Training in progress, epoch 0

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 4.938271604938271,
-    "eval_accuracy": 0.9616724738675958,
-    "eval_loss": 0.11273417621850967,
-    "eval_runtime": 2.4706,
-    "eval_samples_per_second": 116.166,
-    "eval_steps_per_second": 7.286,
+    "eval_accuracy": 0.9156020131629887,
+    "eval_loss": 0.22556395828723907,
+    "eval_runtime": 22.7546,
+    "eval_samples_per_second": 113.516,
+    "eval_steps_per_second": 3.56,
     "total_flos": 3.1727957353537536e+17,
-    "train_loss": 0.06341908916831017,
-    "train_runtime": 245.9692,
-    "train_samples_per_second": 52.507,
-    "train_steps_per_second": 0.813
+    "train_loss": 0.4883076953887939,
+    "train_runtime": 265.3223,
+    "train_samples_per_second": 48.677,
+    "train_steps_per_second": 0.377
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.938271604938271,
-    "eval_accuracy": 0.9616724738675958,
-    "eval_loss": 0.11273417621850967,
-    "eval_runtime": 2.4706,
-    "eval_samples_per_second": 116.166,
-    "eval_steps_per_second": 7.286
+    "eval_accuracy": 0.9372822299651568,
+    "eval_loss": 0.1715514212846756,
+    "eval_runtime": 2.7579,
+    "eval_samples_per_second": 104.066,
+    "eval_steps_per_second": 3.263
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:035707ba7def84500aaa8a6501235bc2da59c7aa8f09dcb90e5701c17922b7e4
+oid sha256:c78aab4276c550f24351312fae389bc05b9ce88326f8ed0162336967cb32961e
 size 110348984

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.938271604938271,
-    "total_flos": 3.1727957353537536e+17,
-    "train_loss": 0.06341908916831017,
-    "train_runtime": 245.9692,
-    "train_samples_per_second": 52.507,
-    "train_steps_per_second": 0.813
+    "eval_accuracy": 0.9156020131629887,
+    "eval_loss": 0.22556395828723907,
+    "eval_runtime": 22.7546,
+    "eval_samples_per_second": 113.516,
+    "eval_steps_per_second": 3.56
 }

trainer_state.json

@@ -1,210 +1,158 @@
 {
-  "best_metric": 0.9651567944250871,
-  "best_model_checkpoint": "brain-tumor-classification/checkpoint-200",
+  "best_metric": 0.9372822299651568,
+  "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": 0.24691358024691357,
-      "grad_norm": 9.583415985107422,
-      "learning_rate": 2.5e-05,
-      "loss": 0.0798,
-      "step": 10
-    },
     {
       "epoch": 0.49382716049382713,
-      "grad_norm": 4.763514518737793,
+      "grad_norm": 11.951680183410645,
       "learning_rate": 5e-05,
-      "loss": 0.0516,
-      "step": 20
-    },
-    {
-      "epoch": 0.7407407407407407,
-      "grad_norm": 3.712924003601074,
-      "learning_rate": 4.722222222222222e-05,
-      "loss": 0.0708,
-      "step": 30
+      "loss": 1.2238,
+      "step": 10
     },
     {
       "epoch": 0.9876543209876543,
-      "grad_norm": 3.5341458320617676,
+      "grad_norm": 11.546021461486816,
       "learning_rate": 4.4444444444444447e-05,
-      "loss": 0.0628,
-      "step": 40
+      "loss": 0.8509,
+      "step": 20
     },
     {
       "epoch": 0.9876543209876543,
-      "eval_accuracy": 0.9512195121951219,
-      "eval_loss": 0.14935272932052612,
-      "eval_runtime": 2.3845,
-      "eval_samples_per_second": 120.36,
-      "eval_steps_per_second": 7.549,
-      "step": 40
-    },
-    {
-      "epoch": 1.2345679012345678,
-      "grad_norm": 2.3918871879577637,
-      "learning_rate": 4.166666666666667e-05,
-      "loss": 0.0562,
-      "step": 50
+      "eval_accuracy": 0.8466898954703833,
+      "eval_loss": 0.5305207967758179,
+      "eval_runtime": 2.7821,
+      "eval_samples_per_second": 103.16,
+      "eval_steps_per_second": 3.235,
+      "step": 20
     },
     {
       "epoch": 1.4814814814814814,
-      "grad_norm": 7.054160118103027,
+      "grad_norm": 13.172475814819336,
       "learning_rate": 3.888888888888889e-05,
-      "loss": 0.0587,
-      "step": 60
-    },
-    {
-      "epoch": 1.7283950617283952,
-      "grad_norm": 16.58973503112793,
-      "learning_rate": 3.611111111111111e-05,
-      "loss": 0.0679,
-      "step": 70
+      "loss": 0.5721,
+      "step": 30
     },
     {
       "epoch": 1.9753086419753085,
-      "grad_norm": 11.970051765441895,
+      "grad_norm": 7.819632530212402,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.0499,
-      "step": 80
-    },
-    {
-      "epoch": 2.0,
-      "eval_accuracy": 0.9372822299651568,
-      "eval_loss": 0.18301770091056824,
-      "eval_runtime": 2.3848,
-      "eval_samples_per_second": 120.346,
-      "eval_steps_per_second": 7.548,
-      "step": 81
+      "loss": 0.4478,
+      "step": 40
     },
     {
-      "epoch": 2.2222222222222223,
-      "grad_norm": 6.739134788513184,
-      "learning_rate": 3.055555555555556e-05,
-      "loss": 0.0352,
-      "step": 90
+      "epoch": 1.9753086419753085,
+      "eval_accuracy": 0.9094076655052264,
+      "eval_loss": 0.3092304468154907,
+      "eval_runtime": 2.5037,
+      "eval_samples_per_second": 114.631,
+      "eval_steps_per_second": 3.595,
+      "step": 40
     },
     {
       "epoch": 2.4691358024691357,
-      "grad_norm": 23.700721740722656,
+      "grad_norm": 10.437914848327637,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 0.1,
-      "step": 100
-    },
-    {
-      "epoch": 2.7160493827160495,
-      "grad_norm": 6.7289581298828125,
-      "learning_rate": 2.5e-05,
-      "loss": 0.0504,
-      "step": 110
+      "loss": 0.3668,
+      "step": 50
     },
     {
       "epoch": 2.962962962962963,
-      "grad_norm": 11.606633186340332,
+      "grad_norm": 9.35677433013916,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.0697,
-      "step": 120
-    },
-    {
-      "epoch": 2.9876543209876543,
-      "eval_accuracy": 0.9547038327526133,
-      "eval_loss": 0.10937974601984024,
-      "eval_runtime": 2.9179,
-      "eval_samples_per_second": 98.36,
-      "eval_steps_per_second": 6.169,
-      "step": 121
+      "loss": 0.3313,
+      "step": 60
     },
     {
-      "epoch": 3.2098765432098766,
-      "grad_norm": 17.90212059020996,
-      "learning_rate": 1.9444444444444445e-05,
-      "loss": 0.0546,
-      "step": 130
+      "epoch": 2.962962962962963,
+      "eval_accuracy": 0.9233449477351916,
+      "eval_loss": 0.24220147728919983,
+      "eval_runtime": 2.4542,
+      "eval_samples_per_second": 116.944,
+      "eval_steps_per_second": 3.667,
+      "step": 60
     },
     {
       "epoch": 3.45679012345679,
-      "grad_norm": 8.192621231079102,
+      "grad_norm": 8.183242797851562,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.0601,
-      "step": 140
-    },
-    {
-      "epoch": 3.7037037037037037,
-      "grad_norm": 3.0072553157806396,
-      "learning_rate": 1.388888888888889e-05,
-      "loss": 0.07,
-      "step": 150
+      "loss": 0.2979,
+      "step": 70
     },
     {
       "epoch": 3.950617283950617,
-      "grad_norm": 5.344573974609375,
+      "grad_norm": 8.208118438720703,
       "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.0646,
-      "step": 160
+      "loss": 0.2777,
+      "step": 80
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9581881533101045,
-      "eval_loss": 0.13176245987415314,
-      "eval_runtime": 2.8306,
-      "eval_samples_per_second": 101.391,
-      "eval_steps_per_second": 6.359,
-      "step": 162
-    },
-    {
-      "epoch": 4.197530864197531,
-      "grad_norm": 12.885249137878418,
-      "learning_rate": 8.333333333333334e-06,
-      "loss": 0.0515,
-      "step": 170
+      "eval_accuracy": 0.9372822299651568,
+      "eval_loss": 0.1715514212846756,
+      "eval_runtime": 2.7133,
+      "eval_samples_per_second": 105.776,
+      "eval_steps_per_second": 3.317,
+      "step": 81
     },
     {
       "epoch": 4.444444444444445,
-      "grad_norm": 13.122868537902832,
+      "grad_norm": 8.672185897827148,
       "learning_rate": 5.555555555555556e-06,
-      "loss": 0.0784,
-      "step": 180
-    },
-    {
-      "epoch": 4.6913580246913575,
-      "grad_norm": 8.814871788024902,
-      "learning_rate": 2.777777777777778e-06,
-      "loss": 0.088,
-      "step": 190
+      "loss": 0.2684,
+      "step": 90
     },
     {
       "epoch": 4.938271604938271,
-      "grad_norm": 1.9182705879211426,
+      "grad_norm": 10.231292724609375,
       "learning_rate": 0.0,
-      "loss": 0.0481,
-      "step": 200
+      "loss": 0.2465,
+      "step": 100
     },
     {
       "epoch": 4.938271604938271,
-      "eval_accuracy": 0.9651567944250871,
-      "eval_loss": 0.11668162047863007,
-      "eval_runtime": 2.6735,
-      "eval_samples_per_second": 107.351,
-      "eval_steps_per_second": 6.733,
-      "step": 200
+      "eval_accuracy": 0.9372822299651568,
+      "eval_loss": 0.16363371908664703,
+      "eval_runtime": 2.7577,
+      "eval_samples_per_second": 104.074,
+      "eval_steps_per_second": 3.264,
+      "step": 100
     },
     {
       "epoch": 4.938271604938271,
-      "step": 200,
+      "step": 100,
       "total_flos": 3.1727957353537536e+17,
-      "train_loss": 0.06341908916831017,
-      "train_runtime": 245.9692,
-      "train_samples_per_second": 52.507,
-      "train_steps_per_second": 0.813
+      "train_loss": 0.4883076953887939,
+      "train_runtime": 265.3223,
+      "train_samples_per_second": 48.677,
+      "train_steps_per_second": 0.377
+    },
+    {
+      "epoch": 4.938271604938271,
+      "eval_accuracy": 0.9372822299651568,
+      "eval_loss": 0.1715514212846756,
+      "eval_runtime": 2.7579,
+      "eval_samples_per_second": 104.066,
+      "eval_steps_per_second": 3.263,
+      "step": 100
+    },
+    {
+      "epoch": 4.938271604938271,
+      "eval_accuracy": 0.9156020131629887,
+      "eval_loss": 0.22556395828723907,
+      "eval_runtime": 22.7546,
+      "eval_samples_per_second": 113.516,
+      "eval_steps_per_second": 3.56,
+      "step": 100
     }
   ],
   "logging_steps": 10,
-  "max_steps": 200,
+  "max_steps": 100,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
   "save_steps": 500,
@@ -221,7 +169,7 @@
     }
   },
   "total_flos": 3.1727957353537536e+17,
-  "train_batch_size": 16,
+  "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7ac1b4cd2d8159af1fc8fd51f45445f24aa9bf4ee3eaf6c19dbdcdb4ce9f1e54
+oid sha256:9df4ade6f59d89ee284f128277b27cd59499314a50c762b40744a848779e24e7
 size 5240