+    },
+    {
+      "epoch": 45.61224489795919,
+      "grad_norm": 3.0227720737457275,
+      "learning_rate": 1.0877551020408163e-05,
+      "loss": 0.502,
+      "step": 8940
+    },
+    {
+      "epoch": 45.66326530612245,
+      "grad_norm": 4.809809684753418,
+      "learning_rate": 1.0867346938775512e-05,
+      "loss": 0.3271,
+      "step": 8950
+    },
+    {
+      "epoch": 45.714285714285715,
+      "grad_norm": 3.6863017082214355,
+      "learning_rate": 1.0857142857142858e-05,
+      "loss": 0.3246,
+      "step": 8960
+    },
+    {
+      "epoch": 45.765306122448976,
+      "grad_norm": 7.67105770111084,
+      "learning_rate": 1.0846938775510205e-05,
+      "loss": 0.2358,
+      "step": 8970
+    },
+    {
+      "epoch": 45.816326530612244,
+      "grad_norm": 1.7396820783615112,
+      "learning_rate": 1.0836734693877551e-05,
+      "loss": 0.2026,
+      "step": 8980
+    },
+    {
+      "epoch": 45.86734693877551,
+      "grad_norm": 2.3171074390411377,
+      "learning_rate": 1.0826530612244899e-05,
+      "loss": 0.148,
+      "step": 8990
+    },
+    {
+      "epoch": 45.91836734693877,
+      "grad_norm": 4.390339374542236,
+      "learning_rate": 1.0816326530612246e-05,
+      "loss": 0.2545,
+      "step": 9000
+    },
+    {
+      "epoch": 45.96938775510204,
+      "grad_norm": 13.35239028930664,
+      "learning_rate": 1.0806122448979592e-05,
+      "loss": 0.1525,
+      "step": 9010
+    },
+    {
+      "epoch": 46.0,
+      "eval_accuracy": 0.9169675090252708,
+      "eval_loss": 0.3231888711452484,
+      "eval_runtime": 0.9348,
+      "eval_samples_per_second": 296.308,
+      "eval_steps_per_second": 37.44,
+      "step": 9016
+    },
+    {
+      "epoch": 46.02040816326531,
+      "grad_norm": 19.203357696533203,
+      "learning_rate": 1.0795918367346941e-05,
+      "loss": 0.4035,
+      "step": 9020
+    },
+    {
+      "epoch": 46.07142857142857,
+      "grad_norm": 2.2039167881011963,
+      "learning_rate": 1.0785714285714287e-05,
+      "loss": 0.3711,
+      "step": 9030
+    },
+    {
+      "epoch": 46.12244897959184,
+      "grad_norm": 1.917140007019043,
+      "learning_rate": 1.0775510204081634e-05,
+      "loss": 0.2079,
+      "step": 9040
+    },
+    {
+      "epoch": 46.173469387755105,
+      "grad_norm": 9.410566329956055,
+      "learning_rate": 1.076530612244898e-05,
+      "loss": 0.2509,
+      "step": 9050
+    },
+    {
+      "epoch": 46.224489795918366,
+      "grad_norm": 4.123296737670898,
+      "learning_rate": 1.0755102040816328e-05,
+      "loss": 0.4548,
+      "step": 9060
+    },
+    {
+      "epoch": 46.275510204081634,
+      "grad_norm": 2.996526002883911,
+      "learning_rate": 1.0744897959183673e-05,
+      "loss": 0.1775,
+      "step": 9070
+    },
+    {
+      "epoch": 46.326530612244895,
+      "grad_norm": 6.14008903503418,
+      "learning_rate": 1.073469387755102e-05,
+      "loss": 0.1807,
+      "step": 9080
+    },
+    {
+      "epoch": 46.37755102040816,
+      "grad_norm": 5.149447441101074,
+      "learning_rate": 1.072448979591837e-05,
+      "loss": 0.2278,
+      "step": 9090
+    },
+    {
+      "epoch": 46.42857142857143,
+      "grad_norm": 0.6238847374916077,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.2204,
+      "step": 9100
+    },
+    {
+      "epoch": 46.47959183673469,
+      "grad_norm": 8.696666717529297,
+      "learning_rate": 1.0704081632653063e-05,
+      "loss": 0.2924,
+      "step": 9110
+    },
+    {
+      "epoch": 46.53061224489796,
+      "grad_norm": 7.7894978523254395,
+      "learning_rate": 1.0693877551020409e-05,
+      "loss": 0.2058,
+      "step": 9120
+    },
+    {
+      "epoch": 46.58163265306123,
+      "grad_norm": 3.7465615272521973,
+      "learning_rate": 1.0683673469387757e-05,
+      "loss": 0.4511,
+      "step": 9130
+    },
+    {
+      "epoch": 46.63265306122449,
+      "grad_norm": 8.284019470214844,
+      "learning_rate": 1.0673469387755102e-05,
+      "loss": 0.2944,
+      "step": 9140
+    },
+    {
+      "epoch": 46.683673469387756,
+      "grad_norm": 4.210254192352295,
+      "learning_rate": 1.066326530612245e-05,
+      "loss": 0.1609,
+      "step": 9150
+    },
+    {
+      "epoch": 46.734693877551024,
+      "grad_norm": 1.5888322591781616,
+      "learning_rate": 1.0653061224489796e-05,
+      "loss": 0.1958,
+      "step": 9160
+    },
+    {
+      "epoch": 46.785714285714285,
+      "grad_norm": 10.489404678344727,
+      "learning_rate": 1.0642857142857143e-05,
+      "loss": 0.2796,
+      "step": 9170
+    },
+    {
+      "epoch": 46.83673469387755,
+      "grad_norm": 1.8088388442993164,
+      "learning_rate": 1.0632653061224492e-05,
+      "loss": 0.1466,
+      "step": 9180
+    },
+    {
+      "epoch": 46.88775510204081,
+      "grad_norm": 4.326778888702393,
+      "learning_rate": 1.0622448979591838e-05,
+      "loss": 0.2615,
+      "step": 9190
+    },
+    {
+      "epoch": 46.93877551020408,
+      "grad_norm": 6.9802632331848145,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 0.3217,
+      "step": 9200
+    },
+    {
+      "epoch": 46.98979591836735,
+      "grad_norm": 9.483171463012695,
+      "learning_rate": 1.0602040816326531e-05,
+      "loss": 0.4022,
+      "step": 9210
+    },
+    {
+      "epoch": 47.0,
+      "eval_accuracy": 0.9169675090252708,
+      "eval_loss": 0.3112559914588928,
+      "eval_runtime": 1.0011,
+      "eval_samples_per_second": 276.691,
+      "eval_steps_per_second": 34.961,
+      "step": 9212
+    },
+    {
+      "epoch": 47.04081632653061,
+      "grad_norm": 5.846846103668213,
+      "learning_rate": 1.0591836734693879e-05,
+      "loss": 0.2519,
+      "step": 9220
+    },
+    {
+      "epoch": 47.09183673469388,
+      "grad_norm": 12.650550842285156,
+      "learning_rate": 1.0581632653061225e-05,
+      "loss": 0.1602,
+      "step": 9230
+    },
+    {
+      "epoch": 47.142857142857146,
+      "grad_norm": 6.220149993896484,
+      "learning_rate": 1.0571428571428572e-05,
+      "loss": 0.3288,
+      "step": 9240
+    },
+    {
+      "epoch": 47.19387755102041,
+      "grad_norm": 3.168201208114624,
+      "learning_rate": 1.0561224489795918e-05,
+      "loss": 0.1582,
+      "step": 9250
+    },
+    {
+      "epoch": 47.244897959183675,
+      "grad_norm": 2.817613124847412,
+      "learning_rate": 1.0551020408163265e-05,
+      "loss": 0.3362,
+      "step": 9260
+    },
+    {
+      "epoch": 47.295918367346935,
+      "grad_norm": 2.279808521270752,
+      "learning_rate": 1.0540816326530615e-05,
+      "loss": 0.3093,
+      "step": 9270
+    },
+    {
+      "epoch": 47.3469387755102,
+      "grad_norm": 9.442809104919434,
+      "learning_rate": 1.053061224489796e-05,
+      "loss": 0.4158,
+      "step": 9280
+    },
+    {
+      "epoch": 47.39795918367347,
+      "grad_norm": 7.1183342933654785,
+      "learning_rate": 1.0520408163265308e-05,
+      "loss": 0.1822,
+      "step": 9290
+    },
+    {
+      "epoch": 47.44897959183673,
+      "grad_norm": 1.518853783607483,
+      "learning_rate": 1.0510204081632654e-05,
+      "loss": 0.1303,
+      "step": 9300
+    },
+    {
+      "epoch": 47.5,
+      "grad_norm": 5.1223602294921875,
+      "learning_rate": 1.0500000000000001e-05,
+      "loss": 0.6394,
+      "step": 9310
+    },
+    {
+      "epoch": 47.55102040816327,
+      "grad_norm": 8.16996955871582,
+      "learning_rate": 1.0489795918367347e-05,
+      "loss": 0.286,
+      "step": 9320
+    },
+    {
+      "epoch": 47.60204081632653,
+      "grad_norm": 7.438451766967773,
+      "learning_rate": 1.0479591836734694e-05,
+      "loss": 0.2902,
+      "step": 9330
+    },
+    {
+      "epoch": 47.6530612244898,
+      "grad_norm": 9.90367603302002,
+      "learning_rate": 1.046938775510204e-05,
+      "loss": 0.1833,
+      "step": 9340
+    },
+    {
+      "epoch": 47.704081632653065,
+      "grad_norm": 1.0031007528305054,
+      "learning_rate": 1.045918367346939e-05,
+      "loss": 0.27,
+      "step": 9350
+    },
+    {
+      "epoch": 47.755102040816325,
+      "grad_norm": 5.630500793457031,
+      "learning_rate": 1.0448979591836737e-05,
+      "loss": 0.2838,
+      "step": 9360
+    },
+    {
+      "epoch": 47.80612244897959,
+      "grad_norm": 3.1275367736816406,
+      "learning_rate": 1.0438775510204083e-05,
+      "loss": 0.1895,
+      "step": 9370
+    },
+    {
+      "epoch": 47.857142857142854,
+      "grad_norm": 4.408709526062012,
+      "learning_rate": 1.042857142857143e-05,
+      "loss": 0.234,
+      "step": 9380
+    },
+    {
+      "epoch": 47.90816326530612,
+      "grad_norm": 1.6241774559020996,
+      "learning_rate": 1.0418367346938776e-05,
+      "loss": 0.0964,
+      "step": 9390
+    },
+    {
+      "epoch": 47.95918367346939,
+      "grad_norm": 13.035674095153809,
+      "learning_rate": 1.0408163265306123e-05,
+      "loss": 0.4994,
+      "step": 9400
+    },
+    {
+      "epoch": 48.0,
+      "eval_accuracy": 0.855595667870036,
+      "eval_loss": 0.4494384229183197,
+      "eval_runtime": 0.943,
+      "eval_samples_per_second": 293.731,
+      "eval_steps_per_second": 37.114,
+      "step": 9408
+    },
+    {
+      "epoch": 48.01020408163265,
+      "grad_norm": 2.8087539672851562,
+      "learning_rate": 1.039795918367347e-05,
+      "loss": 0.2604,
+      "step": 9410
+    },
+    {
+      "epoch": 48.06122448979592,
+      "grad_norm": 2.577768325805664,
+      "learning_rate": 1.0387755102040817e-05,
+      "loss": 0.2844,
+      "step": 9420
+    },
+    {
+      "epoch": 48.11224489795919,
+      "grad_norm": 11.842287063598633,
+      "learning_rate": 1.0377551020408162e-05,
+      "loss": 0.3377,
+      "step": 9430
+    },
+    {
+      "epoch": 48.16326530612245,
+      "grad_norm": 14.558892250061035,
+      "learning_rate": 1.0367346938775512e-05,
+      "loss": 0.2846,
+      "step": 9440
+    },
+    {
+      "epoch": 48.214285714285715,
+      "grad_norm": 2.9422316551208496,
+      "learning_rate": 1.0357142857142859e-05,
+      "loss": 0.2325,
+      "step": 9450
+    },
+    {
+      "epoch": 48.265306122448976,
+      "grad_norm": 2.472435235977173,
+      "learning_rate": 1.0346938775510205e-05,
+      "loss": 0.1945,
+      "step": 9460
+    },
+    {
+      "epoch": 48.316326530612244,
+      "grad_norm": 6.785258769989014,
+      "learning_rate": 1.0336734693877552e-05,
+      "loss": 0.2503,
+      "step": 9470
+    },
+    {
+      "epoch": 48.36734693877551,
+      "grad_norm": 1.3686105012893677,
+      "learning_rate": 1.0326530612244898e-05,
+      "loss": 0.3424,
+      "step": 9480
+    },
+    {
+      "epoch": 48.41836734693877,
+      "grad_norm": 2.2825608253479004,
+      "learning_rate": 1.0316326530612246e-05,
+      "loss": 0.1742,
+      "step": 9490
+    },
+    {
+      "epoch": 48.46938775510204,
+      "grad_norm": 8.19665813446045,
+      "learning_rate": 1.0306122448979591e-05,
+      "loss": 0.2663,
+      "step": 9500
+    },
+    {
+      "epoch": 48.52040816326531,
+      "grad_norm": 5.659435272216797,
+      "learning_rate": 1.029591836734694e-05,
+      "loss": 0.2131,
+      "step": 9510
+    },
+    {
+      "epoch": 48.57142857142857,
+      "grad_norm": 6.981075763702393,
+      "learning_rate": 1.0285714285714285e-05,
+      "loss": 0.209,
+      "step": 9520
+    },
+    {
+      "epoch": 48.62244897959184,
+      "grad_norm": 5.7839202880859375,
+      "learning_rate": 1.0275510204081634e-05,
+      "loss": 0.1524,
+      "step": 9530
+    },
+    {
+      "epoch": 48.673469387755105,
+      "grad_norm": 6.4543890953063965,
+      "learning_rate": 1.0265306122448981e-05,
+      "loss": 0.332,
+      "step": 9540
+    },
+    {
+      "epoch": 48.724489795918366,
+      "grad_norm": 1.987481713294983,
+      "learning_rate": 1.0255102040816327e-05,
+      "loss": 0.2465,
+      "step": 9550
+    },
+    {
+      "epoch": 48.775510204081634,
+      "grad_norm": 4.892277717590332,
+      "learning_rate": 1.0244897959183675e-05,
+      "loss": 0.3518,
+      "step": 9560
+    },
+    {
+      "epoch": 48.826530612244895,
+      "grad_norm": 6.381310939788818,
+      "learning_rate": 1.023469387755102e-05,
+      "loss": 0.2743,
+      "step": 9570
+    },
+    {
+      "epoch": 48.87755102040816,
+      "grad_norm": 8.336471557617188,
+      "learning_rate": 1.0224489795918368e-05,
+      "loss": 0.4257,
+      "step": 9580
+    },
+    {
+      "epoch": 48.92857142857143,
+      "grad_norm": 1.3258341550827026,
+      "learning_rate": 1.0214285714285714e-05,
+      "loss": 0.055,
+      "step": 9590
+    },
+    {
+      "epoch": 48.97959183673469,
+      "grad_norm": 10.289790153503418,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 0.6512,
+      "step": 9600
+    },
+    {
+      "epoch": 49.0,
+      "eval_accuracy": 0.9205776173285198,
+      "eval_loss": 0.37217509746551514,
+      "eval_runtime": 0.9375,
+      "eval_samples_per_second": 295.453,
+      "eval_steps_per_second": 37.332,
+      "step": 9604
+    },
+    {
+      "epoch": 49.03061224489796,
+      "grad_norm": 2.5601608753204346,
+      "learning_rate": 1.0193877551020409e-05,
+      "loss": 0.2695,
+      "step": 9610
+    },
+    {
+      "epoch": 49.08163265306123,
+      "grad_norm": 8.381750106811523,
+      "learning_rate": 1.0183673469387756e-05,
+      "loss": 0.3243,
+      "step": 9620
+    },
+    {
+      "epoch": 49.13265306122449,
+      "grad_norm": 9.810985565185547,
+      "learning_rate": 1.0173469387755104e-05,
+      "loss": 0.3454,
+      "step": 9630
+    },
+    {
+      "epoch": 49.183673469387756,
+      "grad_norm": 10.479945182800293,
+      "learning_rate": 1.016326530612245e-05,
+      "loss": 0.324,
+      "step": 9640
+    },
+    {
+      "epoch": 49.234693877551024,
+      "grad_norm": 9.079980850219727,
+      "learning_rate": 1.0153061224489797e-05,
+      "loss": 0.427,
+      "step": 9650
+    },
+    {
+      "epoch": 49.285714285714285,
+      "grad_norm": 11.63463020324707,
+      "learning_rate": 1.0142857142857143e-05,
+      "loss": 0.3591,
+      "step": 9660
+    },
+    {
+      "epoch": 49.33673469387755,
+      "grad_norm": 2.4312291145324707,
+      "learning_rate": 1.013265306122449e-05,
+      "loss": 0.2909,
+      "step": 9670
+    },
+    {
+      "epoch": 49.38775510204081,
+      "grad_norm": 1.2564700841903687,
+      "learning_rate": 1.0122448979591836e-05,
+      "loss": 0.1863,
+      "step": 9680
+    },
+    {
+      "epoch": 49.43877551020408,
+      "grad_norm": 5.644169330596924,
+      "learning_rate": 1.0112244897959185e-05,
+      "loss": 0.2788,
+      "step": 9690
+    },
+    {
+      "epoch": 49.48979591836735,
+      "grad_norm": 6.595949172973633,
+      "learning_rate": 1.0102040816326531e-05,
+      "loss": 0.3373,
+      "step": 9700
+    },
+    {
+      "epoch": 49.54081632653061,
+      "grad_norm": 0.5495048761367798,
+      "learning_rate": 1.0091836734693879e-05,
+      "loss": 0.1776,
+      "step": 9710
+    },
+    {
+      "epoch": 49.59183673469388,
+      "grad_norm": 11.747842788696289,
+      "learning_rate": 1.0081632653061226e-05,
+      "loss": 0.352,
+      "step": 9720
+    },
+    {
+      "epoch": 49.642857142857146,
+      "grad_norm": 2.48116135597229,
+      "learning_rate": 1.0071428571428572e-05,
+      "loss": 0.1722,
+      "step": 9730
+    },
+    {
+      "epoch": 49.69387755102041,
+      "grad_norm": 8.725930213928223,
+      "learning_rate": 1.006122448979592e-05,
+      "loss": 0.1544,
+      "step": 9740
+    },
+    {
+      "epoch": 49.744897959183675,
+      "grad_norm": 1.277961254119873,
+      "learning_rate": 1.0051020408163265e-05,
+      "loss": 0.2064,
+      "step": 9750
+    },
+    {
+      "epoch": 49.795918367346935,
+      "grad_norm": 0.349727064371109,
+      "learning_rate": 1.0040816326530614e-05,
+      "loss": 0.1728,
+      "step": 9760
+    },
+    {
+      "epoch": 49.8469387755102,
+      "grad_norm": 2.0223069190979004,
+      "learning_rate": 1.003061224489796e-05,
+      "loss": 0.4544,
+      "step": 9770
+    },
+    {
+      "epoch": 49.89795918367347,
+      "grad_norm": 0.8309898972511292,
+      "learning_rate": 1.0020408163265308e-05,
+      "loss": 0.1539,
+      "step": 9780
+    },
+    {
+      "epoch": 49.94897959183673,
+      "grad_norm": 2.0115163326263428,
+      "learning_rate": 1.0010204081632653e-05,
+      "loss": 0.4388,
+      "step": 9790
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 7.568119525909424,
+      "learning_rate": 1e-05,
+      "loss": 0.3152,
+      "step": 9800
+    },
+    {
+      "epoch": 50.0,
+      "eval_accuracy": 0.9097472924187726,
+      "eval_loss": 0.2852139472961426,
+      "eval_runtime": 0.9346,
+      "eval_samples_per_second": 296.382,
+      "eval_steps_per_second": 37.449,
+      "step": 9800
+    },
+    {
+      "epoch": 50.05102040816327,
+      "grad_norm": 1.160243034362793,
+      "learning_rate": 9.989795918367348e-06,
+      "loss": 0.2282,
+      "step": 9810
+    },
+    {
+      "epoch": 50.10204081632653,
+      "grad_norm": 13.818071365356445,
+      "learning_rate": 9.979591836734694e-06,
+      "loss": 0.2916,
+      "step": 9820
+    },
+    {
+      "epoch": 50.1530612244898,
+      "grad_norm": 4.897973537445068,
+      "learning_rate": 9.969387755102042e-06,
+      "loss": 0.2524,
+      "step": 9830
+    },
+    {
+      "epoch": 50.204081632653065,
+      "grad_norm": 5.517409324645996,
+      "learning_rate": 9.959183673469387e-06,
+      "loss": 0.2813,
+      "step": 9840
+    },
+    {
+      "epoch": 50.255102040816325,
+      "grad_norm": 6.2214274406433105,
+      "learning_rate": 9.948979591836737e-06,
+      "loss": 0.2462,
+      "step": 9850
+    },
+    {
+      "epoch": 50.30612244897959,
+      "grad_norm": 0.3451865315437317,
+      "learning_rate": 9.938775510204082e-06,
+      "loss": 0.2339,
+      "step": 9860
+    },
+    {
+      "epoch": 50.357142857142854,
+      "grad_norm": 6.983676433563232,
+      "learning_rate": 9.92857142857143e-06,
+      "loss": 0.1045,
+      "step": 9870
+    },
+    {
+      "epoch": 50.40816326530612,
+      "grad_norm": 1.0857036113739014,
+      "learning_rate": 9.918367346938776e-06,
+      "loss": 0.0822,
+      "step": 9880
+    },
+    {
+      "epoch": 50.45918367346939,
+      "grad_norm": 0.8820812702178955,
+      "learning_rate": 9.908163265306123e-06,
+      "loss": 0.2875,
+      "step": 9890
+    },
+    {
+      "epoch": 50.51020408163265,
+      "grad_norm": 4.251252174377441,
+      "learning_rate": 9.89795918367347e-06,
+      "loss": 0.2354,
+      "step": 9900
+    },
+    {
+      "epoch": 50.56122448979592,
+      "grad_norm": 17.90270233154297,
+      "learning_rate": 9.887755102040816e-06,
+      "loss": 0.4852,
+      "step": 9910
+    },
+    {
+      "epoch": 50.61224489795919,
+      "grad_norm": 11.787466049194336,
+      "learning_rate": 9.877551020408164e-06,
+      "loss": 0.4315,
+      "step": 9920
+    },
+    {
+      "epoch": 50.66326530612245,
+      "grad_norm": 3.597476005554199,
+      "learning_rate": 9.867346938775511e-06,
+      "loss": 0.2477,
+      "step": 9930
+    },
+    {
+      "epoch": 50.714285714285715,
+      "grad_norm": 2.140202522277832,
+      "learning_rate": 9.857142857142859e-06,
+      "loss": 0.1874,
+      "step": 9940
+    },
+    {
+      "epoch": 50.765306122448976,
+      "grad_norm": 11.62266731262207,
+      "learning_rate": 9.846938775510205e-06,
+      "loss": 0.3983,
+      "step": 9950
+    },
+    {
+      "epoch": 50.816326530612244,
+      "grad_norm": 11.113944053649902,
+      "learning_rate": 9.836734693877552e-06,
+      "loss": 0.4397,
+      "step": 9960
+    },
+    {
+      "epoch": 50.86734693877551,
+      "grad_norm": 9.346400260925293,
+      "learning_rate": 9.8265306122449e-06,
+      "loss": 0.2636,
+      "step": 9970
+    },
+    {
+      "epoch": 50.91836734693877,
+      "grad_norm": 1.1383917331695557,
+      "learning_rate": 9.816326530612245e-06,
+      "loss": 0.2994,
+      "step": 9980
+    },
+    {
+      "epoch": 50.96938775510204,
+      "grad_norm": 6.278914451599121,
+      "learning_rate": 9.806122448979593e-06,
+      "loss": 0.1165,
+      "step": 9990
+    },
+    {
+      "epoch": 51.0,
+      "eval_accuracy": 0.8628158844765343,
+      "eval_loss": 0.4137951135635376,
+      "eval_runtime": 0.9343,
+      "eval_samples_per_second": 296.489,
+      "eval_steps_per_second": 37.463,
+      "step": 9996
+    },
+    {
+      "epoch": 51.02040816326531,
+      "grad_norm": 5.278802871704102,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 0.251,
+      "step": 10000
+    },
+    {
+      "epoch": 51.07142857142857,
+      "grad_norm": 10.420355796813965,
+      "learning_rate": 9.785714285714286e-06,
+      "loss": 0.1771,
+      "step": 10010
+    },
+    {
+      "epoch": 51.12244897959184,
+      "grad_norm": 14.361730575561523,
+      "learning_rate": 9.775510204081634e-06,
+      "loss": 0.3886,
+      "step": 10020
+    },
+    {
+      "epoch": 51.173469387755105,
+      "grad_norm": 7.604317665100098,
+      "learning_rate": 9.765306122448981e-06,
+      "loss": 0.1775,
+      "step": 10030
+    },
+    {
+      "epoch": 51.224489795918366,
+      "grad_norm": 1.9585540294647217,
+      "learning_rate": 9.755102040816327e-06,
+      "loss": 0.2672,
+      "step": 10040
+    },
+    {
+      "epoch": 51.275510204081634,
+      "grad_norm": 1.5716756582260132,
+      "learning_rate": 9.744897959183674e-06,
+      "loss": 0.093,
+      "step": 10050
+    },
+    {
+      "epoch": 51.326530612244895,
+      "grad_norm": 10.107017517089844,
+      "learning_rate": 9.734693877551022e-06,
+      "loss": 0.4693,
+      "step": 10060
+    },
+    {
+      "epoch": 51.37755102040816,
+      "grad_norm": 2.6521267890930176,
+      "learning_rate": 9.724489795918368e-06,
+      "loss": 0.2922,
+      "step": 10070
+    },
+    {
+      "epoch": 51.42857142857143,
+      "grad_norm": 3.5833919048309326,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 0.1893,
+      "step": 10080
+    },
+    {
+      "epoch": 51.47959183673469,
+      "grad_norm": 3.322532892227173,
+      "learning_rate": 9.704081632653061e-06,
+      "loss": 0.1881,
+      "step": 10090
+    },
+    {
+      "epoch": 51.53061224489796,
+      "grad_norm": 12.746216773986816,
+      "learning_rate": 9.693877551020408e-06,
+      "loss": 0.4767,
+      "step": 10100
+    },
+    {
+      "epoch": 51.58163265306123,
+      "grad_norm": 4.599057674407959,
+      "learning_rate": 9.683673469387756e-06,
+      "loss": 0.2251,
+      "step": 10110
+    },
+    {
+      "epoch": 51.63265306122449,
+      "grad_norm": 11.859138488769531,
+      "learning_rate": 9.673469387755103e-06,
+      "loss": 0.2792,
+      "step": 10120
+    },
+    {
+      "epoch": 51.683673469387756,
+      "grad_norm": 0.5115078091621399,
+      "learning_rate": 9.663265306122451e-06,
+      "loss": 0.0973,
+      "step": 10130
+    },
+    {
+      "epoch": 51.734693877551024,
+      "grad_norm": 1.4567404985427856,
+      "learning_rate": 9.653061224489797e-06,
+      "loss": 0.1249,
+      "step": 10140
+    },
+    {
+      "epoch": 51.785714285714285,
+      "grad_norm": 1.8316110372543335,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.2642,
+      "step": 10150
+    },
+    {
+      "epoch": 51.83673469387755,
+      "grad_norm": 7.921043395996094,
+      "learning_rate": 9.63265306122449e-06,
+      "loss": 0.2997,
+      "step": 10160
+    },
+    {
+      "epoch": 51.88775510204081,
+      "grad_norm": 0.15132638812065125,
+      "learning_rate": 9.622448979591837e-06,
+      "loss": 0.2621,
+      "step": 10170
+    },
+    {
+      "epoch": 51.93877551020408,
+      "grad_norm": 3.1799213886260986,
+      "learning_rate": 9.612244897959185e-06,
+      "loss": 0.4857,
+      "step": 10180
+    },
+    {
+      "epoch": 51.98979591836735,
+      "grad_norm": 6.599296569824219,
+      "learning_rate": 9.60204081632653e-06,
+      "loss": 0.216,
+      "step": 10190
+    },
+    {
+      "epoch": 52.0,
+      "eval_accuracy": 0.8953068592057761,
+      "eval_loss": 0.34130236506462097,
+      "eval_runtime": 0.9456,
+      "eval_samples_per_second": 292.95,
+      "eval_steps_per_second": 37.015,
+      "step": 10192
+    },
+    {
+      "epoch": 52.04081632653061,
+      "grad_norm": 0.39954033493995667,
+      "learning_rate": 9.591836734693878e-06,
+      "loss": 0.2164,
+      "step": 10200
+    },
+    {
+      "epoch": 52.09183673469388,
+      "grad_norm": 0.30738261342048645,
+      "learning_rate": 9.581632653061226e-06,
+      "loss": 0.314,
+      "step": 10210
+    },
+    {
+      "epoch": 52.142857142857146,
+      "grad_norm": 4.521490573883057,
+      "learning_rate": 9.571428571428573e-06,
+      "loss": 0.4708,
+      "step": 10220
+    },
+    {
+      "epoch": 52.19387755102041,
+      "grad_norm": 8.1461181640625,
+      "learning_rate": 9.561224489795919e-06,
+      "loss": 0.3965,
+      "step": 10230
+    },
+    {
+      "epoch": 52.244897959183675,
+      "grad_norm": 8.73324966430664,
+      "learning_rate": 9.551020408163266e-06,
+      "loss": 0.2251,
+      "step": 10240
+    },
+    {
+      "epoch": 52.295918367346935,
+      "grad_norm": 5.212022304534912,
+      "learning_rate": 9.540816326530612e-06,
+      "loss": 0.0885,
+      "step": 10250
+    },
+    {
+      "epoch": 52.3469387755102,
+      "grad_norm": 3.9524590969085693,
+      "learning_rate": 9.53061224489796e-06,
+      "loss": 0.2023,
+      "step": 10260
+    },
+    {
+      "epoch": 52.39795918367347,
+      "grad_norm": 6.595174789428711,
+      "learning_rate": 9.520408163265307e-06,
+      "loss": 0.1592,
+      "step": 10270
+    },
+    {
+      "epoch": 52.44897959183673,
+      "grad_norm": 10.941976547241211,
+      "learning_rate": 9.510204081632653e-06,
+      "loss": 0.4717,
+      "step": 10280
+    },
+    {
+      "epoch": 52.5,
+      "grad_norm": 8.068604469299316,
+      "learning_rate": 9.5e-06,
+      "loss": 0.3821,
+      "step": 10290
+    },
+    {
+      "epoch": 52.55102040816327,
+      "grad_norm": 6.500422477722168,
+      "learning_rate": 9.489795918367348e-06,
+      "loss": 0.171,
+      "step": 10300
+    },
+    {
+      "epoch": 52.60204081632653,
+      "grad_norm": 18.862512588500977,
+      "learning_rate": 9.479591836734695e-06,
+      "loss": 0.3375,
+      "step": 10310
+    },
+    {
+      "epoch": 52.6530612244898,
+      "grad_norm": 4.508473873138428,
+      "learning_rate": 9.469387755102041e-06,
+      "loss": 0.2942,
+      "step": 10320
+    },
+    {
+      "epoch": 52.704081632653065,
+      "grad_norm": 2.781569004058838,
+      "learning_rate": 9.459183673469389e-06,
+      "loss": 0.2091,
+      "step": 10330
+    },
+    {
+      "epoch": 52.755102040816325,
+      "grad_norm": 14.221521377563477,
+      "learning_rate": 9.448979591836736e-06,
+      "loss": 0.3069,
+      "step": 10340
+    },
+    {
+      "epoch": 52.80612244897959,
+      "grad_norm": 5.661198139190674,
+      "learning_rate": 9.438775510204082e-06,
+      "loss": 0.2134,
+      "step": 10350
+    },
+    {
+      "epoch": 52.857142857142854,
+      "grad_norm": 10.020429611206055,
+      "learning_rate": 9.42857142857143e-06,
+      "loss": 0.2813,
+      "step": 10360
+    },
+    {
+      "epoch": 52.90816326530612,
+      "grad_norm": 10.018217086791992,
+      "learning_rate": 9.418367346938775e-06,
+      "loss": 0.3867,
+      "step": 10370
+    },
+    {
+      "epoch": 52.95918367346939,
+      "grad_norm": 0.26604777574539185,
+      "learning_rate": 9.408163265306123e-06,
+      "loss": 0.1455,
+      "step": 10380
+    },
+    {
+      "epoch": 53.0,
+      "eval_accuracy": 0.9169675090252708,
+      "eval_loss": 0.30462250113487244,
+      "eval_runtime": 0.9334,
+      "eval_samples_per_second": 296.752,
+      "eval_steps_per_second": 37.496,
+      "step": 10388
+    },
+    {
+      "epoch": 53.01020408163265,
+      "grad_norm": 2.2215254306793213,
+      "learning_rate": 9.39795918367347e-06,
+      "loss": 0.418,
+      "step": 10390
+    },
+    {
+      "epoch": 53.06122448979592,
+      "grad_norm": 5.510087966918945,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 0.362,
+      "step": 10400
+    },
+    {
+      "epoch": 53.11224489795919,
+      "grad_norm": 1.2044073343276978,
+      "learning_rate": 9.377551020408164e-06,
+      "loss": 0.2571,
+      "step": 10410
+    },
+    {
+      "epoch": 53.16326530612245,
+      "grad_norm": 6.720277786254883,
+      "learning_rate": 9.367346938775511e-06,
+      "loss": 0.2213,
+      "step": 10420
+    },
+    {
+      "epoch": 53.214285714285715,
+      "grad_norm": 10.51973819732666,
+      "learning_rate": 9.357142857142859e-06,
+      "loss": 0.149,
+      "step": 10430
+    },
+    {
+      "epoch": 53.265306122448976,
+      "grad_norm": 7.785160064697266,
+      "learning_rate": 9.346938775510204e-06,
+      "loss": 0.3354,
+      "step": 10440
+    },
+    {
+      "epoch": 53.316326530612244,
+      "grad_norm": 2.9712700843811035,
+      "learning_rate": 9.336734693877552e-06,
+      "loss": 0.2505,
+      "step": 10450
+    },
+    {
+      "epoch": 53.36734693877551,
+      "grad_norm": 4.9919962882995605,
+      "learning_rate": 9.326530612244898e-06,
+      "loss": 0.2684,
+      "step": 10460
+    },
+    {
+      "epoch": 53.41836734693877,
+      "grad_norm": 2.1093006134033203,
+      "learning_rate": 9.316326530612245e-06,
+      "loss": 0.1597,
+      "step": 10470
+    },
+    {
+      "epoch": 53.46938775510204,
+      "grad_norm": 4.116796016693115,
+      "learning_rate": 9.306122448979593e-06,
+      "loss": 0.2162,
+      "step": 10480
+    },
+    {
+      "epoch": 53.52040816326531,
+      "grad_norm": 4.428661346435547,
+      "learning_rate": 9.29591836734694e-06,
+      "loss": 0.1825,
+      "step": 10490
+    },
+    {
+      "epoch": 53.57142857142857,
+      "grad_norm": 5.626094341278076,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 0.1225,
+      "step": 10500
+    },
+    {
+      "epoch": 53.62244897959184,
+      "grad_norm": 1.5667091608047485,
+      "learning_rate": 9.275510204081633e-06,
+      "loss": 0.2052,
+      "step": 10510
+    },
+    {
+      "epoch": 53.673469387755105,
+      "grad_norm": 3.0137667655944824,
+      "learning_rate": 9.26530612244898e-06,
+      "loss": 0.1321,
+      "step": 10520
+    },
+    {
+      "epoch": 53.724489795918366,
+      "grad_norm": 7.578505516052246,
+      "learning_rate": 9.255102040816327e-06,
+      "loss": 0.2478,
+      "step": 10530
+    },
+    {
+      "epoch": 53.775510204081634,
+      "grad_norm": 2.6894452571868896,
+      "learning_rate": 9.244897959183674e-06,
+      "loss": 0.2133,
+      "step": 10540
+    },
+    {
+      "epoch": 53.826530612244895,
+      "grad_norm": 13.425848007202148,
+      "learning_rate": 9.234693877551022e-06,
+      "loss": 0.2095,
+      "step": 10550
+    },
+    {
+      "epoch": 53.87755102040816,
+      "grad_norm": 0.28253173828125,
+      "learning_rate": 9.224489795918367e-06,
+      "loss": 0.2264,
+      "step": 10560
+    },
+    {
+      "epoch": 53.92857142857143,
+      "grad_norm": 1.276113748550415,
+      "learning_rate": 9.214285714285715e-06,
+      "loss": 0.1314,
+      "step": 10570
+    },
+    {
+      "epoch": 53.97959183673469,
+      "grad_norm": 3.8790574073791504,
+      "learning_rate": 9.204081632653062e-06,
+      "loss": 0.554,
+      "step": 10580
+    },
+    {
+      "epoch": 54.0,
+      "eval_accuracy": 0.8989169675090253,
+      "eval_loss": 0.2849152684211731,
+      "eval_runtime": 0.9325,
+      "eval_samples_per_second": 297.065,
+      "eval_steps_per_second": 37.535,
+      "step": 10584
+    },
+    {
+      "epoch": 54.03061224489796,
+      "grad_norm": 9.941154479980469,
+      "learning_rate": 9.19387755102041e-06,
+      "loss": 0.2049,
+      "step": 10590
+    },
+    {
+      "epoch": 54.08163265306123,
+      "grad_norm": 13.770801544189453,
+      "learning_rate": 9.183673469387756e-06,
+      "loss": 0.1889,
+      "step": 10600
+    },
+    {
+      "epoch": 54.13265306122449,
+      "grad_norm": 0.3031742572784424,
+      "learning_rate": 9.173469387755103e-06,
+      "loss": 0.3012,
+      "step": 10610
+    },
+    {
+      "epoch": 54.183673469387756,
+      "grad_norm": 6.222158908843994,
+      "learning_rate": 9.163265306122449e-06,
+      "loss": 0.1944,
+      "step": 10620
+    },
+    {
+      "epoch": 54.234693877551024,
+      "grad_norm": 0.2729008197784424,
+      "learning_rate": 9.153061224489796e-06,
+      "loss": 0.1875,
+      "step": 10630
+    },
+    {
+      "epoch": 54.285714285714285,
+      "grad_norm": 1.3354541063308716,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 0.1959,
+      "step": 10640
+    },
+    {
+      "epoch": 54.33673469387755,
+      "grad_norm": 3.6860952377319336,
+      "learning_rate": 9.13265306122449e-06,
+      "loss": 0.4611,
+      "step": 10650
+    },
+    {
+      "epoch": 54.38775510204081,
+      "grad_norm": 2.5632665157318115,
+      "learning_rate": 9.122448979591837e-06,
+      "loss": 0.1682,
+      "step": 10660
+    },
+    {
+      "epoch": 54.43877551020408,
+      "grad_norm": 0.9468786120414734,
+      "learning_rate": 9.112244897959185e-06,
+      "loss": 0.2813,
+      "step": 10670
+    },
+    {
+      "epoch": 54.48979591836735,
+      "grad_norm": 7.614074230194092,
+      "learning_rate": 9.102040816326532e-06,
+      "loss": 0.235,
+      "step": 10680
+    },
+    {
+      "epoch": 54.54081632653061,
+      "grad_norm": 10.54065227508545,
+      "learning_rate": 9.091836734693878e-06,
+      "loss": 0.2602,
+      "step": 10690
+    },
+    {
+      "epoch": 54.59183673469388,
+      "grad_norm": 14.505642890930176,
+      "learning_rate": 9.081632653061225e-06,
+      "loss": 0.3777,
+      "step": 10700
+    },
+    {
+      "epoch": 54.642857142857146,
+      "grad_norm": 6.543236255645752,
+      "learning_rate": 9.071428571428573e-06,
+      "loss": 0.2374,
+      "step": 10710
+    },
+    {
+      "epoch": 54.69387755102041,
+      "grad_norm": 1.4957891702651978,
+      "learning_rate": 9.061224489795919e-06,
+      "loss": 0.3301,
+      "step": 10720
+    },
+    {
+      "epoch": 54.744897959183675,
+      "grad_norm": 4.956941604614258,
+      "learning_rate": 9.051020408163266e-06,
+      "loss": 0.2131,
+      "step": 10730
+    },
+    {
+      "epoch": 54.795918367346935,
+      "grad_norm": 16.052499771118164,
+      "learning_rate": 9.040816326530612e-06,
+      "loss": 0.3309,
+      "step": 10740
+    },
+    {
+      "epoch": 54.8469387755102,
+      "grad_norm": 2.2586865425109863,
+      "learning_rate": 9.03061224489796e-06,
+      "loss": 0.2048,
+      "step": 10750
+    },
+    {
+      "epoch": 54.89795918367347,
+      "grad_norm": 6.091680526733398,
+      "learning_rate": 9.020408163265307e-06,
+      "loss": 0.1964,
+      "step": 10760
+    },
+    {
+      "epoch": 54.94897959183673,
+      "grad_norm": 2.350778818130493,
+      "learning_rate": 9.010204081632654e-06,
+      "loss": 0.1545,
+      "step": 10770
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 15.803840637207031,
+      "learning_rate": 9e-06,
+      "loss": 0.3586,
+      "step": 10780
+    },
+    {
+      "epoch": 55.0,
+      "eval_accuracy": 0.9133574007220217,
+      "eval_loss": 0.35168057680130005,
+      "eval_runtime": 0.9347,
+      "eval_samples_per_second": 296.337,
+      "eval_steps_per_second": 37.443,
+      "step": 10780
+    },
+    {
+      "epoch": 55.05102040816327,
+      "grad_norm": 8.534765243530273,
+      "learning_rate": 8.989795918367348e-06,
+      "loss": 0.172,
+      "step": 10790
+    },
+    {
+      "epoch": 55.10204081632653,
+      "grad_norm": 6.930604457855225,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 0.3134,
+      "step": 10800
+    },
+    {
+      "epoch": 55.1530612244898,
+      "grad_norm": 4.500041961669922,
+      "learning_rate": 8.969387755102041e-06,
+      "loss": 0.3951,
+      "step": 10810
+    },
+    {
+      "epoch": 55.204081632653065,
+      "grad_norm": 10.691000938415527,
+      "learning_rate": 8.959183673469388e-06,
+      "loss": 0.556,
+      "step": 10820
+    },
+    {
+      "epoch": 55.255102040816325,
+      "grad_norm": 5.5316853523254395,
+      "learning_rate": 8.948979591836734e-06,
+      "loss": 0.2841,
+      "step": 10830
+    },
+    {
+      "epoch": 55.30612244897959,
+      "grad_norm": 11.046544075012207,
+      "learning_rate": 8.938775510204082e-06,
+      "loss": 0.1667,
+      "step": 10840
+    },
+    {
+      "epoch": 55.357142857142854,
+      "grad_norm": 3.7789947986602783,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 0.2502,
+      "step": 10850
+    },
+    {
+      "epoch": 55.40816326530612,
+      "grad_norm": 3.0397422313690186,
+      "learning_rate": 8.918367346938777e-06,
+      "loss": 0.3261,
+      "step": 10860
+    },
+    {
+      "epoch": 55.45918367346939,
+      "grad_norm": 10.146750450134277,
+      "learning_rate": 8.908163265306124e-06,
+      "loss": 0.4223,
+      "step": 10870
+    },
+    {
+      "epoch": 55.51020408163265,
+      "grad_norm": 14.126100540161133,
+      "learning_rate": 8.89795918367347e-06,
+      "loss": 0.3837,
+      "step": 10880
+    },
+    {
+      "epoch": 55.56122448979592,
+      "grad_norm": 12.385411262512207,
+      "learning_rate": 8.887755102040817e-06,
+      "loss": 0.4422,
+      "step": 10890
+    },
+    {
+      "epoch": 55.61224489795919,
+      "grad_norm": 1.6094768047332764,
+      "learning_rate": 8.877551020408163e-06,
+      "loss": 0.2001,
+      "step": 10900
+    },
+    {
+      "epoch": 55.66326530612245,
+      "grad_norm": 0.18602848052978516,
+      "learning_rate": 8.86734693877551e-06,
+      "loss": 0.2053,
+      "step": 10910
+    },
+    {
+      "epoch": 55.714285714285715,
+      "grad_norm": 7.022019386291504,
+      "learning_rate": 8.857142857142858e-06,
+      "loss": 0.3404,
+      "step": 10920
+    },
+    {
+      "epoch": 55.765306122448976,
+      "grad_norm": 3.6204686164855957,
+      "learning_rate": 8.846938775510204e-06,
+      "loss": 0.1141,
+      "step": 10930
+    },
+    {
+      "epoch": 55.816326530612244,
+      "grad_norm": 0.32874494791030884,
+      "learning_rate": 8.836734693877552e-06,
+      "loss": 0.1693,
+      "step": 10940
+    },
+    {
+      "epoch": 55.86734693877551,
+      "grad_norm": 3.1909286975860596,
+      "learning_rate": 8.826530612244899e-06,
+      "loss": 0.333,
+      "step": 10950
+    },
+    {
+      "epoch": 55.91836734693877,
+      "grad_norm": 6.641414165496826,
+      "learning_rate": 8.816326530612247e-06,
+      "loss": 0.2012,
+      "step": 10960
+    },
+    {
+      "epoch": 55.96938775510204,
+      "grad_norm": 9.18001651763916,
+      "learning_rate": 8.806122448979592e-06,
+      "loss": 0.2239,
+      "step": 10970
+    },
+    {
+      "epoch": 56.0,
+      "eval_accuracy": 0.9025270758122743,
+      "eval_loss": 0.45381447672843933,
+      "eval_runtime": 0.9387,
+      "eval_samples_per_second": 295.088,
+      "eval_steps_per_second": 37.286,
+      "step": 10976
+    },
+    {
+      "epoch": 56.02040816326531,
+      "grad_norm": 7.012698173522949,
+      "learning_rate": 8.79591836734694e-06,
+      "loss": 0.3606,
+      "step": 10980
+    },
+    {
+      "epoch": 56.07142857142857,
+      "grad_norm": 0.33083972334861755,
+      "learning_rate": 8.785714285714286e-06,
+      "loss": 0.164,
+      "step": 10990
+    },
+    {
+      "epoch": 56.12244897959184,
+      "grad_norm": 1.1617504358291626,
+      "learning_rate": 8.775510204081633e-06,
+      "loss": 0.4036,
+      "step": 11000
+    },
+    {
+      "epoch": 56.173469387755105,
+      "grad_norm": 18.734424591064453,
+      "learning_rate": 8.76530612244898e-06,
+      "loss": 0.2242,
+      "step": 11010
+    },
+    {
+      "epoch": 56.224489795918366,
+      "grad_norm": 0.8280305862426758,
+      "learning_rate": 8.755102040816326e-06,
+      "loss": 0.24,
+      "step": 11020
+    },
+    {
+      "epoch": 56.275510204081634,
+      "grad_norm": 1.0710173845291138,
+      "learning_rate": 8.744897959183676e-06,
+      "loss": 0.2093,
+      "step": 11030
+    },
+    {
+      "epoch": 56.326530612244895,
+      "grad_norm": 0.26472461223602295,
+      "learning_rate": 8.734693877551021e-06,
+      "loss": 0.3167,
+      "step": 11040
+    },
+    {
+      "epoch": 56.37755102040816,
+      "grad_norm": 17.181556701660156,
+      "learning_rate": 8.724489795918369e-06,
+      "loss": 0.3231,
+      "step": 11050
+    },
+    {
+      "epoch": 56.42857142857143,
+      "grad_norm": 0.4559863209724426,
+      "learning_rate": 8.714285714285715e-06,
+      "loss": 0.1493,
+      "step": 11060
+    },
+    {
+      "epoch": 56.47959183673469,
+      "grad_norm": 14.466240882873535,
+      "learning_rate": 8.704081632653062e-06,
+      "loss": 0.3536,
+      "step": 11070
+    },
+    {
+      "epoch": 56.53061224489796,
+      "grad_norm": 1.9131622314453125,
+      "learning_rate": 8.69387755102041e-06,
+      "loss": 0.2293,
+      "step": 11080
+    },
+    {
+      "epoch": 56.58163265306123,
+      "grad_norm": 3.1065633296966553,
+      "learning_rate": 8.683673469387755e-06,
+      "loss": 0.1173,
+      "step": 11090
+    },
+    {
+      "epoch": 56.63265306122449,
+      "grad_norm": 12.009686470031738,
+      "learning_rate": 8.673469387755103e-06,
+      "loss": 0.2443,
+      "step": 11100
+    },
+    {
+      "epoch": 56.683673469387756,
+      "grad_norm": 2.6055593490600586,
+      "learning_rate": 8.663265306122449e-06,
+      "loss": 0.4624,
+      "step": 11110
+    },
+    {
+      "epoch": 56.734693877551024,
+      "grad_norm": 6.770671367645264,
+      "learning_rate": 8.653061224489798e-06,
+      "loss": 0.2318,
+      "step": 11120
+    },
+    {
+      "epoch": 56.785714285714285,
+      "grad_norm": 6.624651908874512,
+      "learning_rate": 8.642857142857144e-06,
+      "loss": 0.3801,
+      "step": 11130
+    },
+    {
+      "epoch": 56.83673469387755,
+      "grad_norm": 0.4821954667568207,
+      "learning_rate": 8.632653061224491e-06,
+      "loss": 0.2533,
+      "step": 11140
+    },
+    {
+      "epoch": 56.88775510204081,
+      "grad_norm": 7.2710113525390625,
+      "learning_rate": 8.622448979591837e-06,
+      "loss": 0.2761,
+      "step": 11150
+    },
+    {
+      "epoch": 56.93877551020408,
+      "grad_norm": 20.32927131652832,
+      "learning_rate": 8.612244897959184e-06,
+      "loss": 0.4597,
+      "step": 11160
+    },
+    {
+      "epoch": 56.98979591836735,
+      "grad_norm": 0.4477481245994568,
+      "learning_rate": 8.602040816326532e-06,
+      "loss": 0.1725,
+      "step": 11170
+    },
+    {
+      "epoch": 57.0,
+      "eval_accuracy": 0.8592057761732852,
+      "eval_loss": 0.44916799664497375,
+      "eval_runtime": 0.9302,
+      "eval_samples_per_second": 297.771,
+      "eval_steps_per_second": 37.625,
+      "step": 11172
+    },
+    {
+      "epoch": 57.04081632653061,
+      "grad_norm": 11.276688575744629,
+      "learning_rate": 8.591836734693878e-06,
+      "loss": 0.234,
+      "step": 11180
+    },
+    {
+      "epoch": 57.09183673469388,
+      "grad_norm": 1.517318844795227,
+      "learning_rate": 8.581632653061225e-06,
+      "loss": 0.1559,
+      "step": 11190
+    },
+    {
+      "epoch": 57.142857142857146,
+      "grad_norm": 7.025010585784912,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.2112,
+      "step": 11200
+    },
+    {
+      "epoch": 57.19387755102041,
+      "grad_norm": 10.036983489990234,
+      "learning_rate": 8.56122448979592e-06,
+      "loss": 0.2964,
+      "step": 11210
+    },
+    {
+      "epoch": 57.244897959183675,
+      "grad_norm": 9.007682800292969,
+      "learning_rate": 8.551020408163266e-06,
+      "loss": 0.2345,
+      "step": 11220
+    },
+    {
+      "epoch": 57.295918367346935,
+      "grad_norm": 15.974291801452637,
+      "learning_rate": 8.540816326530613e-06,
+      "loss": 0.5178,
+      "step": 11230
+    },
+    {
+      "epoch": 57.3469387755102,
+      "grad_norm": 0.5210385918617249,
+      "learning_rate": 8.530612244897961e-06,
+      "loss": 0.0861,
+      "step": 11240
+    },
+    {
+      "epoch": 57.39795918367347,
+      "grad_norm": 0.22501814365386963,
+      "learning_rate": 8.520408163265307e-06,
+      "loss": 0.1869,
+      "step": 11250
+    },
+    {
+      "epoch": 57.44897959183673,
+      "grad_norm": 1.9573888778686523,
+      "learning_rate": 8.510204081632654e-06,
+      "loss": 0.2643,
+      "step": 11260
+    },
+    {
+      "epoch": 57.5,
+      "grad_norm": 3.5790092945098877,
+      "learning_rate": 8.5e-06,
+      "loss": 0.1604,
+      "step": 11270
+    },
+    {
+      "epoch": 57.55102040816327,
+      "grad_norm": 10.310325622558594,
+      "learning_rate": 8.489795918367347e-06,
+      "loss": 0.1772,
+      "step": 11280
+    },
+    {
+      "epoch": 57.60204081632653,
+      "grad_norm": 10.196538925170898,
+      "learning_rate": 8.479591836734695e-06,
+      "loss": 0.1882,
+      "step": 11290
+    },
+    {
+      "epoch": 57.6530612244898,
+      "grad_norm": 6.247244358062744,
+      "learning_rate": 8.469387755102042e-06,
+      "loss": 0.2541,
+      "step": 11300
+    },
+    {
+      "epoch": 57.704081632653065,
+      "grad_norm": 11.577144622802734,
+      "learning_rate": 8.459183673469388e-06,
+      "loss": 0.3183,
+      "step": 11310
+    },
+    {
+      "epoch": 57.755102040816325,
+      "grad_norm": 1.9273695945739746,
+      "learning_rate": 8.448979591836736e-06,
+      "loss": 0.1365,
+      "step": 11320
+    },
+    {
+      "epoch": 57.80612244897959,
+      "grad_norm": 3.144540548324585,
+      "learning_rate": 8.438775510204083e-06,
+      "loss": 0.2859,
+      "step": 11330
+    },
+    {
+      "epoch": 57.857142857142854,
+      "grad_norm": 10.630606651306152,
+      "learning_rate": 8.428571428571429e-06,
+      "loss": 0.2768,
+      "step": 11340
+    },
+    {
+      "epoch": 57.90816326530612,
+      "grad_norm": 17.469684600830078,
+      "learning_rate": 8.418367346938776e-06,
+      "loss": 0.3497,
+      "step": 11350
+    },
+    {
+      "epoch": 57.95918367346939,
+      "grad_norm": 1.6836483478546143,
+      "learning_rate": 8.408163265306122e-06,
+      "loss": 0.4689,
+      "step": 11360
+    },
+    {
+      "epoch": 58.0,
+      "eval_accuracy": 0.8628158844765343,
+      "eval_loss": 0.4738902747631073,
+      "eval_runtime": 0.9341,
+      "eval_samples_per_second": 296.536,
+      "eval_steps_per_second": 37.468,
+      "step": 11368
+    },
+    {
+      "epoch": 58.01020408163265,
+      "grad_norm": 12.947834968566895,
+      "learning_rate": 8.39795918367347e-06,
+      "loss": 0.3388,
+      "step": 11370
+    },
+    {
+      "epoch": 58.06122448979592,
+      "grad_norm": 9.694038391113281,
+      "learning_rate": 8.387755102040817e-06,
+      "loss": 0.2179,
+      "step": 11380
+    },
+    {
+      "epoch": 58.11224489795919,
+      "grad_norm": 2.7403390407562256,
+      "learning_rate": 8.377551020408165e-06,
+      "loss": 0.2709,
+      "step": 11390
+    },
+    {
+      "epoch": 58.16326530612245,
+      "grad_norm": 6.763411045074463,
+      "learning_rate": 8.36734693877551e-06,
+      "loss": 0.1844,
+      "step": 11400
+    },
+    {
+      "epoch": 58.214285714285715,
+      "grad_norm": 9.05958080291748,
+      "learning_rate": 8.357142857142858e-06,
+      "loss": 0.1588,
+      "step": 11410
+    },
+    {
+      "epoch": 58.265306122448976,
+      "grad_norm": 8.429780960083008,
+      "learning_rate": 8.346938775510205e-06,
+      "loss": 0.3569,
+      "step": 11420
+    },
+    {
+      "epoch": 58.316326530612244,
+      "grad_norm": 5.592264175415039,
+      "learning_rate": 8.336734693877551e-06,
+      "loss": 0.2468,
+      "step": 11430
+    },
+    {
+      "epoch": 58.36734693877551,
+      "grad_norm": 3.081688642501831,
+      "learning_rate": 8.326530612244899e-06,
+      "loss": 0.0708,
+      "step": 11440
+    },
+    {
+      "epoch": 58.41836734693877,
+      "grad_norm": 1.1242700815200806,
+      "learning_rate": 8.316326530612246e-06,
+      "loss": 0.4146,
+      "step": 11450
+    },
+    {
+      "epoch": 58.46938775510204,
+      "grad_norm": 9.29017162322998,
+      "learning_rate": 8.306122448979592e-06,
+      "loss": 0.2877,
+      "step": 11460
+    },
+    {
+      "epoch": 58.52040816326531,
+      "grad_norm": 1.7449393272399902,
+      "learning_rate": 8.29591836734694e-06,
+      "loss": 0.3654,
+      "step": 11470
+    },
+    {
+      "epoch": 58.57142857142857,
+      "grad_norm": 9.28416919708252,
+      "learning_rate": 8.285714285714287e-06,
+      "loss": 0.2411,
+      "step": 11480
+    },
+    {
+      "epoch": 58.62244897959184,
+      "grad_norm": 2.053839683532715,
+      "learning_rate": 8.275510204081634e-06,
+      "loss": 0.2477,
+      "step": 11490
+    },
+    {
+      "epoch": 58.673469387755105,
+      "grad_norm": 0.5099808573722839,
+      "learning_rate": 8.26530612244898e-06,
+      "loss": 0.086,
+      "step": 11500
+    },
+    {
+      "epoch": 58.724489795918366,
+      "grad_norm": 5.123264789581299,
+      "learning_rate": 8.255102040816328e-06,
+      "loss": 0.1873,
+      "step": 11510
+    },
+    {
+      "epoch": 58.775510204081634,
+      "grad_norm": 0.3827672004699707,
+      "learning_rate": 8.244897959183674e-06,
+      "loss": 0.245,
+      "step": 11520
+    },
+    {
+      "epoch": 58.826530612244895,
+      "grad_norm": 6.846168041229248,
+      "learning_rate": 8.234693877551021e-06,
+      "loss": 0.6233,
+      "step": 11530
+    },
+    {
+      "epoch": 58.87755102040816,
+      "grad_norm": 0.9291605949401855,
+      "learning_rate": 8.224489795918369e-06,
+      "loss": 0.2342,
+      "step": 11540
+    },
+    {
+      "epoch": 58.92857142857143,
+      "grad_norm": 4.6036834716796875,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 0.2538,
+      "step": 11550
+    },
+    {
+      "epoch": 58.97959183673469,
+      "grad_norm": 10.339995384216309,
+      "learning_rate": 8.204081632653062e-06,
+      "loss": 0.3565,
+      "step": 11560
+    },
+    {
+      "epoch": 59.0,
+      "eval_accuracy": 0.9205776173285198,
+      "eval_loss": 0.2831103205680847,
+      "eval_runtime": 0.9374,
+      "eval_samples_per_second": 295.493,
+      "eval_steps_per_second": 37.337,
+      "step": 11564
+    },
+    {
+      "epoch": 59.03061224489796,
+      "grad_norm": 1.1292849779129028,
+      "learning_rate": 8.19387755102041e-06,
+      "loss": 0.1716,
+      "step": 11570
+    },
+    {
+      "epoch": 59.08163265306123,
+      "grad_norm": 0.5711737275123596,
+      "learning_rate": 8.183673469387757e-06,
+      "loss": 0.1703,
+      "step": 11580
+    },
+    {
+      "epoch": 59.13265306122449,
+      "grad_norm": 0.48393628001213074,
+      "learning_rate": 8.173469387755103e-06,
+      "loss": 0.1301,
+      "step": 11590
+    },
+    {
+      "epoch": 59.183673469387756,
+      "grad_norm": 0.1569240391254425,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 0.1911,
+      "step": 11600
+    },
+    {
+      "epoch": 59.234693877551024,
+      "grad_norm": 11.578142166137695,
+      "learning_rate": 8.153061224489796e-06,
+      "loss": 0.1351,
+      "step": 11610
+    },
+    {
+      "epoch": 59.285714285714285,
+      "grad_norm": 11.53231430053711,
+      "learning_rate": 8.142857142857143e-06,
+      "loss": 0.3704,
+      "step": 11620
+    },
+    {
+      "epoch": 59.33673469387755,
+      "grad_norm": 7.75083589553833,
+      "learning_rate": 8.13265306122449e-06,
+      "loss": 0.2867,
+      "step": 11630
+    },
+    {
+      "epoch": 59.38775510204081,
+      "grad_norm": 1.0234230756759644,
+      "learning_rate": 8.122448979591837e-06,
+      "loss": 0.2227,
+      "step": 11640
+    },
+    {
+      "epoch": 59.43877551020408,
+      "grad_norm": 6.651299953460693,
+      "learning_rate": 8.112244897959184e-06,
+      "loss": 0.1946,
+      "step": 11650
+    },
+    {
+      "epoch": 59.48979591836735,
+      "grad_norm": 7.23628568649292,
+      "learning_rate": 8.102040816326532e-06,
+      "loss": 0.1456,
+      "step": 11660
+    },
+    {
+      "epoch": 59.54081632653061,
+      "grad_norm": 2.716782331466675,
+      "learning_rate": 8.091836734693879e-06,
+      "loss": 0.1696,
+      "step": 11670
+    },
+    {
+      "epoch": 59.59183673469388,
+      "grad_norm": 4.466701030731201,
+      "learning_rate": 8.081632653061225e-06,
+      "loss": 0.29,
+      "step": 11680
+    },
+    {
+      "epoch": 59.642857142857146,
+      "grad_norm": 16.736553192138672,
+      "learning_rate": 8.071428571428572e-06,
+      "loss": 0.3853,
+      "step": 11690
+    },
+    {
+      "epoch": 59.69387755102041,
+      "grad_norm": 0.4944998621940613,
+      "learning_rate": 8.06122448979592e-06,
+      "loss": 0.354,
+      "step": 11700
+    },
+    {
+      "epoch": 59.744897959183675,
+      "grad_norm": 3.3472633361816406,
+      "learning_rate": 8.051020408163266e-06,
+      "loss": 0.2751,
+      "step": 11710
+    },
+    {
+      "epoch": 59.795918367346935,
+      "grad_norm": 2.8295209407806396,
+      "learning_rate": 8.040816326530613e-06,
+      "loss": 0.1301,
+      "step": 11720
+    },
+    {
+      "epoch": 59.8469387755102,
+      "grad_norm": 1.9733431339263916,
+      "learning_rate": 8.030612244897959e-06,
+      "loss": 0.5063,
+      "step": 11730
+    },
+    {
+      "epoch": 59.89795918367347,
+      "grad_norm": 0.1797264963388443,
+      "learning_rate": 8.020408163265306e-06,
+      "loss": 0.1427,
+      "step": 11740
+    },
+    {
+      "epoch": 59.94897959183673,
+      "grad_norm": 9.685842514038086,
+      "learning_rate": 8.010204081632654e-06,
+      "loss": 0.2899,
+      "step": 11750
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 13.383129119873047,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.2259,
+      "step": 11760
+    },
+    {
+      "epoch": 60.0,
+      "eval_accuracy": 0.9205776173285198,
+      "eval_loss": 0.34654152393341064,
+      "eval_runtime": 0.9505,
+      "eval_samples_per_second": 291.425,
+      "eval_steps_per_second": 36.823,
+      "step": 11760
+    },
+    {
+      "epoch": 60.05102040816327,
+      "grad_norm": 2.9076764583587646,
+      "learning_rate": 7.989795918367347e-06,
+      "loss": 0.065,
+      "step": 11770
+    },
+    {
+      "epoch": 60.10204081632653,
+      "grad_norm": 8.378006935119629,
+      "learning_rate": 7.979591836734695e-06,
+      "loss": 0.2488,
+      "step": 11780
+    },
+    {
+      "epoch": 60.1530612244898,
+      "grad_norm": 12.380623817443848,
+      "learning_rate": 7.969387755102042e-06,
+      "loss": 0.3169,
+      "step": 11790
+    },
+    {
+      "epoch": 60.204081632653065,
+      "grad_norm": 4.164218425750732,
+      "learning_rate": 7.959183673469388e-06,
+      "loss": 0.1818,
+      "step": 11800
+    },
+    {
+      "epoch": 60.255102040816325,
+      "grad_norm": 14.524949073791504,
+      "learning_rate": 7.948979591836735e-06,
+      "loss": 0.3417,
+      "step": 11810
+    },
+    {
+      "epoch": 60.30612244897959,
+      "grad_norm": 6.435640811920166,
+      "learning_rate": 7.938775510204081e-06,
+      "loss": 0.274,
+      "step": 11820
+    },
+    {
+      "epoch": 60.357142857142854,
+      "grad_norm": 4.574618816375732,
+      "learning_rate": 7.928571428571429e-06,
+      "loss": 0.4462,
+      "step": 11830
+    },
+    {
+      "epoch": 60.40816326530612,
+      "grad_norm": 7.529806137084961,
+      "learning_rate": 7.918367346938776e-06,
+      "loss": 0.3905,
+      "step": 11840
+    },
+    {
+      "epoch": 60.45918367346939,
+      "grad_norm": 3.4086387157440186,
+      "learning_rate": 7.908163265306124e-06,
+      "loss": 0.301,
+      "step": 11850
+    },
+    {
+      "epoch": 60.51020408163265,
+      "grad_norm": 0.847341001033783,
+      "learning_rate": 7.897959183673471e-06,
+      "loss": 0.1661,
+      "step": 11860
+    },
+    {
+      "epoch": 60.56122448979592,
+      "grad_norm": 3.0656235218048096,
+      "learning_rate": 7.887755102040817e-06,
+      "loss": 0.2628,
+      "step": 11870
+    },
+    {
+      "epoch": 60.61224489795919,
+      "grad_norm": 1.6243197917938232,
+      "learning_rate": 7.877551020408164e-06,
+      "loss": 0.1612,
+      "step": 11880
+    },
+    {
+      "epoch": 60.66326530612245,
+      "grad_norm": 11.95365047454834,
+      "learning_rate": 7.86734693877551e-06,
+      "loss": 0.1425,
+      "step": 11890
+    },
+    {
+      "epoch": 60.714285714285715,
+      "grad_norm": 7.095501899719238,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.1915,
+      "step": 11900
+    },
+    {
+      "epoch": 60.765306122448976,
+      "grad_norm": 8.617204666137695,
+      "learning_rate": 7.846938775510205e-06,
+      "loss": 0.1871,
+      "step": 11910
+    },
+    {
+      "epoch": 60.816326530612244,
+      "grad_norm": 7.942294120788574,
+      "learning_rate": 7.836734693877551e-06,
+      "loss": 0.2194,
+      "step": 11920
+    },
+    {
+      "epoch": 60.86734693877551,
+      "grad_norm": 4.334501266479492,
+      "learning_rate": 7.826530612244898e-06,
+      "loss": 0.2936,
+      "step": 11930
+    },
+    {
+      "epoch": 60.91836734693877,
+      "grad_norm": 6.984222888946533,
+      "learning_rate": 7.816326530612246e-06,
+      "loss": 0.2264,
+      "step": 11940
+    },
+    {
+      "epoch": 60.96938775510204,
+      "grad_norm": 10.189035415649414,
+      "learning_rate": 7.806122448979593e-06,
+      "loss": 0.2212,
+      "step": 11950
+    },
+    {
+      "epoch": 61.0,
+      "eval_accuracy": 0.9314079422382672,
+      "eval_loss": 0.28843551874160767,
+      "eval_runtime": 0.9333,
+      "eval_samples_per_second": 296.782,
+      "eval_steps_per_second": 37.5,
+      "step": 11956
+    },
+    {
+      "epoch": 61.02040816326531,
+      "grad_norm": 5.532245635986328,
+      "learning_rate": 7.79591836734694e-06,
+      "loss": 0.2216,
+      "step": 11960
+    },
+    {
+      "epoch": 61.07142857142857,
+      "grad_norm": 2.590909242630005,
+      "learning_rate": 7.785714285714287e-06,
+      "loss": 0.1215,
+      "step": 11970
+    },
+    {
+      "epoch": 61.12244897959184,
+      "grad_norm": 7.619017601013184,
+      "learning_rate": 7.775510204081632e-06,
+      "loss": 0.1876,
+      "step": 11980
+    },
+    {
+      "epoch": 61.173469387755105,
+      "grad_norm": 17.85263442993164,
+      "learning_rate": 7.76530612244898e-06,
+      "loss": 0.2472,
+      "step": 11990
+    },
+    {
+      "epoch": 61.224489795918366,
+      "grad_norm": 5.559810161590576,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 0.3429,
+      "step": 12000
+    },
+    {
+      "epoch": 61.275510204081634,
+      "grad_norm": 3.257960796356201,
+      "learning_rate": 7.744897959183673e-06,
+      "loss": 0.2473,
+      "step": 12010
+    },
+    {
+      "epoch": 61.326530612244895,
+      "grad_norm": 2.2071876525878906,
+      "learning_rate": 7.73469387755102e-06,
+      "loss": 0.1332,
+      "step": 12020
+    },
+    {
+      "epoch": 61.37755102040816,
+      "grad_norm": 0.41261130571365356,
+      "learning_rate": 7.724489795918368e-06,
+      "loss": 0.1469,
+      "step": 12030
+    },
+    {
+      "epoch": 61.42857142857143,
+      "grad_norm": 5.125622272491455,
+      "learning_rate": 7.714285714285716e-06,
+      "loss": 0.3787,
+      "step": 12040
+    },
+    {
+      "epoch": 61.47959183673469,
+      "grad_norm": 0.23421549797058105,
+      "learning_rate": 7.704081632653061e-06,
+      "loss": 0.0996,
+      "step": 12050
+    },
+    {
+      "epoch": 61.53061224489796,
+      "grad_norm": 0.6105390787124634,
+      "learning_rate": 7.693877551020409e-06,
+      "loss": 0.1756,
+      "step": 12060
+    },
+    {
+      "epoch": 61.58163265306123,
+      "grad_norm": 9.674683570861816,
+      "learning_rate": 7.683673469387756e-06,
+      "loss": 0.2785,
+      "step": 12070
+    },
+    {
+      "epoch": 61.63265306122449,
+      "grad_norm": 10.404973030090332,
+      "learning_rate": 7.673469387755102e-06,
+      "loss": 0.3011,
+      "step": 12080
+    },
+    {
+      "epoch": 61.683673469387756,
+      "grad_norm": 22.518468856811523,
+      "learning_rate": 7.66326530612245e-06,
+      "loss": 0.5065,
+      "step": 12090
+    },
+    {
+      "epoch": 61.734693877551024,
+      "grad_norm": 7.028233051300049,
+      "learning_rate": 7.653061224489796e-06,
+      "loss": 0.1553,
+      "step": 12100
+    },
+    {
+      "epoch": 61.785714285714285,
+      "grad_norm": 0.5314661860466003,
+      "learning_rate": 7.642857142857143e-06,
+      "loss": 0.2187,
+      "step": 12110
+    },
+    {
+      "epoch": 61.83673469387755,
+      "grad_norm": 6.046501159667969,
+      "learning_rate": 7.63265306122449e-06,
+      "loss": 0.1042,
+      "step": 12120
+    },
+    {
+      "epoch": 61.88775510204081,
+      "grad_norm": 1.112051248550415,
+      "learning_rate": 7.622448979591838e-06,
+      "loss": 0.3595,
+      "step": 12130
+    },
+    {
+      "epoch": 61.93877551020408,
+      "grad_norm": 2.3171072006225586,
+      "learning_rate": 7.612244897959185e-06,
+      "loss": 0.19,
+      "step": 12140
+    },
+    {
+      "epoch": 61.98979591836735,
+      "grad_norm": 11.073331832885742,
+      "learning_rate": 7.602040816326531e-06,
+      "loss": 0.2648,
+      "step": 12150
+    },
+    {
+      "epoch": 62.0,
+      "eval_accuracy": 0.8447653429602888,
+      "eval_loss": 0.4874745309352875,
+      "eval_runtime": 0.9313,
+      "eval_samples_per_second": 297.448,
+      "eval_steps_per_second": 37.584,
+      "step": 12152
+    },
+    {
+      "epoch": 62.04081632653061,
+      "grad_norm": 2.1476967334747314,
+      "learning_rate": 7.591836734693878e-06,
+      "loss": 0.1787,
+      "step": 12160
+    },
+    {
+      "epoch": 62.09183673469388,
+      "grad_norm": 0.47355252504348755,
+      "learning_rate": 7.581632653061225e-06,
+      "loss": 0.1529,
+      "step": 12170
+    },
+    {
+      "epoch": 62.142857142857146,
+      "grad_norm": 6.79935359954834,
+      "learning_rate": 7.571428571428572e-06,
+      "loss": 0.2228,
+      "step": 12180
+    },
+    {
+      "epoch": 62.19387755102041,
+      "grad_norm": 4.178977966308594,
+      "learning_rate": 7.561224489795919e-06,
+      "loss": 0.1832,
+      "step": 12190
+    },
+    {
+      "epoch": 62.244897959183675,
+      "grad_norm": 12.742783546447754,
+      "learning_rate": 7.551020408163265e-06,
+      "loss": 0.2281,
+      "step": 12200
+    },
+    {
+      "epoch": 62.295918367346935,
+      "grad_norm": 1.5981056690216064,
+      "learning_rate": 7.540816326530614e-06,
+      "loss": 0.2196,
+      "step": 12210
+    },
+    {
+      "epoch": 62.3469387755102,
+      "grad_norm": 5.777743339538574,
+      "learning_rate": 7.53061224489796e-06,
+      "loss": 0.4064,
+      "step": 12220
+    },
+    {
+      "epoch": 62.39795918367347,
+      "grad_norm": 1.1234701871871948,
+      "learning_rate": 7.520408163265307e-06,
+      "loss": 0.2547,
+      "step": 12230
+    },
+    {
+      "epoch": 62.44897959183673,
+      "grad_norm": 3.0036509037017822,
+      "learning_rate": 7.5102040816326536e-06,
+      "loss": 0.2274,
+      "step": 12240
+    },
+    {
+      "epoch": 62.5,
+      "grad_norm": 0.5534256100654602,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1548,
+      "step": 12250
+    },
+    {
+      "epoch": 62.55102040816327,
+      "grad_norm": 1.8934872150421143,
+      "learning_rate": 7.489795918367348e-06,
+      "loss": 0.0968,
+      "step": 12260
+    },
+    {
+      "epoch": 62.60204081632653,
+      "grad_norm": 4.307313919067383,
+      "learning_rate": 7.479591836734694e-06,
+      "loss": 0.2193,
+      "step": 12270
+    },
+    {
+      "epoch": 62.6530612244898,
+      "grad_norm": 11.095629692077637,
+      "learning_rate": 7.469387755102041e-06,
+      "loss": 0.2613,
+      "step": 12280
+    },
+    {
+      "epoch": 62.704081632653065,
+      "grad_norm": 2.2190847396850586,
+      "learning_rate": 7.459183673469388e-06,
+      "loss": 0.1524,
+      "step": 12290
+    },
+    {
+      "epoch": 62.755102040816325,
+      "grad_norm": 1.4255743026733398,
+      "learning_rate": 7.448979591836736e-06,
+      "loss": 0.0702,
+      "step": 12300
+    },
+    {
+      "epoch": 62.80612244897959,
+      "grad_norm": 0.909092903137207,
+      "learning_rate": 7.4387755102040826e-06,
+      "loss": 0.2265,
+      "step": 12310
+    },
+    {
+      "epoch": 62.857142857142854,
+      "grad_norm": 7.196882724761963,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 0.2823,
+      "step": 12320
+    },
+    {
+      "epoch": 62.90816326530612,
+      "grad_norm": 8.029342651367188,
+      "learning_rate": 7.418367346938776e-06,
+      "loss": 0.1778,
+      "step": 12330
+    },
+    {
+      "epoch": 62.95918367346939,
+      "grad_norm": 7.18194055557251,
+      "learning_rate": 7.408163265306123e-06,
+      "loss": 0.3438,
+      "step": 12340
+    },
+    {
+      "epoch": 63.0,
+      "eval_accuracy": 0.9061371841155235,
+      "eval_loss": 0.3988655209541321,
+      "eval_runtime": 0.9937,
+      "eval_samples_per_second": 278.765,
+      "eval_steps_per_second": 35.223,
+      "step": 12348
+    },
+    {
+      "epoch": 63.01020408163265,
+      "grad_norm": 8.065620422363281,
+      "learning_rate": 7.39795918367347e-06,
+      "loss": 0.2821,
+      "step": 12350
+    },
+    {
+      "epoch": 63.06122448979592,
+      "grad_norm": 1.9322210550308228,
+      "learning_rate": 7.387755102040817e-06,
+      "loss": 0.095,
+      "step": 12360
+    },
+    {
+      "epoch": 63.11224489795919,
+      "grad_norm": 3.922755002975464,
+      "learning_rate": 7.377551020408163e-06,
+      "loss": 0.2767,
+      "step": 12370
+    },
+    {
+      "epoch": 63.16326530612245,
+      "grad_norm": 10.300689697265625,
+      "learning_rate": 7.367346938775511e-06,
+      "loss": 0.1315,
+      "step": 12380
+    },
+    {
+      "epoch": 63.214285714285715,
+      "grad_norm": 12.600347518920898,
+      "learning_rate": 7.357142857142858e-06,
+      "loss": 0.2517,
+      "step": 12390
+    },
+    {
+      "epoch": 63.265306122448976,
+      "grad_norm": 7.00731897354126,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 0.1496,
+      "step": 12400
+    },
+    {
+      "epoch": 63.316326530612244,
+      "grad_norm": 9.240813255310059,
+      "learning_rate": 7.3367346938775515e-06,
+      "loss": 0.196,
+      "step": 12410
+    },
+    {
+      "epoch": 63.36734693877551,
+      "grad_norm": 9.863750457763672,
+      "learning_rate": 7.326530612244899e-06,
+      "loss": 0.3422,
+      "step": 12420
+    },
+    {
+      "epoch": 63.41836734693877,
+      "grad_norm": 1.6901395320892334,
+      "learning_rate": 7.316326530612246e-06,
+      "loss": 0.1819,
+      "step": 12430
+    },
+    {
+      "epoch": 63.46938775510204,
+      "grad_norm": 12.429134368896484,
+      "learning_rate": 7.306122448979592e-06,
+      "loss": 0.4074,
+      "step": 12440
+    },
+    {
+      "epoch": 63.52040816326531,
+      "grad_norm": 1.9005409479141235,
+      "learning_rate": 7.295918367346939e-06,
+      "loss": 0.323,
+      "step": 12450
+    },
+    {
+      "epoch": 63.57142857142857,
+      "grad_norm": 5.085525989532471,
+      "learning_rate": 7.285714285714286e-06,
+      "loss": 0.1967,
+      "step": 12460
+    },
+    {
+      "epoch": 63.62244897959184,
+      "grad_norm": 7.6504340171813965,
+      "learning_rate": 7.275510204081633e-06,
+      "loss": 0.4315,
+      "step": 12470
+    },
+    {
+      "epoch": 63.673469387755105,
+      "grad_norm": 0.2195933312177658,
+      "learning_rate": 7.2653061224489805e-06,
+      "loss": 0.183,
+      "step": 12480
+    },
+    {
+      "epoch": 63.724489795918366,
+      "grad_norm": 8.156294822692871,
+      "learning_rate": 7.255102040816327e-06,
+      "loss": 0.1558,
+      "step": 12490
+    },
+    {
+      "epoch": 63.775510204081634,
+      "grad_norm": 3.2828145027160645,
+      "learning_rate": 7.244897959183675e-06,
+      "loss": 0.3911,
+      "step": 12500
+    },
+    {
+      "epoch": 63.826530612244895,
+      "grad_norm": 1.5318927764892578,
+      "learning_rate": 7.234693877551021e-06,
+      "loss": 0.1245,
+      "step": 12510
+    },
+    {
+      "epoch": 63.87755102040816,
+      "grad_norm": 1.8835053443908691,
+      "learning_rate": 7.224489795918368e-06,
+      "loss": 0.2946,
+      "step": 12520
+    },
+    {
+      "epoch": 63.92857142857143,
+      "grad_norm": 10.958850860595703,
+      "learning_rate": 7.2142857142857145e-06,
+      "loss": 0.1734,
+      "step": 12530
+    },
+    {
+      "epoch": 63.97959183673469,
+      "grad_norm": 4.654305934906006,
+      "learning_rate": 7.204081632653061e-06,
+      "loss": 0.4785,
+      "step": 12540
+    },
+    {
+      "epoch": 64.0,
+      "eval_accuracy": 0.851985559566787,
+      "eval_loss": 0.5952923893928528,
+      "eval_runtime": 0.9341,
+      "eval_samples_per_second": 296.557,
+      "eval_steps_per_second": 37.471,
+      "step": 12544
+    },
+    {
+      "epoch": 64.03061224489795,
+      "grad_norm": 8.106365203857422,
+      "learning_rate": 7.193877551020409e-06,
+      "loss": 0.1046,
+      "step": 12550
+    },
+    {
+      "epoch": 64.08163265306122,
+      "grad_norm": 18.67780113220215,
+      "learning_rate": 7.183673469387755e-06,
+      "loss": 0.294,
+      "step": 12560
+    },
+    {
+      "epoch": 64.13265306122449,
+      "grad_norm": 0.6589402556419373,
+      "learning_rate": 7.173469387755103e-06,
+      "loss": 0.2063,
+      "step": 12570
+    },
+    {
+      "epoch": 64.18367346938776,
+      "grad_norm": 17.626237869262695,
+      "learning_rate": 7.16326530612245e-06,
+      "loss": 0.246,
+      "step": 12580
+    },
+    {
+      "epoch": 64.23469387755102,
+      "grad_norm": 6.089940071105957,
+      "learning_rate": 7.153061224489797e-06,
+      "loss": 0.2377,
+      "step": 12590
+    },
+    {
+      "epoch": 64.28571428571429,
+      "grad_norm": 3.529348134994507,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.1449,
+      "step": 12600
+    },
+    {
+      "epoch": 64.33673469387755,
+      "grad_norm": 8.65750789642334,
+      "learning_rate": 7.13265306122449e-06,
+      "loss": 0.3385,
+      "step": 12610
+    },
+    {
+      "epoch": 64.38775510204081,
+      "grad_norm": 1.2404389381408691,
+      "learning_rate": 7.122448979591837e-06,
+      "loss": 0.1215,
+      "step": 12620
+    },
+    {
+      "epoch": 64.43877551020408,
+      "grad_norm": 0.9970741271972656,
+      "learning_rate": 7.112244897959184e-06,
+      "loss": 0.0967,
+      "step": 12630
+    },
+    {
+      "epoch": 64.48979591836735,
+      "grad_norm": 10.057463645935059,
+      "learning_rate": 7.102040816326531e-06,
+      "loss": 0.3337,
+      "step": 12640
+    },
+    {
+      "epoch": 64.54081632653062,
+      "grad_norm": 0.8182612061500549,
+      "learning_rate": 7.091836734693878e-06,
+      "loss": 0.2539,
+      "step": 12650
+    },
+    {
+      "epoch": 64.59183673469387,
+      "grad_norm": 2.4172310829162598,
+      "learning_rate": 7.081632653061226e-06,
+      "loss": 0.2421,
+      "step": 12660
+    },
+    {
+      "epoch": 64.64285714285714,
+      "grad_norm": 12.167418479919434,
+      "learning_rate": 7.0714285714285726e-06,
+      "loss": 0.2537,
+      "step": 12670
+    },
+    {
+      "epoch": 64.6938775510204,
+      "grad_norm": 6.439947605133057,
+      "learning_rate": 7.061224489795919e-06,
+      "loss": 0.471,
+      "step": 12680
+    },
+    {
+      "epoch": 64.74489795918367,
+      "grad_norm": 8.06596565246582,
+      "learning_rate": 7.051020408163266e-06,
+      "loss": 0.4742,
+      "step": 12690
+    },
+    {
+      "epoch": 64.79591836734694,
+      "grad_norm": 6.091670513153076,
+      "learning_rate": 7.0408163265306125e-06,
+      "loss": 0.1066,
+      "step": 12700
+    },
+    {
+      "epoch": 64.84693877551021,
+      "grad_norm": 7.659086227416992,
+      "learning_rate": 7.03061224489796e-06,
+      "loss": 0.1774,
+      "step": 12710
+    },
+    {
+      "epoch": 64.89795918367346,
+      "grad_norm": 4.606033802032471,
+      "learning_rate": 7.020408163265307e-06,
+      "loss": 0.157,
+      "step": 12720
+    },
+    {
+      "epoch": 64.94897959183673,
+      "grad_norm": 7.792344570159912,
+      "learning_rate": 7.010204081632653e-06,
+      "loss": 0.2453,
+      "step": 12730
+    },
+    {
+      "epoch": 65.0,
+      "grad_norm": 9.722793579101562,
+      "learning_rate": 7e-06,
+      "loss": 0.06,
+      "step": 12740
+    },
+    {
+      "epoch": 65.0,
+      "eval_accuracy": 0.927797833935018,
+      "eval_loss": 0.2953914701938629,
+      "eval_runtime": 0.9311,
+      "eval_samples_per_second": 297.487,
+      "eval_steps_per_second": 37.589,
+      "step": 12740
+    },
+    {
+      "epoch": 65.05102040816327,
+      "grad_norm": 1.075190544128418,
+      "learning_rate": 6.989795918367348e-06,
+      "loss": 0.0614,
+      "step": 12750
+    },
+    {
+      "epoch": 65.10204081632654,
+      "grad_norm": 0.13435253500938416,
+      "learning_rate": 6.979591836734695e-06,
+      "loss": 0.2266,
+      "step": 12760
+    },
+    {
+      "epoch": 65.15306122448979,
+      "grad_norm": 8.657276153564453,
+      "learning_rate": 6.9693877551020415e-06,
+      "loss": 0.1568,
+      "step": 12770
+    },
+    {
+      "epoch": 65.20408163265306,
+      "grad_norm": 4.1393656730651855,
+      "learning_rate": 6.959183673469388e-06,
+      "loss": 0.4365,
+      "step": 12780
+    },
+    {
+      "epoch": 65.25510204081633,
+      "grad_norm": 9.038256645202637,
+      "learning_rate": 6.948979591836736e-06,
+      "loss": 0.1867,
+      "step": 12790
+    },
+    {
+      "epoch": 65.3061224489796,
+      "grad_norm": 5.405453681945801,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 0.2024,
+      "step": 12800
+    },
+    {
+      "epoch": 65.35714285714286,
+      "grad_norm": 11.368095397949219,
+      "learning_rate": 6.928571428571429e-06,
+      "loss": 0.1922,
+      "step": 12810
+    },
+    {
+      "epoch": 65.40816326530613,
+      "grad_norm": 2.3432440757751465,
+      "learning_rate": 6.9183673469387755e-06,
+      "loss": 0.3727,
+      "step": 12820
+    },
+    {
+      "epoch": 65.45918367346938,
+      "grad_norm": 6.652347087860107,
+      "learning_rate": 6.908163265306122e-06,
+      "loss": 0.2511,
+      "step": 12830
+    },
+    {
+      "epoch": 65.51020408163265,
+      "grad_norm": 4.0521135330200195,
+      "learning_rate": 6.8979591836734705e-06,
+      "loss": 0.2601,
+      "step": 12840
+    },
+    {
+      "epoch": 65.56122448979592,
+      "grad_norm": 4.677563190460205,
+      "learning_rate": 6.887755102040817e-06,
+      "loss": 0.0881,
+      "step": 12850
+    },
+    {
+      "epoch": 65.61224489795919,
+      "grad_norm": 8.159587860107422,
+      "learning_rate": 6.877551020408164e-06,
+      "loss": 0.135,
+      "step": 12860
+    },
+    {
+      "epoch": 65.66326530612245,
+      "grad_norm": 7.858253479003906,
+      "learning_rate": 6.867346938775511e-06,
+      "loss": 0.188,
+      "step": 12870
+    },
+    {
+      "epoch": 65.71428571428571,
+      "grad_norm": 4.100142478942871,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 0.3009,
+      "step": 12880
+    },
+    {
+      "epoch": 65.76530612244898,
+      "grad_norm": 16.863325119018555,
+      "learning_rate": 6.8469387755102046e-06,
+      "loss": 0.4487,
+      "step": 12890
+    },
+    {
+      "epoch": 65.81632653061224,
+      "grad_norm": 1.883056402206421,
+      "learning_rate": 6.836734693877551e-06,
+      "loss": 0.1038,
+      "step": 12900
+    },
+    {
+      "epoch": 65.86734693877551,
+      "grad_norm": 10.750667572021484,
+      "learning_rate": 6.826530612244898e-06,
+      "loss": 0.2203,
+      "step": 12910
+    },
+    {
+      "epoch": 65.91836734693878,
+      "grad_norm": 11.105408668518066,
+      "learning_rate": 6.816326530612245e-06,
+      "loss": 0.2605,
+      "step": 12920
+    },
+    {
+      "epoch": 65.96938775510205,
+      "grad_norm": 1.4005035161972046,
+      "learning_rate": 6.806122448979592e-06,
+      "loss": 0.1965,
+      "step": 12930
+    },
+    {
+      "epoch": 66.0,
+      "eval_accuracy": 0.851985559566787,
+      "eval_loss": 0.5033073425292969,
+      "eval_runtime": 0.929,
+      "eval_samples_per_second": 298.158,
+      "eval_steps_per_second": 37.673,
+      "step": 12936
+    },
+    {
+      "epoch": 66.0204081632653,
+      "grad_norm": 1.441422462463379,
+      "learning_rate": 6.7959183673469394e-06,
+      "loss": 0.3474,
+      "step": 12940
+    },
+    {
+      "epoch": 66.07142857142857,
+      "grad_norm": 0.8666236996650696,
+      "learning_rate": 6.785714285714287e-06,
+      "loss": 0.2731,
+      "step": 12950
+    },
+    {
+      "epoch": 66.12244897959184,
+      "grad_norm": 1.4008103609085083,
+      "learning_rate": 6.7755102040816336e-06,
+      "loss": 0.1877,
+      "step": 12960
+    },
+    {
+      "epoch": 66.1734693877551,
+      "grad_norm": 3.096766948699951,
+      "learning_rate": 6.76530612244898e-06,
+      "loss": 0.2621,
+      "step": 12970
+    },
+    {
+      "epoch": 66.22448979591837,
+      "grad_norm": 2.314729690551758,
+      "learning_rate": 6.755102040816327e-06,
+      "loss": 0.2495,
+      "step": 12980
+    },
+    {
+      "epoch": 66.27551020408163,
+      "grad_norm": 0.8084143400192261,
+      "learning_rate": 6.7448979591836735e-06,
+      "loss": 0.0909,
+      "step": 12990
+    },
+    {
+      "epoch": 66.3265306122449,
+      "grad_norm": 10.513947486877441,
+      "learning_rate": 6.734693877551021e-06,
+      "loss": 0.4269,
+      "step": 13000
+    },
+    {
+      "epoch": 66.37755102040816,
+      "grad_norm": 21.57809829711914,
+      "learning_rate": 6.724489795918368e-06,
+      "loss": 0.3193,
+      "step": 13010
+    },
+    {
+      "epoch": 66.42857142857143,
+      "grad_norm": 1.8246338367462158,
+      "learning_rate": 6.714285714285714e-06,
+      "loss": 0.0993,
+      "step": 13020
+    },
+    {
+      "epoch": 66.4795918367347,
+      "grad_norm": 10.447412490844727,
+      "learning_rate": 6.704081632653063e-06,
+      "loss": 0.1901,
+      "step": 13030
+    },
+    {
+      "epoch": 66.53061224489795,
+      "grad_norm": 0.9140769243240356,
+      "learning_rate": 6.693877551020409e-06,
+      "loss": 0.0921,
+      "step": 13040
+    },
+    {
+      "epoch": 66.58163265306122,
+      "grad_norm": 7.879315376281738,
+      "learning_rate": 6.683673469387756e-06,
+      "loss": 0.1556,
+      "step": 13050
+    },
+    {
+      "epoch": 66.63265306122449,
+      "grad_norm": 5.356064319610596,
+      "learning_rate": 6.6734693877551025e-06,
+      "loss": 0.1502,
+      "step": 13060
+    },
+    {
+      "epoch": 66.68367346938776,
+      "grad_norm": 3.6449520587921143,
+      "learning_rate": 6.663265306122449e-06,
+      "loss": 0.1708,
+      "step": 13070
+    },
+    {
+      "epoch": 66.73469387755102,
+      "grad_norm": 14.209073066711426,
+      "learning_rate": 6.653061224489797e-06,
+      "loss": 0.2792,
+      "step": 13080
+    },
+    {
+      "epoch": 66.78571428571429,
+      "grad_norm": 0.243174210190773,
+      "learning_rate": 6.642857142857143e-06,
+      "loss": 0.2626,
+      "step": 13090
+    },
+    {
+      "epoch": 66.83673469387755,
+      "grad_norm": 0.8402270674705505,
+      "learning_rate": 6.63265306122449e-06,
+      "loss": 0.341,
+      "step": 13100
+    },
+    {
+      "epoch": 66.88775510204081,
+      "grad_norm": 1.1840009689331055,
+      "learning_rate": 6.6224489795918365e-06,
+      "loss": 0.077,
+      "step": 13110
+    },
+    {
+      "epoch": 66.93877551020408,
+      "grad_norm": 1.2926697731018066,
+      "learning_rate": 6.612244897959185e-06,
+      "loss": 0.1844,
+      "step": 13120
+    },
+    {
+      "epoch": 66.98979591836735,
+      "grad_norm": 0.45596539974212646,
+      "learning_rate": 6.6020408163265315e-06,
+      "loss": 0.3548,
+      "step": 13130
+    },
+    {
+      "epoch": 67.0,
+      "eval_accuracy": 0.8808664259927798,
+      "eval_loss": 0.41321179270744324,
+      "eval_runtime": 0.9321,
+      "eval_samples_per_second": 297.182,
+      "eval_steps_per_second": 37.55,
+      "step": 13132
+    },
+    {
+      "epoch": 67.04081632653062,
+      "grad_norm": 6.747880458831787,
+      "learning_rate": 6.591836734693878e-06,
+      "loss": 0.2125,
+      "step": 13140
+    },
+    {
+      "epoch": 67.09183673469387,
+      "grad_norm": 12.555801391601562,
+      "learning_rate": 6.581632653061225e-06,
+      "loss": 0.1632,
+      "step": 13150
+    },
+    {
+      "epoch": 67.14285714285714,
+      "grad_norm": 11.827412605285645,
+      "learning_rate": 6.571428571428572e-06,
+      "loss": 0.1482,
+      "step": 13160
+    },
+    {
+      "epoch": 67.1938775510204,
+      "grad_norm": 2.0173394680023193,
+      "learning_rate": 6.561224489795919e-06,
+      "loss": 0.3297,
+      "step": 13170
+    },
+    {
+      "epoch": 67.24489795918367,
+      "grad_norm": 7.96342658996582,
+      "learning_rate": 6.5510204081632656e-06,
+      "loss": 0.093,
+      "step": 13180
+    },
+    {
+      "epoch": 67.29591836734694,
+      "grad_norm": 10.090936660766602,
+      "learning_rate": 6.540816326530612e-06,
+      "loss": 0.3328,
+      "step": 13190
+    },
+    {
+      "epoch": 67.34693877551021,
+      "grad_norm": 8.72697925567627,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 0.1716,
+      "step": 13200
+    },
+    {
+      "epoch": 67.39795918367346,
+      "grad_norm": 3.7887697219848633,
+      "learning_rate": 6.520408163265307e-06,
+      "loss": 0.316,
+      "step": 13210
+    },
+    {
+      "epoch": 67.44897959183673,
+      "grad_norm": 9.916276931762695,
+      "learning_rate": 6.510204081632654e-06,
+      "loss": 0.1719,
+      "step": 13220
+    },
+    {
+      "epoch": 67.5,
+      "grad_norm": 0.9465085864067078,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.1347,
+      "step": 13230
+    },
+    {
+      "epoch": 67.55102040816327,
+      "grad_norm": 2.688920259475708,
+      "learning_rate": 6.489795918367348e-06,
+      "loss": 0.2941,
+      "step": 13240
+    },
+    {
+      "epoch": 67.60204081632654,
+      "grad_norm": 4.175118446350098,
+      "learning_rate": 6.4795918367346946e-06,
+      "loss": 0.2066,
+      "step": 13250
+    },
+    {
+      "epoch": 67.65306122448979,
+      "grad_norm": 6.48569917678833,
+      "learning_rate": 6.469387755102041e-06,
+      "loss": 0.2816,
+      "step": 13260
+    },
+    {
+      "epoch": 67.70408163265306,
+      "grad_norm": 2.492340087890625,
+      "learning_rate": 6.459183673469388e-06,
+      "loss": 0.1307,
+      "step": 13270
+    },
+    {
+      "epoch": 67.75510204081633,
+      "grad_norm": 13.288766860961914,
+      "learning_rate": 6.4489795918367345e-06,
+      "loss": 0.2548,
+      "step": 13280
+    },
+    {
+      "epoch": 67.8061224489796,
+      "grad_norm": 2.2034645080566406,
+      "learning_rate": 6.438775510204082e-06,
+      "loss": 0.3435,
+      "step": 13290
+    },
+    {
+      "epoch": 67.85714285714286,
+      "grad_norm": 1.1759686470031738,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 0.2406,
+      "step": 13300
+    },
+    {
+      "epoch": 67.90816326530613,
+      "grad_norm": 17.009254455566406,
+      "learning_rate": 6.418367346938776e-06,
+      "loss": 0.1782,
+      "step": 13310
+    },
+    {
+      "epoch": 67.95918367346938,
+      "grad_norm": 7.602895736694336,
+      "learning_rate": 6.408163265306124e-06,
+      "loss": 0.1279,
+      "step": 13320
+    },
+    {
+      "epoch": 68.0,
+      "eval_accuracy": 0.9169675090252708,
+      "eval_loss": 0.37430626153945923,
+      "eval_runtime": 0.9377,
+      "eval_samples_per_second": 295.419,
+      "eval_steps_per_second": 37.327,
+      "step": 13328
+    },
+    {
+      "epoch": 68.01020408163265,
+      "grad_norm": 0.3676498830318451,
+      "learning_rate": 6.39795918367347e-06,
+      "loss": 0.387,
+      "step": 13330
+    },
+    {
+      "epoch": 68.06122448979592,
+      "grad_norm": 9.916852951049805,
+      "learning_rate": 6.387755102040817e-06,
+      "loss": 0.3069,
+      "step": 13340
+    },
+    {
+      "epoch": 68.11224489795919,
+      "grad_norm": 1.7918875217437744,
+      "learning_rate": 6.3775510204081635e-06,
+      "loss": 0.2864,
+      "step": 13350
+    },
+    {
+      "epoch": 68.16326530612245,
+      "grad_norm": 12.890216827392578,
+      "learning_rate": 6.36734693877551e-06,
+      "loss": 0.2125,
+      "step": 13360
+    },
+    {
+      "epoch": 68.21428571428571,
+      "grad_norm": 7.735411643981934,
+      "learning_rate": 6.357142857142858e-06,
+      "loss": 0.3003,
+      "step": 13370
+    },
+    {
+      "epoch": 68.26530612244898,
+      "grad_norm": 1.7586148977279663,
+      "learning_rate": 6.346938775510204e-06,
+      "loss": 0.1708,
+      "step": 13380
+    },
+    {
+      "epoch": 68.31632653061224,
+      "grad_norm": 2.7423510551452637,
+      "learning_rate": 6.336734693877552e-06,
+      "loss": 0.2475,
+      "step": 13390
+    },
+    {
+      "epoch": 68.36734693877551,
+      "grad_norm": 10.699454307556152,
+      "learning_rate": 6.326530612244899e-06,
+      "loss": 0.2548,
+      "step": 13400
+    },
+    {
+      "epoch": 68.41836734693878,
+      "grad_norm": 18.133024215698242,
+      "learning_rate": 6.316326530612246e-06,
+      "loss": 0.3625,
+      "step": 13410
+    },
+    {
+      "epoch": 68.46938775510205,
+      "grad_norm": 10.453315734863281,
+      "learning_rate": 6.3061224489795925e-06,
+      "loss": 0.1363,
+      "step": 13420
+    },
+    {
+      "epoch": 68.5204081632653,
+      "grad_norm": 17.546966552734375,
+      "learning_rate": 6.295918367346939e-06,
+      "loss": 0.1487,
+      "step": 13430
+    },
+    {
+      "epoch": 68.57142857142857,
+      "grad_norm": 6.659013271331787,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 0.1051,
+      "step": 13440
+    },
+    {
+      "epoch": 68.62244897959184,
+      "grad_norm": 10.408591270446777,
+      "learning_rate": 6.275510204081633e-06,
+      "loss": 0.1431,
+      "step": 13450
+    },
+    {
+      "epoch": 68.6734693877551,
+      "grad_norm": 1.8873507976531982,
+      "learning_rate": 6.26530612244898e-06,
+      "loss": 0.2465,
+      "step": 13460
+    },
+    {
+      "epoch": 68.72448979591837,
+      "grad_norm": 2.6099853515625,
+      "learning_rate": 6.2551020408163266e-06,
+      "loss": 0.1926,
+      "step": 13470
+    },
+    {
+      "epoch": 68.77551020408163,
+      "grad_norm": 6.034646034240723,
+      "learning_rate": 6.244897959183675e-06,
+      "loss": 0.3591,
+      "step": 13480
+    },
+    {
+      "epoch": 68.8265306122449,
+      "grad_norm": 9.02139949798584,
+      "learning_rate": 6.2346938775510215e-06,
+      "loss": 0.2112,
+      "step": 13490
+    },
+    {
+      "epoch": 68.87755102040816,
+      "grad_norm": 3.453343391418457,
+      "learning_rate": 6.224489795918368e-06,
+      "loss": 0.2187,
+      "step": 13500
+    },
+    {
+      "epoch": 68.92857142857143,
+      "grad_norm": 1.6684788465499878,
+      "learning_rate": 6.214285714285715e-06,
+      "loss": 0.2207,
+      "step": 13510
+    },
+    {
+      "epoch": 68.9795918367347,
+      "grad_norm": 0.5747165679931641,
+      "learning_rate": 6.2040816326530614e-06,
+      "loss": 0.2879,
+      "step": 13520
+    },
+    {
+      "epoch": 69.0,
+      "eval_accuracy": 0.776173285198556,
+      "eval_loss": 0.6423271298408508,
+      "eval_runtime": 0.9319,
+      "eval_samples_per_second": 297.254,
+      "eval_steps_per_second": 37.559,
+      "step": 13524
+    },
+    {
+      "epoch": 69.03061224489795,
+      "grad_norm": 14.904623985290527,
+      "learning_rate": 6.193877551020409e-06,
+      "loss": 0.3672,
+      "step": 13530
+    },
+    {
+      "epoch": 69.08163265306122,
+      "grad_norm": 10.467020034790039,
+      "learning_rate": 6.1836734693877556e-06,
+      "loss": 0.17,
+      "step": 13540
+    },
+    {
+      "epoch": 69.13265306122449,
+      "grad_norm": 0.9199422597885132,
+      "learning_rate": 6.173469387755102e-06,
+      "loss": 0.1159,
+      "step": 13550
+    },
+    {
+      "epoch": 69.18367346938776,
+      "grad_norm": 4.06253719329834,
+      "learning_rate": 6.163265306122449e-06,
+      "loss": 0.1432,
+      "step": 13560
+    },
+    {
+      "epoch": 69.23469387755102,
+      "grad_norm": 4.386868476867676,
+      "learning_rate": 6.153061224489797e-06,
+      "loss": 0.2992,
+      "step": 13570
+    },
+    {
+      "epoch": 69.28571428571429,
+      "grad_norm": 12.222877502441406,
+      "learning_rate": 6.142857142857144e-06,
+      "loss": 0.2958,
+      "step": 13580
+    },
+    {
+      "epoch": 69.33673469387755,
+      "grad_norm": 6.182301044464111,
+      "learning_rate": 6.1326530612244905e-06,
+      "loss": 0.1483,
+      "step": 13590
+    },
+    {
+      "epoch": 69.38775510204081,
+      "grad_norm": 0.8524590134620667,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 0.1975,
+      "step": 13600
+    },
+    {
+      "epoch": 69.43877551020408,
+      "grad_norm": 0.8242807388305664,
+      "learning_rate": 6.112244897959185e-06,
+      "loss": 0.719,
+      "step": 13610
+    },
+    {
+      "epoch": 69.48979591836735,
+      "grad_norm": 1.1100050210952759,
+      "learning_rate": 6.102040816326531e-06,
+      "loss": 0.266,
+      "step": 13620
+    },
+    {
+      "epoch": 69.54081632653062,
+      "grad_norm": 1.066702961921692,
+      "learning_rate": 6.091836734693878e-06,
+      "loss": 0.218,
+      "step": 13630
+    },
+    {
+      "epoch": 69.59183673469387,
+      "grad_norm": 2.186619281768799,
+      "learning_rate": 6.0816326530612245e-06,
+      "loss": 0.1463,
+      "step": 13640
+    },
+    {
+      "epoch": 69.64285714285714,
+      "grad_norm": 2.3016552925109863,
+      "learning_rate": 6.071428571428571e-06,
+      "loss": 0.1605,
+      "step": 13650
+    },
+    {
+      "epoch": 69.6938775510204,
+      "grad_norm": 9.424914360046387,
+      "learning_rate": 6.0612244897959195e-06,
+      "loss": 0.1386,
+      "step": 13660
+    },
+    {
+      "epoch": 69.74489795918367,
+      "grad_norm": 1.9103364944458008,
+      "learning_rate": 6.051020408163266e-06,
+      "loss": 0.1881,
+      "step": 13670
+    },
+    {
+      "epoch": 69.79591836734694,
+      "grad_norm": 3.562779188156128,
+      "learning_rate": 6.040816326530613e-06,
+      "loss": 0.146,
+      "step": 13680
+    },
+    {
+      "epoch": 69.84693877551021,
+      "grad_norm": 0.7750720977783203,
+      "learning_rate": 6.03061224489796e-06,
+      "loss": 0.0724,
+      "step": 13690
+    },
+    {
+      "epoch": 69.89795918367346,
+      "grad_norm": 0.9294744729995728,
+      "learning_rate": 6.020408163265307e-06,
+      "loss": 0.2269,
+      "step": 13700
+    },
+    {
+      "epoch": 69.94897959183673,
+      "grad_norm": 0.2986402213573456,
+      "learning_rate": 6.0102040816326535e-06,
+      "loss": 0.1076,
+      "step": 13710
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 5.967097759246826,
+      "learning_rate": 6e-06,
+      "loss": 0.1757,
+      "step": 13720
+    },
+    {
+      "epoch": 70.0,
+      "eval_accuracy": 0.8014440433212996,
+      "eval_loss": 0.5979345440864563,
+      "eval_runtime": 0.9344,
+      "eval_samples_per_second": 296.449,
+      "eval_steps_per_second": 37.457,
+      "step": 13720
+    },
+    {
+      "epoch": 70.05102040816327,
+      "grad_norm": 13.955671310424805,
+      "learning_rate": 5.989795918367347e-06,
+      "loss": 0.294,
+      "step": 13730
+    },
+    {
+      "epoch": 70.10204081632654,
+      "grad_norm": 0.3971206843852997,
+      "learning_rate": 5.979591836734694e-06,
+      "loss": 0.071,
+      "step": 13740
+    },
+    {
+      "epoch": 70.15306122448979,
+      "grad_norm": 6.5772528648376465,
+      "learning_rate": 5.969387755102042e-06,
+      "loss": 0.2846,
+      "step": 13750
+    },
+    {
+      "epoch": 70.20408163265306,
+      "grad_norm": 1.784285068511963,
+      "learning_rate": 5.959183673469388e-06,
+      "loss": 0.2171,
+      "step": 13760
+    },
+    {
+      "epoch": 70.25510204081633,
+      "grad_norm": 2.0860092639923096,
+      "learning_rate": 5.948979591836735e-06,
+      "loss": 0.2162,
+      "step": 13770
+    },
+    {
+      "epoch": 70.3061224489796,
+      "grad_norm": 0.8636461496353149,
+      "learning_rate": 5.9387755102040825e-06,
+      "loss": 0.1208,
+      "step": 13780
+    },
+    {
+      "epoch": 70.35714285714286,
+      "grad_norm": 0.5153712630271912,
+      "learning_rate": 5.928571428571429e-06,
+      "loss": 0.115,
+      "step": 13790
+    },
+    {
+      "epoch": 70.40816326530613,
+      "grad_norm": 9.724133491516113,
+      "learning_rate": 5.918367346938776e-06,
+      "loss": 0.2175,
+      "step": 13800
+    },
+    {
+      "epoch": 70.45918367346938,
+      "grad_norm": 7.4661407470703125,
+      "learning_rate": 5.9081632653061224e-06,
+      "loss": 0.2529,
+      "step": 13810
+    },
+    {
+      "epoch": 70.51020408163265,
+      "grad_norm": 10.019166946411133,
+      "learning_rate": 5.89795918367347e-06,
+      "loss": 0.4082,
+      "step": 13820
+    },
+    {
+      "epoch": 70.56122448979592,
+      "grad_norm": 1.403428316116333,
+      "learning_rate": 5.8877551020408166e-06,
+      "loss": 0.2125,
+      "step": 13830
+    },
+    {
+      "epoch": 70.61224489795919,
+      "grad_norm": 0.517578125,
+      "learning_rate": 5.877551020408164e-06,
+      "loss": 0.3907,
+      "step": 13840
+    },
+    {
+      "epoch": 70.66326530612245,
+      "grad_norm": 9.466553688049316,
+      "learning_rate": 5.867346938775511e-06,
+      "loss": 0.3029,
+      "step": 13850
+    },
+    {
+      "epoch": 70.71428571428571,
+      "grad_norm": 15.54366397857666,
+      "learning_rate": 5.857142857142858e-06,
+      "loss": 0.1536,
+      "step": 13860
+    },
+    {
+      "epoch": 70.76530612244898,
+      "grad_norm": 7.9617228507995605,
+      "learning_rate": 5.846938775510205e-06,
+      "loss": 0.1818,
+      "step": 13870
+    },
+    {
+      "epoch": 70.81632653061224,
+      "grad_norm": 1.013411521911621,
+      "learning_rate": 5.8367346938775515e-06,
+      "loss": 0.3629,
+      "step": 13880
+    },
+    {
+      "epoch": 70.86734693877551,
+      "grad_norm": 0.12895554304122925,
+      "learning_rate": 5.826530612244898e-06,
+      "loss": 0.1266,
+      "step": 13890
+    },
+    {
+      "epoch": 70.91836734693878,
+      "grad_norm": 10.661433219909668,
+      "learning_rate": 5.816326530612246e-06,
+      "loss": 0.2916,
+      "step": 13900
+    },
+    {
+      "epoch": 70.96938775510205,
+      "grad_norm": 16.746891021728516,
+      "learning_rate": 5.806122448979592e-06,
+      "loss": 0.3338,
+      "step": 13910
+    },
+    {
+      "epoch": 71.0,
+      "eval_accuracy": 0.8989169675090253,
+      "eval_loss": 0.43981537222862244,
+      "eval_runtime": 0.9347,
+      "eval_samples_per_second": 296.341,
+      "eval_steps_per_second": 37.444,
+      "step": 13916
+    },
+    {
+      "epoch": 71.0204081632653,
+      "grad_norm": 12.110264778137207,
+      "learning_rate": 5.795918367346939e-06,
+      "loss": 0.3042,
+      "step": 13920
+    },
+    {
+      "epoch": 71.07142857142857,
+      "grad_norm": 4.592965602874756,
+      "learning_rate": 5.785714285714286e-06,
+      "loss": 0.1833,
+      "step": 13930
+    },
+    {
+      "epoch": 71.12244897959184,
+      "grad_norm": 3.0122435092926025,
+      "learning_rate": 5.775510204081634e-06,
+      "loss": 0.3606,
+      "step": 13940
+    },
+    {
+      "epoch": 71.1734693877551,
+      "grad_norm": 2.8679792881011963,
+      "learning_rate": 5.7653061224489805e-06,
+      "loss": 0.2585,
+      "step": 13950
+    },
+    {
+      "epoch": 71.22448979591837,
+      "grad_norm": 1.3968089818954468,
+      "learning_rate": 5.755102040816327e-06,
+      "loss": 0.117,
+      "step": 13960
+    },
+    {
+      "epoch": 71.27551020408163,
+      "grad_norm": 4.6621174812316895,
+      "learning_rate": 5.744897959183674e-06,
+      "loss": 0.1251,
+      "step": 13970
+    },
+    {
+      "epoch": 71.3265306122449,
+      "grad_norm": 4.598996162414551,
+      "learning_rate": 5.73469387755102e-06,
+      "loss": 0.1581,
+      "step": 13980
+    },
+    {
+      "epoch": 71.37755102040816,
+      "grad_norm": 16.922142028808594,
+      "learning_rate": 5.724489795918368e-06,
+      "loss": 0.2025,
+      "step": 13990
+    },
+    {
+      "epoch": 71.42857142857143,
+      "grad_norm": 3.0897347927093506,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.2834,
+      "step": 14000
+    },
+    {
+      "epoch": 71.4795918367347,
+      "grad_norm": 18.263383865356445,
+      "learning_rate": 5.704081632653061e-06,
+      "loss": 0.4577,
+      "step": 14010
+    },
+    {
+      "epoch": 71.53061224489795,
+      "grad_norm": 0.9202457070350647,
+      "learning_rate": 5.6938775510204095e-06,
+      "loss": 0.318,
+      "step": 14020
+    },
+    {
+      "epoch": 71.58163265306122,
+      "grad_norm": 0.4576379358768463,
+      "learning_rate": 5.683673469387756e-06,
+      "loss": 0.1531,
+      "step": 14030
+    },
+    {
+      "epoch": 71.63265306122449,
+      "grad_norm": 17.588665008544922,
+      "learning_rate": 5.673469387755103e-06,
+      "loss": 0.1395,
+      "step": 14040
+    },
+    {
+      "epoch": 71.68367346938776,
+      "grad_norm": 14.638072967529297,
+      "learning_rate": 5.663265306122449e-06,
+      "loss": 0.2767,
+      "step": 14050
+    },
+    {
+      "epoch": 71.73469387755102,
+      "grad_norm": 2.2523481845855713,
+      "learning_rate": 5.653061224489796e-06,
+      "loss": 0.2819,
+      "step": 14060
+    },
+    {
+      "epoch": 71.78571428571429,
+      "grad_norm": 10.696439743041992,
+      "learning_rate": 5.6428571428571435e-06,
+      "loss": 0.1996,
+      "step": 14070
+    },
+    {
+      "epoch": 71.83673469387755,
+      "grad_norm": 13.24187183380127,
+      "learning_rate": 5.63265306122449e-06,
+      "loss": 0.2872,
+      "step": 14080
+    },
+    {
+      "epoch": 71.88775510204081,
+      "grad_norm": 4.308647632598877,
+      "learning_rate": 5.622448979591837e-06,
+      "loss": 0.1112,
+      "step": 14090
+    },
+    {
+      "epoch": 71.93877551020408,
+      "grad_norm": 14.375484466552734,
+      "learning_rate": 5.6122448979591834e-06,
+      "loss": 0.1914,
+      "step": 14100
+    },
+    {
+      "epoch": 71.98979591836735,
+      "grad_norm": 6.740991115570068,
+      "learning_rate": 5.602040816326531e-06,
+      "loss": 0.1604,
+      "step": 14110
+    },
+    {
+      "epoch": 72.0,
+      "eval_accuracy": 0.8231046931407943,
+      "eval_loss": 0.5633912086486816,
+      "eval_runtime": 0.9307,
+      "eval_samples_per_second": 297.639,
+      "eval_steps_per_second": 37.608,
+      "step": 14112
+    },
+    {
+      "epoch": 72.04081632653062,
+      "grad_norm": 0.3390078544616699,
+      "learning_rate": 5.591836734693878e-06,
+      "loss": 0.3075,
+      "step": 14120
+    },
+    {
+      "epoch": 72.09183673469387,
+      "grad_norm": 4.501482009887695,
+      "learning_rate": 5.581632653061225e-06,
+      "loss": 0.1095,
+      "step": 14130
+    },
+    {
+      "epoch": 72.14285714285714,
+      "grad_norm": 3.729031562805176,
+      "learning_rate": 5.571428571428572e-06,
+      "loss": 0.0767,
+      "step": 14140
+    },
+    {
+      "epoch": 72.1938775510204,
+      "grad_norm": 0.17877550423145294,
+      "learning_rate": 5.561224489795919e-06,
+      "loss": 0.3787,
+      "step": 14150
+    },
+    {
+      "epoch": 72.24489795918367,
+      "grad_norm": 5.799465179443359,
+      "learning_rate": 5.551020408163266e-06,
+      "loss": 0.1648,
+      "step": 14160
+    },
+    {
+      "epoch": 72.29591836734694,
+      "grad_norm": 16.26845359802246,
+      "learning_rate": 5.5408163265306125e-06,
+      "loss": 0.1484,
+      "step": 14170
+    },
+    {
+      "epoch": 72.34693877551021,
+      "grad_norm": 3.746767997741699,
+      "learning_rate": 5.530612244897959e-06,
+      "loss": 0.1447,
+      "step": 14180
+    },
+    {
+      "epoch": 72.39795918367346,
+      "grad_norm": 1.8594543933868408,
+      "learning_rate": 5.520408163265306e-06,
+      "loss": 0.1184,
+      "step": 14190
+    },
+    {
+      "epoch": 72.44897959183673,
+      "grad_norm": 10.922441482543945,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 0.1443,
+      "step": 14200
+    },
+    {
+      "epoch": 72.5,
+      "grad_norm": 1.249107003211975,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.189,
+      "step": 14210
+    },
+    {
+      "epoch": 72.55102040816327,
+      "grad_norm": 1.8553344011306763,
+      "learning_rate": 5.489795918367347e-06,
+      "loss": 0.2099,
+      "step": 14220
+    },
+    {
+      "epoch": 72.60204081632654,
+      "grad_norm": 19.02747344970703,
+      "learning_rate": 5.479591836734695e-06,
+      "loss": 0.4534,
+      "step": 14230
+    },
+    {
+      "epoch": 72.65306122448979,
+      "grad_norm": 0.5484607219696045,
+      "learning_rate": 5.4693877551020415e-06,
+      "loss": 0.1987,
+      "step": 14240
+    },
+    {
+      "epoch": 72.70408163265306,
+      "grad_norm": 3.448791265487671,
+      "learning_rate": 5.459183673469388e-06,
+      "loss": 0.1921,
+      "step": 14250
+    },
+    {
+      "epoch": 72.75510204081633,
+      "grad_norm": 0.8846081495285034,
+      "learning_rate": 5.448979591836735e-06,
+      "loss": 0.141,
+      "step": 14260
+    },
+    {
+      "epoch": 72.8061224489796,
+      "grad_norm": 9.476802825927734,
+      "learning_rate": 5.438775510204081e-06,
+      "loss": 0.2398,
+      "step": 14270
+    },
+    {
+      "epoch": 72.85714285714286,
+      "grad_norm": 8.394309997558594,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 0.1556,
+      "step": 14280
+    },
+    {
+      "epoch": 72.90816326530613,
+      "grad_norm": 2.2207627296447754,
+      "learning_rate": 5.4183673469387755e-06,
+      "loss": 0.1218,
+      "step": 14290
+    },
+    {
+      "epoch": 72.95918367346938,
+      "grad_norm": 5.007809162139893,
+      "learning_rate": 5.408163265306123e-06,
+      "loss": 0.1078,
+      "step": 14300
+    },
+    {
+      "epoch": 73.0,
+      "eval_accuracy": 0.776173285198556,
+      "eval_loss": 0.620398223400116,
+      "eval_runtime": 0.9313,
+      "eval_samples_per_second": 297.423,
+      "eval_steps_per_second": 37.581,
+      "step": 14308
+    },
+    {
+      "epoch": 73.01020408163265,
+      "grad_norm": 5.273265838623047,
+      "learning_rate": 5.3979591836734705e-06,
+      "loss": 0.1099,
+      "step": 14310
+    },
+    {
+      "epoch": 73.06122448979592,
+      "grad_norm": 0.09366770088672638,
+      "learning_rate": 5.387755102040817e-06,
+      "loss": 0.3264,
+      "step": 14320
+    },
+    {
+      "epoch": 73.11224489795919,
+      "grad_norm": 12.97093677520752,
+      "learning_rate": 5.377551020408164e-06,
+      "loss": 0.2472,
+      "step": 14330
+    },
+    {
+      "epoch": 73.16326530612245,
+      "grad_norm": 23.67729377746582,
+      "learning_rate": 5.36734693877551e-06,
+      "loss": 0.3731,
+      "step": 14340
+    },
+    {
+      "epoch": 73.21428571428571,
+      "grad_norm": 3.248290777206421,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.1723,
+      "step": 14350
+    },
+    {
+      "epoch": 73.26530612244898,
+      "grad_norm": 0.6866236925125122,
+      "learning_rate": 5.3469387755102045e-06,
+      "loss": 0.2337,
+      "step": 14360
+    },
+    {
+      "epoch": 73.31632653061224,
+      "grad_norm": 11.86717700958252,
+      "learning_rate": 5.336734693877551e-06,
+      "loss": 0.2086,
+      "step": 14370
+    },
+    {
+      "epoch": 73.36734693877551,
+      "grad_norm": 0.9687001705169678,
+      "learning_rate": 5.326530612244898e-06,
+      "loss": 0.1213,
+      "step": 14380
+    },
+    {
+      "epoch": 73.41836734693878,
+      "grad_norm": 1.566986322402954,
+      "learning_rate": 5.316326530612246e-06,
+      "loss": 0.3542,
+      "step": 14390
+    },
+    {
+      "epoch": 73.46938775510205,
+      "grad_norm": 11.866606712341309,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 0.5427,
+      "step": 14400
+    },
+    {
+      "epoch": 73.5204081632653,
+      "grad_norm": 3.2856667041778564,
+      "learning_rate": 5.295918367346939e-06,
+      "loss": 0.1367,
+      "step": 14410
+    },
+    {
+      "epoch": 73.57142857142857,
+      "grad_norm": 7.826207160949707,
+      "learning_rate": 5.285714285714286e-06,
+      "loss": 0.1293,
+      "step": 14420
+    },
+    {
+      "epoch": 73.62244897959184,
+      "grad_norm": 7.829553127288818,
+      "learning_rate": 5.275510204081633e-06,
+      "loss": 0.2473,
+      "step": 14430
+    },
+    {
+      "epoch": 73.6734693877551,
+      "grad_norm": 10.244202613830566,
+      "learning_rate": 5.26530612244898e-06,
+      "loss": 0.3114,
+      "step": 14440
+    },
+    {
+      "epoch": 73.72448979591837,
+      "grad_norm": 15.15924072265625,
+      "learning_rate": 5.255102040816327e-06,
+      "loss": 0.1595,
+      "step": 14450
+    },
+    {
+      "epoch": 73.77551020408163,
+      "grad_norm": 10.309656143188477,
+      "learning_rate": 5.2448979591836735e-06,
+      "loss": 0.3066,
+      "step": 14460
+    },
+    {
+      "epoch": 73.8265306122449,
+      "grad_norm": 2.3161091804504395,
+      "learning_rate": 5.23469387755102e-06,
+      "loss": 0.1034,
+      "step": 14470
+    },
+    {
+      "epoch": 73.87755102040816,
+      "grad_norm": 4.930816173553467,
+      "learning_rate": 5.2244897959183684e-06,
+      "loss": 0.1057,
+      "step": 14480
+    },
+    {
+      "epoch": 73.92857142857143,
+      "grad_norm": 3.2863380908966064,
+      "learning_rate": 5.214285714285715e-06,
+      "loss": 0.1034,
+      "step": 14490
+    },
+    {
+      "epoch": 73.9795918367347,
+      "grad_norm": 3.996938705444336,
+      "learning_rate": 5.204081632653062e-06,
+      "loss": 0.258,
+      "step": 14500
+    },
+    {
+      "epoch": 74.0,
+      "eval_accuracy": 0.8953068592057761,
+      "eval_loss": 0.36853671073913574,
+      "eval_runtime": 0.9322,
+      "eval_samples_per_second": 297.152,
+      "eval_steps_per_second": 37.546,
+      "step": 14504
+    },
+    {
+      "epoch": 74.03061224489795,
+      "grad_norm": 5.564041614532471,
+      "learning_rate": 5.193877551020408e-06,
+      "loss": 0.152,
+      "step": 14510
+    },
+    {
+      "epoch": 74.08163265306122,
+      "grad_norm": 14.177313804626465,
+      "learning_rate": 5.183673469387756e-06,
+      "loss": 0.2189,
+      "step": 14520
+    },
+    {
+      "epoch": 74.13265306122449,
+      "grad_norm": 1.9523788690567017,
+      "learning_rate": 5.1734693877551025e-06,
+      "loss": 0.2653,
+      "step": 14530
+    },
+    {
+      "epoch": 74.18367346938776,
+      "grad_norm": 0.6890932321548462,
+      "learning_rate": 5.163265306122449e-06,
+      "loss": 0.1789,
+      "step": 14540
+    },
+    {
+      "epoch": 74.23469387755102,
+      "grad_norm": 15.715421676635742,
+      "learning_rate": 5.153061224489796e-06,
+      "loss": 0.4299,
+      "step": 14550
+    },
+    {
+      "epoch": 74.28571428571429,
+      "grad_norm": 1.4377721548080444,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 0.1145,
+      "step": 14560
+    },
+    {
+      "epoch": 74.33673469387755,
+      "grad_norm": 21.627445220947266,
+      "learning_rate": 5.132653061224491e-06,
+      "loss": 0.412,
+      "step": 14570
+    },
+    {
+      "epoch": 74.38775510204081,
+      "grad_norm": 6.41222620010376,
+      "learning_rate": 5.122448979591837e-06,
+      "loss": 0.4198,
+      "step": 14580
+    },
+    {
+      "epoch": 74.43877551020408,
+      "grad_norm": 1.1630773544311523,
+      "learning_rate": 5.112244897959184e-06,
+      "loss": 0.1438,
+      "step": 14590
+    },
+    {
+      "epoch": 74.48979591836735,
+      "grad_norm": 11.959744453430176,
+      "learning_rate": 5.1020408163265315e-06,
+      "loss": 0.1354,
+      "step": 14600
+    },
+    {
+      "epoch": 74.54081632653062,
+      "grad_norm": 12.669620513916016,
+      "learning_rate": 5.091836734693878e-06,
+      "loss": 0.2402,
+      "step": 14610
+    },
+    {
+      "epoch": 74.59183673469387,
+      "grad_norm": 11.747788429260254,
+      "learning_rate": 5.081632653061225e-06,
+      "loss": 0.0847,
+      "step": 14620
+    },
+    {
+      "epoch": 74.64285714285714,
+      "grad_norm": 15.508429527282715,
+      "learning_rate": 5.071428571428571e-06,
+      "loss": 0.2417,
+      "step": 14630
+    },
+    {
+      "epoch": 74.6938775510204,
+      "grad_norm": 1.273260474205017,
+      "learning_rate": 5.061224489795918e-06,
+      "loss": 0.2925,
+      "step": 14640
+    },
+    {
+      "epoch": 74.74489795918367,
+      "grad_norm": 0.26550906896591187,
+      "learning_rate": 5.0510204081632655e-06,
+      "loss": 0.2467,
+      "step": 14650
+    },
+    {
+      "epoch": 74.79591836734694,
+      "grad_norm": 5.7561936378479,
+      "learning_rate": 5.040816326530613e-06,
+      "loss": 0.1374,
+      "step": 14660
+    },
+    {
+      "epoch": 74.84693877551021,
+      "grad_norm": 10.998912811279297,
+      "learning_rate": 5.03061224489796e-06,
+      "loss": 0.1987,
+      "step": 14670
+    },
+    {
+      "epoch": 74.89795918367346,
+      "grad_norm": 2.961043357849121,
+      "learning_rate": 5.020408163265307e-06,
+      "loss": 0.1702,
+      "step": 14680
+    },
+    {
+      "epoch": 74.94897959183673,
+      "grad_norm": 11.977360725402832,
+      "learning_rate": 5.010204081632654e-06,
+      "loss": 0.2016,
+      "step": 14690
+    },
+    {
+      "epoch": 75.0,
+      "grad_norm": 0.5454691052436829,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 14700
+    },
+    {
+      "epoch": 75.0,
+      "eval_accuracy": 0.8158844765342961,
+      "eval_loss": 0.7026161551475525,
+      "eval_runtime": 0.9329,
+      "eval_samples_per_second": 296.916,
+      "eval_steps_per_second": 37.517,
+      "step": 14700
+    },
+    {
+      "epoch": 75.05102040816327,
+      "grad_norm": 6.788924217224121,
+      "learning_rate": 4.989795918367347e-06,
+      "loss": 0.1771,
+      "step": 14710
+    },
+    {
+      "epoch": 75.10204081632654,
+      "grad_norm": 0.8358808159828186,
+      "learning_rate": 4.979591836734694e-06,
+      "loss": 0.4148,
+      "step": 14720
+    },
+    {
+      "epoch": 75.15306122448979,
+      "grad_norm": 2.5965681076049805,
+      "learning_rate": 4.969387755102041e-06,
+      "loss": 0.1705,
+      "step": 14730
+    },
+    {
+      "epoch": 75.20408163265306,
+      "grad_norm": 2.074181079864502,
+      "learning_rate": 4.959183673469388e-06,
+      "loss": 0.2377,
+      "step": 14740
+    },
+    {
+      "epoch": 75.25510204081633,
+      "grad_norm": 8.877419471740723,
+      "learning_rate": 4.948979591836735e-06,
+      "loss": 0.2705,
+      "step": 14750
+    },
+    {
+      "epoch": 75.3061224489796,
+      "grad_norm": 2.2369658946990967,
+      "learning_rate": 4.938775510204082e-06,
+      "loss": 0.1848,
+      "step": 14760
+    },
+    {
+      "epoch": 75.35714285714286,
+      "grad_norm": 1.3437224626541138,
+      "learning_rate": 4.928571428571429e-06,
+      "loss": 0.309,
+      "step": 14770
+    },
+    {
+      "epoch": 75.40816326530613,
+      "grad_norm": 12.179243087768555,
+      "learning_rate": 4.918367346938776e-06,
+      "loss": 0.3758,
+      "step": 14780
+    },
+    {
+      "epoch": 75.45918367346938,
+      "grad_norm": 0.31163737177848816,
+      "learning_rate": 4.908163265306123e-06,
+      "loss": 0.0913,
+      "step": 14790
+    },
+    {
+      "epoch": 75.51020408163265,
+      "grad_norm": 6.074469089508057,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 0.1946,
+      "step": 14800
+    },
+    {
+      "epoch": 75.56122448979592,
+      "grad_norm": 0.20314039289951324,
+      "learning_rate": 4.887755102040817e-06,
+      "loss": 0.3418,
+      "step": 14810
+    },
+    {
+      "epoch": 75.61224489795919,
+      "grad_norm": 1.0009397268295288,
+      "learning_rate": 4.8775510204081635e-06,
+      "loss": 0.1232,
+      "step": 14820
+    },
+    {
+      "epoch": 75.66326530612245,
+      "grad_norm": 4.9547505378723145,
+      "learning_rate": 4.867346938775511e-06,
+      "loss": 0.179,
+      "step": 14830
+    },
+    {
+      "epoch": 75.71428571428571,
+      "grad_norm": 14.29842472076416,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 0.2963,
+      "step": 14840
+    },
+    {
+      "epoch": 75.76530612244898,
+      "grad_norm": 0.6081609129905701,
+      "learning_rate": 4.846938775510204e-06,
+      "loss": 0.3408,
+      "step": 14850
+    },
+    {
+      "epoch": 75.81632653061224,
+      "grad_norm": 16.967266082763672,
+      "learning_rate": 4.836734693877552e-06,
+      "loss": 0.2528,
+      "step": 14860
+    },
+    {
+      "epoch": 75.86734693877551,
+      "grad_norm": 4.57008171081543,
+      "learning_rate": 4.826530612244898e-06,
+      "loss": 0.3173,
+      "step": 14870
+    },
+    {
+      "epoch": 75.91836734693878,
+      "grad_norm": 10.530519485473633,
+      "learning_rate": 4.816326530612245e-06,
+      "loss": 0.1651,
+      "step": 14880
+    },
+    {
+      "epoch": 75.96938775510205,
+      "grad_norm": 15.925687789916992,
+      "learning_rate": 4.8061224489795925e-06,
+      "loss": 0.2257,
+      "step": 14890
+    },
+    {
+      "epoch": 76.0,
+      "eval_accuracy": 0.9169675090252708,
+      "eval_loss": 0.40481600165367126,
+      "eval_runtime": 0.942,
+      "eval_samples_per_second": 294.046,
+      "eval_steps_per_second": 37.154,
+      "step": 14896
+    },
+    {
+      "epoch": 76.0204081632653,
+      "grad_norm": 5.312804222106934,
+      "learning_rate": 4.795918367346939e-06,
+      "loss": 0.4119,
+      "step": 14900
+    },
+    {
+      "epoch": 76.07142857142857,
+      "grad_norm": 2.115060567855835,
+      "learning_rate": 4.785714285714287e-06,
+      "loss": 0.1965,
+      "step": 14910
+    },
+    {
+      "epoch": 76.12244897959184,
+      "grad_norm": 2.3201823234558105,
+      "learning_rate": 4.775510204081633e-06,
+      "loss": 0.1222,
+      "step": 14920
+    },
+    {
+      "epoch": 76.1734693877551,
+      "grad_norm": 9.353812217712402,
+      "learning_rate": 4.76530612244898e-06,
+      "loss": 0.1391,
+      "step": 14930
+    },
+    {
+      "epoch": 76.22448979591837,
+      "grad_norm": 11.522284507751465,
+      "learning_rate": 4.7551020408163265e-06,
+      "loss": 0.2193,
+      "step": 14940
+    },
+    {
+      "epoch": 76.27551020408163,
+      "grad_norm": 9.714153289794922,
+      "learning_rate": 4.744897959183674e-06,
+      "loss": 0.3002,
+      "step": 14950
+    },
+    {
+      "epoch": 76.3265306122449,
+      "grad_norm": 1.5741373300552368,
+      "learning_rate": 4.734693877551021e-06,
+      "loss": 0.0989,
+      "step": 14960
+    },
+    {
+      "epoch": 76.37755102040816,
+      "grad_norm": 12.37441635131836,
+      "learning_rate": 4.724489795918368e-06,
+      "loss": 0.2207,
+      "step": 14970
+    },
+    {
+      "epoch": 76.42857142857143,
+      "grad_norm": 1.2197359800338745,
+      "learning_rate": 4.714285714285715e-06,
+      "loss": 0.3594,
+      "step": 14980
+    },
+    {
+      "epoch": 76.4795918367347,
+      "grad_norm": 1.7340853214263916,
+      "learning_rate": 4.704081632653061e-06,
+      "loss": 0.1337,
+      "step": 14990
+    },
+    {
+      "epoch": 76.53061224489795,
+      "grad_norm": 8.35191822052002,
+      "learning_rate": 4.693877551020409e-06,
+      "loss": 0.1877,
+      "step": 15000
+    },
+    {
+      "epoch": 76.58163265306122,
+      "grad_norm": 9.665618896484375,
+      "learning_rate": 4.6836734693877555e-06,
+      "loss": 0.2073,
+      "step": 15010
+    },
+    {
+      "epoch": 76.63265306122449,
+      "grad_norm": 0.48984846472740173,
+      "learning_rate": 4.673469387755102e-06,
+      "loss": 0.3014,
+      "step": 15020
+    },
+    {
+      "epoch": 76.68367346938776,
+      "grad_norm": 2.9640002250671387,
+      "learning_rate": 4.663265306122449e-06,
+      "loss": 0.215,
+      "step": 15030
+    },
+    {
+      "epoch": 76.73469387755102,
+      "grad_norm": 3.568899154663086,
+      "learning_rate": 4.653061224489796e-06,
+      "loss": 0.1687,
+      "step": 15040
+    },
+    {
+      "epoch": 76.78571428571429,
+      "grad_norm": 3.5330944061279297,
+      "learning_rate": 4.642857142857144e-06,
+      "loss": 0.2252,
+      "step": 15050
+    },
+    {
+      "epoch": 76.83673469387755,
+      "grad_norm": 7.781928062438965,
+      "learning_rate": 4.63265306122449e-06,
+      "loss": 0.212,
+      "step": 15060
+    },
+    {
+      "epoch": 76.88775510204081,
+      "grad_norm": 3.0376219749450684,
+      "learning_rate": 4.622448979591837e-06,
+      "loss": 0.141,
+      "step": 15070
+    },
+    {
+      "epoch": 76.93877551020408,
+      "grad_norm": 1.0159159898757935,
+      "learning_rate": 4.612244897959184e-06,
+      "loss": 0.2339,
+      "step": 15080
+    },
+    {
+      "epoch": 76.98979591836735,
+      "grad_norm": 3.5187315940856934,
+      "learning_rate": 4.602040816326531e-06,
+      "loss": 0.1786,
+      "step": 15090
+    },
+    {
+      "epoch": 77.0,
+      "eval_accuracy": 0.8844765342960289,
+      "eval_loss": 0.48912858963012695,
+      "eval_runtime": 1.0003,
+      "eval_samples_per_second": 276.912,
+      "eval_steps_per_second": 34.989,
+      "step": 15092
+    },
+    {
+      "epoch": 77.04081632653062,
+      "grad_norm": 0.9008731842041016,
+      "learning_rate": 4.591836734693878e-06,
+      "loss": 0.2919,
+      "step": 15100
+    },
+    {
+      "epoch": 77.09183673469387,
+      "grad_norm": 7.379002094268799,
+      "learning_rate": 4.5816326530612245e-06,
+      "loss": 0.2478,
+      "step": 15110
+    },
+    {
+      "epoch": 77.14285714285714,
+      "grad_norm": 1.654058575630188,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 0.1717,
+      "step": 15120
+    },
+    {
+      "epoch": 77.1938775510204,
+      "grad_norm": 10.02890396118164,
+      "learning_rate": 4.561224489795919e-06,
+      "loss": 0.1046,
+      "step": 15130
+    },
+    {
+      "epoch": 77.24489795918367,
+      "grad_norm": 1.1923608779907227,
+      "learning_rate": 4.551020408163266e-06,
+      "loss": 0.1507,
+      "step": 15140
+    },
+    {
+      "epoch": 77.29591836734694,
+      "grad_norm": 2.4354844093322754,
+      "learning_rate": 4.540816326530613e-06,
+      "loss": 0.1582,
+      "step": 15150
+    },
+    {
+      "epoch": 77.34693877551021,
+      "grad_norm": 7.890419960021973,
+      "learning_rate": 4.530612244897959e-06,
+      "loss": 0.2618,
+      "step": 15160
+    },
+    {
+      "epoch": 77.39795918367346,
+      "grad_norm": 5.627359867095947,
+      "learning_rate": 4.520408163265306e-06,
+      "loss": 0.1926,
+      "step": 15170
+    },
+    {
+      "epoch": 77.44897959183673,
+      "grad_norm": 3.1545939445495605,
+      "learning_rate": 4.5102040816326535e-06,
+      "loss": 0.2936,
+      "step": 15180
+    },
+    {
+      "epoch": 77.5,
+      "grad_norm": 10.855856895446777,
+      "learning_rate": 4.5e-06,
+      "loss": 0.6137,
+      "step": 15190
+    },
+    {
+      "epoch": 77.55102040816327,
+      "grad_norm": 7.537050247192383,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 0.2718,
+      "step": 15200
+    },
+    {
+      "epoch": 77.60204081632654,
+      "grad_norm": 10.528838157653809,
+      "learning_rate": 4.479591836734694e-06,
+      "loss": 0.2494,
+      "step": 15210
+    },
+    {
+      "epoch": 77.65306122448979,
+      "grad_norm": 0.04168795049190521,
+      "learning_rate": 4.469387755102041e-06,
+      "loss": 0.3021,
+      "step": 15220
+    },
+    {
+      "epoch": 77.70408163265306,
+      "grad_norm": 6.072834014892578,
+      "learning_rate": 4.459183673469388e-06,
+      "loss": 0.1545,
+      "step": 15230
+    },
+    {
+      "epoch": 77.75510204081633,
+      "grad_norm": 1.0188490152359009,
+      "learning_rate": 4.448979591836735e-06,
+      "loss": 0.154,
+      "step": 15240
+    },
+    {
+      "epoch": 77.8061224489796,
+      "grad_norm": 2.8648037910461426,
+      "learning_rate": 4.438775510204082e-06,
+      "loss": 0.1568,
+      "step": 15250
+    },
+    {
+      "epoch": 77.85714285714286,
+      "grad_norm": 2.9121363162994385,
+      "learning_rate": 4.428571428571429e-06,
+      "loss": 0.1795,
+      "step": 15260
+    },
+    {
+      "epoch": 77.90816326530613,
+      "grad_norm": 7.241103172302246,
+      "learning_rate": 4.418367346938776e-06,
+      "loss": 0.1361,
+      "step": 15270
+    },
+    {
+      "epoch": 77.95918367346938,
+      "grad_norm": 18.421598434448242,
+      "learning_rate": 4.408163265306123e-06,
+      "loss": 0.2006,
+      "step": 15280
+    },
+    {
+      "epoch": 78.0,
+      "eval_accuracy": 0.8772563176895307,
+      "eval_loss": 0.42159637808799744,
+      "eval_runtime": 0.9323,
+      "eval_samples_per_second": 297.124,
+      "eval_steps_per_second": 37.543,
+      "step": 15288
+    },
+    {
+      "epoch": 78.01020408163265,
+      "grad_norm": 0.5753517150878906,
+      "learning_rate": 4.39795918367347e-06,
+      "loss": 0.2196,
+      "step": 15290
+    },
+    {
+      "epoch": 78.06122448979592,
+      "grad_norm": 6.517210960388184,
+      "learning_rate": 4.3877551020408165e-06,
+      "loss": 0.1645,
+      "step": 15300
+    },
+    {
+      "epoch": 78.11224489795919,
+      "grad_norm": 1.0163953304290771,
+      "learning_rate": 4.377551020408163e-06,
+      "loss": 0.1594,
+      "step": 15310
+    },
+    {
+      "epoch": 78.16326530612245,
+      "grad_norm": 0.28174296021461487,
+      "learning_rate": 4.367346938775511e-06,
+      "loss": 0.3152,
+      "step": 15320
+    },
+    {
+      "epoch": 78.21428571428571,
+      "grad_norm": 8.987310409545898,
+      "learning_rate": 4.357142857142857e-06,
+      "loss": 0.1839,
+      "step": 15330
+    },
+    {
+      "epoch": 78.26530612244898,
+      "grad_norm": 3.5911307334899902,
+      "learning_rate": 4.346938775510205e-06,
+      "loss": 0.2263,
+      "step": 15340
+    },
+    {
+      "epoch": 78.31632653061224,
+      "grad_norm": 5.655363082885742,
+      "learning_rate": 4.336734693877551e-06,
+      "loss": 0.1143,
+      "step": 15350
+    },
+    {
+      "epoch": 78.36734693877551,
+      "grad_norm": 8.902242660522461,
+      "learning_rate": 4.326530612244899e-06,
+      "loss": 0.3429,
+      "step": 15360
+    },
+    {
+      "epoch": 78.41836734693878,
+      "grad_norm": 0.3062533140182495,
+      "learning_rate": 4.3163265306122455e-06,
+      "loss": 0.1505,
+      "step": 15370
+    },
+    {
+      "epoch": 78.46938775510205,
+      "grad_norm": 2.60772442817688,
+      "learning_rate": 4.306122448979592e-06,
+      "loss": 0.1027,
+      "step": 15380
+    },
+    {
+      "epoch": 78.5204081632653,
+      "grad_norm": 5.816764831542969,
+      "learning_rate": 4.295918367346939e-06,
+      "loss": 0.276,
+      "step": 15390
+    },
+    {
+      "epoch": 78.57142857142857,
+      "grad_norm": 11.878080368041992,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.3061,
+      "step": 15400
+    },
+    {
+      "epoch": 78.62244897959184,
+      "grad_norm": 2.4790613651275635,
+      "learning_rate": 4.275510204081633e-06,
+      "loss": 0.1698,
+      "step": 15410
+    },
+    {
+      "epoch": 78.6734693877551,
+      "grad_norm": 2.452232837677002,
+      "learning_rate": 4.2653061224489804e-06,
+      "loss": 0.0984,
+      "step": 15420
+    },
+    {
+      "epoch": 78.72448979591837,
+      "grad_norm": 3.86421799659729,
+      "learning_rate": 4.255102040816327e-06,
+      "loss": 0.2261,
+      "step": 15430
+    },
+    {
+      "epoch": 78.77551020408163,
+      "grad_norm": 6.461427688598633,
+      "learning_rate": 4.244897959183674e-06,
+      "loss": 0.1279,
+      "step": 15440
+    },
+    {
+      "epoch": 78.8265306122449,
+      "grad_norm": 0.5817276835441589,
+      "learning_rate": 4.234693877551021e-06,
+      "loss": 0.238,
+      "step": 15450
+    },
+    {
+      "epoch": 78.87755102040816,
+      "grad_norm": 1.191446304321289,
+      "learning_rate": 4.224489795918368e-06,
+      "loss": 0.1597,
+      "step": 15460
+    },
+    {
+      "epoch": 78.92857142857143,
+      "grad_norm": 0.9815487265586853,
+      "learning_rate": 4.2142857142857145e-06,
+      "loss": 0.1815,
+      "step": 15470
+    },
+    {
+      "epoch": 78.9795918367347,
+      "grad_norm": 13.951576232910156,
+      "learning_rate": 4.204081632653061e-06,
+      "loss": 0.3144,
+      "step": 15480
+    },
+    {
+      "epoch": 79.0,
+      "eval_accuracy": 0.8953068592057761,
+      "eval_loss": 0.2720978558063507,
+      "eval_runtime": 1.0029,
+      "eval_samples_per_second": 276.21,
+      "eval_steps_per_second": 34.9,
+      "step": 15484
+    },
+    {
+      "epoch": 79.03061224489795,
+      "grad_norm": 0.6935974359512329,
+      "learning_rate": 4.193877551020409e-06,
+      "loss": 0.1547,
+      "step": 15490
+    },
+    {
+      "epoch": 79.08163265306122,
+      "grad_norm": 4.269100666046143,
+      "learning_rate": 4.183673469387755e-06,
+      "loss": 0.2228,
+      "step": 15500
+    },
+    {
+      "epoch": 79.13265306122449,
+      "grad_norm": 16.676679611206055,
+      "learning_rate": 4.173469387755103e-06,
+      "loss": 0.2051,
+      "step": 15510
+    },
+    {
+      "epoch": 79.18367346938776,
+      "grad_norm": 1.743389368057251,
+      "learning_rate": 4.163265306122449e-06,
+      "loss": 0.0857,
+      "step": 15520
+    },
+    {
+      "epoch": 79.23469387755102,
+      "grad_norm": 5.236385822296143,
+      "learning_rate": 4.153061224489796e-06,
+      "loss": 0.1267,
+      "step": 15530
+    },
+    {
+      "epoch": 79.28571428571429,
+      "grad_norm": 6.302170753479004,
+      "learning_rate": 4.1428571428571435e-06,
+      "loss": 0.1847,
+      "step": 15540
+    },
+    {
+      "epoch": 79.33673469387755,
+      "grad_norm": 12.430984497070312,
+      "learning_rate": 4.13265306122449e-06,
+      "loss": 0.2319,
+      "step": 15550
+    },
+    {
+      "epoch": 79.38775510204081,
+      "grad_norm": 0.6306114792823792,
+      "learning_rate": 4.122448979591837e-06,
+      "loss": 0.2798,
+      "step": 15560
+    },
+    {
+      "epoch": 79.43877551020408,
+      "grad_norm": 24.38300323486328,
+      "learning_rate": 4.112244897959184e-06,
+      "loss": 0.2502,
+      "step": 15570
+    },
+    {
+      "epoch": 79.48979591836735,
+      "grad_norm": 8.707317352294922,
+      "learning_rate": 4.102040816326531e-06,
+      "loss": 0.2853,
+      "step": 15580
+    },
+    {
+      "epoch": 79.54081632653062,
+      "grad_norm": 12.659032821655273,
+      "learning_rate": 4.091836734693878e-06,
+      "loss": 0.1358,
+      "step": 15590
+    },
+    {
+      "epoch": 79.59183673469387,
+      "grad_norm": 13.423051834106445,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 0.121,
+      "step": 15600
+    },
+    {
+      "epoch": 79.64285714285714,
+      "grad_norm": 2.983762502670288,
+      "learning_rate": 4.071428571428572e-06,
+      "loss": 0.1123,
+      "step": 15610
+    },
+    {
+      "epoch": 79.6938775510204,
+      "grad_norm": 3.414855480194092,
+      "learning_rate": 4.061224489795918e-06,
+      "loss": 0.1874,
+      "step": 15620
+    },
+    {
+      "epoch": 79.74489795918367,
+      "grad_norm": 13.161895751953125,
+      "learning_rate": 4.051020408163266e-06,
+      "loss": 0.1783,
+      "step": 15630
+    },
+    {
+      "epoch": 79.79591836734694,
+      "grad_norm": 7.434272766113281,
+      "learning_rate": 4.040816326530612e-06,
+      "loss": 0.2543,
+      "step": 15640
+    },
+    {
+      "epoch": 79.84693877551021,
+      "grad_norm": 16.419912338256836,
+      "learning_rate": 4.03061224489796e-06,
+      "loss": 0.3525,
+      "step": 15650
+    },
+    {
+      "epoch": 79.89795918367346,
+      "grad_norm": 3.6615853309631348,
+      "learning_rate": 4.0204081632653065e-06,
+      "loss": 0.1637,
+      "step": 15660
+    },
+    {
+      "epoch": 79.94897959183673,
+      "grad_norm": 2.119905948638916,
+      "learning_rate": 4.010204081632653e-06,
+      "loss": 0.3019,
+      "step": 15670
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 0.8194112181663513,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.1969,
+      "step": 15680
+    },
+    {
+      "epoch": 80.0,
+      "eval_accuracy": 0.8483754512635379,
+      "eval_loss": 0.42701977491378784,
+      "eval_runtime": 0.9397,
+      "eval_samples_per_second": 294.767,
+      "eval_steps_per_second": 37.245,
+      "step": 15680
+    },
+    {
+      "epoch": 80.05102040816327,
+      "grad_norm": 9.35610580444336,
+      "learning_rate": 3.989795918367347e-06,
+      "loss": 0.19,
+      "step": 15690
+    },
+    {
+      "epoch": 80.10204081632654,
+      "grad_norm": 8.794118881225586,
+      "learning_rate": 3.979591836734694e-06,
+      "loss": 0.2407,
+      "step": 15700
+    },
+    {
+      "epoch": 80.15306122448979,
+      "grad_norm": 0.1584535837173462,
+      "learning_rate": 3.969387755102041e-06,
+      "loss": 0.0312,
+      "step": 15710
+    },
+    {
+      "epoch": 80.20408163265306,
+      "grad_norm": 4.157765865325928,
+      "learning_rate": 3.959183673469388e-06,
+      "loss": 0.3281,
+      "step": 15720
+    },
+    {
+      "epoch": 80.25510204081633,
+      "grad_norm": 11.907217979431152,
+      "learning_rate": 3.9489795918367356e-06,
+      "loss": 0.1984,
+      "step": 15730
+    },
+    {
+      "epoch": 80.3061224489796,
+      "grad_norm": 2.841599702835083,
+      "learning_rate": 3.938775510204082e-06,
+      "loss": 0.1251,
+      "step": 15740
+    },
+    {
+      "epoch": 80.35714285714286,
+      "grad_norm": 5.652760982513428,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.2669,
+      "step": 15750
+    },
+    {
+      "epoch": 80.40816326530613,
+      "grad_norm": 3.341134548187256,
+      "learning_rate": 3.9183673469387755e-06,
+      "loss": 0.1515,
+      "step": 15760
+    },
+    {
+      "epoch": 80.45918367346938,
+      "grad_norm": 0.34180694818496704,
+      "learning_rate": 3.908163265306123e-06,
+      "loss": 0.2429,
+      "step": 15770
+    },
+    {
+      "epoch": 80.51020408163265,
+      "grad_norm": 1.8589460849761963,
+      "learning_rate": 3.89795918367347e-06,
+      "loss": 0.1411,
+      "step": 15780
+    },
+    {
+      "epoch": 80.56122448979592,
+      "grad_norm": 1.2255891561508179,
+      "learning_rate": 3.887755102040816e-06,
+      "loss": 0.2386,
+      "step": 15790
+    },
+    {
+      "epoch": 80.61224489795919,
+      "grad_norm": 0.7362205386161804,
+      "learning_rate": 3.877551020408164e-06,
+      "loss": 0.0887,
+      "step": 15800
+    },
+    {
+      "epoch": 80.66326530612245,
+      "grad_norm": 11.652864456176758,
+      "learning_rate": 3.86734693877551e-06,
+      "loss": 0.2231,
+      "step": 15810
+    },
+    {
+      "epoch": 80.71428571428571,
+      "grad_norm": 1.9581044912338257,
+      "learning_rate": 3.857142857142858e-06,
+      "loss": 0.1942,
+      "step": 15820
+    },
+    {
+      "epoch": 80.76530612244898,
+      "grad_norm": 1.3246722221374512,
+      "learning_rate": 3.8469387755102045e-06,
+      "loss": 0.2956,
+      "step": 15830
+    },
+    {
+      "epoch": 80.81632653061224,
+      "grad_norm": 3.7175517082214355,
+      "learning_rate": 3.836734693877551e-06,
+      "loss": 0.1526,
+      "step": 15840
+    },
+    {
+      "epoch": 80.86734693877551,
+      "grad_norm": 1.4184467792510986,
+      "learning_rate": 3.826530612244898e-06,
+      "loss": 0.0829,
+      "step": 15850
+    },
+    {
+      "epoch": 80.91836734693878,
+      "grad_norm": 5.230016231536865,
+      "learning_rate": 3.816326530612245e-06,
+      "loss": 0.1705,
+      "step": 15860
+    },
+    {
+      "epoch": 80.96938775510205,
+      "grad_norm": 0.31383898854255676,
+      "learning_rate": 3.8061224489795923e-06,
+      "loss": 0.1405,
+      "step": 15870
+    },
+    {
+      "epoch": 81.0,
+      "eval_accuracy": 0.7833935018050542,
+      "eval_loss": 0.763225793838501,
+      "eval_runtime": 0.927,
+      "eval_samples_per_second": 298.824,
+      "eval_steps_per_second": 37.758,
+      "step": 15876
+    },
+    {
+      "epoch": 81.0204081632653,
+      "grad_norm": 5.370396614074707,
+      "learning_rate": 3.795918367346939e-06,
+      "loss": 0.1886,
+      "step": 15880
+    },
+    {
+      "epoch": 81.07142857142857,
+      "grad_norm": 0.8666300773620605,
+      "learning_rate": 3.785714285714286e-06,
+      "loss": 0.3796,
+      "step": 15890
+    },
+    {
+      "epoch": 81.12244897959184,
+      "grad_norm": 15.143610000610352,
+      "learning_rate": 3.7755102040816327e-06,
+      "loss": 0.2811,
+      "step": 15900
+    },
+    {
+      "epoch": 81.1734693877551,
+      "grad_norm": 0.3934500515460968,
+      "learning_rate": 3.76530612244898e-06,
+      "loss": 0.1487,
+      "step": 15910
+    },
+    {
+      "epoch": 81.22448979591837,
+      "grad_norm": 1.4975956678390503,
+      "learning_rate": 3.7551020408163268e-06,
+      "loss": 0.1721,
+      "step": 15920
+    },
+    {
+      "epoch": 81.27551020408163,
+      "grad_norm": 2.16916823387146,
+      "learning_rate": 3.744897959183674e-06,
+      "loss": 0.182,
+      "step": 15930
+    },
+    {
+      "epoch": 81.3265306122449,
+      "grad_norm": 8.05843448638916,
+      "learning_rate": 3.7346938775510205e-06,
+      "loss": 0.3428,
+      "step": 15940
+    },
+    {
+      "epoch": 81.37755102040816,
+      "grad_norm": 0.2246151566505432,
+      "learning_rate": 3.724489795918368e-06,
+      "loss": 0.1239,
+      "step": 15950
+    },
+    {
+      "epoch": 81.42857142857143,
+      "grad_norm": 8.20610237121582,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 0.1717,
+      "step": 15960
+    },
+    {
+      "epoch": 81.4795918367347,
+      "grad_norm": 8.254730224609375,
+      "learning_rate": 3.7040816326530617e-06,
+      "loss": 0.3008,
+      "step": 15970
+    },
+    {
+      "epoch": 81.53061224489795,
+      "grad_norm": 0.9908814430236816,
+      "learning_rate": 3.6938775510204083e-06,
+      "loss": 0.0522,
+      "step": 15980
+    },
+    {
+      "epoch": 81.58163265306122,
+      "grad_norm": 3.838620901107788,
+      "learning_rate": 3.6836734693877554e-06,
+      "loss": 0.2645,
+      "step": 15990
+    },
+    {
+      "epoch": 81.63265306122449,
+      "grad_norm": 0.31536588072776794,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 0.3646,
+      "step": 16000
+    },
+    {
+      "epoch": 81.68367346938776,
+      "grad_norm": 0.3082359731197357,
+      "learning_rate": 3.6632653061224495e-06,
+      "loss": 0.1775,
+      "step": 16010
+    },
+    {
+      "epoch": 81.73469387755102,
+      "grad_norm": 12.527641296386719,
+      "learning_rate": 3.653061224489796e-06,
+      "loss": 0.1322,
+      "step": 16020
+    },
+    {
+      "epoch": 81.78571428571429,
+      "grad_norm": 0.1477854698896408,
+      "learning_rate": 3.642857142857143e-06,
+      "loss": 0.112,
+      "step": 16030
+    },
+    {
+      "epoch": 81.83673469387755,
+      "grad_norm": 1.7172036170959473,
+      "learning_rate": 3.6326530612244903e-06,
+      "loss": 0.2226,
+      "step": 16040
+    },
+    {
+      "epoch": 81.88775510204081,
+      "grad_norm": 1.1714388132095337,
+      "learning_rate": 3.6224489795918373e-06,
+      "loss": 0.1911,
+      "step": 16050
+    },
+    {
+      "epoch": 81.93877551020408,
+      "grad_norm": 7.487484931945801,
+      "learning_rate": 3.612244897959184e-06,
+      "loss": 0.1288,
+      "step": 16060
+    },
+    {
+      "epoch": 81.98979591836735,
+      "grad_norm": 3.3677072525024414,
+      "learning_rate": 3.6020408163265306e-06,
+      "loss": 0.1427,
+      "step": 16070
+    },
+    {
+      "epoch": 82.0,
+      "eval_accuracy": 0.9025270758122743,
+      "eval_loss": 0.3249270021915436,
+      "eval_runtime": 0.9346,
+      "eval_samples_per_second": 296.382,
+      "eval_steps_per_second": 37.449,
+      "step": 16072
+    },
+    {
+      "epoch": 82.04081632653062,
+      "grad_norm": 2.0028810501098633,
+      "learning_rate": 3.5918367346938777e-06,
+      "loss": 0.1303,
+      "step": 16080
+    },
+    {
+      "epoch": 82.09183673469387,
+      "grad_norm": 1.0715469121932983,
+      "learning_rate": 3.581632653061225e-06,
+      "loss": 0.232,
+      "step": 16090
+    },
+    {
+      "epoch": 82.14285714285714,
+      "grad_norm": 6.861850261688232,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.1009,
+      "step": 16100
+    },
+    {
+      "epoch": 82.1938775510204,
+      "grad_norm": 0.32505741715431213,
+      "learning_rate": 3.5612244897959184e-06,
+      "loss": 0.1015,
+      "step": 16110
+    },
+    {
+      "epoch": 82.24489795918367,
+      "grad_norm": 11.258378028869629,
+      "learning_rate": 3.5510204081632655e-06,
+      "loss": 0.2765,
+      "step": 16120
+    },
+    {
+      "epoch": 82.29591836734694,
+      "grad_norm": 0.8208149671554565,
+      "learning_rate": 3.540816326530613e-06,
+      "loss": 0.1769,
+      "step": 16130
+    },
+    {
+      "epoch": 82.34693877551021,
+      "grad_norm": 14.091160774230957,
+      "learning_rate": 3.5306122448979596e-06,
+      "loss": 0.1296,
+      "step": 16140
+    },
+    {
+      "epoch": 82.39795918367346,
+      "grad_norm": 7.795243740081787,
+      "learning_rate": 3.5204081632653062e-06,
+      "loss": 0.4265,
+      "step": 16150
+    },
+    {
+      "epoch": 82.44897959183673,
+      "grad_norm": 19.263708114624023,
+      "learning_rate": 3.5102040816326533e-06,
+      "loss": 0.2897,
+      "step": 16160
+    },
+    {
+      "epoch": 82.5,
+      "grad_norm": 5.552840709686279,
+      "learning_rate": 3.5e-06,
+      "loss": 0.148,
+      "step": 16170
+    },
+    {
+      "epoch": 82.55102040816327,
+      "grad_norm": 16.170299530029297,
+      "learning_rate": 3.4897959183673474e-06,
+      "loss": 0.3484,
+      "step": 16180
+    },
+    {
+      "epoch": 82.60204081632654,
+      "grad_norm": 4.0577263832092285,
+      "learning_rate": 3.479591836734694e-06,
+      "loss": 0.2218,
+      "step": 16190
+    },
+    {
+      "epoch": 82.65306122448979,
+      "grad_norm": 5.736649990081787,
+      "learning_rate": 3.469387755102041e-06,
+      "loss": 0.1525,
+      "step": 16200
+    },
+    {
+      "epoch": 82.70408163265306,
+      "grad_norm": 21.887012481689453,
+      "learning_rate": 3.4591836734693878e-06,
+      "loss": 0.276,
+      "step": 16210
+    },
+    {
+      "epoch": 82.75510204081633,
+      "grad_norm": 18.0561580657959,
+      "learning_rate": 3.4489795918367353e-06,
+      "loss": 0.2404,
+      "step": 16220
+    },
+    {
+      "epoch": 82.8061224489796,
+      "grad_norm": 2.2050490379333496,
+      "learning_rate": 3.438775510204082e-06,
+      "loss": 0.1132,
+      "step": 16230
+    },
+    {
+      "epoch": 82.85714285714286,
+      "grad_norm": 1.9476908445358276,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 0.1607,
+      "step": 16240
+    },
+    {
+      "epoch": 82.90816326530613,
+      "grad_norm": 0.8884503841400146,
+      "learning_rate": 3.4183673469387756e-06,
+      "loss": 0.1731,
+      "step": 16250
+    },
+    {
+      "epoch": 82.95918367346938,
+      "grad_norm": 1.7185746431350708,
+      "learning_rate": 3.4081632653061227e-06,
+      "loss": 0.2493,
+      "step": 16260
+    },
+    {
+      "epoch": 83.0,
+      "eval_accuracy": 0.8989169675090253,
+      "eval_loss": 0.38384127616882324,
+      "eval_runtime": 0.9321,
+      "eval_samples_per_second": 297.167,
+      "eval_steps_per_second": 37.548,
+      "step": 16268
+    },
+    {
+      "epoch": 83.01020408163265,
+      "grad_norm": 7.763315200805664,
+      "learning_rate": 3.3979591836734697e-06,
+      "loss": 0.2217,
+      "step": 16270
+    },
+    {
+      "epoch": 83.06122448979592,
+      "grad_norm": 1.9195549488067627,
+      "learning_rate": 3.3877551020408168e-06,
+      "loss": 0.137,
+      "step": 16280
+    },
+    {
+      "epoch": 83.11224489795919,
+      "grad_norm": 1.0584914684295654,
+      "learning_rate": 3.3775510204081634e-06,
+      "loss": 0.3192,
+      "step": 16290
+    },
+    {
+      "epoch": 83.16326530612245,
+      "grad_norm": 0.7407238483428955,
+      "learning_rate": 3.3673469387755105e-06,
+      "loss": 0.2872,
+      "step": 16300
+    },
+    {
+      "epoch": 83.21428571428571,
+      "grad_norm": 23.612741470336914,
+      "learning_rate": 3.357142857142857e-06,
+      "loss": 0.2775,
+      "step": 16310
+    },
+    {
+      "epoch": 83.26530612244898,
+      "grad_norm": 0.3100240230560303,
+      "learning_rate": 3.3469387755102046e-06,
+      "loss": 0.1608,
+      "step": 16320
+    },
+    {
+      "epoch": 83.31632653061224,
+      "grad_norm": 5.569301128387451,
+      "learning_rate": 3.3367346938775513e-06,
+      "loss": 0.2444,
+      "step": 16330
+    },
+    {
+      "epoch": 83.36734693877551,
+      "grad_norm": 14.591517448425293,
+      "learning_rate": 3.3265306122448983e-06,
+      "loss": 0.29,
+      "step": 16340
+    },
+    {
+      "epoch": 83.41836734693878,
+      "grad_norm": 7.931599140167236,
+      "learning_rate": 3.316326530612245e-06,
+      "loss": 0.277,
+      "step": 16350
+    },
+    {
+      "epoch": 83.46938775510205,
+      "grad_norm": 15.914363861083984,
+      "learning_rate": 3.3061224489795924e-06,
+      "loss": 0.1571,
+      "step": 16360
+    },
+    {
+      "epoch": 83.5204081632653,
+      "grad_norm": 13.947531700134277,
+      "learning_rate": 3.295918367346939e-06,
+      "loss": 0.3652,
+      "step": 16370
+    },
+    {
+      "epoch": 83.57142857142857,
+      "grad_norm": 3.993175506591797,
+      "learning_rate": 3.285714285714286e-06,
+      "loss": 0.4236,
+      "step": 16380
+    },
+    {
+      "epoch": 83.62244897959184,
+      "grad_norm": 9.41831111907959,
+      "learning_rate": 3.2755102040816328e-06,
+      "loss": 0.4799,
+      "step": 16390
+    },
+    {
+      "epoch": 83.6734693877551,
+      "grad_norm": 4.541845321655273,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 0.2001,
+      "step": 16400
+    },
+    {
+      "epoch": 83.72448979591837,
+      "grad_norm": 17.684083938598633,
+      "learning_rate": 3.255102040816327e-06,
+      "loss": 0.1741,
+      "step": 16410
+    },
+    {
+      "epoch": 83.77551020408163,
+      "grad_norm": 0.664922297000885,
+      "learning_rate": 3.244897959183674e-06,
+      "loss": 0.1669,
+      "step": 16420
+    },
+    {
+      "epoch": 83.8265306122449,
+      "grad_norm": 4.690441608428955,
+      "learning_rate": 3.2346938775510206e-06,
+      "loss": 0.2087,
+      "step": 16430
+    },
+    {
+      "epoch": 83.87755102040816,
+      "grad_norm": 5.026739120483398,
+      "learning_rate": 3.2244897959183672e-06,
+      "loss": 0.1029,
+      "step": 16440
+    },
+    {
+      "epoch": 83.92857142857143,
+      "grad_norm": 8.364355087280273,
+      "learning_rate": 3.2142857142857147e-06,
+      "loss": 0.1695,
+      "step": 16450
+    },
+    {
+      "epoch": 83.9795918367347,
+      "grad_norm": 1.0005097389221191,
+      "learning_rate": 3.204081632653062e-06,
+      "loss": 0.331,
+      "step": 16460
+    },
+    {
+      "epoch": 84.0,
+      "eval_accuracy": 0.9205776173285198,
+      "eval_loss": 0.3329784572124481,
+      "eval_runtime": 0.9402,
+      "eval_samples_per_second": 294.616,
+      "eval_steps_per_second": 37.226,
+      "step": 16464
+    },
+    {
+      "epoch": 84.03061224489795,
+      "grad_norm": 9.732176780700684,
+      "learning_rate": 3.1938775510204084e-06,
+      "loss": 0.1133,
+      "step": 16470
+    },
+    {
+      "epoch": 84.08163265306122,
+      "grad_norm": 1.5133066177368164,
+      "learning_rate": 3.183673469387755e-06,
+      "loss": 0.1528,
+      "step": 16480
+    },
+    {
+      "epoch": 84.13265306122449,
+      "grad_norm": 6.870362758636475,
+      "learning_rate": 3.173469387755102e-06,
+      "loss": 0.2269,
+      "step": 16490
+    },
+    {
+      "epoch": 84.18367346938776,
+      "grad_norm": 16.802181243896484,
+      "learning_rate": 3.1632653061224496e-06,
+      "loss": 0.3749,
+      "step": 16500
+    },
+    {
+      "epoch": 84.23469387755102,
+      "grad_norm": 8.427709579467773,
+      "learning_rate": 3.1530612244897963e-06,
+      "loss": 0.4603,
+      "step": 16510
+    },
+    {
+      "epoch": 84.28571428571429,
+      "grad_norm": 5.263187885284424,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 0.1094,
+      "step": 16520
+    },
+    {
+      "epoch": 84.33673469387755,
+      "grad_norm": 12.858963012695312,
+      "learning_rate": 3.13265306122449e-06,
+      "loss": 0.2149,
+      "step": 16530
+    },
+    {
+      "epoch": 84.38775510204081,
+      "grad_norm": 4.1460490226745605,
+      "learning_rate": 3.1224489795918374e-06,
+      "loss": 0.1401,
+      "step": 16540
+    },
+    {
+      "epoch": 84.43877551020408,
+      "grad_norm": 16.93732261657715,
+      "learning_rate": 3.112244897959184e-06,
+      "loss": 0.4208,
+      "step": 16550
+    },
+    {
+      "epoch": 84.48979591836735,
+      "grad_norm": 0.18154656887054443,
+      "learning_rate": 3.1020408163265307e-06,
+      "loss": 0.2127,
+      "step": 16560
+    },
+    {
+      "epoch": 84.54081632653062,
+      "grad_norm": 1.393512487411499,
+      "learning_rate": 3.0918367346938778e-06,
+      "loss": 0.1961,
+      "step": 16570
+    },
+    {
+      "epoch": 84.59183673469387,
+      "grad_norm": 3.731569290161133,
+      "learning_rate": 3.0816326530612244e-06,
+      "loss": 0.3392,
+      "step": 16580
+    },
+    {
+      "epoch": 84.64285714285714,
+      "grad_norm": 3.3272573947906494,
+      "learning_rate": 3.071428571428572e-06,
+      "loss": 0.1673,
+      "step": 16590
+    },
+    {
+      "epoch": 84.6938775510204,
+      "grad_norm": 12.289058685302734,
+      "learning_rate": 3.0612244897959185e-06,
+      "loss": 0.1943,
+      "step": 16600
+    },
+    {
+      "epoch": 84.74489795918367,
+      "grad_norm": 1.6420972347259521,
+      "learning_rate": 3.0510204081632656e-06,
+      "loss": 0.128,
+      "step": 16610
+    },
+    {
+      "epoch": 84.79591836734694,
+      "grad_norm": 2.594130754470825,
+      "learning_rate": 3.0408163265306122e-06,
+      "loss": 0.1445,
+      "step": 16620
+    },
+    {
+      "epoch": 84.84693877551021,
+      "grad_norm": 0.44964268803596497,
+      "learning_rate": 3.0306122448979597e-06,
+      "loss": 0.3691,
+      "step": 16630
+    },
+    {
+      "epoch": 84.89795918367346,
+      "grad_norm": 0.9908690452575684,
+      "learning_rate": 3.0204081632653064e-06,
+      "loss": 0.2374,
+      "step": 16640
+    },
+    {
+      "epoch": 84.94897959183673,
+      "grad_norm": 0.32848837971687317,
+      "learning_rate": 3.0102040816326534e-06,
+      "loss": 0.1338,
+      "step": 16650
+    },
+    {
+      "epoch": 85.0,
+      "grad_norm": 1.4169095754623413,
+      "learning_rate": 3e-06,
+      "loss": 0.1231,
+      "step": 16660
+    },
+    {
+      "epoch": 85.0,
+      "eval_accuracy": 0.8700361010830325,
+      "eval_loss": 0.32456666231155396,
+      "eval_runtime": 0.9552,
+      "eval_samples_per_second": 289.984,
+      "eval_steps_per_second": 36.641,
+      "step": 16660
+    },
+    {
+      "epoch": 85.05102040816327,
+      "grad_norm": 9.53570556640625,
+      "learning_rate": 2.989795918367347e-06,
+      "loss": 0.1539,
+      "step": 16670
+    },
+    {
+      "epoch": 85.10204081632654,
+      "grad_norm": 6.8302106857299805,
+      "learning_rate": 2.979591836734694e-06,
+      "loss": 0.1866,
+      "step": 16680
+    },
+    {
+      "epoch": 85.15306122448979,
+      "grad_norm": 0.5330386161804199,
+      "learning_rate": 2.9693877551020413e-06,
+      "loss": 0.1126,
+      "step": 16690
+    },
+    {
+      "epoch": 85.20408163265306,
+      "grad_norm": 0.9756630063056946,
+      "learning_rate": 2.959183673469388e-06,
+      "loss": 0.2992,
+      "step": 16700
+    },
+    {
+      "epoch": 85.25510204081633,
+      "grad_norm": 7.213351726531982,
+      "learning_rate": 2.948979591836735e-06,
+      "loss": 0.2454,
+      "step": 16710
+    },
+    {
+      "epoch": 85.3061224489796,
+      "grad_norm": 6.819469928741455,
+      "learning_rate": 2.938775510204082e-06,
+      "loss": 0.1805,
+      "step": 16720
+    },
+    {
+      "epoch": 85.35714285714286,
+      "grad_norm": 6.094568252563477,
+      "learning_rate": 2.928571428571429e-06,
+      "loss": 0.248,
+      "step": 16730
+    },
+    {
+      "epoch": 85.40816326530613,
+      "grad_norm": 1.5939466953277588,
+      "learning_rate": 2.9183673469387757e-06,
+      "loss": 0.2055,
+      "step": 16740
+    },
+    {
+      "epoch": 85.45918367346938,
+      "grad_norm": 1.6548603773117065,
+      "learning_rate": 2.908163265306123e-06,
+      "loss": 0.1873,
+      "step": 16750
+    },
+    {
+      "epoch": 85.51020408163265,
+      "grad_norm": 2.3544037342071533,
+      "learning_rate": 2.8979591836734694e-06,
+      "loss": 0.084,
+      "step": 16760
+    },
+    {
+      "epoch": 85.56122448979592,
+      "grad_norm": 11.432575225830078,
+      "learning_rate": 2.887755102040817e-06,
+      "loss": 0.3432,
+      "step": 16770
+    },
+    {
+      "epoch": 85.61224489795919,
+      "grad_norm": 3.1924562454223633,
+      "learning_rate": 2.8775510204081636e-06,
+      "loss": 0.4288,
+      "step": 16780
+    },
+    {
+      "epoch": 85.66326530612245,
+      "grad_norm": 0.5942087769508362,
+      "learning_rate": 2.86734693877551e-06,
+      "loss": 0.2546,
+      "step": 16790
+    },
+    {
+      "epoch": 85.71428571428571,
+      "grad_norm": 0.3294694423675537,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.2859,
+      "step": 16800
+    },
+    {
+      "epoch": 85.76530612244898,
+      "grad_norm": 1.8321048021316528,
+      "learning_rate": 2.8469387755102047e-06,
+      "loss": 0.3555,
+      "step": 16810
+    },
+    {
+      "epoch": 85.81632653061224,
+      "grad_norm": 10.187983512878418,
+      "learning_rate": 2.8367346938775514e-06,
+      "loss": 0.2943,
+      "step": 16820
+    },
+    {
+      "epoch": 85.86734693877551,
+      "grad_norm": 10.006871223449707,
+      "learning_rate": 2.826530612244898e-06,
+      "loss": 0.0889,
+      "step": 16830
+    },
+    {
+      "epoch": 85.91836734693878,
+      "grad_norm": 0.727992594242096,
+      "learning_rate": 2.816326530612245e-06,
+      "loss": 0.1329,
+      "step": 16840
+    },
+    {
+      "epoch": 85.96938775510205,
+      "grad_norm": 15.858503341674805,
+      "learning_rate": 2.8061224489795917e-06,
+      "loss": 0.2781,
+      "step": 16850
+    },
+    {
+      "epoch": 86.0,
+      "eval_accuracy": 0.8736462093862816,
+      "eval_loss": 0.3710058927536011,
+      "eval_runtime": 0.9355,
+      "eval_samples_per_second": 296.084,
+      "eval_steps_per_second": 37.411,
+      "step": 16856
+    },
+    {
+      "epoch": 86.0204081632653,
+      "grad_norm": 8.392712593078613,
+      "learning_rate": 2.795918367346939e-06,
+      "loss": 0.1672,
+      "step": 16860
+    },
+    {
+      "epoch": 86.07142857142857,
+      "grad_norm": 13.223836898803711,
+      "learning_rate": 2.785714285714286e-06,
+      "loss": 0.2495,
+      "step": 16870
+    },
+    {
+      "epoch": 86.12244897959184,
+      "grad_norm": 6.981386661529541,
+      "learning_rate": 2.775510204081633e-06,
+      "loss": 0.1335,
+      "step": 16880
+    },
+    {
+      "epoch": 86.1734693877551,
+      "grad_norm": 16.30123519897461,
+      "learning_rate": 2.7653061224489795e-06,
+      "loss": 0.2004,
+      "step": 16890
+    },
+    {
+      "epoch": 86.22448979591837,
+      "grad_norm": 14.89189624786377,
+      "learning_rate": 2.7551020408163266e-06,
+      "loss": 0.201,
+      "step": 16900
+    },
+    {
+      "epoch": 86.27551020408163,
+      "grad_norm": 0.510366678237915,
+      "learning_rate": 2.7448979591836737e-06,
+      "loss": 0.1122,
+      "step": 16910
+    },
+    {
+      "epoch": 86.3265306122449,
+      "grad_norm": 8.832609176635742,
+      "learning_rate": 2.7346938775510207e-06,
+      "loss": 0.2677,
+      "step": 16920
+    },
+    {
+      "epoch": 86.37755102040816,
+      "grad_norm": 0.24932265281677246,
+      "learning_rate": 2.7244897959183674e-06,
+      "loss": 0.2158,
+      "step": 16930
+    },
+    {
+      "epoch": 86.42857142857143,
+      "grad_norm": 6.414112091064453,
+      "learning_rate": 2.7142857142857144e-06,
+      "loss": 0.1902,
+      "step": 16940
+    },
+    {
+      "epoch": 86.4795918367347,
+      "grad_norm": 0.8077690601348877,
+      "learning_rate": 2.7040816326530615e-06,
+      "loss": 0.1207,
+      "step": 16950
+    },
+    {
+      "epoch": 86.53061224489795,
+      "grad_norm": 1.0012266635894775,
+      "learning_rate": 2.6938775510204086e-06,
+      "loss": 0.29,
+      "step": 16960
+    },
+    {
+      "epoch": 86.58163265306122,
+      "grad_norm": 1.7598363161087036,
+      "learning_rate": 2.683673469387755e-06,
+      "loss": 0.4445,
+      "step": 16970
+    },
+    {
+      "epoch": 86.63265306122449,
+      "grad_norm": 11.231893539428711,
+      "learning_rate": 2.6734693877551023e-06,
+      "loss": 0.1348,
+      "step": 16980
+    },
+    {
+      "epoch": 86.68367346938776,
+      "grad_norm": 11.320291519165039,
+      "learning_rate": 2.663265306122449e-06,
+      "loss": 0.1525,
+      "step": 16990
+    },
+    {
+      "epoch": 86.73469387755102,
+      "grad_norm": 0.5130727291107178,
+      "learning_rate": 2.6530612244897964e-06,
+      "loss": 0.2855,
+      "step": 17000
+    },
+    {
+      "epoch": 86.78571428571429,
+      "grad_norm": 8.888845443725586,
+      "learning_rate": 2.642857142857143e-06,
+      "loss": 0.1978,
+      "step": 17010
+    },
+    {
+      "epoch": 86.83673469387755,
+      "grad_norm": 5.790079593658447,
+      "learning_rate": 2.63265306122449e-06,
+      "loss": 0.0955,
+      "step": 17020
+    },
+    {
+      "epoch": 86.88775510204081,
+      "grad_norm": 12.437826156616211,
+      "learning_rate": 2.6224489795918367e-06,
+      "loss": 0.1971,
+      "step": 17030
+    },
+    {
+      "epoch": 86.93877551020408,
+      "grad_norm": 2.4969210624694824,
+      "learning_rate": 2.6122448979591842e-06,
+      "loss": 0.2551,
+      "step": 17040
+    },
+    {
+      "epoch": 86.98979591836735,
+      "grad_norm": 22.423816680908203,
+      "learning_rate": 2.602040816326531e-06,
+      "loss": 0.7193,
+      "step": 17050
+    },
+    {
+      "epoch": 87.0,
+      "eval_accuracy": 0.9061371841155235,
+      "eval_loss": 0.33839184045791626,
+      "eval_runtime": 0.957,
+      "eval_samples_per_second": 289.444,
+      "eval_steps_per_second": 36.572,
+      "step": 17052
+    },
+    {
+      "epoch": 87.04081632653062,
+      "grad_norm": 4.3819355964660645,
+      "learning_rate": 2.591836734693878e-06,
+      "loss": 0.2473,
+      "step": 17060
+    },
+    {
+      "epoch": 87.09183673469387,
+      "grad_norm": 0.5776048302650452,
+      "learning_rate": 2.5816326530612246e-06,
+      "loss": 0.2227,
+      "step": 17070
+    },
+    {
+      "epoch": 87.14285714285714,
+      "grad_norm": 16.769515991210938,
+      "learning_rate": 2.571428571428571e-06,
+      "loss": 0.248,
+      "step": 17080
+    },
+    {
+      "epoch": 87.1938775510204,
+      "grad_norm": 9.135167121887207,
+      "learning_rate": 2.5612244897959187e-06,
+      "loss": 0.2463,
+      "step": 17090
+    },
+    {
+      "epoch": 87.24489795918367,
+      "grad_norm": 11.403244972229004,
+      "learning_rate": 2.5510204081632657e-06,
+      "loss": 0.4687,
+      "step": 17100
+    },
+    {
+      "epoch": 87.29591836734694,
+      "grad_norm": 14.178475379943848,
+      "learning_rate": 2.5408163265306124e-06,
+      "loss": 0.3548,
+      "step": 17110
+    },
+    {
+      "epoch": 87.34693877551021,
+      "grad_norm": 0.38246801495552063,
+      "learning_rate": 2.530612244897959e-06,
+      "loss": 0.0965,
+      "step": 17120
+    },
+    {
+      "epoch": 87.39795918367346,
+      "grad_norm": 0.6379547119140625,
+      "learning_rate": 2.5204081632653065e-06,
+      "loss": 0.0286,
+      "step": 17130
+    },
+    {
+      "epoch": 87.44897959183673,
+      "grad_norm": 1.0402100086212158,
+      "learning_rate": 2.5102040816326536e-06,
+      "loss": 0.134,
+      "step": 17140
+    },
+    {
+      "epoch": 87.5,
+      "grad_norm": 0.46137455105781555,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1412,
+      "step": 17150
+    },
+    {
+      "epoch": 87.55102040816327,
+      "grad_norm": 10.064380645751953,
+      "learning_rate": 2.489795918367347e-06,
+      "loss": 0.264,
+      "step": 17160
+    },
+    {
+      "epoch": 87.60204081632654,
+      "grad_norm": 13.384167671203613,
+      "learning_rate": 2.479591836734694e-06,
+      "loss": 0.2698,
+      "step": 17170
+    },
+    {
+      "epoch": 87.65306122448979,
+      "grad_norm": 10.690115928649902,
+      "learning_rate": 2.469387755102041e-06,
+      "loss": 0.2503,
+      "step": 17180
+    },
+    {
+      "epoch": 87.70408163265306,
+      "grad_norm": 0.05725846067070961,
+      "learning_rate": 2.459183673469388e-06,
+      "loss": 0.2594,
+      "step": 17190
+    },
+    {
+      "epoch": 87.75510204081633,
+      "grad_norm": 0.4028679132461548,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 0.059,
+      "step": 17200
+    },
+    {
+      "epoch": 87.8061224489796,
+      "grad_norm": 13.775506019592285,
+      "learning_rate": 2.4387755102040817e-06,
+      "loss": 0.3321,
+      "step": 17210
+    },
+    {
+      "epoch": 87.85714285714286,
+      "grad_norm": 1.2155430316925049,
+      "learning_rate": 2.428571428571429e-06,
+      "loss": 0.2333,
+      "step": 17220
+    },
+    {
+      "epoch": 87.90816326530613,
+      "grad_norm": 15.729743003845215,
+      "learning_rate": 2.418367346938776e-06,
+      "loss": 0.1389,
+      "step": 17230
+    },
+    {
+      "epoch": 87.95918367346938,
+      "grad_norm": 0.565786600112915,
+      "learning_rate": 2.4081632653061225e-06,
+      "loss": 0.1149,
+      "step": 17240
+    },
+    {
+      "epoch": 88.0,
+      "eval_accuracy": 0.9097472924187726,
+      "eval_loss": 0.3703186511993408,
+      "eval_runtime": 1.0082,
+      "eval_samples_per_second": 274.745,
+      "eval_steps_per_second": 34.715,
+      "step": 17248
+    },
+    {
+      "epoch": 88.01020408163265,
+      "grad_norm": 0.8142939209938049,
+      "learning_rate": 2.3979591836734696e-06,
+      "loss": 0.148,
+      "step": 17250
+    },
+    {
+      "epoch": 88.06122448979592,
+      "grad_norm": 3.3334763050079346,
+      "learning_rate": 2.3877551020408166e-06,
+      "loss": 0.1569,
+      "step": 17260
+    },
+    {
+      "epoch": 88.11224489795919,
+      "grad_norm": 20.91797637939453,
+      "learning_rate": 2.3775510204081633e-06,
+      "loss": 0.1814,
+      "step": 17270
+    },
+    {
+      "epoch": 88.16326530612245,
+      "grad_norm": 15.569923400878906,
+      "learning_rate": 2.3673469387755103e-06,
+      "loss": 0.2453,
+      "step": 17280
+    },
+    {
+      "epoch": 88.21428571428571,
+      "grad_norm": 5.313724994659424,
+      "learning_rate": 2.3571428571428574e-06,
+      "loss": 0.0974,
+      "step": 17290
+    },
+    {
+      "epoch": 88.26530612244898,
+      "grad_norm": 6.935853958129883,
+      "learning_rate": 2.3469387755102044e-06,
+      "loss": 0.1783,
+      "step": 17300
+    },
+    {
+      "epoch": 88.31632653061224,
+      "grad_norm": 8.016472816467285,
+      "learning_rate": 2.336734693877551e-06,
+      "loss": 0.2279,
+      "step": 17310
+    },
+    {
+      "epoch": 88.36734693877551,
+      "grad_norm": 0.44269347190856934,
+      "learning_rate": 2.326530612244898e-06,
+      "loss": 0.2314,
+      "step": 17320
+    },
+    {
+      "epoch": 88.41836734693878,
+      "grad_norm": 4.441252708435059,
+      "learning_rate": 2.316326530612245e-06,
+      "loss": 0.1242,
+      "step": 17330
+    },
+    {
+      "epoch": 88.46938775510205,
+      "grad_norm": 1.3162168264389038,
+      "learning_rate": 2.306122448979592e-06,
+      "loss": 0.731,
+      "step": 17340
+    },
+    {
+      "epoch": 88.5204081632653,
+      "grad_norm": 3.208286762237549,
+      "learning_rate": 2.295918367346939e-06,
+      "loss": 0.357,
+      "step": 17350
+    },
+    {
+      "epoch": 88.57142857142857,
+      "grad_norm": 13.052005767822266,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 0.0895,
+      "step": 17360
+    },
+    {
+      "epoch": 88.62244897959184,
+      "grad_norm": 13.767204284667969,
+      "learning_rate": 2.275510204081633e-06,
+      "loss": 0.2135,
+      "step": 17370
+    },
+    {
+      "epoch": 88.6734693877551,
+      "grad_norm": 0.7811937928199768,
+      "learning_rate": 2.2653061224489797e-06,
+      "loss": 0.3113,
+      "step": 17380
+    },
+    {
+      "epoch": 88.72448979591837,
+      "grad_norm": 10.270852088928223,
+      "learning_rate": 2.2551020408163267e-06,
+      "loss": 0.2986,
+      "step": 17390
+    },
+    {
+      "epoch": 88.77551020408163,
+      "grad_norm": 0.29720714688301086,
+      "learning_rate": 2.244897959183674e-06,
+      "loss": 0.1637,
+      "step": 17400
+    },
+    {
+      "epoch": 88.8265306122449,
+      "grad_norm": 5.928308486938477,
+      "learning_rate": 2.2346938775510204e-06,
+      "loss": 0.1065,
+      "step": 17410
+    },
+    {
+      "epoch": 88.87755102040816,
+      "grad_norm": 5.635687351226807,
+      "learning_rate": 2.2244897959183675e-06,
+      "loss": 0.3591,
+      "step": 17420
+    },
+    {
+      "epoch": 88.92857142857143,
+      "grad_norm": 0.88820880651474,
+      "learning_rate": 2.2142857142857146e-06,
+      "loss": 0.1143,
+      "step": 17430
+    },
+    {
+      "epoch": 88.9795918367347,
+      "grad_norm": 5.052770614624023,
+      "learning_rate": 2.2040816326530616e-06,
+      "loss": 0.0269,
+      "step": 17440
+    },
+    {
+      "epoch": 89.0,
+      "eval_accuracy": 0.8592057761732852,
+      "eval_loss": 0.5013492703437805,
+      "eval_runtime": 0.9338,
+      "eval_samples_per_second": 296.635,
+      "eval_steps_per_second": 37.481,
+      "step": 17444
+    },
+    {
+      "epoch": 89.03061224489795,
+      "grad_norm": 15.623953819274902,
+      "learning_rate": 2.1938775510204083e-06,
+      "loss": 0.2499,
+      "step": 17450
+    },
+    {
+      "epoch": 89.08163265306122,
+      "grad_norm": 3.665088653564453,
+      "learning_rate": 2.1836734693877553e-06,
+      "loss": 0.1199,
+      "step": 17460
+    },
+    {
+      "epoch": 89.13265306122449,
+      "grad_norm": 1.6090409755706787,
+      "learning_rate": 2.1734693877551024e-06,
+      "loss": 0.3602,
+      "step": 17470
+    },
+    {
+      "epoch": 89.18367346938776,
+      "grad_norm": 16.96292495727539,
+      "learning_rate": 2.1632653061224495e-06,
+      "loss": 0.2923,
+      "step": 17480
+    },
+    {
+      "epoch": 89.23469387755102,
+      "grad_norm": 1.3773144483566284,
+      "learning_rate": 2.153061224489796e-06,
+      "loss": 0.2201,
+      "step": 17490
+    },
+    {
+      "epoch": 89.28571428571429,
+      "grad_norm": 1.2361564636230469,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 0.301,
+      "step": 17500
+    },
+    {
+      "epoch": 89.33673469387755,
+      "grad_norm": 0.3144189119338989,
+      "learning_rate": 2.1326530612244902e-06,
+      "loss": 0.1185,
+      "step": 17510
+    },
+    {
+      "epoch": 89.38775510204081,
+      "grad_norm": 2.084606170654297,
+      "learning_rate": 2.122448979591837e-06,
+      "loss": 0.2111,
+      "step": 17520
+    },
+    {
+      "epoch": 89.43877551020408,
+      "grad_norm": 0.8495397567749023,
+      "learning_rate": 2.112244897959184e-06,
+      "loss": 0.2766,
+      "step": 17530
+    },
+    {
+      "epoch": 89.48979591836735,
+      "grad_norm": 1.677476167678833,
+      "learning_rate": 2.1020408163265306e-06,
+      "loss": 0.17,
+      "step": 17540
+    },
+    {
+      "epoch": 89.54081632653062,
+      "grad_norm": 23.526060104370117,
+      "learning_rate": 2.0918367346938776e-06,
+      "loss": 0.272,
+      "step": 17550
+    },
+    {
+      "epoch": 89.59183673469387,
+      "grad_norm": 1.2515920400619507,
+      "learning_rate": 2.0816326530612247e-06,
+      "loss": 0.1553,
+      "step": 17560
+    },
+    {
+      "epoch": 89.64285714285714,
+      "grad_norm": 16.200321197509766,
+      "learning_rate": 2.0714285714285717e-06,
+      "loss": 0.1367,
+      "step": 17570
+    },
+    {
+      "epoch": 89.6938775510204,
+      "grad_norm": 1.1732423305511475,
+      "learning_rate": 2.0612244897959184e-06,
+      "loss": 0.1491,
+      "step": 17580
+    },
+    {
+      "epoch": 89.74489795918367,
+      "grad_norm": 0.4454115331172943,
+      "learning_rate": 2.0510204081632654e-06,
+      "loss": 0.1806,
+      "step": 17590
+    },
+    {
+      "epoch": 89.79591836734694,
+      "grad_norm": 3.885526657104492,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 0.0694,
+      "step": 17600
+    },
+    {
+      "epoch": 89.84693877551021,
+      "grad_norm": 6.656656265258789,
+      "learning_rate": 2.030612244897959e-06,
+      "loss": 0.2663,
+      "step": 17610
+    },
+    {
+      "epoch": 89.89795918367346,
+      "grad_norm": 7.046533584594727,
+      "learning_rate": 2.020408163265306e-06,
+      "loss": 0.3706,
+      "step": 17620
+    },
+    {
+      "epoch": 89.94897959183673,
+      "grad_norm": 1.4293006658554077,
+      "learning_rate": 2.0102040816326533e-06,
+      "loss": 0.1359,
+      "step": 17630
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 10.948988914489746,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.0967,
+      "step": 17640
+    },
+    {
+      "epoch": 90.0,
+      "eval_accuracy": 0.8989169675090253,
+      "eval_loss": 0.3456423580646515,
+      "eval_runtime": 0.9381,
+      "eval_samples_per_second": 295.278,
+      "eval_steps_per_second": 37.309,
+      "step": 17640
+    },
+    {
+      "epoch": 90.05102040816327,
+      "grad_norm": 0.12575684487819672,
+      "learning_rate": 1.989795918367347e-06,
+      "loss": 0.1386,
+      "step": 17650
+    },
+    {
+      "epoch": 90.10204081632654,
+      "grad_norm": 1.654032588005066,
+      "learning_rate": 1.979591836734694e-06,
+      "loss": 0.3257,
+      "step": 17660
+    },
+    {
+      "epoch": 90.15306122448979,
+      "grad_norm": 1.0453482866287231,
+      "learning_rate": 1.969387755102041e-06,
+      "loss": 0.2971,
+      "step": 17670
+    },
+    {
+      "epoch": 90.20408163265306,
+      "grad_norm": 4.922945022583008,
+      "learning_rate": 1.9591836734693877e-06,
+      "loss": 0.302,
+      "step": 17680
+    },
+    {
+      "epoch": 90.25510204081633,
+      "grad_norm": 1.6829240322113037,
+      "learning_rate": 1.948979591836735e-06,
+      "loss": 0.0836,
+      "step": 17690
+    },
+    {
+      "epoch": 90.3061224489796,
+      "grad_norm": 13.822141647338867,
+      "learning_rate": 1.938775510204082e-06,
+      "loss": 0.1002,
+      "step": 17700
+    },
+    {
+      "epoch": 90.35714285714286,
+      "grad_norm": 15.23730754852295,
+      "learning_rate": 1.928571428571429e-06,
+      "loss": 0.3489,
+      "step": 17710
+    },
+    {
+      "epoch": 90.40816326530613,
+      "grad_norm": 1.2752635478973389,
+      "learning_rate": 1.9183673469387756e-06,
+      "loss": 0.0777,
+      "step": 17720
+    },
+    {
+      "epoch": 90.45918367346938,
+      "grad_norm": 1.7229952812194824,
+      "learning_rate": 1.9081632653061226e-06,
+      "loss": 0.0637,
+      "step": 17730
+    },
+    {
+      "epoch": 90.51020408163265,
+      "grad_norm": 4.0512542724609375,
+      "learning_rate": 1.8979591836734695e-06,
+      "loss": 0.3496,
+      "step": 17740
+    },
+    {
+      "epoch": 90.56122448979592,
+      "grad_norm": 1.8018417358398438,
+      "learning_rate": 1.8877551020408163e-06,
+      "loss": 0.1146,
+      "step": 17750
+    },
+    {
+      "epoch": 90.61224489795919,
+      "grad_norm": 1.9088629484176636,
+      "learning_rate": 1.8775510204081634e-06,
+      "loss": 0.1174,
+      "step": 17760
+    },
+    {
+      "epoch": 90.66326530612245,
+      "grad_norm": 0.3909110426902771,
+      "learning_rate": 1.8673469387755102e-06,
+      "loss": 0.3986,
+      "step": 17770
+    },
+    {
+      "epoch": 90.71428571428571,
+      "grad_norm": 16.561473846435547,
+      "learning_rate": 1.8571428571428573e-06,
+      "loss": 0.2499,
+      "step": 17780
+    },
+    {
+      "epoch": 90.76530612244898,
+      "grad_norm": 0.9881889224052429,
+      "learning_rate": 1.8469387755102042e-06,
+      "loss": 0.2025,
+      "step": 17790
+    },
+    {
+      "epoch": 90.81632653061224,
+      "grad_norm": 1.2124401330947876,
+      "learning_rate": 1.8367346938775512e-06,
+      "loss": 0.1655,
+      "step": 17800
+    },
+    {
+      "epoch": 90.86734693877551,
+      "grad_norm": 0.8858227729797363,
+      "learning_rate": 1.826530612244898e-06,
+      "loss": 0.1717,
+      "step": 17810
+    },
+    {
+      "epoch": 90.91836734693878,
+      "grad_norm": 8.273017883300781,
+      "learning_rate": 1.8163265306122451e-06,
+      "loss": 0.1776,
+      "step": 17820
+    },
+    {
+      "epoch": 90.96938775510205,
+      "grad_norm": 16.358352661132812,
+      "learning_rate": 1.806122448979592e-06,
+      "loss": 0.177,
+      "step": 17830
+    },
+    {
+      "epoch": 91.0,
+      "eval_accuracy": 0.8880866425992779,
+      "eval_loss": 0.3798539340496063,
+      "eval_runtime": 0.935,
+      "eval_samples_per_second": 296.266,
+      "eval_steps_per_second": 37.434,
+      "step": 17836
+    },
+    {
+      "epoch": 91.0204081632653,
+      "grad_norm": 11.906256675720215,
+      "learning_rate": 1.7959183673469388e-06,
+      "loss": 0.1371,
+      "step": 17840
+    },
+    {
+      "epoch": 91.07142857142857,
+      "grad_norm": 12.646199226379395,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 0.133,
+      "step": 17850
+    },
+    {
+      "epoch": 91.12244897959184,
+      "grad_norm": 3.5620455741882324,
+      "learning_rate": 1.7755102040816327e-06,
+      "loss": 0.2049,
+      "step": 17860
+    },
+    {
+      "epoch": 91.1734693877551,
+      "grad_norm": 15.82820987701416,
+      "learning_rate": 1.7653061224489798e-06,
+      "loss": 0.1595,
+      "step": 17870
+    },
+    {
+      "epoch": 91.22448979591837,
+      "grad_norm": 0.8638943433761597,
+      "learning_rate": 1.7551020408163267e-06,
+      "loss": 0.1957,
+      "step": 17880
+    },
+    {
+      "epoch": 91.27551020408163,
+      "grad_norm": 6.596780776977539,
+      "learning_rate": 1.7448979591836737e-06,
+      "loss": 0.1857,
+      "step": 17890
+    },
+    {
+      "epoch": 91.3265306122449,
+      "grad_norm": 19.93320083618164,
+      "learning_rate": 1.7346938775510206e-06,
+      "loss": 0.3028,
+      "step": 17900
+    },
+    {
+      "epoch": 91.37755102040816,
+      "grad_norm": 1.593752145767212,
+      "learning_rate": 1.7244897959183676e-06,
+      "loss": 0.1536,
+      "step": 17910
+    },
+    {
+      "epoch": 91.42857142857143,
+      "grad_norm": 6.460892200469971,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 0.1282,
+      "step": 17920
+    },
+    {
+      "epoch": 91.4795918367347,
+      "grad_norm": 0.6627766489982605,
+      "learning_rate": 1.7040816326530613e-06,
+      "loss": 0.3066,
+      "step": 17930
+    },
+    {
+      "epoch": 91.53061224489795,
+      "grad_norm": 5.6217360496521,
+      "learning_rate": 1.6938775510204084e-06,
+      "loss": 0.1714,
+      "step": 17940
+    },
+    {
+      "epoch": 91.58163265306122,
+      "grad_norm": 2.5595052242279053,
+      "learning_rate": 1.6836734693877552e-06,
+      "loss": 0.1745,
+      "step": 17950
+    },
+    {
+      "epoch": 91.63265306122449,
+      "grad_norm": 6.671074867248535,
+      "learning_rate": 1.6734693877551023e-06,
+      "loss": 0.2174,
+      "step": 17960
+    },
+    {
+      "epoch": 91.68367346938776,
+      "grad_norm": 8.686925888061523,
+      "learning_rate": 1.6632653061224492e-06,
+      "loss": 0.2493,
+      "step": 17970
+    },
+    {
+      "epoch": 91.73469387755102,
+      "grad_norm": 0.4992872476577759,
+      "learning_rate": 1.6530612244897962e-06,
+      "loss": 0.0777,
+      "step": 17980
+    },
+    {
+      "epoch": 91.78571428571429,
+      "grad_norm": 0.5224694609642029,
+      "learning_rate": 1.642857142857143e-06,
+      "loss": 0.1589,
+      "step": 17990
+    },
+    {
+      "epoch": 91.83673469387755,
+      "grad_norm": 20.030405044555664,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 0.5603,
+      "step": 18000
+    },
+    {
+      "epoch": 91.88775510204081,
+      "grad_norm": 8.830347061157227,
+      "learning_rate": 1.622448979591837e-06,
+      "loss": 0.093,
+      "step": 18010
+    },
+    {
+      "epoch": 91.93877551020408,
+      "grad_norm": 5.54904317855835,
+      "learning_rate": 1.6122448979591836e-06,
+      "loss": 0.1399,
+      "step": 18020
+    },
+    {
+      "epoch": 91.98979591836735,
+      "grad_norm": 1.547883152961731,
+      "learning_rate": 1.602040816326531e-06,
+      "loss": 0.1917,
+      "step": 18030
+    },
+    {
+      "epoch": 92.0,
+      "eval_accuracy": 0.9061371841155235,
+      "eval_loss": 0.32391440868377686,
+      "eval_runtime": 0.9963,
+      "eval_samples_per_second": 278.038,
+      "eval_steps_per_second": 35.131,
+      "step": 18032
+    },
+    {
+      "epoch": 92.04081632653062,
+      "grad_norm": 0.934727132320404,
+      "learning_rate": 1.5918367346938775e-06,
+      "loss": 0.0907,
+      "step": 18040
+    },
+    {
+      "epoch": 92.09183673469387,
+      "grad_norm": 20.481788635253906,
+      "learning_rate": 1.5816326530612248e-06,
+      "loss": 0.2082,
+      "step": 18050
+    },
+    {
+      "epoch": 92.14285714285714,
+      "grad_norm": 0.2531684935092926,
+      "learning_rate": 1.5714285714285714e-06,
+      "loss": 0.1877,
+      "step": 18060
+    },
+    {
+      "epoch": 92.1938775510204,
+      "grad_norm": 1.0499683618545532,
+      "learning_rate": 1.5612244897959187e-06,
+      "loss": 0.2461,
+      "step": 18070
+    },
+    {
+      "epoch": 92.24489795918367,
+      "grad_norm": 7.276120185852051,
+      "learning_rate": 1.5510204081632654e-06,
+      "loss": 0.2614,
+      "step": 18080
+    },
+    {
+      "epoch": 92.29591836734694,
+      "grad_norm": 11.999966621398926,
+      "learning_rate": 1.5408163265306122e-06,
+      "loss": 0.2402,
+      "step": 18090
+    },
+    {
+      "epoch": 92.34693877551021,
+      "grad_norm": 7.187252998352051,
+      "learning_rate": 1.5306122448979593e-06,
+      "loss": 0.2077,
+      "step": 18100
+    },
+    {
+      "epoch": 92.39795918367346,
+      "grad_norm": 7.240973472595215,
+      "learning_rate": 1.5204081632653061e-06,
+      "loss": 0.298,
+      "step": 18110
+    },
+    {
+      "epoch": 92.44897959183673,
+      "grad_norm": 0.2603718936443329,
+      "learning_rate": 1.5102040816326532e-06,
+      "loss": 0.1392,
+      "step": 18120
+    },
+    {
+      "epoch": 92.5,
+      "grad_norm": 0.6799657344818115,
+      "learning_rate": 1.5e-06,
+      "loss": 0.1727,
+      "step": 18130
+    },
+    {
+      "epoch": 92.55102040816327,
+      "grad_norm": 6.5458574295043945,
+      "learning_rate": 1.489795918367347e-06,
+      "loss": 0.1966,
+      "step": 18140
+    },
+    {
+      "epoch": 92.60204081632654,
+      "grad_norm": 13.007467269897461,
+      "learning_rate": 1.479591836734694e-06,
+      "loss": 0.2052,
+      "step": 18150
+    },
+    {
+      "epoch": 92.65306122448979,
+      "grad_norm": 13.764470100402832,
+      "learning_rate": 1.469387755102041e-06,
+      "loss": 0.2512,
+      "step": 18160
+    },
+    {
+      "epoch": 92.70408163265306,
+      "grad_norm": 0.7824844717979431,
+      "learning_rate": 1.4591836734693879e-06,
+      "loss": 0.0945,
+      "step": 18170
+    },
+    {
+      "epoch": 92.75510204081633,
+      "grad_norm": 16.08002471923828,
+      "learning_rate": 1.4489795918367347e-06,
+      "loss": 0.3292,
+      "step": 18180
+    },
+    {
+      "epoch": 92.8061224489796,
+      "grad_norm": 1.7446142435073853,
+      "learning_rate": 1.4387755102040818e-06,
+      "loss": 0.3982,
+      "step": 18190
+    },
+    {
+      "epoch": 92.85714285714286,
+      "grad_norm": 0.45311930775642395,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.2266,
+      "step": 18200
+    },
+    {
+      "epoch": 92.90816326530613,
+      "grad_norm": 1.0091290473937988,
+      "learning_rate": 1.4183673469387757e-06,
+      "loss": 0.3635,
+      "step": 18210
+    },
+    {
+      "epoch": 92.95918367346938,
+      "grad_norm": 9.684619903564453,
+      "learning_rate": 1.4081632653061225e-06,
+      "loss": 0.2082,
+      "step": 18220
+    },
+    {
+      "epoch": 93.0,
+      "eval_accuracy": 0.8989169675090253,
+      "eval_loss": 0.4861180782318115,
+      "eval_runtime": 0.9336,
+      "eval_samples_per_second": 296.71,
+      "eval_steps_per_second": 37.49,
+      "step": 18228
+    },
+    {
+      "epoch": 93.01020408163265,
+      "grad_norm": 7.2504401206970215,
+      "learning_rate": 1.3979591836734696e-06,
+      "loss": 0.4401,
+      "step": 18230
+    },
+    {
+      "epoch": 93.06122448979592,
+      "grad_norm": 2.0435526371002197,
+      "learning_rate": 1.3877551020408165e-06,
+      "loss": 0.224,
+      "step": 18240
+    },
+    {
+      "epoch": 93.11224489795919,
+      "grad_norm": 11.459671020507812,
+      "learning_rate": 1.3775510204081633e-06,
+      "loss": 0.258,
+      "step": 18250
+    },
+    {
+      "epoch": 93.16326530612245,
+      "grad_norm": 7.774988651275635,
+      "learning_rate": 1.3673469387755104e-06,
+      "loss": 0.1181,
+      "step": 18260
+    },
+    {
+      "epoch": 93.21428571428571,
+      "grad_norm": 11.726682662963867,
+      "learning_rate": 1.3571428571428572e-06,
+      "loss": 0.2191,
+      "step": 18270
+    },
+    {
+      "epoch": 93.26530612244898,
+      "grad_norm": 10.793588638305664,
+      "learning_rate": 1.3469387755102043e-06,
+      "loss": 0.2481,
+      "step": 18280
+    },
+    {
+      "epoch": 93.31632653061224,
+      "grad_norm": 6.10498571395874,
+      "learning_rate": 1.3367346938775511e-06,
+      "loss": 0.4219,
+      "step": 18290
+    },
+    {
+      "epoch": 93.36734693877551,
+      "grad_norm": 2.345142126083374,
+      "learning_rate": 1.3265306122448982e-06,
+      "loss": 0.2457,
+      "step": 18300
+    },
+    {
+      "epoch": 93.41836734693878,
+      "grad_norm": 10.917703628540039,
+      "learning_rate": 1.316326530612245e-06,
+      "loss": 0.2174,
+      "step": 18310
+    },
+    {
+      "epoch": 93.46938775510205,
+      "grad_norm": 1.5860787630081177,
+      "learning_rate": 1.3061224489795921e-06,
+      "loss": 0.1963,
+      "step": 18320
+    },
+    {
+      "epoch": 93.5204081632653,
+      "grad_norm": 0.7468981742858887,
+      "learning_rate": 1.295918367346939e-06,
+      "loss": 0.1721,
+      "step": 18330
+    },
+    {
+      "epoch": 93.57142857142857,
+      "grad_norm": 5.929870128631592,
+      "learning_rate": 1.2857142857142856e-06,
+      "loss": 0.2093,
+      "step": 18340
+    },
+    {
+      "epoch": 93.62244897959184,
+      "grad_norm": 7.6047821044921875,
+      "learning_rate": 1.2755102040816329e-06,
+      "loss": 0.1547,
+      "step": 18350
+    },
+    {
+      "epoch": 93.6734693877551,
+      "grad_norm": 0.3163466155529022,
+      "learning_rate": 1.2653061224489795e-06,
+      "loss": 0.1218,
+      "step": 18360
+    },
+    {
+      "epoch": 93.72448979591837,
+      "grad_norm": 1.5049079656600952,
+      "learning_rate": 1.2551020408163268e-06,
+      "loss": 0.1935,
+      "step": 18370
+    },
+    {
+      "epoch": 93.77551020408163,
+      "grad_norm": 2.0531539916992188,
+      "learning_rate": 1.2448979591836734e-06,
+      "loss": 0.3614,
+      "step": 18380
+    },
+    {
+      "epoch": 93.8265306122449,
+      "grad_norm": 2.066619634628296,
+      "learning_rate": 1.2346938775510205e-06,
+      "loss": 0.1524,
+      "step": 18390
+    },
+    {
+      "epoch": 93.87755102040816,
+      "grad_norm": 0.1106644943356514,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 0.1908,
+      "step": 18400
+    },
+    {
+      "epoch": 93.92857142857143,
+      "grad_norm": 0.2896367609500885,
+      "learning_rate": 1.2142857142857144e-06,
+      "loss": 0.0657,
+      "step": 18410
+    },
+    {
+      "epoch": 93.9795918367347,
+      "grad_norm": 22.562599182128906,
+      "learning_rate": 1.2040816326530612e-06,
+      "loss": 0.3836,
+      "step": 18420
+    },
+    {
+      "epoch": 94.0,
+      "eval_accuracy": 0.8736462093862816,
+      "eval_loss": 0.4443919062614441,
+      "eval_runtime": 0.9771,
+      "eval_samples_per_second": 283.506,
+      "eval_steps_per_second": 35.822,
+      "step": 18424
+    },
+    {
+      "epoch": 94.03061224489795,
+      "grad_norm": 4.277006149291992,
+      "learning_rate": 1.1938775510204083e-06,
+      "loss": 0.1953,
+      "step": 18430
+    },
+    {
+      "epoch": 94.08163265306122,
+      "grad_norm": 5.866278648376465,
+      "learning_rate": 1.1836734693877552e-06,
+      "loss": 0.2129,
+      "step": 18440
+    },
+    {
+      "epoch": 94.13265306122449,
+      "grad_norm": 13.665555000305176,
+      "learning_rate": 1.1734693877551022e-06,
+      "loss": 0.171,
+      "step": 18450
+    },
+    {
+      "epoch": 94.18367346938776,
+      "grad_norm": 6.091044902801514,
+      "learning_rate": 1.163265306122449e-06,
+      "loss": 0.172,
+      "step": 18460
+    },
+    {
+      "epoch": 94.23469387755102,
+      "grad_norm": 13.635805130004883,
+      "learning_rate": 1.153061224489796e-06,
+      "loss": 0.3361,
+      "step": 18470
+    },
+    {
+      "epoch": 94.28571428571429,
+      "grad_norm": 1.790496826171875,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 0.0893,
+      "step": 18480
+    },
+    {
+      "epoch": 94.33673469387755,
+      "grad_norm": 3.7977652549743652,
+      "learning_rate": 1.1326530612244898e-06,
+      "loss": 0.1547,
+      "step": 18490
+    },
+    {
+      "epoch": 94.38775510204081,
+      "grad_norm": 7.650182247161865,
+      "learning_rate": 1.122448979591837e-06,
+      "loss": 0.2096,
+      "step": 18500
+    },
+    {
+      "epoch": 94.43877551020408,
+      "grad_norm": 4.0888352394104,
+      "learning_rate": 1.1122448979591838e-06,
+      "loss": 0.1384,
+      "step": 18510
+    },
+    {
+      "epoch": 94.48979591836735,
+      "grad_norm": 4.018925666809082,
+      "learning_rate": 1.1020408163265308e-06,
+      "loss": 0.1704,
+      "step": 18520
+    },
+    {
+      "epoch": 94.54081632653062,
+      "grad_norm": 3.2753589153289795,
+      "learning_rate": 1.0918367346938777e-06,
+      "loss": 0.1741,
+      "step": 18530
+    },
+    {
+      "epoch": 94.59183673469387,
+      "grad_norm": 18.035402297973633,
+      "learning_rate": 1.0816326530612247e-06,
+      "loss": 0.1346,
+      "step": 18540
+    },
+    {
+      "epoch": 94.64285714285714,
+      "grad_norm": 6.348093032836914,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 0.1886,
+      "step": 18550
+    },
+    {
+      "epoch": 94.6938775510204,
+      "grad_norm": 6.390273571014404,
+      "learning_rate": 1.0612244897959184e-06,
+      "loss": 0.1955,
+      "step": 18560
+    },
+    {
+      "epoch": 94.74489795918367,
+      "grad_norm": 0.4464976489543915,
+      "learning_rate": 1.0510204081632653e-06,
+      "loss": 0.2558,
+      "step": 18570
+    },
+    {
+      "epoch": 94.79591836734694,
+      "grad_norm": 11.005455017089844,
+      "learning_rate": 1.0408163265306123e-06,
+      "loss": 0.2175,
+      "step": 18580
+    },
+    {
+      "epoch": 94.84693877551021,
+      "grad_norm": 5.856806755065918,
+      "learning_rate": 1.0306122448979592e-06,
+      "loss": 0.2982,
+      "step": 18590
+    },
+    {
+      "epoch": 94.89795918367346,
+      "grad_norm": 1.477173089981079,
+      "learning_rate": 1.0204081632653063e-06,
+      "loss": 0.2076,
+      "step": 18600
+    },
+    {
+      "epoch": 94.94897959183673,
+      "grad_norm": 12.501953125,
+      "learning_rate": 1.010204081632653e-06,
+      "loss": 0.1872,
+      "step": 18610
+    },
+    {
+      "epoch": 95.0,
+      "grad_norm": 1.0849741697311401,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.1,
+      "step": 18620
+    },
+    {
+      "epoch": 95.0,
+      "eval_accuracy": 0.8844765342960289,
+      "eval_loss": 0.3712524473667145,
+      "eval_runtime": 0.935,
+      "eval_samples_per_second": 296.245,
+      "eval_steps_per_second": 37.432,
+      "step": 18620
+    },
+    {
+      "epoch": 95.05102040816327,
+      "grad_norm": 0.12624509632587433,
+      "learning_rate": 9.89795918367347e-07,
+      "loss": 0.0413,
+      "step": 18630
+    },
+    {
+      "epoch": 95.10204081632654,
+      "grad_norm": 5.990360736846924,
+      "learning_rate": 9.795918367346939e-07,
+      "loss": 0.2007,
+      "step": 18640
+    },
+    {
+      "epoch": 95.15306122448979,
+      "grad_norm": 10.36894416809082,
+      "learning_rate": 9.69387755102041e-07,
+      "loss": 0.1651,
+      "step": 18650
+    },
+    {
+      "epoch": 95.20408163265306,
+      "grad_norm": 4.233863830566406,
+      "learning_rate": 9.591836734693878e-07,
+      "loss": 0.1365,
+      "step": 18660
+    },
+    {
+      "epoch": 95.25510204081633,
+      "grad_norm": 2.113361358642578,
+      "learning_rate": 9.489795918367347e-07,
+      "loss": 0.2163,
+      "step": 18670
+    },
+    {
+      "epoch": 95.3061224489796,
+      "grad_norm": 1.022985816001892,
+      "learning_rate": 9.387755102040817e-07,
+      "loss": 0.1879,
+      "step": 18680
+    },
+    {
+      "epoch": 95.35714285714286,
+      "grad_norm": 2.9235568046569824,
+      "learning_rate": 9.285714285714287e-07,
+      "loss": 0.0615,
+      "step": 18690
+    },
+    {
+      "epoch": 95.40816326530613,
+      "grad_norm": 3.9105029106140137,
+      "learning_rate": 9.183673469387756e-07,
+      "loss": 0.3205,
+      "step": 18700
+    },
+    {
+      "epoch": 95.45918367346938,
+      "grad_norm": 7.213064670562744,
+      "learning_rate": 9.081632653061226e-07,
+      "loss": 0.3415,
+      "step": 18710
+    },
+    {
+      "epoch": 95.51020408163265,
+      "grad_norm": 3.849168539047241,
+      "learning_rate": 8.979591836734694e-07,
+      "loss": 0.1077,
+      "step": 18720
+    },
+    {
+      "epoch": 95.56122448979592,
+      "grad_norm": 0.49878036975860596,
+      "learning_rate": 8.877551020408164e-07,
+      "loss": 0.2587,
+      "step": 18730
+    },
+    {
+      "epoch": 95.61224489795919,
+      "grad_norm": 18.13332748413086,
+      "learning_rate": 8.775510204081633e-07,
+      "loss": 0.3615,
+      "step": 18740
+    },
+    {
+      "epoch": 95.66326530612245,
+      "grad_norm": 6.391093730926514,
+      "learning_rate": 8.673469387755103e-07,
+      "loss": 0.1355,
+      "step": 18750
+    },
+    {
+      "epoch": 95.71428571428571,
+      "grad_norm": 15.012079238891602,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 0.1515,
+      "step": 18760
+    },
+    {
+      "epoch": 95.76530612244898,
+      "grad_norm": 0.518864631652832,
+      "learning_rate": 8.469387755102042e-07,
+      "loss": 0.2164,
+      "step": 18770
+    },
+    {
+      "epoch": 95.81632653061224,
+      "grad_norm": 21.811342239379883,
+      "learning_rate": 8.367346938775512e-07,
+      "loss": 0.3386,
+      "step": 18780
+    },
+    {
+      "epoch": 95.86734693877551,
+      "grad_norm": 8.917824745178223,
+      "learning_rate": 8.265306122448981e-07,
+      "loss": 0.2287,
+      "step": 18790
+    },
+    {
+      "epoch": 95.91836734693878,
+      "grad_norm": 2.237412214279175,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 0.0678,
+      "step": 18800
+    },
+    {
+      "epoch": 95.96938775510205,
+      "grad_norm": 6.585892677307129,
+      "learning_rate": 8.061224489795918e-07,
+      "loss": 0.1785,
+      "step": 18810
+    },
+    {
+      "epoch": 96.0,
+      "eval_accuracy": 0.8303249097472925,
+      "eval_loss": 0.427941232919693,
+      "eval_runtime": 0.9327,
+      "eval_samples_per_second": 296.984,
+      "eval_steps_per_second": 37.525,
+      "step": 18816
+    },
+    {
+      "epoch": 96.0204081632653,
+      "grad_norm": 0.1639726758003235,
+      "learning_rate": 7.959183673469388e-07,
+      "loss": 0.1298,
+      "step": 18820
+    },
+    {
+      "epoch": 96.07142857142857,
+      "grad_norm": 2.6104323863983154,
+      "learning_rate": 7.857142857142857e-07,
+      "loss": 0.0609,
+      "step": 18830
+    },
+    {
+      "epoch": 96.12244897959184,
+      "grad_norm": 8.791910171508789,
+      "learning_rate": 7.755102040816327e-07,
+      "loss": 0.2782,
+      "step": 18840
+    },
+    {
+      "epoch": 96.1734693877551,
+      "grad_norm": 0.7717019319534302,
+      "learning_rate": 7.653061224489796e-07,
+      "loss": 0.5283,
+      "step": 18850
+    },
+    {
+      "epoch": 96.22448979591837,
+      "grad_norm": 0.5696031451225281,
+      "learning_rate": 7.551020408163266e-07,
+      "loss": 0.1675,
+      "step": 18860
+    },
+    {
+      "epoch": 96.27551020408163,
+      "grad_norm": 3.209054708480835,
+      "learning_rate": 7.448979591836736e-07,
+      "loss": 0.2733,
+      "step": 18870
+    },
+    {
+      "epoch": 96.3265306122449,
+      "grad_norm": 0.5075634121894836,
+      "learning_rate": 7.346938775510205e-07,
+      "loss": 0.0789,
+      "step": 18880
+    },
+    {
+      "epoch": 96.37755102040816,
+      "grad_norm": 6.202861785888672,
+      "learning_rate": 7.244897959183674e-07,
+      "loss": 0.1033,
+      "step": 18890
+    },
+    {
+      "epoch": 96.42857142857143,
+      "grad_norm": 0.5611489415168762,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.201,
+      "step": 18900
+    },
+    {
+      "epoch": 96.4795918367347,
+      "grad_norm": 1.8156996965408325,
+      "learning_rate": 7.040816326530613e-07,
+      "loss": 0.15,
+      "step": 18910
+    },
+    {
+      "epoch": 96.53061224489795,
+      "grad_norm": 8.122031211853027,
+      "learning_rate": 6.938775510204082e-07,
+      "loss": 0.0895,
+      "step": 18920
+    },
+    {
+      "epoch": 96.58163265306122,
+      "grad_norm": 5.461958408355713,
+      "learning_rate": 6.836734693877552e-07,
+      "loss": 0.0865,
+      "step": 18930
+    },
+    {
+      "epoch": 96.63265306122449,
+      "grad_norm": 2.365891456604004,
+      "learning_rate": 6.734693877551021e-07,
+      "loss": 0.2617,
+      "step": 18940
+    },
+    {
+      "epoch": 96.68367346938776,
+      "grad_norm": 6.673534393310547,
+      "learning_rate": 6.632653061224491e-07,
+      "loss": 0.2343,
+      "step": 18950
+    },
+    {
+      "epoch": 96.73469387755102,
+      "grad_norm": 8.938642501831055,
+      "learning_rate": 6.530612244897961e-07,
+      "loss": 0.2346,
+      "step": 18960
+    },
+    {
+      "epoch": 96.78571428571429,
+      "grad_norm": 1.8851449489593506,
+      "learning_rate": 6.428571428571428e-07,
+      "loss": 0.186,
+      "step": 18970
+    },
+    {
+      "epoch": 96.83673469387755,
+      "grad_norm": 4.962350368499756,
+      "learning_rate": 6.326530612244898e-07,
+      "loss": 0.2458,
+      "step": 18980
+    },
+    {
+      "epoch": 96.88775510204081,
+      "grad_norm": 3.843569755554199,
+      "learning_rate": 6.224489795918367e-07,
+      "loss": 0.3118,
+      "step": 18990
+    },
+    {
+      "epoch": 96.93877551020408,
+      "grad_norm": 0.8823779225349426,
+      "learning_rate": 6.122448979591837e-07,
+      "loss": 0.1672,
+      "step": 19000
+    },
+    {
+      "epoch": 96.98979591836735,
+      "grad_norm": 0.39520347118377686,
+      "learning_rate": 6.020408163265306e-07,
+      "loss": 0.19,
+      "step": 19010
+    },
+    {
+      "epoch": 97.0,
+      "eval_accuracy": 0.8411552346570397,
+      "eval_loss": 0.6587985157966614,
+      "eval_runtime": 0.9348,
+      "eval_samples_per_second": 296.305,
+      "eval_steps_per_second": 37.439,
+      "step": 19012
+    },
+    {
+      "epoch": 97.04081632653062,
+      "grad_norm": 1.7464901208877563,
+      "learning_rate": 5.918367346938776e-07,
+      "loss": 0.1802,
+      "step": 19020
+    },
+    {
+      "epoch": 97.09183673469387,
+      "grad_norm": 4.634471893310547,
+      "learning_rate": 5.816326530612245e-07,
+      "loss": 0.1461,
+      "step": 19030
+    },
+    {
+      "epoch": 97.14285714285714,
+      "grad_norm": 0.36374080181121826,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 0.1334,
+      "step": 19040
+    },
+    {
+      "epoch": 97.1938775510204,
+      "grad_norm": 2.8511147499084473,
+      "learning_rate": 5.612244897959184e-07,
+      "loss": 0.1569,
+      "step": 19050
+    },
+    {
+      "epoch": 97.24489795918367,
+      "grad_norm": 2.9540822505950928,
+      "learning_rate": 5.510204081632654e-07,
+      "loss": 0.1175,
+      "step": 19060
+    },
+    {
+      "epoch": 97.29591836734694,
+      "grad_norm": 0.20516809821128845,
+      "learning_rate": 5.408163265306124e-07,
+      "loss": 0.2094,
+      "step": 19070
+    },
+    {
+      "epoch": 97.34693877551021,
+      "grad_norm": 7.4667816162109375,
+      "learning_rate": 5.306122448979592e-07,
+      "loss": 0.1318,
+      "step": 19080
+    },
+    {
+      "epoch": 97.39795918367346,
+      "grad_norm": 0.7376777529716492,
+      "learning_rate": 5.204081632653062e-07,
+      "loss": 0.103,
+      "step": 19090
+    },
+    {
+      "epoch": 97.44897959183673,
+      "grad_norm": 9.432369232177734,
+      "learning_rate": 5.102040816326531e-07,
+      "loss": 0.1561,
+      "step": 19100
+    },
+    {
+      "epoch": 97.5,
+      "grad_norm": 15.827686309814453,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.3899,
+      "step": 19110
+    },
+    {
+      "epoch": 97.55102040816327,
+      "grad_norm": 7.54927396774292,
+      "learning_rate": 4.897959183673469e-07,
+      "loss": 0.0907,
+      "step": 19120
+    },
+    {
+      "epoch": 97.60204081632654,
+      "grad_norm": 14.595918655395508,
+      "learning_rate": 4.795918367346939e-07,
+      "loss": 0.5122,
+      "step": 19130
+    },
+    {
+      "epoch": 97.65306122448979,
+      "grad_norm": 1.9597631692886353,
+      "learning_rate": 4.6938775510204085e-07,
+      "loss": 0.3503,
+      "step": 19140
+    },
+    {
+      "epoch": 97.70408163265306,
+      "grad_norm": 19.388639450073242,
+      "learning_rate": 4.591836734693878e-07,
+      "loss": 0.3461,
+      "step": 19150
+    },
+    {
+      "epoch": 97.75510204081633,
+      "grad_norm": 2.2139663696289062,
+      "learning_rate": 4.489795918367347e-07,
+      "loss": 0.1561,
+      "step": 19160
+    },
+    {
+      "epoch": 97.8061224489796,
+      "grad_norm": 7.7265520095825195,
+      "learning_rate": 4.3877551020408166e-07,
+      "loss": 0.3166,
+      "step": 19170
+    },
+    {
+      "epoch": 97.85714285714286,
+      "grad_norm": 2.6642425060272217,
+      "learning_rate": 4.285714285714286e-07,
+      "loss": 0.0529,
+      "step": 19180
+    },
+    {
+      "epoch": 97.90816326530613,
+      "grad_norm": 5.023648262023926,
+      "learning_rate": 4.183673469387756e-07,
+      "loss": 0.2524,
+      "step": 19190
+    },
+    {
+      "epoch": 97.95918367346938,
+      "grad_norm": 13.295014381408691,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 0.099,
+      "step": 19200
+    },
+    {
+      "epoch": 98.0,
+      "eval_accuracy": 0.8267148014440433,
+      "eval_loss": 0.66323322057724,
+      "eval_runtime": 0.9334,
+      "eval_samples_per_second": 296.754,
+      "eval_steps_per_second": 37.496,
+      "step": 19208
+    },
+    {
+      "epoch": 98.01020408163265,
+      "grad_norm": 0.9346858263015747,
+      "learning_rate": 3.979591836734694e-07,
+      "loss": 0.1299,
+      "step": 19210
+    },
+    {
+      "epoch": 98.06122448979592,
+      "grad_norm": 9.008681297302246,
+      "learning_rate": 3.8775510204081634e-07,
+      "loss": 0.3403,
+      "step": 19220
+    },
+    {
+      "epoch": 98.11224489795919,
+      "grad_norm": 2.1377320289611816,
+      "learning_rate": 3.775510204081633e-07,
+      "loss": 0.1092,
+      "step": 19230
+    },
+    {
+      "epoch": 98.16326530612245,
+      "grad_norm": 21.07135581970215,
+      "learning_rate": 3.6734693877551025e-07,
+      "loss": 0.1871,
+      "step": 19240
+    },
+    {
+      "epoch": 98.21428571428571,
+      "grad_norm": 6.922109127044678,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 0.1675,
+      "step": 19250
+    },
+    {
+      "epoch": 98.26530612244898,
+      "grad_norm": 1.566651463508606,
+      "learning_rate": 3.469387755102041e-07,
+      "loss": 0.2421,
+      "step": 19260
+    },
+    {
+      "epoch": 98.31632653061224,
+      "grad_norm": 9.104406356811523,
+      "learning_rate": 3.3673469387755107e-07,
+      "loss": 0.441,
+      "step": 19270
+    },
+    {
+      "epoch": 98.36734693877551,
+      "grad_norm": 2.4959702491760254,
+      "learning_rate": 3.2653061224489803e-07,
+      "loss": 0.1556,
+      "step": 19280
+    },
+    {
+      "epoch": 98.41836734693878,
+      "grad_norm": 12.012667655944824,
+      "learning_rate": 3.163265306122449e-07,
+      "loss": 0.1268,
+      "step": 19290
+    },
+    {
+      "epoch": 98.46938775510205,
+      "grad_norm": 0.16600196063518524,
+      "learning_rate": 3.0612244897959183e-07,
+      "loss": 0.2915,
+      "step": 19300
+    },
+    {
+      "epoch": 98.5204081632653,
+      "grad_norm": 5.219106197357178,
+      "learning_rate": 2.959183673469388e-07,
+      "loss": 0.2216,
+      "step": 19310
+    },
+    {
+      "epoch": 98.57142857142857,
+      "grad_norm": 0.5317900776863098,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 0.0964,
+      "step": 19320
+    },
+    {
+      "epoch": 98.62244897959184,
+      "grad_norm": 2.0076568126678467,
+      "learning_rate": 2.755102040816327e-07,
+      "loss": 0.112,
+      "step": 19330
+    },
+    {
+      "epoch": 98.6734693877551,
+      "grad_norm": 8.94671630859375,
+      "learning_rate": 2.653061224489796e-07,
+      "loss": 0.2102,
+      "step": 19340
+    },
+    {
+      "epoch": 98.72448979591837,
+      "grad_norm": 1.0493749380111694,
+      "learning_rate": 2.5510204081632656e-07,
+      "loss": 0.1934,
+      "step": 19350
+    },
+    {
+      "epoch": 98.77551020408163,
+      "grad_norm": 0.16692312061786652,
+      "learning_rate": 2.4489795918367347e-07,
+      "loss": 0.1121,
+      "step": 19360
+    },
+    {
+      "epoch": 98.8265306122449,
+      "grad_norm": 14.747842788696289,
+      "learning_rate": 2.3469387755102042e-07,
+      "loss": 0.1758,
+      "step": 19370
+    },
+    {
+      "epoch": 98.87755102040816,
+      "grad_norm": 13.387628555297852,
+      "learning_rate": 2.2448979591836735e-07,
+      "loss": 0.2225,
+      "step": 19380
+    },
+    {
+      "epoch": 98.92857142857143,
+      "grad_norm": 0.7674017548561096,
+      "learning_rate": 2.142857142857143e-07,
+      "loss": 0.2042,
+      "step": 19390
+    },
+    {
+      "epoch": 98.9795918367347,
+      "grad_norm": 6.51180362701416,
+      "learning_rate": 2.0408163265306121e-07,
+      "loss": 0.1467,
+      "step": 19400
+    },
+    {
+      "epoch": 99.0,
+      "eval_accuracy": 0.8808664259927798,
+      "eval_loss": 0.4641692638397217,
+      "eval_runtime": 0.9294,
+      "eval_samples_per_second": 298.047,
+      "eval_steps_per_second": 37.659,
+      "step": 19404
+    },
+    {
+      "epoch": 99.03061224489795,
+      "grad_norm": 9.76559066772461,
+      "learning_rate": 1.9387755102040817e-07,
+      "loss": 0.2495,
+      "step": 19410
+    },
+    {
+      "epoch": 99.08163265306122,
+      "grad_norm": 1.884592056274414,
+      "learning_rate": 1.8367346938775513e-07,
+      "loss": 0.2225,
+      "step": 19420
+    },
+    {
+      "epoch": 99.13265306122449,
+      "grad_norm": 6.432409763336182,
+      "learning_rate": 1.7346938775510206e-07,
+      "loss": 0.2526,
+      "step": 19430
+    },
+    {
+      "epoch": 99.18367346938776,
+      "grad_norm": 4.175636291503906,
+      "learning_rate": 1.6326530612244901e-07,
+      "loss": 0.6109,
+      "step": 19440
+    },
+    {
+      "epoch": 99.23469387755102,
+      "grad_norm": 15.096162796020508,
+      "learning_rate": 1.5306122448979592e-07,
+      "loss": 0.276,
+      "step": 19450
+    },
+    {
+      "epoch": 99.28571428571429,
+      "grad_norm": 7.944954872131348,
+      "learning_rate": 1.4285714285714287e-07,
+      "loss": 0.0812,
+      "step": 19460
+    },
+    {
+      "epoch": 99.33673469387755,
+      "grad_norm": 3.630222797393799,
+      "learning_rate": 1.326530612244898e-07,
+      "loss": 0.2417,
+      "step": 19470
+    },
+    {
+      "epoch": 99.38775510204081,
+      "grad_norm": 1.4471020698547363,
+      "learning_rate": 1.2244897959183673e-07,
+      "loss": 0.1571,
+      "step": 19480
+    },
+    {
+      "epoch": 99.43877551020408,
+      "grad_norm": 1.2315754890441895,
+      "learning_rate": 1.1224489795918368e-07,
+      "loss": 0.199,
+      "step": 19490
+    },
+    {
+      "epoch": 99.48979591836735,
+      "grad_norm": 3.3233890533447266,
+      "learning_rate": 1.0204081632653061e-07,
+      "loss": 0.2458,
+      "step": 19500
+    },
+    {
+      "epoch": 99.54081632653062,
+      "grad_norm": 2.3438429832458496,
+      "learning_rate": 9.183673469387756e-08,
+      "loss": 0.0734,
+      "step": 19510
+    },
+    {
+      "epoch": 99.59183673469387,
+      "grad_norm": 4.115237712860107,
+      "learning_rate": 8.163265306122451e-08,
+      "loss": 0.1425,
+      "step": 19520
+    },
+    {
+      "epoch": 99.64285714285714,
+      "grad_norm": 16.608760833740234,
+      "learning_rate": 7.142857142857144e-08,
+      "loss": 0.1921,
+      "step": 19530
+    },
+    {
+      "epoch": 99.6938775510204,
+      "grad_norm": 6.704372406005859,
+      "learning_rate": 6.122448979591837e-08,
+      "loss": 0.1868,
+      "step": 19540
+    },
+    {
+      "epoch": 99.74489795918367,
+      "grad_norm": 6.689555644989014,
+      "learning_rate": 5.1020408163265303e-08,
+      "loss": 0.0779,
+      "step": 19550
+    },
+    {
+      "epoch": 99.79591836734694,
+      "grad_norm": 1.298974871635437,
+      "learning_rate": 4.0816326530612253e-08,
+      "loss": 0.148,
+      "step": 19560
+    },
+    {
+      "epoch": 99.84693877551021,
+      "grad_norm": 4.185670375823975,
+      "learning_rate": 3.0612244897959183e-08,
+      "loss": 0.2843,
+      "step": 19570
+    },
+    {
+      "epoch": 99.89795918367346,
+      "grad_norm": 2.600241184234619,
+      "learning_rate": 2.0408163265306127e-08,
+      "loss": 0.1492,
+      "step": 19580
+    },
+    {
+      "epoch": 99.94897959183673,
+      "grad_norm": 14.082707405090332,
+      "learning_rate": 1.0204081632653063e-08,
+      "loss": 0.1229,
+      "step": 19590
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 0.1434904783964157,
+      "learning_rate": 0.0,
+      "loss": 0.2617,
+      "step": 19600
+    },
+    {
+      "epoch": 100.0,
+      "eval_accuracy": 0.8808664259927798,
+      "eval_loss": 0.3624129593372345,
+      "eval_runtime": 1.0133,
+      "eval_samples_per_second": 273.363,
+      "eval_steps_per_second": 34.54,
+      "step": 19600
+    },
+    {
+      "epoch": 100.0,
+      "step": 19600,
+      "total_flos": 5.678990728814592e+17,
+      "train_loss": 0.30055975163317455,
+      "train_runtime": 951.6633,
+      "train_samples_per_second": 164.554,
+      "train_steps_per_second": 20.596
     }
   ],
   "logging_steps": 10,
-  "max_steps": 980,
+  "max_steps": 19600,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 100,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -766,7 +14655,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.6633116935737344e+17,
+  "total_flos": 5.678990728814592e+17,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null