Training in progress, step 100

all_results.json

@@ -1,8 +1,13 @@
 {
-    "epoch": 4.0,
-    "total_flos": 2.451846551405691e+17,
-    "train_loss": 0.30905103862285616,
-    "train_runtime": 185.0394,
-    "train_samples_per_second": 17.099,
-    "train_steps_per_second": 1.081
+    "epoch": 10.0,
+    "eval_accuracy": 0.9591836734693877,
+    "eval_loss": 0.17216908931732178,
+    "eval_runtime": 3.4667,
+    "eval_samples_per_second": 28.269,
+    "eval_steps_per_second": 3.75,
+    "total_flos": 6.129616378514227e+17,
+    "train_loss": 0.1937787554860115,
+    "train_runtime": 492.9488,
+    "train_samples_per_second": 16.046,
+    "train_steps_per_second": 1.014
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 10.0,
+    "eval_accuracy": 0.9591836734693877,
+    "eval_loss": 0.17216908931732178,
+    "eval_runtime": 2.3887,
+    "eval_samples_per_second": 41.026,
+    "eval_steps_per_second": 5.442
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ae97fd4bf6040174380657756dbee2a08a8051490a45d569384e15ea6a02ae3
+oid sha256:3ede0e8d4c0f611742667a390b1ea2dce854b835570a79c7e5420295720b0a74
 size 343269162

runs/May15_17-41-07_601c3d0c9bcd/events.out.tfevents.1715795412.601c3d0c9bcd.2403.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87dc7717cf4564b6a5b256aab57e9ec889f1819b796dcc838017ca715c9864a0
+size 734

runs/May15_17-51-41_601c3d0c9bcd/events.out.tfevents.1715795517.601c3d0c9bcd.2403.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e48d72ad24d03f9797125b8afb10a881216de95746fd2cf5ec91f249266fb4a
+size 5801

test_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 10.0,
+    "eval_accuracy": 0.9591836734693877,
+    "eval_loss": 0.17216908931732178,
+    "eval_runtime": 3.4667,
+    "eval_samples_per_second": 28.269,
+    "eval_steps_per_second": 3.75
+}

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.0,
-    "total_flos": 2.451846551405691e+17,
-    "train_loss": 0.30905103862285616,
-    "train_runtime": 185.0394,
-    "train_samples_per_second": 17.099,
-    "train_steps_per_second": 1.081
+    "epoch": 10.0,
+    "total_flos": 6.129616378514227e+17,
+    "train_loss": 0.1937787554860115,
+    "train_runtime": 492.9488,
+    "train_samples_per_second": 16.046,
+    "train_steps_per_second": 1.014
 }

trainer_state.json

@@ -1,163 +1,370 @@
 {
-  "best_metric": 0.13754139840602875,
-  "best_model_checkpoint": "./vit-accident-image/checkpoint-200",
-  "epoch": 4.0,
-  "global_step": 200,
+  "best_metric": 0.17216908931732178,
+  "best_model_checkpoint": "./vit-accident-image/checkpoint-400",
+  "epoch": 10.0,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.2,
-      "learning_rate": 0.00019,
-      "loss": 0.6901,
+      "learning_rate": 0.000196,
+      "loss": 0.6688,
       "step": 10
     },
     {
       "epoch": 0.4,
-      "learning_rate": 0.00018,
-      "loss": 0.6157,
+      "learning_rate": 0.000192,
+      "loss": 0.6088,
       "step": 20
     },
     {
       "epoch": 0.6,
-      "learning_rate": 0.00017,
-      "loss": 0.6145,
+      "learning_rate": 0.000188,
+      "loss": 0.6081,
       "step": 30
     },
     {
       "epoch": 0.8,
-      "learning_rate": 0.00016,
-      "loss": 0.5501,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 0.5727,
       "step": 40
     },
     {
       "epoch": 1.0,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 0.4607,
+      "learning_rate": 0.00018,
+      "loss": 0.5506,
       "step": 50
     },
     {
       "epoch": 1.2,
-      "learning_rate": 0.00014,
-      "loss": 0.27,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 0.3946,
       "step": 60
     },
     {
       "epoch": 1.4,
-      "learning_rate": 0.00013000000000000002,
-      "loss": 0.3907,
+      "learning_rate": 0.000172,
+      "loss": 0.3847,
       "step": 70
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.00012,
-      "loss": 0.5207,
+      "learning_rate": 0.000168,
+      "loss": 0.5908,
       "step": 80
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.00011000000000000002,
-      "loss": 0.2799,
+      "learning_rate": 0.000164,
+      "loss": 0.3266,
       "step": 90
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0001,
-      "loss": 0.3926,
+      "learning_rate": 0.00016,
+      "loss": 0.5025,
       "step": 100
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8979591836734694,
-      "eval_loss": 0.2468928098678589,
-      "eval_runtime": 2.237,
-      "eval_samples_per_second": 43.808,
-      "eval_steps_per_second": 5.811,
+      "eval_accuracy": 0.8775510204081632,
+      "eval_loss": 0.3267006278038025,
+      "eval_runtime": 2.247,
+      "eval_samples_per_second": 43.613,
+      "eval_steps_per_second": 5.785,
       "step": 100
     },
     {
       "epoch": 2.2,
-      "learning_rate": 9e-05,
-      "loss": 0.158,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 0.287,
       "step": 110
     },
     {
       "epoch": 2.4,
-      "learning_rate": 8e-05,
-      "loss": 0.2589,
+      "learning_rate": 0.000152,
+      "loss": 0.2996,
       "step": 120
     },
     {
       "epoch": 2.6,
-      "learning_rate": 7e-05,
-      "loss": 0.1508,
+      "learning_rate": 0.000148,
+      "loss": 0.1329,
       "step": 130
     },
     {
       "epoch": 2.8,
-      "learning_rate": 6e-05,
-      "loss": 0.188,
+      "learning_rate": 0.000144,
+      "loss": 0.2857,
       "step": 140
     },
     {
       "epoch": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1564,
+      "learning_rate": 0.00014,
+      "loss": 0.2885,
       "step": 150
     },
     {
       "epoch": 3.2,
-      "learning_rate": 4e-05,
-      "loss": 0.1226,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 0.244,
       "step": 160
     },
     {
       "epoch": 3.4,
-      "learning_rate": 3e-05,
-      "loss": 0.0921,
+      "learning_rate": 0.000132,
+      "loss": 0.1873,
       "step": 170
     },
     {
       "epoch": 3.6,
-      "learning_rate": 2e-05,
-      "loss": 0.0828,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 0.1446,
       "step": 180
     },
     {
       "epoch": 3.8,
-      "learning_rate": 1e-05,
-      "loss": 0.096,
+      "learning_rate": 0.000124,
+      "loss": 0.1887,
       "step": 190
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.0,
-      "loss": 0.0904,
+      "learning_rate": 0.00012,
+      "loss": 0.204,
       "step": 200
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9285714285714286,
-      "eval_loss": 0.13754139840602875,
-      "eval_runtime": 2.2019,
-      "eval_samples_per_second": 44.506,
-      "eval_steps_per_second": 5.904,
+      "eval_accuracy": 0.9081632653061225,
+      "eval_loss": 0.29408177733421326,
+      "eval_runtime": 3.0909,
+      "eval_samples_per_second": 31.706,
+      "eval_steps_per_second": 4.206,
       "step": 200
     },
     {
-      "epoch": 4.0,
-      "step": 200,
-      "total_flos": 2.451846551405691e+17,
-      "train_loss": 0.30905103862285616,
-      "train_runtime": 185.0394,
-      "train_samples_per_second": 17.099,
-      "train_steps_per_second": 1.081
+      "epoch": 4.2,
+      "learning_rate": 0.000116,
+      "loss": 0.0822,
+      "step": 210
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 0.0912,
+      "step": 220
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 0.1816,
+      "step": 230
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 0.1005,
+      "step": 240
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0001,
+      "loss": 0.1167,
+      "step": 250
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 9.6e-05,
+      "loss": 0.1081,
+      "step": 260
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 0.0973,
+      "step": 270
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 0.1825,
+      "step": 280
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 8.4e-05,
+      "loss": 0.0941,
+      "step": 290
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 8e-05,
+      "loss": 0.1425,
+      "step": 300
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.9081632653061225,
+      "eval_loss": 0.34069785475730896,
+      "eval_runtime": 2.9263,
+      "eval_samples_per_second": 33.489,
+      "eval_steps_per_second": 4.442,
+      "step": 300
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 7.6e-05,
+      "loss": 0.093,
+      "step": 310
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 7.2e-05,
+      "loss": 0.0847,
+      "step": 320
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 0.0561,
+      "step": 330
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.0832,
+      "step": 340
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 6e-05,
+      "loss": 0.0663,
+      "step": 350
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 0.0626,
+      "step": 360
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 0.0282,
+      "step": 370
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 4.8e-05,
+      "loss": 0.0528,
+      "step": 380
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.0416,
+      "step": 390
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4e-05,
+      "loss": 0.091,
+      "step": 400
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9591836734693877,
+      "eval_loss": 0.17216908931732178,
+      "eval_runtime": 2.1538,
+      "eval_samples_per_second": 45.502,
+      "eval_steps_per_second": 6.036,
+      "step": 400
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 3.6e-05,
+      "loss": 0.0602,
+      "step": 410
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.0267,
+      "step": 420
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.0447,
+      "step": 430
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 2.4e-05,
+      "loss": 0.038,
+      "step": 440
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 450
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.0271,
+      "step": 460
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 1.2e-05,
+      "loss": 0.0318,
+      "step": 470
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.0465,
+      "step": 480
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0154,
+      "step": 490
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 0.0328,
+      "step": 500
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9387755102040817,
+      "eval_loss": 0.20507687330245972,
+      "eval_runtime": 3.1293,
+      "eval_samples_per_second": 31.316,
+      "eval_steps_per_second": 4.154,
+      "step": 500
+    },
+    {
+      "epoch": 10.0,
+      "step": 500,
+      "total_flos": 6.129616378514227e+17,
+      "train_loss": 0.1937787554860115,
+      "train_runtime": 492.9488,
+      "train_samples_per_second": 16.046,
+      "train_steps_per_second": 1.014
     }
   ],
-  "max_steps": 200,
-  "num_train_epochs": 4,
-  "total_flos": 2.451846551405691e+17,
+  "max_steps": 500,
+  "num_train_epochs": 10,
+  "total_flos": 6.129616378514227e+17,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:72bf40f34cbaf5f535327170fb1dd632c66cda01a6edf24373636ba45348a414
+oid sha256:daed8b6ce15cc6b3f87734aa2e8979670d0cfb855c83d202deb6e40526c5fca7
 size 4408