End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9809264305177112,
-    "eval_loss": 0.07860878109931946,
-    "eval_runtime": 5.4221,
-    "eval_samples_per_second": 67.687,
-    "eval_steps_per_second": 2.213,
-    "total_flos": 7.678887715683717e+17,
-    "train_loss": 0.36344620203360534,
-    "train_runtime": 415.1296,
-    "train_samples_per_second": 23.87,
-    "train_steps_per_second": 0.188
+    "eval_accuracy": 0.6746987951807228,
+    "eval_loss": 1.1532235145568848,
+    "eval_runtime": 8.0102,
+    "eval_samples_per_second": 62.171,
+    "eval_steps_per_second": 1.997,
+    "total_flos": 1.0402373155785892e+18,
+    "train_loss": 1.5203653063092912,
+    "train_runtime": 531.0704,
+    "train_samples_per_second": 25.273,
+    "train_steps_per_second": 0.198
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9809264305177112,
-    "eval_loss": 0.07860878109931946,
-    "eval_runtime": 5.4221,
-    "eval_samples_per_second": 67.687,
-    "eval_steps_per_second": 2.213
+    "eval_accuracy": 0.6746987951807228,
+    "eval_loss": 1.1532235145568848,
+    "eval_runtime": 8.0102,
+    "eval_samples_per_second": 62.171,
+    "eval_steps_per_second": 1.997
 }

runs/Nov30_13-49-19_4b23d0e9dade/events.out.tfevents.1701352758.4b23d0e9dade.2219.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b42ca4a5a3cabea59983fea349942f7b779dc374cba6baa20644ea1c5e5ec76
+size 405

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "total_flos": 7.678887715683717e+17,
-    "train_loss": 0.36344620203360534,
-    "train_runtime": 415.1296,
-    "train_samples_per_second": 23.87,
-    "train_steps_per_second": 0.188
+    "total_flos": 1.0402373155785892e+18,
+    "train_loss": 1.5203653063092912,
+    "train_runtime": 531.0704,
+    "train_samples_per_second": 25.273,
+    "train_steps_per_second": 0.198
 }

trainer_state.json

@@ -1,97 +1,115 @@
 {
-  "best_metric": 0.9809264305177112,
-  "best_model_checkpoint": "vit-base-patch16-224-finetuned/checkpoint-52",
+  "best_metric": 0.6746987951807228,
+  "best_model_checkpoint": "vit-base-patch16-224-finetuned/checkpoint-105",
   "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 78,
+  "global_step": 105,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.38,
-      "learning_rate": 4.8571428571428576e-05,
-      "loss": 1.5192,
+      "epoch": 0.29,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 2.8668,
       "step": 10
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.1428571428571437e-05,
-      "loss": 0.5888,
+      "epoch": 0.57,
+      "learning_rate": 4.5212765957446815e-05,
+      "loss": 2.2489,
       "step": 20
     },
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.9700272479564033,
-      "eval_loss": 0.17195835709571838,
-      "eval_runtime": 5.5957,
-      "eval_samples_per_second": 65.586,
-      "eval_steps_per_second": 2.145,
-      "step": 26
+      "epoch": 0.86,
+      "learning_rate": 3.9893617021276594e-05,
+      "loss": 1.8046,
+      "step": 30
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 3.428571428571429e-05,
-      "loss": 0.2357,
-      "step": 30
+      "epoch": 1.0,
+      "eval_accuracy": 0.6004016064257028,
+      "eval_loss": 1.5307656526565552,
+      "eval_runtime": 7.8884,
+      "eval_samples_per_second": 63.131,
+      "eval_steps_per_second": 2.028,
+      "step": 35
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 2.714285714285714e-05,
-      "loss": 0.1499,
+      "epoch": 1.14,
+      "learning_rate": 3.4574468085106386e-05,
+      "loss": 1.505,
       "step": 40
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 2e-05,
-      "loss": 0.1027,
+      "epoch": 1.43,
+      "learning_rate": 2.925531914893617e-05,
+      "loss": 1.3769,
       "step": 50
     },
     {
-      "epoch": 2.0,
-      "eval_accuracy": 0.9809264305177112,
-      "eval_loss": 0.07860878109931946,
-      "eval_runtime": 5.4987,
-      "eval_samples_per_second": 66.743,
-      "eval_steps_per_second": 2.182,
-      "step": 52
+      "epoch": 1.71,
+      "learning_rate": 2.393617021276596e-05,
+      "loss": 1.2191,
+      "step": 60
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 1.2857142857142857e-05,
-      "loss": 0.0989,
-      "step": 60
+      "epoch": 2.0,
+      "learning_rate": 1.8617021276595745e-05,
+      "loss": 1.1931,
+      "step": 70
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 5.7142857142857145e-06,
-      "loss": 0.0809,
+      "epoch": 2.0,
+      "eval_accuracy": 0.6526104417670683,
+      "eval_loss": 1.2079730033874512,
+      "eval_runtime": 8.2786,
+      "eval_samples_per_second": 60.155,
+      "eval_steps_per_second": 1.933,
       "step": 70
     },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.3297872340425532e-05,
+      "loss": 1.0954,
+      "step": 80
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.97872340425532e-06,
+      "loss": 1.0719,
+      "step": 90
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 1.0292,
+      "step": 100
+    },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9809264305177112,
-      "eval_loss": 0.07304801046848297,
-      "eval_runtime": 5.4275,
-      "eval_samples_per_second": 67.619,
-      "eval_steps_per_second": 2.211,
-      "step": 78
+      "eval_accuracy": 0.6746987951807228,
+      "eval_loss": 1.1532235145568848,
+      "eval_runtime": 7.6457,
+      "eval_samples_per_second": 65.135,
+      "eval_steps_per_second": 2.093,
+      "step": 105
     },
     {
       "epoch": 3.0,
-      "step": 78,
-      "total_flos": 7.678887715683717e+17,
-      "train_loss": 0.36344620203360534,
-      "train_runtime": 415.1296,
-      "train_samples_per_second": 23.87,
-      "train_steps_per_second": 0.188
+      "step": 105,
+      "total_flos": 1.0402373155785892e+18,
+      "train_loss": 1.5203653063092912,
+      "train_runtime": 531.0704,
+      "train_samples_per_second": 25.273,
+      "train_steps_per_second": 0.198
     }
   ],
   "logging_steps": 10,
-  "max_steps": 78,
+  "max_steps": 105,
   "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 7.678887715683717e+17,
+  "total_flos": 1.0402373155785892e+18,
   "trial_name": null,
   "trial_params": null
 }