End of training

README.md

@@ -32,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1135
+- Loss: 0.1666
 - Accuracy: 0.9765
 
 ## Model description

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 2.79,
-    "eval_accuracy": 0.9803921568627451,
-    "eval_loss": 0.1283220797777176,
-    "eval_runtime": 2.6884,
-    "eval_samples_per_second": 56.912,
-    "eval_steps_per_second": 1.86,
-    "total_flos": 9.510432845082624e+16,
-    "train_loss": 0.4215235114097595,
-    "train_runtime": 104.5277,
-    "train_samples_per_second": 39.291,
-    "train_steps_per_second": 0.287
+    "epoch": 3.0,
+    "eval_accuracy": 0.9764705882352941,
+    "eval_loss": 0.16659174859523773,
+    "eval_runtime": 4.0114,
+    "eval_samples_per_second": 42.379,
+    "eval_steps_per_second": 1.496,
+    "total_flos": 1.137256821195264e+17,
+    "train_loss": 0.5669683615366617,
+    "train_runtime": 135.6407,
+    "train_samples_per_second": 33.729,
+    "train_steps_per_second": 0.265
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.79,
-    "eval_accuracy": 0.9803921568627451,
-    "eval_loss": 0.1283220797777176,
-    "eval_runtime": 2.6884,
-    "eval_samples_per_second": 56.912,
-    "eval_steps_per_second": 1.86
+    "epoch": 3.0,
+    "eval_accuracy": 0.9764705882352941,
+    "eval_loss": 0.16659174859523773,
+    "eval_runtime": 4.0114,
+    "eval_samples_per_second": 42.379,
+    "eval_steps_per_second": 1.496
 }

runs/Apr08_04-52-43_e7c0f18713ab/events.out.tfevents.1712552170.e7c0f18713ab.441.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c175064aa056a701db9282ce5cd1afae3c8913320cd53327ae472815e83961ee
+size 405

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.79,
-    "total_flos": 9.510432845082624e+16,
-    "train_loss": 0.4215235114097595,
-    "train_runtime": 104.5277,
-    "train_samples_per_second": 39.291,
-    "train_steps_per_second": 0.287
+    "epoch": 3.0,
+    "total_flos": 1.137256821195264e+17,
+    "train_loss": 0.5669683615366617,
+    "train_runtime": 135.6407,
+    "train_samples_per_second": 33.729,
+    "train_steps_per_second": 0.265
 }

trainer_state.json

@@ -1,73 +1,78 @@
 {
-  "best_metric": 0.9803921568627451,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-30",
-  "epoch": 2.7906976744186047,
+  "best_metric": 0.9764705882352941,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-24",
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 30,
+  "global_step": 36,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.93,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.9149,
+      "epoch": 0.83,
+      "grad_norm": 4.032827854156494,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 1.2376,
       "step": 10
     },
     {
-      "epoch": 0.93,
-      "eval_accuracy": 0.9084967320261438,
-      "eval_loss": 0.3354261815547943,
-      "eval_runtime": 2.5879,
-      "eval_samples_per_second": 59.121,
-      "eval_steps_per_second": 1.932,
-      "step": 10
+      "epoch": 1.0,
+      "eval_accuracy": 0.7470588235294118,
+      "eval_loss": 0.5182495713233948,
+      "eval_runtime": 3.913,
+      "eval_samples_per_second": 43.445,
+      "eval_steps_per_second": 1.533,
+      "step": 12
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.2319,
+      "epoch": 1.67,
+      "grad_norm": 3.3620471954345703,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4743,
       "step": 20
     },
     {
-      "epoch": 1.95,
-      "eval_accuracy": 0.9738562091503268,
-      "eval_loss": 0.1467364877462387,
-      "eval_runtime": 2.6141,
-      "eval_samples_per_second": 58.53,
-      "eval_steps_per_second": 1.913,
-      "step": 21
+      "epoch": 2.0,
+      "eval_accuracy": 0.9764705882352941,
+      "eval_loss": 0.16659174859523773,
+      "eval_runtime": 3.9553,
+      "eval_samples_per_second": 42.98,
+      "eval_steps_per_second": 1.517,
+      "step": 24
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 0.0,
-      "loss": 0.1178,
+      "epoch": 2.5,
+      "grad_norm": 3.355973720550537,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.2261,
       "step": 30
     },
     {
-      "epoch": 2.79,
-      "eval_accuracy": 0.9803921568627451,
-      "eval_loss": 0.1283220797777176,
-      "eval_runtime": 2.7473,
-      "eval_samples_per_second": 55.692,
-      "eval_steps_per_second": 1.82,
-      "step": 30
+      "epoch": 3.0,
+      "eval_accuracy": 0.9764705882352941,
+      "eval_loss": 0.11346375942230225,
+      "eval_runtime": 4.6585,
+      "eval_samples_per_second": 36.492,
+      "eval_steps_per_second": 1.288,
+      "step": 36
     },
     {
-      "epoch": 2.79,
-      "step": 30,
-      "total_flos": 9.510432845082624e+16,
-      "train_loss": 0.4215235114097595,
-      "train_runtime": 104.5277,
-      "train_samples_per_second": 39.291,
-      "train_steps_per_second": 0.287
+      "epoch": 3.0,
+      "step": 36,
+      "total_flos": 1.137256821195264e+17,
+      "train_loss": 0.5669683615366617,
+      "train_runtime": 135.6407,
+      "train_samples_per_second": 33.729,
+      "train_steps_per_second": 0.265
     }
   ],
   "logging_steps": 10,
-  "max_steps": 30,
+  "max_steps": 36,
+  "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 9.510432845082624e+16,
+  "total_flos": 1.137256821195264e+17,
+  "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null
 }