Training in progress, epoch 0

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 4.888888888888889,
-    "eval_accuracy": 0.9686192468619247,
-    "eval_loss": 0.08668407797813416,
-    "eval_runtime": 2.1418,
-    "eval_samples_per_second": 223.174,
-    "eval_steps_per_second": 7.003,
-    "total_flos": 5.227711603505234e+17,
-    "train_loss": 0.22451729810599125,
-    "train_runtime": 231.4165,
-    "train_samples_per_second": 92.863,
-    "train_steps_per_second": 0.713
+    "epoch": 1.9555555555555557,
+    "eval_accuracy": 0.897489539748954,
+    "eval_loss": 0.23102231323719025,
+    "eval_runtime": 2.1759,
+    "eval_samples_per_second": 219.675,
+    "eval_steps_per_second": 6.894,
+    "total_flos": 2.0943656319482266e+17,
+    "train_loss": 0.3058708075321082,
+    "train_runtime": 94.9406,
+    "train_samples_per_second": 90.541,
+    "train_steps_per_second": 0.695
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.888888888888889,
-    "eval_accuracy": 0.9686192468619247,
-    "eval_loss": 0.08668407797813416,
-    "eval_runtime": 2.1418,
-    "eval_samples_per_second": 223.174,
-    "eval_steps_per_second": 7.003
+    "epoch": 1.9555555555555557,
+    "eval_accuracy": 0.897489539748954,
+    "eval_loss": 0.23102231323719025,
+    "eval_runtime": 2.1759,
+    "eval_samples_per_second": 219.675,
+    "eval_steps_per_second": 6.894
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:00a94d41951c09179b9754f2ab82611067a772369682d2e532748164d380ac95
+oid sha256:4dc0387f629720c63b32bf787672f3035cb9c0eee6dcaa6bb2ec89535e23b34a
 size 110342832

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.888888888888889,
-    "total_flos": 5.227711603505234e+17,
-    "train_loss": 0.22451729810599125,
-    "train_runtime": 231.4165,
-    "train_samples_per_second": 92.863,
-    "train_steps_per_second": 0.713
+    "epoch": 1.9555555555555557,
+    "total_flos": 2.0943656319482266e+17,
+    "train_loss": 0.3058708075321082,
+    "train_runtime": 94.9406,
+    "train_samples_per_second": 90.541,
+    "train_steps_per_second": 0.695
 }

trainer_state.json

@@ -1,184 +1,87 @@
 {
-  "best_metric": 0.9686192468619247,
-  "best_model_checkpoint": "/kaggle/working/swin-brain-abnormalities-classification/checkpoint-135",
-  "epoch": 4.888888888888889,
+  "best_metric": 0.897489539748954,
+  "best_model_checkpoint": "/kaggle/working/swin-brain-abnormalities-classification/checkpoint-33",
+  "epoch": 1.9555555555555557,
   "eval_steps": 500,
-  "global_step": 165,
+  "global_step": 66,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.2962962962962963,
-      "grad_norm": 5.072397232055664,
-      "learning_rate": 2.9411764705882354e-05,
-      "loss": 0.695,
+      "grad_norm": 3.896357297897339,
+      "learning_rate": 4.745762711864407e-05,
+      "loss": 0.6251,
       "step": 10
     },
     {
       "epoch": 0.5925925925925926,
-      "grad_norm": 5.263370037078857,
-      "learning_rate": 4.8986486486486486e-05,
-      "loss": 0.5148,
+      "grad_norm": 8.837748527526855,
+      "learning_rate": 3.898305084745763e-05,
+      "loss": 0.38,
       "step": 20
     },
     {
       "epoch": 0.8888888888888888,
-      "grad_norm": 6.117415428161621,
-      "learning_rate": 4.560810810810811e-05,
-      "loss": 0.3422,
+      "grad_norm": 9.348395347595215,
+      "learning_rate": 3.050847457627119e-05,
+      "loss": 0.2769,
       "step": 30
     },
     {
       "epoch": 0.9777777777777777,
-      "eval_accuracy": 0.9414225941422594,
-      "eval_loss": 0.16747963428497314,
-      "eval_runtime": 2.1191,
-      "eval_samples_per_second": 225.569,
-      "eval_steps_per_second": 7.079,
+      "eval_accuracy": 0.897489539748954,
+      "eval_loss": 0.23102231323719025,
+      "eval_runtime": 2.1587,
+      "eval_samples_per_second": 221.433,
+      "eval_steps_per_second": 6.949,
       "step": 33
     },
     {
       "epoch": 1.1851851851851851,
-      "grad_norm": 6.219008445739746,
-      "learning_rate": 4.222972972972973e-05,
-      "loss": 0.2855,
+      "grad_norm": 8.096166610717773,
+      "learning_rate": 2.2033898305084748e-05,
+      "loss": 0.2364,
       "step": 40
     },
     {
       "epoch": 1.4814814814814814,
-      "grad_norm": 20.622819900512695,
-      "learning_rate": 3.885135135135135e-05,
-      "loss": 0.2291,
+      "grad_norm": 15.118054389953613,
+      "learning_rate": 1.3559322033898305e-05,
+      "loss": 0.2129,
       "step": 50
     },
     {
       "epoch": 1.7777777777777777,
-      "grad_norm": 5.243396759033203,
-      "learning_rate": 3.547297297297297e-05,
-      "loss": 0.217,
+      "grad_norm": 6.056334495544434,
+      "learning_rate": 5.084745762711865e-06,
+      "loss": 0.1861,
       "step": 60
     },
     {
-      "epoch": 1.9851851851851852,
-      "eval_accuracy": 0.9581589958158996,
-      "eval_loss": 0.13918140530586243,
-      "eval_runtime": 2.1072,
-      "eval_samples_per_second": 226.844,
-      "eval_steps_per_second": 7.119,
-      "step": 67
+      "epoch": 1.9555555555555557,
+      "eval_accuracy": 0.897489539748954,
+      "eval_loss": 0.24522443115711212,
+      "eval_runtime": 2.2548,
+      "eval_samples_per_second": 211.994,
+      "eval_steps_per_second": 6.653,
+      "step": 66
     },
     {
-      "epoch": 2.074074074074074,
-      "grad_norm": 4.132438659667969,
-      "learning_rate": 3.20945945945946e-05,
-      "loss": 0.2056,
-      "step": 70
-    },
-    {
-      "epoch": 2.3703703703703702,
-      "grad_norm": 9.177586555480957,
-      "learning_rate": 2.8716216216216217e-05,
-      "loss": 0.1965,
-      "step": 80
-    },
-    {
-      "epoch": 2.6666666666666665,
-      "grad_norm": 4.452145099639893,
-      "learning_rate": 2.533783783783784e-05,
-      "loss": 0.1428,
-      "step": 90
-    },
-    {
-      "epoch": 2.962962962962963,
-      "grad_norm": 8.201363563537598,
-      "learning_rate": 2.195945945945946e-05,
-      "loss": 0.1411,
-      "step": 100
-    },
-    {
-      "epoch": 2.9925925925925925,
-      "eval_accuracy": 0.9539748953974896,
-      "eval_loss": 0.13948805630207062,
-      "eval_runtime": 2.1138,
-      "eval_samples_per_second": 226.133,
-      "eval_steps_per_second": 7.096,
-      "step": 101
-    },
-    {
-      "epoch": 3.259259259259259,
-      "grad_norm": 4.830763816833496,
-      "learning_rate": 1.8581081081081082e-05,
-      "loss": 0.1161,
-      "step": 110
-    },
-    {
-      "epoch": 3.5555555555555554,
-      "grad_norm": 10.040033340454102,
-      "learning_rate": 1.5202702702702704e-05,
-      "loss": 0.1191,
-      "step": 120
-    },
-    {
-      "epoch": 3.851851851851852,
-      "grad_norm": 8.36621379852295,
-      "learning_rate": 1.1824324324324325e-05,
-      "loss": 0.128,
-      "step": 130
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.9686192468619247,
-      "eval_loss": 0.08668407797813416,
-      "eval_runtime": 2.1145,
-      "eval_samples_per_second": 226.057,
-      "eval_steps_per_second": 7.094,
-      "step": 135
-    },
-    {
-      "epoch": 4.148148148148148,
-      "grad_norm": 6.916751861572266,
-      "learning_rate": 8.445945945945946e-06,
-      "loss": 0.0959,
-      "step": 140
-    },
-    {
-      "epoch": 4.444444444444445,
-      "grad_norm": 7.64593505859375,
-      "learning_rate": 5.067567567567568e-06,
-      "loss": 0.1115,
-      "step": 150
-    },
-    {
-      "epoch": 4.7407407407407405,
-      "grad_norm": 4.8151984214782715,
-      "learning_rate": 1.6891891891891894e-06,
-      "loss": 0.1108,
-      "step": 160
-    },
-    {
-      "epoch": 4.888888888888889,
-      "eval_accuracy": 0.9539748953974896,
-      "eval_loss": 0.12008119374513626,
-      "eval_runtime": 2.2552,
-      "eval_samples_per_second": 211.955,
-      "eval_steps_per_second": 6.651,
-      "step": 165
-    },
-    {
-      "epoch": 4.888888888888889,
-      "step": 165,
-      "total_flos": 5.227711603505234e+17,
-      "train_loss": 0.22451729810599125,
-      "train_runtime": 231.4165,
-      "train_samples_per_second": 92.863,
-      "train_steps_per_second": 0.713
+      "epoch": 1.9555555555555557,
+      "step": 66,
+      "total_flos": 2.0943656319482266e+17,
+      "train_loss": 0.3058708075321082,
+      "train_runtime": 94.9406,
+      "train_samples_per_second": 90.541,
+      "train_steps_per_second": 0.695
     }
   ],
   "logging_steps": 10,
-  "max_steps": 165,
+  "max_steps": 66,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 2,
   "save_steps": 500,
   "stateful_callbacks": {
     "CustomEarlyStoppingCallback": {
@@ -201,7 +104,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5.227711603505234e+17,
+  "total_flos": 2.0943656319482266e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null