jmmr-8282/email
Browse files
README.md
CHANGED
|
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss: 0.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
@@ -41,7 +41,7 @@ The following hyperparameters were used during training:
|
|
| 41 |
- seed: 42
|
| 42 |
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 43 |
- lr_scheduler_type: linear
|
| 44 |
-
- num_epochs:
|
| 45 |
- mixed_precision_training: Native AMP
|
| 46 |
|
| 47 |
### Training results
|
|
@@ -53,6 +53,11 @@ The following hyperparameters were used during training:
|
|
| 53 |
| 0.8155 | 3.0 | 1401 | 0.6182 |
|
| 54 |
| 0.6724 | 4.0 | 1868 | 0.5721 |
|
| 55 |
| 0.6073 | 5.0 | 2335 | 0.5628 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 56 |
|
| 57 |
|
| 58 |
### Framework versions
|
|
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.4552
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 41 |
- seed: 42
|
| 42 |
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 43 |
- lr_scheduler_type: linear
|
| 44 |
+
- num_epochs: 10
|
| 45 |
- mixed_precision_training: Native AMP
|
| 46 |
|
| 47 |
### Training results
|
|
|
|
| 53 |
| 0.8155 | 3.0 | 1401 | 0.6182 |
|
| 54 |
| 0.6724 | 4.0 | 1868 | 0.5721 |
|
| 55 |
| 0.6073 | 5.0 | 2335 | 0.5628 |
|
| 56 |
+
| 0.5785 | 6.0 | 2802 | 0.5057 |
|
| 57 |
+
| 0.5418 | 7.0 | 3269 | 0.4829 |
|
| 58 |
+
| 0.5139 | 8.0 | 3736 | 0.4674 |
|
| 59 |
+
| 0.4864 | 9.0 | 4203 | 0.4572 |
|
| 60 |
+
| 0.4845 | 10.0 | 4670 | 0.4552 |
|
| 61 |
|
| 62 |
|
| 63 |
### Framework versions
|