End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 3.0,
-    "eval_accuracy": 0.8685015290519877,
-    "eval_loss": 0.34957513213157654,
-    "eval_runtime": 2.9329,
-    "eval_samples_per_second": 111.492,
-    "eval_steps_per_second": 3.751,
-    "total_flos": 2.2838958959910912e+17,
-    "train_loss": 0.41367715683536255,
-    "train_runtime": 173.7772,
-    "train_samples_per_second": 50.703,
-    "train_steps_per_second": 0.397
+    "epoch": 4.938271604938271,
+    "eval_accuracy": 0.9477351916376306,
+    "eval_loss": 0.11578787863254547,
+    "eval_runtime": 2.7445,
+    "eval_samples_per_second": 104.573,
+    "eval_steps_per_second": 6.559,
+    "total_flos": 3.1727957353537536e+17,
+    "train_loss": 0.41540566325187683,
+    "train_runtime": 248.6508,
+    "train_samples_per_second": 51.94,
+    "train_steps_per_second": 0.804
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "eval_accuracy": 0.8685015290519877,
-    "eval_loss": 0.34957513213157654,
-    "eval_runtime": 2.9329,
-    "eval_samples_per_second": 111.492,
-    "eval_steps_per_second": 3.751
+    "epoch": 4.938271604938271,
+    "eval_accuracy": 0.9477351916376306,
+    "eval_loss": 0.11578787863254547,
+    "eval_runtime": 2.7445,
+    "eval_samples_per_second": 104.573,
+    "eval_steps_per_second": 6.559
 }

runs/Oct10_11-22-23_e11ecca18f17/events.out.tfevents.1728559702.e11ecca18f17.609.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06fbe26e63fdc02be19f4f2eabd2f1523e7ae6faeeb6890b302671a6b7c14b13
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "total_flos": 2.2838958959910912e+17,
-    "train_loss": 0.41367715683536255,
-    "train_runtime": 173.7772,
-    "train_samples_per_second": 50.703,
-    "train_steps_per_second": 0.397
+    "epoch": 4.938271604938271,
+    "total_flos": 3.1727957353537536e+17,
+    "train_loss": 0.41540566325187683,
+    "train_runtime": 248.6508,
+    "train_samples_per_second": 51.94,
+    "train_steps_per_second": 0.804
 }

trainer_state.json

@@ -1,96 +1,212 @@
 {
-  "best_metric": 0.8685015290519877,
-  "best_model_checkpoint": "brain-tumor-classification/checkpoint-69",
-  "epoch": 3.0,
+  "best_metric": 0.9477351916376306,
+  "best_model_checkpoint": "brain-tumor-classification/checkpoint-200",
+  "epoch": 4.938271604938271,
   "eval_steps": 500,
-  "global_step": 69,
+  "global_step": 200,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.43478260869565216,
-      "grad_norm": 7.908116340637207,
-      "learning_rate": 4.7580645161290326e-05,
-      "loss": 0.623,
+      "epoch": 0.24691358024691357,
+      "grad_norm": 11.865983009338379,
+      "learning_rate": 2.5e-05,
+      "loss": 1.335,
       "step": 10
     },
     {
-      "epoch": 0.8695652173913043,
-      "grad_norm": 10.378320693969727,
-      "learning_rate": 3.951612903225806e-05,
-      "loss": 0.5239,
+      "epoch": 0.49382716049382713,
+      "grad_norm": 14.611420631408691,
+      "learning_rate": 5e-05,
+      "loss": 1.0649,
       "step": 20
     },
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.7889908256880734,
-      "eval_loss": 0.5283120274543762,
-      "eval_runtime": 3.5122,
-      "eval_samples_per_second": 93.103,
-      "eval_steps_per_second": 3.132,
-      "step": 23
+      "epoch": 0.7407407407407407,
+      "grad_norm": 14.365787506103516,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 0.7611,
+      "step": 30
     },
     {
-      "epoch": 1.3043478260869565,
-      "grad_norm": 5.665172100067139,
-      "learning_rate": 3.1451612903225806e-05,
-      "loss": 0.4202,
-      "step": 30
+      "epoch": 0.9876543209876543,
+      "grad_norm": 16.896533966064453,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5761,
+      "step": 40
     },
     {
-      "epoch": 1.7391304347826086,
-      "grad_norm": 9.73236083984375,
-      "learning_rate": 2.338709677419355e-05,
-      "loss": 0.3716,
+      "epoch": 0.9876543209876543,
+      "eval_accuracy": 0.8432055749128919,
+      "eval_loss": 0.41130325198173523,
+      "eval_runtime": 2.4201,
+      "eval_samples_per_second": 118.592,
+      "eval_steps_per_second": 7.438,
       "step": 40
     },
     {
-      "epoch": 2.0,
-      "eval_accuracy": 0.8409785932721713,
-      "eval_loss": 0.3933815658092499,
-      "eval_runtime": 2.873,
-      "eval_samples_per_second": 113.816,
-      "eval_steps_per_second": 3.829,
-      "step": 46
-    },
-    {
-      "epoch": 2.1739130434782608,
-      "grad_norm": 8.678784370422363,
-      "learning_rate": 1.5322580645161292e-05,
-      "loss": 0.3549,
+      "epoch": 1.2345679012345678,
+      "grad_norm": 9.885971069335938,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.5027,
       "step": 50
     },
     {
-      "epoch": 2.608695652173913,
-      "grad_norm": 5.479146957397461,
-      "learning_rate": 7.258064516129033e-06,
-      "loss": 0.2964,
+      "epoch": 1.4814814814814814,
+      "grad_norm": 19.173418045043945,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.4857,
       "step": 60
     },
     {
-      "epoch": 3.0,
-      "eval_accuracy": 0.8685015290519877,
-      "eval_loss": 0.34957513213157654,
-      "eval_runtime": 3.2055,
-      "eval_samples_per_second": 102.012,
-      "eval_steps_per_second": 3.432,
-      "step": 69
+      "epoch": 1.7283950617283952,
+      "grad_norm": 10.684455871582031,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 0.4252,
+      "step": 70
+    },
+    {
+      "epoch": 1.9753086419753085,
+      "grad_norm": 11.967300415039062,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3871,
+      "step": 80
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9024390243902439,
+      "eval_loss": 0.25702860951423645,
+      "eval_runtime": 3.0171,
+      "eval_samples_per_second": 95.125,
+      "eval_steps_per_second": 5.966,
+      "step": 81
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 7.449192047119141,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.294,
+      "step": 90
+    },
+    {
+      "epoch": 2.4691358024691357,
+      "grad_norm": 15.003212928771973,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.3583,
+      "step": 100
+    },
+    {
+      "epoch": 2.7160493827160495,
+      "grad_norm": 12.026302337646484,
+      "learning_rate": 2.5e-05,
+      "loss": 0.2584,
+      "step": 110
+    },
+    {
+      "epoch": 2.962962962962963,
+      "grad_norm": 10.30452823638916,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.2586,
+      "step": 120
+    },
+    {
+      "epoch": 2.9876543209876543,
+      "eval_accuracy": 0.9407665505226481,
+      "eval_loss": 0.1910204291343689,
+      "eval_runtime": 2.8228,
+      "eval_samples_per_second": 101.672,
+      "eval_steps_per_second": 6.377,
+      "step": 121
+    },
+    {
+      "epoch": 3.2098765432098766,
+      "grad_norm": 9.547262191772461,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.1965,
+      "step": 130
+    },
+    {
+      "epoch": 3.45679012345679,
+      "grad_norm": 8.193156242370605,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2272,
+      "step": 140
+    },
+    {
+      "epoch": 3.7037037037037037,
+      "grad_norm": 6.964083671569824,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.237,
+      "step": 150
+    },
+    {
+      "epoch": 3.950617283950617,
+      "grad_norm": 15.732324600219727,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.2164,
+      "step": 160
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9442508710801394,
+      "eval_loss": 0.13123387098312378,
+      "eval_runtime": 2.4291,
+      "eval_samples_per_second": 118.153,
+      "eval_steps_per_second": 7.41,
+      "step": 162
+    },
+    {
+      "epoch": 4.197530864197531,
+      "grad_norm": 7.277712821960449,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.1614,
+      "step": 170
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 13.366209030151367,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.1978,
+      "step": 180
+    },
+    {
+      "epoch": 4.6913580246913575,
+      "grad_norm": 9.085039138793945,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 0.189,
+      "step": 190
+    },
+    {
+      "epoch": 4.938271604938271,
+      "grad_norm": 7.736137866973877,
+      "learning_rate": 0.0,
+      "loss": 0.1757,
+      "step": 200
+    },
+    {
+      "epoch": 4.938271604938271,
+      "eval_accuracy": 0.9477351916376306,
+      "eval_loss": 0.11578787863254547,
+      "eval_runtime": 2.7368,
+      "eval_samples_per_second": 104.868,
+      "eval_steps_per_second": 6.577,
+      "step": 200
     },
     {
-      "epoch": 3.0,
-      "step": 69,
-      "total_flos": 2.2838958959910912e+17,
-      "train_loss": 0.41367715683536255,
-      "train_runtime": 173.7772,
-      "train_samples_per_second": 50.703,
-      "train_steps_per_second": 0.397
+      "epoch": 4.938271604938271,
+      "step": 200,
+      "total_flos": 3.1727957353537536e+17,
+      "train_loss": 0.41540566325187683,
+      "train_runtime": 248.6508,
+      "train_samples_per_second": 51.94,
+      "train_steps_per_second": 0.804
     }
   ],
   "logging_steps": 10,
-  "max_steps": 69,
+  "max_steps": 200,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -104,8 +220,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.2838958959910912e+17,
-  "train_batch_size": 32,
+  "total_flos": 3.1727957353537536e+17,
+  "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null
 }