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: 0.8899
-- Accuracy: 0.6
+- Loss: 0.9637
+- Accuracy: 0.6667
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 37.94,
-    "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_accuracy": 0.6666666666666666,
+    "eval_loss": 0.9636775851249695,
+    "eval_runtime": 1.1302,
+    "eval_samples_per_second": 53.09,
+    "eval_steps_per_second": 1.77,
+    "train_loss": 1.86217847324553,
+    "train_runtime": 485.4466,
+    "train_samples_per_second": 84.269,
+    "train_steps_per_second": 0.606
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 37.94,
-    "eval_accuracy": 0.65,
-    "eval_loss": 0.8386151194572449,
-    "eval_runtime": 1.0431,
-    "eval_samples_per_second": 57.519,
-    "eval_steps_per_second": 1.917
+    "eval_accuracy": 0.6666666666666666,
+    "eval_loss": 0.9636775851249695,
+    "eval_runtime": 1.1302,
+    "eval_samples_per_second": 53.09,
+    "eval_steps_per_second": 1.77
 }

runs/Apr19_07-34-21_DESKTOP-SKBE9FB/events.out.tfevents.1713534195.DESKTOP-SKBE9FB.5900.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abadd4302e47ed73108845080e541be64f9777a18297f767915e5a5ab1169c51
+size 411

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 37.94,
-    "train_loss": 1.4638761638784084,
-    "train_runtime": 475.2898,
-    "train_samples_per_second": 86.07,
-    "train_steps_per_second": 0.619
+    "train_loss": 1.86217847324553,
+    "train_runtime": 485.4466,
+    "train_samples_per_second": 84.269,
+    "train_steps_per_second": 0.606
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.65,
+  "best_metric": 0.6666666666666666,
   "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-va-U5-42B\\checkpoint-209",
   "epoch": 37.935483870967744,
   "eval_steps": 500,
@@ -11,497 +11,497 @@
     {
       "epoch": 0.9,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 6.174824237823486,
-      "eval_runtime": 0.845,
-      "eval_samples_per_second": 71.01,
-      "eval_steps_per_second": 2.367,
+      "eval_loss": 7.866283893585205,
+      "eval_runtime": 0.8861,
+      "eval_samples_per_second": 67.712,
+      "eval_steps_per_second": 2.257,
       "step": 7
     },
     {
       "epoch": 1.55,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 5.327,
+      "loss": 6.936,
       "step": 12
     },
     {
       "epoch": 1.94,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 6.066026210784912,
-      "eval_runtime": 0.8511,
-      "eval_samples_per_second": 70.499,
-      "eval_steps_per_second": 2.35,
+      "eval_loss": 7.757172584533691,
+      "eval_runtime": 0.8501,
+      "eval_samples_per_second": 70.577,
+      "eval_steps_per_second": 2.353,
       "step": 15
     },
     {
       "epoch": 2.97,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 5.4902424812316895,
-      "eval_runtime": 0.8648,
-      "eval_samples_per_second": 69.381,
-      "eval_steps_per_second": 2.313,
+      "eval_loss": 7.178952217102051,
+      "eval_runtime": 0.9919,
+      "eval_samples_per_second": 60.493,
+      "eval_steps_per_second": 2.016,
       "step": 23
     },
     {
       "epoch": 3.1,
       "learning_rate": 3.2000000000000005e-05,
-      "loss": 5.0963,
+      "loss": 6.7016,
       "step": 24
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 4.276790142059326,
-      "eval_runtime": 0.8095,
-      "eval_samples_per_second": 74.116,
-      "eval_steps_per_second": 2.471,
+      "eval_loss": 5.903270244598389,
+      "eval_runtime": 0.9119,
+      "eval_samples_per_second": 65.796,
+      "eval_steps_per_second": 2.193,
       "step": 31
     },
     {
       "epoch": 4.65,
       "learning_rate": 3.909090909090909e-05,
-      "loss": 3.9193,
+      "loss": 5.5439,
       "step": 36
     },
     {
       "epoch": 4.9,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 3.0012993812561035,
-      "eval_runtime": 0.8424,
-      "eval_samples_per_second": 71.228,
-      "eval_steps_per_second": 2.374,
+      "eval_loss": 4.611630916595459,
+      "eval_runtime": 0.8484,
+      "eval_samples_per_second": 70.723,
+      "eval_steps_per_second": 2.357,
       "step": 38
     },
     {
       "epoch": 5.94,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 1.9289205074310303,
-      "eval_runtime": 0.8816,
-      "eval_samples_per_second": 68.055,
-      "eval_steps_per_second": 2.269,
+      "eval_loss": 3.2829790115356445,
+      "eval_runtime": 0.8288,
+      "eval_samples_per_second": 72.396,
+      "eval_steps_per_second": 2.413,
       "step": 46
     },
     {
       "epoch": 6.19,
       "learning_rate": 3.7272727272727276e-05,
-      "loss": 2.2222,
+      "loss": 3.6477,
       "step": 48
     },
     {
       "epoch": 6.97,
       "eval_accuracy": 0.11666666666666667,
-      "eval_loss": 1.385749101638794,
-      "eval_runtime": 0.8792,
-      "eval_samples_per_second": 68.244,
-      "eval_steps_per_second": 2.275,
+      "eval_loss": 2.2014169692993164,
+      "eval_runtime": 0.9115,
+      "eval_samples_per_second": 65.826,
+      "eval_steps_per_second": 2.194,
       "step": 54
     },
     {
       "epoch": 7.74,
       "learning_rate": 3.545454545454546e-05,
-      "loss": 1.4465,
+      "loss": 2.2506,
       "step": 60
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.43333333333333335,
-      "eval_loss": 1.342329740524292,
-      "eval_runtime": 0.8314,
-      "eval_samples_per_second": 72.165,
-      "eval_steps_per_second": 2.405,
+      "eval_accuracy": 0.45,
+      "eval_loss": 1.5647218227386475,
+      "eval_runtime": 0.9989,
+      "eval_samples_per_second": 60.066,
+      "eval_steps_per_second": 2.002,
       "step": 62
     },
     {
       "epoch": 8.9,
       "eval_accuracy": 0.45,
-      "eval_loss": 1.2786400318145752,
-      "eval_runtime": 0.8842,
-      "eval_samples_per_second": 67.857,
-      "eval_steps_per_second": 2.262,
+      "eval_loss": 1.3160017728805542,
+      "eval_runtime": 0.8669,
+      "eval_samples_per_second": 69.211,
+      "eval_steps_per_second": 2.307,
       "step": 69
     },
     {
       "epoch": 9.29,
       "learning_rate": 3.363636363636364e-05,
-      "loss": 1.3709,
+      "loss": 1.5088,
       "step": 72
     },
     {
       "epoch": 9.94,
-      "eval_accuracy": 0.4666666666666667,
-      "eval_loss": 1.2653944492340088,
-      "eval_runtime": 0.8365,
-      "eval_samples_per_second": 71.724,
-      "eval_steps_per_second": 2.391,
+      "eval_accuracy": 0.3333333333333333,
+      "eval_loss": 1.3676066398620605,
+      "eval_runtime": 0.8692,
+      "eval_samples_per_second": 69.028,
+      "eval_steps_per_second": 2.301,
       "step": 77
     },
     {
       "epoch": 10.84,
       "learning_rate": 3.181818181818182e-05,
-      "loss": 1.3511,
+      "loss": 1.3868,
       "step": 84
     },
     {
       "epoch": 10.97,
-      "eval_accuracy": 0.4666666666666667,
-      "eval_loss": 1.2605175971984863,
-      "eval_runtime": 0.8152,
-      "eval_samples_per_second": 73.599,
-      "eval_steps_per_second": 2.453,
+      "eval_accuracy": 0.45,
+      "eval_loss": 1.3390430212020874,
+      "eval_runtime": 0.8862,
+      "eval_samples_per_second": 67.708,
+      "eval_steps_per_second": 2.257,
       "step": 85
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.4666666666666667,
-      "eval_loss": 1.2183691263198853,
-      "eval_runtime": 0.8329,
-      "eval_samples_per_second": 72.04,
-      "eval_steps_per_second": 2.401,
+      "eval_accuracy": 0.38333333333333336,
+      "eval_loss": 1.322303056716919,
+      "eval_runtime": 0.8998,
+      "eval_samples_per_second": 66.681,
+      "eval_steps_per_second": 2.223,
       "step": 93
     },
     {
       "epoch": 12.39,
       "learning_rate": 3.0000000000000004e-05,
-      "loss": 1.2749,
+      "loss": 1.351,
       "step": 96
     },
     {
       "epoch": 12.9,
-      "eval_accuracy": 0.5,
-      "eval_loss": 1.289415955543518,
-      "eval_runtime": 0.8348,
-      "eval_samples_per_second": 71.87,
-      "eval_steps_per_second": 2.396,
+      "eval_accuracy": 0.45,
+      "eval_loss": 1.3155896663665771,
+      "eval_runtime": 0.867,
+      "eval_samples_per_second": 69.206,
+      "eval_steps_per_second": 2.307,
       "step": 100
     },
     {
       "epoch": 13.94,
       "learning_rate": 2.8181818181818185e-05,
-      "loss": 1.222,
+      "loss": 1.3271,
       "step": 108
     },
     {
       "epoch": 13.94,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 1.2071518898010254,
-      "eval_runtime": 0.8805,
-      "eval_samples_per_second": 68.142,
-      "eval_steps_per_second": 2.271,
+      "eval_accuracy": 0.48333333333333334,
+      "eval_loss": 1.3484553098678589,
+      "eval_runtime": 0.8698,
+      "eval_samples_per_second": 68.985,
+      "eval_steps_per_second": 2.299,
       "step": 108
     },
     {
       "epoch": 14.97,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 1.1748946905136108,
-      "eval_runtime": 0.8381,
-      "eval_samples_per_second": 71.592,
-      "eval_steps_per_second": 2.386,
+      "eval_accuracy": 0.48333333333333334,
+      "eval_loss": 1.2645785808563232,
+      "eval_runtime": 0.9492,
+      "eval_samples_per_second": 63.21,
+      "eval_steps_per_second": 2.107,
       "step": 116
     },
     {
       "epoch": 15.48,
       "learning_rate": 2.6363636363636365e-05,
-      "loss": 1.1668,
+      "loss": 1.2322,
       "step": 120
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 1.198820948600769,
-      "eval_runtime": 0.8663,
-      "eval_samples_per_second": 69.26,
-      "eval_steps_per_second": 2.309,
+      "eval_accuracy": 0.48333333333333334,
+      "eval_loss": 1.2307511568069458,
+      "eval_runtime": 0.8623,
+      "eval_samples_per_second": 69.584,
+      "eval_steps_per_second": 2.319,
       "step": 124
     },
     {
       "epoch": 16.9,
-      "eval_accuracy": 0.5166666666666667,
-      "eval_loss": 1.2305818796157837,
-      "eval_runtime": 0.8839,
-      "eval_samples_per_second": 67.884,
-      "eval_steps_per_second": 2.263,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.215951681137085,
+      "eval_runtime": 0.8482,
+      "eval_samples_per_second": 70.734,
+      "eval_steps_per_second": 2.358,
       "step": 131
     },
     {
       "epoch": 17.03,
       "learning_rate": 2.454545454545455e-05,
-      "loss": 1.101,
+      "loss": 1.22,
       "step": 132
     },
     {
       "epoch": 17.94,
-      "eval_accuracy": 0.5333333333333333,
-      "eval_loss": 1.1431663036346436,
-      "eval_runtime": 0.8778,
-      "eval_samples_per_second": 68.351,
-      "eval_steps_per_second": 2.278,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.2015122175216675,
+      "eval_runtime": 0.8406,
+      "eval_samples_per_second": 71.379,
+      "eval_steps_per_second": 2.379,
       "step": 139
     },
     {
       "epoch": 18.58,
       "learning_rate": 2.2727272727272733e-05,
-      "loss": 1.029,
+      "loss": 1.1899,
       "step": 144
     },
     {
       "epoch": 18.97,
-      "eval_accuracy": 0.55,
-      "eval_loss": 1.0208261013031006,
-      "eval_runtime": 0.8723,
-      "eval_samples_per_second": 68.786,
-      "eval_steps_per_second": 2.293,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.2008181810379028,
+      "eval_runtime": 0.9372,
+      "eval_samples_per_second": 64.019,
+      "eval_steps_per_second": 2.134,
       "step": 147
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.6166666666666667,
-      "eval_loss": 0.9576696157455444,
-      "eval_runtime": 0.849,
-      "eval_samples_per_second": 70.671,
-      "eval_steps_per_second": 2.356,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.1605515480041504,
+      "eval_runtime": 1.0194,
+      "eval_samples_per_second": 58.858,
+      "eval_steps_per_second": 1.962,
       "step": 155
     },
     {
       "epoch": 20.13,
       "learning_rate": 2.090909090909091e-05,
-      "loss": 0.9403,
+      "loss": 1.109,
       "step": 156
     },
     {
       "epoch": 20.9,
-      "eval_accuracy": 0.5,
-      "eval_loss": 0.9478870034217834,
-      "eval_runtime": 0.8337,
-      "eval_samples_per_second": 71.971,
-      "eval_steps_per_second": 2.399,
+      "eval_accuracy": 0.5666666666666667,
+      "eval_loss": 1.118190050125122,
+      "eval_runtime": 0.8587,
+      "eval_samples_per_second": 69.877,
+      "eval_steps_per_second": 2.329,
       "step": 162
     },
     {
       "epoch": 21.68,
       "learning_rate": 1.9090909090909094e-05,
-      "loss": 0.8887,
+      "loss": 1.0603,
       "step": 168
     },
     {
       "epoch": 21.94,
-      "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 0.8909610509872437,
-      "eval_runtime": 0.8389,
-      "eval_samples_per_second": 71.521,
-      "eval_steps_per_second": 2.384,
+      "eval_accuracy": 0.5333333333333333,
+      "eval_loss": 1.0855048894882202,
+      "eval_runtime": 0.8678,
+      "eval_samples_per_second": 69.144,
+      "eval_steps_per_second": 2.305,
       "step": 170
     },
     {
       "epoch": 22.97,
-      "eval_accuracy": 0.5333333333333333,
-      "eval_loss": 0.9441585540771484,
-      "eval_runtime": 0.8343,
-      "eval_samples_per_second": 71.914,
-      "eval_steps_per_second": 2.397,
+      "eval_accuracy": 0.5666666666666667,
+      "eval_loss": 1.0762869119644165,
+      "eval_runtime": 0.9558,
+      "eval_samples_per_second": 62.772,
+      "eval_steps_per_second": 2.092,
       "step": 178
     },
     {
       "epoch": 23.23,
       "learning_rate": 1.7272727272727274e-05,
-      "loss": 0.8506,
+      "loss": 1.0264,
       "step": 180
     },
     {
       "epoch": 24.0,
-      "eval_accuracy": 0.6,
-      "eval_loss": 0.8923367261886597,
-      "eval_runtime": 0.8341,
-      "eval_samples_per_second": 71.937,
-      "eval_steps_per_second": 2.398,
+      "eval_accuracy": 0.5833333333333334,
+      "eval_loss": 1.1152801513671875,
+      "eval_runtime": 0.9634,
+      "eval_samples_per_second": 62.28,
+      "eval_steps_per_second": 2.076,
       "step": 186
     },
     {
       "epoch": 24.77,
       "learning_rate": 1.5454545454545454e-05,
-      "loss": 0.8064,
+      "loss": 1.0086,
       "step": 192
     },
     {
       "epoch": 24.9,
-      "eval_accuracy": 0.6,
-      "eval_loss": 0.8972867131233215,
-      "eval_runtime": 0.8817,
-      "eval_samples_per_second": 68.052,
-      "eval_steps_per_second": 2.268,
+      "eval_accuracy": 0.65,
+      "eval_loss": 1.0770429372787476,
+      "eval_runtime": 0.8946,
+      "eval_samples_per_second": 67.069,
+      "eval_steps_per_second": 2.236,
       "step": 193
     },
     {
       "epoch": 25.94,
-      "eval_accuracy": 0.55,
-      "eval_loss": 0.907895028591156,
-      "eval_runtime": 0.87,
-      "eval_samples_per_second": 68.962,
-      "eval_steps_per_second": 2.299,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 1.0040525197982788,
+      "eval_runtime": 0.8805,
+      "eval_samples_per_second": 68.142,
+      "eval_steps_per_second": 2.271,
       "step": 201
     },
     {
       "epoch": 26.32,
       "learning_rate": 1.3636363636363637e-05,
-      "loss": 0.7434,
+      "loss": 0.9301,
       "step": 204
     },
     {
       "epoch": 26.97,
-      "eval_accuracy": 0.65,
-      "eval_loss": 0.8386151194572449,
-      "eval_runtime": 0.9123,
-      "eval_samples_per_second": 65.768,
-      "eval_steps_per_second": 2.192,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 0.9636775851249695,
+      "eval_runtime": 0.9061,
+      "eval_samples_per_second": 66.221,
+      "eval_steps_per_second": 2.207,
       "step": 209
     },
     {
       "epoch": 27.87,
       "learning_rate": 1.181818181818182e-05,
-      "loss": 0.7404,
+      "loss": 0.9077,
       "step": 216
     },
     {
       "epoch": 28.0,
-      "eval_accuracy": 0.6166666666666667,
-      "eval_loss": 0.8644894361495972,
-      "eval_runtime": 0.8305,
-      "eval_samples_per_second": 72.242,
-      "eval_steps_per_second": 2.408,
+      "eval_accuracy": 0.5833333333333334,
+      "eval_loss": 0.9824321866035461,
+      "eval_runtime": 0.8887,
+      "eval_samples_per_second": 67.511,
+      "eval_steps_per_second": 2.25,
       "step": 217
     },
     {
       "epoch": 28.9,
-      "eval_accuracy": 0.5666666666666667,
-      "eval_loss": 0.8599310517311096,
-      "eval_runtime": 0.9003,
-      "eval_samples_per_second": 66.644,
-      "eval_steps_per_second": 2.221,
+      "eval_accuracy": 0.6,
+      "eval_loss": 0.9485177397727966,
+      "eval_runtime": 0.9068,
+      "eval_samples_per_second": 66.168,
+      "eval_steps_per_second": 2.206,
       "step": 224
     },
     {
       "epoch": 29.42,
       "learning_rate": 1e-05,
-      "loss": 0.7215,
+      "loss": 0.8725,
       "step": 228
     },
     {
       "epoch": 29.94,
-      "eval_accuracy": 0.65,
-      "eval_loss": 0.8420030474662781,
-      "eval_runtime": 0.8701,
-      "eval_samples_per_second": 68.954,
-      "eval_steps_per_second": 2.298,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 0.9294353723526001,
+      "eval_runtime": 0.8792,
+      "eval_samples_per_second": 68.246,
+      "eval_steps_per_second": 2.275,
       "step": 232
     },
     {
       "epoch": 30.97,
       "learning_rate": 8.181818181818183e-06,
-      "loss": 0.6743,
+      "loss": 0.8203,
       "step": 240
     },
     {
       "epoch": 30.97,
-      "eval_accuracy": 0.5666666666666667,
-      "eval_loss": 0.855254590511322,
-      "eval_runtime": 0.8716,
-      "eval_samples_per_second": 68.841,
-      "eval_steps_per_second": 2.295,
+      "eval_accuracy": 0.6166666666666667,
+      "eval_loss": 0.9347657561302185,
+      "eval_runtime": 0.8842,
+      "eval_samples_per_second": 67.856,
+      "eval_steps_per_second": 2.262,
       "step": 240
     },
     {
       "epoch": 32.0,
-      "eval_accuracy": 0.6166666666666667,
-      "eval_loss": 0.8355186581611633,
-      "eval_runtime": 0.8931,
-      "eval_samples_per_second": 67.182,
-      "eval_steps_per_second": 2.239,
+      "eval_accuracy": 0.6,
+      "eval_loss": 0.9295245409011841,
+      "eval_runtime": 0.8786,
+      "eval_samples_per_second": 68.289,
+      "eval_steps_per_second": 2.276,
       "step": 248
     },
     {
       "epoch": 32.52,
       "learning_rate": 6.363636363636364e-06,
-      "loss": 0.6767,
+      "loss": 0.8211,
       "step": 252
     },
     {
       "epoch": 32.9,
-      "eval_accuracy": 0.5833333333333334,
-      "eval_loss": 0.8694211840629578,
-      "eval_runtime": 0.87,
-      "eval_samples_per_second": 68.966,
-      "eval_steps_per_second": 2.299,
+      "eval_accuracy": 0.6,
+      "eval_loss": 0.9167192578315735,
+      "eval_runtime": 0.8974,
+      "eval_samples_per_second": 66.863,
+      "eval_steps_per_second": 2.229,
       "step": 255
     },
     {
       "epoch": 33.94,
-      "eval_accuracy": 0.65,
-      "eval_loss": 0.8559350371360779,
-      "eval_runtime": 0.8505,
-      "eval_samples_per_second": 70.545,
-      "eval_steps_per_second": 2.351,
+      "eval_accuracy": 0.5833333333333334,
+      "eval_loss": 0.9280623197555542,
+      "eval_runtime": 0.8572,
+      "eval_samples_per_second": 69.993,
+      "eval_steps_per_second": 2.333,
       "step": 263
     },
     {
       "epoch": 34.06,
       "learning_rate": 4.5454545454545455e-06,
-      "loss": 0.6606,
+      "loss": 0.7916,
       "step": 264
     },
     {
       "epoch": 34.97,
-      "eval_accuracy": 0.6166666666666667,
-      "eval_loss": 0.8350656032562256,
-      "eval_runtime": 0.8389,
-      "eval_samples_per_second": 71.519,
-      "eval_steps_per_second": 2.384,
+      "eval_accuracy": 0.6333333333333333,
+      "eval_loss": 0.880310595035553,
+      "eval_runtime": 0.8918,
+      "eval_samples_per_second": 67.281,
+      "eval_steps_per_second": 2.243,
       "step": 271
     },
     {
       "epoch": 35.61,
       "learning_rate": 2.7272727272727272e-06,
-      "loss": 0.6488,
+      "loss": 0.7822,
       "step": 276
     },
     {
       "epoch": 36.0,
       "eval_accuracy": 0.6333333333333333,
-      "eval_loss": 0.8287425637245178,
-      "eval_runtime": 0.8966,
-      "eval_samples_per_second": 66.919,
-      "eval_steps_per_second": 2.231,
+      "eval_loss": 0.8785258531570435,
+      "eval_runtime": 0.8791,
+      "eval_samples_per_second": 68.25,
+      "eval_steps_per_second": 2.275,
       "step": 279
     },
     {
       "epoch": 36.9,
-      "eval_accuracy": 0.6166666666666667,
-      "eval_loss": 0.8377331495285034,
-      "eval_runtime": 0.8452,
-      "eval_samples_per_second": 70.987,
-      "eval_steps_per_second": 2.366,
+      "eval_accuracy": 0.6,
+      "eval_loss": 0.8906211853027344,
+      "eval_runtime": 0.8494,
+      "eval_samples_per_second": 70.635,
+      "eval_steps_per_second": 2.354,
       "step": 286
     },
     {
       "epoch": 37.16,
       "learning_rate": 9.090909090909091e-07,
-      "loss": 0.6544,
+      "loss": 0.7937,
       "step": 288
     },
     {
       "epoch": 37.94,
       "eval_accuracy": 0.6,
-      "eval_loss": 0.8406462073326111,
-      "eval_runtime": 0.8978,
-      "eval_samples_per_second": 66.829,
-      "eval_steps_per_second": 2.228,
+      "eval_loss": 0.889906108379364,
+      "eval_runtime": 0.8734,
+      "eval_samples_per_second": 68.698,
+      "eval_steps_per_second": 2.29,
       "step": 294
     },
     {
       "epoch": 37.94,
       "step": 294,
       "total_flos": 1.2027310550050406e+18,
-      "train_loss": 1.4638761638784084,
-      "train_runtime": 475.2898,
-      "train_samples_per_second": 86.07,
-      "train_steps_per_second": 0.619
+      "train_loss": 1.86217847324553,
+      "train_runtime": 485.4466,
+      "train_samples_per_second": 84.269,
+      "train_steps_per_second": 0.606
     }
   ],
   "logging_steps": 12,