End of training

README.md

@@ -32,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0631
+- Loss: 1.1564
 - Accuracy: 0.8391
 
 ## Model description

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 40.0,
-    "eval_accuracy": 0.8673913043478261,
-    "eval_loss": 0.6579979062080383,
-    "eval_runtime": 4.0597,
-    "eval_samples_per_second": 113.309,
-    "eval_steps_per_second": 3.695,
+    "eval_accuracy": 0.8391304347826087,
+    "eval_loss": 1.1564103364944458,
+    "eval_runtime": 4.5109,
+    "eval_samples_per_second": 101.976,
+    "eval_steps_per_second": 3.325,
     "total_flos": 4.585003492737024e+17,
-    "train_loss": 0.305610773563385,
-    "train_runtime": 583.5167,
-    "train_samples_per_second": 39.416,
-    "train_steps_per_second": 0.257
+    "train_loss": 2.1741886361440024,
+    "train_runtime": 1273.2626,
+    "train_samples_per_second": 18.064,
+    "train_steps_per_second": 0.118
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 40.0,
-    "eval_accuracy": 0.8673913043478261,
-    "eval_loss": 0.6579979062080383,
-    "eval_runtime": 4.0597,
-    "eval_samples_per_second": 113.309,
-    "eval_steps_per_second": 3.695
+    "eval_accuracy": 0.8391304347826087,
+    "eval_loss": 1.1564103364944458,
+    "eval_runtime": 4.5109,
+    "eval_samples_per_second": 101.976,
+    "eval_steps_per_second": 3.325
 }

runs/Apr04_09-11-23_fe56890e08d0/events.out.tfevents.1712223190.fe56890e08d0.1029.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9af311342097efaa8fdc346ed9a898d6e8099390e3983c5607b32d1afa9a6a5c
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 40.0,
     "total_flos": 4.585003492737024e+17,
-    "train_loss": 0.305610773563385,
-    "train_runtime": 583.5167,
-    "train_samples_per_second": 39.416,
-    "train_steps_per_second": 0.257
+    "train_loss": 2.1741886361440024,
+    "train_runtime": 1273.2626,
+    "train_samples_per_second": 18.064,
+    "train_steps_per_second": 0.118
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.8673913043478261,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-teeth_dataset/checkpoint-67",
+  "best_metric": 0.8391304347826087,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-teeth_dataset/checkpoint-120",
   "epoch": 40.0,
   "eval_steps": 500,
   "global_step": 150,
@@ -10,477 +10,477 @@
   "log_history": [
     {
       "epoch": 0.8,
-      "eval_accuracy": 0.7913043478260869,
-      "eval_loss": 1.1515443325042725,
-      "eval_runtime": 4.209,
-      "eval_samples_per_second": 109.29,
-      "eval_steps_per_second": 3.564,
+      "eval_accuracy": 0.015217391304347827,
+      "eval_loss": 4.579592704772949,
+      "eval_runtime": 349.3053,
+      "eval_samples_per_second": 1.317,
+      "eval_steps_per_second": 0.043,
       "step": 3
     },
     {
       "epoch": 1.87,
-      "eval_accuracy": 0.7630434782608696,
-      "eval_loss": 1.1545782089233398,
-      "eval_runtime": 4.4373,
-      "eval_samples_per_second": 103.665,
-      "eval_steps_per_second": 3.38,
+      "eval_accuracy": 0.02608695652173913,
+      "eval_loss": 4.519979000091553,
+      "eval_runtime": 4.4643,
+      "eval_samples_per_second": 103.04,
+      "eval_steps_per_second": 3.36,
       "step": 7
     },
     {
       "epoch": 2.67,
-      "grad_norm": 12.2432222366333,
+      "grad_norm": 4.626802921295166,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.6441,
+      "loss": 4.5616,
       "step": 10
     },
     {
       "epoch": 2.93,
-      "eval_accuracy": 0.8065217391304348,
-      "eval_loss": 1.070844292640686,
-      "eval_runtime": 4.2607,
-      "eval_samples_per_second": 107.964,
-      "eval_steps_per_second": 3.521,
+      "eval_accuracy": 0.03260869565217391,
+      "eval_loss": 4.470459461212158,
+      "eval_runtime": 4.8338,
+      "eval_samples_per_second": 95.164,
+      "eval_steps_per_second": 3.103,
       "step": 11
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8217391304347826,
-      "eval_loss": 0.992546796798706,
-      "eval_runtime": 4.1431,
-      "eval_samples_per_second": 111.027,
-      "eval_steps_per_second": 3.62,
+      "eval_accuracy": 0.06739130434782609,
+      "eval_loss": 4.412718296051025,
+      "eval_runtime": 5.0133,
+      "eval_samples_per_second": 91.756,
+      "eval_steps_per_second": 2.992,
       "step": 15
     },
     {
       "epoch": 4.8,
-      "eval_accuracy": 0.8108695652173913,
-      "eval_loss": 0.99262535572052,
-      "eval_runtime": 3.9996,
-      "eval_samples_per_second": 115.012,
-      "eval_steps_per_second": 3.75,
+      "eval_accuracy": 0.08043478260869565,
+      "eval_loss": 4.349262714385986,
+      "eval_runtime": 3.9868,
+      "eval_samples_per_second": 115.382,
+      "eval_steps_per_second": 3.762,
       "step": 18
     },
     {
       "epoch": 5.33,
-      "grad_norm": 12.744916915893555,
+      "grad_norm": 4.411898136138916,
       "learning_rate": 4.814814814814815e-05,
-      "loss": 0.6042,
+      "loss": 4.44,
       "step": 20
     },
     {
       "epoch": 5.87,
-      "eval_accuracy": 0.7934782608695652,
-      "eval_loss": 1.016530990600586,
-      "eval_runtime": 4.0517,
-      "eval_samples_per_second": 113.532,
-      "eval_steps_per_second": 3.702,
+      "eval_accuracy": 0.11304347826086956,
+      "eval_loss": 4.242533206939697,
+      "eval_runtime": 4.0721,
+      "eval_samples_per_second": 112.963,
+      "eval_steps_per_second": 3.684,
       "step": 22
     },
     {
       "epoch": 6.93,
-      "eval_accuracy": 0.8347826086956521,
-      "eval_loss": 0.8679854273796082,
-      "eval_runtime": 3.9943,
-      "eval_samples_per_second": 115.163,
-      "eval_steps_per_second": 3.755,
+      "eval_accuracy": 0.13695652173913042,
+      "eval_loss": 4.110694885253906,
+      "eval_runtime": 4.2677,
+      "eval_samples_per_second": 107.788,
+      "eval_steps_per_second": 3.515,
       "step": 26
     },
     {
       "epoch": 8.0,
-      "grad_norm": 23.645719528198242,
+      "grad_norm": 10.778473854064941,
       "learning_rate": 4.4444444444444447e-05,
-      "loss": 0.4293,
+      "loss": 4.1823,
       "step": 30
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8347826086956521,
-      "eval_loss": 0.8360960483551025,
-      "eval_runtime": 4.0619,
-      "eval_samples_per_second": 113.248,
-      "eval_steps_per_second": 3.693,
+      "eval_accuracy": 0.1608695652173913,
+      "eval_loss": 3.9340145587921143,
+      "eval_runtime": 4.4318,
+      "eval_samples_per_second": 103.796,
+      "eval_steps_per_second": 3.385,
       "step": 30
     },
     {
       "epoch": 8.8,
-      "eval_accuracy": 0.8565217391304348,
-      "eval_loss": 0.8075701594352722,
-      "eval_runtime": 4.0291,
-      "eval_samples_per_second": 114.17,
-      "eval_steps_per_second": 3.723,
+      "eval_accuracy": 0.1934782608695652,
+      "eval_loss": 3.7821249961853027,
+      "eval_runtime": 4.4828,
+      "eval_samples_per_second": 102.615,
+      "eval_steps_per_second": 3.346,
       "step": 33
     },
     {
       "epoch": 9.87,
-      "eval_accuracy": 0.8565217391304348,
-      "eval_loss": 0.7547765970230103,
-      "eval_runtime": 4.0783,
-      "eval_samples_per_second": 112.793,
-      "eval_steps_per_second": 3.678,
+      "eval_accuracy": 0.2782608695652174,
+      "eval_loss": 3.5313777923583984,
+      "eval_runtime": 4.1068,
+      "eval_samples_per_second": 112.01,
+      "eval_steps_per_second": 3.653,
       "step": 37
     },
     {
       "epoch": 10.67,
-      "grad_norm": 20.158720016479492,
+      "grad_norm": 14.638664245605469,
       "learning_rate": 4.074074074074074e-05,
-      "loss": 0.4471,
+      "loss": 3.6357,
       "step": 40
     },
     {
       "epoch": 10.93,
-      "eval_accuracy": 0.8369565217391305,
-      "eval_loss": 0.7901058793067932,
-      "eval_runtime": 4.2465,
-      "eval_samples_per_second": 108.325,
-      "eval_steps_per_second": 3.532,
+      "eval_accuracy": 0.30434782608695654,
+      "eval_loss": 3.285728693008423,
+      "eval_runtime": 4.0688,
+      "eval_samples_per_second": 113.055,
+      "eval_steps_per_second": 3.687,
       "step": 41
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8543478260869565,
-      "eval_loss": 0.7326875925064087,
-      "eval_runtime": 4.3562,
-      "eval_samples_per_second": 105.597,
-      "eval_steps_per_second": 3.443,
+      "eval_accuracy": 0.3695652173913043,
+      "eval_loss": 3.106356143951416,
+      "eval_runtime": 4.0947,
+      "eval_samples_per_second": 112.339,
+      "eval_steps_per_second": 3.663,
       "step": 45
     },
     {
       "epoch": 12.8,
-      "eval_accuracy": 0.8478260869565217,
-      "eval_loss": 0.748528003692627,
-      "eval_runtime": 4.402,
-      "eval_samples_per_second": 104.498,
-      "eval_steps_per_second": 3.408,
+      "eval_accuracy": 0.3826086956521739,
+      "eval_loss": 2.9712679386138916,
+      "eval_runtime": 4.065,
+      "eval_samples_per_second": 113.162,
+      "eval_steps_per_second": 3.69,
       "step": 48
     },
     {
       "epoch": 13.33,
-      "grad_norm": 11.568655967712402,
+      "grad_norm": 24.2481746673584,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.3436,
+      "loss": 3.0041,
       "step": 50
     },
     {
       "epoch": 13.87,
-      "eval_accuracy": 0.8347826086956521,
-      "eval_loss": 0.7844414710998535,
-      "eval_runtime": 4.2227,
-      "eval_samples_per_second": 108.934,
-      "eval_steps_per_second": 3.552,
+      "eval_accuracy": 0.48695652173913045,
+      "eval_loss": 2.7171707153320312,
+      "eval_runtime": 4.147,
+      "eval_samples_per_second": 110.923,
+      "eval_steps_per_second": 3.617,
       "step": 52
     },
     {
       "epoch": 14.93,
-      "eval_accuracy": 0.8630434782608696,
-      "eval_loss": 0.6644619107246399,
-      "eval_runtime": 4.1242,
-      "eval_samples_per_second": 111.537,
-      "eval_steps_per_second": 3.637,
+      "eval_accuracy": 0.5434782608695652,
+      "eval_loss": 2.511073589324951,
+      "eval_runtime": 4.329,
+      "eval_samples_per_second": 106.261,
+      "eval_steps_per_second": 3.465,
       "step": 56
     },
     {
       "epoch": 16.0,
-      "grad_norm": 14.64358139038086,
+      "grad_norm": 17.775148391723633,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.2813,
+      "loss": 2.4604,
       "step": 60
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.8391304347826087,
-      "eval_loss": 0.7370373010635376,
-      "eval_runtime": 4.0228,
-      "eval_samples_per_second": 114.348,
-      "eval_steps_per_second": 3.729,
+      "eval_accuracy": 0.5695652173913044,
+      "eval_loss": 2.356055736541748,
+      "eval_runtime": 4.5447,
+      "eval_samples_per_second": 101.216,
+      "eval_steps_per_second": 3.301,
       "step": 60
     },
     {
       "epoch": 16.8,
-      "eval_accuracy": 0.8347826086956521,
-      "eval_loss": 0.7598132491111755,
-      "eval_runtime": 3.9928,
-      "eval_samples_per_second": 115.208,
-      "eval_steps_per_second": 3.757,
+      "eval_accuracy": 0.5717391304347826,
+      "eval_loss": 2.2684459686279297,
+      "eval_runtime": 4.3589,
+      "eval_samples_per_second": 105.531,
+      "eval_steps_per_second": 3.441,
       "step": 63
     },
     {
       "epoch": 17.87,
-      "eval_accuracy": 0.8673913043478261,
-      "eval_loss": 0.6579979062080383,
-      "eval_runtime": 4.1194,
-      "eval_samples_per_second": 111.667,
-      "eval_steps_per_second": 3.641,
+      "eval_accuracy": 0.6347826086956522,
+      "eval_loss": 2.0961129665374756,
+      "eval_runtime": 4.0684,
+      "eval_samples_per_second": 113.066,
+      "eval_steps_per_second": 3.687,
       "step": 67
     },
     {
       "epoch": 18.67,
-      "grad_norm": 16.545446395874023,
+      "grad_norm": 28.70151710510254,
       "learning_rate": 2.962962962962963e-05,
-      "loss": 0.275,
+      "loss": 1.971,
       "step": 70
     },
     {
       "epoch": 18.93,
-      "eval_accuracy": 0.8456521739130435,
-      "eval_loss": 0.7026421427726746,
-      "eval_runtime": 4.039,
-      "eval_samples_per_second": 113.889,
-      "eval_steps_per_second": 3.714,
+      "eval_accuracy": 0.6782608695652174,
+      "eval_loss": 1.9555323123931885,
+      "eval_runtime": 4.1018,
+      "eval_samples_per_second": 112.147,
+      "eval_steps_per_second": 3.657,
       "step": 71
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.8369565217391305,
-      "eval_loss": 0.7055774927139282,
-      "eval_runtime": 4.0414,
-      "eval_samples_per_second": 113.823,
-      "eval_steps_per_second": 3.712,
+      "eval_accuracy": 0.6891304347826087,
+      "eval_loss": 1.8400135040283203,
+      "eval_runtime": 4.114,
+      "eval_samples_per_second": 111.814,
+      "eval_steps_per_second": 3.646,
       "step": 75
     },
     {
       "epoch": 20.8,
-      "eval_accuracy": 0.8543478260869565,
-      "eval_loss": 0.6517285108566284,
-      "eval_runtime": 4.1333,
-      "eval_samples_per_second": 111.292,
-      "eval_steps_per_second": 3.629,
+      "eval_accuracy": 0.7239130434782609,
+      "eval_loss": 1.78555166721344,
+      "eval_runtime": 4.1163,
+      "eval_samples_per_second": 111.752,
+      "eval_steps_per_second": 3.644,
       "step": 78
     },
     {
       "epoch": 21.33,
-      "grad_norm": 13.078596115112305,
+      "grad_norm": 25.09538459777832,
       "learning_rate": 2.5925925925925925e-05,
-      "loss": 0.2674,
+      "loss": 1.651,
       "step": 80
     },
     {
       "epoch": 21.87,
-      "eval_accuracy": 0.8608695652173913,
-      "eval_loss": 0.5930755734443665,
-      "eval_runtime": 4.2994,
-      "eval_samples_per_second": 106.991,
-      "eval_steps_per_second": 3.489,
+      "eval_accuracy": 0.7369565217391304,
+      "eval_loss": 1.6797059774398804,
+      "eval_runtime": 4.243,
+      "eval_samples_per_second": 108.413,
+      "eval_steps_per_second": 3.535,
       "step": 82
     },
     {
       "epoch": 22.93,
-      "eval_accuracy": 0.8608695652173913,
-      "eval_loss": 0.5985938310623169,
-      "eval_runtime": 4.3353,
-      "eval_samples_per_second": 106.105,
-      "eval_steps_per_second": 3.46,
+      "eval_accuracy": 0.7717391304347826,
+      "eval_loss": 1.600671410560608,
+      "eval_runtime": 4.4052,
+      "eval_samples_per_second": 104.422,
+      "eval_steps_per_second": 3.405,
       "step": 86
     },
     {
       "epoch": 24.0,
-      "grad_norm": 8.3045015335083,
+      "grad_norm": 30.617403030395508,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.2278,
+      "loss": 1.3665,
       "step": 90
     },
     {
       "epoch": 24.0,
-      "eval_accuracy": 0.8608695652173913,
-      "eval_loss": 0.5982087254524231,
-      "eval_runtime": 4.3231,
-      "eval_samples_per_second": 106.404,
-      "eval_steps_per_second": 3.47,
+      "eval_accuracy": 0.7739130434782608,
+      "eval_loss": 1.5255870819091797,
+      "eval_runtime": 4.5557,
+      "eval_samples_per_second": 100.972,
+      "eval_steps_per_second": 3.293,
       "step": 90
     },
     {
       "epoch": 24.8,
-      "eval_accuracy": 0.8543478260869565,
-      "eval_loss": 0.604324460029602,
-      "eval_runtime": 4.0948,
-      "eval_samples_per_second": 112.337,
-      "eval_steps_per_second": 3.663,
+      "eval_accuracy": 0.7652173913043478,
+      "eval_loss": 1.4875919818878174,
+      "eval_runtime": 4.9234,
+      "eval_samples_per_second": 93.432,
+      "eval_steps_per_second": 3.047,
       "step": 93
     },
     {
       "epoch": 25.87,
-      "eval_accuracy": 0.8521739130434782,
-      "eval_loss": 0.6148393750190735,
-      "eval_runtime": 4.0061,
-      "eval_samples_per_second": 114.824,
-      "eval_steps_per_second": 3.744,
+      "eval_accuracy": 0.7782608695652173,
+      "eval_loss": 1.4394937753677368,
+      "eval_runtime": 4.1294,
+      "eval_samples_per_second": 111.395,
+      "eval_steps_per_second": 3.632,
       "step": 97
     },
     {
       "epoch": 26.67,
-      "grad_norm": 15.322602272033691,
+      "grad_norm": 17.45021629333496,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.2034,
+      "loss": 1.1954,
       "step": 100
     },
     {
       "epoch": 26.93,
-      "eval_accuracy": 0.85,
-      "eval_loss": 0.6160370111465454,
-      "eval_runtime": 4.0736,
-      "eval_samples_per_second": 112.922,
-      "eval_steps_per_second": 3.682,
+      "eval_accuracy": 0.7869565217391304,
+      "eval_loss": 1.3679003715515137,
+      "eval_runtime": 4.1015,
+      "eval_samples_per_second": 112.155,
+      "eval_steps_per_second": 3.657,
       "step": 101
     },
     {
       "epoch": 28.0,
-      "eval_accuracy": 0.8521739130434782,
-      "eval_loss": 0.5903608798980713,
-      "eval_runtime": 4.1333,
-      "eval_samples_per_second": 111.291,
-      "eval_steps_per_second": 3.629,
+      "eval_accuracy": 0.8021739130434783,
+      "eval_loss": 1.304268479347229,
+      "eval_runtime": 4.1469,
+      "eval_samples_per_second": 110.928,
+      "eval_steps_per_second": 3.617,
       "step": 105
     },
     {
       "epoch": 28.8,
-      "eval_accuracy": 0.8608695652173913,
-      "eval_loss": 0.5824848413467407,
-      "eval_runtime": 4.2951,
-      "eval_samples_per_second": 107.1,
-      "eval_steps_per_second": 3.492,
+      "eval_accuracy": 0.8021739130434783,
+      "eval_loss": 1.2906242609024048,
+      "eval_runtime": 4.1542,
+      "eval_samples_per_second": 110.731,
+      "eval_steps_per_second": 3.611,
       "step": 108
     },
     {
       "epoch": 29.33,
-      "grad_norm": 12.949268341064453,
+      "grad_norm": 16.89618492126465,
       "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.2207,
+      "loss": 0.9886,
       "step": 110
     },
     {
       "epoch": 29.87,
-      "eval_accuracy": 0.8478260869565217,
-      "eval_loss": 0.6132237911224365,
-      "eval_runtime": 4.382,
-      "eval_samples_per_second": 104.975,
-      "eval_steps_per_second": 3.423,
+      "eval_accuracy": 0.8108695652173913,
+      "eval_loss": 1.2312711477279663,
+      "eval_runtime": 4.2901,
+      "eval_samples_per_second": 107.224,
+      "eval_steps_per_second": 3.496,
       "step": 112
     },
     {
       "epoch": 30.93,
-      "eval_accuracy": 0.8391304347826087,
-      "eval_loss": 0.6181351542472839,
-      "eval_runtime": 4.27,
-      "eval_samples_per_second": 107.728,
-      "eval_steps_per_second": 3.513,
+      "eval_accuracy": 0.8347826086956521,
+      "eval_loss": 1.182868480682373,
+      "eval_runtime": 4.4407,
+      "eval_samples_per_second": 103.588,
+      "eval_steps_per_second": 3.378,
       "step": 116
     },
     {
       "epoch": 32.0,
-      "grad_norm": 5.803860664367676,
+      "grad_norm": 14.743515968322754,
       "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.1701,
+      "loss": 0.8803,
       "step": 120
     },
     {
       "epoch": 32.0,
-      "eval_accuracy": 0.85,
-      "eval_loss": 0.5688785910606384,
-      "eval_runtime": 4.1338,
-      "eval_samples_per_second": 111.279,
-      "eval_steps_per_second": 3.629,
+      "eval_accuracy": 0.8391304347826087,
+      "eval_loss": 1.1564103364944458,
+      "eval_runtime": 4.4277,
+      "eval_samples_per_second": 103.892,
+      "eval_steps_per_second": 3.388,
       "step": 120
     },
     {
       "epoch": 32.8,
-      "eval_accuracy": 0.8608695652173913,
-      "eval_loss": 0.5633876919746399,
-      "eval_runtime": 4.7704,
-      "eval_samples_per_second": 96.428,
-      "eval_steps_per_second": 3.144,
+      "eval_accuracy": 0.8304347826086956,
+      "eval_loss": 1.1421394348144531,
+      "eval_runtime": 4.2096,
+      "eval_samples_per_second": 109.275,
+      "eval_steps_per_second": 3.563,
       "step": 123
     },
     {
       "epoch": 33.87,
-      "eval_accuracy": 0.8652173913043478,
-      "eval_loss": 0.5659446120262146,
-      "eval_runtime": 4.0029,
-      "eval_samples_per_second": 114.915,
-      "eval_steps_per_second": 3.747,
+      "eval_accuracy": 0.8326086956521739,
+      "eval_loss": 1.114410400390625,
+      "eval_runtime": 4.0812,
+      "eval_samples_per_second": 112.713,
+      "eval_steps_per_second": 3.675,
       "step": 127
     },
     {
       "epoch": 34.67,
-      "grad_norm": 18.19119644165039,
+      "grad_norm": 15.186381340026855,
       "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.1681,
+      "loss": 0.815,
       "step": 130
     },
     {
       "epoch": 34.93,
-      "eval_accuracy": 0.8565217391304348,
-      "eval_loss": 0.5895132422447205,
-      "eval_runtime": 4.0203,
-      "eval_samples_per_second": 114.42,
-      "eval_steps_per_second": 3.731,
+      "eval_accuracy": 0.8304347826086956,
+      "eval_loss": 1.107426404953003,
+      "eval_runtime": 4.1243,
+      "eval_samples_per_second": 111.534,
+      "eval_steps_per_second": 3.637,
       "step": 131
     },
     {
       "epoch": 36.0,
-      "eval_accuracy": 0.85,
-      "eval_loss": 0.6094810366630554,
-      "eval_runtime": 4.1594,
-      "eval_samples_per_second": 110.594,
-      "eval_steps_per_second": 3.606,
+      "eval_accuracy": 0.8282608695652174,
+      "eval_loss": 1.0919322967529297,
+      "eval_runtime": 4.1615,
+      "eval_samples_per_second": 110.537,
+      "eval_steps_per_second": 3.604,
       "step": 135
     },
     {
       "epoch": 36.8,
-      "eval_accuracy": 0.8456521739130435,
-      "eval_loss": 0.614500105381012,
-      "eval_runtime": 4.3427,
-      "eval_samples_per_second": 105.924,
-      "eval_steps_per_second": 3.454,
+      "eval_accuracy": 0.8326086956521739,
+      "eval_loss": 1.082062840461731,
+      "eval_runtime": 4.184,
+      "eval_samples_per_second": 109.942,
+      "eval_steps_per_second": 3.585,
       "step": 138
     },
     {
       "epoch": 37.33,
-      "grad_norm": 7.945148468017578,
+      "grad_norm": 12.77287483215332,
       "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.1435,
+      "loss": 0.7619,
       "step": 140
     },
     {
       "epoch": 37.87,
-      "eval_accuracy": 0.8478260869565217,
-      "eval_loss": 0.6111825704574585,
-      "eval_runtime": 4.3974,
-      "eval_samples_per_second": 104.606,
-      "eval_steps_per_second": 3.411,
+      "eval_accuracy": 0.8347826086956521,
+      "eval_loss": 1.070085883140564,
+      "eval_runtime": 4.3491,
+      "eval_samples_per_second": 105.769,
+      "eval_steps_per_second": 3.449,
       "step": 142
     },
     {
       "epoch": 38.93,
-      "eval_accuracy": 0.8456521739130435,
-      "eval_loss": 0.599669873714447,
-      "eval_runtime": 4.2726,
-      "eval_samples_per_second": 107.662,
-      "eval_steps_per_second": 3.511,
+      "eval_accuracy": 0.8347826086956521,
+      "eval_loss": 1.0642296075820923,
+      "eval_runtime": 4.4959,
+      "eval_samples_per_second": 102.315,
+      "eval_steps_per_second": 3.336,
       "step": 146
     },
     {
       "epoch": 40.0,
-      "grad_norm": 11.281535148620605,
+      "grad_norm": 20.172346115112305,
       "learning_rate": 0.0,
-      "loss": 0.1586,
+      "loss": 0.6991,
       "step": 150
     },
     {
       "epoch": 40.0,
-      "eval_accuracy": 0.85,
-      "eval_loss": 0.5965761542320251,
-      "eval_runtime": 4.0634,
-      "eval_samples_per_second": 113.205,
-      "eval_steps_per_second": 3.691,
+      "eval_accuracy": 0.8391304347826087,
+      "eval_loss": 1.063112497329712,
+      "eval_runtime": 4.3636,
+      "eval_samples_per_second": 105.417,
+      "eval_steps_per_second": 3.438,
       "step": 150
     },
     {
       "epoch": 40.0,
       "step": 150,
       "total_flos": 4.585003492737024e+17,
-      "train_loss": 0.305610773563385,
-      "train_runtime": 583.5167,
-      "train_samples_per_second": 39.416,
-      "train_steps_per_second": 0.257
+      "train_loss": 2.1741886361440024,
+      "train_runtime": 1273.2626,
+      "train_samples_per_second": 18.064,
+      "train_steps_per_second": 0.118
     }
   ],
   "logging_steps": 10,