Model save

README.md

@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.6571
-- Exact Match: 41.5730
-- F1: 0.6310
+- Loss: 5.4813
+- Exact Match: 6.7416
+- F1: 5.2790
 
 ## Model description
 
@@ -39,22 +39,21 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1.5101664377049603e-05
-- train_batch_size: 8
-- eval_batch_size: 8
+- learning_rate: 1.3093619983970319e-05
+- train_batch_size: 32
+- eval_batch_size: 32
 - seed: 42
 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
-- num_epochs: 4
+- num_epochs: 3
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Exact Match | F1     |
 |:-------------:|:-----:|:----:|:---------------:|:-----------:|:------:|
-| No log        | 1.0   | 32   | 4.8737          | 10.1124     | 4.5405 |
-| No log        | 2.0   | 64   | 4.0712          | 13.4831     | 2.6494 |
-| No log        | 3.0   | 96   | 3.7312          | 48.3146     | 0.0    |
-| No log        | 4.0   | 128  | 3.6571          | 41.5730     | 0.6310 |
+| No log        | 1.0   | 6    | 5.8944          | 0.0         | 4.8268 |
+| No log        | 2.0   | 12   | 5.5614          | 4.4944      | 5.0076 |
+| No log        | 3.0   | 18   | 5.4813          | 6.7416      | 5.2790 |
 
 
 ### Framework versions