End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9612,
-    "eval_loss": 0.12175925076007843,
-    "eval_runtime": 32.9858,
-    "eval_samples_per_second": 151.581,
-    "eval_steps_per_second": 1.213,
+    "eval_accuracy": 0.9592,
+    "eval_loss": 0.1276138424873352,
+    "eval_runtime": 27.4656,
+    "eval_samples_per_second": 182.046,
+    "eval_steps_per_second": 1.456,
     "total_flos": 3.35630861217792e+18,
-    "train_loss": 0.6934980005025864,
-    "train_runtime": 2145.9759,
-    "train_samples_per_second": 62.908,
-    "train_steps_per_second": 0.123
+    "train_loss": 0.7078906785358082,
+    "train_runtime": 1898.5865,
+    "train_samples_per_second": 71.106,
+    "train_steps_per_second": 0.139
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9612,
-    "eval_loss": 0.12175925076007843,
-    "eval_runtime": 32.9858,
-    "eval_samples_per_second": 151.581,
-    "eval_steps_per_second": 1.213
+    "eval_accuracy": 0.9592,
+    "eval_loss": 0.1276138424873352,
+    "eval_runtime": 27.4656,
+    "eval_samples_per_second": 182.046,
+    "eval_steps_per_second": 1.456
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
     "total_flos": 3.35630861217792e+18,
-    "train_loss": 0.6934980005025864,
-    "train_runtime": 2145.9759,
-    "train_samples_per_second": 62.908,
-    "train_steps_per_second": 0.123
+    "train_loss": 0.7078906785358082,
+    "train_runtime": 1898.5865,
+    "train_samples_per_second": 71.106,
+    "train_steps_per_second": 0.139
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.9612,
+  "best_metric": 0.9592,
   "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-cifar10/checkpoint-264",
   "epoch": 3.0,
   "eval_steps": 500,
@@ -11,194 +11,194 @@
     {
       "epoch": 0.11,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 2.3155,
+      "loss": 2.3771,
       "step": 10
     },
     {
       "epoch": 0.23,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 2.0565,
+      "loss": 2.1143,
       "step": 20
     },
     {
       "epoch": 0.34,
       "learning_rate": 4.936708860759494e-05,
-      "loss": 1.4845,
+      "loss": 1.5311,
       "step": 30
     },
     {
       "epoch": 0.45,
       "learning_rate": 4.725738396624473e-05,
-      "loss": 1.0076,
+      "loss": 1.0618,
       "step": 40
     },
     {
       "epoch": 0.57,
       "learning_rate": 4.5147679324894514e-05,
-      "loss": 0.8139,
+      "loss": 0.8814,
       "step": 50
     },
     {
       "epoch": 0.68,
       "learning_rate": 4.3037974683544305e-05,
-      "loss": 0.7133,
+      "loss": 0.7527,
       "step": 60
     },
     {
       "epoch": 0.8,
       "learning_rate": 4.0928270042194096e-05,
-      "loss": 0.6384,
+      "loss": 0.6469,
       "step": 70
     },
     {
       "epoch": 0.91,
       "learning_rate": 3.881856540084388e-05,
-      "loss": 0.6141,
+      "loss": 0.6304,
       "step": 80
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.931,
-      "eval_loss": 0.20727095007896423,
-      "eval_runtime": 32.7794,
-      "eval_samples_per_second": 152.535,
-      "eval_steps_per_second": 1.22,
+      "eval_accuracy": 0.93,
+      "eval_loss": 0.21966831386089325,
+      "eval_runtime": 27.3393,
+      "eval_samples_per_second": 182.887,
+      "eval_steps_per_second": 1.463,
       "step": 88
     },
     {
       "epoch": 1.02,
       "learning_rate": 3.670886075949367e-05,
-      "loss": 0.5816,
+      "loss": 0.5971,
       "step": 90
     },
     {
       "epoch": 1.14,
       "learning_rate": 3.459915611814346e-05,
-      "loss": 0.5254,
+      "loss": 0.5672,
       "step": 100
     },
     {
       "epoch": 1.25,
       "learning_rate": 3.248945147679325e-05,
-      "loss": 0.5173,
+      "loss": 0.5286,
       "step": 110
     },
     {
       "epoch": 1.36,
       "learning_rate": 3.0379746835443042e-05,
-      "loss": 0.5545,
+      "loss": 0.5397,
       "step": 120
     },
     {
       "epoch": 1.48,
       "learning_rate": 2.8270042194092826e-05,
-      "loss": 0.5089,
+      "loss": 0.5127,
       "step": 130
     },
     {
       "epoch": 1.59,
       "learning_rate": 2.616033755274262e-05,
-      "loss": 0.4906,
+      "loss": 0.4767,
       "step": 140
     },
     {
       "epoch": 1.7,
       "learning_rate": 2.4050632911392405e-05,
-      "loss": 0.4969,
+      "loss": 0.4775,
       "step": 150
     },
     {
       "epoch": 1.82,
       "learning_rate": 2.1940928270042196e-05,
-      "loss": 0.471,
+      "loss": 0.4686,
       "step": 160
     },
     {
       "epoch": 1.93,
       "learning_rate": 1.9831223628691984e-05,
-      "loss": 0.4767,
+      "loss": 0.4802,
       "step": 170
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.959,
-      "eval_loss": 0.13500115275382996,
-      "eval_runtime": 33.1717,
-      "eval_samples_per_second": 150.731,
-      "eval_steps_per_second": 1.206,
+      "eval_accuracy": 0.9558,
+      "eval_loss": 0.14049997925758362,
+      "eval_runtime": 27.3436,
+      "eval_samples_per_second": 182.858,
+      "eval_steps_per_second": 1.463,
       "step": 176
     },
     {
       "epoch": 2.05,
       "learning_rate": 1.7721518987341772e-05,
-      "loss": 0.424,
+      "loss": 0.4383,
       "step": 180
     },
     {
       "epoch": 2.16,
       "learning_rate": 1.5611814345991563e-05,
-      "loss": 0.4367,
+      "loss": 0.4394,
       "step": 190
     },
     {
       "epoch": 2.27,
       "learning_rate": 1.3502109704641349e-05,
-      "loss": 0.4228,
+      "loss": 0.4426,
       "step": 200
     },
     {
       "epoch": 2.39,
       "learning_rate": 1.139240506329114e-05,
-      "loss": 0.4556,
+      "loss": 0.444,
       "step": 210
     },
     {
       "epoch": 2.5,
       "learning_rate": 9.28270042194093e-06,
-      "loss": 0.4401,
+      "loss": 0.4171,
       "step": 220
     },
     {
       "epoch": 2.61,
       "learning_rate": 7.1729957805907175e-06,
-      "loss": 0.445,
+      "loss": 0.4329,
       "step": 230
     },
     {
       "epoch": 2.73,
       "learning_rate": 5.063291139240506e-06,
-      "loss": 0.4204,
+      "loss": 0.4367,
       "step": 240
     },
     {
       "epoch": 2.84,
       "learning_rate": 2.9535864978902956e-06,
-      "loss": 0.423,
+      "loss": 0.4151,
       "step": 250
     },
     {
       "epoch": 2.95,
       "learning_rate": 8.438818565400843e-07,
-      "loss": 0.4153,
+      "loss": 0.4114,
       "step": 260
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9612,
-      "eval_loss": 0.12175925076007843,
-      "eval_runtime": 32.5962,
-      "eval_samples_per_second": 153.392,
-      "eval_steps_per_second": 1.227,
+      "eval_accuracy": 0.9592,
+      "eval_loss": 0.1276138424873352,
+      "eval_runtime": 27.1749,
+      "eval_samples_per_second": 183.993,
+      "eval_steps_per_second": 1.472,
       "step": 264
     },
     {
       "epoch": 3.0,
       "step": 264,
       "total_flos": 3.35630861217792e+18,
-      "train_loss": 0.6934980005025864,
-      "train_runtime": 2145.9759,
-      "train_samples_per_second": 62.908,
-      "train_steps_per_second": 0.123
+      "train_loss": 0.7078906785358082,
+      "train_runtime": 1898.5865,
+      "train_samples_per_second": 71.106,
+      "train_steps_per_second": 0.139
     }
   ],
   "logging_steps": 10,