Training in progress, epoch 0

all_results.json

@@ -1,13 +1,8 @@
 {
-    "epoch": 9.876543209876543,
-    "eval_accuracy": 0.9721254355400697,
-    "eval_loss": 0.07933783531188965,
-    "eval_runtime": 2.8381,
-    "eval_samples_per_second": 101.125,
-    "eval_steps_per_second": 3.171,
-    "total_flos": 6.343354306682266e+17,
-    "train_loss": 0.2855896496772766,
-    "train_runtime": 464.8772,
-    "train_samples_per_second": 55.563,
-    "train_steps_per_second": 0.43
+    "epoch": 4.938271604938271,
+    "total_flos": 3.1727957353537536e+17,
+    "train_loss": 0.5141408157348633,
+    "train_runtime": 269.7071,
+    "train_samples_per_second": 47.885,
+    "train_steps_per_second": 0.371
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:064f351ddf553d481d1a221ebc9733692dde775d77133e068606f6ddf69ecfed
+oid sha256:d5671ef3cdb7c0d6b39d5800b57e58f8c339a64fc660900287c5c8926e7490cb
 size 110348984

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.876543209876543,
-    "eval_accuracy": 0.9519938056523423,
-    "eval_loss": 0.1284445822238922,
-    "eval_runtime": 22.1837,
-    "eval_samples_per_second": 116.437,
-    "eval_steps_per_second": 3.651
+    "epoch": 4.938271604938271,
+    "total_flos": 3.1727957353537536e+17,
+    "train_loss": 0.5141408157348633,
+    "train_runtime": 269.7071,
+    "train_samples_per_second": 47.885,
+    "train_steps_per_second": 0.371
 }

trainer_state.json

@@ -1,223 +1,107 @@
 {
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 9.876543209876543,
+  "best_metric": 0.9128919860627178,
+  "best_model_checkpoint": "brain-tumor-classification/checkpoint-81",
+  "epoch": 4.938271604938271,
   "eval_steps": 500,
-  "global_step": 200,
+  "global_step": 100,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.9876543209876543,
-      "grad_norm": 8.28109359741211,
-      "learning_rate": 4.5e-05,
-      "loss": 0.8062,
+      "grad_norm": 13.366371154785156,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.0827,
       "step": 20
     },
     {
       "epoch": 0.9876543209876543,
-      "eval_accuracy": 0.818815331010453,
-      "eval_loss": 0.4432358145713806,
-      "eval_runtime": 2.3752,
-      "eval_samples_per_second": 120.833,
-      "eval_steps_per_second": 3.789,
+      "eval_accuracy": 0.7630662020905923,
+      "eval_loss": 0.5737153887748718,
+      "eval_runtime": 2.4539,
+      "eval_samples_per_second": 116.958,
+      "eval_steps_per_second": 3.668,
       "step": 20
     },
     {
       "epoch": 1.9753086419753085,
-      "grad_norm": 9.780616760253906,
-      "learning_rate": 4e-05,
-      "loss": 0.4153,
+      "grad_norm": 7.611490249633789,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5357,
       "step": 40
     },
     {
       "epoch": 1.9753086419753085,
-      "eval_accuracy": 0.8536585365853658,
-      "eval_loss": 0.3407208323478699,
-      "eval_runtime": 2.4373,
-      "eval_samples_per_second": 117.752,
-      "eval_steps_per_second": 3.693,
+      "eval_accuracy": 0.8641114982578397,
+      "eval_loss": 0.3689139187335968,
+      "eval_runtime": 2.4398,
+      "eval_samples_per_second": 117.633,
+      "eval_steps_per_second": 3.689,
       "step": 40
     },
     {
       "epoch": 2.962962962962963,
-      "grad_norm": 6.261844635009766,
-      "learning_rate": 3.5e-05,
-      "loss": 0.3213,
+      "grad_norm": 7.663794040679932,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.3875,
       "step": 60
     },
     {
       "epoch": 2.962962962962963,
-      "eval_accuracy": 0.9372822299651568,
-      "eval_loss": 0.1876000016927719,
-      "eval_runtime": 2.4811,
-      "eval_samples_per_second": 115.675,
-      "eval_steps_per_second": 3.627,
+      "eval_accuracy": 0.8954703832752613,
+      "eval_loss": 0.29916301369667053,
+      "eval_runtime": 2.5481,
+      "eval_samples_per_second": 112.634,
+      "eval_steps_per_second": 3.532,
       "step": 60
     },
     {
       "epoch": 4.0,
-      "grad_norm": 10.462350845336914,
-      "learning_rate": 2.975e-05,
-      "loss": 0.2633,
+      "grad_norm": 14.111477851867676,
+      "learning_rate": 1.0555555555555555e-05,
+      "loss": 0.2921,
       "step": 81
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9442508710801394,
-      "eval_loss": 0.15536989271640778,
-      "eval_runtime": 2.8656,
-      "eval_samples_per_second": 100.154,
-      "eval_steps_per_second": 3.141,
+      "eval_accuracy": 0.9128919860627178,
+      "eval_loss": 0.27263280749320984,
+      "eval_runtime": 2.6327,
+      "eval_samples_per_second": 109.015,
+      "eval_steps_per_second": 3.419,
       "step": 81
     },
     {
-      "epoch": 4.987654320987654,
-      "grad_norm": 6.831620693206787,
-      "learning_rate": 2.4750000000000002e-05,
-      "loss": 0.2201,
-      "step": 101
-    },
-    {
-      "epoch": 4.987654320987654,
-      "eval_accuracy": 0.9547038327526133,
-      "eval_loss": 0.13280798494815826,
-      "eval_runtime": 2.4786,
-      "eval_samples_per_second": 115.792,
-      "eval_steps_per_second": 3.631,
-      "step": 101
-    },
-    {
-      "epoch": 5.9753086419753085,
-      "grad_norm": 8.320969581604004,
-      "learning_rate": 1.9750000000000002e-05,
-      "loss": 0.2087,
-      "step": 121
-    },
-    {
-      "epoch": 5.9753086419753085,
-      "eval_accuracy": 0.9721254355400697,
-      "eval_loss": 0.08554696291685104,
-      "eval_runtime": 2.6901,
-      "eval_samples_per_second": 106.686,
-      "eval_steps_per_second": 3.346,
-      "step": 121
-    },
-    {
-      "epoch": 6.962962962962963,
-      "grad_norm": 5.462257385253906,
-      "learning_rate": 1.475e-05,
-      "loss": 0.1797,
-      "step": 141
-    },
-    {
-      "epoch": 6.962962962962963,
-      "eval_accuracy": 0.9442508710801394,
-      "eval_loss": 0.12809309363365173,
-      "eval_runtime": 2.4948,
-      "eval_samples_per_second": 115.041,
-      "eval_steps_per_second": 3.608,
-      "step": 141
-    },
-    {
-      "epoch": 8.0,
-      "grad_norm": 6.069087982177734,
-      "learning_rate": 9.5e-06,
-      "loss": 0.1478,
-      "step": 162
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.9721254355400697,
-      "eval_loss": 0.08397921919822693,
-      "eval_runtime": 2.5152,
-      "eval_samples_per_second": 114.108,
-      "eval_steps_per_second": 3.578,
-      "step": 162
-    },
-    {
-      "epoch": 8.987654320987655,
-      "grad_norm": 12.428985595703125,
-      "learning_rate": 4.5e-06,
-      "loss": 0.1545,
-      "step": 182
-    },
-    {
-      "epoch": 8.987654320987655,
-      "eval_accuracy": 0.9686411149825784,
-      "eval_loss": 0.08367497473955154,
-      "eval_runtime": 2.4429,
-      "eval_samples_per_second": 117.485,
-      "eval_steps_per_second": 3.684,
-      "step": 182
-    },
-    {
-      "epoch": 9.876543209876543,
-      "grad_norm": 2.292888641357422,
+      "epoch": 4.938271604938271,
+      "grad_norm": 9.39560317993164,
       "learning_rate": 0.0,
-      "loss": 0.1315,
-      "step": 200
-    },
-    {
-      "epoch": 9.876543209876543,
-      "eval_accuracy": 0.9721254355400697,
-      "eval_loss": 0.07933783531188965,
-      "eval_runtime": 2.7805,
-      "eval_samples_per_second": 103.217,
-      "eval_steps_per_second": 3.237,
-      "step": 200
-    },
-    {
-      "epoch": 9.876543209876543,
-      "step": 200,
-      "total_flos": 6.343354306682266e+17,
-      "train_loss": 0.2855896496772766,
-      "train_runtime": 464.8772,
-      "train_samples_per_second": 55.563,
-      "train_steps_per_second": 0.43
-    },
-    {
-      "epoch": 9.876543209876543,
-      "eval_accuracy": 0.9562524196670538,
-      "eval_loss": 0.11971130222082138,
-      "eval_runtime": 24.8934,
-      "eval_samples_per_second": 103.763,
-      "eval_steps_per_second": 3.254,
-      "step": 200
-    },
-    {
-      "epoch": 9.876543209876543,
-      "eval_accuracy": 0.9562524196670538,
-      "eval_loss": 0.12554492056369781,
-      "eval_runtime": 27.9535,
-      "eval_samples_per_second": 92.403,
-      "eval_steps_per_second": 2.898,
-      "step": 200
-    },
-    {
-      "epoch": 9.876543209876543,
-      "eval_accuracy": 0.9721254355400697,
-      "eval_loss": 0.07933783531188965,
-      "eval_runtime": 2.5048,
-      "eval_samples_per_second": 114.582,
-      "eval_steps_per_second": 3.593,
-      "step": 200
-    },
-    {
-      "epoch": 9.876543209876543,
-      "eval_accuracy": 0.9519938056523423,
-      "eval_loss": 0.1284445822238922,
-      "eval_runtime": 22.1837,
-      "eval_samples_per_second": 116.437,
-      "eval_steps_per_second": 3.651,
-      "step": 200
+      "loss": 0.2716,
+      "step": 100
+    },
+    {
+      "epoch": 4.938271604938271,
+      "eval_accuracy": 0.9128919860627178,
+      "eval_loss": 0.23185838758945465,
+      "eval_runtime": 3.5061,
+      "eval_samples_per_second": 81.858,
+      "eval_steps_per_second": 2.567,
+      "step": 100
+    },
+    {
+      "epoch": 4.938271604938271,
+      "step": 100,
+      "total_flos": 3.1727957353537536e+17,
+      "train_loss": 0.5141408157348633,
+      "train_runtime": 269.7071,
+      "train_samples_per_second": 47.885,
+      "train_steps_per_second": 0.371
     }
   ],
-  "logging_steps": 10,
-  "max_steps": 200,
+  "logging_steps": 500,
+  "max_steps": 100,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -231,7 +115,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 6.343354306682266e+17,
+  "total_flos": 3.1727957353537536e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null