🍻 cheers

README.md

@@ -2,6 +2,7 @@
 license: apache-2.0
 base_model: microsoft/swin-tiny-patch4-window7-224
 tags:
+- image-classification
 - generated_from_trainer
 metrics:
 - accuracy
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # swin-tiny-patch4-window7-224-finetuned_ASL_Isolated_Swin_dataset2
 
-This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on an unknown dataset.
+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 ASL_Isolated_Swin_dataset dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0867
+- Loss: 0.1319
 - Accuracy: 0.9769
 
 ## Model description

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 10.0,
-    "eval_accuracy": 0.9846153846153847,
-    "eval_loss": 0.06404715776443481,
-    "eval_runtime": 2.1321,
-    "eval_samples_per_second": 121.943,
-    "eval_steps_per_second": 15.477,
+    "eval_accuracy": 0.9769230769230769,
+    "eval_loss": 0.13189926743507385,
+    "eval_runtime": 1.9108,
+    "eval_samples_per_second": 136.069,
+    "eval_steps_per_second": 17.27,
     "total_flos": 3.651306173035315e+17,
-    "train_loss": 0.6966753153697304,
-    "train_runtime": 217.2288,
-    "train_samples_per_second": 67.579,
-    "train_steps_per_second": 4.235
+    "train_loss": 0.7250858170830685,
+    "train_runtime": 208.4502,
+    "train_samples_per_second": 70.424,
+    "train_steps_per_second": 4.414
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
-    "eval_accuracy": 0.9846153846153847,
-    "eval_loss": 0.06404715776443481,
-    "eval_runtime": 2.1321,
-    "eval_samples_per_second": 121.943,
-    "eval_steps_per_second": 15.477
+    "eval_accuracy": 0.9769230769230769,
+    "eval_loss": 0.13189926743507385,
+    "eval_runtime": 1.9108,
+    "eval_samples_per_second": 136.069,
+    "eval_steps_per_second": 17.27
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
     "total_flos": 3.651306173035315e+17,
-    "train_loss": 0.6966753153697304,
-    "train_runtime": 217.2288,
-    "train_samples_per_second": 67.579,
-    "train_steps_per_second": 4.235
+    "train_loss": 0.7250858170830685,
+    "train_runtime": 208.4502,
+    "train_samples_per_second": 70.424,
+    "train_steps_per_second": 4.414
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.9846153846153847,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned_ASL_Isolated_Swin_dataset2/checkpoint-800",
+  "best_metric": 0.9769230769230769,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned_ASL_Isolated_Swin_dataset2/checkpoint-700",
   "epoch": 10.0,
   "eval_steps": 100,
   "global_step": 920,
@@ -10,645 +10,645 @@
   "log_history": [
     {
       "epoch": 0.11,
-      "learning_rate": 0.00019782608695652175,
-      "loss": 3.3609,
+      "learning_rate": 0.00019804347826086956,
+      "loss": 3.3183,
       "step": 10
     },
     {
       "epoch": 0.22,
-      "learning_rate": 0.0001956521739130435,
-      "loss": 3.1386,
+      "learning_rate": 0.0001958695652173913,
+      "loss": 3.1344,
       "step": 20
     },
     {
       "epoch": 0.33,
-      "learning_rate": 0.00019347826086956523,
-      "loss": 2.9847,
+      "learning_rate": 0.00019369565217391307,
+      "loss": 3.1166,
       "step": 30
     },
     {
       "epoch": 0.43,
-      "learning_rate": 0.00019130434782608697,
-      "loss": 2.7624,
+      "learning_rate": 0.00019152173913043478,
+      "loss": 2.8413,
       "step": 40
     },
     {
       "epoch": 0.54,
-      "learning_rate": 0.0001891304347826087,
-      "loss": 2.2708,
+      "learning_rate": 0.00018934782608695655,
+      "loss": 2.4092,
       "step": 50
     },
     {
       "epoch": 0.65,
       "learning_rate": 0.00018717391304347826,
-      "loss": 2.1984,
+      "loss": 2.173,
       "step": 60
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.00018500000000000002,
-      "loss": 1.8882,
+      "loss": 1.9006,
       "step": 70
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.00018282608695652174,
-      "loss": 1.6971,
+      "loss": 1.8329,
       "step": 80
     },
     {
       "epoch": 0.98,
-      "learning_rate": 0.00018065217391304348,
-      "loss": 1.4803,
+      "learning_rate": 0.00018086956521739132,
+      "loss": 1.7864,
       "step": 90
     },
     {
       "epoch": 1.09,
-      "learning_rate": 0.00017847826086956522,
-      "loss": 1.39,
+      "learning_rate": 0.00017869565217391305,
+      "loss": 1.6751,
       "step": 100
     },
     {
       "epoch": 1.09,
-      "eval_accuracy": 0.6346153846153846,
-      "eval_loss": 1.182677984237671,
-      "eval_runtime": 2.4526,
-      "eval_samples_per_second": 106.009,
-      "eval_steps_per_second": 13.455,
+      "eval_accuracy": 0.6307692307692307,
+      "eval_loss": 1.1246747970581055,
+      "eval_runtime": 1.9197,
+      "eval_samples_per_second": 135.438,
+      "eval_steps_per_second": 17.19,
       "step": 100
     },
     {
       "epoch": 1.2,
-      "learning_rate": 0.00017630434782608695,
-      "loss": 1.2313,
+      "learning_rate": 0.0001765217391304348,
+      "loss": 1.3096,
       "step": 110
     },
     {
       "epoch": 1.3,
-      "learning_rate": 0.00017413043478260872,
-      "loss": 1.2511,
+      "learning_rate": 0.00017434782608695653,
+      "loss": 1.3416,
       "step": 120
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.00017217391304347827,
-      "loss": 1.1545,
+      "loss": 1.4005,
       "step": 130
     },
     {
       "epoch": 1.52,
       "learning_rate": 0.00017,
-      "loss": 1.0298,
+      "loss": 1.1935,
       "step": 140
     },
     {
       "epoch": 1.63,
       "learning_rate": 0.00016782608695652175,
-      "loss": 1.0783,
+      "loss": 1.0311,
       "step": 150
     },
     {
       "epoch": 1.74,
       "learning_rate": 0.0001656521739130435,
-      "loss": 1.0336,
+      "loss": 0.9773,
       "step": 160
     },
     {
       "epoch": 1.85,
       "learning_rate": 0.00016347826086956523,
-      "loss": 1.0377,
+      "loss": 1.0554,
       "step": 170
     },
     {
       "epoch": 1.96,
       "learning_rate": 0.00016130434782608697,
-      "loss": 1.1622,
+      "loss": 0.9596,
       "step": 180
     },
     {
       "epoch": 2.07,
       "learning_rate": 0.00015913043478260868,
-      "loss": 0.9864,
+      "loss": 1.1052,
       "step": 190
     },
     {
       "epoch": 2.17,
       "learning_rate": 0.00015695652173913045,
-      "loss": 0.8972,
+      "loss": 0.928,
       "step": 200
     },
     {
       "epoch": 2.17,
-      "eval_accuracy": 0.7807692307692308,
-      "eval_loss": 0.6287302374839783,
-      "eval_runtime": 2.6916,
-      "eval_samples_per_second": 96.596,
-      "eval_steps_per_second": 12.26,
+      "eval_accuracy": 0.8461538461538461,
+      "eval_loss": 0.5504888892173767,
+      "eval_runtime": 1.8676,
+      "eval_samples_per_second": 139.218,
+      "eval_steps_per_second": 17.67,
       "step": 200
     },
     {
       "epoch": 2.28,
       "learning_rate": 0.0001547826086956522,
-      "loss": 0.7377,
+      "loss": 0.7055,
       "step": 210
     },
     {
       "epoch": 2.39,
       "learning_rate": 0.00015260869565217393,
-      "loss": 0.8899,
+      "loss": 1.0022,
       "step": 220
     },
     {
       "epoch": 2.5,
       "learning_rate": 0.00015043478260869567,
-      "loss": 0.6429,
+      "loss": 0.8177,
       "step": 230
     },
     {
       "epoch": 2.61,
       "learning_rate": 0.0001482608695652174,
-      "loss": 0.8942,
+      "loss": 0.8889,
       "step": 240
     },
     {
       "epoch": 2.72,
       "learning_rate": 0.00014608695652173914,
-      "loss": 0.8716,
+      "loss": 0.7134,
       "step": 250
     },
     {
       "epoch": 2.83,
       "learning_rate": 0.00014391304347826086,
-      "loss": 0.8024,
+      "loss": 0.7063,
       "step": 260
     },
     {
       "epoch": 2.93,
       "learning_rate": 0.00014173913043478262,
-      "loss": 0.5412,
+      "loss": 0.558,
       "step": 270
     },
     {
       "epoch": 3.04,
       "learning_rate": 0.00013956521739130434,
-      "loss": 0.7387,
+      "loss": 0.6297,
       "step": 280
     },
     {
       "epoch": 3.15,
       "learning_rate": 0.0001373913043478261,
-      "loss": 0.5791,
+      "loss": 0.6457,
       "step": 290
     },
     {
       "epoch": 3.26,
       "learning_rate": 0.00013521739130434784,
-      "loss": 0.4884,
+      "loss": 0.4477,
       "step": 300
     },
     {
       "epoch": 3.26,
-      "eval_accuracy": 0.8961538461538462,
-      "eval_loss": 0.29269811511039734,
-      "eval_runtime": 1.9876,
-      "eval_samples_per_second": 130.808,
-      "eval_steps_per_second": 16.603,
+      "eval_accuracy": 0.9038461538461539,
+      "eval_loss": 0.31669339537620544,
+      "eval_runtime": 1.8909,
+      "eval_samples_per_second": 137.504,
+      "eval_steps_per_second": 17.452,
       "step": 300
     },
     {
       "epoch": 3.37,
       "learning_rate": 0.00013304347826086958,
-      "loss": 0.6485,
+      "loss": 0.5847,
       "step": 310
     },
     {
       "epoch": 3.48,
       "learning_rate": 0.00013086956521739132,
-      "loss": 0.7623,
+      "loss": 0.8343,
       "step": 320
     },
     {
       "epoch": 3.59,
       "learning_rate": 0.00012869565217391303,
-      "loss": 0.6887,
+      "loss": 0.7171,
       "step": 330
     },
     {
       "epoch": 3.7,
       "learning_rate": 0.0001265217391304348,
-      "loss": 0.5794,
+      "loss": 0.5016,
       "step": 340
     },
     {
       "epoch": 3.8,
       "learning_rate": 0.0001243478260869565,
-      "loss": 0.6654,
+      "loss": 0.683,
       "step": 350
     },
     {
       "epoch": 3.91,
       "learning_rate": 0.00012217391304347828,
-      "loss": 0.5027,
+      "loss": 0.653,
       "step": 360
     },
     {
       "epoch": 4.02,
       "learning_rate": 0.00012,
-      "loss": 0.4336,
+      "loss": 0.6446,
       "step": 370
     },
     {
       "epoch": 4.13,
       "learning_rate": 0.00011782608695652176,
-      "loss": 0.5295,
+      "loss": 0.449,
       "step": 380
     },
     {
       "epoch": 4.24,
       "learning_rate": 0.00011565217391304348,
-      "loss": 0.6389,
+      "loss": 0.5571,
       "step": 390
     },
     {
       "epoch": 4.35,
       "learning_rate": 0.00011347826086956523,
-      "loss": 0.4179,
+      "loss": 0.5154,
       "step": 400
     },
     {
       "epoch": 4.35,
       "eval_accuracy": 0.9423076923076923,
-      "eval_loss": 0.1795085072517395,
-      "eval_runtime": 2.6974,
-      "eval_samples_per_second": 96.391,
-      "eval_steps_per_second": 12.234,
+      "eval_loss": 0.21009394526481628,
+      "eval_runtime": 1.8763,
+      "eval_samples_per_second": 138.572,
+      "eval_steps_per_second": 17.588,
       "step": 400
     },
     {
       "epoch": 4.46,
-      "learning_rate": 0.00011152173913043477,
-      "loss": 0.5516,
+      "learning_rate": 0.00011130434782608696,
+      "loss": 0.4713,
       "step": 410
     },
     {
       "epoch": 4.57,
-      "learning_rate": 0.00010934782608695653,
-      "loss": 0.4264,
+      "learning_rate": 0.00010913043478260869,
+      "loss": 0.451,
       "step": 420
     },
     {
       "epoch": 4.67,
-      "learning_rate": 0.00010717391304347825,
-      "loss": 0.4827,
+      "learning_rate": 0.00010695652173913044,
+      "loss": 0.5028,
       "step": 430
     },
     {
       "epoch": 4.78,
-      "learning_rate": 0.000105,
-      "loss": 0.4434,
+      "learning_rate": 0.00010478260869565218,
+      "loss": 0.4907,
       "step": 440
     },
     {
       "epoch": 4.89,
-      "learning_rate": 0.00010282608695652174,
-      "loss": 0.4812,
+      "learning_rate": 0.00010260869565217393,
+      "loss": 0.4996,
       "step": 450
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.0001006521739130435,
-      "loss": 0.5592,
+      "learning_rate": 0.00010043478260869566,
+      "loss": 0.5479,
       "step": 460
     },
     {
       "epoch": 5.11,
-      "learning_rate": 9.847826086956522e-05,
-      "loss": 0.4754,
+      "learning_rate": 9.82608695652174e-05,
+      "loss": 0.4643,
       "step": 470
     },
     {
       "epoch": 5.22,
-      "learning_rate": 9.630434782608696e-05,
-      "loss": 0.3776,
+      "learning_rate": 9.608695652173914e-05,
+      "loss": 0.5042,
       "step": 480
     },
     {
       "epoch": 5.33,
-      "learning_rate": 9.41304347826087e-05,
-      "loss": 0.4084,
+      "learning_rate": 9.391304347826087e-05,
+      "loss": 0.4327,
       "step": 490
     },
     {
       "epoch": 5.43,
-      "learning_rate": 9.195652173913044e-05,
-      "loss": 0.4169,
+      "learning_rate": 9.173913043478261e-05,
+      "loss": 0.4627,
       "step": 500
     },
     {
       "epoch": 5.43,
-      "eval_accuracy": 0.95,
-      "eval_loss": 0.1563507616519928,
-      "eval_runtime": 1.8046,
-      "eval_samples_per_second": 144.078,
-      "eval_steps_per_second": 18.287,
+      "eval_accuracy": 0.9346153846153846,
+      "eval_loss": 0.2289218157529831,
+      "eval_runtime": 2.011,
+      "eval_samples_per_second": 129.291,
+      "eval_steps_per_second": 16.41,
       "step": 500
     },
     {
       "epoch": 5.54,
-      "learning_rate": 8.978260869565218e-05,
-      "loss": 0.371,
+      "learning_rate": 8.956521739130435e-05,
+      "loss": 0.391,
       "step": 510
     },
     {
       "epoch": 5.65,
-      "learning_rate": 8.760869565217392e-05,
-      "loss": 0.4745,
+      "learning_rate": 8.739130434782609e-05,
+      "loss": 0.4032,
       "step": 520
     },
     {
       "epoch": 5.76,
-      "learning_rate": 8.543478260869566e-05,
-      "loss": 0.4209,
+      "learning_rate": 8.521739130434783e-05,
+      "loss": 0.3878,
       "step": 530
     },
     {
       "epoch": 5.87,
-      "learning_rate": 8.32608695652174e-05,
-      "loss": 0.4645,
+      "learning_rate": 8.304347826086957e-05,
+      "loss": 0.5423,
       "step": 540
     },
     {
       "epoch": 5.98,
-      "learning_rate": 8.108695652173914e-05,
-      "loss": 0.4322,
+      "learning_rate": 8.086956521739131e-05,
+      "loss": 0.4408,
       "step": 550
     },
     {
       "epoch": 6.09,
-      "learning_rate": 7.891304347826088e-05,
-      "loss": 0.489,
+      "learning_rate": 7.869565217391305e-05,
+      "loss": 0.5748,
       "step": 560
     },
     {
       "epoch": 6.2,
-      "learning_rate": 7.673913043478262e-05,
-      "loss": 0.4531,
+      "learning_rate": 7.652173913043479e-05,
+      "loss": 0.4865,
       "step": 570
     },
     {
       "epoch": 6.3,
-      "learning_rate": 7.456521739130435e-05,
-      "loss": 0.2837,
+      "learning_rate": 7.434782608695653e-05,
+      "loss": 0.2866,
       "step": 580
     },
     {
       "epoch": 6.41,
-      "learning_rate": 7.23913043478261e-05,
-      "loss": 0.37,
+      "learning_rate": 7.217391304347827e-05,
+      "loss": 0.3527,
       "step": 590
     },
     {
       "epoch": 6.52,
-      "learning_rate": 7.021739130434783e-05,
-      "loss": 0.3427,
+      "learning_rate": 7e-05,
+      "loss": 0.3103,
       "step": 600
     },
     {
       "epoch": 6.52,
-      "eval_accuracy": 0.95,
-      "eval_loss": 0.14258505403995514,
-      "eval_runtime": 1.9248,
-      "eval_samples_per_second": 135.078,
-      "eval_steps_per_second": 17.145,
+      "eval_accuracy": 0.9615384615384616,
+      "eval_loss": 0.15010055899620056,
+      "eval_runtime": 1.8802,
+      "eval_samples_per_second": 138.282,
+      "eval_steps_per_second": 17.551,
       "step": 600
     },
     {
       "epoch": 6.63,
-      "learning_rate": 6.804347826086956e-05,
-      "loss": 0.3713,
+      "learning_rate": 6.782608695652173e-05,
+      "loss": 0.4045,
       "step": 610
     },
     {
       "epoch": 6.74,
-      "learning_rate": 6.58695652173913e-05,
-      "loss": 0.2211,
+      "learning_rate": 6.565217391304349e-05,
+      "loss": 0.2758,
       "step": 620
     },
     {
       "epoch": 6.85,
-      "learning_rate": 6.369565217391305e-05,
-      "loss": 0.2769,
+      "learning_rate": 6.347826086956523e-05,
+      "loss": 0.3185,
       "step": 630
     },
     {
       "epoch": 6.96,
-      "learning_rate": 6.152173913043479e-05,
-      "loss": 0.2778,
+      "learning_rate": 6.130434782608696e-05,
+      "loss": 0.3133,
       "step": 640
     },
     {
       "epoch": 7.07,
-      "learning_rate": 5.934782608695652e-05,
-      "loss": 0.3071,
+      "learning_rate": 5.9130434782608704e-05,
+      "loss": 0.2535,
       "step": 650
     },
     {
       "epoch": 7.17,
-      "learning_rate": 5.717391304347827e-05,
-      "loss": 0.2988,
+      "learning_rate": 5.695652173913044e-05,
+      "loss": 0.2885,
       "step": 660
     },
     {
       "epoch": 7.28,
-      "learning_rate": 5.500000000000001e-05,
-      "loss": 0.2773,
+      "learning_rate": 5.478260869565217e-05,
+      "loss": 0.3363,
       "step": 670
     },
     {
       "epoch": 7.39,
-      "learning_rate": 5.2826086956521735e-05,
-      "loss": 0.2628,
+      "learning_rate": 5.260869565217391e-05,
+      "loss": 0.3127,
       "step": 680
     },
     {
       "epoch": 7.5,
-      "learning_rate": 5.0652173913043474e-05,
-      "loss": 0.4372,
+      "learning_rate": 5.0434782608695655e-05,
+      "loss": 0.3392,
       "step": 690
     },
     {
       "epoch": 7.61,
-      "learning_rate": 4.847826086956522e-05,
-      "loss": 0.2939,
+      "learning_rate": 4.8260869565217394e-05,
+      "loss": 0.3637,
       "step": 700
     },
     {
       "epoch": 7.61,
-      "eval_accuracy": 0.9730769230769231,
-      "eval_loss": 0.11737523972988129,
-      "eval_runtime": 1.9862,
-      "eval_samples_per_second": 130.905,
-      "eval_steps_per_second": 16.615,
+      "eval_accuracy": 0.9769230769230769,
+      "eval_loss": 0.13189926743507385,
+      "eval_runtime": 2.2676,
+      "eval_samples_per_second": 114.659,
+      "eval_steps_per_second": 14.553,
       "step": 700
     },
     {
       "epoch": 7.72,
-      "learning_rate": 4.630434782608696e-05,
-      "loss": 0.3101,
+      "learning_rate": 4.608695652173913e-05,
+      "loss": 0.3521,
       "step": 710
     },
     {
       "epoch": 7.83,
-      "learning_rate": 4.41304347826087e-05,
-      "loss": 0.1584,
+      "learning_rate": 4.391304347826087e-05,
+      "loss": 0.2299,
       "step": 720
     },
     {
       "epoch": 7.93,
-      "learning_rate": 4.195652173913044e-05,
-      "loss": 0.3143,
+      "learning_rate": 4.1739130434782605e-05,
+      "loss": 0.3716,
       "step": 730
     },
     {
       "epoch": 8.04,
-      "learning_rate": 3.978260869565217e-05,
-      "loss": 0.3705,
+      "learning_rate": 3.956521739130435e-05,
+      "loss": 0.3325,
       "step": 740
     },
     {
       "epoch": 8.15,
-      "learning_rate": 3.7608695652173917e-05,
-      "loss": 0.3003,
+      "learning_rate": 3.739130434782609e-05,
+      "loss": 0.2256,
       "step": 750
     },
     {
       "epoch": 8.26,
-      "learning_rate": 3.5434782608695656e-05,
-      "loss": 0.3864,
+      "learning_rate": 3.521739130434783e-05,
+      "loss": 0.4014,
       "step": 760
     },
     {
       "epoch": 8.37,
-      "learning_rate": 3.3260869565217395e-05,
-      "loss": 0.1961,
+      "learning_rate": 3.304347826086956e-05,
+      "loss": 0.2468,
       "step": 770
     },
     {
       "epoch": 8.48,
-      "learning_rate": 3.108695652173913e-05,
-      "loss": 0.2859,
+      "learning_rate": 3.086956521739131e-05,
+      "loss": 0.3243,
       "step": 780
     },
     {
       "epoch": 8.59,
-      "learning_rate": 2.891304347826087e-05,
-      "loss": 0.3521,
+      "learning_rate": 2.8695652173913044e-05,
+      "loss": 0.298,
       "step": 790
     },
     {
       "epoch": 8.7,
-      "learning_rate": 2.673913043478261e-05,
-      "loss": 0.1605,
+      "learning_rate": 2.6521739130434787e-05,
+      "loss": 0.1941,
       "step": 800
     },
     {
       "epoch": 8.7,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.06404715776443481,
-      "eval_runtime": 1.8828,
-      "eval_samples_per_second": 138.091,
-      "eval_steps_per_second": 17.527,
+      "eval_accuracy": 0.9769230769230769,
+      "eval_loss": 0.08613809198141098,
+      "eval_runtime": 1.8262,
+      "eval_samples_per_second": 142.375,
+      "eval_steps_per_second": 18.071,
       "step": 800
     },
     {
       "epoch": 8.8,
-      "learning_rate": 2.456521739130435e-05,
-      "loss": 0.1828,
+      "learning_rate": 2.4347826086956523e-05,
+      "loss": 0.1929,
       "step": 810
     },
     {
       "epoch": 8.91,
-      "learning_rate": 2.239130434782609e-05,
-      "loss": 0.1448,
+      "learning_rate": 2.2173913043478262e-05,
+      "loss": 0.1907,
       "step": 820
     },
     {
       "epoch": 9.02,
-      "learning_rate": 2.0217391304347828e-05,
-      "loss": 0.2442,
+      "learning_rate": 2e-05,
+      "loss": 0.2992,
       "step": 830
     },
     {
       "epoch": 9.13,
-      "learning_rate": 1.8043478260869567e-05,
-      "loss": 0.2182,
+      "learning_rate": 1.782608695652174e-05,
+      "loss": 0.2516,
       "step": 840
     },
     {
       "epoch": 9.24,
-      "learning_rate": 1.5869565217391306e-05,
-      "loss": 0.1857,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 0.2113,
       "step": 850
     },
     {
       "epoch": 9.35,
-      "learning_rate": 1.3695652173913042e-05,
-      "loss": 0.123,
+      "learning_rate": 1.3478260869565218e-05,
+      "loss": 0.1503,
       "step": 860
     },
     {
       "epoch": 9.46,
-      "learning_rate": 1.1521739130434783e-05,
-      "loss": 0.1286,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 0.1664,
       "step": 870
     },
     {
       "epoch": 9.57,
-      "learning_rate": 9.347826086956523e-06,
-      "loss": 0.3302,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 0.398,
       "step": 880
     },
     {
       "epoch": 9.67,
-      "learning_rate": 7.173913043478261e-06,
-      "loss": 0.3513,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.3558,
       "step": 890
     },
     {
       "epoch": 9.78,
-      "learning_rate": 5e-06,
-      "loss": 0.1865,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 0.1663,
       "step": 900
     },
     {
       "epoch": 9.78,
-      "eval_accuracy": 0.9807692307692307,
-      "eval_loss": 0.07018034905195236,
-      "eval_runtime": 2.3303,
-      "eval_samples_per_second": 111.572,
-      "eval_steps_per_second": 14.161,
+      "eval_accuracy": 0.9769230769230769,
+      "eval_loss": 0.08665793389081955,
+      "eval_runtime": 2.6289,
+      "eval_samples_per_second": 98.9,
+      "eval_steps_per_second": 12.553,
       "step": 900
     },
     {
       "epoch": 9.89,
-      "learning_rate": 2.8260869565217393e-06,
-      "loss": 0.1316,
+      "learning_rate": 2.608695652173913e-06,
+      "loss": 0.1478,
       "step": 910
     },
     {
       "epoch": 10.0,
-      "learning_rate": 6.521739130434782e-07,
-      "loss": 0.1979,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 0.2479,
       "step": 920
     },
     {
       "epoch": 10.0,
       "step": 920,
       "total_flos": 3.651306173035315e+17,
-      "train_loss": 0.6966753153697304,
-      "train_runtime": 217.2288,
-      "train_samples_per_second": 67.579,
-      "train_steps_per_second": 4.235
+      "train_loss": 0.7250858170830685,
+      "train_runtime": 208.4502,
+      "train_samples_per_second": 70.424,
+      "train_steps_per_second": 4.414
     }
   ],
   "logging_steps": 10,