End of training

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swinv2-tiny-patch4-window8-256](https://huggingface.co/microsoft/swinv2-tiny-patch4-window8-256) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8406
-- Accuracy: 0.6
+- Loss: 0.8386
+- Accuracy: 0.65
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 37.94,
-    "eval_accuracy": 0.6333333333333333,
-    "eval_loss": 1.0633025169372559,
-    "eval_runtime": 1.2452,
-    "eval_samples_per_second": 48.183,
-    "eval_steps_per_second": 1.606,
-    "train_loss": 1.8558632776039798,
-    "train_runtime": 524.2652,
-    "train_samples_per_second": 78.029,
-    "train_steps_per_second": 0.561
+    "eval_accuracy": 0.65,
+    "eval_loss": 0.8386151194572449,
+    "eval_runtime": 1.0431,
+    "eval_samples_per_second": 57.519,
+    "eval_steps_per_second": 1.917,
+    "train_loss": 1.4638761638784084,
+    "train_runtime": 475.2898,
+    "train_samples_per_second": 86.07,
+    "train_steps_per_second": 0.619
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 37.94,
-    "eval_accuracy": 0.6333333333333333,
-    "eval_loss": 1.0633025169372559,
-    "eval_runtime": 1.2452,
-    "eval_samples_per_second": 48.183,
-    "eval_steps_per_second": 1.606
+    "eval_accuracy": 0.65,
+    "eval_loss": 0.8386151194572449,
+    "eval_runtime": 1.0431,
+    "eval_samples_per_second": 57.519,
+    "eval_steps_per_second": 1.917
 }

runs/Apr19_07-22-40_DESKTOP-SKBE9FB/events.out.tfevents.1713533467.DESKTOP-SKBE9FB.604.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9abb5481ec5302aeeefad24dfe4bac914b2a5a15e4a2d9b332729db181a9738
+size 411

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 37.94,
-    "train_loss": 1.8558632776039798,
-    "train_runtime": 524.2652,
-    "train_samples_per_second": 78.029,
-    "train_steps_per_second": 0.561
+    "train_loss": 1.4638761638784084,
+    "train_runtime": 475.2898,
+    "train_samples_per_second": 86.07,
+    "train_steps_per_second": 0.619
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.6333333333333333,
-  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-va-U5-42B\\checkpoint-193",
+  "best_metric": 0.65,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-va-U5-42B\\checkpoint-209",
   "epoch": 37.935483870967744,
   "eval_steps": 500,
   "global_step": 294,
@@ -11,497 +11,497 @@
     {
       "epoch": 0.9,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 7.866283893585205,
-      "eval_runtime": 1.0387,
-      "eval_samples_per_second": 57.764,
-      "eval_steps_per_second": 1.925,
+      "eval_loss": 6.174824237823486,
+      "eval_runtime": 0.845,
+      "eval_samples_per_second": 71.01,
+      "eval_steps_per_second": 2.367,
       "step": 7
     },
     {
       "epoch": 1.55,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 6.936,
+      "loss": 5.327,
       "step": 12
     },
     {
       "epoch": 1.94,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 7.757172584533691,
-      "eval_runtime": 0.9676,
-      "eval_samples_per_second": 62.01,
-      "eval_steps_per_second": 2.067,
+      "eval_loss": 6.066026210784912,
+      "eval_runtime": 0.8511,
+      "eval_samples_per_second": 70.499,
+      "eval_steps_per_second": 2.35,
       "step": 15
     },
     {
       "epoch": 2.97,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 7.17894983291626,
-      "eval_runtime": 1.0595,
-      "eval_samples_per_second": 56.633,
-      "eval_steps_per_second": 1.888,
+      "eval_loss": 5.4902424812316895,
+      "eval_runtime": 0.8648,
+      "eval_samples_per_second": 69.381,
+      "eval_steps_per_second": 2.313,
       "step": 23
     },
     {
       "epoch": 3.1,
       "learning_rate": 3.2000000000000005e-05,
-      "loss": 6.7016,
+      "loss": 5.0963,
       "step": 24
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 5.914210796356201,
-      "eval_runtime": 1.1717,
-      "eval_samples_per_second": 51.21,
-      "eval_steps_per_second": 1.707,
+      "eval_loss": 4.276790142059326,
+      "eval_runtime": 0.8095,
+      "eval_samples_per_second": 74.116,
+      "eval_steps_per_second": 2.471,
       "step": 31
     },
     {
       "epoch": 4.65,
       "learning_rate": 3.909090909090909e-05,
-      "loss": 5.5418,
+      "loss": 3.9193,
       "step": 36
     },
     {
       "epoch": 4.9,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 4.606503486633301,
-      "eval_runtime": 0.9707,
-      "eval_samples_per_second": 61.813,
-      "eval_steps_per_second": 2.06,
+      "eval_loss": 3.0012993812561035,
+      "eval_runtime": 0.8424,
+      "eval_samples_per_second": 71.228,
+      "eval_steps_per_second": 2.374,
       "step": 38
     },
     {
       "epoch": 5.94,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 3.2782602310180664,
-      "eval_runtime": 1.0414,
-      "eval_samples_per_second": 57.616,
-      "eval_steps_per_second": 1.921,
+      "eval_loss": 1.9289205074310303,
+      "eval_runtime": 0.8816,
+      "eval_samples_per_second": 68.055,
+      "eval_steps_per_second": 2.269,
       "step": 46
     },
     {
       "epoch": 6.19,
       "learning_rate": 3.7272727272727276e-05,
-      "loss": 3.6439,
+      "loss": 2.2222,
       "step": 48
     },
     {
       "epoch": 6.97,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 2.1983864307403564,
-      "eval_runtime": 1.0791,
-      "eval_samples_per_second": 55.604,
-      "eval_steps_per_second": 1.853,
+      "eval_loss": 1.385749101638794,
+      "eval_runtime": 0.8792,
+      "eval_samples_per_second": 68.244,
+      "eval_steps_per_second": 2.275,
       "step": 54
     },
     {
       "epoch": 7.74,
       "learning_rate": 3.545454545454546e-05,
-      "loss": 2.2477,
+      "loss": 1.4465,
       "step": 60
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.45,
-      "eval_loss": 1.5630935430526733,
-      "eval_runtime": 1.2494,
-      "eval_samples_per_second": 48.021,
-      "eval_steps_per_second": 1.601,
+      "eval_accuracy": 0.43333333333333335,
+      "eval_loss": 1.342329740524292,
+      "eval_runtime": 0.8314,
+      "eval_samples_per_second": 72.165,
+      "eval_steps_per_second": 2.405,
       "step": 62
     },
     {
       "epoch": 8.9,
       "eval_accuracy": 0.45,
-      "eval_loss": 1.315753698348999,
-      "eval_runtime": 1.0403,
-      "eval_samples_per_second": 57.675,
-      "eval_steps_per_second": 1.923,
+      "eval_loss": 1.2786400318145752,
+      "eval_runtime": 0.8842,
+      "eval_samples_per_second": 67.857,
+      "eval_steps_per_second": 2.262,
       "step": 69
     },
     {
       "epoch": 9.29,
       "learning_rate": 3.363636363636364e-05,
-      "loss": 1.5076,
+      "loss": 1.3709,
       "step": 72
     },
     {
       "epoch": 9.94,
-      "eval_accuracy": 0.3333333333333333,
-      "eval_loss": 1.3664613962173462,
-      "eval_runtime": 1.076,
-      "eval_samples_per_second": 55.76,
-      "eval_steps_per_second": 1.859,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 1.2653944492340088,
+      "eval_runtime": 0.8365,
+      "eval_samples_per_second": 71.724,
+      "eval_steps_per_second": 2.391,
       "step": 77
     },
     {
       "epoch": 10.84,
       "learning_rate": 3.181818181818182e-05,
-      "loss": 1.3865,
+      "loss": 1.3511,
       "step": 84
     },
     {
       "epoch": 10.97,
-      "eval_accuracy": 0.35,
-      "eval_loss": 1.3560707569122314,
-      "eval_runtime": 1.2112,
-      "eval_samples_per_second": 49.538,
-      "eval_steps_per_second": 1.651,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 1.2605175971984863,
+      "eval_runtime": 0.8152,
+      "eval_samples_per_second": 73.599,
+      "eval_steps_per_second": 2.453,
       "step": 85
     },
     {
       "epoch": 12.0,
       "eval_accuracy": 0.4666666666666667,
-      "eval_loss": 1.2673075199127197,
-      "eval_runtime": 0.9395,
-      "eval_samples_per_second": 63.867,
-      "eval_steps_per_second": 2.129,
+      "eval_loss": 1.2183691263198853,
+      "eval_runtime": 0.8329,
+      "eval_samples_per_second": 72.04,
+      "eval_steps_per_second": 2.401,
       "step": 93
     },
     {
       "epoch": 12.39,
       "learning_rate": 3.0000000000000004e-05,
-      "loss": 1.3436,
+      "loss": 1.2749,
       "step": 96
     },
     {
       "epoch": 12.9,
-      "eval_accuracy": 0.2833333333333333,
-      "eval_loss": 1.5020203590393066,
-      "eval_runtime": 1.0828,
-      "eval_samples_per_second": 55.411,
-      "eval_steps_per_second": 1.847,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.289415955543518,
+      "eval_runtime": 0.8348,
+      "eval_samples_per_second": 71.87,
+      "eval_steps_per_second": 2.396,
       "step": 100
     },
     {
       "epoch": 13.94,
       "learning_rate": 2.8181818181818185e-05,
-      "loss": 1.3187,
+      "loss": 1.222,
       "step": 108
     },
     {
       "epoch": 13.94,
-      "eval_accuracy": 0.4666666666666667,
-      "eval_loss": 1.3018145561218262,
-      "eval_runtime": 0.9637,
-      "eval_samples_per_second": 62.261,
-      "eval_steps_per_second": 2.075,
+      "eval_accuracy": 0.5166666666666667,
+      "eval_loss": 1.2071518898010254,
+      "eval_runtime": 0.8805,
+      "eval_samples_per_second": 68.142,
+      "eval_steps_per_second": 2.271,
       "step": 108
     },
     {
       "epoch": 14.97,
-      "eval_accuracy": 0.48333333333333334,
-      "eval_loss": 1.2581952810287476,
-      "eval_runtime": 0.968,
-      "eval_samples_per_second": 61.985,
-      "eval_steps_per_second": 2.066,
+      "eval_accuracy": 0.5166666666666667,
+      "eval_loss": 1.1748946905136108,
+      "eval_runtime": 0.8381,
+      "eval_samples_per_second": 71.592,
+      "eval_steps_per_second": 2.386,
       "step": 116
     },
     {
       "epoch": 15.48,
       "learning_rate": 2.6363636363636365e-05,
-      "loss": 1.2132,
+      "loss": 1.1668,
       "step": 120
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.5,
-      "eval_loss": 1.2250442504882812,
-      "eval_runtime": 1.0265,
-      "eval_samples_per_second": 58.453,
-      "eval_steps_per_second": 1.948,
+      "eval_accuracy": 0.5166666666666667,
+      "eval_loss": 1.198820948600769,
+      "eval_runtime": 0.8663,
+      "eval_samples_per_second": 69.26,
+      "eval_steps_per_second": 2.309,
       "step": 124
     },
     {
       "epoch": 16.9,
       "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 1.2070510387420654,
-      "eval_runtime": 0.8885,
-      "eval_samples_per_second": 67.53,
-      "eval_steps_per_second": 2.251,
+      "eval_loss": 1.2305818796157837,
+      "eval_runtime": 0.8839,
+      "eval_samples_per_second": 67.884,
+      "eval_steps_per_second": 2.263,
       "step": 131
     },
     {
       "epoch": 17.03,
       "learning_rate": 2.454545454545455e-05,
-      "loss": 1.2041,
+      "loss": 1.101,
       "step": 132
     },
     {
       "epoch": 17.94,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 1.1806073188781738,
-      "eval_runtime": 1.0077,
-      "eval_samples_per_second": 59.543,
-      "eval_steps_per_second": 1.985,
+      "eval_accuracy": 0.5333333333333333,
+      "eval_loss": 1.1431663036346436,
+      "eval_runtime": 0.8778,
+      "eval_samples_per_second": 68.351,
+      "eval_steps_per_second": 2.278,
       "step": 139
     },
     {
       "epoch": 18.58,
       "learning_rate": 2.2727272727272733e-05,
-      "loss": 1.1756,
+      "loss": 1.029,
       "step": 144
     },
     {
       "epoch": 18.97,
       "eval_accuracy": 0.55,
-      "eval_loss": 1.1720372438430786,
-      "eval_runtime": 1.0797,
-      "eval_samples_per_second": 55.573,
-      "eval_steps_per_second": 1.852,
+      "eval_loss": 1.0208261013031006,
+      "eval_runtime": 0.8723,
+      "eval_samples_per_second": 68.786,
+      "eval_steps_per_second": 2.293,
       "step": 147
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.5333333333333333,
-      "eval_loss": 1.1318727731704712,
-      "eval_runtime": 0.8755,
-      "eval_samples_per_second": 68.529,
-      "eval_steps_per_second": 2.284,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 0.9576696157455444,
+      "eval_runtime": 0.849,
+      "eval_samples_per_second": 70.671,
+      "eval_steps_per_second": 2.356,
       "step": 155
     },
     {
       "epoch": 20.13,
       "learning_rate": 2.090909090909091e-05,
-      "loss": 1.1107,
+      "loss": 0.9403,
       "step": 156
     },
     {
       "epoch": 20.9,
-      "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 1.0850528478622437,
-      "eval_runtime": 0.9417,
-      "eval_samples_per_second": 63.716,
-      "eval_steps_per_second": 2.124,
+      "eval_accuracy": 0.5,
+      "eval_loss": 0.9478870034217834,
+      "eval_runtime": 0.8337,
+      "eval_samples_per_second": 71.971,
+      "eval_steps_per_second": 2.399,
       "step": 162
     },
     {
       "epoch": 21.68,
       "learning_rate": 1.9090909090909094e-05,
-      "loss": 1.0651,
+      "loss": 0.8887,
       "step": 168
     },
     {
       "epoch": 21.94,
       "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 1.0553667545318604,
-      "eval_runtime": 0.9498,
-      "eval_samples_per_second": 63.168,
-      "eval_steps_per_second": 2.106,
+      "eval_loss": 0.8909610509872437,
+      "eval_runtime": 0.8389,
+      "eval_samples_per_second": 71.521,
+      "eval_steps_per_second": 2.384,
       "step": 170
     },
     {
       "epoch": 22.97,
-      "eval_accuracy": 0.5,
-      "eval_loss": 1.0579936504364014,
-      "eval_runtime": 0.9211,
-      "eval_samples_per_second": 65.141,
-      "eval_steps_per_second": 2.171,
+      "eval_accuracy": 0.5333333333333333,
+      "eval_loss": 0.9441585540771484,
+      "eval_runtime": 0.8343,
+      "eval_samples_per_second": 71.914,
+      "eval_steps_per_second": 2.397,
       "step": 178
     },
     {
       "epoch": 23.23,
       "learning_rate": 1.7272727272727274e-05,
-      "loss": 1.0419,
+      "loss": 0.8506,
       "step": 180
     },
     {
       "epoch": 24.0,
-      "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 1.1471267938613892,
-      "eval_runtime": 0.9539,
-      "eval_samples_per_second": 62.901,
-      "eval_steps_per_second": 2.097,
+      "eval_accuracy": 0.6,
+      "eval_loss": 0.8923367261886597,
+      "eval_runtime": 0.8341,
+      "eval_samples_per_second": 71.937,
+      "eval_steps_per_second": 2.398,
       "step": 186
     },
     {
       "epoch": 24.77,
       "learning_rate": 1.5454545454545454e-05,
-      "loss": 0.9804,
+      "loss": 0.8064,
       "step": 192
     },
     {
       "epoch": 24.9,
-      "eval_accuracy": 0.6333333333333333,
-      "eval_loss": 1.0633025169372559,
-      "eval_runtime": 0.9613,
-      "eval_samples_per_second": 62.413,
-      "eval_steps_per_second": 2.08,
+      "eval_accuracy": 0.6,
+      "eval_loss": 0.8972867131233215,
+      "eval_runtime": 0.8817,
+      "eval_samples_per_second": 68.052,
+      "eval_steps_per_second": 2.268,
       "step": 193
     },
     {
       "epoch": 25.94,
-      "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 1.0597257614135742,
-      "eval_runtime": 1.9689,
-      "eval_samples_per_second": 30.474,
-      "eval_steps_per_second": 1.016,
+      "eval_accuracy": 0.55,
+      "eval_loss": 0.907895028591156,
+      "eval_runtime": 0.87,
+      "eval_samples_per_second": 68.962,
+      "eval_steps_per_second": 2.299,
       "step": 201
     },
     {
       "epoch": 26.32,
       "learning_rate": 1.3636363636363637e-05,
-      "loss": 0.9195,
+      "loss": 0.7434,
       "step": 204
     },
     {
       "epoch": 26.97,
-      "eval_accuracy": 0.6333333333333333,
-      "eval_loss": 0.9563372135162354,
-      "eval_runtime": 0.9494,
-      "eval_samples_per_second": 63.199,
-      "eval_steps_per_second": 2.107,
+      "eval_accuracy": 0.65,
+      "eval_loss": 0.8386151194572449,
+      "eval_runtime": 0.9123,
+      "eval_samples_per_second": 65.768,
+      "eval_steps_per_second": 2.192,
       "step": 209
     },
     {
       "epoch": 27.87,
       "learning_rate": 1.181818181818182e-05,
-      "loss": 0.9053,
+      "loss": 0.7404,
       "step": 216
     },
     {
       "epoch": 28.0,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 0.9838621020317078,
-      "eval_runtime": 0.8939,
-      "eval_samples_per_second": 67.124,
-      "eval_steps_per_second": 2.237,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 0.8644894361495972,
+      "eval_runtime": 0.8305,
+      "eval_samples_per_second": 72.242,
+      "eval_steps_per_second": 2.408,
       "step": 217
     },
     {
       "epoch": 28.9,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 0.9913681745529175,
-      "eval_runtime": 0.8926,
-      "eval_samples_per_second": 67.223,
-      "eval_steps_per_second": 2.241,
+      "eval_accuracy": 0.5666666666666667,
+      "eval_loss": 0.8599310517311096,
+      "eval_runtime": 0.9003,
+      "eval_samples_per_second": 66.644,
+      "eval_steps_per_second": 2.221,
       "step": 224
     },
     {
       "epoch": 29.42,
       "learning_rate": 1e-05,
-      "loss": 0.8645,
+      "loss": 0.7215,
       "step": 228
     },
     {
       "epoch": 29.94,
-      "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 0.9519664645195007,
-      "eval_runtime": 0.9221,
-      "eval_samples_per_second": 65.066,
-      "eval_steps_per_second": 2.169,
+      "eval_accuracy": 0.65,
+      "eval_loss": 0.8420030474662781,
+      "eval_runtime": 0.8701,
+      "eval_samples_per_second": 68.954,
+      "eval_steps_per_second": 2.298,
       "step": 232
     },
     {
       "epoch": 30.97,
       "learning_rate": 8.181818181818183e-06,
-      "loss": 0.8139,
+      "loss": 0.6743,
       "step": 240
     },
     {
       "epoch": 30.97,
-      "eval_accuracy": 0.5333333333333333,
-      "eval_loss": 0.9572514295578003,
-      "eval_runtime": 0.8817,
-      "eval_samples_per_second": 68.051,
-      "eval_steps_per_second": 2.268,
+      "eval_accuracy": 0.5666666666666667,
+      "eval_loss": 0.855254590511322,
+      "eval_runtime": 0.8716,
+      "eval_samples_per_second": 68.841,
+      "eval_steps_per_second": 2.295,
       "step": 240
     },
     {
       "epoch": 32.0,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 0.9509603977203369,
-      "eval_runtime": 0.9232,
-      "eval_samples_per_second": 64.991,
-      "eval_steps_per_second": 2.166,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 0.8355186581611633,
+      "eval_runtime": 0.8931,
+      "eval_samples_per_second": 67.182,
+      "eval_steps_per_second": 2.239,
       "step": 248
     },
     {
       "epoch": 32.52,
       "learning_rate": 6.363636363636364e-06,
-      "loss": 0.8151,
+      "loss": 0.6767,
       "step": 252
     },
     {
       "epoch": 32.9,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 0.9468602538108826,
-      "eval_runtime": 0.8657,
-      "eval_samples_per_second": 69.308,
-      "eval_steps_per_second": 2.31,
+      "eval_accuracy": 0.5833333333333334,
+      "eval_loss": 0.8694211840629578,
+      "eval_runtime": 0.87,
+      "eval_samples_per_second": 68.966,
+      "eval_steps_per_second": 2.299,
       "step": 255
     },
     {
       "epoch": 33.94,
-      "eval_accuracy": 0.5333333333333333,
-      "eval_loss": 0.9872100353240967,
-      "eval_runtime": 0.9376,
-      "eval_samples_per_second": 63.99,
-      "eval_steps_per_second": 2.133,
+      "eval_accuracy": 0.65,
+      "eval_loss": 0.8559350371360779,
+      "eval_runtime": 0.8505,
+      "eval_samples_per_second": 70.545,
+      "eval_steps_per_second": 2.351,
       "step": 263
     },
     {
       "epoch": 34.06,
       "learning_rate": 4.5454545454545455e-06,
-      "loss": 0.7837,
+      "loss": 0.6606,
       "step": 264
     },
     {
       "epoch": 34.97,
-      "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 0.9278779625892639,
-      "eval_runtime": 1.3601,
-      "eval_samples_per_second": 44.113,
-      "eval_steps_per_second": 1.47,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 0.8350656032562256,
+      "eval_runtime": 0.8389,
+      "eval_samples_per_second": 71.519,
+      "eval_steps_per_second": 2.384,
       "step": 271
     },
     {
       "epoch": 35.61,
       "learning_rate": 2.7272727272727272e-06,
-      "loss": 0.7659,
+      "loss": 0.6488,
       "step": 276
     },
     {
       "epoch": 36.0,
-      "eval_accuracy": 0.5666666666666667,
-      "eval_loss": 0.9173651337623596,
-      "eval_runtime": 0.9252,
-      "eval_samples_per_second": 64.854,
-      "eval_steps_per_second": 2.162,
+      "eval_accuracy": 0.6333333333333333,
+      "eval_loss": 0.8287425637245178,
+      "eval_runtime": 0.8966,
+      "eval_samples_per_second": 66.919,
+      "eval_steps_per_second": 2.231,
       "step": 279
     },
     {
       "epoch": 36.9,
-      "eval_accuracy": 0.5666666666666667,
-      "eval_loss": 0.934675931930542,
-      "eval_runtime": 1.2264,
-      "eval_samples_per_second": 48.923,
-      "eval_steps_per_second": 1.631,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 0.8377331495285034,
+      "eval_runtime": 0.8452,
+      "eval_samples_per_second": 70.987,
+      "eval_steps_per_second": 2.366,
       "step": 286
     },
     {
       "epoch": 37.16,
       "learning_rate": 9.090909090909091e-07,
-      "loss": 0.7835,
+      "loss": 0.6544,
       "step": 288
     },
     {
       "epoch": 37.94,
-      "eval_accuracy": 0.5666666666666667,
-      "eval_loss": 0.9372119903564453,
-      "eval_runtime": 0.9199,
-      "eval_samples_per_second": 65.224,
-      "eval_steps_per_second": 2.174,
+      "eval_accuracy": 0.6,
+      "eval_loss": 0.8406462073326111,
+      "eval_runtime": 0.8978,
+      "eval_samples_per_second": 66.829,
+      "eval_steps_per_second": 2.228,
       "step": 294
     },
     {
       "epoch": 37.94,
       "step": 294,
       "total_flos": 1.2027310550050406e+18,
-      "train_loss": 1.8558632776039798,
-      "train_runtime": 524.2652,
-      "train_samples_per_second": 78.029,
-      "train_steps_per_second": 0.561
+      "train_loss": 1.4638761638784084,
+      "train_runtime": 475.2898,
+      "train_samples_per_second": 86.07,
+      "train_steps_per_second": 0.619
     }
   ],
   "logging_steps": 12,