End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 2.0,
-    "eval_accuracy": 1.0,
-    "eval_loss": 0.008575703017413616,
-    "eval_runtime": 1.7087,
-    "eval_samples_per_second": 122.9,
-    "eval_steps_per_second": 4.097,
+    "eval_accuracy": 0.9904761904761905,
+    "eval_loss": 0.017159689217805862,
+    "eval_runtime": 1.7705,
+    "eval_samples_per_second": 118.612,
+    "eval_steps_per_second": 3.954,
     "total_flos": 9.395826371334144e+16,
-    "train_loss": 0.33677051266034447,
-    "train_runtime": 52.9479,
-    "train_samples_per_second": 178.477,
-    "train_steps_per_second": 1.416
+    "train_loss": 0.323590816060702,
+    "train_runtime": 55.2419,
+    "train_samples_per_second": 171.066,
+    "train_steps_per_second": 1.358
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.0,
-    "eval_accuracy": 1.0,
-    "eval_loss": 0.008575703017413616,
-    "eval_runtime": 1.7087,
-    "eval_samples_per_second": 122.9,
-    "eval_steps_per_second": 4.097
+    "eval_accuracy": 0.9904761904761905,
+    "eval_loss": 0.017159689217805862,
+    "eval_runtime": 1.7705,
+    "eval_samples_per_second": 118.612,
+    "eval_steps_per_second": 3.954
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.0,
     "total_flos": 9.395826371334144e+16,
-    "train_loss": 0.33677051266034447,
-    "train_runtime": 52.9479,
-    "train_samples_per_second": 178.477,
-    "train_steps_per_second": 1.416
+    "train_loss": 0.323590816060702,
+    "train_runtime": 55.2419,
+    "train_samples_per_second": 171.066,
+    "train_steps_per_second": 1.358
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 1.0,
+  "best_metric": 0.9904761904761905,
   "best_model_checkpoint": "/kaggle/working/swin-brain-modality-classification/checkpoint-30",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,51 +10,51 @@
   "log_history": [
     {
       "epoch": 0.6666666666666666,
-      "grad_norm": 5.258427619934082,
+      "grad_norm": 5.664389610290527,
       "learning_rate": 4.850746268656717e-05,
-      "loss": 0.8454,
+      "loss": 0.803,
       "step": 10
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9761904761904762,
-      "eval_loss": 0.11262203752994537,
-      "eval_runtime": 1.7281,
-      "eval_samples_per_second": 121.524,
-      "eval_steps_per_second": 4.051,
+      "eval_accuracy": 0.9619047619047619,
+      "eval_loss": 0.1342908889055252,
+      "eval_runtime": 1.7227,
+      "eval_samples_per_second": 121.902,
+      "eval_steps_per_second": 4.063,
       "step": 15
     },
     {
       "epoch": 1.3333333333333333,
-      "grad_norm": 3.9709157943725586,
+      "grad_norm": 3.640326976776123,
       "learning_rate": 4.104477611940299e-05,
-      "loss": 0.1186,
+      "loss": 0.1269,
       "step": 20
     },
     {
       "epoch": 2.0,
-      "grad_norm": 1.1410547494888306,
+      "grad_norm": 1.9675283432006836,
       "learning_rate": 3.358208955223881e-05,
-      "loss": 0.0463,
+      "loss": 0.0409,
       "step": 30
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 1.0,
-      "eval_loss": 0.008575703017413616,
-      "eval_runtime": 1.7615,
-      "eval_samples_per_second": 119.214,
-      "eval_steps_per_second": 3.974,
+      "eval_accuracy": 0.9904761904761905,
+      "eval_loss": 0.017159689217805862,
+      "eval_runtime": 1.722,
+      "eval_samples_per_second": 121.952,
+      "eval_steps_per_second": 4.065,
       "step": 30
     },
     {
       "epoch": 2.0,
       "step": 30,
       "total_flos": 9.395826371334144e+16,
-      "train_loss": 0.33677051266034447,
-      "train_runtime": 52.9479,
-      "train_samples_per_second": 178.477,
-      "train_steps_per_second": 1.416
+      "train_loss": 0.323590816060702,
+      "train_runtime": 55.2419,
+      "train_samples_per_second": 171.066,
+      "train_steps_per_second": 1.358
     }
   ],
   "logging_steps": 10,