End of training

README.md

@@ -22,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.85
+      value: 0.8673913043478261
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ 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: 0.5966
-- Accuracy: 0.85
+- Loss: 0.6580
+- Accuracy: 0.8674
 
 ## Model description
 

all_results.json

@@ -1,8 +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,
     "total_flos": 4.585003492737024e+17,
-    "train_loss": 2.1046753883361817,
-    "train_runtime": 578.2992,
-    "train_samples_per_second": 39.772,
-    "train_steps_per_second": 0.259
+    "train_loss": 0.305610773563385,
+    "train_runtime": 583.5167,
+    "train_samples_per_second": 39.416,
+    "train_steps_per_second": 0.257
 }

eval_results.json

@@ -0,0 +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
+}

runs/Apr04_05-20-25_d6892a113cb9/events.out.tfevents.1712209271.d6892a113cb9.1199.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e09a31eba0739a424a4d98a485229276ed5c340526f16822c28012afbba52f1
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 40.0,
     "total_flos": 4.585003492737024e+17,
-    "train_loss": 2.1046753883361817,
-    "train_runtime": 578.2992,
-    "train_samples_per_second": 39.772,
-    "train_steps_per_second": 0.259
+    "train_loss": 0.305610773563385,
+    "train_runtime": 583.5167,
+    "train_samples_per_second": 39.416,
+    "train_steps_per_second": 0.257
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.7978260869565217,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-teeth_dataset/checkpoint-146",
+  "best_metric": 0.8673913043478261,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-teeth_dataset/checkpoint-67",
   "epoch": 40.0,
   "eval_steps": 500,
   "global_step": 150,
@@ -10,477 +10,477 @@
   "log_history": [
     {
       "epoch": 0.8,
-      "eval_accuracy": 0.017391304347826087,
-      "eval_loss": 4.560667037963867,
-      "eval_runtime": 4.0314,
-      "eval_samples_per_second": 114.103,
-      "eval_steps_per_second": 3.721,
+      "eval_accuracy": 0.7913043478260869,
+      "eval_loss": 1.1515443325042725,
+      "eval_runtime": 4.209,
+      "eval_samples_per_second": 109.29,
+      "eval_steps_per_second": 3.564,
       "step": 3
     },
     {
       "epoch": 1.87,
-      "eval_accuracy": 0.021739130434782608,
-      "eval_loss": 4.518674850463867,
-      "eval_runtime": 4.0009,
-      "eval_samples_per_second": 114.973,
-      "eval_steps_per_second": 3.749,
+      "eval_accuracy": 0.7630434782608696,
+      "eval_loss": 1.1545782089233398,
+      "eval_runtime": 4.4373,
+      "eval_samples_per_second": 103.665,
+      "eval_steps_per_second": 3.38,
       "step": 7
     },
     {
       "epoch": 2.67,
-      "grad_norm": 3.8695149421691895,
+      "grad_norm": 12.2432222366333,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 4.5558,
+      "loss": 0.6441,
       "step": 10
     },
     {
       "epoch": 2.93,
-      "eval_accuracy": 0.02826086956521739,
-      "eval_loss": 4.462986469268799,
-      "eval_runtime": 3.9817,
-      "eval_samples_per_second": 115.528,
-      "eval_steps_per_second": 3.767,
+      "eval_accuracy": 0.8065217391304348,
+      "eval_loss": 1.070844292640686,
+      "eval_runtime": 4.2607,
+      "eval_samples_per_second": 107.964,
+      "eval_steps_per_second": 3.521,
       "step": 11
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.04565217391304348,
-      "eval_loss": 4.393328666687012,
-      "eval_runtime": 4.1763,
-      "eval_samples_per_second": 110.145,
-      "eval_steps_per_second": 3.592,
+      "eval_accuracy": 0.8217391304347826,
+      "eval_loss": 0.992546796798706,
+      "eval_runtime": 4.1431,
+      "eval_samples_per_second": 111.027,
+      "eval_steps_per_second": 3.62,
       "step": 15
     },
     {
       "epoch": 4.8,
-      "eval_accuracy": 0.04782608695652174,
-      "eval_loss": 4.3319196701049805,
-      "eval_runtime": 4.2196,
-      "eval_samples_per_second": 109.014,
-      "eval_steps_per_second": 3.555,
+      "eval_accuracy": 0.8108695652173913,
+      "eval_loss": 0.99262535572052,
+      "eval_runtime": 3.9996,
+      "eval_samples_per_second": 115.012,
+      "eval_steps_per_second": 3.75,
       "step": 18
     },
     {
       "epoch": 5.33,
-      "grad_norm": 6.7950005531311035,
+      "grad_norm": 12.744916915893555,
       "learning_rate": 4.814814814814815e-05,
-      "loss": 4.418,
+      "loss": 0.6042,
       "step": 20
     },
     {
       "epoch": 5.87,
-      "eval_accuracy": 0.07173913043478261,
-      "eval_loss": 4.238537311553955,
-      "eval_runtime": 4.3495,
-      "eval_samples_per_second": 105.759,
-      "eval_steps_per_second": 3.449,
+      "eval_accuracy": 0.7934782608695652,
+      "eval_loss": 1.016530990600586,
+      "eval_runtime": 4.0517,
+      "eval_samples_per_second": 113.532,
+      "eval_steps_per_second": 3.702,
       "step": 22
     },
     {
       "epoch": 6.93,
-      "eval_accuracy": 0.11956521739130435,
-      "eval_loss": 4.0838470458984375,
-      "eval_runtime": 4.4141,
-      "eval_samples_per_second": 104.212,
-      "eval_steps_per_second": 3.398,
+      "eval_accuracy": 0.8347826086956521,
+      "eval_loss": 0.8679854273796082,
+      "eval_runtime": 3.9943,
+      "eval_samples_per_second": 115.163,
+      "eval_steps_per_second": 3.755,
       "step": 26
     },
     {
       "epoch": 8.0,
-      "grad_norm": 16.177133560180664,
+      "grad_norm": 23.645719528198242,
       "learning_rate": 4.4444444444444447e-05,
-      "loss": 4.1135,
+      "loss": 0.4293,
       "step": 30
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.15217391304347827,
-      "eval_loss": 3.9196059703826904,
-      "eval_runtime": 4.1806,
-      "eval_samples_per_second": 110.032,
-      "eval_steps_per_second": 3.588,
+      "eval_accuracy": 0.8347826086956521,
+      "eval_loss": 0.8360960483551025,
+      "eval_runtime": 4.0619,
+      "eval_samples_per_second": 113.248,
+      "eval_steps_per_second": 3.693,
       "step": 30
     },
     {
       "epoch": 8.8,
-      "eval_accuracy": 0.20434782608695654,
-      "eval_loss": 3.774428129196167,
-      "eval_runtime": 3.9808,
-      "eval_samples_per_second": 115.556,
-      "eval_steps_per_second": 3.768,
+      "eval_accuracy": 0.8565217391304348,
+      "eval_loss": 0.8075701594352722,
+      "eval_runtime": 4.0291,
+      "eval_samples_per_second": 114.17,
+      "eval_steps_per_second": 3.723,
       "step": 33
     },
     {
       "epoch": 9.87,
-      "eval_accuracy": 0.2543478260869565,
-      "eval_loss": 3.5351600646972656,
-      "eval_runtime": 3.9835,
-      "eval_samples_per_second": 115.475,
-      "eval_steps_per_second": 3.766,
+      "eval_accuracy": 0.8565217391304348,
+      "eval_loss": 0.7547765970230103,
+      "eval_runtime": 4.0783,
+      "eval_samples_per_second": 112.793,
+      "eval_steps_per_second": 3.678,
       "step": 37
     },
     {
       "epoch": 10.67,
-      "grad_norm": 15.815583229064941,
+      "grad_norm": 20.158720016479492,
       "learning_rate": 4.074074074074074e-05,
-      "loss": 3.5597,
+      "loss": 0.4471,
       "step": 40
     },
     {
       "epoch": 10.93,
-      "eval_accuracy": 0.2891304347826087,
-      "eval_loss": 3.324174642562866,
-      "eval_runtime": 4.0012,
-      "eval_samples_per_second": 114.967,
-      "eval_steps_per_second": 3.749,
+      "eval_accuracy": 0.8369565217391305,
+      "eval_loss": 0.7901058793067932,
+      "eval_runtime": 4.2465,
+      "eval_samples_per_second": 108.325,
+      "eval_steps_per_second": 3.532,
       "step": 41
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.3565217391304348,
-      "eval_loss": 3.090582847595215,
-      "eval_runtime": 4.0067,
-      "eval_samples_per_second": 114.807,
-      "eval_steps_per_second": 3.744,
+      "eval_accuracy": 0.8543478260869565,
+      "eval_loss": 0.7326875925064087,
+      "eval_runtime": 4.3562,
+      "eval_samples_per_second": 105.597,
+      "eval_steps_per_second": 3.443,
       "step": 45
     },
     {
       "epoch": 12.8,
-      "eval_accuracy": 0.4,
-      "eval_loss": 2.915708541870117,
-      "eval_runtime": 4.0294,
-      "eval_samples_per_second": 114.16,
-      "eval_steps_per_second": 3.723,
+      "eval_accuracy": 0.8478260869565217,
+      "eval_loss": 0.748528003692627,
+      "eval_runtime": 4.402,
+      "eval_samples_per_second": 104.498,
+      "eval_steps_per_second": 3.408,
       "step": 48
     },
     {
       "epoch": 13.33,
-      "grad_norm": 17.922029495239258,
+      "grad_norm": 11.568655967712402,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 2.9063,
+      "loss": 0.3436,
       "step": 50
     },
     {
       "epoch": 13.87,
-      "eval_accuracy": 0.45869565217391306,
-      "eval_loss": 2.70074462890625,
-      "eval_runtime": 4.2235,
-      "eval_samples_per_second": 108.915,
+      "eval_accuracy": 0.8347826086956521,
+      "eval_loss": 0.7844414710998535,
+      "eval_runtime": 4.2227,
+      "eval_samples_per_second": 108.934,
       "eval_steps_per_second": 3.552,
       "step": 52
     },
     {
       "epoch": 14.93,
-      "eval_accuracy": 0.4956521739130435,
-      "eval_loss": 2.491408348083496,
-      "eval_runtime": 4.3767,
-      "eval_samples_per_second": 105.102,
-      "eval_steps_per_second": 3.427,
+      "eval_accuracy": 0.8630434782608696,
+      "eval_loss": 0.6644619107246399,
+      "eval_runtime": 4.1242,
+      "eval_samples_per_second": 111.537,
+      "eval_steps_per_second": 3.637,
       "step": 56
     },
     {
       "epoch": 16.0,
-      "grad_norm": 22.626970291137695,
+      "grad_norm": 14.64358139038086,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.3716,
+      "loss": 0.2813,
       "step": 60
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.5673913043478261,
-      "eval_loss": 2.3103415966033936,
-      "eval_runtime": 4.34,
-      "eval_samples_per_second": 105.991,
-      "eval_steps_per_second": 3.456,
+      "eval_accuracy": 0.8391304347826087,
+      "eval_loss": 0.7370373010635376,
+      "eval_runtime": 4.0228,
+      "eval_samples_per_second": 114.348,
+      "eval_steps_per_second": 3.729,
       "step": 60
     },
     {
       "epoch": 16.8,
-      "eval_accuracy": 0.5847826086956521,
-      "eval_loss": 2.233128309249878,
-      "eval_runtime": 4.137,
-      "eval_samples_per_second": 111.191,
-      "eval_steps_per_second": 3.626,
+      "eval_accuracy": 0.8347826086956521,
+      "eval_loss": 0.7598132491111755,
+      "eval_runtime": 3.9928,
+      "eval_samples_per_second": 115.208,
+      "eval_steps_per_second": 3.757,
       "step": 63
     },
     {
       "epoch": 17.87,
-      "eval_accuracy": 0.6347826086956522,
-      "eval_loss": 2.067967176437378,
-      "eval_runtime": 4.0035,
-      "eval_samples_per_second": 114.898,
-      "eval_steps_per_second": 3.747,
+      "eval_accuracy": 0.8673913043478261,
+      "eval_loss": 0.6579979062080383,
+      "eval_runtime": 4.1194,
+      "eval_samples_per_second": 111.667,
+      "eval_steps_per_second": 3.641,
       "step": 67
     },
     {
       "epoch": 18.67,
-      "grad_norm": 27.876663208007812,
+      "grad_norm": 16.545446395874023,
       "learning_rate": 2.962962962962963e-05,
-      "loss": 1.8836,
+      "loss": 0.275,
       "step": 70
     },
     {
       "epoch": 18.93,
-      "eval_accuracy": 0.6347826086956522,
-      "eval_loss": 1.9850654602050781,
-      "eval_runtime": 3.9694,
-      "eval_samples_per_second": 115.886,
-      "eval_steps_per_second": 3.779,
+      "eval_accuracy": 0.8456521739130435,
+      "eval_loss": 0.7026421427726746,
+      "eval_runtime": 4.039,
+      "eval_samples_per_second": 113.889,
+      "eval_steps_per_second": 3.714,
       "step": 71
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.6630434782608695,
-      "eval_loss": 1.8823285102844238,
-      "eval_runtime": 3.9914,
-      "eval_samples_per_second": 115.249,
-      "eval_steps_per_second": 3.758,
+      "eval_accuracy": 0.8369565217391305,
+      "eval_loss": 0.7055774927139282,
+      "eval_runtime": 4.0414,
+      "eval_samples_per_second": 113.823,
+      "eval_steps_per_second": 3.712,
       "step": 75
     },
     {
       "epoch": 20.8,
-      "eval_accuracy": 0.6978260869565217,
-      "eval_loss": 1.7874654531478882,
-      "eval_runtime": 3.9997,
-      "eval_samples_per_second": 115.008,
-      "eval_steps_per_second": 3.75,
+      "eval_accuracy": 0.8543478260869565,
+      "eval_loss": 0.6517285108566284,
+      "eval_runtime": 4.1333,
+      "eval_samples_per_second": 111.292,
+      "eval_steps_per_second": 3.629,
       "step": 78
     },
     {
       "epoch": 21.33,
-      "grad_norm": 15.183324813842773,
+      "grad_norm": 13.078596115112305,
       "learning_rate": 2.5925925925925925e-05,
-      "loss": 1.5373,
+      "loss": 0.2674,
       "step": 80
     },
     {
       "epoch": 21.87,
-      "eval_accuracy": 0.7043478260869566,
-      "eval_loss": 1.7380263805389404,
-      "eval_runtime": 4.08,
-      "eval_samples_per_second": 112.744,
-      "eval_steps_per_second": 3.676,
+      "eval_accuracy": 0.8608695652173913,
+      "eval_loss": 0.5930755734443665,
+      "eval_runtime": 4.2994,
+      "eval_samples_per_second": 106.991,
+      "eval_steps_per_second": 3.489,
       "step": 82
     },
     {
       "epoch": 22.93,
-      "eval_accuracy": 0.717391304347826,
-      "eval_loss": 1.6596403121948242,
-      "eval_runtime": 4.1994,
-      "eval_samples_per_second": 109.54,
-      "eval_steps_per_second": 3.572,
+      "eval_accuracy": 0.8608695652173913,
+      "eval_loss": 0.5985938310623169,
+      "eval_runtime": 4.3353,
+      "eval_samples_per_second": 106.105,
+      "eval_steps_per_second": 3.46,
       "step": 86
     },
     {
       "epoch": 24.0,
-      "grad_norm": 29.264057159423828,
+      "grad_norm": 8.3045015335083,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 1.2639,
+      "loss": 0.2278,
       "step": 90
     },
     {
       "epoch": 24.0,
-      "eval_accuracy": 0.717391304347826,
-      "eval_loss": 1.53962242603302,
-      "eval_runtime": 4.3346,
-      "eval_samples_per_second": 106.122,
-      "eval_steps_per_second": 3.46,
+      "eval_accuracy": 0.8608695652173913,
+      "eval_loss": 0.5982087254524231,
+      "eval_runtime": 4.3231,
+      "eval_samples_per_second": 106.404,
+      "eval_steps_per_second": 3.47,
       "step": 90
     },
     {
       "epoch": 24.8,
-      "eval_accuracy": 0.7152173913043478,
-      "eval_loss": 1.5275580883026123,
-      "eval_runtime": 4.361,
-      "eval_samples_per_second": 105.48,
-      "eval_steps_per_second": 3.44,
+      "eval_accuracy": 0.8543478260869565,
+      "eval_loss": 0.604324460029602,
+      "eval_runtime": 4.0948,
+      "eval_samples_per_second": 112.337,
+      "eval_steps_per_second": 3.663,
       "step": 93
     },
     {
       "epoch": 25.87,
-      "eval_accuracy": 0.7434782608695653,
-      "eval_loss": 1.4459482431411743,
-      "eval_runtime": 4.2974,
-      "eval_samples_per_second": 107.043,
-      "eval_steps_per_second": 3.491,
+      "eval_accuracy": 0.8521739130434782,
+      "eval_loss": 0.6148393750190735,
+      "eval_runtime": 4.0061,
+      "eval_samples_per_second": 114.824,
+      "eval_steps_per_second": 3.744,
       "step": 97
     },
     {
       "epoch": 26.67,
-      "grad_norm": 17.330795288085938,
+      "grad_norm": 15.322602272033691,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 1.1322,
+      "loss": 0.2034,
       "step": 100
     },
     {
       "epoch": 26.93,
-      "eval_accuracy": 0.7478260869565218,
-      "eval_loss": 1.4325517416000366,
-      "eval_runtime": 4.0947,
-      "eval_samples_per_second": 112.341,
-      "eval_steps_per_second": 3.663,
+      "eval_accuracy": 0.85,
+      "eval_loss": 0.6160370111465454,
+      "eval_runtime": 4.0736,
+      "eval_samples_per_second": 112.922,
+      "eval_steps_per_second": 3.682,
       "step": 101
     },
     {
       "epoch": 28.0,
-      "eval_accuracy": 0.7521739130434782,
-      "eval_loss": 1.4265176057815552,
-      "eval_runtime": 4.0215,
-      "eval_samples_per_second": 114.385,
-      "eval_steps_per_second": 3.73,
+      "eval_accuracy": 0.8521739130434782,
+      "eval_loss": 0.5903608798980713,
+      "eval_runtime": 4.1333,
+      "eval_samples_per_second": 111.291,
+      "eval_steps_per_second": 3.629,
       "step": 105
     },
     {
       "epoch": 28.8,
-      "eval_accuracy": 0.7652173913043478,
-      "eval_loss": 1.3482317924499512,
-      "eval_runtime": 4.0579,
-      "eval_samples_per_second": 113.358,
-      "eval_steps_per_second": 3.696,
+      "eval_accuracy": 0.8608695652173913,
+      "eval_loss": 0.5824848413467407,
+      "eval_runtime": 4.2951,
+      "eval_samples_per_second": 107.1,
+      "eval_steps_per_second": 3.492,
       "step": 108
     },
     {
       "epoch": 29.33,
-      "grad_norm": 11.148159980773926,
+      "grad_norm": 12.949268341064453,
       "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.9104,
+      "loss": 0.2207,
       "step": 110
     },
     {
       "epoch": 29.87,
-      "eval_accuracy": 0.7717391304347826,
-      "eval_loss": 1.281937837600708,
-      "eval_runtime": 4.0153,
-      "eval_samples_per_second": 114.561,
-      "eval_steps_per_second": 3.736,
+      "eval_accuracy": 0.8478260869565217,
+      "eval_loss": 0.6132237911224365,
+      "eval_runtime": 4.382,
+      "eval_samples_per_second": 104.975,
+      "eval_steps_per_second": 3.423,
       "step": 112
     },
     {
       "epoch": 30.93,
-      "eval_accuracy": 0.7717391304347826,
-      "eval_loss": 1.272769808769226,
-      "eval_runtime": 4.0262,
-      "eval_samples_per_second": 114.25,
-      "eval_steps_per_second": 3.726,
+      "eval_accuracy": 0.8391304347826087,
+      "eval_loss": 0.6181351542472839,
+      "eval_runtime": 4.27,
+      "eval_samples_per_second": 107.728,
+      "eval_steps_per_second": 3.513,
       "step": 116
     },
     {
       "epoch": 32.0,
-      "grad_norm": 14.274454116821289,
+      "grad_norm": 5.803860664367676,
       "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.8025,
+      "loss": 0.1701,
       "step": 120
     },
     {
       "epoch": 32.0,
-      "eval_accuracy": 0.7673913043478261,
-      "eval_loss": 1.2269138097763062,
-      "eval_runtime": 4.0744,
-      "eval_samples_per_second": 112.901,
-      "eval_steps_per_second": 3.682,
+      "eval_accuracy": 0.85,
+      "eval_loss": 0.5688785910606384,
+      "eval_runtime": 4.1338,
+      "eval_samples_per_second": 111.279,
+      "eval_steps_per_second": 3.629,
       "step": 120
     },
     {
       "epoch": 32.8,
-      "eval_accuracy": 0.7847826086956522,
-      "eval_loss": 1.1989597082138062,
-      "eval_runtime": 4.1887,
-      "eval_samples_per_second": 109.82,
-      "eval_steps_per_second": 3.581,
+      "eval_accuracy": 0.8608695652173913,
+      "eval_loss": 0.5633876919746399,
+      "eval_runtime": 4.7704,
+      "eval_samples_per_second": 96.428,
+      "eval_steps_per_second": 3.144,
       "step": 123
     },
     {
       "epoch": 33.87,
-      "eval_accuracy": 0.782608695652174,
-      "eval_loss": 1.186194658279419,
-      "eval_runtime": 4.3153,
-      "eval_samples_per_second": 106.598,
-      "eval_steps_per_second": 3.476,
+      "eval_accuracy": 0.8652173913043478,
+      "eval_loss": 0.5659446120262146,
+      "eval_runtime": 4.0029,
+      "eval_samples_per_second": 114.915,
+      "eval_steps_per_second": 3.747,
       "step": 127
     },
     {
       "epoch": 34.67,
-      "grad_norm": 17.313024520874023,
+      "grad_norm": 18.19119644165039,
       "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.7499,
+      "loss": 0.1681,
       "step": 130
     },
     {
       "epoch": 34.93,
-      "eval_accuracy": 0.7869565217391304,
-      "eval_loss": 1.1800005435943604,
-      "eval_runtime": 4.3789,
-      "eval_samples_per_second": 105.049,
-      "eval_steps_per_second": 3.426,
+      "eval_accuracy": 0.8565217391304348,
+      "eval_loss": 0.5895132422447205,
+      "eval_runtime": 4.0203,
+      "eval_samples_per_second": 114.42,
+      "eval_steps_per_second": 3.731,
       "step": 131
     },
     {
       "epoch": 36.0,
-      "eval_accuracy": 0.7913043478260869,
-      "eval_loss": 1.1722314357757568,
-      "eval_runtime": 4.2382,
-      "eval_samples_per_second": 108.537,
-      "eval_steps_per_second": 3.539,
+      "eval_accuracy": 0.85,
+      "eval_loss": 0.6094810366630554,
+      "eval_runtime": 4.1594,
+      "eval_samples_per_second": 110.594,
+      "eval_steps_per_second": 3.606,
       "step": 135
     },
     {
       "epoch": 36.8,
-      "eval_accuracy": 0.7891304347826087,
-      "eval_loss": 1.1541346311569214,
-      "eval_runtime": 4.0299,
-      "eval_samples_per_second": 114.148,
-      "eval_steps_per_second": 3.722,
+      "eval_accuracy": 0.8456521739130435,
+      "eval_loss": 0.614500105381012,
+      "eval_runtime": 4.3427,
+      "eval_samples_per_second": 105.924,
+      "eval_steps_per_second": 3.454,
       "step": 138
     },
     {
       "epoch": 37.33,
-      "grad_norm": 13.900386810302734,
+      "grad_norm": 7.945148468017578,
       "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.7032,
+      "loss": 0.1435,
       "step": 140
     },
     {
       "epoch": 37.87,
-      "eval_accuracy": 0.7934782608695652,
-      "eval_loss": 1.1435534954071045,
-      "eval_runtime": 3.9993,
-      "eval_samples_per_second": 115.021,
-      "eval_steps_per_second": 3.751,
+      "eval_accuracy": 0.8478260869565217,
+      "eval_loss": 0.6111825704574585,
+      "eval_runtime": 4.3974,
+      "eval_samples_per_second": 104.606,
+      "eval_steps_per_second": 3.411,
       "step": 142
     },
     {
       "epoch": 38.93,
-      "eval_accuracy": 0.7978260869565217,
-      "eval_loss": 1.1396770477294922,
-      "eval_runtime": 3.9973,
-      "eval_samples_per_second": 115.078,
-      "eval_steps_per_second": 3.753,
+      "eval_accuracy": 0.8456521739130435,
+      "eval_loss": 0.599669873714447,
+      "eval_runtime": 4.2726,
+      "eval_samples_per_second": 107.662,
+      "eval_steps_per_second": 3.511,
       "step": 146
     },
     {
       "epoch": 40.0,
-      "grad_norm": 18.35047149658203,
+      "grad_norm": 11.281535148620605,
       "learning_rate": 0.0,
-      "loss": 0.6621,
+      "loss": 0.1586,
       "step": 150
     },
     {
       "epoch": 40.0,
-      "eval_accuracy": 0.7978260869565217,
-      "eval_loss": 1.1415306329727173,
-      "eval_runtime": 4.0474,
-      "eval_samples_per_second": 113.654,
-      "eval_steps_per_second": 3.706,
+      "eval_accuracy": 0.85,
+      "eval_loss": 0.5965761542320251,
+      "eval_runtime": 4.0634,
+      "eval_samples_per_second": 113.205,
+      "eval_steps_per_second": 3.691,
       "step": 150
     },
     {
       "epoch": 40.0,
       "step": 150,
       "total_flos": 4.585003492737024e+17,
-      "train_loss": 2.1046753883361817,
-      "train_runtime": 578.2992,
-      "train_samples_per_second": 39.772,
-      "train_steps_per_second": 0.259
+      "train_loss": 0.305610773563385,
+      "train_runtime": 583.5167,
+      "train_samples_per_second": 39.416,
+      "train_steps_per_second": 0.257
     }
   ],
   "logging_steps": 10,