Model save

README.md

@@ -19,10 +19,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [xlnet-large-cased](https://huggingface.co/xlnet-large-cased) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5708
-- F1: 0.1476
-- Roc Auc: 0.4998
-- Accuracy: 0.1465
+- Loss: 0.7955
+- F1: 0.7396
+- Roc Auc: 0.8018
+- Accuracy: 0.5192
 
 ## Model description
 
@@ -42,8 +42,8 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 2e-05
-- train_batch_size: 32
-- eval_batch_size: 32
+- train_batch_size: 16
+- eval_batch_size: 16
 - seed: 42
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
@@ -54,10 +54,36 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | F1     | Roc Auc | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:------:|:-------:|:--------:|
-| 0.5836        | 1.0   | 88   | 0.5708          | 0.1476 | 0.4998  | 0.1465   |
-| 0.5786        | 2.0   | 176  | 0.5738          | 0.1476 | 0.5     | 0.1479   |
-| 0.5752        | 3.0   | 264  | 0.5617          | 0.1476 | 0.5     | 0.1479   |
-| 0.5701        | 4.0   | 352  | 0.5603          | 0.1476 | 0.5     | 0.1479   |
+| 0.5878        | 1.0   | 176  | 0.5881          | 0.1451 | 0.5     | 0.1351   |
+| 0.5732        | 2.0   | 352  | 0.5805          | 0.1451 | 0.5     | 0.1351   |
+| 0.557         | 3.0   | 528  | 0.5823          | 0.1451 | 0.5     | 0.1351   |
+| 0.5772        | 4.0   | 704  | 0.5767          | 0.1452 | 0.5003  | 0.1351   |
+| 0.5465        | 5.0   | 880  | 0.5267          | 0.2391 | 0.5664  | 0.2048   |
+| 0.4598        | 6.0   | 1056 | 0.4979          | 0.3903 | 0.6249  | 0.2774   |
+| 0.4303        | 7.0   | 1232 | 0.4610          | 0.5396 | 0.6945  | 0.3556   |
+| 0.3532        | 8.0   | 1408 | 0.4440          | 0.5847 | 0.7137  | 0.3841   |
+| 0.2993        | 9.0   | 1584 | 0.4478          | 0.6193 | 0.7243  | 0.4168   |
+| 0.2826        | 10.0  | 1760 | 0.4795          | 0.6142 | 0.7301  | 0.4196   |
+| 0.2466        | 11.0  | 1936 | 0.4835          | 0.6709 | 0.7516  | 0.4481   |
+| 0.1609        | 12.0  | 2112 | 0.4983          | 0.6965 | 0.7764  | 0.4637   |
+| 0.1089        | 13.0  | 2288 | 0.5277          | 0.7061 | 0.7775  | 0.4666   |
+| 0.0849        | 14.0  | 2464 | 0.5735          | 0.7163 | 0.7840  | 0.4609   |
+| 0.0884        | 15.0  | 2640 | 0.6126          | 0.7061 | 0.7699  | 0.4822   |
+| 0.04          | 16.0  | 2816 | 0.6565          | 0.7173 | 0.7884  | 0.4908   |
+| 0.063         | 17.0  | 2992 | 0.6826          | 0.7215 | 0.7863  | 0.4979   |
+| 0.0378        | 18.0  | 3168 | 0.6910          | 0.7319 | 0.7974  | 0.5121   |
+| 0.0263        | 19.0  | 3344 | 0.7434          | 0.7230 | 0.7901  | 0.4936   |
+| 0.0229        | 20.0  | 3520 | 0.7325          | 0.7376 | 0.8036  | 0.5050   |
+| 0.0133        | 21.0  | 3696 | 0.7438          | 0.7364 | 0.8012  | 0.5092   |
+| 0.0065        | 22.0  | 3872 | 0.7647          | 0.7334 | 0.8001  | 0.5135   |
+| 0.0056        | 23.0  | 4048 | 0.7734          | 0.7374 | 0.8018  | 0.5164   |
+| 0.0032        | 24.0  | 4224 | 0.7828          | 0.7382 | 0.8014  | 0.5206   |
+| 0.0064        | 25.0  | 4400 | 0.7855          | 0.7352 | 0.7987  | 0.5149   |
+| 0.005         | 26.0  | 4576 | 0.7907          | 0.7347 | 0.7967  | 0.5149   |
+| 0.0097        | 27.0  | 4752 | 0.7985          | 0.7362 | 0.7996  | 0.5149   |
+| 0.0032        | 28.0  | 4928 | 0.7953          | 0.7385 | 0.8011  | 0.5192   |
+| 0.0027        | 29.0  | 5104 | 0.7954          | 0.7378 | 0.8005  | 0.5178   |
+| 0.0027        | 30.0  | 5280 | 0.7955          | 0.7396 | 0.8018  | 0.5192   |
 
 
 ### Framework versions