diff --git a/all_results.json b/all_results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7cee57f317119bc7dce107069e74104d4f6d5e4
--- /dev/null
+++ b/all_results.json
@@ -0,0 +1,8 @@
+{
+    "epoch": 49.02,
+    "eval_accuracy": 0.8875,
+    "eval_loss": 0.4278368353843689,
+    "eval_runtime": 1123.0215,
+    "eval_samples_per_second": 0.57,
+    "eval_steps_per_second": 0.071
+}
\ No newline at end of file
diff --git a/runs/Apr24_20-05-06_blackhorse/events.out.tfevents.1714244811.blackhorse.9061.1 b/runs/Apr24_20-05-06_blackhorse/events.out.tfevents.1714244811.blackhorse.9061.1
index b5bc02fa96ed1bf18d795c489b310dd2bf997378..e2b00eb7fda540bd19c35dbd50bd7692c54ada1b 100644
--- a/runs/Apr24_20-05-06_blackhorse/events.out.tfevents.1714244811.blackhorse.9061.1
+++ b/runs/Apr24_20-05-06_blackhorse/events.out.tfevents.1714244811.blackhorse.9061.1
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d58156d37c20e78424840356fe1b1b1f834d116f1a8653284b358a053ea5a8de
-size 411
+oid sha256:83d67358f4de06b288aa9d32aa3642fd3b42e99f24fcdcab8e824f7aaeb18629
+size 734
diff --git a/test_results.json b/test_results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7cee57f317119bc7dce107069e74104d4f6d5e4
--- /dev/null
+++ b/test_results.json
@@ -0,0 +1,8 @@
+{
+    "epoch": 49.02,
+    "eval_accuracy": 0.8875,
+    "eval_loss": 0.4278368353843689,
+    "eval_runtime": 1123.0215,
+    "eval_samples_per_second": 0.57,
+    "eval_steps_per_second": 0.071
+}
\ No newline at end of file
diff --git a/trainer_state.json b/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7a9b4c31d59c8f7468d5fd303f94028a2954438
--- /dev/null
+++ b/trainer_state.json
@@ -0,0 +1,11558 @@
+{
+  "best_metric": 0.9769673704414588,
+  "best_model_checkpoint": "videomae-base-finetuned-isl-numbers-alphabet-nouns/checkpoint-13904",
+  "epoch": 49.02,
+  "eval_steps": 500,
+  "global_step": 15800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006329113924050633,
+      "grad_norm": 7.342864990234375,
+      "learning_rate": 3.1645569620253163e-07,
+      "loss": 5.0266,
+      "step": 10
+    },
+    {
+      "epoch": 0.0012658227848101266,
+      "grad_norm": 7.16239070892334,
+      "learning_rate": 6.329113924050633e-07,
+      "loss": 5.0833,
+      "step": 20
+    },
+    {
+      "epoch": 0.0018987341772151898,
+      "grad_norm": 7.36508846282959,
+      "learning_rate": 9.493670886075951e-07,
+      "loss": 5.0852,
+      "step": 30
+    },
+    {
+      "epoch": 0.002531645569620253,
+      "grad_norm": 7.174118995666504,
+      "learning_rate": 1.2658227848101265e-06,
+      "loss": 5.1223,
+      "step": 40
+    },
+    {
+      "epoch": 0.0031645569620253164,
+      "grad_norm": 7.650392055511475,
+      "learning_rate": 1.5822784810126583e-06,
+      "loss": 5.0765,
+      "step": 50
+    },
+    {
+      "epoch": 0.0037974683544303796,
+      "grad_norm": 7.764711856842041,
+      "learning_rate": 1.8987341772151901e-06,
+      "loss": 5.0422,
+      "step": 60
+    },
+    {
+      "epoch": 0.004430379746835443,
+      "grad_norm": 11.08586311340332,
+      "learning_rate": 2.2151898734177215e-06,
+      "loss": 5.0661,
+      "step": 70
+    },
+    {
+      "epoch": 0.005063291139240506,
+      "grad_norm": 7.217874050140381,
+      "learning_rate": 2.531645569620253e-06,
+      "loss": 5.0943,
+      "step": 80
+    },
+    {
+      "epoch": 0.00569620253164557,
+      "grad_norm": 7.349112510681152,
+      "learning_rate": 2.848101265822785e-06,
+      "loss": 5.1201,
+      "step": 90
+    },
+    {
+      "epoch": 0.006329113924050633,
+      "grad_norm": 7.260085582733154,
+      "learning_rate": 3.1645569620253167e-06,
+      "loss": 5.0797,
+      "step": 100
+    },
+    {
+      "epoch": 0.006962025316455696,
+      "grad_norm": 7.140063762664795,
+      "learning_rate": 3.4810126582278482e-06,
+      "loss": 5.0803,
+      "step": 110
+    },
+    {
+      "epoch": 0.007594936708860759,
+      "grad_norm": 8.30470085144043,
+      "learning_rate": 3.7974683544303802e-06,
+      "loss": 5.061,
+      "step": 120
+    },
+    {
+      "epoch": 0.008227848101265823,
+      "grad_norm": 8.53849983215332,
+      "learning_rate": 4.113924050632911e-06,
+      "loss": 5.0407,
+      "step": 130
+    },
+    {
+      "epoch": 0.008860759493670886,
+      "grad_norm": 7.535670757293701,
+      "learning_rate": 4.430379746835443e-06,
+      "loss": 5.0505,
+      "step": 140
+    },
+    {
+      "epoch": 0.00949367088607595,
+      "grad_norm": 8.011333465576172,
+      "learning_rate": 4.746835443037975e-06,
+      "loss": 5.0643,
+      "step": 150
+    },
+    {
+      "epoch": 0.010126582278481013,
+      "grad_norm": 8.219789505004883,
+      "learning_rate": 5.063291139240506e-06,
+      "loss": 4.9714,
+      "step": 160
+    },
+    {
+      "epoch": 0.010759493670886076,
+      "grad_norm": 8.42809009552002,
+      "learning_rate": 5.379746835443038e-06,
+      "loss": 4.9477,
+      "step": 170
+    },
+    {
+      "epoch": 0.01139240506329114,
+      "grad_norm": 11.367712020874023,
+      "learning_rate": 5.69620253164557e-06,
+      "loss": 4.9509,
+      "step": 180
+    },
+    {
+      "epoch": 0.012025316455696202,
+      "grad_norm": 9.428906440734863,
+      "learning_rate": 6.012658227848101e-06,
+      "loss": 4.9267,
+      "step": 190
+    },
+    {
+      "epoch": 0.012658227848101266,
+      "grad_norm": 8.949156761169434,
+      "learning_rate": 6.329113924050633e-06,
+      "loss": 4.8931,
+      "step": 200
+    },
+    {
+      "epoch": 0.013291139240506329,
+      "grad_norm": 8.922019004821777,
+      "learning_rate": 6.6455696202531645e-06,
+      "loss": 4.921,
+      "step": 210
+    },
+    {
+      "epoch": 0.013924050632911392,
+      "grad_norm": 10.950883865356445,
+      "learning_rate": 6.9620253164556965e-06,
+      "loss": 4.781,
+      "step": 220
+    },
+    {
+      "epoch": 0.014556962025316455,
+      "grad_norm": 10.333183288574219,
+      "learning_rate": 7.2784810126582285e-06,
+      "loss": 4.8136,
+      "step": 230
+    },
+    {
+      "epoch": 0.015189873417721518,
+      "grad_norm": 11.363033294677734,
+      "learning_rate": 7.5949367088607605e-06,
+      "loss": 4.8154,
+      "step": 240
+    },
+    {
+      "epoch": 0.015822784810126583,
+      "grad_norm": 10.919289588928223,
+      "learning_rate": 7.911392405063292e-06,
+      "loss": 4.8229,
+      "step": 250
+    },
+    {
+      "epoch": 0.016455696202531647,
+      "grad_norm": 13.473514556884766,
+      "learning_rate": 8.227848101265822e-06,
+      "loss": 4.6251,
+      "step": 260
+    },
+    {
+      "epoch": 0.01708860759493671,
+      "grad_norm": 13.167325973510742,
+      "learning_rate": 8.544303797468354e-06,
+      "loss": 4.6354,
+      "step": 270
+    },
+    {
+      "epoch": 0.017721518987341773,
+      "grad_norm": 12.691937446594238,
+      "learning_rate": 8.860759493670886e-06,
+      "loss": 4.6029,
+      "step": 280
+    },
+    {
+      "epoch": 0.018354430379746836,
+      "grad_norm": 13.479981422424316,
+      "learning_rate": 9.177215189873418e-06,
+      "loss": 4.621,
+      "step": 290
+    },
+    {
+      "epoch": 0.0189873417721519,
+      "grad_norm": 12.953887939453125,
+      "learning_rate": 9.49367088607595e-06,
+      "loss": 4.5616,
+      "step": 300
+    },
+    {
+      "epoch": 0.019620253164556962,
+      "grad_norm": 13.775838851928711,
+      "learning_rate": 9.81012658227848e-06,
+      "loss": 4.5228,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "eval_accuracy": 0.2533589251439539,
+      "eval_loss": 4.3514204025268555,
+      "eval_runtime": 834.4919,
+      "eval_samples_per_second": 0.624,
+      "eval_steps_per_second": 0.079,
+      "step": 316
+    },
+    {
+      "epoch": 1.0002531645569621,
+      "grad_norm": 12.405595779418945,
+      "learning_rate": 1.0126582278481012e-05,
+      "loss": 4.3858,
+      "step": 320
+    },
+    {
+      "epoch": 1.0008860759493672,
+      "grad_norm": 14.15708065032959,
+      "learning_rate": 1.0443037974683544e-05,
+      "loss": 4.3131,
+      "step": 330
+    },
+    {
+      "epoch": 1.0015189873417722,
+      "grad_norm": 12.798722267150879,
+      "learning_rate": 1.0759493670886076e-05,
+      "loss": 4.3399,
+      "step": 340
+    },
+    {
+      "epoch": 1.0021518987341773,
+      "grad_norm": 15.9908447265625,
+      "learning_rate": 1.1075949367088608e-05,
+      "loss": 4.2376,
+      "step": 350
+    },
+    {
+      "epoch": 1.0027848101265824,
+      "grad_norm": 12.808874130249023,
+      "learning_rate": 1.139240506329114e-05,
+      "loss": 4.1385,
+      "step": 360
+    },
+    {
+      "epoch": 1.0034177215189874,
+      "grad_norm": 15.553874969482422,
+      "learning_rate": 1.170886075949367e-05,
+      "loss": 4.202,
+      "step": 370
+    },
+    {
+      "epoch": 1.0040506329113925,
+      "grad_norm": 12.774052619934082,
+      "learning_rate": 1.2025316455696203e-05,
+      "loss": 4.0002,
+      "step": 380
+    },
+    {
+      "epoch": 1.0046835443037974,
+      "grad_norm": 15.000925064086914,
+      "learning_rate": 1.2341772151898735e-05,
+      "loss": 4.0083,
+      "step": 390
+    },
+    {
+      "epoch": 1.0053164556962024,
+      "grad_norm": 13.524576187133789,
+      "learning_rate": 1.2658227848101267e-05,
+      "loss": 4.086,
+      "step": 400
+    },
+    {
+      "epoch": 1.0059493670886075,
+      "grad_norm": 14.740311622619629,
+      "learning_rate": 1.2974683544303799e-05,
+      "loss": 4.0018,
+      "step": 410
+    },
+    {
+      "epoch": 1.0065822784810126,
+      "grad_norm": 13.781682968139648,
+      "learning_rate": 1.3291139240506329e-05,
+      "loss": 3.9788,
+      "step": 420
+    },
+    {
+      "epoch": 1.0072151898734176,
+      "grad_norm": 17.60359764099121,
+      "learning_rate": 1.3607594936708861e-05,
+      "loss": 3.7957,
+      "step": 430
+    },
+    {
+      "epoch": 1.0078481012658227,
+      "grad_norm": 15.860666275024414,
+      "learning_rate": 1.3924050632911393e-05,
+      "loss": 3.8316,
+      "step": 440
+    },
+    {
+      "epoch": 1.0084810126582278,
+      "grad_norm": 18.55211639404297,
+      "learning_rate": 1.4240506329113925e-05,
+      "loss": 3.7908,
+      "step": 450
+    },
+    {
+      "epoch": 1.0091139240506328,
+      "grad_norm": 12.959016799926758,
+      "learning_rate": 1.4556962025316457e-05,
+      "loss": 3.6121,
+      "step": 460
+    },
+    {
+      "epoch": 1.009746835443038,
+      "grad_norm": 14.276836395263672,
+      "learning_rate": 1.4873417721518987e-05,
+      "loss": 3.6771,
+      "step": 470
+    },
+    {
+      "epoch": 1.010379746835443,
+      "grad_norm": 15.222319602966309,
+      "learning_rate": 1.5189873417721521e-05,
+      "loss": 3.695,
+      "step": 480
+    },
+    {
+      "epoch": 1.011012658227848,
+      "grad_norm": 12.566011428833008,
+      "learning_rate": 1.550632911392405e-05,
+      "loss": 3.6071,
+      "step": 490
+    },
+    {
+      "epoch": 1.011645569620253,
+      "grad_norm": 13.210318565368652,
+      "learning_rate": 1.5822784810126583e-05,
+      "loss": 3.4652,
+      "step": 500
+    },
+    {
+      "epoch": 1.0122784810126582,
+      "grad_norm": 14.671469688415527,
+      "learning_rate": 1.6139240506329115e-05,
+      "loss": 3.457,
+      "step": 510
+    },
+    {
+      "epoch": 1.0129113924050632,
+      "grad_norm": 12.707144737243652,
+      "learning_rate": 1.6455696202531644e-05,
+      "loss": 3.5506,
+      "step": 520
+    },
+    {
+      "epoch": 1.0135443037974683,
+      "grad_norm": 15.84700870513916,
+      "learning_rate": 1.677215189873418e-05,
+      "loss": 3.433,
+      "step": 530
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 13.466840744018555,
+      "learning_rate": 1.7088607594936708e-05,
+      "loss": 3.4244,
+      "step": 540
+    },
+    {
+      "epoch": 1.0148101265822784,
+      "grad_norm": 12.980550765991211,
+      "learning_rate": 1.7405063291139243e-05,
+      "loss": 3.3659,
+      "step": 550
+    },
+    {
+      "epoch": 1.0154430379746835,
+      "grad_norm": 12.236871719360352,
+      "learning_rate": 1.7721518987341772e-05,
+      "loss": 3.3406,
+      "step": 560
+    },
+    {
+      "epoch": 1.0160759493670886,
+      "grad_norm": 14.126879692077637,
+      "learning_rate": 1.8037974683544304e-05,
+      "loss": 3.2676,
+      "step": 570
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 14.927568435668945,
+      "learning_rate": 1.8354430379746836e-05,
+      "loss": 3.1862,
+      "step": 580
+    },
+    {
+      "epoch": 1.0173417721518987,
+      "grad_norm": 14.192325592041016,
+      "learning_rate": 1.8670886075949368e-05,
+      "loss": 3.2144,
+      "step": 590
+    },
+    {
+      "epoch": 1.0179746835443038,
+      "grad_norm": 12.938292503356934,
+      "learning_rate": 1.89873417721519e-05,
+      "loss": 3.1199,
+      "step": 600
+    },
+    {
+      "epoch": 1.0186075949367088,
+      "grad_norm": 15.724618911743164,
+      "learning_rate": 1.9303797468354432e-05,
+      "loss": 3.1761,
+      "step": 610
+    },
+    {
+      "epoch": 1.019240506329114,
+      "grad_norm": 12.795076370239258,
+      "learning_rate": 1.962025316455696e-05,
+      "loss": 3.0823,
+      "step": 620
+    },
+    {
+      "epoch": 1.019873417721519,
+      "grad_norm": 16.50122833251953,
+      "learning_rate": 1.9936708860759496e-05,
+      "loss": 3.0795,
+      "step": 630
+    },
+    {
+      "epoch": 1.02,
+      "eval_accuracy": 0.581573896353167,
+      "eval_loss": 2.8514750003814697,
+      "eval_runtime": 833.5163,
+      "eval_samples_per_second": 0.625,
+      "eval_steps_per_second": 0.079,
+      "step": 632
+    },
+    {
+      "epoch": 2.0005063291139242,
+      "grad_norm": 16.488386154174805,
+      "learning_rate": 2.0253164556962025e-05,
+      "loss": 2.7661,
+      "step": 640
+    },
+    {
+      "epoch": 2.0011392405063293,
+      "grad_norm": 13.507458686828613,
+      "learning_rate": 2.056962025316456e-05,
+      "loss": 2.6532,
+      "step": 650
+    },
+    {
+      "epoch": 2.0017721518987344,
+      "grad_norm": 12.340897560119629,
+      "learning_rate": 2.088607594936709e-05,
+      "loss": 2.6855,
+      "step": 660
+    },
+    {
+      "epoch": 2.0024050632911394,
+      "grad_norm": 13.872140884399414,
+      "learning_rate": 2.120253164556962e-05,
+      "loss": 2.8458,
+      "step": 670
+    },
+    {
+      "epoch": 2.0030379746835445,
+      "grad_norm": 12.209596633911133,
+      "learning_rate": 2.1518987341772153e-05,
+      "loss": 2.6217,
+      "step": 680
+    },
+    {
+      "epoch": 2.0036708860759496,
+      "grad_norm": 12.713300704956055,
+      "learning_rate": 2.1835443037974685e-05,
+      "loss": 2.6449,
+      "step": 690
+    },
+    {
+      "epoch": 2.0043037974683546,
+      "grad_norm": 13.209062576293945,
+      "learning_rate": 2.2151898734177217e-05,
+      "loss": 2.6258,
+      "step": 700
+    },
+    {
+      "epoch": 2.0049367088607597,
+      "grad_norm": 13.877063751220703,
+      "learning_rate": 2.246835443037975e-05,
+      "loss": 2.5975,
+      "step": 710
+    },
+    {
+      "epoch": 2.0055696202531648,
+      "grad_norm": 15.146308898925781,
+      "learning_rate": 2.278481012658228e-05,
+      "loss": 2.4536,
+      "step": 720
+    },
+    {
+      "epoch": 2.00620253164557,
+      "grad_norm": 12.468453407287598,
+      "learning_rate": 2.3101265822784813e-05,
+      "loss": 2.4758,
+      "step": 730
+    },
+    {
+      "epoch": 2.006835443037975,
+      "grad_norm": 13.691295623779297,
+      "learning_rate": 2.341772151898734e-05,
+      "loss": 2.6211,
+      "step": 740
+    },
+    {
+      "epoch": 2.00746835443038,
+      "grad_norm": 13.543225288391113,
+      "learning_rate": 2.3734177215189873e-05,
+      "loss": 2.5074,
+      "step": 750
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 12.138303756713867,
+      "learning_rate": 2.4050632911392405e-05,
+      "loss": 2.2878,
+      "step": 760
+    },
+    {
+      "epoch": 2.0087341772151897,
+      "grad_norm": 12.926602363586426,
+      "learning_rate": 2.4367088607594937e-05,
+      "loss": 2.3154,
+      "step": 770
+    },
+    {
+      "epoch": 2.0093670886075947,
+      "grad_norm": 13.598923683166504,
+      "learning_rate": 2.468354430379747e-05,
+      "loss": 2.2751,
+      "step": 780
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.883562088012695,
+      "learning_rate": 2.5e-05,
+      "loss": 2.3078,
+      "step": 790
+    },
+    {
+      "epoch": 2.010632911392405,
+      "grad_norm": 13.546457290649414,
+      "learning_rate": 2.5316455696202533e-05,
+      "loss": 2.5038,
+      "step": 800
+    },
+    {
+      "epoch": 2.01126582278481,
+      "grad_norm": 15.554511070251465,
+      "learning_rate": 2.5632911392405062e-05,
+      "loss": 2.3539,
+      "step": 810
+    },
+    {
+      "epoch": 2.011898734177215,
+      "grad_norm": 12.446627616882324,
+      "learning_rate": 2.5949367088607597e-05,
+      "loss": 2.4136,
+      "step": 820
+    },
+    {
+      "epoch": 2.01253164556962,
+      "grad_norm": 17.281230926513672,
+      "learning_rate": 2.626582278481013e-05,
+      "loss": 2.2076,
+      "step": 830
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 11.104803085327148,
+      "learning_rate": 2.6582278481012658e-05,
+      "loss": 2.1716,
+      "step": 840
+    },
+    {
+      "epoch": 2.01379746835443,
+      "grad_norm": 15.020671844482422,
+      "learning_rate": 2.689873417721519e-05,
+      "loss": 2.3163,
+      "step": 850
+    },
+    {
+      "epoch": 2.0144303797468353,
+      "grad_norm": 12.115602493286133,
+      "learning_rate": 2.7215189873417722e-05,
+      "loss": 2.1664,
+      "step": 860
+    },
+    {
+      "epoch": 2.0150632911392403,
+      "grad_norm": 13.693445205688477,
+      "learning_rate": 2.7531645569620257e-05,
+      "loss": 2.1606,
+      "step": 870
+    },
+    {
+      "epoch": 2.0156962025316454,
+      "grad_norm": 10.924906730651855,
+      "learning_rate": 2.7848101265822786e-05,
+      "loss": 2.136,
+      "step": 880
+    },
+    {
+      "epoch": 2.0163291139240505,
+      "grad_norm": 11.92796802520752,
+      "learning_rate": 2.8164556962025318e-05,
+      "loss": 2.1733,
+      "step": 890
+    },
+    {
+      "epoch": 2.0169620253164555,
+      "grad_norm": 12.855241775512695,
+      "learning_rate": 2.848101265822785e-05,
+      "loss": 2.07,
+      "step": 900
+    },
+    {
+      "epoch": 2.0175949367088606,
+      "grad_norm": 14.710478782653809,
+      "learning_rate": 2.879746835443038e-05,
+      "loss": 2.1004,
+      "step": 910
+    },
+    {
+      "epoch": 2.0182278481012657,
+      "grad_norm": 14.612150192260742,
+      "learning_rate": 2.9113924050632914e-05,
+      "loss": 2.011,
+      "step": 920
+    },
+    {
+      "epoch": 2.0188607594936707,
+      "grad_norm": 15.704828262329102,
+      "learning_rate": 2.9430379746835446e-05,
+      "loss": 2.0263,
+      "step": 930
+    },
+    {
+      "epoch": 2.019493670886076,
+      "grad_norm": 12.45781421661377,
+      "learning_rate": 2.9746835443037974e-05,
+      "loss": 1.8438,
+      "step": 940
+    },
+    {
+      "epoch": 2.02,
+      "eval_accuracy": 0.7332053742802304,
+      "eval_loss": 1.7508126497268677,
+      "eval_runtime": 838.4982,
+      "eval_samples_per_second": 0.621,
+      "eval_steps_per_second": 0.079,
+      "step": 948
+    },
+    {
+      "epoch": 3.000126582278481,
+      "grad_norm": 11.997598648071289,
+      "learning_rate": 3.0063291139240506e-05,
+      "loss": 1.8218,
+      "step": 950
+    },
+    {
+      "epoch": 3.000759493670886,
+      "grad_norm": 10.052958488464355,
+      "learning_rate": 3.0379746835443042e-05,
+      "loss": 1.8534,
+      "step": 960
+    },
+    {
+      "epoch": 3.001392405063291,
+      "grad_norm": 12.40282154083252,
+      "learning_rate": 3.0696202531645574e-05,
+      "loss": 1.7654,
+      "step": 970
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 12.564291000366211,
+      "learning_rate": 3.10126582278481e-05,
+      "loss": 1.537,
+      "step": 980
+    },
+    {
+      "epoch": 3.002658227848101,
+      "grad_norm": 12.120141983032227,
+      "learning_rate": 3.132911392405064e-05,
+      "loss": 1.7112,
+      "step": 990
+    },
+    {
+      "epoch": 3.003291139240506,
+      "grad_norm": 9.203033447265625,
+      "learning_rate": 3.1645569620253167e-05,
+      "loss": 1.5344,
+      "step": 1000
+    },
+    {
+      "epoch": 3.0039240506329112,
+      "grad_norm": 10.776932716369629,
+      "learning_rate": 3.1962025316455695e-05,
+      "loss": 1.6746,
+      "step": 1010
+    },
+    {
+      "epoch": 3.0045569620253163,
+      "grad_norm": 13.07652759552002,
+      "learning_rate": 3.227848101265823e-05,
+      "loss": 1.7123,
+      "step": 1020
+    },
+    {
+      "epoch": 3.0051898734177214,
+      "grad_norm": 12.412099838256836,
+      "learning_rate": 3.2594936708860766e-05,
+      "loss": 1.6235,
+      "step": 1030
+    },
+    {
+      "epoch": 3.0058227848101264,
+      "grad_norm": 11.809402465820312,
+      "learning_rate": 3.291139240506329e-05,
+      "loss": 1.7162,
+      "step": 1040
+    },
+    {
+      "epoch": 3.0064556962025315,
+      "grad_norm": 10.014476776123047,
+      "learning_rate": 3.322784810126582e-05,
+      "loss": 1.5555,
+      "step": 1050
+    },
+    {
+      "epoch": 3.0070886075949366,
+      "grad_norm": 11.16119384765625,
+      "learning_rate": 3.354430379746836e-05,
+      "loss": 1.5755,
+      "step": 1060
+    },
+    {
+      "epoch": 3.0077215189873416,
+      "grad_norm": 11.865368843078613,
+      "learning_rate": 3.386075949367089e-05,
+      "loss": 1.4931,
+      "step": 1070
+    },
+    {
+      "epoch": 3.0083544303797467,
+      "grad_norm": 10.88665771484375,
+      "learning_rate": 3.4177215189873416e-05,
+      "loss": 1.4321,
+      "step": 1080
+    },
+    {
+      "epoch": 3.0089873417721518,
+      "grad_norm": 15.595149993896484,
+      "learning_rate": 3.449367088607595e-05,
+      "loss": 1.7013,
+      "step": 1090
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 12.852721214294434,
+      "learning_rate": 3.4810126582278487e-05,
+      "loss": 1.4012,
+      "step": 1100
+    },
+    {
+      "epoch": 3.010253164556962,
+      "grad_norm": 11.31406307220459,
+      "learning_rate": 3.5126582278481015e-05,
+      "loss": 1.3529,
+      "step": 1110
+    },
+    {
+      "epoch": 3.010886075949367,
+      "grad_norm": 13.643885612487793,
+      "learning_rate": 3.5443037974683544e-05,
+      "loss": 1.6591,
+      "step": 1120
+    },
+    {
+      "epoch": 3.011518987341772,
+      "grad_norm": 8.9163818359375,
+      "learning_rate": 3.575949367088608e-05,
+      "loss": 1.5645,
+      "step": 1130
+    },
+    {
+      "epoch": 3.012151898734177,
+      "grad_norm": 11.10364818572998,
+      "learning_rate": 3.607594936708861e-05,
+      "loss": 1.573,
+      "step": 1140
+    },
+    {
+      "epoch": 3.012784810126582,
+      "grad_norm": 12.710495948791504,
+      "learning_rate": 3.639240506329114e-05,
+      "loss": 1.4811,
+      "step": 1150
+    },
+    {
+      "epoch": 3.0134177215189872,
+      "grad_norm": 9.857568740844727,
+      "learning_rate": 3.670886075949367e-05,
+      "loss": 1.3618,
+      "step": 1160
+    },
+    {
+      "epoch": 3.0140506329113923,
+      "grad_norm": 12.947936058044434,
+      "learning_rate": 3.70253164556962e-05,
+      "loss": 1.4507,
+      "step": 1170
+    },
+    {
+      "epoch": 3.0146835443037974,
+      "grad_norm": 7.388261795043945,
+      "learning_rate": 3.7341772151898736e-05,
+      "loss": 1.368,
+      "step": 1180
+    },
+    {
+      "epoch": 3.0153164556962024,
+      "grad_norm": 12.413993835449219,
+      "learning_rate": 3.765822784810127e-05,
+      "loss": 1.2501,
+      "step": 1190
+    },
+    {
+      "epoch": 3.0159493670886075,
+      "grad_norm": 9.680545806884766,
+      "learning_rate": 3.79746835443038e-05,
+      "loss": 1.3826,
+      "step": 1200
+    },
+    {
+      "epoch": 3.0165822784810126,
+      "grad_norm": 10.841100692749023,
+      "learning_rate": 3.829113924050633e-05,
+      "loss": 1.3361,
+      "step": 1210
+    },
+    {
+      "epoch": 3.0172151898734176,
+      "grad_norm": 10.57774829864502,
+      "learning_rate": 3.8607594936708864e-05,
+      "loss": 1.4978,
+      "step": 1220
+    },
+    {
+      "epoch": 3.0178481012658227,
+      "grad_norm": 20.876358032226562,
+      "learning_rate": 3.89240506329114e-05,
+      "loss": 1.4616,
+      "step": 1230
+    },
+    {
+      "epoch": 3.0184810126582278,
+      "grad_norm": 13.871617317199707,
+      "learning_rate": 3.924050632911392e-05,
+      "loss": 1.2869,
+      "step": 1240
+    },
+    {
+      "epoch": 3.019113924050633,
+      "grad_norm": 10.208392143249512,
+      "learning_rate": 3.9556962025316456e-05,
+      "loss": 1.4173,
+      "step": 1250
+    },
+    {
+      "epoch": 3.019746835443038,
+      "grad_norm": 13.292877197265625,
+      "learning_rate": 3.987341772151899e-05,
+      "loss": 1.1451,
+      "step": 1260
+    },
+    {
+      "epoch": 3.02,
+      "eval_accuracy": 0.7389635316698656,
+      "eval_loss": 1.146372675895691,
+      "eval_runtime": 855.4763,
+      "eval_samples_per_second": 0.609,
+      "eval_steps_per_second": 0.077,
+      "step": 1264
+    },
+    {
+      "epoch": 4.000379746835443,
+      "grad_norm": 12.077232360839844,
+      "learning_rate": 4.018987341772152e-05,
+      "loss": 1.186,
+      "step": 1270
+    },
+    {
+      "epoch": 4.0010126582278485,
+      "grad_norm": 11.689604759216309,
+      "learning_rate": 4.050632911392405e-05,
+      "loss": 1.2715,
+      "step": 1280
+    },
+    {
+      "epoch": 4.001645569620253,
+      "grad_norm": 10.000913619995117,
+      "learning_rate": 4.0822784810126584e-05,
+      "loss": 1.1501,
+      "step": 1290
+    },
+    {
+      "epoch": 4.002278481012659,
+      "grad_norm": 11.559700965881348,
+      "learning_rate": 4.113924050632912e-05,
+      "loss": 1.3049,
+      "step": 1300
+    },
+    {
+      "epoch": 4.002911392405063,
+      "grad_norm": 9.99675178527832,
+      "learning_rate": 4.145569620253165e-05,
+      "loss": 1.1664,
+      "step": 1310
+    },
+    {
+      "epoch": 4.003544303797469,
+      "grad_norm": 10.537801742553711,
+      "learning_rate": 4.177215189873418e-05,
+      "loss": 1.0544,
+      "step": 1320
+    },
+    {
+      "epoch": 4.004177215189873,
+      "grad_norm": 7.792250633239746,
+      "learning_rate": 4.208860759493671e-05,
+      "loss": 1.185,
+      "step": 1330
+    },
+    {
+      "epoch": 4.004810126582279,
+      "grad_norm": 7.729872226715088,
+      "learning_rate": 4.240506329113924e-05,
+      "loss": 0.958,
+      "step": 1340
+    },
+    {
+      "epoch": 4.0054430379746835,
+      "grad_norm": 8.17205810546875,
+      "learning_rate": 4.2721518987341776e-05,
+      "loss": 1.1805,
+      "step": 1350
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 8.281828880310059,
+      "learning_rate": 4.3037974683544305e-05,
+      "loss": 1.0389,
+      "step": 1360
+    },
+    {
+      "epoch": 4.006708860759494,
+      "grad_norm": 10.64874267578125,
+      "learning_rate": 4.3354430379746834e-05,
+      "loss": 1.1096,
+      "step": 1370
+    },
+    {
+      "epoch": 4.007341772151899,
+      "grad_norm": 10.857294082641602,
+      "learning_rate": 4.367088607594937e-05,
+      "loss": 1.0934,
+      "step": 1380
+    },
+    {
+      "epoch": 4.007974683544304,
+      "grad_norm": 10.056556701660156,
+      "learning_rate": 4.3987341772151904e-05,
+      "loss": 1.0382,
+      "step": 1390
+    },
+    {
+      "epoch": 4.008607594936709,
+      "grad_norm": 5.310425281524658,
+      "learning_rate": 4.430379746835443e-05,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 4.009240506329114,
+      "grad_norm": 9.902403831481934,
+      "learning_rate": 4.462025316455696e-05,
+      "loss": 1.1619,
+      "step": 1410
+    },
+    {
+      "epoch": 4.009873417721519,
+      "grad_norm": 10.88439655303955,
+      "learning_rate": 4.49367088607595e-05,
+      "loss": 1.1363,
+      "step": 1420
+    },
+    {
+      "epoch": 4.010506329113924,
+      "grad_norm": 8.662186622619629,
+      "learning_rate": 4.525316455696203e-05,
+      "loss": 1.0104,
+      "step": 1430
+    },
+    {
+      "epoch": 4.0111392405063295,
+      "grad_norm": 7.63654899597168,
+      "learning_rate": 4.556962025316456e-05,
+      "loss": 0.9476,
+      "step": 1440
+    },
+    {
+      "epoch": 4.011772151898734,
+      "grad_norm": 15.624666213989258,
+      "learning_rate": 4.588607594936709e-05,
+      "loss": 1.0386,
+      "step": 1450
+    },
+    {
+      "epoch": 4.01240506329114,
+      "grad_norm": 10.169722557067871,
+      "learning_rate": 4.6202531645569625e-05,
+      "loss": 1.0147,
+      "step": 1460
+    },
+    {
+      "epoch": 4.013037974683544,
+      "grad_norm": 10.036338806152344,
+      "learning_rate": 4.6518987341772154e-05,
+      "loss": 1.0159,
+      "step": 1470
+    },
+    {
+      "epoch": 4.01367088607595,
+      "grad_norm": 10.780523300170898,
+      "learning_rate": 4.683544303797468e-05,
+      "loss": 1.0246,
+      "step": 1480
+    },
+    {
+      "epoch": 4.014303797468354,
+      "grad_norm": 12.800423622131348,
+      "learning_rate": 4.715189873417722e-05,
+      "loss": 0.9578,
+      "step": 1490
+    },
+    {
+      "epoch": 4.01493670886076,
+      "grad_norm": 11.1471586227417,
+      "learning_rate": 4.7468354430379746e-05,
+      "loss": 0.9799,
+      "step": 1500
+    },
+    {
+      "epoch": 4.0155696202531646,
+      "grad_norm": 5.804954528808594,
+      "learning_rate": 4.778481012658228e-05,
+      "loss": 0.9171,
+      "step": 1510
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 14.239872932434082,
+      "learning_rate": 4.810126582278481e-05,
+      "loss": 1.1415,
+      "step": 1520
+    },
+    {
+      "epoch": 4.016835443037975,
+      "grad_norm": 6.897063255310059,
+      "learning_rate": 4.8417721518987346e-05,
+      "loss": 1.0636,
+      "step": 1530
+    },
+    {
+      "epoch": 4.017468354430379,
+      "grad_norm": 12.96029281616211,
+      "learning_rate": 4.8734177215189874e-05,
+      "loss": 0.9399,
+      "step": 1540
+    },
+    {
+      "epoch": 4.018101265822785,
+      "grad_norm": 17.250343322753906,
+      "learning_rate": 4.905063291139241e-05,
+      "loss": 0.8518,
+      "step": 1550
+    },
+    {
+      "epoch": 4.018734177215189,
+      "grad_norm": 10.851552963256836,
+      "learning_rate": 4.936708860759494e-05,
+      "loss": 1.091,
+      "step": 1560
+    },
+    {
+      "epoch": 4.019367088607595,
+      "grad_norm": 16.084505081176758,
+      "learning_rate": 4.968354430379747e-05,
+      "loss": 1.0378,
+      "step": 1570
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 11.509261131286621,
+      "learning_rate": 5e-05,
+      "loss": 1.0637,
+      "step": 1580
+    },
+    {
+      "epoch": 4.02,
+      "eval_accuracy": 0.7773512476007678,
+      "eval_loss": 0.7994989156723022,
+      "eval_runtime": 871.4001,
+      "eval_samples_per_second": 0.598,
+      "eval_steps_per_second": 0.076,
+      "step": 1580
+    },
+    {
+      "epoch": 5.000632911392405,
+      "grad_norm": 8.971628189086914,
+      "learning_rate": 4.99648382559775e-05,
+      "loss": 0.8316,
+      "step": 1590
+    },
+    {
+      "epoch": 5.00126582278481,
+      "grad_norm": 13.378620147705078,
+      "learning_rate": 4.9929676511955e-05,
+      "loss": 0.8914,
+      "step": 1600
+    },
+    {
+      "epoch": 5.001898734177215,
+      "grad_norm": 9.115056037902832,
+      "learning_rate": 4.989451476793249e-05,
+      "loss": 0.9715,
+      "step": 1610
+    },
+    {
+      "epoch": 5.00253164556962,
+      "grad_norm": 20.863187789916992,
+      "learning_rate": 4.985935302390999e-05,
+      "loss": 0.7779,
+      "step": 1620
+    },
+    {
+      "epoch": 5.003164556962025,
+      "grad_norm": 10.088607788085938,
+      "learning_rate": 4.982419127988748e-05,
+      "loss": 0.8775,
+      "step": 1630
+    },
+    {
+      "epoch": 5.00379746835443,
+      "grad_norm": 6.800666332244873,
+      "learning_rate": 4.9789029535864986e-05,
+      "loss": 0.7337,
+      "step": 1640
+    },
+    {
+      "epoch": 5.004430379746835,
+      "grad_norm": 12.704118728637695,
+      "learning_rate": 4.975386779184248e-05,
+      "loss": 0.7668,
+      "step": 1650
+    },
+    {
+      "epoch": 5.0050632911392405,
+      "grad_norm": 7.471581935882568,
+      "learning_rate": 4.9718706047819975e-05,
+      "loss": 0.7418,
+      "step": 1660
+    },
+    {
+      "epoch": 5.005696202531645,
+      "grad_norm": 11.495733261108398,
+      "learning_rate": 4.968354430379747e-05,
+      "loss": 0.8782,
+      "step": 1670
+    },
+    {
+      "epoch": 5.006329113924051,
+      "grad_norm": 7.495334148406982,
+      "learning_rate": 4.964838255977497e-05,
+      "loss": 0.7234,
+      "step": 1680
+    },
+    {
+      "epoch": 5.006962025316455,
+      "grad_norm": 9.78715991973877,
+      "learning_rate": 4.9613220815752464e-05,
+      "loss": 0.8924,
+      "step": 1690
+    },
+    {
+      "epoch": 5.007594936708861,
+      "grad_norm": 6.057085990905762,
+      "learning_rate": 4.957805907172996e-05,
+      "loss": 0.6553,
+      "step": 1700
+    },
+    {
+      "epoch": 5.008227848101265,
+      "grad_norm": 7.595339298248291,
+      "learning_rate": 4.9542897327707454e-05,
+      "loss": 0.9012,
+      "step": 1710
+    },
+    {
+      "epoch": 5.008860759493671,
+      "grad_norm": 11.763632774353027,
+      "learning_rate": 4.950773558368496e-05,
+      "loss": 0.7808,
+      "step": 1720
+    },
+    {
+      "epoch": 5.0094936708860756,
+      "grad_norm": 9.754287719726562,
+      "learning_rate": 4.947257383966245e-05,
+      "loss": 0.8886,
+      "step": 1730
+    },
+    {
+      "epoch": 5.010126582278481,
+      "grad_norm": 7.873773097991943,
+      "learning_rate": 4.943741209563995e-05,
+      "loss": 0.6798,
+      "step": 1740
+    },
+    {
+      "epoch": 5.010759493670886,
+      "grad_norm": 9.378247261047363,
+      "learning_rate": 4.940225035161744e-05,
+      "loss": 0.763,
+      "step": 1750
+    },
+    {
+      "epoch": 5.011392405063291,
+      "grad_norm": 13.096108436584473,
+      "learning_rate": 4.936708860759494e-05,
+      "loss": 0.7834,
+      "step": 1760
+    },
+    {
+      "epoch": 5.012025316455696,
+      "grad_norm": 5.072068691253662,
+      "learning_rate": 4.933192686357244e-05,
+      "loss": 0.6282,
+      "step": 1770
+    },
+    {
+      "epoch": 5.012658227848101,
+      "grad_norm": 14.543283462524414,
+      "learning_rate": 4.929676511954993e-05,
+      "loss": 0.6679,
+      "step": 1780
+    },
+    {
+      "epoch": 5.013291139240506,
+      "grad_norm": 11.026363372802734,
+      "learning_rate": 4.9261603375527427e-05,
+      "loss": 0.7401,
+      "step": 1790
+    },
+    {
+      "epoch": 5.0139240506329115,
+      "grad_norm": 5.937496662139893,
+      "learning_rate": 4.9226441631504925e-05,
+      "loss": 0.7408,
+      "step": 1800
+    },
+    {
+      "epoch": 5.014556962025316,
+      "grad_norm": 9.093450546264648,
+      "learning_rate": 4.919127988748242e-05,
+      "loss": 0.9351,
+      "step": 1810
+    },
+    {
+      "epoch": 5.015189873417722,
+      "grad_norm": 6.913345813751221,
+      "learning_rate": 4.9156118143459915e-05,
+      "loss": 0.6614,
+      "step": 1820
+    },
+    {
+      "epoch": 5.015822784810126,
+      "grad_norm": 5.915560245513916,
+      "learning_rate": 4.912095639943741e-05,
+      "loss": 0.6584,
+      "step": 1830
+    },
+    {
+      "epoch": 5.016455696202532,
+      "grad_norm": 9.257479667663574,
+      "learning_rate": 4.908579465541491e-05,
+      "loss": 0.6744,
+      "step": 1840
+    },
+    {
+      "epoch": 5.017088607594936,
+      "grad_norm": 12.13564395904541,
+      "learning_rate": 4.905063291139241e-05,
+      "loss": 0.7035,
+      "step": 1850
+    },
+    {
+      "epoch": 5.017721518987342,
+      "grad_norm": 6.930849075317383,
+      "learning_rate": 4.90154711673699e-05,
+      "loss": 0.7628,
+      "step": 1860
+    },
+    {
+      "epoch": 5.0183544303797465,
+      "grad_norm": 13.400097846984863,
+      "learning_rate": 4.89803094233474e-05,
+      "loss": 0.6906,
+      "step": 1870
+    },
+    {
+      "epoch": 5.018987341772152,
+      "grad_norm": 3.020031452178955,
+      "learning_rate": 4.89451476793249e-05,
+      "loss": 0.8423,
+      "step": 1880
+    },
+    {
+      "epoch": 5.019620253164557,
+      "grad_norm": 11.807697296142578,
+      "learning_rate": 4.8909985935302396e-05,
+      "loss": 0.7795,
+      "step": 1890
+    },
+    {
+      "epoch": 5.02,
+      "eval_accuracy": 0.8829174664107485,
+      "eval_loss": 0.4938121438026428,
+      "eval_runtime": 888.2928,
+      "eval_samples_per_second": 0.587,
+      "eval_steps_per_second": 0.074,
+      "step": 1896
+    },
+    {
+      "epoch": 6.000253164556962,
+      "grad_norm": 1.9311364889144897,
+      "learning_rate": 4.887482419127989e-05,
+      "loss": 0.5037,
+      "step": 1900
+    },
+    {
+      "epoch": 6.000886075949367,
+      "grad_norm": 2.175624370574951,
+      "learning_rate": 4.8839662447257386e-05,
+      "loss": 0.5447,
+      "step": 1910
+    },
+    {
+      "epoch": 6.001518987341772,
+      "grad_norm": 6.488412380218506,
+      "learning_rate": 4.8804500703234885e-05,
+      "loss": 0.4324,
+      "step": 1920
+    },
+    {
+      "epoch": 6.002151898734177,
+      "grad_norm": 7.727587699890137,
+      "learning_rate": 4.876933895921238e-05,
+      "loss": 0.7189,
+      "step": 1930
+    },
+    {
+      "epoch": 6.002784810126582,
+      "grad_norm": 9.56769847869873,
+      "learning_rate": 4.8734177215189874e-05,
+      "loss": 0.4081,
+      "step": 1940
+    },
+    {
+      "epoch": 6.0034177215189874,
+      "grad_norm": 4.022727012634277,
+      "learning_rate": 4.869901547116737e-05,
+      "loss": 0.6191,
+      "step": 1950
+    },
+    {
+      "epoch": 6.004050632911392,
+      "grad_norm": 4.674831867218018,
+      "learning_rate": 4.866385372714487e-05,
+      "loss": 0.523,
+      "step": 1960
+    },
+    {
+      "epoch": 6.004683544303798,
+      "grad_norm": 9.083635330200195,
+      "learning_rate": 4.862869198312236e-05,
+      "loss": 0.5281,
+      "step": 1970
+    },
+    {
+      "epoch": 6.005316455696202,
+      "grad_norm": 11.267608642578125,
+      "learning_rate": 4.859353023909986e-05,
+      "loss": 0.3717,
+      "step": 1980
+    },
+    {
+      "epoch": 6.005949367088608,
+      "grad_norm": 4.938966751098633,
+      "learning_rate": 4.855836849507735e-05,
+      "loss": 0.4737,
+      "step": 1990
+    },
+    {
+      "epoch": 6.006582278481012,
+      "grad_norm": 7.700178623199463,
+      "learning_rate": 4.852320675105486e-05,
+      "loss": 0.507,
+      "step": 2000
+    },
+    {
+      "epoch": 6.007215189873418,
+      "grad_norm": 13.916314125061035,
+      "learning_rate": 4.848804500703235e-05,
+      "loss": 0.4627,
+      "step": 2010
+    },
+    {
+      "epoch": 6.0078481012658225,
+      "grad_norm": 9.637331008911133,
+      "learning_rate": 4.845288326300985e-05,
+      "loss": 0.509,
+      "step": 2020
+    },
+    {
+      "epoch": 6.008481012658228,
+      "grad_norm": 6.41615629196167,
+      "learning_rate": 4.8417721518987346e-05,
+      "loss": 0.4817,
+      "step": 2030
+    },
+    {
+      "epoch": 6.009113924050633,
+      "grad_norm": 4.391960144042969,
+      "learning_rate": 4.8382559774964844e-05,
+      "loss": 0.6306,
+      "step": 2040
+    },
+    {
+      "epoch": 6.009746835443038,
+      "grad_norm": 3.4830005168914795,
+      "learning_rate": 4.8347398030942336e-05,
+      "loss": 0.6192,
+      "step": 2050
+    },
+    {
+      "epoch": 6.010379746835443,
+      "grad_norm": 8.927491188049316,
+      "learning_rate": 4.8312236286919834e-05,
+      "loss": 0.3656,
+      "step": 2060
+    },
+    {
+      "epoch": 6.011012658227848,
+      "grad_norm": 6.872843265533447,
+      "learning_rate": 4.827707454289733e-05,
+      "loss": 0.4649,
+      "step": 2070
+    },
+    {
+      "epoch": 6.011645569620253,
+      "grad_norm": 5.7255682945251465,
+      "learning_rate": 4.824191279887483e-05,
+      "loss": 0.5626,
+      "step": 2080
+    },
+    {
+      "epoch": 6.012278481012658,
+      "grad_norm": 12.707902908325195,
+      "learning_rate": 4.820675105485232e-05,
+      "loss": 0.7542,
+      "step": 2090
+    },
+    {
+      "epoch": 6.012911392405063,
+      "grad_norm": 8.386948585510254,
+      "learning_rate": 4.817158931082982e-05,
+      "loss": 0.5574,
+      "step": 2100
+    },
+    {
+      "epoch": 6.0135443037974685,
+      "grad_norm": 5.364466667175293,
+      "learning_rate": 4.813642756680732e-05,
+      "loss": 0.5088,
+      "step": 2110
+    },
+    {
+      "epoch": 6.014177215189873,
+      "grad_norm": 7.0593791007995605,
+      "learning_rate": 4.810126582278481e-05,
+      "loss": 0.4507,
+      "step": 2120
+    },
+    {
+      "epoch": 6.014810126582279,
+      "grad_norm": 4.73020076751709,
+      "learning_rate": 4.806610407876231e-05,
+      "loss": 0.4129,
+      "step": 2130
+    },
+    {
+      "epoch": 6.015443037974683,
+      "grad_norm": 9.857553482055664,
+      "learning_rate": 4.80309423347398e-05,
+      "loss": 0.6213,
+      "step": 2140
+    },
+    {
+      "epoch": 6.016075949367089,
+      "grad_norm": 3.092639446258545,
+      "learning_rate": 4.7995780590717305e-05,
+      "loss": 0.465,
+      "step": 2150
+    },
+    {
+      "epoch": 6.016708860759493,
+      "grad_norm": 10.670005798339844,
+      "learning_rate": 4.79606188466948e-05,
+      "loss": 0.5071,
+      "step": 2160
+    },
+    {
+      "epoch": 6.017341772151899,
+      "grad_norm": 10.013630867004395,
+      "learning_rate": 4.7925457102672295e-05,
+      "loss": 0.4878,
+      "step": 2170
+    },
+    {
+      "epoch": 6.0179746835443035,
+      "grad_norm": 5.06149959564209,
+      "learning_rate": 4.789029535864979e-05,
+      "loss": 0.4294,
+      "step": 2180
+    },
+    {
+      "epoch": 6.018607594936709,
+      "grad_norm": 6.84969425201416,
+      "learning_rate": 4.785513361462729e-05,
+      "loss": 0.4591,
+      "step": 2190
+    },
+    {
+      "epoch": 6.019240506329114,
+      "grad_norm": 16.97979736328125,
+      "learning_rate": 4.7819971870604783e-05,
+      "loss": 0.3969,
+      "step": 2200
+    },
+    {
+      "epoch": 6.019873417721519,
+      "grad_norm": 4.75033712387085,
+      "learning_rate": 4.778481012658228e-05,
+      "loss": 0.4484,
+      "step": 2210
+    },
+    {
+      "epoch": 6.02,
+      "eval_accuracy": 0.8829174664107485,
+      "eval_loss": 0.38333675265312195,
+      "eval_runtime": 852.5332,
+      "eval_samples_per_second": 0.611,
+      "eval_steps_per_second": 0.077,
+      "step": 2212
+    },
+    {
+      "epoch": 7.000506329113924,
+      "grad_norm": 14.740522384643555,
+      "learning_rate": 4.774964838255977e-05,
+      "loss": 0.4139,
+      "step": 2220
+    },
+    {
+      "epoch": 7.001139240506329,
+      "grad_norm": 2.6119625568389893,
+      "learning_rate": 4.771448663853728e-05,
+      "loss": 0.3493,
+      "step": 2230
+    },
+    {
+      "epoch": 7.001772151898734,
+      "grad_norm": 1.260722279548645,
+      "learning_rate": 4.767932489451477e-05,
+      "loss": 0.2562,
+      "step": 2240
+    },
+    {
+      "epoch": 7.002405063291139,
+      "grad_norm": 2.5697829723358154,
+      "learning_rate": 4.764416315049227e-05,
+      "loss": 0.3227,
+      "step": 2250
+    },
+    {
+      "epoch": 7.0030379746835445,
+      "grad_norm": 5.382594585418701,
+      "learning_rate": 4.760900140646976e-05,
+      "loss": 0.42,
+      "step": 2260
+    },
+    {
+      "epoch": 7.003670886075949,
+      "grad_norm": 3.63082218170166,
+      "learning_rate": 4.757383966244726e-05,
+      "loss": 0.3715,
+      "step": 2270
+    },
+    {
+      "epoch": 7.004303797468355,
+      "grad_norm": 4.501291751861572,
+      "learning_rate": 4.7538677918424756e-05,
+      "loss": 0.2527,
+      "step": 2280
+    },
+    {
+      "epoch": 7.004936708860759,
+      "grad_norm": 6.318272590637207,
+      "learning_rate": 4.7503516174402255e-05,
+      "loss": 0.2922,
+      "step": 2290
+    },
+    {
+      "epoch": 7.005569620253165,
+      "grad_norm": 15.981465339660645,
+      "learning_rate": 4.7468354430379746e-05,
+      "loss": 0.4605,
+      "step": 2300
+    },
+    {
+      "epoch": 7.006202531645569,
+      "grad_norm": 7.8365020751953125,
+      "learning_rate": 4.7433192686357245e-05,
+      "loss": 0.4728,
+      "step": 2310
+    },
+    {
+      "epoch": 7.006835443037975,
+      "grad_norm": 6.1255388259887695,
+      "learning_rate": 4.739803094233474e-05,
+      "loss": 0.3578,
+      "step": 2320
+    },
+    {
+      "epoch": 7.0074683544303795,
+      "grad_norm": 12.867307662963867,
+      "learning_rate": 4.7362869198312235e-05,
+      "loss": 0.3075,
+      "step": 2330
+    },
+    {
+      "epoch": 7.008101265822785,
+      "grad_norm": 3.2056193351745605,
+      "learning_rate": 4.732770745428973e-05,
+      "loss": 0.4281,
+      "step": 2340
+    },
+    {
+      "epoch": 7.00873417721519,
+      "grad_norm": 14.265457153320312,
+      "learning_rate": 4.729254571026723e-05,
+      "loss": 0.584,
+      "step": 2350
+    },
+    {
+      "epoch": 7.009367088607595,
+      "grad_norm": 5.145318984985352,
+      "learning_rate": 4.725738396624473e-05,
+      "loss": 0.3264,
+      "step": 2360
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 12.106173515319824,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 0.4977,
+      "step": 2370
+    },
+    {
+      "epoch": 7.010632911392405,
+      "grad_norm": 2.557343006134033,
+      "learning_rate": 4.718706047819972e-05,
+      "loss": 0.2676,
+      "step": 2380
+    },
+    {
+      "epoch": 7.01126582278481,
+      "grad_norm": 9.406204223632812,
+      "learning_rate": 4.715189873417722e-05,
+      "loss": 0.3632,
+      "step": 2390
+    },
+    {
+      "epoch": 7.011898734177215,
+      "grad_norm": 0.7909258604049683,
+      "learning_rate": 4.7116736990154716e-05,
+      "loss": 0.2164,
+      "step": 2400
+    },
+    {
+      "epoch": 7.01253164556962,
+      "grad_norm": 0.8705269694328308,
+      "learning_rate": 4.708157524613221e-05,
+      "loss": 0.3072,
+      "step": 2410
+    },
+    {
+      "epoch": 7.013164556962026,
+      "grad_norm": 3.603971481323242,
+      "learning_rate": 4.704641350210971e-05,
+      "loss": 0.2117,
+      "step": 2420
+    },
+    {
+      "epoch": 7.01379746835443,
+      "grad_norm": 7.447606563568115,
+      "learning_rate": 4.7011251758087204e-05,
+      "loss": 0.32,
+      "step": 2430
+    },
+    {
+      "epoch": 7.014430379746836,
+      "grad_norm": 7.379356384277344,
+      "learning_rate": 4.69760900140647e-05,
+      "loss": 0.3718,
+      "step": 2440
+    },
+    {
+      "epoch": 7.01506329113924,
+      "grad_norm": 17.737926483154297,
+      "learning_rate": 4.6940928270042194e-05,
+      "loss": 0.4529,
+      "step": 2450
+    },
+    {
+      "epoch": 7.015696202531646,
+      "grad_norm": 3.0855658054351807,
+      "learning_rate": 4.690576652601969e-05,
+      "loss": 0.2227,
+      "step": 2460
+    },
+    {
+      "epoch": 7.0163291139240505,
+      "grad_norm": 14.56204605102539,
+      "learning_rate": 4.687060478199719e-05,
+      "loss": 0.3014,
+      "step": 2470
+    },
+    {
+      "epoch": 7.016962025316456,
+      "grad_norm": 3.587031841278076,
+      "learning_rate": 4.683544303797468e-05,
+      "loss": 0.3073,
+      "step": 2480
+    },
+    {
+      "epoch": 7.017594936708861,
+      "grad_norm": 3.022749423980713,
+      "learning_rate": 4.680028129395218e-05,
+      "loss": 0.2003,
+      "step": 2490
+    },
+    {
+      "epoch": 7.018227848101266,
+      "grad_norm": 8.433323860168457,
+      "learning_rate": 4.676511954992968e-05,
+      "loss": 0.3411,
+      "step": 2500
+    },
+    {
+      "epoch": 7.018860759493671,
+      "grad_norm": 4.128199100494385,
+      "learning_rate": 4.672995780590718e-05,
+      "loss": 0.3518,
+      "step": 2510
+    },
+    {
+      "epoch": 7.019493670886076,
+      "grad_norm": 11.647120475769043,
+      "learning_rate": 4.669479606188467e-05,
+      "loss": 0.2162,
+      "step": 2520
+    },
+    {
+      "epoch": 7.02,
+      "eval_accuracy": 0.9155470249520153,
+      "eval_loss": 0.25118356943130493,
+      "eval_runtime": 859.0348,
+      "eval_samples_per_second": 0.606,
+      "eval_steps_per_second": 0.077,
+      "step": 2528
+    },
+    {
+      "epoch": 8.000126582278481,
+      "grad_norm": 21.305402755737305,
+      "learning_rate": 4.665963431786217e-05,
+      "loss": 0.3546,
+      "step": 2530
+    },
+    {
+      "epoch": 8.000759493670886,
+      "grad_norm": 9.045007705688477,
+      "learning_rate": 4.6624472573839666e-05,
+      "loss": 0.2433,
+      "step": 2540
+    },
+    {
+      "epoch": 8.00139240506329,
+      "grad_norm": 6.032439708709717,
+      "learning_rate": 4.6589310829817164e-05,
+      "loss": 0.2786,
+      "step": 2550
+    },
+    {
+      "epoch": 8.002025316455697,
+      "grad_norm": 5.416832447052002,
+      "learning_rate": 4.6554149085794655e-05,
+      "loss": 0.2595,
+      "step": 2560
+    },
+    {
+      "epoch": 8.002658227848102,
+      "grad_norm": 2.8014585971832275,
+      "learning_rate": 4.6518987341772154e-05,
+      "loss": 0.301,
+      "step": 2570
+    },
+    {
+      "epoch": 8.003291139240506,
+      "grad_norm": 1.782472014427185,
+      "learning_rate": 4.648382559774965e-05,
+      "loss": 0.1689,
+      "step": 2580
+    },
+    {
+      "epoch": 8.00392405063291,
+      "grad_norm": 2.679903507232666,
+      "learning_rate": 4.644866385372715e-05,
+      "loss": 0.2884,
+      "step": 2590
+    },
+    {
+      "epoch": 8.004556962025317,
+      "grad_norm": 2.2325003147125244,
+      "learning_rate": 4.641350210970464e-05,
+      "loss": 0.2047,
+      "step": 2600
+    },
+    {
+      "epoch": 8.005189873417722,
+      "grad_norm": 9.163021087646484,
+      "learning_rate": 4.637834036568214e-05,
+      "loss": 0.1786,
+      "step": 2610
+    },
+    {
+      "epoch": 8.005822784810126,
+      "grad_norm": 3.2019553184509277,
+      "learning_rate": 4.634317862165964e-05,
+      "loss": 0.2159,
+      "step": 2620
+    },
+    {
+      "epoch": 8.006455696202531,
+      "grad_norm": 4.9009904861450195,
+      "learning_rate": 4.630801687763714e-05,
+      "loss": 0.249,
+      "step": 2630
+    },
+    {
+      "epoch": 8.007088607594937,
+      "grad_norm": 4.865252494812012,
+      "learning_rate": 4.627285513361463e-05,
+      "loss": 0.3338,
+      "step": 2640
+    },
+    {
+      "epoch": 8.007721518987342,
+      "grad_norm": 1.178444504737854,
+      "learning_rate": 4.623769338959213e-05,
+      "loss": 0.2716,
+      "step": 2650
+    },
+    {
+      "epoch": 8.008354430379747,
+      "grad_norm": 2.2412354946136475,
+      "learning_rate": 4.6202531645569625e-05,
+      "loss": 0.1516,
+      "step": 2660
+    },
+    {
+      "epoch": 8.008987341772151,
+      "grad_norm": 11.467605590820312,
+      "learning_rate": 4.616736990154712e-05,
+      "loss": 0.3324,
+      "step": 2670
+    },
+    {
+      "epoch": 8.009620253164558,
+      "grad_norm": 4.130499362945557,
+      "learning_rate": 4.6132208157524615e-05,
+      "loss": 0.3092,
+      "step": 2680
+    },
+    {
+      "epoch": 8.010253164556962,
+      "grad_norm": 8.060324668884277,
+      "learning_rate": 4.6097046413502107e-05,
+      "loss": 0.2931,
+      "step": 2690
+    },
+    {
+      "epoch": 8.010886075949367,
+      "grad_norm": 18.59467124938965,
+      "learning_rate": 4.606188466947961e-05,
+      "loss": 0.2696,
+      "step": 2700
+    },
+    {
+      "epoch": 8.011518987341772,
+      "grad_norm": 0.5714246034622192,
+      "learning_rate": 4.60267229254571e-05,
+      "loss": 0.2422,
+      "step": 2710
+    },
+    {
+      "epoch": 8.012151898734178,
+      "grad_norm": 2.771367311477661,
+      "learning_rate": 4.59915611814346e-05,
+      "loss": 0.2716,
+      "step": 2720
+    },
+    {
+      "epoch": 8.012784810126583,
+      "grad_norm": 1.5201733112335205,
+      "learning_rate": 4.595639943741209e-05,
+      "loss": 0.1204,
+      "step": 2730
+    },
+    {
+      "epoch": 8.013417721518987,
+      "grad_norm": 15.473625183105469,
+      "learning_rate": 4.59212376933896e-05,
+      "loss": 0.2585,
+      "step": 2740
+    },
+    {
+      "epoch": 8.014050632911392,
+      "grad_norm": 3.226699113845825,
+      "learning_rate": 4.588607594936709e-05,
+      "loss": 0.1866,
+      "step": 2750
+    },
+    {
+      "epoch": 8.014683544303798,
+      "grad_norm": 1.999643087387085,
+      "learning_rate": 4.585091420534459e-05,
+      "loss": 0.237,
+      "step": 2760
+    },
+    {
+      "epoch": 8.015316455696203,
+      "grad_norm": 18.86202049255371,
+      "learning_rate": 4.581575246132208e-05,
+      "loss": 0.2425,
+      "step": 2770
+    },
+    {
+      "epoch": 8.015949367088608,
+      "grad_norm": 0.8307034969329834,
+      "learning_rate": 4.5780590717299585e-05,
+      "loss": 0.1629,
+      "step": 2780
+    },
+    {
+      "epoch": 8.016582278481012,
+      "grad_norm": 15.315855979919434,
+      "learning_rate": 4.5745428973277076e-05,
+      "loss": 0.2366,
+      "step": 2790
+    },
+    {
+      "epoch": 8.017215189873419,
+      "grad_norm": 16.721681594848633,
+      "learning_rate": 4.5710267229254575e-05,
+      "loss": 0.3524,
+      "step": 2800
+    },
+    {
+      "epoch": 8.017848101265823,
+      "grad_norm": 0.7722905278205872,
+      "learning_rate": 4.5675105485232066e-05,
+      "loss": 0.1772,
+      "step": 2810
+    },
+    {
+      "epoch": 8.018481012658228,
+      "grad_norm": 1.7419816255569458,
+      "learning_rate": 4.5639943741209564e-05,
+      "loss": 0.206,
+      "step": 2820
+    },
+    {
+      "epoch": 8.019113924050632,
+      "grad_norm": 4.165656089782715,
+      "learning_rate": 4.560478199718706e-05,
+      "loss": 0.2458,
+      "step": 2830
+    },
+    {
+      "epoch": 8.019746835443039,
+      "grad_norm": 2.085017681121826,
+      "learning_rate": 4.556962025316456e-05,
+      "loss": 0.228,
+      "step": 2840
+    },
+    {
+      "epoch": 8.02,
+      "eval_accuracy": 0.9309021113243762,
+      "eval_loss": 0.19722820818424225,
+      "eval_runtime": 905.8613,
+      "eval_samples_per_second": 0.575,
+      "eval_steps_per_second": 0.073,
+      "step": 2844
+    },
+    {
+      "epoch": 9.000379746835444,
+      "grad_norm": 13.498887062072754,
+      "learning_rate": 4.553445850914206e-05,
+      "loss": 0.1674,
+      "step": 2850
+    },
+    {
+      "epoch": 9.001012658227848,
+      "grad_norm": 0.7413739562034607,
+      "learning_rate": 4.549929676511955e-05,
+      "loss": 0.1371,
+      "step": 2860
+    },
+    {
+      "epoch": 9.001645569620253,
+      "grad_norm": 1.669324278831482,
+      "learning_rate": 4.546413502109705e-05,
+      "loss": 0.146,
+      "step": 2870
+    },
+    {
+      "epoch": 9.002278481012658,
+      "grad_norm": 2.6705691814422607,
+      "learning_rate": 4.542897327707454e-05,
+      "loss": 0.2867,
+      "step": 2880
+    },
+    {
+      "epoch": 9.002911392405064,
+      "grad_norm": 3.0164263248443604,
+      "learning_rate": 4.5393811533052046e-05,
+      "loss": 0.2902,
+      "step": 2890
+    },
+    {
+      "epoch": 9.003544303797469,
+      "grad_norm": 6.8492302894592285,
+      "learning_rate": 4.535864978902954e-05,
+      "loss": 0.1386,
+      "step": 2900
+    },
+    {
+      "epoch": 9.004177215189873,
+      "grad_norm": 0.835591733455658,
+      "learning_rate": 4.5323488045007036e-05,
+      "loss": 0.1567,
+      "step": 2910
+    },
+    {
+      "epoch": 9.004810126582278,
+      "grad_norm": 4.592350959777832,
+      "learning_rate": 4.528832630098453e-05,
+      "loss": 0.2066,
+      "step": 2920
+    },
+    {
+      "epoch": 9.005443037974684,
+      "grad_norm": 5.640721797943115,
+      "learning_rate": 4.525316455696203e-05,
+      "loss": 0.0528,
+      "step": 2930
+    },
+    {
+      "epoch": 9.006075949367089,
+      "grad_norm": 13.428877830505371,
+      "learning_rate": 4.5218002812939524e-05,
+      "loss": 0.1163,
+      "step": 2940
+    },
+    {
+      "epoch": 9.006708860759494,
+      "grad_norm": 0.22825801372528076,
+      "learning_rate": 4.518284106891702e-05,
+      "loss": 0.1288,
+      "step": 2950
+    },
+    {
+      "epoch": 9.007341772151898,
+      "grad_norm": 1.8029659986495972,
+      "learning_rate": 4.5147679324894514e-05,
+      "loss": 0.2082,
+      "step": 2960
+    },
+    {
+      "epoch": 9.007974683544305,
+      "grad_norm": 0.23065458238124847,
+      "learning_rate": 4.511251758087202e-05,
+      "loss": 0.2156,
+      "step": 2970
+    },
+    {
+      "epoch": 9.00860759493671,
+      "grad_norm": 3.216405153274536,
+      "learning_rate": 4.507735583684951e-05,
+      "loss": 0.2436,
+      "step": 2980
+    },
+    {
+      "epoch": 9.009240506329114,
+      "grad_norm": 0.4626835584640503,
+      "learning_rate": 4.504219409282701e-05,
+      "loss": 0.2139,
+      "step": 2990
+    },
+    {
+      "epoch": 9.009873417721519,
+      "grad_norm": 5.6381516456604,
+      "learning_rate": 4.50070323488045e-05,
+      "loss": 0.1326,
+      "step": 3000
+    },
+    {
+      "epoch": 9.010506329113925,
+      "grad_norm": 7.001626968383789,
+      "learning_rate": 4.4971870604782e-05,
+      "loss": 0.1666,
+      "step": 3010
+    },
+    {
+      "epoch": 9.01113924050633,
+      "grad_norm": 3.0966169834136963,
+      "learning_rate": 4.49367088607595e-05,
+      "loss": 0.2111,
+      "step": 3020
+    },
+    {
+      "epoch": 9.011772151898734,
+      "grad_norm": 7.884695053100586,
+      "learning_rate": 4.490154711673699e-05,
+      "loss": 0.2074,
+      "step": 3030
+    },
+    {
+      "epoch": 9.012405063291139,
+      "grad_norm": 18.61140251159668,
+      "learning_rate": 4.486638537271449e-05,
+      "loss": 0.1748,
+      "step": 3040
+    },
+    {
+      "epoch": 9.013037974683545,
+      "grad_norm": 23.860849380493164,
+      "learning_rate": 4.4831223628691985e-05,
+      "loss": 0.1367,
+      "step": 3050
+    },
+    {
+      "epoch": 9.01367088607595,
+      "grad_norm": 6.8134942054748535,
+      "learning_rate": 4.4796061884669484e-05,
+      "loss": 0.1653,
+      "step": 3060
+    },
+    {
+      "epoch": 9.014303797468354,
+      "grad_norm": 21.31640625,
+      "learning_rate": 4.4760900140646975e-05,
+      "loss": 0.3217,
+      "step": 3070
+    },
+    {
+      "epoch": 9.014936708860759,
+      "grad_norm": 3.0076091289520264,
+      "learning_rate": 4.4725738396624474e-05,
+      "loss": 0.2275,
+      "step": 3080
+    },
+    {
+      "epoch": 9.015569620253165,
+      "grad_norm": 0.38252460956573486,
+      "learning_rate": 4.469057665260197e-05,
+      "loss": 0.0889,
+      "step": 3090
+    },
+    {
+      "epoch": 9.01620253164557,
+      "grad_norm": 3.001718282699585,
+      "learning_rate": 4.465541490857947e-05,
+      "loss": 0.2065,
+      "step": 3100
+    },
+    {
+      "epoch": 9.016835443037975,
+      "grad_norm": 5.9051923751831055,
+      "learning_rate": 4.462025316455696e-05,
+      "loss": 0.2604,
+      "step": 3110
+    },
+    {
+      "epoch": 9.01746835443038,
+      "grad_norm": 1.5905555486679077,
+      "learning_rate": 4.458509142053446e-05,
+      "loss": 0.2658,
+      "step": 3120
+    },
+    {
+      "epoch": 9.018101265822784,
+      "grad_norm": 2.863093137741089,
+      "learning_rate": 4.454992967651196e-05,
+      "loss": 0.174,
+      "step": 3130
+    },
+    {
+      "epoch": 9.01873417721519,
+      "grad_norm": 2.637539863586426,
+      "learning_rate": 4.451476793248946e-05,
+      "loss": 0.1515,
+      "step": 3140
+    },
+    {
+      "epoch": 9.019367088607595,
+      "grad_norm": 0.3857173025608063,
+      "learning_rate": 4.447960618846695e-05,
+      "loss": 0.0724,
+      "step": 3150
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 1.806871771812439,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.1711,
+      "step": 3160
+    },
+    {
+      "epoch": 9.02,
+      "eval_accuracy": 0.9481765834932822,
+      "eval_loss": 0.14262719452381134,
+      "eval_runtime": 899.0913,
+      "eval_samples_per_second": 0.579,
+      "eval_steps_per_second": 0.073,
+      "step": 3160
+    },
+    {
+      "epoch": 10.000632911392405,
+      "grad_norm": 5.219705104827881,
+      "learning_rate": 4.4409282700421945e-05,
+      "loss": 0.1425,
+      "step": 3170
+    },
+    {
+      "epoch": 10.00126582278481,
+      "grad_norm": 12.191892623901367,
+      "learning_rate": 4.4374120956399436e-05,
+      "loss": 0.1729,
+      "step": 3180
+    },
+    {
+      "epoch": 10.001898734177216,
+      "grad_norm": 1.2383445501327515,
+      "learning_rate": 4.4338959212376935e-05,
+      "loss": 0.2565,
+      "step": 3190
+    },
+    {
+      "epoch": 10.00253164556962,
+      "grad_norm": 0.5347968935966492,
+      "learning_rate": 4.430379746835443e-05,
+      "loss": 0.1981,
+      "step": 3200
+    },
+    {
+      "epoch": 10.003164556962025,
+      "grad_norm": 0.8782184720039368,
+      "learning_rate": 4.426863572433193e-05,
+      "loss": 0.0789,
+      "step": 3210
+    },
+    {
+      "epoch": 10.00379746835443,
+      "grad_norm": 0.14596301317214966,
+      "learning_rate": 4.423347398030942e-05,
+      "loss": 0.1853,
+      "step": 3220
+    },
+    {
+      "epoch": 10.004430379746836,
+      "grad_norm": 0.4674586057662964,
+      "learning_rate": 4.419831223628692e-05,
+      "loss": 0.1213,
+      "step": 3230
+    },
+    {
+      "epoch": 10.00506329113924,
+      "grad_norm": 5.122977256774902,
+      "learning_rate": 4.416315049226441e-05,
+      "loss": 0.1708,
+      "step": 3240
+    },
+    {
+      "epoch": 10.005696202531645,
+      "grad_norm": 2.872215509414673,
+      "learning_rate": 4.412798874824192e-05,
+      "loss": 0.2037,
+      "step": 3250
+    },
+    {
+      "epoch": 10.00632911392405,
+      "grad_norm": 3.31655216217041,
+      "learning_rate": 4.409282700421941e-05,
+      "loss": 0.1574,
+      "step": 3260
+    },
+    {
+      "epoch": 10.006962025316456,
+      "grad_norm": 0.5382121801376343,
+      "learning_rate": 4.405766526019691e-05,
+      "loss": 0.1031,
+      "step": 3270
+    },
+    {
+      "epoch": 10.00759493670886,
+      "grad_norm": 1.605431079864502,
+      "learning_rate": 4.4022503516174406e-05,
+      "loss": 0.1636,
+      "step": 3280
+    },
+    {
+      "epoch": 10.008227848101265,
+      "grad_norm": 9.264389038085938,
+      "learning_rate": 4.3987341772151904e-05,
+      "loss": 0.2556,
+      "step": 3290
+    },
+    {
+      "epoch": 10.00886075949367,
+      "grad_norm": 0.28090256452560425,
+      "learning_rate": 4.3952180028129396e-05,
+      "loss": 0.1123,
+      "step": 3300
+    },
+    {
+      "epoch": 10.009493670886076,
+      "grad_norm": 2.213951349258423,
+      "learning_rate": 4.3917018284106894e-05,
+      "loss": 0.0749,
+      "step": 3310
+    },
+    {
+      "epoch": 10.010126582278481,
+      "grad_norm": 6.6392035484313965,
+      "learning_rate": 4.388185654008439e-05,
+      "loss": 0.1676,
+      "step": 3320
+    },
+    {
+      "epoch": 10.010759493670886,
+      "grad_norm": 0.35711225867271423,
+      "learning_rate": 4.384669479606189e-05,
+      "loss": 0.0756,
+      "step": 3330
+    },
+    {
+      "epoch": 10.01139240506329,
+      "grad_norm": 0.24826167523860931,
+      "learning_rate": 4.381153305203938e-05,
+      "loss": 0.0955,
+      "step": 3340
+    },
+    {
+      "epoch": 10.012025316455697,
+      "grad_norm": 5.6375579833984375,
+      "learning_rate": 4.377637130801688e-05,
+      "loss": 0.1472,
+      "step": 3350
+    },
+    {
+      "epoch": 10.012658227848101,
+      "grad_norm": 0.28531956672668457,
+      "learning_rate": 4.374120956399438e-05,
+      "loss": 0.0843,
+      "step": 3360
+    },
+    {
+      "epoch": 10.013291139240506,
+      "grad_norm": 5.148308277130127,
+      "learning_rate": 4.370604781997187e-05,
+      "loss": 0.222,
+      "step": 3370
+    },
+    {
+      "epoch": 10.01392405063291,
+      "grad_norm": 0.3964751064777374,
+      "learning_rate": 4.367088607594937e-05,
+      "loss": 0.1279,
+      "step": 3380
+    },
+    {
+      "epoch": 10.014556962025317,
+      "grad_norm": 4.597806930541992,
+      "learning_rate": 4.363572433192686e-05,
+      "loss": 0.1258,
+      "step": 3390
+    },
+    {
+      "epoch": 10.015189873417722,
+      "grad_norm": 1.5554261207580566,
+      "learning_rate": 4.3600562587904366e-05,
+      "loss": 0.1193,
+      "step": 3400
+    },
+    {
+      "epoch": 10.015822784810126,
+      "grad_norm": 0.0851406529545784,
+      "learning_rate": 4.356540084388186e-05,
+      "loss": 0.1182,
+      "step": 3410
+    },
+    {
+      "epoch": 10.01645569620253,
+      "grad_norm": 3.5009474754333496,
+      "learning_rate": 4.3530239099859356e-05,
+      "loss": 0.155,
+      "step": 3420
+    },
+    {
+      "epoch": 10.017088607594937,
+      "grad_norm": 2.019500732421875,
+      "learning_rate": 4.349507735583685e-05,
+      "loss": 0.1827,
+      "step": 3430
+    },
+    {
+      "epoch": 10.017721518987342,
+      "grad_norm": 1.833369255065918,
+      "learning_rate": 4.345991561181435e-05,
+      "loss": 0.0991,
+      "step": 3440
+    },
+    {
+      "epoch": 10.018354430379746,
+      "grad_norm": 0.4851783215999603,
+      "learning_rate": 4.3424753867791844e-05,
+      "loss": 0.2478,
+      "step": 3450
+    },
+    {
+      "epoch": 10.018987341772151,
+      "grad_norm": 2.687387466430664,
+      "learning_rate": 4.338959212376934e-05,
+      "loss": 0.0995,
+      "step": 3460
+    },
+    {
+      "epoch": 10.019620253164558,
+      "grad_norm": 11.315146446228027,
+      "learning_rate": 4.3354430379746834e-05,
+      "loss": 0.2251,
+      "step": 3470
+    },
+    {
+      "epoch": 10.02,
+      "eval_accuracy": 0.9558541266794626,
+      "eval_loss": 0.0965082123875618,
+      "eval_runtime": 881.0116,
+      "eval_samples_per_second": 0.591,
+      "eval_steps_per_second": 0.075,
+      "step": 3476
+    },
+    {
+      "epoch": 11.000253164556963,
+      "grad_norm": 8.084904670715332,
+      "learning_rate": 4.331926863572434e-05,
+      "loss": 0.1515,
+      "step": 3480
+    },
+    {
+      "epoch": 11.000886075949367,
+      "grad_norm": 27.45550537109375,
+      "learning_rate": 4.328410689170183e-05,
+      "loss": 0.1284,
+      "step": 3490
+    },
+    {
+      "epoch": 11.001518987341772,
+      "grad_norm": 8.539572715759277,
+      "learning_rate": 4.324894514767933e-05,
+      "loss": 0.1598,
+      "step": 3500
+    },
+    {
+      "epoch": 11.002151898734176,
+      "grad_norm": 0.08583386987447739,
+      "learning_rate": 4.321378340365682e-05,
+      "loss": 0.0517,
+      "step": 3510
+    },
+    {
+      "epoch": 11.002784810126583,
+      "grad_norm": 0.14813897013664246,
+      "learning_rate": 4.317862165963432e-05,
+      "loss": 0.1125,
+      "step": 3520
+    },
+    {
+      "epoch": 11.003417721518987,
+      "grad_norm": 2.4185800552368164,
+      "learning_rate": 4.314345991561182e-05,
+      "loss": 0.1984,
+      "step": 3530
+    },
+    {
+      "epoch": 11.004050632911392,
+      "grad_norm": 0.21438997983932495,
+      "learning_rate": 4.3108298171589315e-05,
+      "loss": 0.0713,
+      "step": 3540
+    },
+    {
+      "epoch": 11.004683544303797,
+      "grad_norm": 0.394996702671051,
+      "learning_rate": 4.307313642756681e-05,
+      "loss": 0.1331,
+      "step": 3550
+    },
+    {
+      "epoch": 11.005316455696203,
+      "grad_norm": 1.3863253593444824,
+      "learning_rate": 4.3037974683544305e-05,
+      "loss": 0.0773,
+      "step": 3560
+    },
+    {
+      "epoch": 11.005949367088608,
+      "grad_norm": 1.7728908061981201,
+      "learning_rate": 4.3002812939521803e-05,
+      "loss": 0.1833,
+      "step": 3570
+    },
+    {
+      "epoch": 11.006582278481012,
+      "grad_norm": 0.08379014581441879,
+      "learning_rate": 4.2967651195499295e-05,
+      "loss": 0.0516,
+      "step": 3580
+    },
+    {
+      "epoch": 11.007215189873417,
+      "grad_norm": 2.6824705600738525,
+      "learning_rate": 4.293248945147679e-05,
+      "loss": 0.1616,
+      "step": 3590
+    },
+    {
+      "epoch": 11.007848101265823,
+      "grad_norm": 2.570587158203125,
+      "learning_rate": 4.289732770745429e-05,
+      "loss": 0.1306,
+      "step": 3600
+    },
+    {
+      "epoch": 11.008481012658228,
+      "grad_norm": 14.159221649169922,
+      "learning_rate": 4.286216596343179e-05,
+      "loss": 0.1083,
+      "step": 3610
+    },
+    {
+      "epoch": 11.009113924050633,
+      "grad_norm": 9.88803482055664,
+      "learning_rate": 4.282700421940928e-05,
+      "loss": 0.1391,
+      "step": 3620
+    },
+    {
+      "epoch": 11.009746835443037,
+      "grad_norm": 5.837044715881348,
+      "learning_rate": 4.279184247538678e-05,
+      "loss": 0.1622,
+      "step": 3630
+    },
+    {
+      "epoch": 11.010379746835444,
+      "grad_norm": 0.1928025782108307,
+      "learning_rate": 4.275668073136428e-05,
+      "loss": 0.1777,
+      "step": 3640
+    },
+    {
+      "epoch": 11.011012658227848,
+      "grad_norm": 0.26675212383270264,
+      "learning_rate": 4.2721518987341776e-05,
+      "loss": 0.088,
+      "step": 3650
+    },
+    {
+      "epoch": 11.011645569620253,
+      "grad_norm": 0.1680004894733429,
+      "learning_rate": 4.268635724331927e-05,
+      "loss": 0.1404,
+      "step": 3660
+    },
+    {
+      "epoch": 11.012278481012657,
+      "grad_norm": 2.046947717666626,
+      "learning_rate": 4.2651195499296766e-05,
+      "loss": 0.1198,
+      "step": 3670
+    },
+    {
+      "epoch": 11.012911392405064,
+      "grad_norm": 1.6072787046432495,
+      "learning_rate": 4.2616033755274265e-05,
+      "loss": 0.085,
+      "step": 3680
+    },
+    {
+      "epoch": 11.013544303797469,
+      "grad_norm": 0.38024991750717163,
+      "learning_rate": 4.258087201125176e-05,
+      "loss": 0.0692,
+      "step": 3690
+    },
+    {
+      "epoch": 11.014177215189873,
+      "grad_norm": 0.4724186062812805,
+      "learning_rate": 4.2545710267229255e-05,
+      "loss": 0.1008,
+      "step": 3700
+    },
+    {
+      "epoch": 11.014810126582278,
+      "grad_norm": 4.140766143798828,
+      "learning_rate": 4.251054852320675e-05,
+      "loss": 0.1582,
+      "step": 3710
+    },
+    {
+      "epoch": 11.015443037974684,
+      "grad_norm": 2.098039388656616,
+      "learning_rate": 4.247538677918425e-05,
+      "loss": 0.1359,
+      "step": 3720
+    },
+    {
+      "epoch": 11.016075949367089,
+      "grad_norm": 1.2859419584274292,
+      "learning_rate": 4.244022503516174e-05,
+      "loss": 0.0806,
+      "step": 3730
+    },
+    {
+      "epoch": 11.016708860759493,
+      "grad_norm": 4.074576377868652,
+      "learning_rate": 4.240506329113924e-05,
+      "loss": 0.2069,
+      "step": 3740
+    },
+    {
+      "epoch": 11.017341772151898,
+      "grad_norm": 11.83212947845459,
+      "learning_rate": 4.236990154711674e-05,
+      "loss": 0.1142,
+      "step": 3750
+    },
+    {
+      "epoch": 11.017974683544304,
+      "grad_norm": 2.487534999847412,
+      "learning_rate": 4.233473980309424e-05,
+      "loss": 0.1747,
+      "step": 3760
+    },
+    {
+      "epoch": 11.018607594936709,
+      "grad_norm": 6.369086265563965,
+      "learning_rate": 4.229957805907173e-05,
+      "loss": 0.2394,
+      "step": 3770
+    },
+    {
+      "epoch": 11.019240506329114,
+      "grad_norm": 14.795226097106934,
+      "learning_rate": 4.226441631504923e-05,
+      "loss": 0.1638,
+      "step": 3780
+    },
+    {
+      "epoch": 11.019873417721518,
+      "grad_norm": 2.7429358959198,
+      "learning_rate": 4.2229254571026726e-05,
+      "loss": 0.1697,
+      "step": 3790
+    },
+    {
+      "epoch": 11.02,
+      "eval_accuracy": 0.9539347408829175,
+      "eval_loss": 0.11410919576883316,
+      "eval_runtime": 878.2486,
+      "eval_samples_per_second": 0.593,
+      "eval_steps_per_second": 0.075,
+      "step": 3792
+    },
+    {
+      "epoch": 12.000506329113923,
+      "grad_norm": 6.182743072509766,
+      "learning_rate": 4.2194092827004224e-05,
+      "loss": 0.1566,
+      "step": 3800
+    },
+    {
+      "epoch": 12.00113924050633,
+      "grad_norm": 3.826575994491577,
+      "learning_rate": 4.2158931082981716e-05,
+      "loss": 0.0969,
+      "step": 3810
+    },
+    {
+      "epoch": 12.001772151898734,
+      "grad_norm": 0.3016466498374939,
+      "learning_rate": 4.2123769338959214e-05,
+      "loss": 0.2192,
+      "step": 3820
+    },
+    {
+      "epoch": 12.002405063291139,
+      "grad_norm": 2.917588710784912,
+      "learning_rate": 4.208860759493671e-05,
+      "loss": 0.0723,
+      "step": 3830
+    },
+    {
+      "epoch": 12.003037974683544,
+      "grad_norm": 0.10706198215484619,
+      "learning_rate": 4.205344585091421e-05,
+      "loss": 0.136,
+      "step": 3840
+    },
+    {
+      "epoch": 12.00367088607595,
+      "grad_norm": 1.9893161058425903,
+      "learning_rate": 4.20182841068917e-05,
+      "loss": 0.1361,
+      "step": 3850
+    },
+    {
+      "epoch": 12.004303797468355,
+      "grad_norm": 1.9333717823028564,
+      "learning_rate": 4.19831223628692e-05,
+      "loss": 0.1467,
+      "step": 3860
+    },
+    {
+      "epoch": 12.00493670886076,
+      "grad_norm": 0.13750018179416656,
+      "learning_rate": 4.19479606188467e-05,
+      "loss": 0.072,
+      "step": 3870
+    },
+    {
+      "epoch": 12.005569620253164,
+      "grad_norm": 3.1651651859283447,
+      "learning_rate": 4.19127988748242e-05,
+      "loss": 0.1241,
+      "step": 3880
+    },
+    {
+      "epoch": 12.00620253164557,
+      "grad_norm": 1.7854634523391724,
+      "learning_rate": 4.187763713080169e-05,
+      "loss": 0.1204,
+      "step": 3890
+    },
+    {
+      "epoch": 12.006835443037975,
+      "grad_norm": 2.030898332595825,
+      "learning_rate": 4.184247538677919e-05,
+      "loss": 0.1028,
+      "step": 3900
+    },
+    {
+      "epoch": 12.00746835443038,
+      "grad_norm": 2.5038902759552,
+      "learning_rate": 4.1807313642756686e-05,
+      "loss": 0.0767,
+      "step": 3910
+    },
+    {
+      "epoch": 12.008101265822784,
+      "grad_norm": 0.07481967657804489,
+      "learning_rate": 4.177215189873418e-05,
+      "loss": 0.1092,
+      "step": 3920
+    },
+    {
+      "epoch": 12.00873417721519,
+      "grad_norm": 15.426689147949219,
+      "learning_rate": 4.1736990154711675e-05,
+      "loss": 0.1277,
+      "step": 3930
+    },
+    {
+      "epoch": 12.009367088607595,
+      "grad_norm": 0.38814952969551086,
+      "learning_rate": 4.170182841068917e-05,
+      "loss": 0.0369,
+      "step": 3940
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 3.7652478218078613,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.0799,
+      "step": 3950
+    },
+    {
+      "epoch": 12.010632911392404,
+      "grad_norm": 0.06228223815560341,
+      "learning_rate": 4.1631504922644164e-05,
+      "loss": 0.1217,
+      "step": 3960
+    },
+    {
+      "epoch": 12.01126582278481,
+      "grad_norm": 0.3022870719432831,
+      "learning_rate": 4.159634317862166e-05,
+      "loss": 0.0405,
+      "step": 3970
+    },
+    {
+      "epoch": 12.011898734177215,
+      "grad_norm": 2.400820255279541,
+      "learning_rate": 4.1561181434599153e-05,
+      "loss": 0.0291,
+      "step": 3980
+    },
+    {
+      "epoch": 12.01253164556962,
+      "grad_norm": 2.68705415725708,
+      "learning_rate": 4.152601969057666e-05,
+      "loss": 0.1775,
+      "step": 3990
+    },
+    {
+      "epoch": 12.013164556962025,
+      "grad_norm": 0.5902582406997681,
+      "learning_rate": 4.149085794655415e-05,
+      "loss": 0.1165,
+      "step": 4000
+    },
+    {
+      "epoch": 12.013797468354431,
+      "grad_norm": 0.9047210812568665,
+      "learning_rate": 4.145569620253165e-05,
+      "loss": 0.07,
+      "step": 4010
+    },
+    {
+      "epoch": 12.014430379746836,
+      "grad_norm": 0.09479296952486038,
+      "learning_rate": 4.142053445850914e-05,
+      "loss": 0.0495,
+      "step": 4020
+    },
+    {
+      "epoch": 12.01506329113924,
+      "grad_norm": 1.401077151298523,
+      "learning_rate": 4.1385372714486645e-05,
+      "loss": 0.0204,
+      "step": 4030
+    },
+    {
+      "epoch": 12.015696202531645,
+      "grad_norm": 4.037631988525391,
+      "learning_rate": 4.135021097046414e-05,
+      "loss": 0.1883,
+      "step": 4040
+    },
+    {
+      "epoch": 12.016329113924051,
+      "grad_norm": 1.7805005311965942,
+      "learning_rate": 4.1315049226441635e-05,
+      "loss": 0.1802,
+      "step": 4050
+    },
+    {
+      "epoch": 12.016962025316456,
+      "grad_norm": 3.6074600219726562,
+      "learning_rate": 4.1279887482419127e-05,
+      "loss": 0.0523,
+      "step": 4060
+    },
+    {
+      "epoch": 12.01759493670886,
+      "grad_norm": 0.1754535734653473,
+      "learning_rate": 4.1244725738396625e-05,
+      "loss": 0.0763,
+      "step": 4070
+    },
+    {
+      "epoch": 12.018227848101265,
+      "grad_norm": 1.0715147256851196,
+      "learning_rate": 4.120956399437412e-05,
+      "loss": 0.1114,
+      "step": 4080
+    },
+    {
+      "epoch": 12.018860759493672,
+      "grad_norm": 0.06572480499744415,
+      "learning_rate": 4.1174402250351615e-05,
+      "loss": 0.0499,
+      "step": 4090
+    },
+    {
+      "epoch": 12.019493670886076,
+      "grad_norm": 2.0521042346954346,
+      "learning_rate": 4.113924050632912e-05,
+      "loss": 0.1229,
+      "step": 4100
+    },
+    {
+      "epoch": 12.02,
+      "eval_accuracy": 0.9539347408829175,
+      "eval_loss": 0.13622045516967773,
+      "eval_runtime": 873.7358,
+      "eval_samples_per_second": 0.596,
+      "eval_steps_per_second": 0.076,
+      "step": 4108
+    },
+    {
+      "epoch": 13.000126582278481,
+      "grad_norm": 0.08451604843139648,
+      "learning_rate": 4.110407876230661e-05,
+      "loss": 0.119,
+      "step": 4110
+    },
+    {
+      "epoch": 13.000759493670886,
+      "grad_norm": 1.373328685760498,
+      "learning_rate": 4.106891701828411e-05,
+      "loss": 0.0773,
+      "step": 4120
+    },
+    {
+      "epoch": 13.00139240506329,
+      "grad_norm": 5.43549919128418,
+      "learning_rate": 4.10337552742616e-05,
+      "loss": 0.1298,
+      "step": 4130
+    },
+    {
+      "epoch": 13.002025316455697,
+      "grad_norm": 1.0948410034179688,
+      "learning_rate": 4.0998593530239106e-05,
+      "loss": 0.1255,
+      "step": 4140
+    },
+    {
+      "epoch": 13.002658227848102,
+      "grad_norm": 0.060861099511384964,
+      "learning_rate": 4.09634317862166e-05,
+      "loss": 0.1304,
+      "step": 4150
+    },
+    {
+      "epoch": 13.003291139240506,
+      "grad_norm": 15.299447059631348,
+      "learning_rate": 4.0928270042194096e-05,
+      "loss": 0.1789,
+      "step": 4160
+    },
+    {
+      "epoch": 13.00392405063291,
+      "grad_norm": 0.08800447732210159,
+      "learning_rate": 4.089310829817159e-05,
+      "loss": 0.0934,
+      "step": 4170
+    },
+    {
+      "epoch": 13.004556962025317,
+      "grad_norm": 2.429713010787964,
+      "learning_rate": 4.085794655414909e-05,
+      "loss": 0.1312,
+      "step": 4180
+    },
+    {
+      "epoch": 13.005189873417722,
+      "grad_norm": 0.05040478706359863,
+      "learning_rate": 4.0822784810126584e-05,
+      "loss": 0.0764,
+      "step": 4190
+    },
+    {
+      "epoch": 13.005822784810126,
+      "grad_norm": 12.875263214111328,
+      "learning_rate": 4.078762306610408e-05,
+      "loss": 0.2538,
+      "step": 4200
+    },
+    {
+      "epoch": 13.006455696202531,
+      "grad_norm": 1.7678494453430176,
+      "learning_rate": 4.0752461322081574e-05,
+      "loss": 0.0656,
+      "step": 4210
+    },
+    {
+      "epoch": 13.007088607594937,
+      "grad_norm": 0.05160669609904289,
+      "learning_rate": 4.071729957805907e-05,
+      "loss": 0.1236,
+      "step": 4220
+    },
+    {
+      "epoch": 13.007721518987342,
+      "grad_norm": 2.3818519115448,
+      "learning_rate": 4.068213783403657e-05,
+      "loss": 0.1056,
+      "step": 4230
+    },
+    {
+      "epoch": 13.008354430379747,
+      "grad_norm": 1.16487717628479,
+      "learning_rate": 4.064697609001407e-05,
+      "loss": 0.1062,
+      "step": 4240
+    },
+    {
+      "epoch": 13.008987341772151,
+      "grad_norm": 1.0950814485549927,
+      "learning_rate": 4.061181434599156e-05,
+      "loss": 0.114,
+      "step": 4250
+    },
+    {
+      "epoch": 13.009620253164558,
+      "grad_norm": 2.856937885284424,
+      "learning_rate": 4.057665260196906e-05,
+      "loss": 0.1701,
+      "step": 4260
+    },
+    {
+      "epoch": 13.010253164556962,
+      "grad_norm": 0.057572536170482635,
+      "learning_rate": 4.054149085794656e-05,
+      "loss": 0.1006,
+      "step": 4270
+    },
+    {
+      "epoch": 13.010886075949367,
+      "grad_norm": 0.05074339732527733,
+      "learning_rate": 4.050632911392405e-05,
+      "loss": 0.1008,
+      "step": 4280
+    },
+    {
+      "epoch": 13.011518987341772,
+      "grad_norm": 3.1123783588409424,
+      "learning_rate": 4.047116736990155e-05,
+      "loss": 0.0352,
+      "step": 4290
+    },
+    {
+      "epoch": 13.012151898734178,
+      "grad_norm": 2.150893211364746,
+      "learning_rate": 4.0436005625879046e-05,
+      "loss": 0.0824,
+      "step": 4300
+    },
+    {
+      "epoch": 13.012784810126583,
+      "grad_norm": 2.5087695121765137,
+      "learning_rate": 4.0400843881856544e-05,
+      "loss": 0.1343,
+      "step": 4310
+    },
+    {
+      "epoch": 13.013417721518987,
+      "grad_norm": 1.6841331720352173,
+      "learning_rate": 4.0365682137834036e-05,
+      "loss": 0.1258,
+      "step": 4320
+    },
+    {
+      "epoch": 13.014050632911392,
+      "grad_norm": 0.3586716949939728,
+      "learning_rate": 4.0330520393811534e-05,
+      "loss": 0.1204,
+      "step": 4330
+    },
+    {
+      "epoch": 13.014683544303798,
+      "grad_norm": 0.045751988887786865,
+      "learning_rate": 4.029535864978903e-05,
+      "loss": 0.0682,
+      "step": 4340
+    },
+    {
+      "epoch": 13.015316455696203,
+      "grad_norm": 3.5271830558776855,
+      "learning_rate": 4.026019690576653e-05,
+      "loss": 0.261,
+      "step": 4350
+    },
+    {
+      "epoch": 13.015949367088608,
+      "grad_norm": 3.505953550338745,
+      "learning_rate": 4.022503516174402e-05,
+      "loss": 0.1309,
+      "step": 4360
+    },
+    {
+      "epoch": 13.016582278481012,
+      "grad_norm": 0.1506178230047226,
+      "learning_rate": 4.018987341772152e-05,
+      "loss": 0.0923,
+      "step": 4370
+    },
+    {
+      "epoch": 13.017215189873419,
+      "grad_norm": 0.3128964304924011,
+      "learning_rate": 4.015471167369902e-05,
+      "loss": 0.0573,
+      "step": 4380
+    },
+    {
+      "epoch": 13.017848101265823,
+      "grad_norm": 2.3400206565856934,
+      "learning_rate": 4.011954992967652e-05,
+      "loss": 0.099,
+      "step": 4390
+    },
+    {
+      "epoch": 13.018481012658228,
+      "grad_norm": 0.4466443955898285,
+      "learning_rate": 4.008438818565401e-05,
+      "loss": 0.0531,
+      "step": 4400
+    },
+    {
+      "epoch": 13.019113924050632,
+      "grad_norm": 2.060894250869751,
+      "learning_rate": 4.004922644163151e-05,
+      "loss": 0.0446,
+      "step": 4410
+    },
+    {
+      "epoch": 13.019746835443039,
+      "grad_norm": 1.2128475904464722,
+      "learning_rate": 4.0014064697609005e-05,
+      "loss": 0.0676,
+      "step": 4420
+    },
+    {
+      "epoch": 13.02,
+      "eval_accuracy": 0.9654510556621881,
+      "eval_loss": 0.07451339066028595,
+      "eval_runtime": 911.5098,
+      "eval_samples_per_second": 0.572,
+      "eval_steps_per_second": 0.072,
+      "step": 4424
+    },
+    {
+      "epoch": 14.000379746835444,
+      "grad_norm": 0.05884459242224693,
+      "learning_rate": 3.99789029535865e-05,
+      "loss": 0.0896,
+      "step": 4430
+    },
+    {
+      "epoch": 14.001012658227848,
+      "grad_norm": 0.05943896621465683,
+      "learning_rate": 3.9943741209563995e-05,
+      "loss": 0.1132,
+      "step": 4440
+    },
+    {
+      "epoch": 14.001645569620253,
+      "grad_norm": 0.13569866120815277,
+      "learning_rate": 3.9908579465541493e-05,
+      "loss": 0.091,
+      "step": 4450
+    },
+    {
+      "epoch": 14.002278481012658,
+      "grad_norm": 0.12505175173282623,
+      "learning_rate": 3.987341772151899e-05,
+      "loss": 0.0719,
+      "step": 4460
+    },
+    {
+      "epoch": 14.002911392405064,
+      "grad_norm": 1.747598648071289,
+      "learning_rate": 3.983825597749648e-05,
+      "loss": 0.0855,
+      "step": 4470
+    },
+    {
+      "epoch": 14.003544303797469,
+      "grad_norm": 0.03635944798588753,
+      "learning_rate": 3.980309423347398e-05,
+      "loss": 0.1261,
+      "step": 4480
+    },
+    {
+      "epoch": 14.004177215189873,
+      "grad_norm": 1.7012404203414917,
+      "learning_rate": 3.976793248945147e-05,
+      "loss": 0.1321,
+      "step": 4490
+    },
+    {
+      "epoch": 14.004810126582278,
+      "grad_norm": 0.040289971977472305,
+      "learning_rate": 3.973277074542898e-05,
+      "loss": 0.0378,
+      "step": 4500
+    },
+    {
+      "epoch": 14.005443037974684,
+      "grad_norm": 2.7338550090789795,
+      "learning_rate": 3.969760900140647e-05,
+      "loss": 0.0765,
+      "step": 4510
+    },
+    {
+      "epoch": 14.006075949367089,
+      "grad_norm": 3.6036548614501953,
+      "learning_rate": 3.966244725738397e-05,
+      "loss": 0.1414,
+      "step": 4520
+    },
+    {
+      "epoch": 14.006708860759494,
+      "grad_norm": 0.07101106643676758,
+      "learning_rate": 3.9627285513361467e-05,
+      "loss": 0.0402,
+      "step": 4530
+    },
+    {
+      "epoch": 14.007341772151898,
+      "grad_norm": 2.1385743618011475,
+      "learning_rate": 3.9592123769338965e-05,
+      "loss": 0.082,
+      "step": 4540
+    },
+    {
+      "epoch": 14.007974683544305,
+      "grad_norm": 1.6399197578430176,
+      "learning_rate": 3.9556962025316456e-05,
+      "loss": 0.1278,
+      "step": 4550
+    },
+    {
+      "epoch": 14.00860759493671,
+      "grad_norm": 0.14580035209655762,
+      "learning_rate": 3.9521800281293955e-05,
+      "loss": 0.097,
+      "step": 4560
+    },
+    {
+      "epoch": 14.009240506329114,
+      "grad_norm": 0.044299304485321045,
+      "learning_rate": 3.948663853727145e-05,
+      "loss": 0.1258,
+      "step": 4570
+    },
+    {
+      "epoch": 14.009873417721519,
+      "grad_norm": 0.06222519651055336,
+      "learning_rate": 3.945147679324895e-05,
+      "loss": 0.0526,
+      "step": 4580
+    },
+    {
+      "epoch": 14.010506329113925,
+      "grad_norm": 2.3559911251068115,
+      "learning_rate": 3.941631504922644e-05,
+      "loss": 0.0934,
+      "step": 4590
+    },
+    {
+      "epoch": 14.01113924050633,
+      "grad_norm": 0.762189507484436,
+      "learning_rate": 3.938115330520394e-05,
+      "loss": 0.0937,
+      "step": 4600
+    },
+    {
+      "epoch": 14.011772151898734,
+      "grad_norm": 0.03613949567079544,
+      "learning_rate": 3.934599156118144e-05,
+      "loss": 0.1002,
+      "step": 4610
+    },
+    {
+      "epoch": 14.012405063291139,
+      "grad_norm": 0.0497271791100502,
+      "learning_rate": 3.931082981715893e-05,
+      "loss": 0.1768,
+      "step": 4620
+    },
+    {
+      "epoch": 14.013037974683545,
+      "grad_norm": 0.0881432294845581,
+      "learning_rate": 3.927566807313643e-05,
+      "loss": 0.111,
+      "step": 4630
+    },
+    {
+      "epoch": 14.01367088607595,
+      "grad_norm": 1.8317328691482544,
+      "learning_rate": 3.924050632911392e-05,
+      "loss": 0.0322,
+      "step": 4640
+    },
+    {
+      "epoch": 14.014303797468354,
+      "grad_norm": 0.1062496155500412,
+      "learning_rate": 3.9205344585091426e-05,
+      "loss": 0.0409,
+      "step": 4650
+    },
+    {
+      "epoch": 14.014936708860759,
+      "grad_norm": 0.05037945136427879,
+      "learning_rate": 3.917018284106892e-05,
+      "loss": 0.158,
+      "step": 4660
+    },
+    {
+      "epoch": 14.015569620253165,
+      "grad_norm": 1.9508050680160522,
+      "learning_rate": 3.9135021097046416e-05,
+      "loss": 0.1135,
+      "step": 4670
+    },
+    {
+      "epoch": 14.01620253164557,
+      "grad_norm": 0.03313547745347023,
+      "learning_rate": 3.909985935302391e-05,
+      "loss": 0.0667,
+      "step": 4680
+    },
+    {
+      "epoch": 14.016835443037975,
+      "grad_norm": 0.0592011958360672,
+      "learning_rate": 3.906469760900141e-05,
+      "loss": 0.096,
+      "step": 4690
+    },
+    {
+      "epoch": 14.01746835443038,
+      "grad_norm": 0.03745780512690544,
+      "learning_rate": 3.9029535864978904e-05,
+      "loss": 0.1084,
+      "step": 4700
+    },
+    {
+      "epoch": 14.018101265822784,
+      "grad_norm": 2.3232529163360596,
+      "learning_rate": 3.89943741209564e-05,
+      "loss": 0.0571,
+      "step": 4710
+    },
+    {
+      "epoch": 14.01873417721519,
+      "grad_norm": 0.03583463653922081,
+      "learning_rate": 3.8959212376933894e-05,
+      "loss": 0.0409,
+      "step": 4720
+    },
+    {
+      "epoch": 14.019367088607595,
+      "grad_norm": 3.5223681926727295,
+      "learning_rate": 3.89240506329114e-05,
+      "loss": 0.0784,
+      "step": 4730
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 0.07330437749624252,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.1228,
+      "step": 4740
+    },
+    {
+      "epoch": 14.02,
+      "eval_accuracy": 0.963531669865643,
+      "eval_loss": 0.08169866353273392,
+      "eval_runtime": 921.6107,
+      "eval_samples_per_second": 0.565,
+      "eval_steps_per_second": 0.072,
+      "step": 4740
+    },
+    {
+      "epoch": 15.000632911392405,
+      "grad_norm": 0.6237608194351196,
+      "learning_rate": 3.885372714486639e-05,
+      "loss": 0.0534,
+      "step": 4750
+    },
+    {
+      "epoch": 15.00126582278481,
+      "grad_norm": 1.1300462484359741,
+      "learning_rate": 3.881856540084388e-05,
+      "loss": 0.1765,
+      "step": 4760
+    },
+    {
+      "epoch": 15.001898734177216,
+      "grad_norm": 0.06267885118722916,
+      "learning_rate": 3.878340365682138e-05,
+      "loss": 0.0799,
+      "step": 4770
+    },
+    {
+      "epoch": 15.00253164556962,
+      "grad_norm": 0.0465877391397953,
+      "learning_rate": 3.874824191279888e-05,
+      "loss": 0.0156,
+      "step": 4780
+    },
+    {
+      "epoch": 15.003164556962025,
+      "grad_norm": 0.08922750502824783,
+      "learning_rate": 3.8713080168776376e-05,
+      "loss": 0.1283,
+      "step": 4790
+    },
+    {
+      "epoch": 15.00379746835443,
+      "grad_norm": 3.4005231857299805,
+      "learning_rate": 3.867791842475387e-05,
+      "loss": 0.0989,
+      "step": 4800
+    },
+    {
+      "epoch": 15.004430379746836,
+      "grad_norm": 0.07256881147623062,
+      "learning_rate": 3.8642756680731365e-05,
+      "loss": 0.099,
+      "step": 4810
+    },
+    {
+      "epoch": 15.00506329113924,
+      "grad_norm": 0.05675378069281578,
+      "learning_rate": 3.8607594936708864e-05,
+      "loss": 0.1094,
+      "step": 4820
+    },
+    {
+      "epoch": 15.005696202531645,
+      "grad_norm": 1.7418971061706543,
+      "learning_rate": 3.8572433192686355e-05,
+      "loss": 0.0712,
+      "step": 4830
+    },
+    {
+      "epoch": 15.00632911392405,
+      "grad_norm": 0.034461941570043564,
+      "learning_rate": 3.8537271448663854e-05,
+      "loss": 0.1568,
+      "step": 4840
+    },
+    {
+      "epoch": 15.006962025316456,
+      "grad_norm": 0.12834057211875916,
+      "learning_rate": 3.850210970464135e-05,
+      "loss": 0.071,
+      "step": 4850
+    },
+    {
+      "epoch": 15.00759493670886,
+      "grad_norm": 2.435114622116089,
+      "learning_rate": 3.846694796061885e-05,
+      "loss": 0.0736,
+      "step": 4860
+    },
+    {
+      "epoch": 15.008227848101265,
+      "grad_norm": 0.10457431524991989,
+      "learning_rate": 3.843178621659634e-05,
+      "loss": 0.0941,
+      "step": 4870
+    },
+    {
+      "epoch": 15.00886075949367,
+      "grad_norm": 1.1757208108901978,
+      "learning_rate": 3.839662447257384e-05,
+      "loss": 0.089,
+      "step": 4880
+    },
+    {
+      "epoch": 15.009493670886076,
+      "grad_norm": 0.11455094069242477,
+      "learning_rate": 3.836146272855134e-05,
+      "loss": 0.146,
+      "step": 4890
+    },
+    {
+      "epoch": 15.010126582278481,
+      "grad_norm": 0.047735344618558884,
+      "learning_rate": 3.832630098452884e-05,
+      "loss": 0.0979,
+      "step": 4900
+    },
+    {
+      "epoch": 15.010759493670886,
+      "grad_norm": 2.26092267036438,
+      "learning_rate": 3.829113924050633e-05,
+      "loss": 0.0598,
+      "step": 4910
+    },
+    {
+      "epoch": 15.01139240506329,
+      "grad_norm": 0.04012997820973396,
+      "learning_rate": 3.825597749648383e-05,
+      "loss": 0.0603,
+      "step": 4920
+    },
+    {
+      "epoch": 15.012025316455697,
+      "grad_norm": 0.060564495623111725,
+      "learning_rate": 3.8220815752461325e-05,
+      "loss": 0.0079,
+      "step": 4930
+    },
+    {
+      "epoch": 15.012658227848101,
+      "grad_norm": 1.690059781074524,
+      "learning_rate": 3.8185654008438823e-05,
+      "loss": 0.1119,
+      "step": 4940
+    },
+    {
+      "epoch": 15.013291139240506,
+      "grad_norm": 0.04470343515276909,
+      "learning_rate": 3.8150492264416315e-05,
+      "loss": 0.0421,
+      "step": 4950
+    },
+    {
+      "epoch": 15.01392405063291,
+      "grad_norm": 2.4390032291412354,
+      "learning_rate": 3.811533052039381e-05,
+      "loss": 0.1052,
+      "step": 4960
+    },
+    {
+      "epoch": 15.014556962025317,
+      "grad_norm": 0.034953054040670395,
+      "learning_rate": 3.808016877637131e-05,
+      "loss": 0.0869,
+      "step": 4970
+    },
+    {
+      "epoch": 15.015189873417722,
+      "grad_norm": 1.8463102579116821,
+      "learning_rate": 3.80450070323488e-05,
+      "loss": 0.0601,
+      "step": 4980
+    },
+    {
+      "epoch": 15.015822784810126,
+      "grad_norm": 0.025022268295288086,
+      "learning_rate": 3.80098452883263e-05,
+      "loss": 0.1042,
+      "step": 4990
+    },
+    {
+      "epoch": 15.01645569620253,
+      "grad_norm": 2.0151307582855225,
+      "learning_rate": 3.79746835443038e-05,
+      "loss": 0.1235,
+      "step": 5000
+    },
+    {
+      "epoch": 15.017088607594937,
+      "grad_norm": 2.130545139312744,
+      "learning_rate": 3.79395218002813e-05,
+      "loss": 0.1464,
+      "step": 5010
+    },
+    {
+      "epoch": 15.017721518987342,
+      "grad_norm": 0.02256944589316845,
+      "learning_rate": 3.790436005625879e-05,
+      "loss": 0.057,
+      "step": 5020
+    },
+    {
+      "epoch": 15.018354430379746,
+      "grad_norm": 0.2824922204017639,
+      "learning_rate": 3.786919831223629e-05,
+      "loss": 0.0738,
+      "step": 5030
+    },
+    {
+      "epoch": 15.018987341772151,
+      "grad_norm": 0.2349025160074234,
+      "learning_rate": 3.7834036568213786e-05,
+      "loss": 0.125,
+      "step": 5040
+    },
+    {
+      "epoch": 15.019620253164558,
+      "grad_norm": 1.1086318492889404,
+      "learning_rate": 3.7798874824191285e-05,
+      "loss": 0.0143,
+      "step": 5050
+    },
+    {
+      "epoch": 15.02,
+      "eval_accuracy": 0.9692898272552783,
+      "eval_loss": 0.061516787856817245,
+      "eval_runtime": 919.3915,
+      "eval_samples_per_second": 0.567,
+      "eval_steps_per_second": 0.072,
+      "step": 5056
+    },
+    {
+      "epoch": 16.000253164556963,
+      "grad_norm": 0.026827219873666763,
+      "learning_rate": 3.7763713080168776e-05,
+      "loss": 0.0445,
+      "step": 5060
+    },
+    {
+      "epoch": 16.000886075949367,
+      "grad_norm": 0.04409428685903549,
+      "learning_rate": 3.7728551336146275e-05,
+      "loss": 0.0879,
+      "step": 5070
+    },
+    {
+      "epoch": 16.001518987341772,
+      "grad_norm": 0.038987431675195694,
+      "learning_rate": 3.769338959212377e-05,
+      "loss": 0.0599,
+      "step": 5080
+    },
+    {
+      "epoch": 16.002151898734176,
+      "grad_norm": 4.49123477935791,
+      "learning_rate": 3.765822784810127e-05,
+      "loss": 0.0703,
+      "step": 5090
+    },
+    {
+      "epoch": 16.00278481012658,
+      "grad_norm": 0.03677366301417351,
+      "learning_rate": 3.762306610407876e-05,
+      "loss": 0.075,
+      "step": 5100
+    },
+    {
+      "epoch": 16.003417721518986,
+      "grad_norm": 3.2981064319610596,
+      "learning_rate": 3.758790436005626e-05,
+      "loss": 0.1017,
+      "step": 5110
+    },
+    {
+      "epoch": 16.004050632911394,
+      "grad_norm": 1.8348667621612549,
+      "learning_rate": 3.755274261603376e-05,
+      "loss": 0.0831,
+      "step": 5120
+    },
+    {
+      "epoch": 16.0046835443038,
+      "grad_norm": 0.023824410513043404,
+      "learning_rate": 3.751758087201125e-05,
+      "loss": 0.0696,
+      "step": 5130
+    },
+    {
+      "epoch": 16.005316455696203,
+      "grad_norm": 0.04150988906621933,
+      "learning_rate": 3.748241912798875e-05,
+      "loss": 0.0828,
+      "step": 5140
+    },
+    {
+      "epoch": 16.005949367088608,
+      "grad_norm": 0.020287124440073967,
+      "learning_rate": 3.744725738396625e-05,
+      "loss": 0.0027,
+      "step": 5150
+    },
+    {
+      "epoch": 16.006582278481012,
+      "grad_norm": 0.999756932258606,
+      "learning_rate": 3.7412095639943746e-05,
+      "loss": 0.0737,
+      "step": 5160
+    },
+    {
+      "epoch": 16.007215189873417,
+      "grad_norm": 0.02667284570634365,
+      "learning_rate": 3.737693389592124e-05,
+      "loss": 0.0686,
+      "step": 5170
+    },
+    {
+      "epoch": 16.00784810126582,
+      "grad_norm": 2.912930727005005,
+      "learning_rate": 3.7341772151898736e-05,
+      "loss": 0.0712,
+      "step": 5180
+    },
+    {
+      "epoch": 16.008481012658226,
+      "grad_norm": 0.03943086788058281,
+      "learning_rate": 3.730661040787623e-05,
+      "loss": 0.1255,
+      "step": 5190
+    },
+    {
+      "epoch": 16.009113924050634,
+      "grad_norm": 0.02671695314347744,
+      "learning_rate": 3.727144866385373e-05,
+      "loss": 0.1072,
+      "step": 5200
+    },
+    {
+      "epoch": 16.00974683544304,
+      "grad_norm": 0.052541956305503845,
+      "learning_rate": 3.7236286919831224e-05,
+      "loss": 0.0429,
+      "step": 5210
+    },
+    {
+      "epoch": 16.010379746835444,
+      "grad_norm": 1.465926170349121,
+      "learning_rate": 3.720112517580872e-05,
+      "loss": 0.0737,
+      "step": 5220
+    },
+    {
+      "epoch": 16.01101265822785,
+      "grad_norm": 0.02939579077064991,
+      "learning_rate": 3.7165963431786214e-05,
+      "loss": 0.026,
+      "step": 5230
+    },
+    {
+      "epoch": 16.011645569620253,
+      "grad_norm": 15.934955596923828,
+      "learning_rate": 3.713080168776372e-05,
+      "loss": 0.1409,
+      "step": 5240
+    },
+    {
+      "epoch": 16.012278481012657,
+      "grad_norm": 0.3337193727493286,
+      "learning_rate": 3.709563994374121e-05,
+      "loss": 0.1064,
+      "step": 5250
+    },
+    {
+      "epoch": 16.012911392405062,
+      "grad_norm": 1.509924292564392,
+      "learning_rate": 3.706047819971871e-05,
+      "loss": 0.1198,
+      "step": 5260
+    },
+    {
+      "epoch": 16.013544303797467,
+      "grad_norm": 1.6039562225341797,
+      "learning_rate": 3.70253164556962e-05,
+      "loss": 0.1751,
+      "step": 5270
+    },
+    {
+      "epoch": 16.014177215189875,
+      "grad_norm": 0.024508150294423103,
+      "learning_rate": 3.6990154711673706e-05,
+      "loss": 0.2054,
+      "step": 5280
+    },
+    {
+      "epoch": 16.01481012658228,
+      "grad_norm": 0.10826098173856735,
+      "learning_rate": 3.69549929676512e-05,
+      "loss": 0.0953,
+      "step": 5290
+    },
+    {
+      "epoch": 16.015443037974684,
+      "grad_norm": 0.1340492069721222,
+      "learning_rate": 3.6919831223628695e-05,
+      "loss": 0.1661,
+      "step": 5300
+    },
+    {
+      "epoch": 16.01607594936709,
+      "grad_norm": 0.20521284639835358,
+      "learning_rate": 3.688466947960619e-05,
+      "loss": 0.0446,
+      "step": 5310
+    },
+    {
+      "epoch": 16.016708860759493,
+      "grad_norm": 2.418673276901245,
+      "learning_rate": 3.6849507735583685e-05,
+      "loss": 0.1078,
+      "step": 5320
+    },
+    {
+      "epoch": 16.017341772151898,
+      "grad_norm": 0.0307689867913723,
+      "learning_rate": 3.6814345991561184e-05,
+      "loss": 0.0744,
+      "step": 5330
+    },
+    {
+      "epoch": 16.017974683544303,
+      "grad_norm": 12.458963394165039,
+      "learning_rate": 3.6779184247538675e-05,
+      "loss": 0.1081,
+      "step": 5340
+    },
+    {
+      "epoch": 16.018607594936707,
+      "grad_norm": 0.07104455679655075,
+      "learning_rate": 3.674402250351618e-05,
+      "loss": 0.0946,
+      "step": 5350
+    },
+    {
+      "epoch": 16.019240506329115,
+      "grad_norm": 0.0781354159116745,
+      "learning_rate": 3.670886075949367e-05,
+      "loss": 0.0375,
+      "step": 5360
+    },
+    {
+      "epoch": 16.01987341772152,
+      "grad_norm": 0.041876938194036484,
+      "learning_rate": 3.667369901547117e-05,
+      "loss": 0.0621,
+      "step": 5370
+    },
+    {
+      "epoch": 16.02,
+      "eval_accuracy": 0.9596928982725528,
+      "eval_loss": 0.07680489867925644,
+      "eval_runtime": 887.3026,
+      "eval_samples_per_second": 0.587,
+      "eval_steps_per_second": 0.074,
+      "step": 5372
+    },
+    {
+      "epoch": 17.000506329113925,
+      "grad_norm": 0.03362284228205681,
+      "learning_rate": 3.663853727144866e-05,
+      "loss": 0.0939,
+      "step": 5380
+    },
+    {
+      "epoch": 17.00113924050633,
+      "grad_norm": 2.931182384490967,
+      "learning_rate": 3.660337552742617e-05,
+      "loss": 0.0942,
+      "step": 5390
+    },
+    {
+      "epoch": 17.001772151898734,
+      "grad_norm": 9.610499382019043,
+      "learning_rate": 3.656821378340366e-05,
+      "loss": 0.2421,
+      "step": 5400
+    },
+    {
+      "epoch": 17.00240506329114,
+      "grad_norm": 2.01033091545105,
+      "learning_rate": 3.653305203938116e-05,
+      "loss": 0.0545,
+      "step": 5410
+    },
+    {
+      "epoch": 17.003037974683544,
+      "grad_norm": 0.5791252851486206,
+      "learning_rate": 3.649789029535865e-05,
+      "loss": 0.0443,
+      "step": 5420
+    },
+    {
+      "epoch": 17.00367088607595,
+      "grad_norm": 2.2349610328674316,
+      "learning_rate": 3.646272855133615e-05,
+      "loss": 0.1949,
+      "step": 5430
+    },
+    {
+      "epoch": 17.004303797468353,
+      "grad_norm": 0.06409902125597,
+      "learning_rate": 3.6427566807313645e-05,
+      "loss": 0.0731,
+      "step": 5440
+    },
+    {
+      "epoch": 17.00493670886076,
+      "grad_norm": 0.11862684041261673,
+      "learning_rate": 3.639240506329114e-05,
+      "loss": 0.0298,
+      "step": 5450
+    },
+    {
+      "epoch": 17.005569620253166,
+      "grad_norm": 1.8723206520080566,
+      "learning_rate": 3.6357243319268635e-05,
+      "loss": 0.1438,
+      "step": 5460
+    },
+    {
+      "epoch": 17.00620253164557,
+      "grad_norm": 0.04506813734769821,
+      "learning_rate": 3.632208157524613e-05,
+      "loss": 0.1927,
+      "step": 5470
+    },
+    {
+      "epoch": 17.006835443037975,
+      "grad_norm": 0.07324240356683731,
+      "learning_rate": 3.628691983122363e-05,
+      "loss": 0.2084,
+      "step": 5480
+    },
+    {
+      "epoch": 17.00746835443038,
+      "grad_norm": 16.583223342895508,
+      "learning_rate": 3.625175808720113e-05,
+      "loss": 0.1256,
+      "step": 5490
+    },
+    {
+      "epoch": 17.008101265822784,
+      "grad_norm": 2.8780710697174072,
+      "learning_rate": 3.621659634317862e-05,
+      "loss": 0.1533,
+      "step": 5500
+    },
+    {
+      "epoch": 17.00873417721519,
+      "grad_norm": 0.049451783299446106,
+      "learning_rate": 3.618143459915612e-05,
+      "loss": 0.0673,
+      "step": 5510
+    },
+    {
+      "epoch": 17.009367088607593,
+      "grad_norm": 1.1349679231643677,
+      "learning_rate": 3.614627285513362e-05,
+      "loss": 0.009,
+      "step": 5520
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 0.08764008432626724,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 0.0392,
+      "step": 5530
+    },
+    {
+      "epoch": 17.010632911392406,
+      "grad_norm": 0.04414183646440506,
+      "learning_rate": 3.607594936708861e-05,
+      "loss": 0.1909,
+      "step": 5540
+    },
+    {
+      "epoch": 17.01126582278481,
+      "grad_norm": 2.099726676940918,
+      "learning_rate": 3.6040787623066106e-05,
+      "loss": 0.3287,
+      "step": 5550
+    },
+    {
+      "epoch": 17.011898734177215,
+      "grad_norm": 0.1605502963066101,
+      "learning_rate": 3.6005625879043604e-05,
+      "loss": 0.1195,
+      "step": 5560
+    },
+    {
+      "epoch": 17.01253164556962,
+      "grad_norm": 0.11115490645170212,
+      "learning_rate": 3.5970464135021096e-05,
+      "loss": 0.1011,
+      "step": 5570
+    },
+    {
+      "epoch": 17.013164556962025,
+      "grad_norm": 1.932108998298645,
+      "learning_rate": 3.5935302390998594e-05,
+      "loss": 0.1557,
+      "step": 5580
+    },
+    {
+      "epoch": 17.01379746835443,
+      "grad_norm": 0.022036854177713394,
+      "learning_rate": 3.590014064697609e-05,
+      "loss": 0.0526,
+      "step": 5590
+    },
+    {
+      "epoch": 17.014430379746834,
+      "grad_norm": 0.0311493631452322,
+      "learning_rate": 3.586497890295359e-05,
+      "loss": 0.1158,
+      "step": 5600
+    },
+    {
+      "epoch": 17.015063291139242,
+      "grad_norm": 0.37227359414100647,
+      "learning_rate": 3.582981715893108e-05,
+      "loss": 0.0607,
+      "step": 5610
+    },
+    {
+      "epoch": 17.015696202531647,
+      "grad_norm": 0.047853514552116394,
+      "learning_rate": 3.579465541490858e-05,
+      "loss": 0.1834,
+      "step": 5620
+    },
+    {
+      "epoch": 17.01632911392405,
+      "grad_norm": 0.798313558101654,
+      "learning_rate": 3.575949367088608e-05,
+      "loss": 0.0497,
+      "step": 5630
+    },
+    {
+      "epoch": 17.016962025316456,
+      "grad_norm": 0.020088857039809227,
+      "learning_rate": 3.572433192686358e-05,
+      "loss": 0.0975,
+      "step": 5640
+    },
+    {
+      "epoch": 17.01759493670886,
+      "grad_norm": 3.423915386199951,
+      "learning_rate": 3.568917018284107e-05,
+      "loss": 0.1255,
+      "step": 5650
+    },
+    {
+      "epoch": 17.018227848101265,
+      "grad_norm": 2.1361560821533203,
+      "learning_rate": 3.565400843881857e-05,
+      "loss": 0.1294,
+      "step": 5660
+    },
+    {
+      "epoch": 17.01886075949367,
+      "grad_norm": 3.7424023151397705,
+      "learning_rate": 3.5618846694796066e-05,
+      "loss": 0.21,
+      "step": 5670
+    },
+    {
+      "epoch": 17.019493670886074,
+      "grad_norm": 0.9021095633506775,
+      "learning_rate": 3.558368495077356e-05,
+      "loss": 0.0597,
+      "step": 5680
+    },
+    {
+      "epoch": 17.02,
+      "eval_accuracy": 0.963531669865643,
+      "eval_loss": 0.08725160360336304,
+      "eval_runtime": 887.9543,
+      "eval_samples_per_second": 0.587,
+      "eval_steps_per_second": 0.074,
+      "step": 5688
+    },
+    {
+      "epoch": 18.00012658227848,
+      "grad_norm": 9.254701614379883,
+      "learning_rate": 3.5548523206751056e-05,
+      "loss": 0.0679,
+      "step": 5690
+    },
+    {
+      "epoch": 18.000759493670888,
+      "grad_norm": 0.09168372303247452,
+      "learning_rate": 3.551336146272855e-05,
+      "loss": 0.0487,
+      "step": 5700
+    },
+    {
+      "epoch": 18.001392405063292,
+      "grad_norm": 1.7651177644729614,
+      "learning_rate": 3.547819971870605e-05,
+      "loss": 0.1359,
+      "step": 5710
+    },
+    {
+      "epoch": 18.002025316455697,
+      "grad_norm": 1.6308825016021729,
+      "learning_rate": 3.5443037974683544e-05,
+      "loss": 0.0374,
+      "step": 5720
+    },
+    {
+      "epoch": 18.0026582278481,
+      "grad_norm": 0.029071198776364326,
+      "learning_rate": 3.540787623066104e-05,
+      "loss": 0.0867,
+      "step": 5730
+    },
+    {
+      "epoch": 18.003291139240506,
+      "grad_norm": 1.1404354572296143,
+      "learning_rate": 3.5372714486638534e-05,
+      "loss": 0.1032,
+      "step": 5740
+    },
+    {
+      "epoch": 18.00392405063291,
+      "grad_norm": 0.060701146721839905,
+      "learning_rate": 3.533755274261604e-05,
+      "loss": 0.0807,
+      "step": 5750
+    },
+    {
+      "epoch": 18.004556962025315,
+      "grad_norm": 0.023883167654275894,
+      "learning_rate": 3.530239099859353e-05,
+      "loss": 0.0917,
+      "step": 5760
+    },
+    {
+      "epoch": 18.00518987341772,
+      "grad_norm": 2.6466054916381836,
+      "learning_rate": 3.526722925457103e-05,
+      "loss": 0.078,
+      "step": 5770
+    },
+    {
+      "epoch": 18.005822784810128,
+      "grad_norm": 0.2958976626396179,
+      "learning_rate": 3.523206751054853e-05,
+      "loss": 0.0468,
+      "step": 5780
+    },
+    {
+      "epoch": 18.006455696202533,
+      "grad_norm": 0.08598524332046509,
+      "learning_rate": 3.5196905766526025e-05,
+      "loss": 0.0925,
+      "step": 5790
+    },
+    {
+      "epoch": 18.007088607594937,
+      "grad_norm": 0.02307640202343464,
+      "learning_rate": 3.516174402250352e-05,
+      "loss": 0.0784,
+      "step": 5800
+    },
+    {
+      "epoch": 18.007721518987342,
+      "grad_norm": 1.9384537935256958,
+      "learning_rate": 3.5126582278481015e-05,
+      "loss": 0.0645,
+      "step": 5810
+    },
+    {
+      "epoch": 18.008354430379747,
+      "grad_norm": 0.06027079373598099,
+      "learning_rate": 3.5091420534458513e-05,
+      "loss": 0.1358,
+      "step": 5820
+    },
+    {
+      "epoch": 18.00898734177215,
+      "grad_norm": 1.6383880376815796,
+      "learning_rate": 3.505625879043601e-05,
+      "loss": 0.0505,
+      "step": 5830
+    },
+    {
+      "epoch": 18.009620253164556,
+      "grad_norm": 1.355348825454712,
+      "learning_rate": 3.50210970464135e-05,
+      "loss": 0.0574,
+      "step": 5840
+    },
+    {
+      "epoch": 18.01025316455696,
+      "grad_norm": 2.113496780395508,
+      "learning_rate": 3.4985935302391e-05,
+      "loss": 0.1005,
+      "step": 5850
+    },
+    {
+      "epoch": 18.01088607594937,
+      "grad_norm": 0.022542107850313187,
+      "learning_rate": 3.49507735583685e-05,
+      "loss": 0.0671,
+      "step": 5860
+    },
+    {
+      "epoch": 18.011518987341773,
+      "grad_norm": 0.05946587771177292,
+      "learning_rate": 3.491561181434599e-05,
+      "loss": 0.0933,
+      "step": 5870
+    },
+    {
+      "epoch": 18.012151898734178,
+      "grad_norm": 1.936155915260315,
+      "learning_rate": 3.488045007032349e-05,
+      "loss": 0.0975,
+      "step": 5880
+    },
+    {
+      "epoch": 18.012784810126583,
+      "grad_norm": 0.03399858996272087,
+      "learning_rate": 3.484528832630098e-05,
+      "loss": 0.0909,
+      "step": 5890
+    },
+    {
+      "epoch": 18.013417721518987,
+      "grad_norm": 0.030509311705827713,
+      "learning_rate": 3.4810126582278487e-05,
+      "loss": 0.0685,
+      "step": 5900
+    },
+    {
+      "epoch": 18.014050632911392,
+      "grad_norm": 0.016794001683592796,
+      "learning_rate": 3.477496483825598e-05,
+      "loss": 0.0481,
+      "step": 5910
+    },
+    {
+      "epoch": 18.014683544303796,
+      "grad_norm": 2.368962049484253,
+      "learning_rate": 3.4739803094233476e-05,
+      "loss": 0.2066,
+      "step": 5920
+    },
+    {
+      "epoch": 18.0153164556962,
+      "grad_norm": 1.5589197874069214,
+      "learning_rate": 3.470464135021097e-05,
+      "loss": 0.0798,
+      "step": 5930
+    },
+    {
+      "epoch": 18.01594936708861,
+      "grad_norm": 2.055568218231201,
+      "learning_rate": 3.466947960618847e-05,
+      "loss": 0.0384,
+      "step": 5940
+    },
+    {
+      "epoch": 18.016582278481014,
+      "grad_norm": 2.212440013885498,
+      "learning_rate": 3.4634317862165965e-05,
+      "loss": 0.0979,
+      "step": 5950
+    },
+    {
+      "epoch": 18.01721518987342,
+      "grad_norm": 0.022211147472262383,
+      "learning_rate": 3.459915611814346e-05,
+      "loss": 0.0575,
+      "step": 5960
+    },
+    {
+      "epoch": 18.017848101265823,
+      "grad_norm": 2.110724687576294,
+      "learning_rate": 3.4563994374120954e-05,
+      "loss": 0.0568,
+      "step": 5970
+    },
+    {
+      "epoch": 18.018481012658228,
+      "grad_norm": 0.025756409391760826,
+      "learning_rate": 3.452883263009846e-05,
+      "loss": 0.0429,
+      "step": 5980
+    },
+    {
+      "epoch": 18.019113924050632,
+      "grad_norm": 2.1070611476898193,
+      "learning_rate": 3.449367088607595e-05,
+      "loss": 0.0821,
+      "step": 5990
+    },
+    {
+      "epoch": 18.019746835443037,
+      "grad_norm": 0.043322015553712845,
+      "learning_rate": 3.445850914205345e-05,
+      "loss": 0.0696,
+      "step": 6000
+    },
+    {
+      "epoch": 18.02,
+      "eval_accuracy": 0.9539347408829175,
+      "eval_loss": 0.11077545583248138,
+      "eval_runtime": 890.6431,
+      "eval_samples_per_second": 0.585,
+      "eval_steps_per_second": 0.074,
+      "step": 6004
+    },
+    {
+      "epoch": 19.000379746835442,
+      "grad_norm": 1.7157703638076782,
+      "learning_rate": 3.442334739803094e-05,
+      "loss": 0.184,
+      "step": 6010
+    },
+    {
+      "epoch": 19.001012658227847,
+      "grad_norm": 0.03436505049467087,
+      "learning_rate": 3.438818565400844e-05,
+      "loss": 0.1121,
+      "step": 6020
+    },
+    {
+      "epoch": 19.001645569620255,
+      "grad_norm": 0.02937289886176586,
+      "learning_rate": 3.435302390998594e-05,
+      "loss": 0.1085,
+      "step": 6030
+    },
+    {
+      "epoch": 19.00227848101266,
+      "grad_norm": 2.090529680252075,
+      "learning_rate": 3.431786216596343e-05,
+      "loss": 0.1076,
+      "step": 6040
+    },
+    {
+      "epoch": 19.002911392405064,
+      "grad_norm": 2.621086835861206,
+      "learning_rate": 3.428270042194093e-05,
+      "loss": 0.0475,
+      "step": 6050
+    },
+    {
+      "epoch": 19.00354430379747,
+      "grad_norm": 0.081081323325634,
+      "learning_rate": 3.4247538677918426e-05,
+      "loss": 0.1206,
+      "step": 6060
+    },
+    {
+      "epoch": 19.004177215189873,
+      "grad_norm": 0.9470970034599304,
+      "learning_rate": 3.4212376933895924e-05,
+      "loss": 0.0197,
+      "step": 6070
+    },
+    {
+      "epoch": 19.004810126582278,
+      "grad_norm": 0.6718599200248718,
+      "learning_rate": 3.4177215189873416e-05,
+      "loss": 0.0415,
+      "step": 6080
+    },
+    {
+      "epoch": 19.005443037974683,
+      "grad_norm": 0.017539095133543015,
+      "learning_rate": 3.4142053445850914e-05,
+      "loss": 0.0924,
+      "step": 6090
+    },
+    {
+      "epoch": 19.006075949367087,
+      "grad_norm": 0.04726846516132355,
+      "learning_rate": 3.410689170182841e-05,
+      "loss": 0.0091,
+      "step": 6100
+    },
+    {
+      "epoch": 19.006708860759495,
+      "grad_norm": 1.5941773653030396,
+      "learning_rate": 3.407172995780591e-05,
+      "loss": 0.1435,
+      "step": 6110
+    },
+    {
+      "epoch": 19.0073417721519,
+      "grad_norm": 2.2858848571777344,
+      "learning_rate": 3.40365682137834e-05,
+      "loss": 0.0776,
+      "step": 6120
+    },
+    {
+      "epoch": 19.007974683544305,
+      "grad_norm": 0.03234981372952461,
+      "learning_rate": 3.40014064697609e-05,
+      "loss": 0.0483,
+      "step": 6130
+    },
+    {
+      "epoch": 19.00860759493671,
+      "grad_norm": 0.07574064284563065,
+      "learning_rate": 3.39662447257384e-05,
+      "loss": 0.0784,
+      "step": 6140
+    },
+    {
+      "epoch": 19.009240506329114,
+      "grad_norm": 0.016080401837825775,
+      "learning_rate": 3.39310829817159e-05,
+      "loss": 0.0479,
+      "step": 6150
+    },
+    {
+      "epoch": 19.00987341772152,
+      "grad_norm": 0.546974241733551,
+      "learning_rate": 3.389592123769339e-05,
+      "loss": 0.0931,
+      "step": 6160
+    },
+    {
+      "epoch": 19.010506329113923,
+      "grad_norm": 18.685945510864258,
+      "learning_rate": 3.386075949367089e-05,
+      "loss": 0.0846,
+      "step": 6170
+    },
+    {
+      "epoch": 19.011139240506328,
+      "grad_norm": 3.7618796825408936,
+      "learning_rate": 3.3825597749648385e-05,
+      "loss": 0.1187,
+      "step": 6180
+    },
+    {
+      "epoch": 19.011772151898736,
+      "grad_norm": 3.0678319931030273,
+      "learning_rate": 3.3790436005625884e-05,
+      "loss": 0.2192,
+      "step": 6190
+    },
+    {
+      "epoch": 19.01240506329114,
+      "grad_norm": 0.021779673174023628,
+      "learning_rate": 3.3755274261603375e-05,
+      "loss": 0.0531,
+      "step": 6200
+    },
+    {
+      "epoch": 19.013037974683545,
+      "grad_norm": 2.7021806240081787,
+      "learning_rate": 3.3720112517580874e-05,
+      "loss": 0.0932,
+      "step": 6210
+    },
+    {
+      "epoch": 19.01367088607595,
+      "grad_norm": 2.251648426055908,
+      "learning_rate": 3.368495077355837e-05,
+      "loss": 0.1122,
+      "step": 6220
+    },
+    {
+      "epoch": 19.014303797468354,
+      "grad_norm": 18.46161651611328,
+      "learning_rate": 3.3649789029535864e-05,
+      "loss": 0.2242,
+      "step": 6230
+    },
+    {
+      "epoch": 19.01493670886076,
+      "grad_norm": 1.9724690914154053,
+      "learning_rate": 3.361462728551336e-05,
+      "loss": 0.1283,
+      "step": 6240
+    },
+    {
+      "epoch": 19.015569620253164,
+      "grad_norm": 1.8774776458740234,
+      "learning_rate": 3.357946554149086e-05,
+      "loss": 0.0629,
+      "step": 6250
+    },
+    {
+      "epoch": 19.01620253164557,
+      "grad_norm": 0.08820010721683502,
+      "learning_rate": 3.354430379746836e-05,
+      "loss": 0.0487,
+      "step": 6260
+    },
+    {
+      "epoch": 19.016835443037976,
+      "grad_norm": 0.9228050708770752,
+      "learning_rate": 3.350914205344585e-05,
+      "loss": 0.1959,
+      "step": 6270
+    },
+    {
+      "epoch": 19.01746835443038,
+      "grad_norm": 0.061690233647823334,
+      "learning_rate": 3.347398030942335e-05,
+      "loss": 0.1182,
+      "step": 6280
+    },
+    {
+      "epoch": 19.018101265822786,
+      "grad_norm": 23.2619686126709,
+      "learning_rate": 3.343881856540085e-05,
+      "loss": 0.1069,
+      "step": 6290
+    },
+    {
+      "epoch": 19.01873417721519,
+      "grad_norm": 0.0341399684548378,
+      "learning_rate": 3.3403656821378345e-05,
+      "loss": 0.1889,
+      "step": 6300
+    },
+    {
+      "epoch": 19.019367088607595,
+      "grad_norm": 14.88992691040039,
+      "learning_rate": 3.3368495077355837e-05,
+      "loss": 0.1285,
+      "step": 6310
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 0.3174104392528534,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.2761,
+      "step": 6320
+    },
+    {
+      "epoch": 19.02,
+      "eval_accuracy": 0.9520153550863724,
+      "eval_loss": 0.14126819372177124,
+      "eval_runtime": 862.4572,
+      "eval_samples_per_second": 0.604,
+      "eval_steps_per_second": 0.077,
+      "step": 6320
+    },
+    {
+      "epoch": 20.000632911392405,
+      "grad_norm": 15.443863868713379,
+      "learning_rate": 3.329817158931083e-05,
+      "loss": 0.2663,
+      "step": 6330
+    },
+    {
+      "epoch": 20.00126582278481,
+      "grad_norm": 0.15935222804546356,
+      "learning_rate": 3.326300984528833e-05,
+      "loss": 0.0412,
+      "step": 6340
+    },
+    {
+      "epoch": 20.001898734177214,
+      "grad_norm": 2.3951077461242676,
+      "learning_rate": 3.322784810126582e-05,
+      "loss": 0.1944,
+      "step": 6350
+    },
+    {
+      "epoch": 20.00253164556962,
+      "grad_norm": 0.055507030338048935,
+      "learning_rate": 3.319268635724332e-05,
+      "loss": 0.0692,
+      "step": 6360
+    },
+    {
+      "epoch": 20.003164556962027,
+      "grad_norm": 15.818922996520996,
+      "learning_rate": 3.315752461322082e-05,
+      "loss": 0.1544,
+      "step": 6370
+    },
+    {
+      "epoch": 20.00379746835443,
+      "grad_norm": 0.0613020621240139,
+      "learning_rate": 3.312236286919831e-05,
+      "loss": 0.0584,
+      "step": 6380
+    },
+    {
+      "epoch": 20.004430379746836,
+      "grad_norm": 1.5182740688323975,
+      "learning_rate": 3.308720112517581e-05,
+      "loss": 0.0507,
+      "step": 6390
+    },
+    {
+      "epoch": 20.00506329113924,
+      "grad_norm": 0.17965848743915558,
+      "learning_rate": 3.305203938115331e-05,
+      "loss": 0.1409,
+      "step": 6400
+    },
+    {
+      "epoch": 20.005696202531645,
+      "grad_norm": 13.088379859924316,
+      "learning_rate": 3.3016877637130806e-05,
+      "loss": 0.1933,
+      "step": 6410
+    },
+    {
+      "epoch": 20.00632911392405,
+      "grad_norm": 3.075198173522949,
+      "learning_rate": 3.29817158931083e-05,
+      "loss": 0.0366,
+      "step": 6420
+    },
+    {
+      "epoch": 20.006962025316454,
+      "grad_norm": 0.25585731863975525,
+      "learning_rate": 3.2946554149085796e-05,
+      "loss": 0.0869,
+      "step": 6430
+    },
+    {
+      "epoch": 20.00759493670886,
+      "grad_norm": 1.9562506675720215,
+      "learning_rate": 3.291139240506329e-05,
+      "loss": 0.0486,
+      "step": 6440
+    },
+    {
+      "epoch": 20.008227848101267,
+      "grad_norm": 1.8931715488433838,
+      "learning_rate": 3.287623066104079e-05,
+      "loss": 0.1142,
+      "step": 6450
+    },
+    {
+      "epoch": 20.008860759493672,
+      "grad_norm": 1.9258003234863281,
+      "learning_rate": 3.2841068917018284e-05,
+      "loss": 0.0716,
+      "step": 6460
+    },
+    {
+      "epoch": 20.009493670886076,
+      "grad_norm": 0.2476690262556076,
+      "learning_rate": 3.280590717299578e-05,
+      "loss": 0.0593,
+      "step": 6470
+    },
+    {
+      "epoch": 20.01012658227848,
+      "grad_norm": 0.041350286453962326,
+      "learning_rate": 3.2770745428973274e-05,
+      "loss": 0.0521,
+      "step": 6480
+    },
+    {
+      "epoch": 20.010759493670886,
+      "grad_norm": 6.784064769744873,
+      "learning_rate": 3.273558368495078e-05,
+      "loss": 0.0805,
+      "step": 6490
+    },
+    {
+      "epoch": 20.01139240506329,
+      "grad_norm": 1.9818538427352905,
+      "learning_rate": 3.270042194092827e-05,
+      "loss": 0.0722,
+      "step": 6500
+    },
+    {
+      "epoch": 20.012025316455695,
+      "grad_norm": 2.0829291343688965,
+      "learning_rate": 3.266526019690577e-05,
+      "loss": 0.1185,
+      "step": 6510
+    },
+    {
+      "epoch": 20.0126582278481,
+      "grad_norm": 0.06409807503223419,
+      "learning_rate": 3.263009845288326e-05,
+      "loss": 0.0777,
+      "step": 6520
+    },
+    {
+      "epoch": 20.013291139240508,
+      "grad_norm": 0.052046943455934525,
+      "learning_rate": 3.2594936708860766e-05,
+      "loss": 0.053,
+      "step": 6530
+    },
+    {
+      "epoch": 20.013924050632912,
+      "grad_norm": 6.423180103302002,
+      "learning_rate": 3.255977496483826e-05,
+      "loss": 0.1333,
+      "step": 6540
+    },
+    {
+      "epoch": 20.014556962025317,
+      "grad_norm": 0.04574073851108551,
+      "learning_rate": 3.2524613220815756e-05,
+      "loss": 0.1059,
+      "step": 6550
+    },
+    {
+      "epoch": 20.01518987341772,
+      "grad_norm": 5.25015115737915,
+      "learning_rate": 3.248945147679325e-05,
+      "loss": 0.1471,
+      "step": 6560
+    },
+    {
+      "epoch": 20.015822784810126,
+      "grad_norm": 0.036584123969078064,
+      "learning_rate": 3.2454289732770746e-05,
+      "loss": 0.2641,
+      "step": 6570
+    },
+    {
+      "epoch": 20.01645569620253,
+      "grad_norm": 10.215600967407227,
+      "learning_rate": 3.2419127988748244e-05,
+      "loss": 0.204,
+      "step": 6580
+    },
+    {
+      "epoch": 20.017088607594935,
+      "grad_norm": 0.7967722415924072,
+      "learning_rate": 3.2383966244725736e-05,
+      "loss": 0.1105,
+      "step": 6590
+    },
+    {
+      "epoch": 20.01772151898734,
+      "grad_norm": 28.42278480529785,
+      "learning_rate": 3.234880450070324e-05,
+      "loss": 0.1647,
+      "step": 6600
+    },
+    {
+      "epoch": 20.01835443037975,
+      "grad_norm": 0.14231210947036743,
+      "learning_rate": 3.231364275668073e-05,
+      "loss": 0.0553,
+      "step": 6610
+    },
+    {
+      "epoch": 20.018987341772153,
+      "grad_norm": 0.05637258663773537,
+      "learning_rate": 3.227848101265823e-05,
+      "loss": 0.074,
+      "step": 6620
+    },
+    {
+      "epoch": 20.019620253164558,
+      "grad_norm": 0.2689402401447296,
+      "learning_rate": 3.224331926863572e-05,
+      "loss": 0.129,
+      "step": 6630
+    },
+    {
+      "epoch": 20.02,
+      "eval_accuracy": 0.9520153550863724,
+      "eval_loss": 0.1470969319343567,
+      "eval_runtime": 903.3982,
+      "eval_samples_per_second": 0.577,
+      "eval_steps_per_second": 0.073,
+      "step": 6636
+    },
+    {
+      "epoch": 21.000253164556963,
+      "grad_norm": 1.2210781574249268,
+      "learning_rate": 3.220815752461323e-05,
+      "loss": 0.1749,
+      "step": 6640
+    },
+    {
+      "epoch": 21.000886075949367,
+      "grad_norm": 1.918101191520691,
+      "learning_rate": 3.217299578059072e-05,
+      "loss": 0.109,
+      "step": 6650
+    },
+    {
+      "epoch": 21.001518987341772,
+      "grad_norm": 1.1142789125442505,
+      "learning_rate": 3.213783403656822e-05,
+      "loss": 0.2533,
+      "step": 6660
+    },
+    {
+      "epoch": 21.002151898734176,
+      "grad_norm": 2.3127224445343018,
+      "learning_rate": 3.210267229254571e-05,
+      "loss": 0.1336,
+      "step": 6670
+    },
+    {
+      "epoch": 21.00278481012658,
+      "grad_norm": 2.306687116622925,
+      "learning_rate": 3.2067510548523214e-05,
+      "loss": 0.1947,
+      "step": 6680
+    },
+    {
+      "epoch": 21.003417721518986,
+      "grad_norm": 0.024246342480182648,
+      "learning_rate": 3.2032348804500705e-05,
+      "loss": 0.0446,
+      "step": 6690
+    },
+    {
+      "epoch": 21.004050632911394,
+      "grad_norm": 0.0923205316066742,
+      "learning_rate": 3.1997187060478204e-05,
+      "loss": 0.0261,
+      "step": 6700
+    },
+    {
+      "epoch": 21.0046835443038,
+      "grad_norm": 0.020252803340554237,
+      "learning_rate": 3.1962025316455695e-05,
+      "loss": 0.0664,
+      "step": 6710
+    },
+    {
+      "epoch": 21.005316455696203,
+      "grad_norm": 1.8104360103607178,
+      "learning_rate": 3.1926863572433193e-05,
+      "loss": 0.0501,
+      "step": 6720
+    },
+    {
+      "epoch": 21.005949367088608,
+      "grad_norm": 0.13536378741264343,
+      "learning_rate": 3.189170182841069e-05,
+      "loss": 0.1107,
+      "step": 6730
+    },
+    {
+      "epoch": 21.006582278481012,
+      "grad_norm": 0.023490697145462036,
+      "learning_rate": 3.185654008438819e-05,
+      "loss": 0.0581,
+      "step": 6740
+    },
+    {
+      "epoch": 21.007215189873417,
+      "grad_norm": 0.011167807504534721,
+      "learning_rate": 3.182137834036568e-05,
+      "loss": 0.0835,
+      "step": 6750
+    },
+    {
+      "epoch": 21.00784810126582,
+      "grad_norm": 2.162050485610962,
+      "learning_rate": 3.178621659634318e-05,
+      "loss": 0.0804,
+      "step": 6760
+    },
+    {
+      "epoch": 21.008481012658226,
+      "grad_norm": 0.026973918080329895,
+      "learning_rate": 3.175105485232068e-05,
+      "loss": 0.0677,
+      "step": 6770
+    },
+    {
+      "epoch": 21.009113924050634,
+      "grad_norm": 2.1330974102020264,
+      "learning_rate": 3.171589310829817e-05,
+      "loss": 0.1016,
+      "step": 6780
+    },
+    {
+      "epoch": 21.00974683544304,
+      "grad_norm": 1.0756934881210327,
+      "learning_rate": 3.168073136427567e-05,
+      "loss": 0.0232,
+      "step": 6790
+    },
+    {
+      "epoch": 21.010379746835444,
+      "grad_norm": 18.678178787231445,
+      "learning_rate": 3.1645569620253167e-05,
+      "loss": 0.0977,
+      "step": 6800
+    },
+    {
+      "epoch": 21.01101265822785,
+      "grad_norm": 1.5976709127426147,
+      "learning_rate": 3.1610407876230665e-05,
+      "loss": 0.069,
+      "step": 6810
+    },
+    {
+      "epoch": 21.011645569620253,
+      "grad_norm": 0.5849307775497437,
+      "learning_rate": 3.1575246132208156e-05,
+      "loss": 0.1015,
+      "step": 6820
+    },
+    {
+      "epoch": 21.012278481012657,
+      "grad_norm": 3.8058197498321533,
+      "learning_rate": 3.1540084388185655e-05,
+      "loss": 0.1165,
+      "step": 6830
+    },
+    {
+      "epoch": 21.012911392405062,
+      "grad_norm": 1.8044683933258057,
+      "learning_rate": 3.150492264416315e-05,
+      "loss": 0.0801,
+      "step": 6840
+    },
+    {
+      "epoch": 21.013544303797467,
+      "grad_norm": 1.89737868309021,
+      "learning_rate": 3.146976090014065e-05,
+      "loss": 0.0453,
+      "step": 6850
+    },
+    {
+      "epoch": 21.014177215189875,
+      "grad_norm": 2.135648250579834,
+      "learning_rate": 3.143459915611814e-05,
+      "loss": 0.0314,
+      "step": 6860
+    },
+    {
+      "epoch": 21.01481012658228,
+      "grad_norm": 0.015429515391588211,
+      "learning_rate": 3.139943741209564e-05,
+      "loss": 0.1356,
+      "step": 6870
+    },
+    {
+      "epoch": 21.015443037974684,
+      "grad_norm": 0.029538467526435852,
+      "learning_rate": 3.136427566807314e-05,
+      "loss": 0.0375,
+      "step": 6880
+    },
+    {
+      "epoch": 21.01607594936709,
+      "grad_norm": 0.013729414902627468,
+      "learning_rate": 3.132911392405064e-05,
+      "loss": 0.154,
+      "step": 6890
+    },
+    {
+      "epoch": 21.016708860759493,
+      "grad_norm": 1.6405558586120605,
+      "learning_rate": 3.129395218002813e-05,
+      "loss": 0.0856,
+      "step": 6900
+    },
+    {
+      "epoch": 21.017341772151898,
+      "grad_norm": 0.02224159613251686,
+      "learning_rate": 3.125879043600563e-05,
+      "loss": 0.0482,
+      "step": 6910
+    },
+    {
+      "epoch": 21.017974683544303,
+      "grad_norm": 0.024815354496240616,
+      "learning_rate": 3.1223628691983126e-05,
+      "loss": 0.0808,
+      "step": 6920
+    },
+    {
+      "epoch": 21.018607594936707,
+      "grad_norm": 2.0164098739624023,
+      "learning_rate": 3.118846694796062e-05,
+      "loss": 0.0392,
+      "step": 6930
+    },
+    {
+      "epoch": 21.019240506329115,
+      "grad_norm": 1.8977909088134766,
+      "learning_rate": 3.1153305203938116e-05,
+      "loss": 0.0415,
+      "step": 6940
+    },
+    {
+      "epoch": 21.01987341772152,
+      "grad_norm": 2.1623473167419434,
+      "learning_rate": 3.111814345991561e-05,
+      "loss": 0.0828,
+      "step": 6950
+    },
+    {
+      "epoch": 21.02,
+      "eval_accuracy": 0.9673704414587332,
+      "eval_loss": 0.06080710142850876,
+      "eval_runtime": 897.4486,
+      "eval_samples_per_second": 0.581,
+      "eval_steps_per_second": 0.074,
+      "step": 6952
+    },
+    {
+      "epoch": 22.000506329113925,
+      "grad_norm": 1.7337490320205688,
+      "learning_rate": 3.108298171589311e-05,
+      "loss": 0.0741,
+      "step": 6960
+    },
+    {
+      "epoch": 22.00113924050633,
+      "grad_norm": 0.029260125011205673,
+      "learning_rate": 3.1047819971870604e-05,
+      "loss": 0.0384,
+      "step": 6970
+    },
+    {
+      "epoch": 22.001772151898734,
+      "grad_norm": 0.014634775929152966,
+      "learning_rate": 3.10126582278481e-05,
+      "loss": 0.0472,
+      "step": 6980
+    },
+    {
+      "epoch": 22.00240506329114,
+      "grad_norm": 0.031179388985037804,
+      "learning_rate": 3.0977496483825594e-05,
+      "loss": 0.0498,
+      "step": 6990
+    },
+    {
+      "epoch": 22.003037974683544,
+      "grad_norm": 2.017439126968384,
+      "learning_rate": 3.09423347398031e-05,
+      "loss": 0.0726,
+      "step": 7000
+    },
+    {
+      "epoch": 22.00367088607595,
+      "grad_norm": 0.01198955811560154,
+      "learning_rate": 3.090717299578059e-05,
+      "loss": 0.0546,
+      "step": 7010
+    },
+    {
+      "epoch": 22.004303797468353,
+      "grad_norm": 1.72632896900177,
+      "learning_rate": 3.087201125175809e-05,
+      "loss": 0.0686,
+      "step": 7020
+    },
+    {
+      "epoch": 22.00493670886076,
+      "grad_norm": 0.013193360529839993,
+      "learning_rate": 3.083684950773559e-05,
+      "loss": 0.0543,
+      "step": 7030
+    },
+    {
+      "epoch": 22.005569620253166,
+      "grad_norm": 0.2840830385684967,
+      "learning_rate": 3.0801687763713086e-05,
+      "loss": 0.0632,
+      "step": 7040
+    },
+    {
+      "epoch": 22.00620253164557,
+      "grad_norm": 3.4681308269500732,
+      "learning_rate": 3.076652601969058e-05,
+      "loss": 0.1418,
+      "step": 7050
+    },
+    {
+      "epoch": 22.006835443037975,
+      "grad_norm": 2.013949155807495,
+      "learning_rate": 3.0731364275668076e-05,
+      "loss": 0.1189,
+      "step": 7060
+    },
+    {
+      "epoch": 22.00746835443038,
+      "grad_norm": 0.03054201602935791,
+      "learning_rate": 3.0696202531645574e-05,
+      "loss": 0.0485,
+      "step": 7070
+    },
+    {
+      "epoch": 22.008101265822784,
+      "grad_norm": 0.01309112273156643,
+      "learning_rate": 3.0661040787623065e-05,
+      "loss": 0.0337,
+      "step": 7080
+    },
+    {
+      "epoch": 22.00873417721519,
+      "grad_norm": 0.009347977116703987,
+      "learning_rate": 3.0625879043600564e-05,
+      "loss": 0.1198,
+      "step": 7090
+    },
+    {
+      "epoch": 22.009367088607593,
+      "grad_norm": 1.491429328918457,
+      "learning_rate": 3.059071729957806e-05,
+      "loss": 0.0425,
+      "step": 7100
+    },
+    {
+      "epoch": 22.01,
+      "grad_norm": 2.0944862365722656,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.0553,
+      "step": 7110
+    },
+    {
+      "epoch": 22.010632911392406,
+      "grad_norm": 0.02638574317097664,
+      "learning_rate": 3.052039381153305e-05,
+      "loss": 0.034,
+      "step": 7120
+    },
+    {
+      "epoch": 22.01126582278481,
+      "grad_norm": 0.053611740469932556,
+      "learning_rate": 3.048523206751055e-05,
+      "loss": 0.0355,
+      "step": 7130
+    },
+    {
+      "epoch": 22.011898734177215,
+      "grad_norm": 0.020069239661097527,
+      "learning_rate": 3.0450070323488045e-05,
+      "loss": 0.0657,
+      "step": 7140
+    },
+    {
+      "epoch": 22.01253164556962,
+      "grad_norm": 0.7463811635971069,
+      "learning_rate": 3.0414908579465547e-05,
+      "loss": 0.119,
+      "step": 7150
+    },
+    {
+      "epoch": 22.013164556962025,
+      "grad_norm": 1.8718035221099854,
+      "learning_rate": 3.0379746835443042e-05,
+      "loss": 0.0826,
+      "step": 7160
+    },
+    {
+      "epoch": 22.01379746835443,
+      "grad_norm": 1.7672007083892822,
+      "learning_rate": 3.0344585091420537e-05,
+      "loss": 0.0616,
+      "step": 7170
+    },
+    {
+      "epoch": 22.014430379746834,
+      "grad_norm": 0.014087089337408543,
+      "learning_rate": 3.0309423347398032e-05,
+      "loss": 0.0753,
+      "step": 7180
+    },
+    {
+      "epoch": 22.015063291139242,
+      "grad_norm": 2.0793707370758057,
+      "learning_rate": 3.027426160337553e-05,
+      "loss": 0.0782,
+      "step": 7190
+    },
+    {
+      "epoch": 22.015696202531647,
+      "grad_norm": 0.040184978395700455,
+      "learning_rate": 3.0239099859353025e-05,
+      "loss": 0.0132,
+      "step": 7200
+    },
+    {
+      "epoch": 22.01632911392405,
+      "grad_norm": 3.020726203918457,
+      "learning_rate": 3.020393811533052e-05,
+      "loss": 0.0371,
+      "step": 7210
+    },
+    {
+      "epoch": 22.016962025316456,
+      "grad_norm": 0.02047577127814293,
+      "learning_rate": 3.0168776371308015e-05,
+      "loss": 0.0535,
+      "step": 7220
+    },
+    {
+      "epoch": 22.01759493670886,
+      "grad_norm": 0.014698415994644165,
+      "learning_rate": 3.0133614627285517e-05,
+      "loss": 0.063,
+      "step": 7230
+    },
+    {
+      "epoch": 22.018227848101265,
+      "grad_norm": 0.016574831679463387,
+      "learning_rate": 3.009845288326301e-05,
+      "loss": 0.0531,
+      "step": 7240
+    },
+    {
+      "epoch": 22.01886075949367,
+      "grad_norm": 0.010280176065862179,
+      "learning_rate": 3.0063291139240506e-05,
+      "loss": 0.0952,
+      "step": 7250
+    },
+    {
+      "epoch": 22.019493670886074,
+      "grad_norm": 0.03817346692085266,
+      "learning_rate": 3.0028129395218e-05,
+      "loss": 0.0544,
+      "step": 7260
+    },
+    {
+      "epoch": 22.02,
+      "eval_accuracy": 0.9712092130518234,
+      "eval_loss": 0.05330738425254822,
+      "eval_runtime": 868.0763,
+      "eval_samples_per_second": 0.6,
+      "eval_steps_per_second": 0.076,
+      "step": 7268
+    },
+    {
+      "epoch": 23.00012658227848,
+      "grad_norm": 0.014792737551033497,
+      "learning_rate": 2.9992967651195503e-05,
+      "loss": 0.0958,
+      "step": 7270
+    },
+    {
+      "epoch": 23.000759493670888,
+      "grad_norm": 0.01068217121064663,
+      "learning_rate": 2.9957805907172998e-05,
+      "loss": 0.0747,
+      "step": 7280
+    },
+    {
+      "epoch": 23.001392405063292,
+      "grad_norm": 1.8503497838974,
+      "learning_rate": 2.9922644163150493e-05,
+      "loss": 0.0916,
+      "step": 7290
+    },
+    {
+      "epoch": 23.002025316455697,
+      "grad_norm": 1.8655881881713867,
+      "learning_rate": 2.9887482419127988e-05,
+      "loss": 0.0526,
+      "step": 7300
+    },
+    {
+      "epoch": 23.0026582278481,
+      "grad_norm": 2.7641124725341797,
+      "learning_rate": 2.985232067510549e-05,
+      "loss": 0.0862,
+      "step": 7310
+    },
+    {
+      "epoch": 23.003291139240506,
+      "grad_norm": 0.013922685757279396,
+      "learning_rate": 2.9817158931082985e-05,
+      "loss": 0.0557,
+      "step": 7320
+    },
+    {
+      "epoch": 23.00392405063291,
+      "grad_norm": 1.8126795291900635,
+      "learning_rate": 2.978199718706048e-05,
+      "loss": 0.1077,
+      "step": 7330
+    },
+    {
+      "epoch": 23.004556962025315,
+      "grad_norm": 0.016684675589203835,
+      "learning_rate": 2.9746835443037974e-05,
+      "loss": 0.1824,
+      "step": 7340
+    },
+    {
+      "epoch": 23.00518987341772,
+      "grad_norm": 0.010096821002662182,
+      "learning_rate": 2.9711673699015473e-05,
+      "loss": 0.0616,
+      "step": 7350
+    },
+    {
+      "epoch": 23.005822784810128,
+      "grad_norm": 0.015037915669381618,
+      "learning_rate": 2.9676511954992968e-05,
+      "loss": 0.029,
+      "step": 7360
+    },
+    {
+      "epoch": 23.006455696202533,
+      "grad_norm": 0.34261658787727356,
+      "learning_rate": 2.9641350210970466e-05,
+      "loss": 0.0256,
+      "step": 7370
+    },
+    {
+      "epoch": 23.007088607594937,
+      "grad_norm": 2.0921599864959717,
+      "learning_rate": 2.960618846694796e-05,
+      "loss": 0.0621,
+      "step": 7380
+    },
+    {
+      "epoch": 23.007721518987342,
+      "grad_norm": 0.009613982401788235,
+      "learning_rate": 2.957102672292546e-05,
+      "loss": 0.0634,
+      "step": 7390
+    },
+    {
+      "epoch": 23.008354430379747,
+      "grad_norm": 0.009403366595506668,
+      "learning_rate": 2.9535864978902954e-05,
+      "loss": 0.1097,
+      "step": 7400
+    },
+    {
+      "epoch": 23.00898734177215,
+      "grad_norm": 0.009238903410732746,
+      "learning_rate": 2.950070323488045e-05,
+      "loss": 0.0313,
+      "step": 7410
+    },
+    {
+      "epoch": 23.009620253164556,
+      "grad_norm": 0.01602065935730934,
+      "learning_rate": 2.9465541490857944e-05,
+      "loss": 0.0149,
+      "step": 7420
+    },
+    {
+      "epoch": 23.01025316455696,
+      "grad_norm": 0.01591380685567856,
+      "learning_rate": 2.9430379746835446e-05,
+      "loss": 0.0474,
+      "step": 7430
+    },
+    {
+      "epoch": 23.01088607594937,
+      "grad_norm": 1.879072666168213,
+      "learning_rate": 2.939521800281294e-05,
+      "loss": 0.0615,
+      "step": 7440
+    },
+    {
+      "epoch": 23.011518987341773,
+      "grad_norm": 0.014801949262619019,
+      "learning_rate": 2.9360056258790436e-05,
+      "loss": 0.0654,
+      "step": 7450
+    },
+    {
+      "epoch": 23.012151898734178,
+      "grad_norm": 0.05049045756459236,
+      "learning_rate": 2.9324894514767937e-05,
+      "loss": 0.025,
+      "step": 7460
+    },
+    {
+      "epoch": 23.012784810126583,
+      "grad_norm": 1.6079883575439453,
+      "learning_rate": 2.9289732770745432e-05,
+      "loss": 0.0663,
+      "step": 7470
+    },
+    {
+      "epoch": 23.013417721518987,
+      "grad_norm": 0.018271761015057564,
+      "learning_rate": 2.9254571026722927e-05,
+      "loss": 0.0164,
+      "step": 7480
+    },
+    {
+      "epoch": 23.014050632911392,
+      "grad_norm": 0.9706446528434753,
+      "learning_rate": 2.9219409282700422e-05,
+      "loss": 0.0981,
+      "step": 7490
+    },
+    {
+      "epoch": 23.014683544303796,
+      "grad_norm": 0.010271470062434673,
+      "learning_rate": 2.9184247538677924e-05,
+      "loss": 0.0293,
+      "step": 7500
+    },
+    {
+      "epoch": 23.0153164556962,
+      "grad_norm": 1.9835656881332397,
+      "learning_rate": 2.914908579465542e-05,
+      "loss": 0.1089,
+      "step": 7510
+    },
+    {
+      "epoch": 23.01594936708861,
+      "grad_norm": 0.01825304701924324,
+      "learning_rate": 2.9113924050632914e-05,
+      "loss": 0.0524,
+      "step": 7520
+    },
+    {
+      "epoch": 23.016582278481014,
+      "grad_norm": 0.009788533672690392,
+      "learning_rate": 2.907876230661041e-05,
+      "loss": 0.065,
+      "step": 7530
+    },
+    {
+      "epoch": 23.01721518987342,
+      "grad_norm": 2.2043983936309814,
+      "learning_rate": 2.9043600562587907e-05,
+      "loss": 0.1945,
+      "step": 7540
+    },
+    {
+      "epoch": 23.017848101265823,
+      "grad_norm": 0.00593506870791316,
+      "learning_rate": 2.9008438818565402e-05,
+      "loss": 0.0846,
+      "step": 7550
+    },
+    {
+      "epoch": 23.018481012658228,
+      "grad_norm": 0.016028795391321182,
+      "learning_rate": 2.8973277074542897e-05,
+      "loss": 0.0697,
+      "step": 7560
+    },
+    {
+      "epoch": 23.019113924050632,
+      "grad_norm": 2.0164456367492676,
+      "learning_rate": 2.8938115330520392e-05,
+      "loss": 0.0579,
+      "step": 7570
+    },
+    {
+      "epoch": 23.019746835443037,
+      "grad_norm": 0.009185468778014183,
+      "learning_rate": 2.8902953586497894e-05,
+      "loss": 0.0509,
+      "step": 7580
+    },
+    {
+      "epoch": 23.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.04993312060832977,
+      "eval_runtime": 884.1639,
+      "eval_samples_per_second": 0.589,
+      "eval_steps_per_second": 0.075,
+      "step": 7584
+    },
+    {
+      "epoch": 24.000379746835442,
+      "grad_norm": 0.007606441620737314,
+      "learning_rate": 2.886779184247539e-05,
+      "loss": 0.086,
+      "step": 7590
+    },
+    {
+      "epoch": 24.001012658227847,
+      "grad_norm": 1.587971806526184,
+      "learning_rate": 2.8832630098452884e-05,
+      "loss": 0.0666,
+      "step": 7600
+    },
+    {
+      "epoch": 24.001645569620255,
+      "grad_norm": 0.06458862125873566,
+      "learning_rate": 2.879746835443038e-05,
+      "loss": 0.1167,
+      "step": 7610
+    },
+    {
+      "epoch": 24.00227848101266,
+      "grad_norm": 0.008511193096637726,
+      "learning_rate": 2.876230661040788e-05,
+      "loss": 0.0582,
+      "step": 7620
+    },
+    {
+      "epoch": 24.002911392405064,
+      "grad_norm": 0.01144248154014349,
+      "learning_rate": 2.8727144866385375e-05,
+      "loss": 0.0537,
+      "step": 7630
+    },
+    {
+      "epoch": 24.00354430379747,
+      "grad_norm": 2.393695116043091,
+      "learning_rate": 2.869198312236287e-05,
+      "loss": 0.0521,
+      "step": 7640
+    },
+    {
+      "epoch": 24.004177215189873,
+      "grad_norm": 0.010813113301992416,
+      "learning_rate": 2.8656821378340365e-05,
+      "loss": 0.0567,
+      "step": 7650
+    },
+    {
+      "epoch": 24.004810126582278,
+      "grad_norm": 0.6152679920196533,
+      "learning_rate": 2.8621659634317867e-05,
+      "loss": 0.0387,
+      "step": 7660
+    },
+    {
+      "epoch": 24.005443037974683,
+      "grad_norm": 1.79938542842865,
+      "learning_rate": 2.858649789029536e-05,
+      "loss": 0.0198,
+      "step": 7670
+    },
+    {
+      "epoch": 24.006075949367087,
+      "grad_norm": 2.5060055255889893,
+      "learning_rate": 2.8551336146272857e-05,
+      "loss": 0.142,
+      "step": 7680
+    },
+    {
+      "epoch": 24.006708860759495,
+      "grad_norm": 0.009323079138994217,
+      "learning_rate": 2.851617440225035e-05,
+      "loss": 0.0479,
+      "step": 7690
+    },
+    {
+      "epoch": 24.0073417721519,
+      "grad_norm": 2.835371732711792,
+      "learning_rate": 2.848101265822785e-05,
+      "loss": 0.0361,
+      "step": 7700
+    },
+    {
+      "epoch": 24.007974683544305,
+      "grad_norm": 2.476736545562744,
+      "learning_rate": 2.8445850914205345e-05,
+      "loss": 0.1028,
+      "step": 7710
+    },
+    {
+      "epoch": 24.00860759493671,
+      "grad_norm": 0.011100457981228828,
+      "learning_rate": 2.8410689170182843e-05,
+      "loss": 0.109,
+      "step": 7720
+    },
+    {
+      "epoch": 24.009240506329114,
+      "grad_norm": 0.009698892012238503,
+      "learning_rate": 2.8375527426160338e-05,
+      "loss": 0.0402,
+      "step": 7730
+    },
+    {
+      "epoch": 24.00987341772152,
+      "grad_norm": 0.32084423303604126,
+      "learning_rate": 2.8340365682137836e-05,
+      "loss": 0.0261,
+      "step": 7740
+    },
+    {
+      "epoch": 24.010506329113923,
+      "grad_norm": 0.03804817050695419,
+      "learning_rate": 2.830520393811533e-05,
+      "loss": 0.0519,
+      "step": 7750
+    },
+    {
+      "epoch": 24.011139240506328,
+      "grad_norm": 0.02210908569395542,
+      "learning_rate": 2.8270042194092826e-05,
+      "loss": 0.1201,
+      "step": 7760
+    },
+    {
+      "epoch": 24.011772151898736,
+      "grad_norm": 0.013488766737282276,
+      "learning_rate": 2.823488045007032e-05,
+      "loss": 0.1115,
+      "step": 7770
+    },
+    {
+      "epoch": 24.01240506329114,
+      "grad_norm": 0.01102654542773962,
+      "learning_rate": 2.8199718706047823e-05,
+      "loss": 0.0377,
+      "step": 7780
+    },
+    {
+      "epoch": 24.013037974683545,
+      "grad_norm": 2.595425844192505,
+      "learning_rate": 2.8164556962025318e-05,
+      "loss": 0.2162,
+      "step": 7790
+    },
+    {
+      "epoch": 24.01367088607595,
+      "grad_norm": 0.3236597180366516,
+      "learning_rate": 2.8129395218002813e-05,
+      "loss": 0.107,
+      "step": 7800
+    },
+    {
+      "epoch": 24.014303797468354,
+      "grad_norm": 0.769616961479187,
+      "learning_rate": 2.8094233473980308e-05,
+      "loss": 0.1602,
+      "step": 7810
+    },
+    {
+      "epoch": 24.01493670886076,
+      "grad_norm": 0.10283089429140091,
+      "learning_rate": 2.805907172995781e-05,
+      "loss": 0.0909,
+      "step": 7820
+    },
+    {
+      "epoch": 24.015569620253164,
+      "grad_norm": 0.02031938172876835,
+      "learning_rate": 2.8023909985935304e-05,
+      "loss": 0.1172,
+      "step": 7830
+    },
+    {
+      "epoch": 24.01620253164557,
+      "grad_norm": 1.7063891887664795,
+      "learning_rate": 2.79887482419128e-05,
+      "loss": 0.0797,
+      "step": 7840
+    },
+    {
+      "epoch": 24.016835443037976,
+      "grad_norm": 1.6002981662750244,
+      "learning_rate": 2.7953586497890294e-05,
+      "loss": 0.0447,
+      "step": 7850
+    },
+    {
+      "epoch": 24.01746835443038,
+      "grad_norm": 1.945346474647522,
+      "learning_rate": 2.7918424753867796e-05,
+      "loss": 0.0461,
+      "step": 7860
+    },
+    {
+      "epoch": 24.018101265822786,
+      "grad_norm": 2.0498995780944824,
+      "learning_rate": 2.788326300984529e-05,
+      "loss": 0.1525,
+      "step": 7870
+    },
+    {
+      "epoch": 24.01873417721519,
+      "grad_norm": 0.018395045772194862,
+      "learning_rate": 2.7848101265822786e-05,
+      "loss": 0.1554,
+      "step": 7880
+    },
+    {
+      "epoch": 24.019367088607595,
+      "grad_norm": 2.4818601608276367,
+      "learning_rate": 2.7812939521800284e-05,
+      "loss": 0.2115,
+      "step": 7890
+    },
+    {
+      "epoch": 24.02,
+      "grad_norm": 0.01920860819518566,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.0308,
+      "step": 7900
+    },
+    {
+      "epoch": 24.02,
+      "eval_accuracy": 0.9596928982725528,
+      "eval_loss": 0.09555886685848236,
+      "eval_runtime": 895.4045,
+      "eval_samples_per_second": 0.582,
+      "eval_steps_per_second": 0.074,
+      "step": 7900
+    },
+    {
+      "epoch": 25.000632911392405,
+      "grad_norm": 0.9809036254882812,
+      "learning_rate": 2.7742616033755274e-05,
+      "loss": 0.0833,
+      "step": 7910
+    },
+    {
+      "epoch": 25.00126582278481,
+      "grad_norm": 0.0848427340388298,
+      "learning_rate": 2.770745428973277e-05,
+      "loss": 0.04,
+      "step": 7920
+    },
+    {
+      "epoch": 25.001898734177214,
+      "grad_norm": 1.635724663734436,
+      "learning_rate": 2.767229254571027e-05,
+      "loss": 0.1677,
+      "step": 7930
+    },
+    {
+      "epoch": 25.00253164556962,
+      "grad_norm": 1.4175573587417603,
+      "learning_rate": 2.7637130801687766e-05,
+      "loss": 0.0733,
+      "step": 7940
+    },
+    {
+      "epoch": 25.003164556962027,
+      "grad_norm": 2.555398464202881,
+      "learning_rate": 2.760196905766526e-05,
+      "loss": 0.0981,
+      "step": 7950
+    },
+    {
+      "epoch": 25.00379746835443,
+      "grad_norm": 0.024168817326426506,
+      "learning_rate": 2.7566807313642756e-05,
+      "loss": 0.0517,
+      "step": 7960
+    },
+    {
+      "epoch": 25.004430379746836,
+      "grad_norm": 2.4232094287872314,
+      "learning_rate": 2.7531645569620257e-05,
+      "loss": 0.1317,
+      "step": 7970
+    },
+    {
+      "epoch": 25.00506329113924,
+      "grad_norm": 0.030434638261795044,
+      "learning_rate": 2.7496483825597752e-05,
+      "loss": 0.162,
+      "step": 7980
+    },
+    {
+      "epoch": 25.005696202531645,
+      "grad_norm": 0.24577154219150543,
+      "learning_rate": 2.7461322081575247e-05,
+      "loss": 0.0346,
+      "step": 7990
+    },
+    {
+      "epoch": 25.00632911392405,
+      "grad_norm": 0.009798028506338596,
+      "learning_rate": 2.7426160337552742e-05,
+      "loss": 0.0209,
+      "step": 8000
+    },
+    {
+      "epoch": 25.006962025316454,
+      "grad_norm": 0.009709338657557964,
+      "learning_rate": 2.7390998593530244e-05,
+      "loss": 0.0329,
+      "step": 8010
+    },
+    {
+      "epoch": 25.00759493670886,
+      "grad_norm": 0.025753848254680634,
+      "learning_rate": 2.735583684950774e-05,
+      "loss": 0.0368,
+      "step": 8020
+    },
+    {
+      "epoch": 25.008227848101267,
+      "grad_norm": 1.9421815872192383,
+      "learning_rate": 2.7320675105485234e-05,
+      "loss": 0.0755,
+      "step": 8030
+    },
+    {
+      "epoch": 25.008860759493672,
+      "grad_norm": 0.00867847353219986,
+      "learning_rate": 2.728551336146273e-05,
+      "loss": 0.0265,
+      "step": 8040
+    },
+    {
+      "epoch": 25.009493670886076,
+      "grad_norm": 0.008214855566620827,
+      "learning_rate": 2.7250351617440227e-05,
+      "loss": 0.0698,
+      "step": 8050
+    },
+    {
+      "epoch": 25.01012658227848,
+      "grad_norm": 0.013393707573413849,
+      "learning_rate": 2.7215189873417722e-05,
+      "loss": 0.1019,
+      "step": 8060
+    },
+    {
+      "epoch": 25.010759493670886,
+      "grad_norm": 0.007839508354663849,
+      "learning_rate": 2.718002812939522e-05,
+      "loss": 0.0401,
+      "step": 8070
+    },
+    {
+      "epoch": 25.01139240506329,
+      "grad_norm": 1.8051772117614746,
+      "learning_rate": 2.7144866385372715e-05,
+      "loss": 0.0971,
+      "step": 8080
+    },
+    {
+      "epoch": 25.012025316455695,
+      "grad_norm": 1.6515543460845947,
+      "learning_rate": 2.7109704641350213e-05,
+      "loss": 0.097,
+      "step": 8090
+    },
+    {
+      "epoch": 25.0126582278481,
+      "grad_norm": 1.732775330543518,
+      "learning_rate": 2.707454289732771e-05,
+      "loss": 0.0661,
+      "step": 8100
+    },
+    {
+      "epoch": 25.013291139240508,
+      "grad_norm": 2.0589871406555176,
+      "learning_rate": 2.7039381153305203e-05,
+      "loss": 0.0727,
+      "step": 8110
+    },
+    {
+      "epoch": 25.013924050632912,
+      "grad_norm": 0.017162639647722244,
+      "learning_rate": 2.7004219409282698e-05,
+      "loss": 0.0666,
+      "step": 8120
+    },
+    {
+      "epoch": 25.014556962025317,
+      "grad_norm": 0.025243617594242096,
+      "learning_rate": 2.69690576652602e-05,
+      "loss": 0.1164,
+      "step": 8130
+    },
+    {
+      "epoch": 25.01518987341772,
+      "grad_norm": 0.011467033065855503,
+      "learning_rate": 2.6933895921237695e-05,
+      "loss": 0.0974,
+      "step": 8140
+    },
+    {
+      "epoch": 25.015822784810126,
+      "grad_norm": 2.5216453075408936,
+      "learning_rate": 2.689873417721519e-05,
+      "loss": 0.1147,
+      "step": 8150
+    },
+    {
+      "epoch": 25.01645569620253,
+      "grad_norm": 0.011125218123197556,
+      "learning_rate": 2.6863572433192685e-05,
+      "loss": 0.011,
+      "step": 8160
+    },
+    {
+      "epoch": 25.017088607594935,
+      "grad_norm": 0.0575890839099884,
+      "learning_rate": 2.6828410689170186e-05,
+      "loss": 0.0757,
+      "step": 8170
+    },
+    {
+      "epoch": 25.01772151898734,
+      "grad_norm": 0.01529190968722105,
+      "learning_rate": 2.679324894514768e-05,
+      "loss": 0.1118,
+      "step": 8180
+    },
+    {
+      "epoch": 25.01835443037975,
+      "grad_norm": 1.783360481262207,
+      "learning_rate": 2.6758087201125176e-05,
+      "loss": 0.0738,
+      "step": 8190
+    },
+    {
+      "epoch": 25.018987341772153,
+      "grad_norm": 0.2645050585269928,
+      "learning_rate": 2.672292545710267e-05,
+      "loss": 0.0719,
+      "step": 8200
+    },
+    {
+      "epoch": 25.019620253164558,
+      "grad_norm": 1.2059509754180908,
+      "learning_rate": 2.6687763713080173e-05,
+      "loss": 0.0729,
+      "step": 8210
+    },
+    {
+      "epoch": 25.02,
+      "eval_accuracy": 0.9731285988483686,
+      "eval_loss": 0.07531290501356125,
+      "eval_runtime": 883.3121,
+      "eval_samples_per_second": 0.59,
+      "eval_steps_per_second": 0.075,
+      "step": 8216
+    },
+    {
+      "epoch": 26.000253164556963,
+      "grad_norm": 0.11341290175914764,
+      "learning_rate": 2.6652601969057668e-05,
+      "loss": 0.1398,
+      "step": 8220
+    },
+    {
+      "epoch": 26.000886075949367,
+      "grad_norm": 0.1513560563325882,
+      "learning_rate": 2.6617440225035163e-05,
+      "loss": 0.0599,
+      "step": 8230
+    },
+    {
+      "epoch": 26.001518987341772,
+      "grad_norm": 0.008346611633896828,
+      "learning_rate": 2.6582278481012658e-05,
+      "loss": 0.0526,
+      "step": 8240
+    },
+    {
+      "epoch": 26.002151898734176,
+      "grad_norm": 0.04121720790863037,
+      "learning_rate": 2.6547116736990156e-05,
+      "loss": 0.0438,
+      "step": 8250
+    },
+    {
+      "epoch": 26.00278481012658,
+      "grad_norm": 0.019222108647227287,
+      "learning_rate": 2.651195499296765e-05,
+      "loss": 0.0679,
+      "step": 8260
+    },
+    {
+      "epoch": 26.003417721518986,
+      "grad_norm": 2.1048424243927,
+      "learning_rate": 2.6476793248945146e-05,
+      "loss": 0.0947,
+      "step": 8270
+    },
+    {
+      "epoch": 26.004050632911394,
+      "grad_norm": 2.955946683883667,
+      "learning_rate": 2.6441631504922648e-05,
+      "loss": 0.1254,
+      "step": 8280
+    },
+    {
+      "epoch": 26.0046835443038,
+      "grad_norm": 0.013249853625893593,
+      "learning_rate": 2.6406469760900143e-05,
+      "loss": 0.0344,
+      "step": 8290
+    },
+    {
+      "epoch": 26.005316455696203,
+      "grad_norm": 0.008436400443315506,
+      "learning_rate": 2.6371308016877638e-05,
+      "loss": 0.0611,
+      "step": 8300
+    },
+    {
+      "epoch": 26.005949367088608,
+      "grad_norm": 1.8616793155670166,
+      "learning_rate": 2.6336146272855133e-05,
+      "loss": 0.0528,
+      "step": 8310
+    },
+    {
+      "epoch": 26.006582278481012,
+      "grad_norm": 0.030170224606990814,
+      "learning_rate": 2.6300984528832634e-05,
+      "loss": 0.0981,
+      "step": 8320
+    },
+    {
+      "epoch": 26.007215189873417,
+      "grad_norm": 0.011277982033789158,
+      "learning_rate": 2.626582278481013e-05,
+      "loss": 0.0216,
+      "step": 8330
+    },
+    {
+      "epoch": 26.00784810126582,
+      "grad_norm": 0.009784871712327003,
+      "learning_rate": 2.6230661040787624e-05,
+      "loss": 0.0485,
+      "step": 8340
+    },
+    {
+      "epoch": 26.008481012658226,
+      "grad_norm": 4.80241584777832,
+      "learning_rate": 2.619549929676512e-05,
+      "loss": 0.1301,
+      "step": 8350
+    },
+    {
+      "epoch": 26.009113924050634,
+      "grad_norm": 2.4445245265960693,
+      "learning_rate": 2.616033755274262e-05,
+      "loss": 0.0571,
+      "step": 8360
+    },
+    {
+      "epoch": 26.00974683544304,
+      "grad_norm": 0.009524138644337654,
+      "learning_rate": 2.6125175808720116e-05,
+      "loss": 0.0229,
+      "step": 8370
+    },
+    {
+      "epoch": 26.010379746835444,
+      "grad_norm": 0.010364298708736897,
+      "learning_rate": 2.609001406469761e-05,
+      "loss": 0.0559,
+      "step": 8380
+    },
+    {
+      "epoch": 26.01101265822785,
+      "grad_norm": 0.047712892293930054,
+      "learning_rate": 2.6054852320675106e-05,
+      "loss": 0.1409,
+      "step": 8390
+    },
+    {
+      "epoch": 26.011645569620253,
+      "grad_norm": 0.010216879658401012,
+      "learning_rate": 2.6019690576652604e-05,
+      "loss": 0.0701,
+      "step": 8400
+    },
+    {
+      "epoch": 26.012278481012657,
+      "grad_norm": 1.5211375951766968,
+      "learning_rate": 2.5984528832630102e-05,
+      "loss": 0.0774,
+      "step": 8410
+    },
+    {
+      "epoch": 26.012911392405062,
+      "grad_norm": 4.259005546569824,
+      "learning_rate": 2.5949367088607597e-05,
+      "loss": 0.0846,
+      "step": 8420
+    },
+    {
+      "epoch": 26.013544303797467,
+      "grad_norm": 0.008692068047821522,
+      "learning_rate": 2.5914205344585092e-05,
+      "loss": 0.2192,
+      "step": 8430
+    },
+    {
+      "epoch": 26.014177215189875,
+      "grad_norm": 0.9112133383750916,
+      "learning_rate": 2.587904360056259e-05,
+      "loss": 0.112,
+      "step": 8440
+    },
+    {
+      "epoch": 26.01481012658228,
+      "grad_norm": 1.4133566617965698,
+      "learning_rate": 2.5843881856540085e-05,
+      "loss": 0.0895,
+      "step": 8450
+    },
+    {
+      "epoch": 26.015443037974684,
+      "grad_norm": 0.019907064735889435,
+      "learning_rate": 2.580872011251758e-05,
+      "loss": 0.0752,
+      "step": 8460
+    },
+    {
+      "epoch": 26.01607594936709,
+      "grad_norm": 0.22535677254199982,
+      "learning_rate": 2.5773558368495075e-05,
+      "loss": 0.1745,
+      "step": 8470
+    },
+    {
+      "epoch": 26.016708860759493,
+      "grad_norm": 0.013840797357261181,
+      "learning_rate": 2.5738396624472577e-05,
+      "loss": 0.0569,
+      "step": 8480
+    },
+    {
+      "epoch": 26.017341772151898,
+      "grad_norm": 1.7949068546295166,
+      "learning_rate": 2.5703234880450072e-05,
+      "loss": 0.1005,
+      "step": 8490
+    },
+    {
+      "epoch": 26.017974683544303,
+      "grad_norm": 5.203663349151611,
+      "learning_rate": 2.5668073136427567e-05,
+      "loss": 0.1344,
+      "step": 8500
+    },
+    {
+      "epoch": 26.018607594936707,
+      "grad_norm": 1.7371711730957031,
+      "learning_rate": 2.5632911392405062e-05,
+      "loss": 0.049,
+      "step": 8510
+    },
+    {
+      "epoch": 26.019240506329115,
+      "grad_norm": 0.03200330212712288,
+      "learning_rate": 2.5597749648382564e-05,
+      "loss": 0.0778,
+      "step": 8520
+    },
+    {
+      "epoch": 26.01987341772152,
+      "grad_norm": 16.085002899169922,
+      "learning_rate": 2.556258790436006e-05,
+      "loss": 0.2328,
+      "step": 8530
+    },
+    {
+      "epoch": 26.02,
+      "eval_accuracy": 0.9654510556621881,
+      "eval_loss": 0.07737769931554794,
+      "eval_runtime": 872.9767,
+      "eval_samples_per_second": 0.597,
+      "eval_steps_per_second": 0.076,
+      "step": 8532
+    },
+    {
+      "epoch": 27.000506329113925,
+      "grad_norm": 2.101548194885254,
+      "learning_rate": 2.5527426160337553e-05,
+      "loss": 0.0209,
+      "step": 8540
+    },
+    {
+      "epoch": 27.00113924050633,
+      "grad_norm": 2.257962942123413,
+      "learning_rate": 2.549226441631505e-05,
+      "loss": 0.2402,
+      "step": 8550
+    },
+    {
+      "epoch": 27.001772151898734,
+      "grad_norm": 1.685132622718811,
+      "learning_rate": 2.545710267229255e-05,
+      "loss": 0.0636,
+      "step": 8560
+    },
+    {
+      "epoch": 27.00240506329114,
+      "grad_norm": 1.7050256729125977,
+      "learning_rate": 2.5421940928270045e-05,
+      "loss": 0.0449,
+      "step": 8570
+    },
+    {
+      "epoch": 27.003037974683544,
+      "grad_norm": 1.991898775100708,
+      "learning_rate": 2.538677918424754e-05,
+      "loss": 0.0455,
+      "step": 8580
+    },
+    {
+      "epoch": 27.00367088607595,
+      "grad_norm": 1.5045628547668457,
+      "learning_rate": 2.5351617440225035e-05,
+      "loss": 0.0672,
+      "step": 8590
+    },
+    {
+      "epoch": 27.004303797468353,
+      "grad_norm": 0.01319907046854496,
+      "learning_rate": 2.5316455696202533e-05,
+      "loss": 0.094,
+      "step": 8600
+    },
+    {
+      "epoch": 27.00493670886076,
+      "grad_norm": 0.01158622931689024,
+      "learning_rate": 2.5281293952180028e-05,
+      "loss": 0.068,
+      "step": 8610
+    },
+    {
+      "epoch": 27.005569620253166,
+      "grad_norm": 1.6995807886123657,
+      "learning_rate": 2.5246132208157523e-05,
+      "loss": 0.1678,
+      "step": 8620
+    },
+    {
+      "epoch": 27.00620253164557,
+      "grad_norm": 0.012468636967241764,
+      "learning_rate": 2.521097046413502e-05,
+      "loss": 0.0623,
+      "step": 8630
+    },
+    {
+      "epoch": 27.006835443037975,
+      "grad_norm": 1.8979823589324951,
+      "learning_rate": 2.517580872011252e-05,
+      "loss": 0.1167,
+      "step": 8640
+    },
+    {
+      "epoch": 27.00746835443038,
+      "grad_norm": 0.03327177092432976,
+      "learning_rate": 2.5140646976090015e-05,
+      "loss": 0.1498,
+      "step": 8650
+    },
+    {
+      "epoch": 27.008101265822784,
+      "grad_norm": 0.013360361568629742,
+      "learning_rate": 2.510548523206751e-05,
+      "loss": 0.149,
+      "step": 8660
+    },
+    {
+      "epoch": 27.00873417721519,
+      "grad_norm": 0.04175138100981712,
+      "learning_rate": 2.5070323488045005e-05,
+      "loss": 0.0581,
+      "step": 8670
+    },
+    {
+      "epoch": 27.009367088607593,
+      "grad_norm": 1.2819527387619019,
+      "learning_rate": 2.5035161744022506e-05,
+      "loss": 0.0572,
+      "step": 8680
+    },
+    {
+      "epoch": 27.01,
+      "grad_norm": 3.946152448654175,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1165,
+      "step": 8690
+    },
+    {
+      "epoch": 27.010632911392406,
+      "grad_norm": 1.6609853506088257,
+      "learning_rate": 2.49648382559775e-05,
+      "loss": 0.0659,
+      "step": 8700
+    },
+    {
+      "epoch": 27.01126582278481,
+      "grad_norm": 0.019263017922639847,
+      "learning_rate": 2.4929676511954994e-05,
+      "loss": 0.1063,
+      "step": 8710
+    },
+    {
+      "epoch": 27.011898734177215,
+      "grad_norm": 0.01123301312327385,
+      "learning_rate": 2.4894514767932493e-05,
+      "loss": 0.0287,
+      "step": 8720
+    },
+    {
+      "epoch": 27.01253164556962,
+      "grad_norm": 0.03187470883131027,
+      "learning_rate": 2.4859353023909988e-05,
+      "loss": 0.0525,
+      "step": 8730
+    },
+    {
+      "epoch": 27.013164556962025,
+      "grad_norm": 0.030269309878349304,
+      "learning_rate": 2.4824191279887486e-05,
+      "loss": 0.0155,
+      "step": 8740
+    },
+    {
+      "epoch": 27.01379746835443,
+      "grad_norm": 0.04713929817080498,
+      "learning_rate": 2.478902953586498e-05,
+      "loss": 0.103,
+      "step": 8750
+    },
+    {
+      "epoch": 27.014430379746834,
+      "grad_norm": 0.007141268812119961,
+      "learning_rate": 2.475386779184248e-05,
+      "loss": 0.0338,
+      "step": 8760
+    },
+    {
+      "epoch": 27.015063291139242,
+      "grad_norm": 0.008212805725634098,
+      "learning_rate": 2.4718706047819974e-05,
+      "loss": 0.0764,
+      "step": 8770
+    },
+    {
+      "epoch": 27.015696202531647,
+      "grad_norm": 0.007574188522994518,
+      "learning_rate": 2.468354430379747e-05,
+      "loss": 0.0772,
+      "step": 8780
+    },
+    {
+      "epoch": 27.01632911392405,
+      "grad_norm": 1.347557544708252,
+      "learning_rate": 2.4648382559774964e-05,
+      "loss": 0.0767,
+      "step": 8790
+    },
+    {
+      "epoch": 27.016962025316456,
+      "grad_norm": 0.008004224859178066,
+      "learning_rate": 2.4613220815752462e-05,
+      "loss": 0.0405,
+      "step": 8800
+    },
+    {
+      "epoch": 27.01759493670886,
+      "grad_norm": 0.06255421787500381,
+      "learning_rate": 2.4578059071729957e-05,
+      "loss": 0.0634,
+      "step": 8810
+    },
+    {
+      "epoch": 27.018227848101265,
+      "grad_norm": 0.014694114215672016,
+      "learning_rate": 2.4542897327707456e-05,
+      "loss": 0.0704,
+      "step": 8820
+    },
+    {
+      "epoch": 27.01886075949367,
+      "grad_norm": 0.0064834184013307095,
+      "learning_rate": 2.450773558368495e-05,
+      "loss": 0.0492,
+      "step": 8830
+    },
+    {
+      "epoch": 27.019493670886074,
+      "grad_norm": 0.01545525249093771,
+      "learning_rate": 2.447257383966245e-05,
+      "loss": 0.1085,
+      "step": 8840
+    },
+    {
+      "epoch": 27.02,
+      "eval_accuracy": 0.9692898272552783,
+      "eval_loss": 0.06086457893252373,
+      "eval_runtime": 940.0459,
+      "eval_samples_per_second": 0.554,
+      "eval_steps_per_second": 0.07,
+      "step": 8848
+    },
+    {
+      "epoch": 28.00012658227848,
+      "grad_norm": 1.5009100437164307,
+      "learning_rate": 2.4437412095639944e-05,
+      "loss": 0.0116,
+      "step": 8850
+    },
+    {
+      "epoch": 28.000759493670888,
+      "grad_norm": 1.8109427690505981,
+      "learning_rate": 2.4402250351617442e-05,
+      "loss": 0.1101,
+      "step": 8860
+    },
+    {
+      "epoch": 28.001392405063292,
+      "grad_norm": 0.016067614778876305,
+      "learning_rate": 2.4367088607594937e-05,
+      "loss": 0.0632,
+      "step": 8870
+    },
+    {
+      "epoch": 28.002025316455697,
+      "grad_norm": 1.9723069667816162,
+      "learning_rate": 2.4331926863572436e-05,
+      "loss": 0.0556,
+      "step": 8880
+    },
+    {
+      "epoch": 28.0026582278481,
+      "grad_norm": 6.840577125549316,
+      "learning_rate": 2.429676511954993e-05,
+      "loss": 0.0614,
+      "step": 8890
+    },
+    {
+      "epoch": 28.003291139240506,
+      "grad_norm": 0.00987607054412365,
+      "learning_rate": 2.426160337552743e-05,
+      "loss": 0.0619,
+      "step": 8900
+    },
+    {
+      "epoch": 28.00392405063291,
+      "grad_norm": 1.5094603300094604,
+      "learning_rate": 2.4226441631504924e-05,
+      "loss": 0.0817,
+      "step": 8910
+    },
+    {
+      "epoch": 28.004556962025315,
+      "grad_norm": 0.0056548151187598705,
+      "learning_rate": 2.4191279887482422e-05,
+      "loss": 0.0395,
+      "step": 8920
+    },
+    {
+      "epoch": 28.00518987341772,
+      "grad_norm": 1.6165566444396973,
+      "learning_rate": 2.4156118143459917e-05,
+      "loss": 0.0196,
+      "step": 8930
+    },
+    {
+      "epoch": 28.005822784810128,
+      "grad_norm": 0.011191487312316895,
+      "learning_rate": 2.4120956399437415e-05,
+      "loss": 0.0488,
+      "step": 8940
+    },
+    {
+      "epoch": 28.006455696202533,
+      "grad_norm": 0.013476500287652016,
+      "learning_rate": 2.408579465541491e-05,
+      "loss": 0.077,
+      "step": 8950
+    },
+    {
+      "epoch": 28.007088607594937,
+      "grad_norm": 0.01594310626387596,
+      "learning_rate": 2.4050632911392405e-05,
+      "loss": 0.0764,
+      "step": 8960
+    },
+    {
+      "epoch": 28.007721518987342,
+      "grad_norm": 2.4658524990081787,
+      "learning_rate": 2.40154711673699e-05,
+      "loss": 0.0816,
+      "step": 8970
+    },
+    {
+      "epoch": 28.008354430379747,
+      "grad_norm": 0.050282131880521774,
+      "learning_rate": 2.39803094233474e-05,
+      "loss": 0.0483,
+      "step": 8980
+    },
+    {
+      "epoch": 28.00898734177215,
+      "grad_norm": 0.0738360807299614,
+      "learning_rate": 2.3945147679324893e-05,
+      "loss": 0.0325,
+      "step": 8990
+    },
+    {
+      "epoch": 28.009620253164556,
+      "grad_norm": 2.087075710296631,
+      "learning_rate": 2.3909985935302392e-05,
+      "loss": 0.1095,
+      "step": 9000
+    },
+    {
+      "epoch": 28.01025316455696,
+      "grad_norm": 0.012715667486190796,
+      "learning_rate": 2.3874824191279887e-05,
+      "loss": 0.0548,
+      "step": 9010
+    },
+    {
+      "epoch": 28.01088607594937,
+      "grad_norm": 2.233593463897705,
+      "learning_rate": 2.3839662447257385e-05,
+      "loss": 0.0342,
+      "step": 9020
+    },
+    {
+      "epoch": 28.011518987341773,
+      "grad_norm": 0.016205696389079094,
+      "learning_rate": 2.380450070323488e-05,
+      "loss": 0.0916,
+      "step": 9030
+    },
+    {
+      "epoch": 28.012151898734178,
+      "grad_norm": 0.008019771426916122,
+      "learning_rate": 2.3769338959212378e-05,
+      "loss": 0.0851,
+      "step": 9040
+    },
+    {
+      "epoch": 28.012784810126583,
+      "grad_norm": 1.6921287775039673,
+      "learning_rate": 2.3734177215189873e-05,
+      "loss": 0.0543,
+      "step": 9050
+    },
+    {
+      "epoch": 28.013417721518987,
+      "grad_norm": 0.03432793170213699,
+      "learning_rate": 2.369901547116737e-05,
+      "loss": 0.1162,
+      "step": 9060
+    },
+    {
+      "epoch": 28.014050632911392,
+      "grad_norm": 4.5309648513793945,
+      "learning_rate": 2.3663853727144866e-05,
+      "loss": 0.1159,
+      "step": 9070
+    },
+    {
+      "epoch": 28.014683544303796,
+      "grad_norm": 1.7229171991348267,
+      "learning_rate": 2.3628691983122365e-05,
+      "loss": 0.0453,
+      "step": 9080
+    },
+    {
+      "epoch": 28.0153164556962,
+      "grad_norm": 1.9000073671340942,
+      "learning_rate": 2.359353023909986e-05,
+      "loss": 0.0311,
+      "step": 9090
+    },
+    {
+      "epoch": 28.01594936708861,
+      "grad_norm": 23.23397445678711,
+      "learning_rate": 2.3558368495077358e-05,
+      "loss": 0.1287,
+      "step": 9100
+    },
+    {
+      "epoch": 28.016582278481014,
+      "grad_norm": 0.14218182861804962,
+      "learning_rate": 2.3523206751054856e-05,
+      "loss": 0.0783,
+      "step": 9110
+    },
+    {
+      "epoch": 28.01721518987342,
+      "grad_norm": 1.8179298639297485,
+      "learning_rate": 2.348804500703235e-05,
+      "loss": 0.0228,
+      "step": 9120
+    },
+    {
+      "epoch": 28.017848101265823,
+      "grad_norm": 0.01124257780611515,
+      "learning_rate": 2.3452883263009846e-05,
+      "loss": 0.0606,
+      "step": 9130
+    },
+    {
+      "epoch": 28.018481012658228,
+      "grad_norm": 0.06378079950809479,
+      "learning_rate": 2.341772151898734e-05,
+      "loss": 0.0613,
+      "step": 9140
+    },
+    {
+      "epoch": 28.019113924050632,
+      "grad_norm": 0.5730488896369934,
+      "learning_rate": 2.338255977496484e-05,
+      "loss": 0.0922,
+      "step": 9150
+    },
+    {
+      "epoch": 28.019746835443037,
+      "grad_norm": 0.1582580953836441,
+      "learning_rate": 2.3347398030942334e-05,
+      "loss": 0.099,
+      "step": 9160
+    },
+    {
+      "epoch": 28.02,
+      "eval_accuracy": 0.9673704414587332,
+      "eval_loss": 0.06770172715187073,
+      "eval_runtime": 874.5137,
+      "eval_samples_per_second": 0.596,
+      "eval_steps_per_second": 0.075,
+      "step": 9164
+    },
+    {
+      "epoch": 29.000379746835442,
+      "grad_norm": 1.0581729412078857,
+      "learning_rate": 2.3312236286919833e-05,
+      "loss": 0.0999,
+      "step": 9170
+    },
+    {
+      "epoch": 29.001012658227847,
+      "grad_norm": 1.9762561321258545,
+      "learning_rate": 2.3277074542897328e-05,
+      "loss": 0.0571,
+      "step": 9180
+    },
+    {
+      "epoch": 29.001645569620255,
+      "grad_norm": 1.5621942281723022,
+      "learning_rate": 2.3241912798874826e-05,
+      "loss": 0.0549,
+      "step": 9190
+    },
+    {
+      "epoch": 29.00227848101266,
+      "grad_norm": 0.007985897362232208,
+      "learning_rate": 2.320675105485232e-05,
+      "loss": 0.0897,
+      "step": 9200
+    },
+    {
+      "epoch": 29.002911392405064,
+      "grad_norm": 1.7005807161331177,
+      "learning_rate": 2.317158931082982e-05,
+      "loss": 0.0953,
+      "step": 9210
+    },
+    {
+      "epoch": 29.00354430379747,
+      "grad_norm": 0.01191030628979206,
+      "learning_rate": 2.3136427566807314e-05,
+      "loss": 0.0495,
+      "step": 9220
+    },
+    {
+      "epoch": 29.004177215189873,
+      "grad_norm": 2.738387107849121,
+      "learning_rate": 2.3101265822784813e-05,
+      "loss": 0.0748,
+      "step": 9230
+    },
+    {
+      "epoch": 29.004810126582278,
+      "grad_norm": 1.7861073017120361,
+      "learning_rate": 2.3066104078762308e-05,
+      "loss": 0.1279,
+      "step": 9240
+    },
+    {
+      "epoch": 29.005443037974683,
+      "grad_norm": 2.4535388946533203,
+      "learning_rate": 2.3030942334739806e-05,
+      "loss": 0.1033,
+      "step": 9250
+    },
+    {
+      "epoch": 29.006075949367087,
+      "grad_norm": 0.028340207412838936,
+      "learning_rate": 2.29957805907173e-05,
+      "loss": 0.2441,
+      "step": 9260
+    },
+    {
+      "epoch": 29.006708860759495,
+      "grad_norm": 0.020006628707051277,
+      "learning_rate": 2.29606188466948e-05,
+      "loss": 0.1339,
+      "step": 9270
+    },
+    {
+      "epoch": 29.0073417721519,
+      "grad_norm": 0.06801651418209076,
+      "learning_rate": 2.2925457102672294e-05,
+      "loss": 0.0383,
+      "step": 9280
+    },
+    {
+      "epoch": 29.007974683544305,
+      "grad_norm": 0.03369883447885513,
+      "learning_rate": 2.2890295358649792e-05,
+      "loss": 0.0552,
+      "step": 9290
+    },
+    {
+      "epoch": 29.00860759493671,
+      "grad_norm": 1.2586785554885864,
+      "learning_rate": 2.2855133614627287e-05,
+      "loss": 0.0745,
+      "step": 9300
+    },
+    {
+      "epoch": 29.009240506329114,
+      "grad_norm": 0.008519505150616169,
+      "learning_rate": 2.2819971870604782e-05,
+      "loss": 0.1307,
+      "step": 9310
+    },
+    {
+      "epoch": 29.00987341772152,
+      "grad_norm": 0.04472700506448746,
+      "learning_rate": 2.278481012658228e-05,
+      "loss": 0.0757,
+      "step": 9320
+    },
+    {
+      "epoch": 29.010506329113923,
+      "grad_norm": 1.7951302528381348,
+      "learning_rate": 2.2749648382559775e-05,
+      "loss": 0.0504,
+      "step": 9330
+    },
+    {
+      "epoch": 29.011139240506328,
+      "grad_norm": 7.45163106918335,
+      "learning_rate": 2.271448663853727e-05,
+      "loss": 0.0637,
+      "step": 9340
+    },
+    {
+      "epoch": 29.011772151898736,
+      "grad_norm": 1.5808919668197632,
+      "learning_rate": 2.267932489451477e-05,
+      "loss": 0.0681,
+      "step": 9350
+    },
+    {
+      "epoch": 29.01240506329114,
+      "grad_norm": 0.009355315938591957,
+      "learning_rate": 2.2644163150492264e-05,
+      "loss": 0.0789,
+      "step": 9360
+    },
+    {
+      "epoch": 29.013037974683545,
+      "grad_norm": 36.12007522583008,
+      "learning_rate": 2.2609001406469762e-05,
+      "loss": 0.1642,
+      "step": 9370
+    },
+    {
+      "epoch": 29.01367088607595,
+      "grad_norm": 0.007421193178743124,
+      "learning_rate": 2.2573839662447257e-05,
+      "loss": 0.0126,
+      "step": 9380
+    },
+    {
+      "epoch": 29.014303797468354,
+      "grad_norm": 1.9807149171829224,
+      "learning_rate": 2.2538677918424755e-05,
+      "loss": 0.1069,
+      "step": 9390
+    },
+    {
+      "epoch": 29.01493670886076,
+      "grad_norm": 0.016664525493979454,
+      "learning_rate": 2.250351617440225e-05,
+      "loss": 0.174,
+      "step": 9400
+    },
+    {
+      "epoch": 29.015569620253164,
+      "grad_norm": 0.015380149707198143,
+      "learning_rate": 2.246835443037975e-05,
+      "loss": 0.0959,
+      "step": 9410
+    },
+    {
+      "epoch": 29.01620253164557,
+      "grad_norm": 1.9330257177352905,
+      "learning_rate": 2.2433192686357243e-05,
+      "loss": 0.0604,
+      "step": 9420
+    },
+    {
+      "epoch": 29.016835443037976,
+      "grad_norm": 55.1077880859375,
+      "learning_rate": 2.2398030942334742e-05,
+      "loss": 0.0694,
+      "step": 9430
+    },
+    {
+      "epoch": 29.01746835443038,
+      "grad_norm": 1.9939329624176025,
+      "learning_rate": 2.2362869198312237e-05,
+      "loss": 0.1348,
+      "step": 9440
+    },
+    {
+      "epoch": 29.018101265822786,
+      "grad_norm": 0.015108354389667511,
+      "learning_rate": 2.2327707454289735e-05,
+      "loss": 0.0449,
+      "step": 9450
+    },
+    {
+      "epoch": 29.01873417721519,
+      "grad_norm": 12.801469802856445,
+      "learning_rate": 2.229254571026723e-05,
+      "loss": 0.1694,
+      "step": 9460
+    },
+    {
+      "epoch": 29.019367088607595,
+      "grad_norm": 2.3264384269714355,
+      "learning_rate": 2.225738396624473e-05,
+      "loss": 0.044,
+      "step": 9470
+    },
+    {
+      "epoch": 29.02,
+      "grad_norm": 0.13212990760803223,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.1988,
+      "step": 9480
+    },
+    {
+      "epoch": 29.02,
+      "eval_accuracy": 0.9558541266794626,
+      "eval_loss": 0.14145594835281372,
+      "eval_runtime": 928.9271,
+      "eval_samples_per_second": 0.561,
+      "eval_steps_per_second": 0.071,
+      "step": 9480
+    },
+    {
+      "epoch": 30.000632911392405,
+      "grad_norm": 2.014655113220215,
+      "learning_rate": 2.2187060478199718e-05,
+      "loss": 0.0823,
+      "step": 9490
+    },
+    {
+      "epoch": 30.00126582278481,
+      "grad_norm": 7.12656831741333,
+      "learning_rate": 2.2151898734177217e-05,
+      "loss": 0.0963,
+      "step": 9500
+    },
+    {
+      "epoch": 30.001898734177214,
+      "grad_norm": 1.7992146015167236,
+      "learning_rate": 2.211673699015471e-05,
+      "loss": 0.1107,
+      "step": 9510
+    },
+    {
+      "epoch": 30.00253164556962,
+      "grad_norm": 7.597923755645752,
+      "learning_rate": 2.2081575246132206e-05,
+      "loss": 0.102,
+      "step": 9520
+    },
+    {
+      "epoch": 30.003164556962027,
+      "grad_norm": 0.026418205350637436,
+      "learning_rate": 2.2046413502109705e-05,
+      "loss": 0.1202,
+      "step": 9530
+    },
+    {
+      "epoch": 30.00379746835443,
+      "grad_norm": 0.01834811083972454,
+      "learning_rate": 2.2011251758087203e-05,
+      "loss": 0.0728,
+      "step": 9540
+    },
+    {
+      "epoch": 30.004430379746836,
+      "grad_norm": 0.02628573216497898,
+      "learning_rate": 2.1976090014064698e-05,
+      "loss": 0.0948,
+      "step": 9550
+    },
+    {
+      "epoch": 30.00506329113924,
+      "grad_norm": 0.023844990879297256,
+      "learning_rate": 2.1940928270042196e-05,
+      "loss": 0.0601,
+      "step": 9560
+    },
+    {
+      "epoch": 30.005696202531645,
+      "grad_norm": 0.019756905734539032,
+      "learning_rate": 2.190576652601969e-05,
+      "loss": 0.0759,
+      "step": 9570
+    },
+    {
+      "epoch": 30.00632911392405,
+      "grad_norm": 6.277482986450195,
+      "learning_rate": 2.187060478199719e-05,
+      "loss": 0.053,
+      "step": 9580
+    },
+    {
+      "epoch": 30.006962025316454,
+      "grad_norm": 0.5496724843978882,
+      "learning_rate": 2.1835443037974685e-05,
+      "loss": 0.0653,
+      "step": 9590
+    },
+    {
+      "epoch": 30.00759493670886,
+      "grad_norm": 1.7114007472991943,
+      "learning_rate": 2.1800281293952183e-05,
+      "loss": 0.0466,
+      "step": 9600
+    },
+    {
+      "epoch": 30.008227848101267,
+      "grad_norm": 2.283554792404175,
+      "learning_rate": 2.1765119549929678e-05,
+      "loss": 0.0608,
+      "step": 9610
+    },
+    {
+      "epoch": 30.008860759493672,
+      "grad_norm": 4.336071014404297,
+      "learning_rate": 2.1729957805907176e-05,
+      "loss": 0.0702,
+      "step": 9620
+    },
+    {
+      "epoch": 30.009493670886076,
+      "grad_norm": 0.2585821747779846,
+      "learning_rate": 2.169479606188467e-05,
+      "loss": 0.0716,
+      "step": 9630
+    },
+    {
+      "epoch": 30.01012658227848,
+      "grad_norm": 2.0628931522369385,
+      "learning_rate": 2.165963431786217e-05,
+      "loss": 0.0975,
+      "step": 9640
+    },
+    {
+      "epoch": 30.010759493670886,
+      "grad_norm": 0.9379414319992065,
+      "learning_rate": 2.1624472573839664e-05,
+      "loss": 0.0643,
+      "step": 9650
+    },
+    {
+      "epoch": 30.01139240506329,
+      "grad_norm": 1.865263819694519,
+      "learning_rate": 2.158931082981716e-05,
+      "loss": 0.089,
+      "step": 9660
+    },
+    {
+      "epoch": 30.012025316455695,
+      "grad_norm": 0.01576172187924385,
+      "learning_rate": 2.1554149085794658e-05,
+      "loss": 0.0863,
+      "step": 9670
+    },
+    {
+      "epoch": 30.0126582278481,
+      "grad_norm": 8.34028434753418,
+      "learning_rate": 2.1518987341772153e-05,
+      "loss": 0.0266,
+      "step": 9680
+    },
+    {
+      "epoch": 30.013291139240508,
+      "grad_norm": 2.5415472984313965,
+      "learning_rate": 2.1483825597749647e-05,
+      "loss": 0.1101,
+      "step": 9690
+    },
+    {
+      "epoch": 30.013924050632912,
+      "grad_norm": 0.010155543684959412,
+      "learning_rate": 2.1448663853727146e-05,
+      "loss": 0.037,
+      "step": 9700
+    },
+    {
+      "epoch": 30.014556962025317,
+      "grad_norm": 2.1569736003875732,
+      "learning_rate": 2.141350210970464e-05,
+      "loss": 0.03,
+      "step": 9710
+    },
+    {
+      "epoch": 30.01518987341772,
+      "grad_norm": 2.8239643573760986,
+      "learning_rate": 2.137834036568214e-05,
+      "loss": 0.1167,
+      "step": 9720
+    },
+    {
+      "epoch": 30.015822784810126,
+      "grad_norm": 1.6771701574325562,
+      "learning_rate": 2.1343178621659634e-05,
+      "loss": 0.1783,
+      "step": 9730
+    },
+    {
+      "epoch": 30.01645569620253,
+      "grad_norm": 1.3176885843276978,
+      "learning_rate": 2.1308016877637132e-05,
+      "loss": 0.0518,
+      "step": 9740
+    },
+    {
+      "epoch": 30.017088607594935,
+      "grad_norm": 2.2316536903381348,
+      "learning_rate": 2.1272855133614627e-05,
+      "loss": 0.0688,
+      "step": 9750
+    },
+    {
+      "epoch": 30.01772151898734,
+      "grad_norm": 0.02535848319530487,
+      "learning_rate": 2.1237693389592126e-05,
+      "loss": 0.1062,
+      "step": 9760
+    },
+    {
+      "epoch": 30.01835443037975,
+      "grad_norm": 0.026239003986120224,
+      "learning_rate": 2.120253164556962e-05,
+      "loss": 0.0434,
+      "step": 9770
+    },
+    {
+      "epoch": 30.018987341772153,
+      "grad_norm": 0.008352408185601234,
+      "learning_rate": 2.116736990154712e-05,
+      "loss": 0.1434,
+      "step": 9780
+    },
+    {
+      "epoch": 30.019620253164558,
+      "grad_norm": 0.008794278837740421,
+      "learning_rate": 2.1132208157524614e-05,
+      "loss": 0.0747,
+      "step": 9790
+    },
+    {
+      "epoch": 30.02,
+      "eval_accuracy": 0.9712092130518234,
+      "eval_loss": 0.058070629835128784,
+      "eval_runtime": 882.2085,
+      "eval_samples_per_second": 0.591,
+      "eval_steps_per_second": 0.075,
+      "step": 9796
+    },
+    {
+      "epoch": 31.000253164556963,
+      "grad_norm": 0.15315893292427063,
+      "learning_rate": 2.1097046413502112e-05,
+      "loss": 0.1227,
+      "step": 9800
+    },
+    {
+      "epoch": 31.000886075949367,
+      "grad_norm": 0.007744469679892063,
+      "learning_rate": 2.1061884669479607e-05,
+      "loss": 0.0588,
+      "step": 9810
+    },
+    {
+      "epoch": 31.001518987341772,
+      "grad_norm": 1.7301479578018188,
+      "learning_rate": 2.1026722925457105e-05,
+      "loss": 0.1045,
+      "step": 9820
+    },
+    {
+      "epoch": 31.002151898734176,
+      "grad_norm": 0.016438119113445282,
+      "learning_rate": 2.09915611814346e-05,
+      "loss": 0.1089,
+      "step": 9830
+    },
+    {
+      "epoch": 31.00278481012658,
+      "grad_norm": 2.2902982234954834,
+      "learning_rate": 2.09563994374121e-05,
+      "loss": 0.0569,
+      "step": 9840
+    },
+    {
+      "epoch": 31.003417721518986,
+      "grad_norm": 0.01934031955897808,
+      "learning_rate": 2.0921237693389594e-05,
+      "loss": 0.0613,
+      "step": 9850
+    },
+    {
+      "epoch": 31.004050632911394,
+      "grad_norm": 0.10376526415348053,
+      "learning_rate": 2.088607594936709e-05,
+      "loss": 0.0278,
+      "step": 9860
+    },
+    {
+      "epoch": 31.0046835443038,
+      "grad_norm": 0.03141545131802559,
+      "learning_rate": 2.0850914205344583e-05,
+      "loss": 0.0557,
+      "step": 9870
+    },
+    {
+      "epoch": 31.005316455696203,
+      "grad_norm": 0.024358389899134636,
+      "learning_rate": 2.0815752461322082e-05,
+      "loss": 0.0358,
+      "step": 9880
+    },
+    {
+      "epoch": 31.005949367088608,
+      "grad_norm": 0.09153765439987183,
+      "learning_rate": 2.0780590717299577e-05,
+      "loss": 0.0682,
+      "step": 9890
+    },
+    {
+      "epoch": 31.006582278481012,
+      "grad_norm": 0.012928824871778488,
+      "learning_rate": 2.0745428973277075e-05,
+      "loss": 0.044,
+      "step": 9900
+    },
+    {
+      "epoch": 31.007215189873417,
+      "grad_norm": 0.02423839271068573,
+      "learning_rate": 2.071026722925457e-05,
+      "loss": 0.029,
+      "step": 9910
+    },
+    {
+      "epoch": 31.00784810126582,
+      "grad_norm": 0.007467347197234631,
+      "learning_rate": 2.067510548523207e-05,
+      "loss": 0.1689,
+      "step": 9920
+    },
+    {
+      "epoch": 31.008481012658226,
+      "grad_norm": 61.641075134277344,
+      "learning_rate": 2.0639943741209563e-05,
+      "loss": 0.0393,
+      "step": 9930
+    },
+    {
+      "epoch": 31.009113924050634,
+      "grad_norm": 0.013753454200923443,
+      "learning_rate": 2.060478199718706e-05,
+      "loss": 0.0425,
+      "step": 9940
+    },
+    {
+      "epoch": 31.00974683544304,
+      "grad_norm": 2.26615047454834,
+      "learning_rate": 2.056962025316456e-05,
+      "loss": 0.0482,
+      "step": 9950
+    },
+    {
+      "epoch": 31.010379746835444,
+      "grad_norm": 17.814420700073242,
+      "learning_rate": 2.0534458509142055e-05,
+      "loss": 0.1064,
+      "step": 9960
+    },
+    {
+      "epoch": 31.01101265822785,
+      "grad_norm": 0.0065340083092451096,
+      "learning_rate": 2.0499296765119553e-05,
+      "loss": 0.0222,
+      "step": 9970
+    },
+    {
+      "epoch": 31.011645569620253,
+      "grad_norm": 0.004512071143835783,
+      "learning_rate": 2.0464135021097048e-05,
+      "loss": 0.0561,
+      "step": 9980
+    },
+    {
+      "epoch": 31.012278481012657,
+      "grad_norm": 2.7562315464019775,
+      "learning_rate": 2.0428973277074546e-05,
+      "loss": 0.0865,
+      "step": 9990
+    },
+    {
+      "epoch": 31.012911392405062,
+      "grad_norm": 0.01689945161342621,
+      "learning_rate": 2.039381153305204e-05,
+      "loss": 0.04,
+      "step": 10000
+    },
+    {
+      "epoch": 31.013544303797467,
+      "grad_norm": 2.3072574138641357,
+      "learning_rate": 2.0358649789029536e-05,
+      "loss": 0.1108,
+      "step": 10010
+    },
+    {
+      "epoch": 31.014177215189875,
+      "grad_norm": 2.39740252494812,
+      "learning_rate": 2.0323488045007035e-05,
+      "loss": 0.0603,
+      "step": 10020
+    },
+    {
+      "epoch": 31.01481012658228,
+      "grad_norm": 2.2645068168640137,
+      "learning_rate": 2.028832630098453e-05,
+      "loss": 0.1595,
+      "step": 10030
+    },
+    {
+      "epoch": 31.015443037974684,
+      "grad_norm": 0.020067648962140083,
+      "learning_rate": 2.0253164556962025e-05,
+      "loss": 0.0824,
+      "step": 10040
+    },
+    {
+      "epoch": 31.01607594936709,
+      "grad_norm": 1.732828974723816,
+      "learning_rate": 2.0218002812939523e-05,
+      "loss": 0.0979,
+      "step": 10050
+    },
+    {
+      "epoch": 31.016708860759493,
+      "grad_norm": 0.006261592730879784,
+      "learning_rate": 2.0182841068917018e-05,
+      "loss": 0.0294,
+      "step": 10060
+    },
+    {
+      "epoch": 31.017341772151898,
+      "grad_norm": 0.009185651317238808,
+      "learning_rate": 2.0147679324894516e-05,
+      "loss": 0.0684,
+      "step": 10070
+    },
+    {
+      "epoch": 31.017974683544303,
+      "grad_norm": 2.076765298843384,
+      "learning_rate": 2.011251758087201e-05,
+      "loss": 0.0763,
+      "step": 10080
+    },
+    {
+      "epoch": 31.018607594936707,
+      "grad_norm": 0.010820266790688038,
+      "learning_rate": 2.007735583684951e-05,
+      "loss": 0.1151,
+      "step": 10090
+    },
+    {
+      "epoch": 31.019240506329115,
+      "grad_norm": 0.9778680801391602,
+      "learning_rate": 2.0042194092827004e-05,
+      "loss": 0.0274,
+      "step": 10100
+    },
+    {
+      "epoch": 31.01987341772152,
+      "grad_norm": 0.03660624101758003,
+      "learning_rate": 2.0007032348804503e-05,
+      "loss": 0.0556,
+      "step": 10110
+    },
+    {
+      "epoch": 31.02,
+      "eval_accuracy": 0.9692898272552783,
+      "eval_loss": 0.051862768828868866,
+      "eval_runtime": 895.6108,
+      "eval_samples_per_second": 0.582,
+      "eval_steps_per_second": 0.074,
+      "step": 10112
+    },
+    {
+      "epoch": 32.000506329113925,
+      "grad_norm": 0.014388811774551868,
+      "learning_rate": 1.9971870604781998e-05,
+      "loss": 0.0289,
+      "step": 10120
+    },
+    {
+      "epoch": 32.001139240506326,
+      "grad_norm": 1.698613166809082,
+      "learning_rate": 1.9936708860759496e-05,
+      "loss": 0.0312,
+      "step": 10130
+    },
+    {
+      "epoch": 32.001772151898734,
+      "grad_norm": 0.026271585375070572,
+      "learning_rate": 1.990154711673699e-05,
+      "loss": 0.0451,
+      "step": 10140
+    },
+    {
+      "epoch": 32.00240506329114,
+      "grad_norm": 0.00675854692235589,
+      "learning_rate": 1.986638537271449e-05,
+      "loss": 0.0247,
+      "step": 10150
+    },
+    {
+      "epoch": 32.003037974683544,
+      "grad_norm": 0.022877847775816917,
+      "learning_rate": 1.9831223628691984e-05,
+      "loss": 0.0235,
+      "step": 10160
+    },
+    {
+      "epoch": 32.00367088607595,
+      "grad_norm": 0.008142875507473946,
+      "learning_rate": 1.9796061884669482e-05,
+      "loss": 0.0699,
+      "step": 10170
+    },
+    {
+      "epoch": 32.00430379746835,
+      "grad_norm": 0.004865671973675489,
+      "learning_rate": 1.9760900140646977e-05,
+      "loss": 0.0645,
+      "step": 10180
+    },
+    {
+      "epoch": 32.00493670886076,
+      "grad_norm": 0.05031589791178703,
+      "learning_rate": 1.9725738396624476e-05,
+      "loss": 0.0675,
+      "step": 10190
+    },
+    {
+      "epoch": 32.00556962025316,
+      "grad_norm": 0.006223399192094803,
+      "learning_rate": 1.969057665260197e-05,
+      "loss": 0.0812,
+      "step": 10200
+    },
+    {
+      "epoch": 32.00620253164557,
+      "grad_norm": 0.008529792539775372,
+      "learning_rate": 1.9655414908579466e-05,
+      "loss": 0.1027,
+      "step": 10210
+    },
+    {
+      "epoch": 32.00683544303797,
+      "grad_norm": 0.004334003198891878,
+      "learning_rate": 1.962025316455696e-05,
+      "loss": 0.0511,
+      "step": 10220
+    },
+    {
+      "epoch": 32.00746835443038,
+      "grad_norm": 0.004973968956619501,
+      "learning_rate": 1.958509142053446e-05,
+      "loss": 0.0556,
+      "step": 10230
+    },
+    {
+      "epoch": 32.00810126582279,
+      "grad_norm": 0.005091523285955191,
+      "learning_rate": 1.9549929676511954e-05,
+      "loss": 0.0889,
+      "step": 10240
+    },
+    {
+      "epoch": 32.00873417721519,
+      "grad_norm": 0.004862161818891764,
+      "learning_rate": 1.9514767932489452e-05,
+      "loss": 0.0728,
+      "step": 10250
+    },
+    {
+      "epoch": 32.0093670886076,
+      "grad_norm": 0.43511274456977844,
+      "learning_rate": 1.9479606188466947e-05,
+      "loss": 0.0398,
+      "step": 10260
+    },
+    {
+      "epoch": 32.01,
+      "grad_norm": 0.006799314171075821,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.0543,
+      "step": 10270
+    },
+    {
+      "epoch": 32.010632911392406,
+      "grad_norm": 2.4008827209472656,
+      "learning_rate": 1.940928270042194e-05,
+      "loss": 0.0623,
+      "step": 10280
+    },
+    {
+      "epoch": 32.01126582278481,
+      "grad_norm": 0.00461224839091301,
+      "learning_rate": 1.937412095639944e-05,
+      "loss": 0.0434,
+      "step": 10290
+    },
+    {
+      "epoch": 32.011898734177215,
+      "grad_norm": 0.004693826660513878,
+      "learning_rate": 1.9338959212376934e-05,
+      "loss": 0.0904,
+      "step": 10300
+    },
+    {
+      "epoch": 32.012531645569624,
+      "grad_norm": 0.006937893573194742,
+      "learning_rate": 1.9303797468354432e-05,
+      "loss": 0.0769,
+      "step": 10310
+    },
+    {
+      "epoch": 32.013164556962025,
+      "grad_norm": 0.005765580106526613,
+      "learning_rate": 1.9268635724331927e-05,
+      "loss": 0.0743,
+      "step": 10320
+    },
+    {
+      "epoch": 32.01379746835443,
+      "grad_norm": 0.003979605156928301,
+      "learning_rate": 1.9233473980309425e-05,
+      "loss": 0.0406,
+      "step": 10330
+    },
+    {
+      "epoch": 32.014430379746834,
+      "grad_norm": 1.9673800468444824,
+      "learning_rate": 1.919831223628692e-05,
+      "loss": 0.0928,
+      "step": 10340
+    },
+    {
+      "epoch": 32.01506329113924,
+      "grad_norm": 2.0642683506011963,
+      "learning_rate": 1.916315049226442e-05,
+      "loss": 0.0872,
+      "step": 10350
+    },
+    {
+      "epoch": 32.01569620253164,
+      "grad_norm": 2.0600333213806152,
+      "learning_rate": 1.9127988748241913e-05,
+      "loss": 0.0747,
+      "step": 10360
+    },
+    {
+      "epoch": 32.01632911392405,
+      "grad_norm": 2.7668960094451904,
+      "learning_rate": 1.9092827004219412e-05,
+      "loss": 0.0496,
+      "step": 10370
+    },
+    {
+      "epoch": 32.01696202531645,
+      "grad_norm": 0.0035948660224676132,
+      "learning_rate": 1.9057665260196907e-05,
+      "loss": 0.0713,
+      "step": 10380
+    },
+    {
+      "epoch": 32.01759493670886,
+      "grad_norm": 1.5214300155639648,
+      "learning_rate": 1.90225035161744e-05,
+      "loss": 0.0854,
+      "step": 10390
+    },
+    {
+      "epoch": 32.01822784810127,
+      "grad_norm": 1.6459559202194214,
+      "learning_rate": 1.89873417721519e-05,
+      "loss": 0.0765,
+      "step": 10400
+    },
+    {
+      "epoch": 32.01886075949367,
+      "grad_norm": 1.9180864095687866,
+      "learning_rate": 1.8952180028129395e-05,
+      "loss": 0.0367,
+      "step": 10410
+    },
+    {
+      "epoch": 32.01949367088608,
+      "grad_norm": 2.7413058280944824,
+      "learning_rate": 1.8917018284106893e-05,
+      "loss": 0.0763,
+      "step": 10420
+    },
+    {
+      "epoch": 32.02,
+      "eval_accuracy": 0.9731285988483686,
+      "eval_loss": 0.050615016371011734,
+      "eval_runtime": 876.7902,
+      "eval_samples_per_second": 0.594,
+      "eval_steps_per_second": 0.075,
+      "step": 10428
+    },
+    {
+      "epoch": 33.00012658227848,
+      "grad_norm": 0.003906856290996075,
+      "learning_rate": 1.8881856540084388e-05,
+      "loss": 0.0514,
+      "step": 10430
+    },
+    {
+      "epoch": 33.000759493670884,
+      "grad_norm": 1.0877649784088135,
+      "learning_rate": 1.8846694796061886e-05,
+      "loss": 0.0264,
+      "step": 10440
+    },
+    {
+      "epoch": 33.00139240506329,
+      "grad_norm": 0.0038005015812814236,
+      "learning_rate": 1.881153305203938e-05,
+      "loss": 0.1031,
+      "step": 10450
+    },
+    {
+      "epoch": 33.00202531645569,
+      "grad_norm": 0.0038557087536901236,
+      "learning_rate": 1.877637130801688e-05,
+      "loss": 0.055,
+      "step": 10460
+    },
+    {
+      "epoch": 33.0026582278481,
+      "grad_norm": 0.005424036644399166,
+      "learning_rate": 1.8741209563994375e-05,
+      "loss": 0.0734,
+      "step": 10470
+    },
+    {
+      "epoch": 33.00329113924051,
+      "grad_norm": 0.003683975664898753,
+      "learning_rate": 1.8706047819971873e-05,
+      "loss": 0.0477,
+      "step": 10480
+    },
+    {
+      "epoch": 33.00392405063291,
+      "grad_norm": 0.3991844952106476,
+      "learning_rate": 1.8670886075949368e-05,
+      "loss": 0.0014,
+      "step": 10490
+    },
+    {
+      "epoch": 33.00455696202532,
+      "grad_norm": 1.7069263458251953,
+      "learning_rate": 1.8635724331926866e-05,
+      "loss": 0.0679,
+      "step": 10500
+    },
+    {
+      "epoch": 33.00518987341772,
+      "grad_norm": 1.4248631000518799,
+      "learning_rate": 1.860056258790436e-05,
+      "loss": 0.0478,
+      "step": 10510
+    },
+    {
+      "epoch": 33.00582278481013,
+      "grad_norm": 0.006337147206068039,
+      "learning_rate": 1.856540084388186e-05,
+      "loss": 0.0769,
+      "step": 10520
+    },
+    {
+      "epoch": 33.00645569620253,
+      "grad_norm": 0.003814426949247718,
+      "learning_rate": 1.8530239099859354e-05,
+      "loss": 0.1142,
+      "step": 10530
+    },
+    {
+      "epoch": 33.00708860759494,
+      "grad_norm": 0.003283862257376313,
+      "learning_rate": 1.8495077355836853e-05,
+      "loss": 0.0826,
+      "step": 10540
+    },
+    {
+      "epoch": 33.00772151898734,
+      "grad_norm": 0.7462538480758667,
+      "learning_rate": 1.8459915611814348e-05,
+      "loss": 0.078,
+      "step": 10550
+    },
+    {
+      "epoch": 33.00835443037975,
+      "grad_norm": 1.6937042474746704,
+      "learning_rate": 1.8424753867791843e-05,
+      "loss": 0.0805,
+      "step": 10560
+    },
+    {
+      "epoch": 33.008987341772155,
+      "grad_norm": 2.729776382446289,
+      "learning_rate": 1.8389592123769338e-05,
+      "loss": 0.0982,
+      "step": 10570
+    },
+    {
+      "epoch": 33.009620253164556,
+      "grad_norm": 0.005251895170658827,
+      "learning_rate": 1.8354430379746836e-05,
+      "loss": 0.0348,
+      "step": 10580
+    },
+    {
+      "epoch": 33.010253164556964,
+      "grad_norm": 0.005169942043721676,
+      "learning_rate": 1.831926863572433e-05,
+      "loss": 0.1224,
+      "step": 10590
+    },
+    {
+      "epoch": 33.010886075949365,
+      "grad_norm": 2.058396577835083,
+      "learning_rate": 1.828410689170183e-05,
+      "loss": 0.0732,
+      "step": 10600
+    },
+    {
+      "epoch": 33.01151898734177,
+      "grad_norm": 0.004916292615234852,
+      "learning_rate": 1.8248945147679324e-05,
+      "loss": 0.0203,
+      "step": 10610
+    },
+    {
+      "epoch": 33.012151898734174,
+      "grad_norm": 0.004748203791677952,
+      "learning_rate": 1.8213783403656822e-05,
+      "loss": 0.0726,
+      "step": 10620
+    },
+    {
+      "epoch": 33.01278481012658,
+      "grad_norm": 0.004022388719022274,
+      "learning_rate": 1.8178621659634317e-05,
+      "loss": 0.0105,
+      "step": 10630
+    },
+    {
+      "epoch": 33.01341772151899,
+      "grad_norm": 0.002989129861816764,
+      "learning_rate": 1.8143459915611816e-05,
+      "loss": 0.0093,
+      "step": 10640
+    },
+    {
+      "epoch": 33.01405063291139,
+      "grad_norm": 0.0042097545228898525,
+      "learning_rate": 1.810829817158931e-05,
+      "loss": 0.0425,
+      "step": 10650
+    },
+    {
+      "epoch": 33.0146835443038,
+      "grad_norm": 0.008987381123006344,
+      "learning_rate": 1.807313642756681e-05,
+      "loss": 0.0721,
+      "step": 10660
+    },
+    {
+      "epoch": 33.0153164556962,
+      "grad_norm": 0.004144416656345129,
+      "learning_rate": 1.8037974683544304e-05,
+      "loss": 0.0474,
+      "step": 10670
+    },
+    {
+      "epoch": 33.01594936708861,
+      "grad_norm": 0.0029516241047531366,
+      "learning_rate": 1.8002812939521802e-05,
+      "loss": 0.0501,
+      "step": 10680
+    },
+    {
+      "epoch": 33.01658227848101,
+      "grad_norm": 0.00844433344900608,
+      "learning_rate": 1.7967651195499297e-05,
+      "loss": 0.0652,
+      "step": 10690
+    },
+    {
+      "epoch": 33.01721518987342,
+      "grad_norm": 2.04638409614563,
+      "learning_rate": 1.7932489451476795e-05,
+      "loss": 0.084,
+      "step": 10700
+    },
+    {
+      "epoch": 33.01784810126582,
+      "grad_norm": 0.004865339025855064,
+      "learning_rate": 1.789732770745429e-05,
+      "loss": 0.0698,
+      "step": 10710
+    },
+    {
+      "epoch": 33.01848101265823,
+      "grad_norm": 2.136470079421997,
+      "learning_rate": 1.786216596343179e-05,
+      "loss": 0.0798,
+      "step": 10720
+    },
+    {
+      "epoch": 33.019113924050636,
+      "grad_norm": 0.004824280273169279,
+      "learning_rate": 1.7827004219409284e-05,
+      "loss": 0.0244,
+      "step": 10730
+    },
+    {
+      "epoch": 33.01974683544304,
+      "grad_norm": 2.0193657875061035,
+      "learning_rate": 1.779184247538678e-05,
+      "loss": 0.0635,
+      "step": 10740
+    },
+    {
+      "epoch": 33.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.049176860600709915,
+      "eval_runtime": 874.5591,
+      "eval_samples_per_second": 0.596,
+      "eval_steps_per_second": 0.075,
+      "step": 10744
+    },
+    {
+      "epoch": 34.00037974683544,
+      "grad_norm": 0.004216327331960201,
+      "learning_rate": 1.7756680731364274e-05,
+      "loss": 0.0312,
+      "step": 10750
+    },
+    {
+      "epoch": 34.00101265822785,
+      "grad_norm": 0.004103939048945904,
+      "learning_rate": 1.7721518987341772e-05,
+      "loss": 0.041,
+      "step": 10760
+    },
+    {
+      "epoch": 34.00164556962025,
+      "grad_norm": 0.014023522846400738,
+      "learning_rate": 1.7686357243319267e-05,
+      "loss": 0.0759,
+      "step": 10770
+    },
+    {
+      "epoch": 34.00227848101266,
+      "grad_norm": 0.003603527322411537,
+      "learning_rate": 1.7651195499296765e-05,
+      "loss": 0.0688,
+      "step": 10780
+    },
+    {
+      "epoch": 34.00291139240506,
+      "grad_norm": 0.0033313506282866,
+      "learning_rate": 1.7616033755274263e-05,
+      "loss": 0.0465,
+      "step": 10790
+    },
+    {
+      "epoch": 34.00354430379747,
+      "grad_norm": 1.8435585498809814,
+      "learning_rate": 1.758087201125176e-05,
+      "loss": 0.0949,
+      "step": 10800
+    },
+    {
+      "epoch": 34.00417721518988,
+      "grad_norm": 1.6445870399475098,
+      "learning_rate": 1.7545710267229257e-05,
+      "loss": 0.1285,
+      "step": 10810
+    },
+    {
+      "epoch": 34.00481012658228,
+      "grad_norm": 0.002922611776739359,
+      "learning_rate": 1.751054852320675e-05,
+      "loss": 0.0706,
+      "step": 10820
+    },
+    {
+      "epoch": 34.005443037974686,
+      "grad_norm": 2.673938274383545,
+      "learning_rate": 1.747538677918425e-05,
+      "loss": 0.0942,
+      "step": 10830
+    },
+    {
+      "epoch": 34.00607594936709,
+      "grad_norm": 0.004203031305223703,
+      "learning_rate": 1.7440225035161745e-05,
+      "loss": 0.0326,
+      "step": 10840
+    },
+    {
+      "epoch": 34.006708860759495,
+      "grad_norm": 1.8749343156814575,
+      "learning_rate": 1.7405063291139243e-05,
+      "loss": 0.1136,
+      "step": 10850
+    },
+    {
+      "epoch": 34.0073417721519,
+      "grad_norm": 0.004128487780690193,
+      "learning_rate": 1.7369901547116738e-05,
+      "loss": 0.0677,
+      "step": 10860
+    },
+    {
+      "epoch": 34.007974683544305,
+      "grad_norm": 0.00483922241255641,
+      "learning_rate": 1.7334739803094237e-05,
+      "loss": 0.0719,
+      "step": 10870
+    },
+    {
+      "epoch": 34.008607594936706,
+      "grad_norm": 1.308681607246399,
+      "learning_rate": 1.729957805907173e-05,
+      "loss": 0.0435,
+      "step": 10880
+    },
+    {
+      "epoch": 34.009240506329114,
+      "grad_norm": 0.0035531967878341675,
+      "learning_rate": 1.726441631504923e-05,
+      "loss": 0.0415,
+      "step": 10890
+    },
+    {
+      "epoch": 34.00987341772152,
+      "grad_norm": 1.937238335609436,
+      "learning_rate": 1.7229254571026725e-05,
+      "loss": 0.0518,
+      "step": 10900
+    },
+    {
+      "epoch": 34.01050632911392,
+      "grad_norm": 0.0039803506806492805,
+      "learning_rate": 1.719409282700422e-05,
+      "loss": 0.065,
+      "step": 10910
+    },
+    {
+      "epoch": 34.01113924050633,
+      "grad_norm": 1.8601768016815186,
+      "learning_rate": 1.7158931082981715e-05,
+      "loss": 0.0599,
+      "step": 10920
+    },
+    {
+      "epoch": 34.01177215189873,
+      "grad_norm": 0.003120964393019676,
+      "learning_rate": 1.7123769338959213e-05,
+      "loss": 0.0747,
+      "step": 10930
+    },
+    {
+      "epoch": 34.01240506329114,
+      "grad_norm": 2.323869228363037,
+      "learning_rate": 1.7088607594936708e-05,
+      "loss": 0.0448,
+      "step": 10940
+    },
+    {
+      "epoch": 34.01303797468354,
+      "grad_norm": 0.007831540890038013,
+      "learning_rate": 1.7053445850914206e-05,
+      "loss": 0.1065,
+      "step": 10950
+    },
+    {
+      "epoch": 34.01367088607595,
+      "grad_norm": 0.007483489811420441,
+      "learning_rate": 1.70182841068917e-05,
+      "loss": 0.057,
+      "step": 10960
+    },
+    {
+      "epoch": 34.01430379746836,
+      "grad_norm": 1.7984932661056519,
+      "learning_rate": 1.69831223628692e-05,
+      "loss": 0.1047,
+      "step": 10970
+    },
+    {
+      "epoch": 34.01493670886076,
+      "grad_norm": 0.011337379924952984,
+      "learning_rate": 1.6947960618846694e-05,
+      "loss": 0.0281,
+      "step": 10980
+    },
+    {
+      "epoch": 34.01556962025317,
+      "grad_norm": 0.004797440022230148,
+      "learning_rate": 1.6912798874824193e-05,
+      "loss": 0.0295,
+      "step": 10990
+    },
+    {
+      "epoch": 34.01620253164557,
+      "grad_norm": 0.0023510728497058153,
+      "learning_rate": 1.6877637130801688e-05,
+      "loss": 0.0391,
+      "step": 11000
+    },
+    {
+      "epoch": 34.01683544303798,
+      "grad_norm": 0.015410896390676498,
+      "learning_rate": 1.6842475386779186e-05,
+      "loss": 0.0316,
+      "step": 11010
+    },
+    {
+      "epoch": 34.01746835443038,
+      "grad_norm": 0.008044656366109848,
+      "learning_rate": 1.680731364275668e-05,
+      "loss": 0.036,
+      "step": 11020
+    },
+    {
+      "epoch": 34.018101265822786,
+      "grad_norm": 0.004154231399297714,
+      "learning_rate": 1.677215189873418e-05,
+      "loss": 0.0852,
+      "step": 11030
+    },
+    {
+      "epoch": 34.01873417721519,
+      "grad_norm": 0.006638134364038706,
+      "learning_rate": 1.6736990154711674e-05,
+      "loss": 0.0142,
+      "step": 11040
+    },
+    {
+      "epoch": 34.019367088607595,
+      "grad_norm": 0.0038269851356744766,
+      "learning_rate": 1.6701828410689173e-05,
+      "loss": 0.0155,
+      "step": 11050
+    },
+    {
+      "epoch": 34.02,
+      "grad_norm": 0.0027346955612301826,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.0729,
+      "step": 11060
+    },
+    {
+      "epoch": 34.02,
+      "eval_accuracy": 0.9692898272552783,
+      "eval_loss": 0.048282425850629807,
+      "eval_runtime": 866.1958,
+      "eval_samples_per_second": 0.601,
+      "eval_steps_per_second": 0.076,
+      "step": 11060
+    },
+    {
+      "epoch": 35.00063291139241,
+      "grad_norm": 0.0029359341133385897,
+      "learning_rate": 1.6631504922644166e-05,
+      "loss": 0.0964,
+      "step": 11070
+    },
+    {
+      "epoch": 35.00126582278481,
+      "grad_norm": 0.0031327735632658005,
+      "learning_rate": 1.659634317862166e-05,
+      "loss": 0.057,
+      "step": 11080
+    },
+    {
+      "epoch": 35.00189873417722,
+      "grad_norm": 0.004360498860478401,
+      "learning_rate": 1.6561181434599156e-05,
+      "loss": 0.0264,
+      "step": 11090
+    },
+    {
+      "epoch": 35.00253164556962,
+      "grad_norm": 0.013268140144646168,
+      "learning_rate": 1.6526019690576654e-05,
+      "loss": 0.0785,
+      "step": 11100
+    },
+    {
+      "epoch": 35.00316455696203,
+      "grad_norm": 0.0029599510598927736,
+      "learning_rate": 1.649085794655415e-05,
+      "loss": 0.0914,
+      "step": 11110
+    },
+    {
+      "epoch": 35.00379746835443,
+      "grad_norm": 0.00652879336848855,
+      "learning_rate": 1.6455696202531644e-05,
+      "loss": 0.0423,
+      "step": 11120
+    },
+    {
+      "epoch": 35.004430379746836,
+      "grad_norm": 0.004876864142715931,
+      "learning_rate": 1.6420534458509142e-05,
+      "loss": 0.0367,
+      "step": 11130
+    },
+    {
+      "epoch": 35.00506329113924,
+      "grad_norm": 0.0030262030195444822,
+      "learning_rate": 1.6385372714486637e-05,
+      "loss": 0.0579,
+      "step": 11140
+    },
+    {
+      "epoch": 35.005696202531645,
+      "grad_norm": 0.005096379201859236,
+      "learning_rate": 1.6350210970464135e-05,
+      "loss": 0.0162,
+      "step": 11150
+    },
+    {
+      "epoch": 35.00632911392405,
+      "grad_norm": 0.005958447232842445,
+      "learning_rate": 1.631504922644163e-05,
+      "loss": 0.0704,
+      "step": 11160
+    },
+    {
+      "epoch": 35.006962025316454,
+      "grad_norm": 0.09998659044504166,
+      "learning_rate": 1.627988748241913e-05,
+      "loss": 0.0396,
+      "step": 11170
+    },
+    {
+      "epoch": 35.00759493670886,
+      "grad_norm": 2.7944371700286865,
+      "learning_rate": 1.6244725738396624e-05,
+      "loss": 0.0813,
+      "step": 11180
+    },
+    {
+      "epoch": 35.008227848101264,
+      "grad_norm": 0.00864170491695404,
+      "learning_rate": 1.6209563994374122e-05,
+      "loss": 0.069,
+      "step": 11190
+    },
+    {
+      "epoch": 35.00886075949367,
+      "grad_norm": 0.008192246779799461,
+      "learning_rate": 1.617440225035162e-05,
+      "loss": 0.1284,
+      "step": 11200
+    },
+    {
+      "epoch": 35.00949367088607,
+      "grad_norm": 0.007694208063185215,
+      "learning_rate": 1.6139240506329115e-05,
+      "loss": 0.0198,
+      "step": 11210
+    },
+    {
+      "epoch": 35.01012658227848,
+      "grad_norm": 0.0043876878917217255,
+      "learning_rate": 1.6104078762306614e-05,
+      "loss": 0.0759,
+      "step": 11220
+    },
+    {
+      "epoch": 35.01075949367089,
+      "grad_norm": 5.103017330169678,
+      "learning_rate": 1.606891701828411e-05,
+      "loss": 0.0263,
+      "step": 11230
+    },
+    {
+      "epoch": 35.01139240506329,
+      "grad_norm": 0.038662366569042206,
+      "learning_rate": 1.6033755274261607e-05,
+      "loss": 0.1215,
+      "step": 11240
+    },
+    {
+      "epoch": 35.0120253164557,
+      "grad_norm": 0.0065099457278847694,
+      "learning_rate": 1.5998593530239102e-05,
+      "loss": 0.1129,
+      "step": 11250
+    },
+    {
+      "epoch": 35.0126582278481,
+      "grad_norm": 3.32816219329834,
+      "learning_rate": 1.5963431786216597e-05,
+      "loss": 0.1243,
+      "step": 11260
+    },
+    {
+      "epoch": 35.01329113924051,
+      "grad_norm": 2.0049562454223633,
+      "learning_rate": 1.5928270042194095e-05,
+      "loss": 0.0641,
+      "step": 11270
+    },
+    {
+      "epoch": 35.01392405063291,
+      "grad_norm": 0.7978969216346741,
+      "learning_rate": 1.589310829817159e-05,
+      "loss": 0.0625,
+      "step": 11280
+    },
+    {
+      "epoch": 35.01455696202532,
+      "grad_norm": 0.004707363899797201,
+      "learning_rate": 1.5857946554149085e-05,
+      "loss": 0.064,
+      "step": 11290
+    },
+    {
+      "epoch": 35.01518987341772,
+      "grad_norm": 0.0142443822696805,
+      "learning_rate": 1.5822784810126583e-05,
+      "loss": 0.0638,
+      "step": 11300
+    },
+    {
+      "epoch": 35.015822784810126,
+      "grad_norm": 0.011079243384301662,
+      "learning_rate": 1.5787623066104078e-05,
+      "loss": 0.0004,
+      "step": 11310
+    },
+    {
+      "epoch": 35.016455696202534,
+      "grad_norm": 2.3243215084075928,
+      "learning_rate": 1.5752461322081577e-05,
+      "loss": 0.0786,
+      "step": 11320
+    },
+    {
+      "epoch": 35.017088607594935,
+      "grad_norm": 0.004174171946942806,
+      "learning_rate": 1.571729957805907e-05,
+      "loss": 0.0151,
+      "step": 11330
+    },
+    {
+      "epoch": 35.017721518987344,
+      "grad_norm": 0.006861343514174223,
+      "learning_rate": 1.568213783403657e-05,
+      "loss": 0.1092,
+      "step": 11340
+    },
+    {
+      "epoch": 35.018354430379745,
+      "grad_norm": 2.0469024181365967,
+      "learning_rate": 1.5646976090014065e-05,
+      "loss": 0.035,
+      "step": 11350
+    },
+    {
+      "epoch": 35.01898734177215,
+      "grad_norm": 2.0664782524108887,
+      "learning_rate": 1.5611814345991563e-05,
+      "loss": 0.0619,
+      "step": 11360
+    },
+    {
+      "epoch": 35.019620253164554,
+      "grad_norm": 0.004111067857593298,
+      "learning_rate": 1.5576652601969058e-05,
+      "loss": 0.0692,
+      "step": 11370
+    },
+    {
+      "epoch": 35.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.04807919263839722,
+      "eval_runtime": 894.7526,
+      "eval_samples_per_second": 0.582,
+      "eval_steps_per_second": 0.074,
+      "step": 11376
+    },
+    {
+      "epoch": 36.00025316455696,
+      "grad_norm": 0.0060658580623567104,
+      "learning_rate": 1.5541490857946556e-05,
+      "loss": 0.0883,
+      "step": 11380
+    },
+    {
+      "epoch": 36.00088607594937,
+      "grad_norm": 2.9155092239379883,
+      "learning_rate": 1.550632911392405e-05,
+      "loss": 0.0509,
+      "step": 11390
+    },
+    {
+      "epoch": 36.001518987341775,
+      "grad_norm": 0.004734630696475506,
+      "learning_rate": 1.547116736990155e-05,
+      "loss": 0.0554,
+      "step": 11400
+    },
+    {
+      "epoch": 36.002151898734176,
+      "grad_norm": 2.3105368614196777,
+      "learning_rate": 1.5436005625879045e-05,
+      "loss": 0.0768,
+      "step": 11410
+    },
+    {
+      "epoch": 36.002784810126585,
+      "grad_norm": 0.0034461403265595436,
+      "learning_rate": 1.5400843881856543e-05,
+      "loss": 0.0336,
+      "step": 11420
+    },
+    {
+      "epoch": 36.003417721518986,
+      "grad_norm": 2.3787004947662354,
+      "learning_rate": 1.5365682137834038e-05,
+      "loss": 0.0541,
+      "step": 11430
+    },
+    {
+      "epoch": 36.004050632911394,
+      "grad_norm": 0.003510306589305401,
+      "learning_rate": 1.5330520393811533e-05,
+      "loss": 0.0132,
+      "step": 11440
+    },
+    {
+      "epoch": 36.004683544303795,
+      "grad_norm": 2.900848627090454,
+      "learning_rate": 1.529535864978903e-05,
+      "loss": 0.0892,
+      "step": 11450
+    },
+    {
+      "epoch": 36.0053164556962,
+      "grad_norm": 0.0045495242811739445,
+      "learning_rate": 1.5260196905766526e-05,
+      "loss": 0.0582,
+      "step": 11460
+    },
+    {
+      "epoch": 36.005949367088604,
+      "grad_norm": 2.1150054931640625,
+      "learning_rate": 1.5225035161744023e-05,
+      "loss": 0.0567,
+      "step": 11470
+    },
+    {
+      "epoch": 36.00658227848101,
+      "grad_norm": 2.303497791290283,
+      "learning_rate": 1.5189873417721521e-05,
+      "loss": 0.0471,
+      "step": 11480
+    },
+    {
+      "epoch": 36.00721518987342,
+      "grad_norm": 1.5887198448181152,
+      "learning_rate": 1.5154711673699016e-05,
+      "loss": 0.0762,
+      "step": 11490
+    },
+    {
+      "epoch": 36.00784810126582,
+      "grad_norm": 0.009701371192932129,
+      "learning_rate": 1.5119549929676513e-05,
+      "loss": 0.0994,
+      "step": 11500
+    },
+    {
+      "epoch": 36.00848101265823,
+      "grad_norm": 1.875978708267212,
+      "learning_rate": 1.5084388185654007e-05,
+      "loss": 0.037,
+      "step": 11510
+    },
+    {
+      "epoch": 36.00911392405063,
+      "grad_norm": 1.7351700067520142,
+      "learning_rate": 1.5049226441631506e-05,
+      "loss": 0.1036,
+      "step": 11520
+    },
+    {
+      "epoch": 36.00974683544304,
+      "grad_norm": 2.9241271018981934,
+      "learning_rate": 1.5014064697609e-05,
+      "loss": 0.0546,
+      "step": 11530
+    },
+    {
+      "epoch": 36.01037974683544,
+      "grad_norm": 0.00652047386392951,
+      "learning_rate": 1.4978902953586499e-05,
+      "loss": 0.0837,
+      "step": 11540
+    },
+    {
+      "epoch": 36.01101265822785,
+      "grad_norm": 0.005075695458799601,
+      "learning_rate": 1.4943741209563994e-05,
+      "loss": 0.0413,
+      "step": 11550
+    },
+    {
+      "epoch": 36.011645569620256,
+      "grad_norm": 0.005239785648882389,
+      "learning_rate": 1.4908579465541492e-05,
+      "loss": 0.0714,
+      "step": 11560
+    },
+    {
+      "epoch": 36.01227848101266,
+      "grad_norm": 0.0033316484186798334,
+      "learning_rate": 1.4873417721518987e-05,
+      "loss": 0.0297,
+      "step": 11570
+    },
+    {
+      "epoch": 36.012911392405066,
+      "grad_norm": 2.3135130405426025,
+      "learning_rate": 1.4838255977496484e-05,
+      "loss": 0.1521,
+      "step": 11580
+    },
+    {
+      "epoch": 36.01354430379747,
+      "grad_norm": 2.01491117477417,
+      "learning_rate": 1.480309423347398e-05,
+      "loss": 0.0542,
+      "step": 11590
+    },
+    {
+      "epoch": 36.014177215189875,
+      "grad_norm": 0.003011218272149563,
+      "learning_rate": 1.4767932489451477e-05,
+      "loss": 0.0605,
+      "step": 11600
+    },
+    {
+      "epoch": 36.014810126582276,
+      "grad_norm": 0.003551116678863764,
+      "learning_rate": 1.4732770745428972e-05,
+      "loss": 0.0359,
+      "step": 11610
+    },
+    {
+      "epoch": 36.015443037974684,
+      "grad_norm": 0.004775399807840586,
+      "learning_rate": 1.469760900140647e-05,
+      "loss": 0.0181,
+      "step": 11620
+    },
+    {
+      "epoch": 36.016075949367085,
+      "grad_norm": 0.0036576108541339636,
+      "learning_rate": 1.4662447257383969e-05,
+      "loss": 0.1157,
+      "step": 11630
+    },
+    {
+      "epoch": 36.01670886075949,
+      "grad_norm": 0.00402354309335351,
+      "learning_rate": 1.4627285513361464e-05,
+      "loss": 0.0341,
+      "step": 11640
+    },
+    {
+      "epoch": 36.0173417721519,
+      "grad_norm": 0.003662722185254097,
+      "learning_rate": 1.4592123769338962e-05,
+      "loss": 0.0393,
+      "step": 11650
+    },
+    {
+      "epoch": 36.0179746835443,
+      "grad_norm": 2.0367825031280518,
+      "learning_rate": 1.4556962025316457e-05,
+      "loss": 0.0609,
+      "step": 11660
+    },
+    {
+      "epoch": 36.01860759493671,
+      "grad_norm": 0.4210270345211029,
+      "learning_rate": 1.4521800281293954e-05,
+      "loss": 0.0793,
+      "step": 11670
+    },
+    {
+      "epoch": 36.01924050632911,
+      "grad_norm": 2.6557528972625732,
+      "learning_rate": 1.4486638537271449e-05,
+      "loss": 0.0707,
+      "step": 11680
+    },
+    {
+      "epoch": 36.01987341772152,
+      "grad_norm": 2.917741537094116,
+      "learning_rate": 1.4451476793248947e-05,
+      "loss": 0.1023,
+      "step": 11690
+    },
+    {
+      "epoch": 36.02,
+      "eval_accuracy": 0.9712092130518234,
+      "eval_loss": 0.047843087464571,
+      "eval_runtime": 935.2983,
+      "eval_samples_per_second": 0.557,
+      "eval_steps_per_second": 0.071,
+      "step": 11692
+    },
+    {
+      "epoch": 37.000506329113925,
+      "grad_norm": 0.004461715929210186,
+      "learning_rate": 1.4416315049226442e-05,
+      "loss": 0.0132,
+      "step": 11700
+    },
+    {
+      "epoch": 37.001139240506326,
+      "grad_norm": 2.282515525817871,
+      "learning_rate": 1.438115330520394e-05,
+      "loss": 0.097,
+      "step": 11710
+    },
+    {
+      "epoch": 37.001772151898734,
+      "grad_norm": 1.2947165966033936,
+      "learning_rate": 1.4345991561181435e-05,
+      "loss": 0.0536,
+      "step": 11720
+    },
+    {
+      "epoch": 37.00240506329114,
+      "grad_norm": 0.003223339095711708,
+      "learning_rate": 1.4310829817158933e-05,
+      "loss": 0.0295,
+      "step": 11730
+    },
+    {
+      "epoch": 37.003037974683544,
+      "grad_norm": 0.003211404662579298,
+      "learning_rate": 1.4275668073136428e-05,
+      "loss": 0.0408,
+      "step": 11740
+    },
+    {
+      "epoch": 37.00367088607595,
+      "grad_norm": 0.004511035978794098,
+      "learning_rate": 1.4240506329113925e-05,
+      "loss": 0.0594,
+      "step": 11750
+    },
+    {
+      "epoch": 37.00430379746835,
+      "grad_norm": 0.9744784235954285,
+      "learning_rate": 1.4205344585091422e-05,
+      "loss": 0.0235,
+      "step": 11760
+    },
+    {
+      "epoch": 37.00493670886076,
+      "grad_norm": 0.0027787466533482075,
+      "learning_rate": 1.4170182841068918e-05,
+      "loss": 0.0331,
+      "step": 11770
+    },
+    {
+      "epoch": 37.00556962025316,
+      "grad_norm": 2.0447912216186523,
+      "learning_rate": 1.4135021097046413e-05,
+      "loss": 0.0689,
+      "step": 11780
+    },
+    {
+      "epoch": 37.00620253164557,
+      "grad_norm": 0.0037356666289269924,
+      "learning_rate": 1.4099859353023911e-05,
+      "loss": 0.1165,
+      "step": 11790
+    },
+    {
+      "epoch": 37.00683544303797,
+      "grad_norm": 0.00903794914484024,
+      "learning_rate": 1.4064697609001406e-05,
+      "loss": 0.0184,
+      "step": 11800
+    },
+    {
+      "epoch": 37.00746835443038,
+      "grad_norm": 2.4592278003692627,
+      "learning_rate": 1.4029535864978905e-05,
+      "loss": 0.0623,
+      "step": 11810
+    },
+    {
+      "epoch": 37.00810126582279,
+      "grad_norm": 30.086641311645508,
+      "learning_rate": 1.39943741209564e-05,
+      "loss": 0.0916,
+      "step": 11820
+    },
+    {
+      "epoch": 37.00873417721519,
+      "grad_norm": 0.006698968820273876,
+      "learning_rate": 1.3959212376933898e-05,
+      "loss": 0.0514,
+      "step": 11830
+    },
+    {
+      "epoch": 37.0093670886076,
+      "grad_norm": 2.0741055011749268,
+      "learning_rate": 1.3924050632911393e-05,
+      "loss": 0.1023,
+      "step": 11840
+    },
+    {
+      "epoch": 37.01,
+      "grad_norm": 3.4092001914978027,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.1079,
+      "step": 11850
+    },
+    {
+      "epoch": 37.010632911392406,
+      "grad_norm": 1.7051777839660645,
+      "learning_rate": 1.3853727144866384e-05,
+      "loss": 0.1146,
+      "step": 11860
+    },
+    {
+      "epoch": 37.01126582278481,
+      "grad_norm": 0.007281308062374592,
+      "learning_rate": 1.3818565400843883e-05,
+      "loss": 0.0593,
+      "step": 11870
+    },
+    {
+      "epoch": 37.011898734177215,
+      "grad_norm": 0.01118686143308878,
+      "learning_rate": 1.3783403656821378e-05,
+      "loss": 0.0342,
+      "step": 11880
+    },
+    {
+      "epoch": 37.012531645569624,
+      "grad_norm": 0.014248767867684364,
+      "learning_rate": 1.3748241912798876e-05,
+      "loss": 0.0685,
+      "step": 11890
+    },
+    {
+      "epoch": 37.013164556962025,
+      "grad_norm": 1.9155502319335938,
+      "learning_rate": 1.3713080168776371e-05,
+      "loss": 0.0562,
+      "step": 11900
+    },
+    {
+      "epoch": 37.01379746835443,
+      "grad_norm": 2.529233694076538,
+      "learning_rate": 1.367791842475387e-05,
+      "loss": 0.0588,
+      "step": 11910
+    },
+    {
+      "epoch": 37.014430379746834,
+      "grad_norm": 0.006225684192031622,
+      "learning_rate": 1.3642756680731364e-05,
+      "loss": 0.046,
+      "step": 11920
+    },
+    {
+      "epoch": 37.01506329113924,
+      "grad_norm": 1.680939793586731,
+      "learning_rate": 1.3607594936708861e-05,
+      "loss": 0.0775,
+      "step": 11930
+    },
+    {
+      "epoch": 37.01569620253164,
+      "grad_norm": 1.7170394659042358,
+      "learning_rate": 1.3572433192686358e-05,
+      "loss": 0.0498,
+      "step": 11940
+    },
+    {
+      "epoch": 37.01632911392405,
+      "grad_norm": 5.160343170166016,
+      "learning_rate": 1.3537271448663854e-05,
+      "loss": 0.0707,
+      "step": 11950
+    },
+    {
+      "epoch": 37.01696202531645,
+      "grad_norm": 2.1870486736297607,
+      "learning_rate": 1.3502109704641349e-05,
+      "loss": 0.0583,
+      "step": 11960
+    },
+    {
+      "epoch": 37.01759493670886,
+      "grad_norm": 2.745607852935791,
+      "learning_rate": 1.3466947960618847e-05,
+      "loss": 0.0737,
+      "step": 11970
+    },
+    {
+      "epoch": 37.01822784810127,
+      "grad_norm": 0.005020118784159422,
+      "learning_rate": 1.3431786216596342e-05,
+      "loss": 0.0867,
+      "step": 11980
+    },
+    {
+      "epoch": 37.01886075949367,
+      "grad_norm": 2.2842442989349365,
+      "learning_rate": 1.339662447257384e-05,
+      "loss": 0.0794,
+      "step": 11990
+    },
+    {
+      "epoch": 37.01949367088608,
+      "grad_norm": 0.011187170632183552,
+      "learning_rate": 1.3361462728551336e-05,
+      "loss": 0.0863,
+      "step": 12000
+    },
+    {
+      "epoch": 37.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.04793384671211243,
+      "eval_runtime": 885.4471,
+      "eval_samples_per_second": 0.588,
+      "eval_steps_per_second": 0.075,
+      "step": 12008
+    },
+    {
+      "epoch": 38.00012658227848,
+      "grad_norm": 0.004787384066730738,
+      "learning_rate": 1.3326300984528834e-05,
+      "loss": 0.0004,
+      "step": 12010
+    },
+    {
+      "epoch": 38.000759493670884,
+      "grad_norm": 0.006256919354200363,
+      "learning_rate": 1.3291139240506329e-05,
+      "loss": 0.0641,
+      "step": 12020
+    },
+    {
+      "epoch": 38.00139240506329,
+      "grad_norm": 2.1292147636413574,
+      "learning_rate": 1.3255977496483826e-05,
+      "loss": 0.0635,
+      "step": 12030
+    },
+    {
+      "epoch": 38.00202531645569,
+      "grad_norm": 1.8341985940933228,
+      "learning_rate": 1.3220815752461324e-05,
+      "loss": 0.0635,
+      "step": 12040
+    },
+    {
+      "epoch": 38.0026582278481,
+      "grad_norm": 0.012860477901995182,
+      "learning_rate": 1.3185654008438819e-05,
+      "loss": 0.0439,
+      "step": 12050
+    },
+    {
+      "epoch": 38.00329113924051,
+      "grad_norm": 0.005083575379103422,
+      "learning_rate": 1.3150492264416317e-05,
+      "loss": 0.0468,
+      "step": 12060
+    },
+    {
+      "epoch": 38.00392405063291,
+      "grad_norm": 1.4542759656906128,
+      "learning_rate": 1.3115330520393812e-05,
+      "loss": 0.0388,
+      "step": 12070
+    },
+    {
+      "epoch": 38.00455696202532,
+      "grad_norm": 1.959895372390747,
+      "learning_rate": 1.308016877637131e-05,
+      "loss": 0.0416,
+      "step": 12080
+    },
+    {
+      "epoch": 38.00518987341772,
+      "grad_norm": 2.081799030303955,
+      "learning_rate": 1.3045007032348805e-05,
+      "loss": 0.1381,
+      "step": 12090
+    },
+    {
+      "epoch": 38.00582278481013,
+      "grad_norm": 0.004892106633633375,
+      "learning_rate": 1.3009845288326302e-05,
+      "loss": 0.0568,
+      "step": 12100
+    },
+    {
+      "epoch": 38.00645569620253,
+      "grad_norm": 0.005953035783022642,
+      "learning_rate": 1.2974683544303799e-05,
+      "loss": 0.0619,
+      "step": 12110
+    },
+    {
+      "epoch": 38.00708860759494,
+      "grad_norm": 0.00476424815133214,
+      "learning_rate": 1.2939521800281295e-05,
+      "loss": 0.0487,
+      "step": 12120
+    },
+    {
+      "epoch": 38.00772151898734,
+      "grad_norm": 0.004643861670047045,
+      "learning_rate": 1.290436005625879e-05,
+      "loss": 0.078,
+      "step": 12130
+    },
+    {
+      "epoch": 38.00835443037975,
+      "grad_norm": 0.004535979591310024,
+      "learning_rate": 1.2869198312236289e-05,
+      "loss": 0.1059,
+      "step": 12140
+    },
+    {
+      "epoch": 38.008987341772155,
+      "grad_norm": 0.9413653612136841,
+      "learning_rate": 1.2834036568213783e-05,
+      "loss": 0.0808,
+      "step": 12150
+    },
+    {
+      "epoch": 38.009620253164556,
+      "grad_norm": 0.23469361662864685,
+      "learning_rate": 1.2798874824191282e-05,
+      "loss": 0.0552,
+      "step": 12160
+    },
+    {
+      "epoch": 38.010253164556964,
+      "grad_norm": 9.23499870300293,
+      "learning_rate": 1.2763713080168777e-05,
+      "loss": 0.1727,
+      "step": 12170
+    },
+    {
+      "epoch": 38.010886075949365,
+      "grad_norm": 2.0388455390930176,
+      "learning_rate": 1.2728551336146275e-05,
+      "loss": 0.1044,
+      "step": 12180
+    },
+    {
+      "epoch": 38.01151898734177,
+      "grad_norm": 1.715149164199829,
+      "learning_rate": 1.269338959212377e-05,
+      "loss": 0.0649,
+      "step": 12190
+    },
+    {
+      "epoch": 38.012151898734174,
+      "grad_norm": 0.0031002135947346687,
+      "learning_rate": 1.2658227848101267e-05,
+      "loss": 0.0792,
+      "step": 12200
+    },
+    {
+      "epoch": 38.01278481012658,
+      "grad_norm": 0.00621592253446579,
+      "learning_rate": 1.2623066104078762e-05,
+      "loss": 0.0711,
+      "step": 12210
+    },
+    {
+      "epoch": 38.01341772151899,
+      "grad_norm": 0.010905969887971878,
+      "learning_rate": 1.258790436005626e-05,
+      "loss": 0.05,
+      "step": 12220
+    },
+    {
+      "epoch": 38.01405063291139,
+      "grad_norm": 0.014951992779970169,
+      "learning_rate": 1.2552742616033755e-05,
+      "loss": 0.0523,
+      "step": 12230
+    },
+    {
+      "epoch": 38.0146835443038,
+      "grad_norm": 1.8830513954162598,
+      "learning_rate": 1.2517580872011253e-05,
+      "loss": 0.0341,
+      "step": 12240
+    },
+    {
+      "epoch": 38.0153164556962,
+      "grad_norm": 2.3067173957824707,
+      "learning_rate": 1.248241912798875e-05,
+      "loss": 0.0422,
+      "step": 12250
+    },
+    {
+      "epoch": 38.01594936708861,
+      "grad_norm": 0.008932225406169891,
+      "learning_rate": 1.2447257383966246e-05,
+      "loss": 0.0639,
+      "step": 12260
+    },
+    {
+      "epoch": 38.01658227848101,
+      "grad_norm": 0.00562430452555418,
+      "learning_rate": 1.2412095639943743e-05,
+      "loss": 0.0559,
+      "step": 12270
+    },
+    {
+      "epoch": 38.01721518987342,
+      "grad_norm": 0.003999199718236923,
+      "learning_rate": 1.237693389592124e-05,
+      "loss": 0.0529,
+      "step": 12280
+    },
+    {
+      "epoch": 38.01784810126582,
+      "grad_norm": 0.01232972927391529,
+      "learning_rate": 1.2341772151898735e-05,
+      "loss": 0.0269,
+      "step": 12290
+    },
+    {
+      "epoch": 38.01848101265823,
+      "grad_norm": 0.004044785164296627,
+      "learning_rate": 1.2306610407876231e-05,
+      "loss": 0.0154,
+      "step": 12300
+    },
+    {
+      "epoch": 38.019113924050636,
+      "grad_norm": 0.0040607457049191,
+      "learning_rate": 1.2271448663853728e-05,
+      "loss": 0.0157,
+      "step": 12310
+    },
+    {
+      "epoch": 38.01974683544304,
+      "grad_norm": 1.6415541172027588,
+      "learning_rate": 1.2236286919831224e-05,
+      "loss": 0.0934,
+      "step": 12320
+    },
+    {
+      "epoch": 38.02,
+      "eval_accuracy": 0.9712092130518234,
+      "eval_loss": 0.04641543701291084,
+      "eval_runtime": 886.1854,
+      "eval_samples_per_second": 0.588,
+      "eval_steps_per_second": 0.074,
+      "step": 12324
+    },
+    {
+      "epoch": 39.00037974683544,
+      "grad_norm": 1.787951111793518,
+      "learning_rate": 1.2201125175808721e-05,
+      "loss": 0.049,
+      "step": 12330
+    },
+    {
+      "epoch": 39.00101265822785,
+      "grad_norm": 0.0033679732587188482,
+      "learning_rate": 1.2165963431786218e-05,
+      "loss": 0.0464,
+      "step": 12340
+    },
+    {
+      "epoch": 39.00164556962025,
+      "grad_norm": 0.003388448618352413,
+      "learning_rate": 1.2130801687763714e-05,
+      "loss": 0.0743,
+      "step": 12350
+    },
+    {
+      "epoch": 39.00227848101266,
+      "grad_norm": 0.004701158031821251,
+      "learning_rate": 1.2095639943741211e-05,
+      "loss": 0.0753,
+      "step": 12360
+    },
+    {
+      "epoch": 39.00291139240506,
+      "grad_norm": 4.90346097946167,
+      "learning_rate": 1.2060478199718708e-05,
+      "loss": 0.0743,
+      "step": 12370
+    },
+    {
+      "epoch": 39.00354430379747,
+      "grad_norm": 1.0210288763046265,
+      "learning_rate": 1.2025316455696203e-05,
+      "loss": 0.0239,
+      "step": 12380
+    },
+    {
+      "epoch": 39.00417721518988,
+      "grad_norm": 0.0035583917051553726,
+      "learning_rate": 1.19901547116737e-05,
+      "loss": 0.0558,
+      "step": 12390
+    },
+    {
+      "epoch": 39.00481012658228,
+      "grad_norm": 0.010518099181354046,
+      "learning_rate": 1.1954992967651196e-05,
+      "loss": 0.0485,
+      "step": 12400
+    },
+    {
+      "epoch": 39.005443037974686,
+      "grad_norm": 1.8886620998382568,
+      "learning_rate": 1.1919831223628692e-05,
+      "loss": 0.0778,
+      "step": 12410
+    },
+    {
+      "epoch": 39.00607594936709,
+      "grad_norm": 1.6945433616638184,
+      "learning_rate": 1.1884669479606189e-05,
+      "loss": 0.0958,
+      "step": 12420
+    },
+    {
+      "epoch": 39.006708860759495,
+      "grad_norm": 0.005481211934238672,
+      "learning_rate": 1.1849507735583686e-05,
+      "loss": 0.072,
+      "step": 12430
+    },
+    {
+      "epoch": 39.0073417721519,
+      "grad_norm": 0.009708485566079617,
+      "learning_rate": 1.1814345991561182e-05,
+      "loss": 0.0492,
+      "step": 12440
+    },
+    {
+      "epoch": 39.007974683544305,
+      "grad_norm": 2.023264169692993,
+      "learning_rate": 1.1779184247538679e-05,
+      "loss": 0.0423,
+      "step": 12450
+    },
+    {
+      "epoch": 39.008607594936706,
+      "grad_norm": 0.002309863455593586,
+      "learning_rate": 1.1744022503516176e-05,
+      "loss": 0.0424,
+      "step": 12460
+    },
+    {
+      "epoch": 39.009240506329114,
+      "grad_norm": 3.2563633918762207,
+      "learning_rate": 1.170886075949367e-05,
+      "loss": 0.0603,
+      "step": 12470
+    },
+    {
+      "epoch": 39.00987341772152,
+      "grad_norm": 0.005214280914515257,
+      "learning_rate": 1.1673699015471167e-05,
+      "loss": 0.0973,
+      "step": 12480
+    },
+    {
+      "epoch": 39.01050632911392,
+      "grad_norm": 0.0032180920243263245,
+      "learning_rate": 1.1638537271448664e-05,
+      "loss": 0.0352,
+      "step": 12490
+    },
+    {
+      "epoch": 39.01113924050633,
+      "grad_norm": 0.0023686427157372236,
+      "learning_rate": 1.160337552742616e-05,
+      "loss": 0.0459,
+      "step": 12500
+    },
+    {
+      "epoch": 39.01177215189873,
+      "grad_norm": 0.0045258644968271255,
+      "learning_rate": 1.1568213783403657e-05,
+      "loss": 0.0683,
+      "step": 12510
+    },
+    {
+      "epoch": 39.01240506329114,
+      "grad_norm": 0.0025700768455863,
+      "learning_rate": 1.1533052039381154e-05,
+      "loss": 0.0412,
+      "step": 12520
+    },
+    {
+      "epoch": 39.01303797468354,
+      "grad_norm": 0.0044273072853684425,
+      "learning_rate": 1.149789029535865e-05,
+      "loss": 0.0338,
+      "step": 12530
+    },
+    {
+      "epoch": 39.01367088607595,
+      "grad_norm": 0.007956129498779774,
+      "learning_rate": 1.1462728551336147e-05,
+      "loss": 0.0658,
+      "step": 12540
+    },
+    {
+      "epoch": 39.01430379746836,
+      "grad_norm": 0.005596678238362074,
+      "learning_rate": 1.1427566807313644e-05,
+      "loss": 0.0658,
+      "step": 12550
+    },
+    {
+      "epoch": 39.01493670886076,
+      "grad_norm": 0.003668359015136957,
+      "learning_rate": 1.139240506329114e-05,
+      "loss": 0.0757,
+      "step": 12560
+    },
+    {
+      "epoch": 39.01556962025317,
+      "grad_norm": 0.0037127614486962557,
+      "learning_rate": 1.1357243319268635e-05,
+      "loss": 0.0477,
+      "step": 12570
+    },
+    {
+      "epoch": 39.01620253164557,
+      "grad_norm": 1.9813635349273682,
+      "learning_rate": 1.1322081575246132e-05,
+      "loss": 0.0453,
+      "step": 12580
+    },
+    {
+      "epoch": 39.01683544303798,
+      "grad_norm": 0.004934421740472317,
+      "learning_rate": 1.1286919831223628e-05,
+      "loss": 0.0507,
+      "step": 12590
+    },
+    {
+      "epoch": 39.01746835443038,
+      "grad_norm": 0.0019371870439499617,
+      "learning_rate": 1.1251758087201125e-05,
+      "loss": 0.053,
+      "step": 12600
+    },
+    {
+      "epoch": 39.018101265822786,
+      "grad_norm": 0.011108608916401863,
+      "learning_rate": 1.1216596343178622e-05,
+      "loss": 0.0283,
+      "step": 12610
+    },
+    {
+      "epoch": 39.01873417721519,
+      "grad_norm": 2.060152530670166,
+      "learning_rate": 1.1181434599156118e-05,
+      "loss": 0.0567,
+      "step": 12620
+    },
+    {
+      "epoch": 39.019367088607595,
+      "grad_norm": 0.002437671646475792,
+      "learning_rate": 1.1146272855133615e-05,
+      "loss": 0.0331,
+      "step": 12630
+    },
+    {
+      "epoch": 39.02,
+      "grad_norm": 0.0036968346685171127,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.0927,
+      "step": 12640
+    },
+    {
+      "epoch": 39.02,
+      "eval_accuracy": 0.9712092130518234,
+      "eval_loss": 0.04621642827987671,
+      "eval_runtime": 901.4604,
+      "eval_samples_per_second": 0.578,
+      "eval_steps_per_second": 0.073,
+      "step": 12640
+    },
+    {
+      "epoch": 40.00063291139241,
+      "grad_norm": 0.0033778073266148567,
+      "learning_rate": 1.1075949367088608e-05,
+      "loss": 0.0145,
+      "step": 12650
+    },
+    {
+      "epoch": 40.00126582278481,
+      "grad_norm": 2.246983528137207,
+      "learning_rate": 1.1040787623066103e-05,
+      "loss": 0.0527,
+      "step": 12660
+    },
+    {
+      "epoch": 40.00189873417722,
+      "grad_norm": 0.002830845071002841,
+      "learning_rate": 1.1005625879043602e-05,
+      "loss": 0.0437,
+      "step": 12670
+    },
+    {
+      "epoch": 40.00253164556962,
+      "grad_norm": 1.5313416719436646,
+      "learning_rate": 1.0970464135021098e-05,
+      "loss": 0.0223,
+      "step": 12680
+    },
+    {
+      "epoch": 40.00316455696203,
+      "grad_norm": 1.529272437095642,
+      "learning_rate": 1.0935302390998595e-05,
+      "loss": 0.0464,
+      "step": 12690
+    },
+    {
+      "epoch": 40.00379746835443,
+      "grad_norm": 0.0037761295679956675,
+      "learning_rate": 1.0900140646976091e-05,
+      "loss": 0.0859,
+      "step": 12700
+    },
+    {
+      "epoch": 40.004430379746836,
+      "grad_norm": 0.003269642125815153,
+      "learning_rate": 1.0864978902953588e-05,
+      "loss": 0.068,
+      "step": 12710
+    },
+    {
+      "epoch": 40.00506329113924,
+      "grad_norm": 2.6645495891571045,
+      "learning_rate": 1.0829817158931085e-05,
+      "loss": 0.0689,
+      "step": 12720
+    },
+    {
+      "epoch": 40.005696202531645,
+      "grad_norm": 2.0125551223754883,
+      "learning_rate": 1.079465541490858e-05,
+      "loss": 0.0139,
+      "step": 12730
+    },
+    {
+      "epoch": 40.00632911392405,
+      "grad_norm": 0.002355450764298439,
+      "learning_rate": 1.0759493670886076e-05,
+      "loss": 0.0098,
+      "step": 12740
+    },
+    {
+      "epoch": 40.006962025316454,
+      "grad_norm": 0.004568996839225292,
+      "learning_rate": 1.0724331926863573e-05,
+      "loss": 0.0692,
+      "step": 12750
+    },
+    {
+      "epoch": 40.00759493670886,
+      "grad_norm": 0.020498152822256088,
+      "learning_rate": 1.068917018284107e-05,
+      "loss": 0.0148,
+      "step": 12760
+    },
+    {
+      "epoch": 40.008227848101264,
+      "grad_norm": 0.006843405310064554,
+      "learning_rate": 1.0654008438818566e-05,
+      "loss": 0.0895,
+      "step": 12770
+    },
+    {
+      "epoch": 40.00886075949367,
+      "grad_norm": 1.5150411128997803,
+      "learning_rate": 1.0618846694796063e-05,
+      "loss": 0.0739,
+      "step": 12780
+    },
+    {
+      "epoch": 40.00949367088607,
+      "grad_norm": 0.003931212704628706,
+      "learning_rate": 1.058368495077356e-05,
+      "loss": 0.1223,
+      "step": 12790
+    },
+    {
+      "epoch": 40.01012658227848,
+      "grad_norm": 0.007164576090872288,
+      "learning_rate": 1.0548523206751056e-05,
+      "loss": 0.0627,
+      "step": 12800
+    },
+    {
+      "epoch": 40.01075949367089,
+      "grad_norm": 0.009334239177405834,
+      "learning_rate": 1.0513361462728553e-05,
+      "loss": 0.0764,
+      "step": 12810
+    },
+    {
+      "epoch": 40.01139240506329,
+      "grad_norm": 5.856595516204834,
+      "learning_rate": 1.047819971870605e-05,
+      "loss": 0.1549,
+      "step": 12820
+    },
+    {
+      "epoch": 40.0120253164557,
+      "grad_norm": 0.006400711834430695,
+      "learning_rate": 1.0443037974683544e-05,
+      "loss": 0.1071,
+      "step": 12830
+    },
+    {
+      "epoch": 40.0126582278481,
+      "grad_norm": 0.005136122461408377,
+      "learning_rate": 1.0407876230661041e-05,
+      "loss": 0.0683,
+      "step": 12840
+    },
+    {
+      "epoch": 40.01329113924051,
+      "grad_norm": 1.7356547117233276,
+      "learning_rate": 1.0372714486638538e-05,
+      "loss": 0.0767,
+      "step": 12850
+    },
+    {
+      "epoch": 40.01392405063291,
+      "grad_norm": 0.008359666913747787,
+      "learning_rate": 1.0337552742616034e-05,
+      "loss": 0.0188,
+      "step": 12860
+    },
+    {
+      "epoch": 40.01455696202532,
+      "grad_norm": 0.0056666117161512375,
+      "learning_rate": 1.030239099859353e-05,
+      "loss": 0.1222,
+      "step": 12870
+    },
+    {
+      "epoch": 40.01518987341772,
+      "grad_norm": 3.0140154361724854,
+      "learning_rate": 1.0267229254571027e-05,
+      "loss": 0.0681,
+      "step": 12880
+    },
+    {
+      "epoch": 40.015822784810126,
+      "grad_norm": 2.062131881713867,
+      "learning_rate": 1.0232067510548524e-05,
+      "loss": 0.1132,
+      "step": 12890
+    },
+    {
+      "epoch": 40.016455696202534,
+      "grad_norm": 0.027009891346096992,
+      "learning_rate": 1.019690576652602e-05,
+      "loss": 0.0292,
+      "step": 12900
+    },
+    {
+      "epoch": 40.017088607594935,
+      "grad_norm": 0.004601217340677977,
+      "learning_rate": 1.0161744022503517e-05,
+      "loss": 0.0366,
+      "step": 12910
+    },
+    {
+      "epoch": 40.017721518987344,
+      "grad_norm": 0.004661747720092535,
+      "learning_rate": 1.0126582278481012e-05,
+      "loss": 0.0386,
+      "step": 12920
+    },
+    {
+      "epoch": 40.018354430379745,
+      "grad_norm": 0.00786674115806818,
+      "learning_rate": 1.0091420534458509e-05,
+      "loss": 0.0872,
+      "step": 12930
+    },
+    {
+      "epoch": 40.01898734177215,
+      "grad_norm": 0.00579045619815588,
+      "learning_rate": 1.0056258790436006e-05,
+      "loss": 0.0319,
+      "step": 12940
+    },
+    {
+      "epoch": 40.019620253164554,
+      "grad_norm": 1.1140903234481812,
+      "learning_rate": 1.0021097046413502e-05,
+      "loss": 0.0254,
+      "step": 12950
+    },
+    {
+      "epoch": 40.02,
+      "eval_accuracy": 0.9731285988483686,
+      "eval_loss": 0.04482452943921089,
+      "eval_runtime": 863.7445,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.076,
+      "step": 12956
+    },
+    {
+      "epoch": 41.00025316455696,
+      "grad_norm": 1.9278745651245117,
+      "learning_rate": 9.985935302390999e-06,
+      "loss": 0.1063,
+      "step": 12960
+    },
+    {
+      "epoch": 41.00088607594937,
+      "grad_norm": 0.011978060938417912,
+      "learning_rate": 9.950773558368495e-06,
+      "loss": 0.047,
+      "step": 12970
+    },
+    {
+      "epoch": 41.001518987341775,
+      "grad_norm": 0.0027657151222229004,
+      "learning_rate": 9.915611814345992e-06,
+      "loss": 0.0534,
+      "step": 12980
+    },
+    {
+      "epoch": 41.002151898734176,
+      "grad_norm": 0.003485744819045067,
+      "learning_rate": 9.880450070323489e-06,
+      "loss": 0.045,
+      "step": 12990
+    },
+    {
+      "epoch": 41.002784810126585,
+      "grad_norm": 0.019399795681238174,
+      "learning_rate": 9.845288326300985e-06,
+      "loss": 0.0372,
+      "step": 13000
+    },
+    {
+      "epoch": 41.003417721518986,
+      "grad_norm": 2.4368183612823486,
+      "learning_rate": 9.81012658227848e-06,
+      "loss": 0.0384,
+      "step": 13010
+    },
+    {
+      "epoch": 41.004050632911394,
+      "grad_norm": 0.0074830991216003895,
+      "learning_rate": 9.774964838255977e-06,
+      "loss": 0.0164,
+      "step": 13020
+    },
+    {
+      "epoch": 41.004683544303795,
+      "grad_norm": 1.7207592725753784,
+      "learning_rate": 9.739803094233474e-06,
+      "loss": 0.025,
+      "step": 13030
+    },
+    {
+      "epoch": 41.0053164556962,
+      "grad_norm": 2.901007652282715,
+      "learning_rate": 9.70464135021097e-06,
+      "loss": 0.0496,
+      "step": 13040
+    },
+    {
+      "epoch": 41.005949367088604,
+      "grad_norm": 2.028984785079956,
+      "learning_rate": 9.669479606188467e-06,
+      "loss": 0.0475,
+      "step": 13050
+    },
+    {
+      "epoch": 41.00658227848101,
+      "grad_norm": 2.09635066986084,
+      "learning_rate": 9.634317862165963e-06,
+      "loss": 0.1207,
+      "step": 13060
+    },
+    {
+      "epoch": 41.00721518987342,
+      "grad_norm": 1.6683522462844849,
+      "learning_rate": 9.59915611814346e-06,
+      "loss": 0.0922,
+      "step": 13070
+    },
+    {
+      "epoch": 41.00784810126582,
+      "grad_norm": 5.185677528381348,
+      "learning_rate": 9.563994374120957e-06,
+      "loss": 0.0592,
+      "step": 13080
+    },
+    {
+      "epoch": 41.00848101265823,
+      "grad_norm": 1.2516071796417236,
+      "learning_rate": 9.528832630098453e-06,
+      "loss": 0.0768,
+      "step": 13090
+    },
+    {
+      "epoch": 41.00911392405063,
+      "grad_norm": 0.0021934949327260256,
+      "learning_rate": 9.49367088607595e-06,
+      "loss": 0.1049,
+      "step": 13100
+    },
+    {
+      "epoch": 41.00974683544304,
+      "grad_norm": 0.0028072514105588198,
+      "learning_rate": 9.458509142053447e-06,
+      "loss": 0.0724,
+      "step": 13110
+    },
+    {
+      "epoch": 41.01037974683544,
+      "grad_norm": 0.0021622704807668924,
+      "learning_rate": 9.423347398030943e-06,
+      "loss": 0.05,
+      "step": 13120
+    },
+    {
+      "epoch": 41.01101265822785,
+      "grad_norm": 0.0022336526308208704,
+      "learning_rate": 9.38818565400844e-06,
+      "loss": 0.0391,
+      "step": 13130
+    },
+    {
+      "epoch": 41.011645569620256,
+      "grad_norm": 3.3134849071502686,
+      "learning_rate": 9.353023909985936e-06,
+      "loss": 0.0497,
+      "step": 13140
+    },
+    {
+      "epoch": 41.01227848101266,
+      "grad_norm": 2.5143470764160156,
+      "learning_rate": 9.317862165963433e-06,
+      "loss": 0.0941,
+      "step": 13150
+    },
+    {
+      "epoch": 41.012911392405066,
+      "grad_norm": 0.003348015947267413,
+      "learning_rate": 9.28270042194093e-06,
+      "loss": 0.0363,
+      "step": 13160
+    },
+    {
+      "epoch": 41.01354430379747,
+      "grad_norm": 0.00388367404229939,
+      "learning_rate": 9.247538677918426e-06,
+      "loss": 0.0579,
+      "step": 13170
+    },
+    {
+      "epoch": 41.014177215189875,
+      "grad_norm": 2.559396266937256,
+      "learning_rate": 9.212376933895921e-06,
+      "loss": 0.0345,
+      "step": 13180
+    },
+    {
+      "epoch": 41.014810126582276,
+      "grad_norm": 0.0016660054679960012,
+      "learning_rate": 9.177215189873418e-06,
+      "loss": 0.0792,
+      "step": 13190
+    },
+    {
+      "epoch": 41.015443037974684,
+      "grad_norm": 1.8110253810882568,
+      "learning_rate": 9.142053445850915e-06,
+      "loss": 0.0731,
+      "step": 13200
+    },
+    {
+      "epoch": 41.016075949367085,
+      "grad_norm": 1.7615784406661987,
+      "learning_rate": 9.106891701828411e-06,
+      "loss": 0.0562,
+      "step": 13210
+    },
+    {
+      "epoch": 41.01670886075949,
+      "grad_norm": 0.002290283562615514,
+      "learning_rate": 9.071729957805908e-06,
+      "loss": 0.0677,
+      "step": 13220
+    },
+    {
+      "epoch": 41.0173417721519,
+      "grad_norm": 1.738335132598877,
+      "learning_rate": 9.036568213783404e-06,
+      "loss": 0.0458,
+      "step": 13230
+    },
+    {
+      "epoch": 41.0179746835443,
+      "grad_norm": 0.003094849642366171,
+      "learning_rate": 9.001406469760901e-06,
+      "loss": 0.0521,
+      "step": 13240
+    },
+    {
+      "epoch": 41.01860759493671,
+      "grad_norm": 2.013655662536621,
+      "learning_rate": 8.966244725738398e-06,
+      "loss": 0.0521,
+      "step": 13250
+    },
+    {
+      "epoch": 41.01924050632911,
+      "grad_norm": 0.0036079958081245422,
+      "learning_rate": 8.931082981715894e-06,
+      "loss": 0.07,
+      "step": 13260
+    },
+    {
+      "epoch": 41.01987341772152,
+      "grad_norm": 0.0022296609822660685,
+      "learning_rate": 8.89592123769339e-06,
+      "loss": 0.043,
+      "step": 13270
+    },
+    {
+      "epoch": 41.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.044986799359321594,
+      "eval_runtime": 875.8809,
+      "eval_samples_per_second": 0.595,
+      "eval_steps_per_second": 0.075,
+      "step": 13272
+    },
+    {
+      "epoch": 42.000506329113925,
+      "grad_norm": 0.004048566333949566,
+      "learning_rate": 8.860759493670886e-06,
+      "loss": 0.0185,
+      "step": 13280
+    },
+    {
+      "epoch": 42.001139240506326,
+      "grad_norm": 0.003146181348711252,
+      "learning_rate": 8.825597749648383e-06,
+      "loss": 0.0577,
+      "step": 13290
+    },
+    {
+      "epoch": 42.001772151898734,
+      "grad_norm": 1.8363380432128906,
+      "learning_rate": 8.79043600562588e-06,
+      "loss": 0.0858,
+      "step": 13300
+    },
+    {
+      "epoch": 42.00240506329114,
+      "grad_norm": 0.010127292014658451,
+      "learning_rate": 8.755274261603376e-06,
+      "loss": 0.0343,
+      "step": 13310
+    },
+    {
+      "epoch": 42.003037974683544,
+      "grad_norm": 0.003768622875213623,
+      "learning_rate": 8.720112517580872e-06,
+      "loss": 0.0568,
+      "step": 13320
+    },
+    {
+      "epoch": 42.00367088607595,
+      "grad_norm": 4.169876575469971,
+      "learning_rate": 8.684950773558369e-06,
+      "loss": 0.0695,
+      "step": 13330
+    },
+    {
+      "epoch": 42.00430379746835,
+      "grad_norm": 1.0784434080123901,
+      "learning_rate": 8.649789029535866e-06,
+      "loss": 0.0755,
+      "step": 13340
+    },
+    {
+      "epoch": 42.00493670886076,
+      "grad_norm": 2.58248233795166,
+      "learning_rate": 8.614627285513362e-06,
+      "loss": 0.051,
+      "step": 13350
+    },
+    {
+      "epoch": 42.00556962025316,
+      "grad_norm": 0.001903594471514225,
+      "learning_rate": 8.579465541490857e-06,
+      "loss": 0.0168,
+      "step": 13360
+    },
+    {
+      "epoch": 42.00620253164557,
+      "grad_norm": 0.0027827962767332792,
+      "learning_rate": 8.544303797468354e-06,
+      "loss": 0.015,
+      "step": 13370
+    },
+    {
+      "epoch": 42.00683544303797,
+      "grad_norm": 1.1959846019744873,
+      "learning_rate": 8.50914205344585e-06,
+      "loss": 0.0271,
+      "step": 13380
+    },
+    {
+      "epoch": 42.00746835443038,
+      "grad_norm": 0.0016241382109001279,
+      "learning_rate": 8.473980309423347e-06,
+      "loss": 0.0164,
+      "step": 13390
+    },
+    {
+      "epoch": 42.00810126582279,
+      "grad_norm": 1.522375226020813,
+      "learning_rate": 8.438818565400844e-06,
+      "loss": 0.0359,
+      "step": 13400
+    },
+    {
+      "epoch": 42.00873417721519,
+      "grad_norm": 1.8420836925506592,
+      "learning_rate": 8.40365682137834e-06,
+      "loss": 0.0759,
+      "step": 13410
+    },
+    {
+      "epoch": 42.0093670886076,
+      "grad_norm": 0.0029541512485593557,
+      "learning_rate": 8.368495077355837e-06,
+      "loss": 0.0002,
+      "step": 13420
+    },
+    {
+      "epoch": 42.01,
+      "grad_norm": 1.9524579048156738,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.0445,
+      "step": 13430
+    },
+    {
+      "epoch": 42.010632911392406,
+      "grad_norm": 1.539594292640686,
+      "learning_rate": 8.29817158931083e-06,
+      "loss": 0.1011,
+      "step": 13440
+    },
+    {
+      "epoch": 42.01126582278481,
+      "grad_norm": 2.301950693130493,
+      "learning_rate": 8.263009845288327e-06,
+      "loss": 0.1041,
+      "step": 13450
+    },
+    {
+      "epoch": 42.011898734177215,
+      "grad_norm": 0.004067489877343178,
+      "learning_rate": 8.227848101265822e-06,
+      "loss": 0.086,
+      "step": 13460
+    },
+    {
+      "epoch": 42.012531645569624,
+      "grad_norm": 2.2229349613189697,
+      "learning_rate": 8.192686357243319e-06,
+      "loss": 0.1346,
+      "step": 13470
+    },
+    {
+      "epoch": 42.013164556962025,
+      "grad_norm": 0.005787982139736414,
+      "learning_rate": 8.157524613220815e-06,
+      "loss": 0.0442,
+      "step": 13480
+    },
+    {
+      "epoch": 42.01379746835443,
+      "grad_norm": 0.0025440517347306013,
+      "learning_rate": 8.122362869198312e-06,
+      "loss": 0.0091,
+      "step": 13490
+    },
+    {
+      "epoch": 42.014430379746834,
+      "grad_norm": 3.941728115081787,
+      "learning_rate": 8.08720112517581e-06,
+      "loss": 0.0812,
+      "step": 13500
+    },
+    {
+      "epoch": 42.01506329113924,
+      "grad_norm": 0.6516626477241516,
+      "learning_rate": 8.052039381153307e-06,
+      "loss": 0.0892,
+      "step": 13510
+    },
+    {
+      "epoch": 42.01569620253164,
+      "grad_norm": 0.002771928673610091,
+      "learning_rate": 8.016877637130803e-06,
+      "loss": 0.0256,
+      "step": 13520
+    },
+    {
+      "epoch": 42.01632911392405,
+      "grad_norm": 1.8547269105911255,
+      "learning_rate": 7.981715893108298e-06,
+      "loss": 0.0839,
+      "step": 13530
+    },
+    {
+      "epoch": 42.01696202531645,
+      "grad_norm": 0.0014696965226903558,
+      "learning_rate": 7.946554149085795e-06,
+      "loss": 0.0356,
+      "step": 13540
+    },
+    {
+      "epoch": 42.01759493670886,
+      "grad_norm": 1.8392219543457031,
+      "learning_rate": 7.911392405063292e-06,
+      "loss": 0.0614,
+      "step": 13550
+    },
+    {
+      "epoch": 42.01822784810127,
+      "grad_norm": 0.0020308021921664476,
+      "learning_rate": 7.876230661040788e-06,
+      "loss": 0.0192,
+      "step": 13560
+    },
+    {
+      "epoch": 42.01886075949367,
+      "grad_norm": 1.8742437362670898,
+      "learning_rate": 7.841068917018285e-06,
+      "loss": 0.0993,
+      "step": 13570
+    },
+    {
+      "epoch": 42.01949367088608,
+      "grad_norm": 2.1382617950439453,
+      "learning_rate": 7.805907172995782e-06,
+      "loss": 0.0695,
+      "step": 13580
+    },
+    {
+      "epoch": 42.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.04481815919280052,
+      "eval_runtime": 864.6404,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.076,
+      "step": 13588
+    },
+    {
+      "epoch": 43.00012658227848,
+      "grad_norm": 0.0016333996318280697,
+      "learning_rate": 7.770745428973278e-06,
+      "loss": 0.0536,
+      "step": 13590
+    },
+    {
+      "epoch": 43.000759493670884,
+      "grad_norm": 2.334437131881714,
+      "learning_rate": 7.735583684950775e-06,
+      "loss": 0.0156,
+      "step": 13600
+    },
+    {
+      "epoch": 43.00139240506329,
+      "grad_norm": 1.4097012281417847,
+      "learning_rate": 7.700421940928271e-06,
+      "loss": 0.0638,
+      "step": 13610
+    },
+    {
+      "epoch": 43.00202531645569,
+      "grad_norm": 0.0024437035899609327,
+      "learning_rate": 7.665260196905766e-06,
+      "loss": 0.0184,
+      "step": 13620
+    },
+    {
+      "epoch": 43.0026582278481,
+      "grad_norm": 1.7008681297302246,
+      "learning_rate": 7.630098452883263e-06,
+      "loss": 0.0664,
+      "step": 13630
+    },
+    {
+      "epoch": 43.00329113924051,
+      "grad_norm": 0.0019239940447732806,
+      "learning_rate": 7.5949367088607605e-06,
+      "loss": 0.1075,
+      "step": 13640
+    },
+    {
+      "epoch": 43.00392405063291,
+      "grad_norm": 0.007515426259487867,
+      "learning_rate": 7.559774964838256e-06,
+      "loss": 0.0423,
+      "step": 13650
+    },
+    {
+      "epoch": 43.00455696202532,
+      "grad_norm": 2.043806552886963,
+      "learning_rate": 7.524613220815753e-06,
+      "loss": 0.057,
+      "step": 13660
+    },
+    {
+      "epoch": 43.00518987341772,
+      "grad_norm": 1.7065999507904053,
+      "learning_rate": 7.4894514767932495e-06,
+      "loss": 0.0867,
+      "step": 13670
+    },
+    {
+      "epoch": 43.00582278481013,
+      "grad_norm": 0.02369196154177189,
+      "learning_rate": 7.454289732770746e-06,
+      "loss": 0.035,
+      "step": 13680
+    },
+    {
+      "epoch": 43.00645569620253,
+      "grad_norm": 2.008984088897705,
+      "learning_rate": 7.419127988748242e-06,
+      "loss": 0.0754,
+      "step": 13690
+    },
+    {
+      "epoch": 43.00708860759494,
+      "grad_norm": 0.002092506969347596,
+      "learning_rate": 7.3839662447257386e-06,
+      "loss": 0.0776,
+      "step": 13700
+    },
+    {
+      "epoch": 43.00772151898734,
+      "grad_norm": 2.8015787601470947,
+      "learning_rate": 7.348804500703235e-06,
+      "loss": 0.0597,
+      "step": 13710
+    },
+    {
+      "epoch": 43.00835443037975,
+      "grad_norm": 0.002701831515878439,
+      "learning_rate": 7.313642756680732e-06,
+      "loss": 0.0401,
+      "step": 13720
+    },
+    {
+      "epoch": 43.008987341772155,
+      "grad_norm": 0.002448461716994643,
+      "learning_rate": 7.2784810126582285e-06,
+      "loss": 0.061,
+      "step": 13730
+    },
+    {
+      "epoch": 43.009620253164556,
+      "grad_norm": 0.001818045973777771,
+      "learning_rate": 7.243319268635724e-06,
+      "loss": 0.0198,
+      "step": 13740
+    },
+    {
+      "epoch": 43.010253164556964,
+      "grad_norm": 1.9535473585128784,
+      "learning_rate": 7.208157524613221e-06,
+      "loss": 0.1018,
+      "step": 13750
+    },
+    {
+      "epoch": 43.010886075949365,
+      "grad_norm": 0.0013298611156642437,
+      "learning_rate": 7.1729957805907175e-06,
+      "loss": 0.0232,
+      "step": 13760
+    },
+    {
+      "epoch": 43.01151898734177,
+      "grad_norm": 0.002159226918593049,
+      "learning_rate": 7.137834036568214e-06,
+      "loss": 0.0652,
+      "step": 13770
+    },
+    {
+      "epoch": 43.012151898734174,
+      "grad_norm": 2.004695177078247,
+      "learning_rate": 7.102672292545711e-06,
+      "loss": 0.0787,
+      "step": 13780
+    },
+    {
+      "epoch": 43.01278481012658,
+      "grad_norm": 0.002241474576294422,
+      "learning_rate": 7.0675105485232066e-06,
+      "loss": 0.0564,
+      "step": 13790
+    },
+    {
+      "epoch": 43.01341772151899,
+      "grad_norm": 0.0015954429982230067,
+      "learning_rate": 7.032348804500703e-06,
+      "loss": 0.0715,
+      "step": 13800
+    },
+    {
+      "epoch": 43.01405063291139,
+      "grad_norm": 0.04250750690698624,
+      "learning_rate": 6.9971870604782e-06,
+      "loss": 0.0477,
+      "step": 13810
+    },
+    {
+      "epoch": 43.0146835443038,
+      "grad_norm": 0.002124907448887825,
+      "learning_rate": 6.9620253164556965e-06,
+      "loss": 0.0152,
+      "step": 13820
+    },
+    {
+      "epoch": 43.0153164556962,
+      "grad_norm": 0.00806543417274952,
+      "learning_rate": 6.926863572433192e-06,
+      "loss": 0.0002,
+      "step": 13830
+    },
+    {
+      "epoch": 43.01594936708861,
+      "grad_norm": 0.0016234181821346283,
+      "learning_rate": 6.891701828410689e-06,
+      "loss": 0.0443,
+      "step": 13840
+    },
+    {
+      "epoch": 43.01658227848101,
+      "grad_norm": 0.0016976363258436322,
+      "learning_rate": 6.8565400843881855e-06,
+      "loss": 0.0811,
+      "step": 13850
+    },
+    {
+      "epoch": 43.01721518987342,
+      "grad_norm": 0.002512180246412754,
+      "learning_rate": 6.821378340365682e-06,
+      "loss": 0.0635,
+      "step": 13860
+    },
+    {
+      "epoch": 43.01784810126582,
+      "grad_norm": 0.014850087463855743,
+      "learning_rate": 6.786216596343179e-06,
+      "loss": 0.1057,
+      "step": 13870
+    },
+    {
+      "epoch": 43.01848101265823,
+      "grad_norm": 1.433249831199646,
+      "learning_rate": 6.7510548523206746e-06,
+      "loss": 0.0757,
+      "step": 13880
+    },
+    {
+      "epoch": 43.019113924050636,
+      "grad_norm": 0.0025778491981327534,
+      "learning_rate": 6.715893108298171e-06,
+      "loss": 0.01,
+      "step": 13890
+    },
+    {
+      "epoch": 43.01974683544304,
+      "grad_norm": 0.0015471552032977343,
+      "learning_rate": 6.680731364275668e-06,
+      "loss": 0.0398,
+      "step": 13900
+    },
+    {
+      "epoch": 43.02,
+      "eval_accuracy": 0.9769673704414588,
+      "eval_loss": 0.04403243213891983,
+      "eval_runtime": 918.4271,
+      "eval_samples_per_second": 0.567,
+      "eval_steps_per_second": 0.072,
+      "step": 13904
+    },
+    {
+      "epoch": 44.00037974683544,
+      "grad_norm": 0.002836668398231268,
+      "learning_rate": 6.6455696202531645e-06,
+      "loss": 0.0786,
+      "step": 13910
+    },
+    {
+      "epoch": 44.00101265822785,
+      "grad_norm": 1.8331170082092285,
+      "learning_rate": 6.610407876230662e-06,
+      "loss": 0.0305,
+      "step": 13920
+    },
+    {
+      "epoch": 44.00164556962025,
+      "grad_norm": 0.002194299828261137,
+      "learning_rate": 6.5752461322081586e-06,
+      "loss": 0.032,
+      "step": 13930
+    },
+    {
+      "epoch": 44.00227848101266,
+      "grad_norm": 0.19766800105571747,
+      "learning_rate": 6.540084388185655e-06,
+      "loss": 0.0536,
+      "step": 13940
+    },
+    {
+      "epoch": 44.00291139240506,
+      "grad_norm": 0.0020373559091240168,
+      "learning_rate": 6.504922644163151e-06,
+      "loss": 0.0457,
+      "step": 13950
+    },
+    {
+      "epoch": 44.00354430379747,
+      "grad_norm": 1.5892828702926636,
+      "learning_rate": 6.469760900140648e-06,
+      "loss": 0.091,
+      "step": 13960
+    },
+    {
+      "epoch": 44.00417721518988,
+      "grad_norm": 0.002311758464202285,
+      "learning_rate": 6.434599156118144e-06,
+      "loss": 0.0358,
+      "step": 13970
+    },
+    {
+      "epoch": 44.00481012658228,
+      "grad_norm": 1.6445496082305908,
+      "learning_rate": 6.399437412095641e-06,
+      "loss": 0.063,
+      "step": 13980
+    },
+    {
+      "epoch": 44.005443037974686,
+      "grad_norm": 0.012705209665000439,
+      "learning_rate": 6.3642756680731375e-06,
+      "loss": 0.0365,
+      "step": 13990
+    },
+    {
+      "epoch": 44.00607594936709,
+      "grad_norm": 2.040945053100586,
+      "learning_rate": 6.329113924050633e-06,
+      "loss": 0.094,
+      "step": 14000
+    },
+    {
+      "epoch": 44.006708860759495,
+      "grad_norm": 0.0017350486014038324,
+      "learning_rate": 6.29395218002813e-06,
+      "loss": 0.0668,
+      "step": 14010
+    },
+    {
+      "epoch": 44.0073417721519,
+      "grad_norm": 1.4773507118225098,
+      "learning_rate": 6.2587904360056266e-06,
+      "loss": 0.0389,
+      "step": 14020
+    },
+    {
+      "epoch": 44.007974683544305,
+      "grad_norm": 1.701064109802246,
+      "learning_rate": 6.223628691983123e-06,
+      "loss": 0.0213,
+      "step": 14030
+    },
+    {
+      "epoch": 44.008607594936706,
+      "grad_norm": 2.707864284515381,
+      "learning_rate": 6.18846694796062e-06,
+      "loss": 0.1098,
+      "step": 14040
+    },
+    {
+      "epoch": 44.009240506329114,
+      "grad_norm": 0.10958682745695114,
+      "learning_rate": 6.153305203938116e-06,
+      "loss": 0.1058,
+      "step": 14050
+    },
+    {
+      "epoch": 44.00987341772152,
+      "grad_norm": 1.8932132720947266,
+      "learning_rate": 6.118143459915612e-06,
+      "loss": 0.0379,
+      "step": 14060
+    },
+    {
+      "epoch": 44.01050632911392,
+      "grad_norm": 0.0022401248570531607,
+      "learning_rate": 6.082981715893109e-06,
+      "loss": 0.0445,
+      "step": 14070
+    },
+    {
+      "epoch": 44.01113924050633,
+      "grad_norm": 0.00331929256208241,
+      "learning_rate": 6.0478199718706055e-06,
+      "loss": 0.0807,
+      "step": 14080
+    },
+    {
+      "epoch": 44.01177215189873,
+      "grad_norm": 1.2681527137756348,
+      "learning_rate": 6.012658227848101e-06,
+      "loss": 0.0283,
+      "step": 14090
+    },
+    {
+      "epoch": 44.01240506329114,
+      "grad_norm": 0.0029717902652919292,
+      "learning_rate": 5.977496483825598e-06,
+      "loss": 0.0867,
+      "step": 14100
+    },
+    {
+      "epoch": 44.01303797468354,
+      "grad_norm": 1.9497371912002563,
+      "learning_rate": 5.9423347398030946e-06,
+      "loss": 0.0512,
+      "step": 14110
+    },
+    {
+      "epoch": 44.01367088607595,
+      "grad_norm": 0.0076379780657589436,
+      "learning_rate": 5.907172995780591e-06,
+      "loss": 0.0647,
+      "step": 14120
+    },
+    {
+      "epoch": 44.01430379746836,
+      "grad_norm": 0.0024882254656404257,
+      "learning_rate": 5.872011251758088e-06,
+      "loss": 0.0205,
+      "step": 14130
+    },
+    {
+      "epoch": 44.01493670886076,
+      "grad_norm": 0.0016955797327682376,
+      "learning_rate": 5.836849507735584e-06,
+      "loss": 0.0855,
+      "step": 14140
+    },
+    {
+      "epoch": 44.01556962025317,
+      "grad_norm": 1.9731398820877075,
+      "learning_rate": 5.80168776371308e-06,
+      "loss": 0.0623,
+      "step": 14150
+    },
+    {
+      "epoch": 44.01620253164557,
+      "grad_norm": 0.0020470027811825275,
+      "learning_rate": 5.766526019690577e-06,
+      "loss": 0.0151,
+      "step": 14160
+    },
+    {
+      "epoch": 44.01683544303798,
+      "grad_norm": 0.0018880012212321162,
+      "learning_rate": 5.7313642756680735e-06,
+      "loss": 0.0393,
+      "step": 14170
+    },
+    {
+      "epoch": 44.01746835443038,
+      "grad_norm": 0.003355634631589055,
+      "learning_rate": 5.69620253164557e-06,
+      "loss": 0.024,
+      "step": 14180
+    },
+    {
+      "epoch": 44.018101265822786,
+      "grad_norm": 0.0030694929882884026,
+      "learning_rate": 5.661040787623066e-06,
+      "loss": 0.0285,
+      "step": 14190
+    },
+    {
+      "epoch": 44.01873417721519,
+      "grad_norm": 0.001790488138794899,
+      "learning_rate": 5.6258790436005626e-06,
+      "loss": 0.0361,
+      "step": 14200
+    },
+    {
+      "epoch": 44.019367088607595,
+      "grad_norm": 0.0015849280171096325,
+      "learning_rate": 5.590717299578059e-06,
+      "loss": 0.0828,
+      "step": 14210
+    },
+    {
+      "epoch": 44.02,
+      "grad_norm": 1.4295841455459595,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.0455,
+      "step": 14220
+    },
+    {
+      "epoch": 44.02,
+      "eval_accuracy": 0.9769673704414588,
+      "eval_loss": 0.04359356313943863,
+      "eval_runtime": 920.1088,
+      "eval_samples_per_second": 0.566,
+      "eval_steps_per_second": 0.072,
+      "step": 14220
+    },
+    {
+      "epoch": 45.00063291139241,
+      "grad_norm": 0.006928425282239914,
+      "learning_rate": 5.520393811533052e-06,
+      "loss": 0.0489,
+      "step": 14230
+    },
+    {
+      "epoch": 45.00126582278481,
+      "grad_norm": 0.0019474881701171398,
+      "learning_rate": 5.485232067510549e-06,
+      "loss": 0.0236,
+      "step": 14240
+    },
+    {
+      "epoch": 45.00189873417722,
+      "grad_norm": 1.7010918855667114,
+      "learning_rate": 5.450070323488046e-06,
+      "loss": 0.0395,
+      "step": 14250
+    },
+    {
+      "epoch": 45.00253164556962,
+      "grad_norm": 0.0026303452905267477,
+      "learning_rate": 5.414908579465542e-06,
+      "loss": 0.0223,
+      "step": 14260
+    },
+    {
+      "epoch": 45.00316455696203,
+      "grad_norm": 0.0020397694315761328,
+      "learning_rate": 5.379746835443038e-06,
+      "loss": 0.0431,
+      "step": 14270
+    },
+    {
+      "epoch": 45.00379746835443,
+      "grad_norm": 2.7060606479644775,
+      "learning_rate": 5.344585091420535e-06,
+      "loss": 0.1383,
+      "step": 14280
+    },
+    {
+      "epoch": 45.004430379746836,
+      "grad_norm": 0.0019224517745897174,
+      "learning_rate": 5.309423347398031e-06,
+      "loss": 0.0401,
+      "step": 14290
+    },
+    {
+      "epoch": 45.00506329113924,
+      "grad_norm": 1.5191059112548828,
+      "learning_rate": 5.274261603375528e-06,
+      "loss": 0.0563,
+      "step": 14300
+    },
+    {
+      "epoch": 45.005696202531645,
+      "grad_norm": 1.7753527164459229,
+      "learning_rate": 5.239099859353025e-06,
+      "loss": 0.0452,
+      "step": 14310
+    },
+    {
+      "epoch": 45.00632911392405,
+      "grad_norm": 0.0027101896703243256,
+      "learning_rate": 5.2039381153305205e-06,
+      "loss": 0.0733,
+      "step": 14320
+    },
+    {
+      "epoch": 45.006962025316454,
+      "grad_norm": 0.0018217426259070635,
+      "learning_rate": 5.168776371308017e-06,
+      "loss": 0.0601,
+      "step": 14330
+    },
+    {
+      "epoch": 45.00759493670886,
+      "grad_norm": 1.4243252277374268,
+      "learning_rate": 5.133614627285514e-06,
+      "loss": 0.0995,
+      "step": 14340
+    },
+    {
+      "epoch": 45.008227848101264,
+      "grad_norm": 0.007520051673054695,
+      "learning_rate": 5.09845288326301e-06,
+      "loss": 0.0881,
+      "step": 14350
+    },
+    {
+      "epoch": 45.00886075949367,
+      "grad_norm": 1.565136194229126,
+      "learning_rate": 5.063291139240506e-06,
+      "loss": 0.0648,
+      "step": 14360
+    },
+    {
+      "epoch": 45.00949367088607,
+      "grad_norm": 0.0016765049658715725,
+      "learning_rate": 5.028129395218003e-06,
+      "loss": 0.0751,
+      "step": 14370
+    },
+    {
+      "epoch": 45.01012658227848,
+      "grad_norm": 0.0014801392098888755,
+      "learning_rate": 4.992967651195499e-06,
+      "loss": 0.0246,
+      "step": 14380
+    },
+    {
+      "epoch": 45.01075949367089,
+      "grad_norm": 0.0016110733849927783,
+      "learning_rate": 4.957805907172996e-06,
+      "loss": 0.0396,
+      "step": 14390
+    },
+    {
+      "epoch": 45.01139240506329,
+      "grad_norm": 0.002271553035825491,
+      "learning_rate": 4.922644163150493e-06,
+      "loss": 0.0268,
+      "step": 14400
+    },
+    {
+      "epoch": 45.0120253164557,
+      "grad_norm": 0.002145587233826518,
+      "learning_rate": 4.8874824191279884e-06,
+      "loss": 0.077,
+      "step": 14410
+    },
+    {
+      "epoch": 45.0126582278481,
+      "grad_norm": 0.001231012400239706,
+      "learning_rate": 4.852320675105485e-06,
+      "loss": 0.0633,
+      "step": 14420
+    },
+    {
+      "epoch": 45.01329113924051,
+      "grad_norm": 2.76226544380188,
+      "learning_rate": 4.817158931082982e-06,
+      "loss": 0.0718,
+      "step": 14430
+    },
+    {
+      "epoch": 45.01392405063291,
+      "grad_norm": 0.002031494863331318,
+      "learning_rate": 4.781997187060478e-06,
+      "loss": 0.0752,
+      "step": 14440
+    },
+    {
+      "epoch": 45.01455696202532,
+      "grad_norm": 2.092021942138672,
+      "learning_rate": 4.746835443037975e-06,
+      "loss": 0.0301,
+      "step": 14450
+    },
+    {
+      "epoch": 45.01518987341772,
+      "grad_norm": 0.0018518840661272407,
+      "learning_rate": 4.711673699015472e-06,
+      "loss": 0.0203,
+      "step": 14460
+    },
+    {
+      "epoch": 45.015822784810126,
+      "grad_norm": 0.0020814728923141956,
+      "learning_rate": 4.676511954992968e-06,
+      "loss": 0.0284,
+      "step": 14470
+    },
+    {
+      "epoch": 45.016455696202534,
+      "grad_norm": 0.001710868556983769,
+      "learning_rate": 4.641350210970465e-06,
+      "loss": 0.0455,
+      "step": 14480
+    },
+    {
+      "epoch": 45.017088607594935,
+      "grad_norm": 0.002616223180666566,
+      "learning_rate": 4.606188466947961e-06,
+      "loss": 0.0875,
+      "step": 14490
+    },
+    {
+      "epoch": 45.017721518987344,
+      "grad_norm": 0.0022610658779740334,
+      "learning_rate": 4.571026722925457e-06,
+      "loss": 0.0313,
+      "step": 14500
+    },
+    {
+      "epoch": 45.018354430379745,
+      "grad_norm": 2.640284776687622,
+      "learning_rate": 4.535864978902954e-06,
+      "loss": 0.073,
+      "step": 14510
+    },
+    {
+      "epoch": 45.01898734177215,
+      "grad_norm": 0.0019336823606863618,
+      "learning_rate": 4.5007032348804506e-06,
+      "loss": 0.0416,
+      "step": 14520
+    },
+    {
+      "epoch": 45.019620253164554,
+      "grad_norm": 0.0023440378718078136,
+      "learning_rate": 4.465541490857947e-06,
+      "loss": 0.0423,
+      "step": 14530
+    },
+    {
+      "epoch": 45.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.043731071054935455,
+      "eval_runtime": 852.3259,
+      "eval_samples_per_second": 0.611,
+      "eval_steps_per_second": 0.077,
+      "step": 14536
+    },
+    {
+      "epoch": 46.00025316455696,
+      "grad_norm": 0.0025269004981964827,
+      "learning_rate": 4.430379746835443e-06,
+      "loss": 0.0336,
+      "step": 14540
+    },
+    {
+      "epoch": 46.00088607594937,
+      "grad_norm": 0.0019006689544767141,
+      "learning_rate": 4.39521800281294e-06,
+      "loss": 0.006,
+      "step": 14550
+    },
+    {
+      "epoch": 46.001518987341775,
+      "grad_norm": 0.0017826940165832639,
+      "learning_rate": 4.360056258790436e-06,
+      "loss": 0.0263,
+      "step": 14560
+    },
+    {
+      "epoch": 46.002151898734176,
+      "grad_norm": 0.0013272733194753528,
+      "learning_rate": 4.324894514767933e-06,
+      "loss": 0.0522,
+      "step": 14570
+    },
+    {
+      "epoch": 46.002784810126585,
+      "grad_norm": 1.84543776512146,
+      "learning_rate": 4.289732770745429e-06,
+      "loss": 0.0337,
+      "step": 14580
+    },
+    {
+      "epoch": 46.003417721518986,
+      "grad_norm": 2.3933355808258057,
+      "learning_rate": 4.254571026722925e-06,
+      "loss": 0.0667,
+      "step": 14590
+    },
+    {
+      "epoch": 46.004050632911394,
+      "grad_norm": 1.3483564853668213,
+      "learning_rate": 4.219409282700422e-06,
+      "loss": 0.0358,
+      "step": 14600
+    },
+    {
+      "epoch": 46.004683544303795,
+      "grad_norm": 0.0019236382795497775,
+      "learning_rate": 4.1842475386779186e-06,
+      "loss": 0.105,
+      "step": 14610
+    },
+    {
+      "epoch": 46.0053164556962,
+      "grad_norm": 1.8564695119857788,
+      "learning_rate": 4.149085794655415e-06,
+      "loss": 0.0618,
+      "step": 14620
+    },
+    {
+      "epoch": 46.005949367088604,
+      "grad_norm": 0.0019053075229749084,
+      "learning_rate": 4.113924050632911e-06,
+      "loss": 0.0308,
+      "step": 14630
+    },
+    {
+      "epoch": 46.00658227848101,
+      "grad_norm": 0.888851523399353,
+      "learning_rate": 4.078762306610408e-06,
+      "loss": 0.0046,
+      "step": 14640
+    },
+    {
+      "epoch": 46.00721518987342,
+      "grad_norm": 0.0015094269765540957,
+      "learning_rate": 4.043600562587905e-06,
+      "loss": 0.0597,
+      "step": 14650
+    },
+    {
+      "epoch": 46.00784810126582,
+      "grad_norm": 0.0023652520030736923,
+      "learning_rate": 4.008438818565402e-06,
+      "loss": 0.0719,
+      "step": 14660
+    },
+    {
+      "epoch": 46.00848101265823,
+      "grad_norm": 0.0014642721507698298,
+      "learning_rate": 3.9732770745428975e-06,
+      "loss": 0.0657,
+      "step": 14670
+    },
+    {
+      "epoch": 46.00911392405063,
+      "grad_norm": 2.2750301361083984,
+      "learning_rate": 3.938115330520394e-06,
+      "loss": 0.0324,
+      "step": 14680
+    },
+    {
+      "epoch": 46.00974683544304,
+      "grad_norm": 0.007223762571811676,
+      "learning_rate": 3.902953586497891e-06,
+      "loss": 0.0269,
+      "step": 14690
+    },
+    {
+      "epoch": 46.01037974683544,
+      "grad_norm": 0.0021259100176393986,
+      "learning_rate": 3.867791842475387e-06,
+      "loss": 0.0067,
+      "step": 14700
+    },
+    {
+      "epoch": 46.01101265822785,
+      "grad_norm": 0.00181837088894099,
+      "learning_rate": 3.832630098452883e-06,
+      "loss": 0.0759,
+      "step": 14710
+    },
+    {
+      "epoch": 46.011645569620256,
+      "grad_norm": 0.006951657589524984,
+      "learning_rate": 3.7974683544303802e-06,
+      "loss": 0.0472,
+      "step": 14720
+    },
+    {
+      "epoch": 46.01227848101266,
+      "grad_norm": 0.0019946214742958546,
+      "learning_rate": 3.7623066104078764e-06,
+      "loss": 0.0398,
+      "step": 14730
+    },
+    {
+      "epoch": 46.012911392405066,
+      "grad_norm": 0.0024334420450031757,
+      "learning_rate": 3.727144866385373e-06,
+      "loss": 0.0486,
+      "step": 14740
+    },
+    {
+      "epoch": 46.01354430379747,
+      "grad_norm": 1.671848177909851,
+      "learning_rate": 3.6919831223628693e-06,
+      "loss": 0.0479,
+      "step": 14750
+    },
+    {
+      "epoch": 46.014177215189875,
+      "grad_norm": 0.0020218833815306425,
+      "learning_rate": 3.656821378340366e-06,
+      "loss": 0.0696,
+      "step": 14760
+    },
+    {
+      "epoch": 46.014810126582276,
+      "grad_norm": 2.070733070373535,
+      "learning_rate": 3.621659634317862e-06,
+      "loss": 0.0623,
+      "step": 14770
+    },
+    {
+      "epoch": 46.015443037974684,
+      "grad_norm": 0.00212723552249372,
+      "learning_rate": 3.5864978902953588e-06,
+      "loss": 0.0344,
+      "step": 14780
+    },
+    {
+      "epoch": 46.016075949367085,
+      "grad_norm": 0.007334825582802296,
+      "learning_rate": 3.5513361462728554e-06,
+      "loss": 0.0943,
+      "step": 14790
+    },
+    {
+      "epoch": 46.01670886075949,
+      "grad_norm": 0.00477689690887928,
+      "learning_rate": 3.5161744022503516e-06,
+      "loss": 0.0391,
+      "step": 14800
+    },
+    {
+      "epoch": 46.0173417721519,
+      "grad_norm": 0.0014756337041035295,
+      "learning_rate": 3.4810126582278482e-06,
+      "loss": 0.0578,
+      "step": 14810
+    },
+    {
+      "epoch": 46.0179746835443,
+      "grad_norm": 0.0018279188079759479,
+      "learning_rate": 3.4458509142053444e-06,
+      "loss": 0.1147,
+      "step": 14820
+    },
+    {
+      "epoch": 46.01860759493671,
+      "grad_norm": 2.317160129547119,
+      "learning_rate": 3.410689170182841e-06,
+      "loss": 0.0934,
+      "step": 14830
+    },
+    {
+      "epoch": 46.01924050632911,
+      "grad_norm": 0.007261245045810938,
+      "learning_rate": 3.3755274261603373e-06,
+      "loss": 0.0732,
+      "step": 14840
+    },
+    {
+      "epoch": 46.01987341772152,
+      "grad_norm": 2.7471020221710205,
+      "learning_rate": 3.340365682137834e-06,
+      "loss": 0.0602,
+      "step": 14850
+    },
+    {
+      "epoch": 46.02,
+      "eval_accuracy": 0.9769673704414588,
+      "eval_loss": 0.0437638945877552,
+      "eval_runtime": 913.7302,
+      "eval_samples_per_second": 0.57,
+      "eval_steps_per_second": 0.072,
+      "step": 14852
+    },
+    {
+      "epoch": 47.000506329113925,
+      "grad_norm": 1.6660641431808472,
+      "learning_rate": 3.305203938115331e-06,
+      "loss": 0.0939,
+      "step": 14860
+    },
+    {
+      "epoch": 47.001139240506326,
+      "grad_norm": 1.8603593111038208,
+      "learning_rate": 3.2700421940928276e-06,
+      "loss": 0.1073,
+      "step": 14870
+    },
+    {
+      "epoch": 47.001772151898734,
+      "grad_norm": 0.0009881872683763504,
+      "learning_rate": 3.234880450070324e-06,
+      "loss": 0.059,
+      "step": 14880
+    },
+    {
+      "epoch": 47.00240506329114,
+      "grad_norm": 0.0015653116861358285,
+      "learning_rate": 3.1997187060478204e-06,
+      "loss": 0.0548,
+      "step": 14890
+    },
+    {
+      "epoch": 47.003037974683544,
+      "grad_norm": 1.5998444557189941,
+      "learning_rate": 3.1645569620253167e-06,
+      "loss": 0.0465,
+      "step": 14900
+    },
+    {
+      "epoch": 47.00367088607595,
+      "grad_norm": 0.0016954662278294563,
+      "learning_rate": 3.1293952180028133e-06,
+      "loss": 0.0684,
+      "step": 14910
+    },
+    {
+      "epoch": 47.00430379746835,
+      "grad_norm": 1.6929208040237427,
+      "learning_rate": 3.09423347398031e-06,
+      "loss": 0.0489,
+      "step": 14920
+    },
+    {
+      "epoch": 47.00493670886076,
+      "grad_norm": 1.6551687717437744,
+      "learning_rate": 3.059071729957806e-06,
+      "loss": 0.0799,
+      "step": 14930
+    },
+    {
+      "epoch": 47.00556962025316,
+      "grad_norm": 0.0021378027740865946,
+      "learning_rate": 3.0239099859353028e-06,
+      "loss": 0.0199,
+      "step": 14940
+    },
+    {
+      "epoch": 47.00620253164557,
+      "grad_norm": 0.001334396773017943,
+      "learning_rate": 2.988748241912799e-06,
+      "loss": 0.0181,
+      "step": 14950
+    },
+    {
+      "epoch": 47.00683544303797,
+      "grad_norm": 0.007229403592646122,
+      "learning_rate": 2.9535864978902956e-06,
+      "loss": 0.0588,
+      "step": 14960
+    },
+    {
+      "epoch": 47.00746835443038,
+      "grad_norm": 0.0016747723566368222,
+      "learning_rate": 2.918424753867792e-06,
+      "loss": 0.034,
+      "step": 14970
+    },
+    {
+      "epoch": 47.00810126582279,
+      "grad_norm": 0.8091289401054382,
+      "learning_rate": 2.8832630098452884e-06,
+      "loss": 0.0476,
+      "step": 14980
+    },
+    {
+      "epoch": 47.00873417721519,
+      "grad_norm": 1.8234349489212036,
+      "learning_rate": 2.848101265822785e-06,
+      "loss": 0.0711,
+      "step": 14990
+    },
+    {
+      "epoch": 47.0093670886076,
+      "grad_norm": 1.9740657806396484,
+      "learning_rate": 2.8129395218002813e-06,
+      "loss": 0.0338,
+      "step": 15000
+    },
+    {
+      "epoch": 47.01,
+      "grad_norm": 0.0016077395994216204,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 0.0454,
+      "step": 15010
+    },
+    {
+      "epoch": 47.010632911392406,
+      "grad_norm": 0.004215624183416367,
+      "learning_rate": 2.7426160337552745e-06,
+      "loss": 0.0438,
+      "step": 15020
+    },
+    {
+      "epoch": 47.01126582278481,
+      "grad_norm": 0.0015969170490279794,
+      "learning_rate": 2.707454289732771e-06,
+      "loss": 0.0418,
+      "step": 15030
+    },
+    {
+      "epoch": 47.011898734177215,
+      "grad_norm": 1.9868484735488892,
+      "learning_rate": 2.6722925457102674e-06,
+      "loss": 0.0777,
+      "step": 15040
+    },
+    {
+      "epoch": 47.012531645569624,
+      "grad_norm": 0.006545115727931261,
+      "learning_rate": 2.637130801687764e-06,
+      "loss": 0.0402,
+      "step": 15050
+    },
+    {
+      "epoch": 47.013164556962025,
+      "grad_norm": 0.008419793099164963,
+      "learning_rate": 2.6019690576652602e-06,
+      "loss": 0.0617,
+      "step": 15060
+    },
+    {
+      "epoch": 47.01379746835443,
+      "grad_norm": 1.8613364696502686,
+      "learning_rate": 2.566807313642757e-06,
+      "loss": 0.0477,
+      "step": 15070
+    },
+    {
+      "epoch": 47.014430379746834,
+      "grad_norm": 1.6971709728240967,
+      "learning_rate": 2.531645569620253e-06,
+      "loss": 0.0393,
+      "step": 15080
+    },
+    {
+      "epoch": 47.01506329113924,
+      "grad_norm": 0.0011392009910196066,
+      "learning_rate": 2.4964838255977497e-06,
+      "loss": 0.0481,
+      "step": 15090
+    },
+    {
+      "epoch": 47.01569620253164,
+      "grad_norm": 2.0638561248779297,
+      "learning_rate": 2.4613220815752463e-06,
+      "loss": 0.0648,
+      "step": 15100
+    },
+    {
+      "epoch": 47.01632911392405,
+      "grad_norm": 0.007941323332488537,
+      "learning_rate": 2.4261603375527425e-06,
+      "loss": 0.07,
+      "step": 15110
+    },
+    {
+      "epoch": 47.01696202531645,
+      "grad_norm": 0.0015764615964144468,
+      "learning_rate": 2.390998593530239e-06,
+      "loss": 0.0537,
+      "step": 15120
+    },
+    {
+      "epoch": 47.01759493670886,
+      "grad_norm": 1.6305996179580688,
+      "learning_rate": 2.355836849507736e-06,
+      "loss": 0.0408,
+      "step": 15130
+    },
+    {
+      "epoch": 47.01822784810127,
+      "grad_norm": 1.698041558265686,
+      "learning_rate": 2.3206751054852324e-06,
+      "loss": 0.0421,
+      "step": 15140
+    },
+    {
+      "epoch": 47.01886075949367,
+      "grad_norm": 1.406259536743164,
+      "learning_rate": 2.2855133614627286e-06,
+      "loss": 0.0367,
+      "step": 15150
+    },
+    {
+      "epoch": 47.01949367088608,
+      "grad_norm": 0.0014187191845849156,
+      "learning_rate": 2.2503516174402253e-06,
+      "loss": 0.0407,
+      "step": 15160
+    },
+    {
+      "epoch": 47.02,
+      "eval_accuracy": 0.9750479846449136,
+      "eval_loss": 0.04366590827703476,
+      "eval_runtime": 877.8986,
+      "eval_samples_per_second": 0.593,
+      "eval_steps_per_second": 0.075,
+      "step": 15168
+    },
+    {
+      "epoch": 48.00012658227848,
+      "grad_norm": 0.0016180831007659435,
+      "learning_rate": 2.2151898734177215e-06,
+      "loss": 0.0144,
+      "step": 15170
+    },
+    {
+      "epoch": 48.000759493670884,
+      "grad_norm": 1.924263596534729,
+      "learning_rate": 2.180028129395218e-06,
+      "loss": 0.0882,
+      "step": 15180
+    },
+    {
+      "epoch": 48.00139240506329,
+      "grad_norm": 1.9763191938400269,
+      "learning_rate": 2.1448663853727143e-06,
+      "loss": 0.051,
+      "step": 15190
+    },
+    {
+      "epoch": 48.00202531645569,
+      "grad_norm": 2.0363032817840576,
+      "learning_rate": 2.109704641350211e-06,
+      "loss": 0.048,
+      "step": 15200
+    },
+    {
+      "epoch": 48.0026582278481,
+      "grad_norm": 1.593593716621399,
+      "learning_rate": 2.0745428973277076e-06,
+      "loss": 0.0373,
+      "step": 15210
+    },
+    {
+      "epoch": 48.00329113924051,
+      "grad_norm": 1.4322104454040527,
+      "learning_rate": 2.039381153305204e-06,
+      "loss": 0.0681,
+      "step": 15220
+    },
+    {
+      "epoch": 48.00392405063291,
+      "grad_norm": 0.0023278342559933662,
+      "learning_rate": 2.004219409282701e-06,
+      "loss": 0.0085,
+      "step": 15230
+    },
+    {
+      "epoch": 48.00455696202532,
+      "grad_norm": 2.485316514968872,
+      "learning_rate": 1.969057665260197e-06,
+      "loss": 0.0902,
+      "step": 15240
+    },
+    {
+      "epoch": 48.00518987341772,
+      "grad_norm": 0.0025991136208176613,
+      "learning_rate": 1.9338959212376937e-06,
+      "loss": 0.0845,
+      "step": 15250
+    },
+    {
+      "epoch": 48.00582278481013,
+      "grad_norm": 0.001590245054103434,
+      "learning_rate": 1.8987341772151901e-06,
+      "loss": 0.0218,
+      "step": 15260
+    },
+    {
+      "epoch": 48.00645569620253,
+      "grad_norm": 0.001682686386629939,
+      "learning_rate": 1.8635724331926865e-06,
+      "loss": 0.0623,
+      "step": 15270
+    },
+    {
+      "epoch": 48.00708860759494,
+      "grad_norm": 1.803706169128418,
+      "learning_rate": 1.828410689170183e-06,
+      "loss": 0.0568,
+      "step": 15280
+    },
+    {
+      "epoch": 48.00772151898734,
+      "grad_norm": 0.003340956987813115,
+      "learning_rate": 1.7932489451476794e-06,
+      "loss": 0.0419,
+      "step": 15290
+    },
+    {
+      "epoch": 48.00835443037975,
+      "grad_norm": 0.0032536042854189873,
+      "learning_rate": 1.7580872011251758e-06,
+      "loss": 0.0861,
+      "step": 15300
+    },
+    {
+      "epoch": 48.008987341772155,
+      "grad_norm": 1.4462651014328003,
+      "learning_rate": 1.7229254571026722e-06,
+      "loss": 0.0639,
+      "step": 15310
+    },
+    {
+      "epoch": 48.009620253164556,
+      "grad_norm": 0.0021976635325700045,
+      "learning_rate": 1.6877637130801686e-06,
+      "loss": 0.0624,
+      "step": 15320
+    },
+    {
+      "epoch": 48.010253164556964,
+      "grad_norm": 0.07273641228675842,
+      "learning_rate": 1.6526019690576655e-06,
+      "loss": 0.0662,
+      "step": 15330
+    },
+    {
+      "epoch": 48.010886075949365,
+      "grad_norm": 0.0021416894160211086,
+      "learning_rate": 1.617440225035162e-06,
+      "loss": 0.0562,
+      "step": 15340
+    },
+    {
+      "epoch": 48.01151898734177,
+      "grad_norm": 1.788458228111267,
+      "learning_rate": 1.5822784810126583e-06,
+      "loss": 0.047,
+      "step": 15350
+    },
+    {
+      "epoch": 48.012151898734174,
+      "grad_norm": 0.03488551825284958,
+      "learning_rate": 1.547116736990155e-06,
+      "loss": 0.0406,
+      "step": 15360
+    },
+    {
+      "epoch": 48.01278481012658,
+      "grad_norm": 0.002074979245662689,
+      "learning_rate": 1.5119549929676514e-06,
+      "loss": 0.0363,
+      "step": 15370
+    },
+    {
+      "epoch": 48.01341772151899,
+      "grad_norm": 0.00142192211933434,
+      "learning_rate": 1.4767932489451478e-06,
+      "loss": 0.0473,
+      "step": 15380
+    },
+    {
+      "epoch": 48.01405063291139,
+      "grad_norm": 1.6816853284835815,
+      "learning_rate": 1.4416315049226442e-06,
+      "loss": 0.0692,
+      "step": 15390
+    },
+    {
+      "epoch": 48.0146835443038,
+      "grad_norm": 0.0019629504531621933,
+      "learning_rate": 1.4064697609001406e-06,
+      "loss": 0.0566,
+      "step": 15400
+    },
+    {
+      "epoch": 48.0153164556962,
+      "grad_norm": 1.898903250694275,
+      "learning_rate": 1.3713080168776373e-06,
+      "loss": 0.0501,
+      "step": 15410
+    },
+    {
+      "epoch": 48.01594936708861,
+      "grad_norm": 0.0017211647937074304,
+      "learning_rate": 1.3361462728551337e-06,
+      "loss": 0.0268,
+      "step": 15420
+    },
+    {
+      "epoch": 48.01658227848101,
+      "grad_norm": 0.0072492752224206924,
+      "learning_rate": 1.3009845288326301e-06,
+      "loss": 0.0376,
+      "step": 15430
+    },
+    {
+      "epoch": 48.01721518987342,
+      "grad_norm": 0.0022765600588172674,
+      "learning_rate": 1.2658227848101265e-06,
+      "loss": 0.0423,
+      "step": 15440
+    },
+    {
+      "epoch": 48.01784810126582,
+      "grad_norm": 0.001874367124401033,
+      "learning_rate": 1.2306610407876232e-06,
+      "loss": 0.0247,
+      "step": 15450
+    },
+    {
+      "epoch": 48.01848101265823,
+      "grad_norm": 0.0012237579794600606,
+      "learning_rate": 1.1954992967651196e-06,
+      "loss": 0.0355,
+      "step": 15460
+    },
+    {
+      "epoch": 48.019113924050636,
+      "grad_norm": 1.6173341274261475,
+      "learning_rate": 1.1603375527426162e-06,
+      "loss": 0.07,
+      "step": 15470
+    },
+    {
+      "epoch": 48.01974683544304,
+      "grad_norm": 0.001786409760825336,
+      "learning_rate": 1.1251758087201126e-06,
+      "loss": 0.0435,
+      "step": 15480
+    },
+    {
+      "epoch": 48.02,
+      "eval_accuracy": 0.9769673704414588,
+      "eval_loss": 0.043503157794475555,
+      "eval_runtime": 856.2754,
+      "eval_samples_per_second": 0.608,
+      "eval_steps_per_second": 0.077,
+      "step": 15484
+    },
+    {
+      "epoch": 49.00037974683544,
+      "grad_norm": 1.8200874328613281,
+      "learning_rate": 1.090014064697609e-06,
+      "loss": 0.0552,
+      "step": 15490
+    },
+    {
+      "epoch": 49.00101265822785,
+      "grad_norm": 0.002268126467242837,
+      "learning_rate": 1.0548523206751055e-06,
+      "loss": 0.0194,
+      "step": 15500
+    },
+    {
+      "epoch": 49.00164556962025,
+      "grad_norm": 0.7153782844543457,
+      "learning_rate": 1.019690576652602e-06,
+      "loss": 0.104,
+      "step": 15510
+    },
+    {
+      "epoch": 49.00227848101266,
+      "grad_norm": 0.0011382299708202481,
+      "learning_rate": 9.845288326300985e-07,
+      "loss": 0.0602,
+      "step": 15520
+    },
+    {
+      "epoch": 49.00291139240506,
+      "grad_norm": 1.8290748596191406,
+      "learning_rate": 9.493670886075951e-07,
+      "loss": 0.0151,
+      "step": 15530
+    },
+    {
+      "epoch": 49.00354430379747,
+      "grad_norm": 0.0012674119789153337,
+      "learning_rate": 9.142053445850915e-07,
+      "loss": 0.0333,
+      "step": 15540
+    },
+    {
+      "epoch": 49.00417721518988,
+      "grad_norm": 1.7633377313613892,
+      "learning_rate": 8.790436005625879e-07,
+      "loss": 0.0616,
+      "step": 15550
+    },
+    {
+      "epoch": 49.00481012658228,
+      "grad_norm": 1.8792239427566528,
+      "learning_rate": 8.438818565400843e-07,
+      "loss": 0.0621,
+      "step": 15560
+    },
+    {
+      "epoch": 49.005443037974686,
+      "grad_norm": 0.002730746753513813,
+      "learning_rate": 8.08720112517581e-07,
+      "loss": 0.0706,
+      "step": 15570
+    },
+    {
+      "epoch": 49.00607594936709,
+      "grad_norm": 0.0031078814063221216,
+      "learning_rate": 7.735583684950775e-07,
+      "loss": 0.0644,
+      "step": 15580
+    },
+    {
+      "epoch": 49.006708860759495,
+      "grad_norm": 0.00169693015050143,
+      "learning_rate": 7.383966244725739e-07,
+      "loss": 0.0387,
+      "step": 15590
+    },
+    {
+      "epoch": 49.0073417721519,
+      "grad_norm": 0.0015373198548331857,
+      "learning_rate": 7.032348804500703e-07,
+      "loss": 0.0601,
+      "step": 15600
+    },
+    {
+      "epoch": 49.007974683544305,
+      "grad_norm": 0.008119100704789162,
+      "learning_rate": 6.680731364275668e-07,
+      "loss": 0.0289,
+      "step": 15610
+    },
+    {
+      "epoch": 49.008607594936706,
+      "grad_norm": 2.2671926021575928,
+      "learning_rate": 6.329113924050633e-07,
+      "loss": 0.0466,
+      "step": 15620
+    },
+    {
+      "epoch": 49.009240506329114,
+      "grad_norm": 0.0015134834684431553,
+      "learning_rate": 5.977496483825598e-07,
+      "loss": 0.0606,
+      "step": 15630
+    },
+    {
+      "epoch": 49.00987341772152,
+      "grad_norm": 1.3951467275619507,
+      "learning_rate": 5.625879043600563e-07,
+      "loss": 0.0563,
+      "step": 15640
+    },
+    {
+      "epoch": 49.01050632911392,
+      "grad_norm": 0.0020348818507045507,
+      "learning_rate": 5.274261603375527e-07,
+      "loss": 0.0662,
+      "step": 15650
+    },
+    {
+      "epoch": 49.01113924050633,
+      "grad_norm": 2.6892151832580566,
+      "learning_rate": 4.922644163150493e-07,
+      "loss": 0.0594,
+      "step": 15660
+    },
+    {
+      "epoch": 49.01177215189873,
+      "grad_norm": 0.0018980724271386862,
+      "learning_rate": 4.5710267229254574e-07,
+      "loss": 0.0116,
+      "step": 15670
+    },
+    {
+      "epoch": 49.01240506329114,
+      "grad_norm": 0.0024212722200900316,
+      "learning_rate": 4.2194092827004216e-07,
+      "loss": 0.0489,
+      "step": 15680
+    },
+    {
+      "epoch": 49.01303797468354,
+      "grad_norm": 0.0014871886232867837,
+      "learning_rate": 3.8677918424753874e-07,
+      "loss": 0.0409,
+      "step": 15690
+    },
+    {
+      "epoch": 49.01367088607595,
+      "grad_norm": 0.0018424575682729483,
+      "learning_rate": 3.5161744022503516e-07,
+      "loss": 0.0436,
+      "step": 15700
+    },
+    {
+      "epoch": 49.01430379746836,
+      "grad_norm": 0.0016427412629127502,
+      "learning_rate": 3.1645569620253163e-07,
+      "loss": 0.0623,
+      "step": 15710
+    },
+    {
+      "epoch": 49.01493670886076,
+      "grad_norm": 0.00207283697091043,
+      "learning_rate": 2.8129395218002816e-07,
+      "loss": 0.099,
+      "step": 15720
+    },
+    {
+      "epoch": 49.01556962025317,
+      "grad_norm": 0.001088398857973516,
+      "learning_rate": 2.4613220815752463e-07,
+      "loss": 0.0557,
+      "step": 15730
+    },
+    {
+      "epoch": 49.01620253164557,
+      "grad_norm": 1.5488436222076416,
+      "learning_rate": 2.1097046413502108e-07,
+      "loss": 0.0514,
+      "step": 15740
+    },
+    {
+      "epoch": 49.01683544303798,
+      "grad_norm": 0.0019001478794962168,
+      "learning_rate": 1.7580872011251758e-07,
+      "loss": 0.0426,
+      "step": 15750
+    },
+    {
+      "epoch": 49.01746835443038,
+      "grad_norm": 0.0017688849475234747,
+      "learning_rate": 1.4064697609001408e-07,
+      "loss": 0.0131,
+      "step": 15760
+    },
+    {
+      "epoch": 49.018101265822786,
+      "grad_norm": 0.0013972530141472816,
+      "learning_rate": 1.0548523206751054e-07,
+      "loss": 0.0471,
+      "step": 15770
+    },
+    {
+      "epoch": 49.01873417721519,
+      "grad_norm": 0.0011928863823413849,
+      "learning_rate": 7.032348804500704e-08,
+      "loss": 0.0233,
+      "step": 15780
+    },
+    {
+      "epoch": 49.019367088607595,
+      "grad_norm": 0.0023855739273130894,
+      "learning_rate": 3.516174402250352e-08,
+      "loss": 0.0767,
+      "step": 15790
+    },
+    {
+      "epoch": 49.02,
+      "grad_norm": 0.0015424606390297413,
+      "learning_rate": 0.0,
+      "loss": 0.0463,
+      "step": 15800
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.9769673704414588,
+      "eval_loss": 0.04361777752637863,
+      "eval_runtime": 919.5227,
+      "eval_samples_per_second": 0.567,
+      "eval_steps_per_second": 0.072,
+      "step": 15800
+    },
+    {
+      "epoch": 49.02,
+      "step": 15800,
+      "total_flos": 1.5772028175840707e+20,
+      "train_loss": 0.3727006455846838,
+      "train_runtime": 274155.4834,
+      "train_samples_per_second": 0.461,
+      "train_steps_per_second": 0.058
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.8875,
+      "eval_loss": 0.4278368353843689,
+      "eval_runtime": 1123.5362,
+      "eval_samples_per_second": 0.57,
+      "eval_steps_per_second": 0.071,
+      "step": 15800
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.8875,
+      "eval_loss": 0.4278368353843689,
+      "eval_runtime": 1123.0215,
+      "eval_samples_per_second": 0.57,
+      "eval_steps_per_second": 0.071,
+      "step": 15800
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 15800,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "total_flos": 1.5772028175840707e+20,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}