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 [facebook/timesformer-base-finetuned-k400](https://huggingface.co/facebook/timesformer-base-finetuned-k400) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8131
-- Accuracy: 0.7849
+- Loss: 0.7582
+- Accuracy: 0.7928
 
 ## Model description
 
@@ -45,13 +45,14 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
-- training_steps: 101
+- training_steps: 366
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 0.8303        | 1.0   | 101  | 0.1456          | 0.9902   |
+| 1.0449        | 0.5   | 183  | 0.0172          | 0.9939   |
+| 0.0056        | 1.5   | 366  | 0.0024          | 1.0      |
 
 
 ### Framework versions

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.0,
-    "eval_accuracy": 0.7848605577689243,
-    "eval_loss": 0.8131076693534851,
-    "eval_runtime": 82.7652,
-    "eval_samples_per_second": 3.033,
-    "eval_steps_per_second": 0.387
+    "epoch": 1.5,
+    "eval_accuracy": 0.7928286852589641,
+    "eval_loss": 0.758155345916748,
+    "eval_runtime": 81.0092,
+    "eval_samples_per_second": 3.098,
+    "eval_steps_per_second": 0.395
 }

test_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.0,
-    "eval_accuracy": 0.7848605577689243,
-    "eval_loss": 0.8131076693534851,
-    "eval_runtime": 82.7652,
-    "eval_samples_per_second": 3.033,
-    "eval_steps_per_second": 0.387
+    "epoch": 1.5,
+    "eval_accuracy": 0.7928286852589641,
+    "eval_loss": 0.758155345916748,
+    "eval_runtime": 81.0092,
+    "eval_samples_per_second": 3.098,
+    "eval_steps_per_second": 0.395
 }

trainer_state.json

@@ -1,50 +1,73 @@
 {
-  "best_metric": 0.990159901599016,
-  "best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/10 fold timesformer/Timesformer_default_fold_10_10_epoch_noAug_batch8_codecheck/checkpoint-101",
-  "epoch": 1.0,
+  "best_metric": 1.0,
+  "best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/10 fold timesformer/Timesformer_default_fold_10_10_epoch_noAug_batch8_codecheck/checkpoint-366",
+  "epoch": 1.5,
   "eval_steps": 500,
-  "global_step": 101,
+  "global_step": 366,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.9900990099009901,
-      "grad_norm": 7.038530349731445,
-      "learning_rate": 5.555555555555556e-07,
-      "loss": 0.8303,
+      "epoch": 0.273224043715847,
+      "grad_norm": 3.3262431621551514,
+      "learning_rate": 4.0425531914893614e-05,
+      "loss": 1.0449,
       "step": 100
     },
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.990159901599016,
-      "eval_loss": 0.14564795792102814,
-      "eval_runtime": 267.0652,
-      "eval_samples_per_second": 3.044,
-      "eval_steps_per_second": 0.382,
-      "step": 101
+      "epoch": 0.5,
+      "eval_accuracy": 0.9938650306748467,
+      "eval_loss": 0.017213916406035423,
+      "eval_runtime": 55.9924,
+      "eval_samples_per_second": 2.911,
+      "eval_steps_per_second": 0.375,
+      "step": 183
     },
     {
-      "epoch": 1.0,
-      "step": 101,
-      "total_flos": 7.07971920321577e+17,
-      "train_loss": 0.8236100951043686,
-      "train_runtime": 621.4991,
-      "train_samples_per_second": 1.3,
-      "train_steps_per_second": 0.163
+      "epoch": 1.046448087431694,
+      "grad_norm": 0.36886924505233765,
+      "learning_rate": 2.5227963525835867e-05,
+      "loss": 0.038,
+      "step": 200
     },
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.7848605577689243,
-      "eval_loss": 0.8131076693534851,
-      "eval_runtime": 82.7652,
-      "eval_samples_per_second": 3.033,
-      "eval_steps_per_second": 0.387,
-      "step": 101
+      "epoch": 1.319672131147541,
+      "grad_norm": 0.04929419606924057,
+      "learning_rate": 1.0030395136778116e-05,
+      "loss": 0.0056,
+      "step": 300
+    },
+    {
+      "epoch": 1.5,
+      "eval_accuracy": 1.0,
+      "eval_loss": 0.0024167420342564583,
+      "eval_runtime": 62.5343,
+      "eval_samples_per_second": 2.607,
+      "eval_steps_per_second": 0.336,
+      "step": 366
+    },
+    {
+      "epoch": 1.5,
+      "step": 366,
+      "total_flos": 2.5655220082940314e+18,
+      "train_loss": 0.2977769932623118,
+      "train_runtime": 1336.9205,
+      "train_samples_per_second": 2.19,
+      "train_steps_per_second": 0.274
+    },
+    {
+      "epoch": 1.5,
+      "eval_accuracy": 0.7928286852589641,
+      "eval_loss": 0.758155345916748,
+      "eval_runtime": 81.0092,
+      "eval_samples_per_second": 3.098,
+      "eval_steps_per_second": 0.395,
+      "step": 366
     }
   ],
   "logging_steps": 100,
-  "max_steps": 101,
+  "max_steps": 366,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9223372036854775807,
   "save_steps": 500,
@@ -60,7 +83,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 7.07971920321577e+17,
+  "total_flos": 2.5655220082940314e+18,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null