End of training

README.md

@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0620
-- Accuracy: 0.9778
+- Loss: 0.0472
+- Accuracy: 0.9852
 
 ## Model description
 

all_results.json

@@ -1,8 +1,13 @@
 {
-    "epoch": 8.0,
-    "total_flos": 3.121999450369229e+16,
-    "train_loss": 0.3916594386100769,
-    "train_runtime": 22.6465,
-    "train_samples_per_second": 69.326,
-    "train_steps_per_second": 0.442
+    "epoch": 10.0,
+    "eval_accuracy": 0.9851851851851852,
+    "eval_loss": 0.04721328616142273,
+    "eval_runtime": 0.7379,
+    "eval_samples_per_second": 182.957,
+    "eval_steps_per_second": 6.776,
+    "total_flos": 1.565971061889024e+17,
+    "train_loss": 0.394845541715622,
+    "train_runtime": 86.4231,
+    "train_samples_per_second": 72.897,
+    "train_steps_per_second": 0.579
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 8.0,
-    "eval_accuracy": 0.9090909090909091,
-    "eval_loss": 0.5003632307052612,
-    "eval_runtime": 0.1947,
-    "eval_samples_per_second": 169.491,
-    "eval_steps_per_second": 10.272
+    "epoch": 10.0,
+    "eval_accuracy": 0.9851851851851852,
+    "eval_loss": 0.05875137820839882,
+    "eval_runtime": 0.7872,
+    "eval_samples_per_second": 171.496,
+    "eval_steps_per_second": 6.352
 }

test_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 8.0,
-    "eval_accuracy": 0.8484848484848485,
-    "eval_loss": 0.5320981740951538,
-    "eval_runtime": 0.1987,
-    "eval_samples_per_second": 166.048,
-    "eval_steps_per_second": 10.064
+    "epoch": 10.0,
+    "eval_accuracy": 0.9851851851851852,
+    "eval_loss": 0.04721328616142273,
+    "eval_runtime": 0.7379,
+    "eval_samples_per_second": 182.957,
+    "eval_steps_per_second": 6.776
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 8.0,
-    "total_flos": 3.121999450369229e+16,
-    "train_loss": 0.3916594386100769,
-    "train_runtime": 22.6465,
-    "train_samples_per_second": 69.326,
-    "train_steps_per_second": 0.442
+    "epoch": 10.0,
+    "total_flos": 1.565971061889024e+17,
+    "train_loss": 0.394845541715622,
+    "train_runtime": 86.4231,
+    "train_samples_per_second": 72.897,
+    "train_steps_per_second": 0.579
 }

trainer_state.json

@@ -1,153 +1,185 @@
 {
-  "best_metric": 0.9696969696969697,
-  "best_model_checkpoint": "/kaggle/working/swin-brain-plane-classification/checkpoint-6",
-  "epoch": 8.0,
+  "best_metric": 0.9851851851851852,
+  "best_model_checkpoint": "/kaggle/working/swin-brain-plane-classification/checkpoint-45",
+  "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 10,
+  "global_step": 50,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.8,
-      "grad_norm": 4.3462605476379395,
+      "epoch": 1.0,
+      "grad_norm": 5.199264049530029,
       "learning_rate": 5e-05,
-      "loss": 0.7819,
-      "step": 1
+      "loss": 1.1833,
+      "step": 5
     },
     {
-      "epoch": 0.8,
-      "eval_accuracy": 0.8484848484848485,
-      "eval_loss": 0.5965157151222229,
-      "eval_runtime": 0.2012,
-      "eval_samples_per_second": 164.047,
-      "eval_steps_per_second": 9.942,
-      "step": 1
+      "epoch": 1.0,
+      "eval_accuracy": 0.6074074074074074,
+      "eval_loss": 0.9675570130348206,
+      "eval_runtime": 0.7682,
+      "eval_samples_per_second": 175.727,
+      "eval_steps_per_second": 6.508,
+      "step": 5
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 4.70327091217041,
+      "epoch": 2.0,
+      "grad_norm": 5.081261157989502,
       "learning_rate": 4.4444444444444447e-05,
-      "loss": 0.689,
-      "step": 2
+      "loss": 0.8399,
+      "step": 10
     },
     {
-      "epoch": 1.6,
-      "eval_accuracy": 0.8484848484848485,
-      "eval_loss": 0.5047040581703186,
-      "eval_runtime": 0.1988,
-      "eval_samples_per_second": 165.985,
-      "eval_steps_per_second": 10.06,
-      "step": 2
+      "epoch": 2.0,
+      "eval_accuracy": 0.8592592592592593,
+      "eval_loss": 0.5346508026123047,
+      "eval_runtime": 0.7504,
+      "eval_samples_per_second": 179.9,
+      "eval_steps_per_second": 6.663,
+      "step": 10
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 5.389795780181885,
+      "epoch": 3.0,
+      "grad_norm": 7.020309925079346,
       "learning_rate": 3.888888888888889e-05,
-      "loss": 0.58,
-      "step": 3
+      "loss": 0.5341,
+      "step": 15
     },
     {
-      "epoch": 2.4,
-      "eval_accuracy": 0.9090909090909091,
-      "eval_loss": 0.3954509198665619,
-      "eval_runtime": 0.196,
-      "eval_samples_per_second": 168.338,
-      "eval_steps_per_second": 10.202,
-      "step": 3
+      "epoch": 3.0,
+      "eval_accuracy": 0.9481481481481482,
+      "eval_loss": 0.2379625141620636,
+      "eval_runtime": 0.7591,
+      "eval_samples_per_second": 177.839,
+      "eval_steps_per_second": 6.587,
+      "step": 15
     },
     {
       "epoch": 4.0,
-      "grad_norm": 4.907708644866943,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 0.2252,
-      "step": 5
+      "grad_norm": 9.26108455657959,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3583,
+      "step": 20
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9393939393939394,
-      "eval_loss": 0.2680495083332062,
-      "eval_runtime": 0.2061,
-      "eval_samples_per_second": 160.109,
-      "eval_steps_per_second": 9.704,
-      "step": 5
+      "eval_accuracy": 0.9629629629629629,
+      "eval_loss": 0.13244035840034485,
+      "eval_runtime": 0.7541,
+      "eval_samples_per_second": 179.024,
+      "eval_steps_per_second": 6.631,
+      "step": 20
     },
     {
-      "epoch": 4.8,
-      "grad_norm": 6.3704328536987305,
+      "epoch": 5.0,
+      "grad_norm": 5.511073112487793,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.2197,
+      "step": 25
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.1311105489730835,
+      "eval_runtime": 0.7578,
+      "eval_samples_per_second": 178.154,
+      "eval_steps_per_second": 6.598,
+      "step": 25
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 9.359658241271973,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.3826,
-      "step": 6
+      "loss": 0.1995,
+      "step": 30
     },
     {
-      "epoch": 4.8,
-      "eval_accuracy": 0.9696969696969697,
-      "eval_loss": 0.2426709532737732,
-      "eval_runtime": 0.2047,
-      "eval_samples_per_second": 161.236,
-      "eval_steps_per_second": 9.772,
-      "step": 6
+      "epoch": 6.0,
+      "eval_accuracy": 0.9777777777777777,
+      "eval_loss": 0.08695662766695023,
+      "eval_runtime": 0.7501,
+      "eval_samples_per_second": 179.972,
+      "eval_steps_per_second": 6.666,
+      "step": 30
     },
     {
-      "epoch": 5.6,
-      "grad_norm": 5.068804740905762,
+      "epoch": 7.0,
+      "grad_norm": 5.365365028381348,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.3551,
-      "step": 7
+      "loss": 0.1485,
+      "step": 35
     },
     {
-      "epoch": 5.6,
-      "eval_accuracy": 0.9090909090909091,
-      "eval_loss": 0.22610554099082947,
-      "eval_runtime": 0.2073,
-      "eval_samples_per_second": 159.187,
-      "eval_steps_per_second": 9.648,
-      "step": 7
+      "epoch": 7.0,
+      "eval_accuracy": 0.9777777777777777,
+      "eval_loss": 0.11073008179664612,
+      "eval_runtime": 0.7524,
+      "eval_samples_per_second": 179.431,
+      "eval_steps_per_second": 6.646,
+      "step": 35
     },
     {
-      "epoch": 6.4,
-      "grad_norm": 3.142571449279785,
+      "epoch": 8.0,
+      "grad_norm": 5.843384265899658,
       "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.3171,
-      "step": 8
+      "loss": 0.1689,
+      "step": 40
     },
     {
-      "epoch": 6.4,
-      "eval_accuracy": 0.9393939393939394,
-      "eval_loss": 0.21792414784431458,
-      "eval_runtime": 0.1962,
-      "eval_samples_per_second": 168.154,
-      "eval_steps_per_second": 10.191,
-      "step": 8
+      "epoch": 8.0,
+      "eval_accuracy": 0.9777777777777777,
+      "eval_loss": 0.07976409047842026,
+      "eval_runtime": 0.7575,
+      "eval_samples_per_second": 178.206,
+      "eval_steps_per_second": 6.6,
+      "step": 40
     },
     {
-      "epoch": 8.0,
-      "grad_norm": 4.538871765136719,
-      "learning_rate": 0.0,
-      "loss": 0.1802,
-      "step": 10
+      "epoch": 9.0,
+      "grad_norm": 4.395670413970947,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.1339,
+      "step": 45
     },
     {
-      "epoch": 8.0,
-      "eval_accuracy": 0.9393939393939394,
-      "eval_loss": 0.21566466987133026,
-      "eval_runtime": 0.219,
-      "eval_samples_per_second": 150.655,
-      "eval_steps_per_second": 9.131,
-      "step": 10
+      "epoch": 9.0,
+      "eval_accuracy": 0.9851851851851852,
+      "eval_loss": 0.05875137820839882,
+      "eval_runtime": 0.7546,
+      "eval_samples_per_second": 178.893,
+      "eval_steps_per_second": 6.626,
+      "step": 45
     },
     {
-      "epoch": 8.0,
-      "step": 10,
-      "total_flos": 3.121999450369229e+16,
-      "train_loss": 0.3916594386100769,
-      "train_runtime": 22.6465,
-      "train_samples_per_second": 69.326,
-      "train_steps_per_second": 0.442
+      "epoch": 10.0,
+      "grad_norm": 7.337897777557373,
+      "learning_rate": 0.0,
+      "loss": 0.1623,
+      "step": 50
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9777777777777777,
+      "eval_loss": 0.06197558343410492,
+      "eval_runtime": 0.7834,
+      "eval_samples_per_second": 172.335,
+      "eval_steps_per_second": 6.383,
+      "step": 50
+    },
+    {
+      "epoch": 10.0,
+      "step": 50,
+      "total_flos": 1.565971061889024e+17,
+      "train_loss": 0.394845541715622,
+      "train_runtime": 86.4231,
+      "train_samples_per_second": 72.897,
+      "train_steps_per_second": 0.579
     }
   ],
   "logging_steps": 1,
-  "max_steps": 10,
+  "max_steps": 50,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
@@ -163,7 +195,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.121999450369229e+16,
+  "total_flos": 1.565971061889024e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null