End of training

README.md

@@ -17,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-base-patch4-window7-224-in22k](https://huggingface.co/microsoft/swin-base-patch4-window7-224-in22k) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0126
+- Loss: 0.0062
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 8.42,
-    "eval_loss": 0.05487216264009476,
-    "eval_runtime": 2.638,
-    "eval_samples_per_second": 50.417,
-    "eval_steps_per_second": 1.137,
+    "eval_loss": 0.006183081306517124,
+    "eval_runtime": 2.9022,
+    "eval_samples_per_second": 45.828,
+    "eval_steps_per_second": 1.034,
     "total_flos": 7.88469671840514e+17,
-    "train_loss": 0.18382581882178783,
-    "train_runtime": 367.9939,
-    "train_samples_per_second": 32.446,
-    "train_steps_per_second": 0.109
+    "train_loss": 0.14818734787404536,
+    "train_runtime": 550.7447,
+    "train_samples_per_second": 21.68,
+    "train_steps_per_second": 0.073
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 8.42,
-    "eval_loss": 0.05487216264009476,
-    "eval_runtime": 2.638,
-    "eval_samples_per_second": 50.417,
-    "eval_steps_per_second": 1.137
+    "eval_loss": 0.006183081306517124,
+    "eval_runtime": 2.9022,
+    "eval_samples_per_second": 45.828,
+    "eval_steps_per_second": 1.034
 }

runs/Apr23_14-17-42_31541765f4b7/events.out.tfevents.1713882471.31541765f4b7.34.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c8c0e2e7c6a140c2c1688fe32b05ce4012ac6d2a1ad08c77b236c8eb981385c
+size 354

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 8.42,
     "total_flos": 7.88469671840514e+17,
-    "train_loss": 0.18382581882178783,
-    "train_runtime": 367.9939,
-    "train_samples_per_second": 32.446,
-    "train_steps_per_second": 0.109
+    "train_loss": 0.14818734787404536,
+    "train_runtime": 550.7447,
+    "train_samples_per_second": 21.68,
+    "train_steps_per_second": 0.073
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.05487216264009476,
-  "best_model_checkpoint": "swin-base-patch4-window7-224-in22k-Kontur-competition-1.4K/checkpoint-23",
+  "best_metric": 0.006183081306517124,
+  "best_model_checkpoint": "swin-base-patch4-window7-224-in22k-Kontur-competition-1.4K/checkpoint-19",
   "epoch": 8.421052631578947,
   "eval_steps": 500,
   "global_step": 40,
@@ -10,112 +10,112 @@
   "log_history": [
     {
       "epoch": 0.84,
-      "eval_loss": 0.5903199911117554,
-      "eval_runtime": 3.2368,
-      "eval_samples_per_second": 41.09,
-      "eval_steps_per_second": 0.927,
+      "eval_loss": 0.5152684450149536,
+      "eval_runtime": 5.7612,
+      "eval_samples_per_second": 23.086,
+      "eval_steps_per_second": 0.521,
       "step": 4
     },
     {
       "epoch": 1.89,
-      "eval_loss": 0.32031941413879395,
-      "eval_runtime": 2.5544,
-      "eval_samples_per_second": 52.067,
-      "eval_steps_per_second": 1.174,
+      "eval_loss": 0.14250300824642181,
+      "eval_runtime": 3.5004,
+      "eval_samples_per_second": 37.995,
+      "eval_steps_per_second": 0.857,
       "step": 9
     },
     {
       "epoch": 2.11,
-      "grad_norm": 12.904247283935547,
+      "grad_norm": 2.0780093669891357,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 0.5582,
+      "loss": 0.4963,
       "step": 10
     },
     {
       "epoch": 2.95,
-      "eval_loss": 0.2519628703594208,
-      "eval_runtime": 2.7088,
-      "eval_samples_per_second": 49.1,
-      "eval_steps_per_second": 1.108,
+      "eval_loss": 0.03258229419589043,
+      "eval_runtime": 2.856,
+      "eval_samples_per_second": 46.569,
+      "eval_steps_per_second": 1.05,
       "step": 14
     },
     {
       "epoch": 4.0,
-      "eval_loss": 0.09062427282333374,
-      "eval_runtime": 2.6817,
-      "eval_samples_per_second": 49.596,
-      "eval_steps_per_second": 1.119,
+      "eval_loss": 0.006183081306517124,
+      "eval_runtime": 2.8656,
+      "eval_samples_per_second": 46.412,
+      "eval_steps_per_second": 1.047,
       "step": 19
     },
     {
       "epoch": 4.21,
-      "grad_norm": 2.254042148590088,
+      "grad_norm": 4.404883861541748,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 0.1094,
+      "loss": 0.0546,
       "step": 20
     },
     {
       "epoch": 4.84,
-      "eval_loss": 0.05487216264009476,
-      "eval_runtime": 2.5622,
-      "eval_samples_per_second": 51.908,
-      "eval_steps_per_second": 1.171,
+      "eval_loss": 0.11129578948020935,
+      "eval_runtime": 2.942,
+      "eval_samples_per_second": 45.207,
+      "eval_steps_per_second": 1.02,
       "step": 23
     },
     {
       "epoch": 5.89,
-      "eval_loss": 0.14387370645999908,
-      "eval_runtime": 2.8209,
-      "eval_samples_per_second": 47.149,
-      "eval_steps_per_second": 1.064,
+      "eval_loss": 0.019418848678469658,
+      "eval_runtime": 3.0717,
+      "eval_samples_per_second": 43.299,
+      "eval_steps_per_second": 0.977,
       "step": 28
     },
     {
       "epoch": 6.32,
-      "grad_norm": 2.9556210041046143,
+      "grad_norm": 4.566842079162598,
       "learning_rate": 1.388888888888889e-05,
-      "loss": 0.0533,
+      "loss": 0.0216,
       "step": 30
     },
     {
       "epoch": 6.95,
-      "eval_loss": 0.06752961128950119,
-      "eval_runtime": 2.5525,
-      "eval_samples_per_second": 52.106,
-      "eval_steps_per_second": 1.175,
+      "eval_loss": 0.006876213010400534,
+      "eval_runtime": 2.9022,
+      "eval_samples_per_second": 45.828,
+      "eval_steps_per_second": 1.034,
       "step": 33
     },
     {
       "epoch": 8.0,
-      "eval_loss": 0.12144926190376282,
-      "eval_runtime": 2.5532,
-      "eval_samples_per_second": 52.091,
-      "eval_steps_per_second": 1.175,
+      "eval_loss": 0.009907796047627926,
+      "eval_runtime": 2.917,
+      "eval_samples_per_second": 45.594,
+      "eval_steps_per_second": 1.028,
       "step": 38
     },
     {
       "epoch": 8.42,
-      "grad_norm": 0.5887470245361328,
+      "grad_norm": 2.6469647884368896,
       "learning_rate": 0.0,
-      "loss": 0.0144,
+      "loss": 0.0203,
       "step": 40
     },
     {
       "epoch": 8.42,
-      "eval_loss": 0.1230710819363594,
-      "eval_runtime": 2.6049,
-      "eval_samples_per_second": 51.058,
-      "eval_steps_per_second": 1.152,
+      "eval_loss": 0.012587403878569603,
+      "eval_runtime": 2.9395,
+      "eval_samples_per_second": 45.246,
+      "eval_steps_per_second": 1.021,
       "step": 40
     },
     {
       "epoch": 8.42,
       "step": 40,
       "total_flos": 7.88469671840514e+17,
-      "train_loss": 0.18382581882178783,
-      "train_runtime": 367.9939,
-      "train_samples_per_second": 32.446,
-      "train_steps_per_second": 0.109
+      "train_loss": 0.14818734787404536,
+      "train_runtime": 550.7447,
+      "train_samples_per_second": 21.68,
+      "train_steps_per_second": 0.073
     }
   ],
   "logging_steps": 10,