End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "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
+    "epoch": 4.97,
+    "eval_accuracy": 0.9983948635634029,
+    "eval_loss": 0.010104374028742313,
+    "eval_runtime": 7.7407,
+    "eval_samples_per_second": 80.484,
+    "eval_steps_per_second": 1.292,
+    "total_flos": 1.461912118361211e+18,
+    "train_loss": 0.25554134793903516,
+    "train_runtime": 1204.2538,
+    "train_samples_per_second": 49.088,
+    "train_steps_per_second": 0.191
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "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
+    "epoch": 4.97,
+    "eval_accuracy": 0.9983948635634029,
+    "eval_loss": 0.010104374028742313,
+    "eval_runtime": 7.7407,
+    "eval_samples_per_second": 80.484,
+    "eval_steps_per_second": 1.292
 }

runs/Feb23_07-09-09_b3d631a6a648/events.out.tfevents.1708673367.b3d631a6a648.34.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0cc9886dc65c3eed28dd79bb4903aa1b0422d867885448a0804744c149d10a5
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "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
+    "epoch": 4.97,
+    "total_flos": 1.461912118361211e+18,
+    "train_loss": 0.25554134793903516,
+    "train_runtime": 1204.2538,
+    "train_samples_per_second": 49.088,
+    "train_steps_per_second": 0.191
 }

trainer_state.json

@@ -1,157 +1,141 @@
 {
-  "best_metric": 0.8653366583541147,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-mobile-eye-tracking-dataset-v2/checkpoint-59",
-  "epoch": 4.957983193277311,
+  "best_metric": 0.9983948635634029,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-mobile-eye-tracking-dataset-v2/checkpoint-230",
+  "epoch": 4.972972972972973,
   "eval_steps": 500,
-  "global_step": 295,
+  "global_step": 230,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.34,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.4662,
+      "epoch": 0.43,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 1.2017,
       "step": 20
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.811320754716982e-05,
-      "loss": 0.3599,
+      "epoch": 0.86,
+      "learning_rate": 4.589371980676328e-05,
+      "loss": 0.5175,
       "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
+      "eval_accuracy": 0.9069020866773676,
+      "eval_loss": 0.2164475917816162,
+      "eval_runtime": 7.2993,
+      "eval_samples_per_second": 85.35,
+      "eval_steps_per_second": 1.37,
+      "step": 46
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.433962264150944e-05,
-      "loss": 0.3968,
+      "epoch": 1.3,
+      "learning_rate": 4.106280193236715e-05,
+      "loss": 0.2849,
       "step": 60
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.0566037735849064e-05,
-      "loss": 0.3828,
+      "epoch": 1.73,
+      "learning_rate": 3.6231884057971014e-05,
+      "loss": 0.1932,
       "step": 80
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.679245283018868e-05,
-      "loss": 0.3768,
-      "step": 100
+      "epoch": 1.99,
+      "eval_accuracy": 0.9919743178170144,
+      "eval_loss": 0.0469740591943264,
+      "eval_runtime": 7.3311,
+      "eval_samples_per_second": 84.981,
+      "eval_steps_per_second": 1.364,
+      "step": 92
     },
     {
-      "epoch": 2.0,
-      "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": 2.16,
+      "learning_rate": 3.140096618357488e-05,
+      "loss": 0.1412,
+      "step": 100
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.30188679245283e-05,
-      "loss": 0.3674,
+      "epoch": 2.59,
+      "learning_rate": 2.6570048309178748e-05,
+      "loss": 0.1321,
       "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": 2.98,
+      "eval_accuracy": 0.9919743178170144,
+      "eval_loss": 0.03293607011437416,
+      "eval_runtime": 7.4116,
+      "eval_samples_per_second": 84.058,
+      "eval_steps_per_second": 1.349,
+      "step": 138
     },
     {
       "epoch": 3.03,
-      "learning_rate": 2.1698113207547172e-05,
-      "loss": 0.382,
-      "step": 180
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.1005,
+      "step": 140
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.7924528301886792e-05,
-      "loss": 0.3594,
-      "step": 200
+      "epoch": 3.46,
+      "learning_rate": 1.6908212560386476e-05,
+      "loss": 0.0922,
+      "step": 160
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 1.4150943396226415e-05,
-      "loss": 0.3838,
-      "step": 220
+      "epoch": 3.89,
+      "learning_rate": 1.2077294685990338e-05,
+      "loss": 0.0924,
+      "step": 180
     },
     {
       "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
+      "eval_accuracy": 0.9967897271268058,
+      "eval_loss": 0.01582668349146843,
+      "eval_runtime": 7.5144,
+      "eval_samples_per_second": 82.908,
+      "eval_steps_per_second": 1.331,
+      "step": 185
     },
     {
-      "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": 4.32,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 0.08,
+      "step": 200
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 2.830188679245283e-06,
-      "loss": 0.3531,
-      "step": 280
+      "epoch": 4.76,
+      "learning_rate": 2.4154589371980677e-06,
+      "loss": 0.0725,
+      "step": 220
     },
     {
-      "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": 4.97,
+      "eval_accuracy": 0.9983948635634029,
+      "eval_loss": 0.010104374028742313,
+      "eval_runtime": 7.9377,
+      "eval_samples_per_second": 78.486,
+      "eval_steps_per_second": 1.26,
+      "step": 230
     },
     {
-      "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
+      "epoch": 4.97,
+      "step": 230,
+      "total_flos": 1.461912118361211e+18,
+      "train_loss": 0.25554134793903516,
+      "train_runtime": 1204.2538,
+      "train_samples_per_second": 49.088,
+      "train_steps_per_second": 0.191
     }
   ],
   "logging_steps": 20,
-  "max_steps": 295,
+  "max_steps": 230,
+  "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
   "save_steps": 500,
-  "total_flos": 9.374684805905449e+17,
+  "total_flos": 1.461912118361211e+18,
+  "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null
 }