End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 9.9,
-    "eval_accuracy": 0.9777131782945736,
-    "eval_loss": 0.0747998058795929,
-    "eval_runtime": 6.6002,
-    "eval_samples_per_second": 156.358,
-    "eval_steps_per_second": 5.0,
-    "total_flos": 2.2837931240094351e+18,
-    "train_loss": 0.11106966810507907,
-    "train_runtime": 1435.0993,
-    "train_samples_per_second": 64.671,
-    "train_steps_per_second": 0.502
+    "epoch": 4.95,
+    "eval_accuracy": 0.9738372093023255,
+    "eval_loss": 0.10162452608346939,
+    "eval_runtime": 6.5939,
+    "eval_samples_per_second": 156.509,
+    "eval_steps_per_second": 5.005,
+    "total_flos": 1.1422818298339983e+18,
+    "train_loss": 0.08449313590923944,
+    "train_runtime": 710.4711,
+    "train_samples_per_second": 65.316,
+    "train_steps_per_second": 0.507
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.9,
-    "eval_accuracy": 0.9777131782945736,
-    "eval_loss": 0.0747998058795929,
-    "eval_runtime": 6.6002,
-    "eval_samples_per_second": 156.358,
-    "eval_steps_per_second": 5.0
+    "epoch": 4.95,
+    "eval_accuracy": 0.9738372093023255,
+    "eval_loss": 0.10162452608346939,
+    "eval_runtime": 6.5939,
+    "eval_samples_per_second": 156.509,
+    "eval_steps_per_second": 5.005
 }

runs/Dec15_11-18-43_666f530b4c9c/events.out.tfevents.1702639864.666f530b4c9c.589.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:739ee0253a33996c15a8032ef95137aaca4b82a4837ba8c4f69dd3214592bb6c
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.9,
-    "total_flos": 2.2837931240094351e+18,
-    "train_loss": 0.11106966810507907,
-    "train_runtime": 1435.0993,
-    "train_samples_per_second": 64.671,
-    "train_steps_per_second": 0.502
+    "epoch": 4.95,
+    "total_flos": 1.1422818298339983e+18,
+    "train_loss": 0.08449313590923944,
+    "train_runtime": 710.4711,
+    "train_samples_per_second": 65.316,
+    "train_steps_per_second": 0.507
 }

trainer_state.json

@@ -1,982 +1,505 @@
 {
-  "best_metric": 0.9777131782945736,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-blank_img/checkpoint-291",
-  "epoch": 9.896907216494846,
+  "best_metric": 0.9738372093023255,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-blank_img/checkpoint-360",
+  "epoch": 4.948453608247423,
   "eval_steps": 500,
-  "global_step": 720,
+  "global_step": 360,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.07,
-      "learning_rate": 3.4722222222222224e-06,
-      "loss": 0.6582,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.0881,
       "step": 5
     },
     {
       "epoch": 0.14,
-      "learning_rate": 6.944444444444445e-06,
-      "loss": 0.6212,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.0624,
       "step": 10
     },
     {
       "epoch": 0.21,
-      "learning_rate": 1.0416666666666668e-05,
-      "loss": 0.513,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.063,
       "step": 15
     },
     {
       "epoch": 0.27,
-      "learning_rate": 1.388888888888889e-05,
-      "loss": 0.3706,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.0901,
       "step": 20
     },
     {
       "epoch": 0.34,
-      "learning_rate": 1.736111111111111e-05,
-      "loss": 0.2605,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 0.0872,
       "step": 25
     },
     {
       "epoch": 0.41,
-      "learning_rate": 2.0833333333333336e-05,
-      "loss": 0.2189,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.1016,
       "step": 30
     },
     {
       "epoch": 0.48,
-      "learning_rate": 2.4305555555555558e-05,
-      "loss": 0.1749,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 0.0905,
       "step": 35
     },
     {
       "epoch": 0.55,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 0.1763,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 0.1495,
       "step": 40
     },
     {
       "epoch": 0.62,
-      "learning_rate": 3.125e-05,
-      "loss": 0.136,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 0.0949,
       "step": 45
     },
     {
       "epoch": 0.69,
-      "learning_rate": 3.472222222222222e-05,
-      "loss": 0.1538,
+      "learning_rate": 4.783950617283951e-05,
+      "loss": 0.1178,
       "step": 50
     },
     {
       "epoch": 0.76,
-      "learning_rate": 3.8194444444444444e-05,
-      "loss": 0.1284,
+      "learning_rate": 4.70679012345679e-05,
+      "loss": 0.08,
       "step": 55
     },
     {
       "epoch": 0.82,
-      "learning_rate": 4.166666666666667e-05,
-      "loss": 0.1346,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.0919,
       "step": 60
     },
     {
       "epoch": 0.89,
-      "learning_rate": 4.5138888888888894e-05,
-      "loss": 0.1098,
+      "learning_rate": 4.5524691358024696e-05,
+      "loss": 0.0773,
       "step": 65
     },
     {
       "epoch": 0.96,
-      "learning_rate": 4.8611111111111115e-05,
-      "loss": 0.0898,
+      "learning_rate": 4.4753086419753084e-05,
+      "loss": 0.0502,
       "step": 70
     },
     {
       "epoch": 0.99,
-      "eval_accuracy": 0.9428294573643411,
-      "eval_loss": 0.1245269775390625,
-      "eval_runtime": 7.009,
-      "eval_samples_per_second": 147.239,
-      "eval_steps_per_second": 4.708,
+      "eval_accuracy": 0.9651162790697675,
+      "eval_loss": 0.12999847531318665,
+      "eval_runtime": 6.6038,
+      "eval_samples_per_second": 156.273,
+      "eval_steps_per_second": 4.997,
       "step": 72
     },
     {
       "epoch": 1.03,
-      "learning_rate": 4.976851851851852e-05,
-      "loss": 0.1117,
+      "learning_rate": 4.3981481481481486e-05,
+      "loss": 0.0838,
       "step": 75
     },
     {
       "epoch": 1.1,
-      "learning_rate": 4.938271604938271e-05,
-      "loss": 0.0963,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 0.067,
       "step": 80
     },
     {
       "epoch": 1.17,
-      "learning_rate": 4.899691358024692e-05,
-      "loss": 0.1203,
+      "learning_rate": 4.243827160493827e-05,
+      "loss": 0.0932,
       "step": 85
     },
     {
       "epoch": 1.24,
-      "learning_rate": 4.8611111111111115e-05,
-      "loss": 0.126,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.1037,
       "step": 90
     },
     {
       "epoch": 1.31,
-      "learning_rate": 4.8225308641975306e-05,
-      "loss": 0.1428,
+      "learning_rate": 4.089506172839506e-05,
+      "loss": 0.1057,
       "step": 95
     },
     {
       "epoch": 1.37,
-      "learning_rate": 4.783950617283951e-05,
-      "loss": 0.1215,
+      "learning_rate": 4.012345679012346e-05,
+      "loss": 0.0774,
       "step": 100
     },
     {
       "epoch": 1.44,
-      "learning_rate": 4.745370370370371e-05,
-      "loss": 0.1363,
+      "learning_rate": 3.935185185185186e-05,
+      "loss": 0.1107,
       "step": 105
     },
     {
       "epoch": 1.51,
-      "learning_rate": 4.70679012345679e-05,
-      "loss": 0.1432,
+      "learning_rate": 3.8580246913580246e-05,
+      "loss": 0.1109,
       "step": 110
     },
     {
       "epoch": 1.58,
-      "learning_rate": 4.66820987654321e-05,
-      "loss": 0.1444,
+      "learning_rate": 3.780864197530865e-05,
+      "loss": 0.1078,
       "step": 115
     },
     {
       "epoch": 1.65,
-      "learning_rate": 4.62962962962963e-05,
-      "loss": 0.1221,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.0935,
       "step": 120
     },
     {
       "epoch": 1.72,
-      "learning_rate": 4.591049382716049e-05,
-      "loss": 0.1159,
+      "learning_rate": 3.626543209876543e-05,
+      "loss": 0.0741,
       "step": 125
     },
     {
       "epoch": 1.79,
-      "learning_rate": 4.5524691358024696e-05,
-      "loss": 0.1311,
+      "learning_rate": 3.5493827160493834e-05,
+      "loss": 0.1031,
       "step": 130
     },
     {
       "epoch": 1.86,
-      "learning_rate": 4.5138888888888894e-05,
-      "loss": 0.0965,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 0.0638,
       "step": 135
     },
     {
       "epoch": 1.92,
-      "learning_rate": 4.4753086419753084e-05,
-      "loss": 0.1129,
+      "learning_rate": 3.395061728395062e-05,
+      "loss": 0.0915,
       "step": 140
     },
     {
       "epoch": 1.99,
-      "learning_rate": 4.436728395061729e-05,
-      "loss": 0.152,
+      "learning_rate": 3.317901234567901e-05,
+      "loss": 0.1107,
       "step": 145
     },
     {
       "epoch": 1.99,
-      "eval_accuracy": 0.9748062015503876,
-      "eval_loss": 0.08107755333185196,
-      "eval_runtime": 7.1051,
-      "eval_samples_per_second": 145.247,
-      "eval_steps_per_second": 4.645,
+      "eval_accuracy": 0.9728682170542635,
+      "eval_loss": 0.10234300047159195,
+      "eval_runtime": 7.16,
+      "eval_samples_per_second": 144.135,
+      "eval_steps_per_second": 4.609,
       "step": 145
     },
     {
       "epoch": 2.06,
-      "learning_rate": 4.3981481481481486e-05,
-      "loss": 0.1105,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 0.0784,
       "step": 150
     },
     {
       "epoch": 2.13,
-      "learning_rate": 4.359567901234568e-05,
-      "loss": 0.0788,
+      "learning_rate": 3.16358024691358e-05,
+      "loss": 0.0672,
       "step": 155
     },
     {
       "epoch": 2.2,
-      "learning_rate": 4.3209876543209875e-05,
-      "loss": 0.0988,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.0598,
       "step": 160
     },
     {
       "epoch": 2.27,
-      "learning_rate": 4.282407407407408e-05,
-      "loss": 0.0963,
+      "learning_rate": 3.0092592592592593e-05,
+      "loss": 0.0641,
       "step": 165
     },
     {
       "epoch": 2.34,
-      "learning_rate": 4.243827160493827e-05,
-      "loss": 0.1201,
+      "learning_rate": 2.9320987654320992e-05,
+      "loss": 0.0805,
       "step": 170
     },
     {
       "epoch": 2.41,
-      "learning_rate": 4.205246913580247e-05,
-      "loss": 0.1001,
+      "learning_rate": 2.8549382716049384e-05,
+      "loss": 0.0864,
       "step": 175
     },
     {
       "epoch": 2.47,
-      "learning_rate": 4.166666666666667e-05,
-      "loss": 0.1118,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.0954,
       "step": 180
     },
     {
       "epoch": 2.54,
-      "learning_rate": 4.128086419753087e-05,
-      "loss": 0.0948,
+      "learning_rate": 2.700617283950617e-05,
+      "loss": 0.0717,
       "step": 185
     },
     {
       "epoch": 2.61,
-      "learning_rate": 4.089506172839506e-05,
-      "loss": 0.0941,
+      "learning_rate": 2.623456790123457e-05,
+      "loss": 0.0634,
       "step": 190
     },
     {
       "epoch": 2.68,
-      "learning_rate": 4.0509259259259265e-05,
-      "loss": 0.1224,
+      "learning_rate": 2.5462962962962965e-05,
+      "loss": 0.1082,
       "step": 195
     },
     {
       "epoch": 2.75,
-      "learning_rate": 4.012345679012346e-05,
-      "loss": 0.0919,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.0748,
       "step": 200
     },
     {
       "epoch": 2.82,
-      "learning_rate": 3.973765432098765e-05,
-      "loss": 0.1715,
+      "learning_rate": 2.3919753086419755e-05,
+      "loss": 0.1221,
       "step": 205
     },
     {
       "epoch": 2.89,
-      "learning_rate": 3.935185185185186e-05,
-      "loss": 0.1294,
+      "learning_rate": 2.314814814814815e-05,
+      "loss": 0.0757,
       "step": 210
     },
     {
       "epoch": 2.96,
-      "learning_rate": 3.8966049382716055e-05,
-      "loss": 0.1235,
+      "learning_rate": 2.2376543209876542e-05,
+      "loss": 0.0917,
       "step": 215
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9699612403100775,
-      "eval_loss": 0.09580326825380325,
-      "eval_runtime": 6.6469,
-      "eval_samples_per_second": 155.26,
-      "eval_steps_per_second": 4.965,
+      "eval_accuracy": 0.9651162790697675,
+      "eval_loss": 0.12774048745632172,
+      "eval_runtime": 6.6558,
+      "eval_samples_per_second": 155.053,
+      "eval_steps_per_second": 4.958,
       "step": 218
     },
     {
       "epoch": 3.02,
-      "learning_rate": 3.8580246913580246e-05,
-      "loss": 0.0777,
+      "learning_rate": 2.1604938271604937e-05,
+      "loss": 0.0553,
       "step": 220
     },
     {
       "epoch": 3.09,
-      "learning_rate": 3.8194444444444444e-05,
-      "loss": 0.1301,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.0942,
       "step": 225
     },
     {
       "epoch": 3.16,
-      "learning_rate": 3.780864197530865e-05,
-      "loss": 0.1285,
+      "learning_rate": 2.006172839506173e-05,
+      "loss": 0.0834,
       "step": 230
     },
     {
       "epoch": 3.23,
-      "learning_rate": 3.742283950617284e-05,
-      "loss": 0.0848,
+      "learning_rate": 1.9290123456790123e-05,
+      "loss": 0.057,
       "step": 235
     },
     {
       "epoch": 3.3,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.1008,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.0773,
       "step": 240
     },
     {
       "epoch": 3.37,
-      "learning_rate": 3.665123456790124e-05,
-      "loss": 0.0919,
+      "learning_rate": 1.7746913580246917e-05,
+      "loss": 0.0617,
       "step": 245
     },
     {
       "epoch": 3.44,
-      "learning_rate": 3.626543209876543e-05,
-      "loss": 0.0997,
+      "learning_rate": 1.697530864197531e-05,
+      "loss": 0.0785,
       "step": 250
     },
     {
       "epoch": 3.51,
-      "learning_rate": 3.587962962962963e-05,
-      "loss": 0.0777,
+      "learning_rate": 1.6203703703703704e-05,
+      "loss": 0.057,
       "step": 255
     },
     {
       "epoch": 3.57,
-      "learning_rate": 3.5493827160493834e-05,
-      "loss": 0.1069,
+      "learning_rate": 1.54320987654321e-05,
+      "loss": 0.0746,
       "step": 260
     },
     {
       "epoch": 3.64,
-      "learning_rate": 3.5108024691358025e-05,
-      "loss": 0.1284,
+      "learning_rate": 1.4660493827160496e-05,
+      "loss": 0.1117,
       "step": 265
     },
     {
       "epoch": 3.71,
-      "learning_rate": 3.472222222222222e-05,
-      "loss": 0.1037,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.0923,
       "step": 270
     },
     {
       "epoch": 3.78,
-      "learning_rate": 3.4336419753086427e-05,
-      "loss": 0.0799,
+      "learning_rate": 1.3117283950617285e-05,
+      "loss": 0.0599,
       "step": 275
     },
     {
       "epoch": 3.85,
-      "learning_rate": 3.395061728395062e-05,
-      "loss": 0.1172,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 0.0895,
       "step": 280
     },
     {
       "epoch": 3.92,
-      "learning_rate": 3.3564814814814815e-05,
-      "loss": 0.1136,
+      "learning_rate": 1.1574074074074075e-05,
+      "loss": 0.0933,
       "step": 285
     },
     {
       "epoch": 3.99,
-      "learning_rate": 3.317901234567901e-05,
-      "loss": 0.1065,
+      "learning_rate": 1.0802469135802469e-05,
+      "loss": 0.1022,
       "step": 290
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9777131782945736,
-      "eval_loss": 0.0747998058795929,
-      "eval_runtime": 6.5479,
-      "eval_samples_per_second": 157.608,
-      "eval_steps_per_second": 5.04,
+      "eval_accuracy": 0.9718992248062015,
+      "eval_loss": 0.12579788267612457,
+      "eval_runtime": 7.0589,
+      "eval_samples_per_second": 146.199,
+      "eval_steps_per_second": 4.675,
       "step": 291
     },
     {
       "epoch": 4.05,
-      "learning_rate": 3.279320987654321e-05,
-      "loss": 0.0848,
+      "learning_rate": 1.0030864197530866e-05,
+      "loss": 0.0889,
       "step": 295
     },
     {
       "epoch": 4.12,
-      "learning_rate": 3.240740740740741e-05,
-      "loss": 0.1073,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.088,
       "step": 300
     },
     {
       "epoch": 4.19,
-      "learning_rate": 3.2021604938271605e-05,
-      "loss": 0.0821,
+      "learning_rate": 8.487654320987654e-06,
+      "loss": 0.0679,
       "step": 305
     },
     {
       "epoch": 4.26,
-      "learning_rate": 3.16358024691358e-05,
-      "loss": 0.0662,
+      "learning_rate": 7.71604938271605e-06,
+      "loss": 0.0632,
       "step": 310
     },
     {
       "epoch": 4.33,
-      "learning_rate": 3.125e-05,
-      "loss": 0.123,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.095,
       "step": 315
     },
     {
       "epoch": 4.4,
-      "learning_rate": 3.08641975308642e-05,
-      "loss": 0.1168,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 0.084,
       "step": 320
     },
     {
       "epoch": 4.47,
-      "learning_rate": 3.04783950617284e-05,
-      "loss": 0.0906,
+      "learning_rate": 5.401234567901234e-06,
+      "loss": 0.07,
       "step": 325
     },
     {
       "epoch": 4.54,
-      "learning_rate": 3.0092592592592593e-05,
-      "loss": 0.0801,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 0.0692,
       "step": 330
     },
     {
       "epoch": 4.6,
-      "learning_rate": 2.970679012345679e-05,
-      "loss": 0.0894,
+      "learning_rate": 3.858024691358025e-06,
+      "loss": 0.0575,
       "step": 335
     },
     {
       "epoch": 4.67,
-      "learning_rate": 2.9320987654320992e-05,
-      "loss": 0.0927,
+      "learning_rate": 3.0864197530864196e-06,
+      "loss": 0.0625,
       "step": 340
     },
     {
       "epoch": 4.74,
-      "learning_rate": 2.8935185185185186e-05,
-      "loss": 0.0803,
+      "learning_rate": 2.3148148148148148e-06,
+      "loss": 0.0715,
       "step": 345
     },
     {
       "epoch": 4.81,
-      "learning_rate": 2.8549382716049384e-05,
-      "loss": 0.0642,
+      "learning_rate": 1.5432098765432098e-06,
+      "loss": 0.0651,
       "step": 350
     },
     {
       "epoch": 4.88,
-      "learning_rate": 2.8163580246913578e-05,
-      "loss": 0.1422,
+      "learning_rate": 7.716049382716049e-07,
+      "loss": 0.1432,
       "step": 355
     },
     {
       "epoch": 4.95,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 0.1115,
+      "learning_rate": 0.0,
+      "loss": 0.0888,
       "step": 360
     },
     {
-      "epoch": 4.99,
-      "eval_accuracy": 0.9728682170542635,
-      "eval_loss": 0.09473349153995514,
-      "eval_runtime": 6.8174,
-      "eval_samples_per_second": 151.376,
-      "eval_steps_per_second": 4.841,
-      "step": 363
-    },
-    {
-      "epoch": 5.02,
-      "learning_rate": 2.7391975308641977e-05,
-      "loss": 0.0944,
-      "step": 365
-    },
-    {
-      "epoch": 5.09,
-      "learning_rate": 2.700617283950617e-05,
-      "loss": 0.0553,
-      "step": 370
-    },
-    {
-      "epoch": 5.15,
-      "learning_rate": 2.6620370370370372e-05,
-      "loss": 0.0994,
-      "step": 375
-    },
-    {
-      "epoch": 5.22,
-      "learning_rate": 2.623456790123457e-05,
-      "loss": 0.0802,
-      "step": 380
-    },
-    {
-      "epoch": 5.29,
-      "learning_rate": 2.5848765432098764e-05,
-      "loss": 0.1078,
-      "step": 385
-    },
-    {
-      "epoch": 5.36,
-      "learning_rate": 2.5462962962962965e-05,
-      "loss": 0.0746,
-      "step": 390
-    },
-    {
-      "epoch": 5.43,
-      "learning_rate": 2.5077160493827162e-05,
-      "loss": 0.0987,
-      "step": 395
-    },
-    {
-      "epoch": 5.5,
-      "learning_rate": 2.4691358024691357e-05,
-      "loss": 0.0885,
-      "step": 400
-    },
-    {
-      "epoch": 5.57,
-      "learning_rate": 2.4305555555555558e-05,
-      "loss": 0.0873,
-      "step": 405
-    },
-    {
-      "epoch": 5.64,
-      "learning_rate": 2.3919753086419755e-05,
-      "loss": 0.0859,
-      "step": 410
-    },
-    {
-      "epoch": 5.7,
-      "learning_rate": 2.353395061728395e-05,
-      "loss": 0.0871,
-      "step": 415
-    },
-    {
-      "epoch": 5.77,
-      "learning_rate": 2.314814814814815e-05,
-      "loss": 0.0847,
-      "step": 420
-    },
-    {
-      "epoch": 5.84,
-      "learning_rate": 2.2762345679012348e-05,
-      "loss": 0.1074,
-      "step": 425
-    },
-    {
-      "epoch": 5.91,
-      "learning_rate": 2.2376543209876542e-05,
-      "loss": 0.1017,
-      "step": 430
-    },
-    {
-      "epoch": 5.98,
-      "learning_rate": 2.1990740740740743e-05,
-      "loss": 0.0804,
-      "step": 435
-    },
-    {
-      "epoch": 5.99,
-      "eval_accuracy": 0.9757751937984496,
-      "eval_loss": 0.08881077915430069,
-      "eval_runtime": 7.2471,
-      "eval_samples_per_second": 142.401,
-      "eval_steps_per_second": 4.554,
-      "step": 436
-    },
-    {
-      "epoch": 6.05,
-      "learning_rate": 2.1604938271604937e-05,
-      "loss": 0.0592,
-      "step": 440
-    },
-    {
-      "epoch": 6.12,
-      "learning_rate": 2.1219135802469135e-05,
-      "loss": 0.1185,
-      "step": 445
-    },
-    {
-      "epoch": 6.19,
-      "learning_rate": 2.0833333333333336e-05,
-      "loss": 0.0868,
-      "step": 450
-    },
-    {
-      "epoch": 6.25,
-      "learning_rate": 2.044753086419753e-05,
-      "loss": 0.0883,
-      "step": 455
-    },
-    {
-      "epoch": 6.32,
-      "learning_rate": 2.006172839506173e-05,
-      "loss": 0.0835,
-      "step": 460
-    },
-    {
-      "epoch": 6.39,
-      "learning_rate": 1.967592592592593e-05,
-      "loss": 0.0781,
-      "step": 465
-    },
-    {
-      "epoch": 6.46,
-      "learning_rate": 1.9290123456790123e-05,
-      "loss": 0.1039,
-      "step": 470
-    },
-    {
-      "epoch": 6.53,
-      "learning_rate": 1.8904320987654324e-05,
-      "loss": 0.0823,
-      "step": 475
-    },
-    {
-      "epoch": 6.6,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.0842,
-      "step": 480
-    },
-    {
-      "epoch": 6.67,
-      "learning_rate": 1.8132716049382716e-05,
-      "loss": 0.0859,
-      "step": 485
-    },
-    {
-      "epoch": 6.74,
-      "learning_rate": 1.7746913580246917e-05,
-      "loss": 0.0751,
-      "step": 490
-    },
-    {
-      "epoch": 6.8,
-      "learning_rate": 1.736111111111111e-05,
-      "loss": 0.081,
-      "step": 495
-    },
-    {
-      "epoch": 6.87,
-      "learning_rate": 1.697530864197531e-05,
-      "loss": 0.1262,
-      "step": 500
-    },
-    {
-      "epoch": 6.94,
-      "learning_rate": 1.6589506172839506e-05,
-      "loss": 0.0722,
-      "step": 505
-    },
-    {
-      "epoch": 7.0,
-      "eval_accuracy": 0.9757751937984496,
-      "eval_loss": 0.08266593515872955,
-      "eval_runtime": 7.0856,
-      "eval_samples_per_second": 145.648,
-      "eval_steps_per_second": 4.657,
-      "step": 509
-    },
-    {
-      "epoch": 7.01,
-      "learning_rate": 1.6203703703703704e-05,
-      "loss": 0.1235,
-      "step": 510
-    },
-    {
-      "epoch": 7.08,
-      "learning_rate": 1.58179012345679e-05,
-      "loss": 0.0734,
-      "step": 515
-    },
-    {
-      "epoch": 7.15,
-      "learning_rate": 1.54320987654321e-05,
-      "loss": 0.0819,
-      "step": 520
-    },
-    {
-      "epoch": 7.22,
-      "learning_rate": 1.5046296296296297e-05,
-      "loss": 0.0916,
-      "step": 525
-    },
-    {
-      "epoch": 7.29,
-      "learning_rate": 1.4660493827160496e-05,
-      "loss": 0.0859,
-      "step": 530
-    },
-    {
-      "epoch": 7.35,
-      "learning_rate": 1.4274691358024692e-05,
-      "loss": 0.1009,
-      "step": 535
-    },
-    {
-      "epoch": 7.42,
-      "learning_rate": 1.388888888888889e-05,
-      "loss": 0.0996,
-      "step": 540
-    },
-    {
-      "epoch": 7.49,
-      "learning_rate": 1.3503086419753085e-05,
-      "loss": 0.1062,
-      "step": 545
-    },
-    {
-      "epoch": 7.56,
-      "learning_rate": 1.3117283950617285e-05,
-      "loss": 0.0759,
-      "step": 550
-    },
-    {
-      "epoch": 7.63,
-      "learning_rate": 1.2731481481481482e-05,
-      "loss": 0.0658,
-      "step": 555
-    },
-    {
-      "epoch": 7.7,
-      "learning_rate": 1.2345679012345678e-05,
-      "loss": 0.1081,
-      "step": 560
-    },
-    {
-      "epoch": 7.77,
-      "learning_rate": 1.1959876543209878e-05,
-      "loss": 0.0866,
-      "step": 565
-    },
-    {
-      "epoch": 7.84,
-      "learning_rate": 1.1574074074074075e-05,
-      "loss": 0.0783,
-      "step": 570
-    },
-    {
-      "epoch": 7.9,
-      "learning_rate": 1.1188271604938271e-05,
-      "loss": 0.0868,
-      "step": 575
-    },
-    {
-      "epoch": 7.97,
-      "learning_rate": 1.0802469135802469e-05,
-      "loss": 0.061,
-      "step": 580
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.9757751937984496,
-      "eval_loss": 0.0899147242307663,
-      "eval_runtime": 6.5821,
-      "eval_samples_per_second": 156.788,
-      "eval_steps_per_second": 5.014,
-      "step": 582
-    },
-    {
-      "epoch": 8.04,
-      "learning_rate": 1.0416666666666668e-05,
-      "loss": 0.053,
-      "step": 585
-    },
-    {
-      "epoch": 8.11,
-      "learning_rate": 1.0030864197530866e-05,
-      "loss": 0.0749,
-      "step": 590
-    },
-    {
-      "epoch": 8.18,
-      "learning_rate": 9.645061728395062e-06,
-      "loss": 0.0607,
-      "step": 595
-    },
-    {
-      "epoch": 8.25,
-      "learning_rate": 9.259259259259259e-06,
-      "loss": 0.0772,
-      "step": 600
-    },
-    {
-      "epoch": 8.32,
-      "learning_rate": 8.873456790123458e-06,
-      "loss": 0.055,
-      "step": 605
-    },
-    {
-      "epoch": 8.38,
-      "learning_rate": 8.487654320987654e-06,
-      "loss": 0.0647,
-      "step": 610
-    },
-    {
-      "epoch": 8.45,
-      "learning_rate": 8.101851851851852e-06,
-      "loss": 0.1082,
-      "step": 615
-    },
-    {
-      "epoch": 8.52,
-      "learning_rate": 7.71604938271605e-06,
-      "loss": 0.09,
-      "step": 620
-    },
-    {
-      "epoch": 8.59,
-      "learning_rate": 7.330246913580248e-06,
-      "loss": 0.0739,
-      "step": 625
-    },
-    {
-      "epoch": 8.66,
-      "learning_rate": 6.944444444444445e-06,
-      "loss": 0.0632,
-      "step": 630
-    },
-    {
-      "epoch": 8.73,
-      "learning_rate": 6.558641975308642e-06,
-      "loss": 0.0662,
-      "step": 635
-    },
-    {
-      "epoch": 8.8,
-      "learning_rate": 6.172839506172839e-06,
-      "loss": 0.0691,
-      "step": 640
-    },
-    {
-      "epoch": 8.87,
-      "learning_rate": 5.787037037037038e-06,
-      "loss": 0.0745,
-      "step": 645
-    },
-    {
-      "epoch": 8.93,
-      "learning_rate": 5.401234567901234e-06,
-      "loss": 0.0706,
-      "step": 650
-    },
-    {
-      "epoch": 8.99,
-      "eval_accuracy": 0.9757751937984496,
-      "eval_loss": 0.09157940745353699,
-      "eval_runtime": 6.6599,
-      "eval_samples_per_second": 154.957,
-      "eval_steps_per_second": 4.955,
-      "step": 654
-    },
-    {
-      "epoch": 9.0,
-      "learning_rate": 5.015432098765433e-06,
-      "loss": 0.0871,
-      "step": 655
-    },
-    {
-      "epoch": 9.07,
-      "learning_rate": 4.6296296296296296e-06,
-      "loss": 0.101,
-      "step": 660
-    },
-    {
-      "epoch": 9.14,
-      "learning_rate": 4.243827160493827e-06,
-      "loss": 0.0738,
-      "step": 665
-    },
-    {
-      "epoch": 9.21,
-      "learning_rate": 3.858024691358025e-06,
-      "loss": 0.0803,
-      "step": 670
-    },
-    {
-      "epoch": 9.28,
-      "learning_rate": 3.4722222222222224e-06,
-      "loss": 0.0787,
-      "step": 675
-    },
-    {
-      "epoch": 9.35,
-      "learning_rate": 3.0864197530864196e-06,
-      "loss": 0.0661,
-      "step": 680
-    },
-    {
-      "epoch": 9.42,
-      "learning_rate": 2.700617283950617e-06,
-      "loss": 0.055,
-      "step": 685
-    },
-    {
-      "epoch": 9.48,
-      "learning_rate": 2.3148148148148148e-06,
-      "loss": 0.0752,
-      "step": 690
-    },
-    {
-      "epoch": 9.55,
-      "learning_rate": 1.9290123456790124e-06,
-      "loss": 0.0933,
-      "step": 695
-    },
-    {
-      "epoch": 9.62,
-      "learning_rate": 1.5432098765432098e-06,
-      "loss": 0.1046,
-      "step": 700
-    },
-    {
-      "epoch": 9.69,
-      "learning_rate": 1.1574074074074074e-06,
-      "loss": 0.0467,
-      "step": 705
-    },
-    {
-      "epoch": 9.76,
-      "learning_rate": 7.716049382716049e-07,
-      "loss": 0.0739,
-      "step": 710
-    },
-    {
-      "epoch": 9.83,
-      "learning_rate": 3.8580246913580245e-07,
-      "loss": 0.068,
-      "step": 715
+      "epoch": 4.95,
+      "eval_accuracy": 0.9738372093023255,
+      "eval_loss": 0.10162452608346939,
+      "eval_runtime": 6.9642,
+      "eval_samples_per_second": 148.186,
+      "eval_steps_per_second": 4.739,
+      "step": 360
     },
     {
-      "epoch": 9.9,
-      "learning_rate": 0.0,
-      "loss": 0.0633,
-      "step": 720
-    },
-    {
-      "epoch": 9.9,
-      "eval_accuracy": 0.9757751937984496,
-      "eval_loss": 0.09366881847381592,
-      "eval_runtime": 6.8299,
-      "eval_samples_per_second": 151.1,
-      "eval_steps_per_second": 4.832,
-      "step": 720
-    },
-    {
-      "epoch": 9.9,
-      "step": 720,
-      "total_flos": 2.2837931240094351e+18,
-      "train_loss": 0.11106966810507907,
-      "train_runtime": 1435.0993,
-      "train_samples_per_second": 64.671,
-      "train_steps_per_second": 0.502
+      "epoch": 4.95,
+      "step": 360,
+      "total_flos": 1.1422818298339983e+18,
+      "train_loss": 0.08449313590923944,
+      "train_runtime": 710.4711,
+      "train_samples_per_second": 65.316,
+      "train_steps_per_second": 0.507
     }
   ],
   "logging_steps": 5,
-  "max_steps": 720,
-  "num_train_epochs": 10,
+  "max_steps": 360,
+  "num_train_epochs": 5,
   "save_steps": 500,
-  "total_flos": 2.2837931240094351e+18,
+  "total_flos": 1.1422818298339983e+18,
   "trial_name": null,
   "trial_params": null
 }