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: 0.3903
+- Loss: 0.4013
 - Accuracy: 0.8653
 
 ## Model description

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 3.0,
-    "eval_accuracy": 0.7692307692307693,
-    "eval_loss": 0.5808867812156677,
-    "eval_runtime": 3.3955,
-    "eval_samples_per_second": 3.829,
-    "eval_steps_per_second": 0.295,
-    "total_flos": 1.789631206981632e+16,
-    "train_loss": 0.6519472996393839,
-    "train_runtime": 550.2032,
-    "train_samples_per_second": 1.309,
-    "train_steps_per_second": 0.011
+    "epoch": 4.96,
+    "eval_accuracy": 0.8653366583541147,
+    "eval_loss": 0.4013111889362335,
+    "eval_runtime": 20.9016,
+    "eval_samples_per_second": 19.185,
+    "eval_steps_per_second": 0.622,
+    "total_flos": 9.374684805905449e+17,
+    "train_loss": 0.37977570679228184,
+    "train_runtime": 2821.3451,
+    "train_samples_per_second": 13.478,
+    "train_steps_per_second": 0.105
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "eval_accuracy": 0.7692307692307693,
-    "eval_loss": 0.5808867812156677,
-    "eval_runtime": 3.3955,
-    "eval_samples_per_second": 3.829,
-    "eval_steps_per_second": 0.295
+    "epoch": 4.96,
+    "eval_accuracy": 0.8653366583541147,
+    "eval_loss": 0.4013111889362335,
+    "eval_runtime": 20.9016,
+    "eval_samples_per_second": 19.185,
+    "eval_steps_per_second": 0.622
 }

runs/Jan13_13-19-29_f0d5ed160523/events.out.tfevents.1705154820.f0d5ed160523.4585.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7905d485ffc31d74a909d398949af39a0146618dd21809c17c54440365e03084
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "total_flos": 1.789631206981632e+16,
-    "train_loss": 0.6519472996393839,
-    "train_runtime": 550.2032,
-    "train_samples_per_second": 1.309,
-    "train_steps_per_second": 0.011
+    "epoch": 4.96,
+    "total_flos": 9.374684805905449e+17,
+    "train_loss": 0.37977570679228184,
+    "train_runtime": 2821.3451,
+    "train_samples_per_second": 13.478,
+    "train_steps_per_second": 0.105
 }

trainer_state.json

@@ -1,55 +1,157 @@
 {
-  "best_metric": 0.7692307692307693,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-mobile-eye-tracking-dataset-v2/checkpoint-4",
-  "epoch": 3.0,
+  "best_metric": 0.8653366583541147,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-mobile-eye-tracking-dataset-v2/checkpoint-59",
+  "epoch": 4.957983193277311,
   "eval_steps": 500,
-  "global_step": 6,
+  "global_step": 295,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.6923076923076923,
-      "eval_loss": 0.6332569718360901,
-      "eval_runtime": 2.7762,
-      "eval_samples_per_second": 4.683,
-      "eval_steps_per_second": 0.36,
-      "step": 2
+      "epoch": 0.34,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4662,
+      "step": 20
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.811320754716982e-05,
+      "loss": 0.3599,
+      "step": 40
+    },
+    {
+      "epoch": 0.99,
+      "eval_accuracy": 0.8653366583541147,
+      "eval_loss": 0.4013111889362335,
+      "eval_runtime": 51.6469,
+      "eval_samples_per_second": 7.764,
+      "eval_steps_per_second": 0.252,
+      "step": 59
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433962264150944e-05,
+      "loss": 0.3968,
+      "step": 60
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.0566037735849064e-05,
+      "loss": 0.3828,
+      "step": 80
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.679245283018868e-05,
+      "loss": 0.3768,
+      "step": 100
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7692307692307693,
-      "eval_loss": 0.5808867812156677,
-      "eval_runtime": 2.7727,
-      "eval_samples_per_second": 4.689,
-      "eval_steps_per_second": 0.361,
-      "step": 4
-    },
-    {
-      "epoch": 3.0,
-      "eval_accuracy": 0.7692307692307693,
-      "eval_loss": 0.5728916525840759,
-      "eval_runtime": 4.1167,
-      "eval_samples_per_second": 3.158,
-      "eval_steps_per_second": 0.243,
-      "step": 6
-    },
-    {
-      "epoch": 3.0,
-      "step": 6,
-      "total_flos": 1.789631206981632e+16,
-      "train_loss": 0.6519472996393839,
-      "train_runtime": 550.2032,
-      "train_samples_per_second": 1.309,
-      "train_steps_per_second": 0.011
+      "eval_accuracy": 0.8653366583541147,
+      "eval_loss": 0.39926877617836,
+      "eval_runtime": 20.5275,
+      "eval_samples_per_second": 19.535,
+      "eval_steps_per_second": 0.633,
+      "step": 119
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.30188679245283e-05,
+      "loss": 0.3674,
+      "step": 120
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.9245283018867926e-05,
+      "loss": 0.3861,
+      "step": 140
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.547169811320755e-05,
+      "loss": 0.3549,
+      "step": 160
+    },
+    {
+      "epoch": 2.99,
+      "eval_accuracy": 0.8653366583541147,
+      "eval_loss": 0.39215829968452454,
+      "eval_runtime": 22.3126,
+      "eval_samples_per_second": 17.972,
+      "eval_steps_per_second": 0.583,
+      "step": 178
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.1698113207547172e-05,
+      "loss": 0.382,
+      "step": 180
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.7924528301886792e-05,
+      "loss": 0.3594,
+      "step": 200
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.3838,
+      "step": 220
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8653366583541147,
+      "eval_loss": 0.38929104804992676,
+      "eval_runtime": 21.0865,
+      "eval_samples_per_second": 19.017,
+      "eval_steps_per_second": 0.617,
+      "step": 238
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.376,
+      "step": 240
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.3763,
+      "step": 260
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.3531,
+      "step": 280
+    },
+    {
+      "epoch": 4.96,
+      "eval_accuracy": 0.8653366583541147,
+      "eval_loss": 0.39027392864227295,
+      "eval_runtime": 21.4803,
+      "eval_samples_per_second": 18.668,
+      "eval_steps_per_second": 0.605,
+      "step": 295
+    },
+    {
+      "epoch": 4.96,
+      "step": 295,
+      "total_flos": 9.374684805905449e+17,
+      "train_loss": 0.37977570679228184,
+      "train_runtime": 2821.3451,
+      "train_samples_per_second": 13.478,
+      "train_steps_per_second": 0.105
     }
   ],
-  "logging_steps": 10,
-  "max_steps": 6,
-  "num_train_epochs": 3,
+  "logging_steps": 20,
+  "max_steps": 295,
+  "num_train_epochs": 5,
   "save_steps": 500,
-  "total_flos": 1.789631206981632e+16,
+  "total_flos": 9.374684805905449e+17,
   "trial_name": null,
   "trial_params": null
 }