Training in progress, epoch 1

all_results.json

@@ -1,8 +1,13 @@
 {
     "epoch": 8.0,
+    "eval_loss": 0.446413516998291,
+    "eval_roc_auc": 0.7681704260651628,
+    "eval_runtime": 0.4149,
+    "eval_samples_per_second": 96.41,
+    "eval_steps_per_second": 4.821,
     "total_flos": 7.078985663171789e+16,
-    "train_loss": 0.5772393196821213,
-    "train_runtime": 61.0545,
-    "train_samples_per_second": 46.647,
-    "train_steps_per_second": 0.393
+    "train_loss": 0.4116581827402115,
+    "train_runtime": 69.7528,
+    "train_samples_per_second": 40.83,
+    "train_steps_per_second": 0.344
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "eval_loss": 0.446413516998291,
+    "eval_roc_auc": 0.7681704260651628,
+    "eval_runtime": 0.4149,
+    "eval_samples_per_second": 96.41,
+    "eval_steps_per_second": 4.821
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cacf0d6edf3dc7b8e2f8686a95e060e12d60468ed93f4ee6899c2c548e5baa96
+oid sha256:3c1106773d6025eb34315bde144095028f612768318d1f1761ab93a0a934f1ec
 size 110394865

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 8.0,
     "total_flos": 7.078985663171789e+16,
-    "train_loss": 0.5772393196821213,
-    "train_runtime": 61.0545,
-    "train_samples_per_second": 46.647,
-    "train_steps_per_second": 0.393
+    "train_loss": 0.4116581827402115,
+    "train_runtime": 69.7528,
+    "train_samples_per_second": 40.83,
+    "train_steps_per_second": 0.344
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.7631578947368421,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetunedo/checkpoint-24",
+  "best_metric": 0.7681704260651628,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetunedo/checkpoint-12",
   "epoch": 8.0,
   "eval_steps": 500,
   "global_step": 24,
@@ -10,96 +10,96 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_loss": 0.7092229723930359,
-      "eval_roc_auc": 0.581453634085213,
-      "eval_runtime": 0.3335,
-      "eval_samples_per_second": 119.954,
-      "eval_steps_per_second": 5.998,
+      "eval_loss": 0.5081062316894531,
+      "eval_roc_auc": 0.736842105263158,
+      "eval_runtime": 0.5379,
+      "eval_samples_per_second": 74.363,
+      "eval_steps_per_second": 3.718,
       "step": 3
     },
     {
       "epoch": 2.0,
-      "eval_loss": 0.6675325632095337,
-      "eval_roc_auc": 0.5526315789473684,
-      "eval_runtime": 0.2781,
-      "eval_samples_per_second": 143.82,
-      "eval_steps_per_second": 7.191,
+      "eval_loss": 0.43417197465896606,
+      "eval_roc_auc": 0.7631578947368421,
+      "eval_runtime": 0.5781,
+      "eval_samples_per_second": 69.192,
+      "eval_steps_per_second": 3.46,
       "step": 6
     },
     {
       "epoch": 3.0,
-      "eval_loss": 0.6448550224304199,
-      "eval_roc_auc": 0.5789473684210527,
-      "eval_runtime": 0.3227,
-      "eval_samples_per_second": 123.942,
-      "eval_steps_per_second": 6.197,
+      "eval_loss": 0.43640345335006714,
+      "eval_roc_auc": 0.7493734335839599,
+      "eval_runtime": 0.3468,
+      "eval_samples_per_second": 115.327,
+      "eval_steps_per_second": 5.766,
       "step": 9
     },
     {
       "epoch": 3.33,
-      "learning_rate": 2.6666666666666667e-05,
-      "loss": 0.6541,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4067,
       "step": 10
     },
     {
       "epoch": 4.0,
-      "eval_loss": 0.6078753471374512,
-      "eval_roc_auc": 0.7105263157894737,
-      "eval_runtime": 0.2706,
-      "eval_samples_per_second": 147.82,
-      "eval_steps_per_second": 7.391,
+      "eval_loss": 0.446413516998291,
+      "eval_roc_auc": 0.7681704260651628,
+      "eval_runtime": 0.2866,
+      "eval_samples_per_second": 139.576,
+      "eval_steps_per_second": 6.979,
       "step": 12
     },
     {
       "epoch": 5.0,
-      "eval_loss": 0.5968624353408813,
-      "eval_roc_auc": 0.7105263157894737,
-      "eval_runtime": 0.3291,
-      "eval_samples_per_second": 121.557,
-      "eval_steps_per_second": 6.078,
+      "eval_loss": 0.46299201250076294,
+      "eval_roc_auc": 0.7656641604010025,
+      "eval_runtime": 0.2863,
+      "eval_samples_per_second": 139.701,
+      "eval_steps_per_second": 6.985,
       "step": 15
     },
     {
       "epoch": 6.0,
-      "eval_loss": 0.564300537109375,
-      "eval_roc_auc": 0.7105263157894737,
-      "eval_runtime": 0.2928,
-      "eval_samples_per_second": 136.627,
-      "eval_steps_per_second": 6.831,
+      "eval_loss": 0.45169758796691895,
+      "eval_roc_auc": 0.744360902255639,
+      "eval_runtime": 0.3942,
+      "eval_samples_per_second": 101.474,
+      "eval_steps_per_second": 5.074,
       "step": 18
     },
     {
       "epoch": 6.67,
-      "learning_rate": 7.61904761904762e-06,
-      "loss": 0.5334,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.4211,
       "step": 20
     },
     {
       "epoch": 7.0,
-      "eval_loss": 0.5328788161277771,
-      "eval_roc_auc": 0.7105263157894737,
-      "eval_runtime": 0.2769,
-      "eval_samples_per_second": 144.431,
-      "eval_steps_per_second": 7.222,
+      "eval_loss": 0.4539761543273926,
+      "eval_roc_auc": 0.7180451127819549,
+      "eval_runtime": 0.2978,
+      "eval_samples_per_second": 134.324,
+      "eval_steps_per_second": 6.716,
       "step": 21
     },
     {
       "epoch": 8.0,
-      "eval_loss": 0.5234358906745911,
-      "eval_roc_auc": 0.7631578947368421,
-      "eval_runtime": 0.2763,
-      "eval_samples_per_second": 144.791,
-      "eval_steps_per_second": 7.24,
+      "eval_loss": 0.45407772064208984,
+      "eval_roc_auc": 0.7180451127819549,
+      "eval_runtime": 0.2973,
+      "eval_samples_per_second": 134.55,
+      "eval_steps_per_second": 6.728,
       "step": 24
     },
     {
       "epoch": 8.0,
       "step": 24,
       "total_flos": 7.078985663171789e+16,
-      "train_loss": 0.5772393196821213,
-      "train_runtime": 61.0545,
-      "train_samples_per_second": 46.647,
-      "train_steps_per_second": 0.393
+      "train_loss": 0.4116581827402115,
+      "train_runtime": 69.7528,
+      "train_samples_per_second": 40.83,
+      "train_steps_per_second": 0.344
     }
   ],
   "logging_steps": 10,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2ba6304179519bcbae92d5008c20f1e1ef48f8904f31d69f6a4e4cadb6285358
+oid sha256:1279cbb2da12d0b87c1cb3f82e87a85c6eda9e64d36cfe842b264fd4214434f9
 size 4091