Masaki Eguchi
commited on
Commit
•
4ae4344
1
Parent(s):
835132c
update model card README.md
Browse files
README.md
CHANGED
|
@@ -14,12 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on an unknown dataset.
|
| 16 |
It achieves the following results on the evaluation set:
|
| 17 |
-
-
|
| 18 |
-
- eval_runtime: 654.089
|
| 19 |
-
- eval_samples_per_second: 8.303
|
| 20 |
-
- eval_steps_per_second: 0.26
|
| 21 |
-
- epoch: 4.28
|
| 22 |
-
- step: 1450
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
@@ -38,18 +33,39 @@ More information needed
|
|
| 38 |
### Training hyperparameters
|
| 39 |
|
| 40 |
The following hyperparameters were used during training:
|
| 41 |
-
- learning_rate:
|
| 42 |
-
- train_batch_size:
|
| 43 |
-
- eval_batch_size:
|
| 44 |
- seed: 42
|
| 45 |
-
- gradient_accumulation_steps:
|
| 46 |
- total_train_batch_size: 256
|
| 47 |
- optimizer: Adam with betas=(0.9,0.99) and epsilon=1e-08
|
| 48 |
-
- lr_scheduler_type:
|
| 49 |
-
-
|
| 50 |
-
- num_epochs:
|
| 51 |
- mixed_precision_training: Native AMP
|
| 52 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53 |
### Framework versions
|
| 54 |
|
| 55 |
- Transformers 4.25.1
|
|
|
|
| 14 |
|
| 15 |
This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on an unknown dataset.
|
| 16 |
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 1.4621
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
|
|
|
| 33 |
### Training hyperparameters
|
| 34 |
|
| 35 |
The following hyperparameters were used during training:
|
| 36 |
+
- learning_rate: 5e-05
|
| 37 |
+
- train_batch_size: 8
|
| 38 |
+
- eval_batch_size: 8
|
| 39 |
- seed: 42
|
| 40 |
+
- gradient_accumulation_steps: 32
|
| 41 |
- total_train_batch_size: 256
|
| 42 |
- optimizer: Adam with betas=(0.9,0.99) and epsilon=1e-08
|
| 43 |
+
- lr_scheduler_type: cosine
|
| 44 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 45 |
+
- num_epochs: 15
|
| 46 |
- mixed_precision_training: Native AMP
|
| 47 |
|
| 48 |
+
### Training results
|
| 49 |
+
|
| 50 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 51 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 52 |
+
| 1.668 | 1.0 | 169 | 1.5412 |
|
| 53 |
+
| 1.6261 | 2.0 | 338 | 1.5138 |
|
| 54 |
+
| 1.5967 | 3.0 | 507 | 1.4943 |
|
| 55 |
+
| 1.5745 | 4.0 | 676 | 1.4913 |
|
| 56 |
+
| 1.5549 | 5.0 | 845 | 1.4738 |
|
| 57 |
+
| 1.5445 | 6.0 | 1014 | 1.4671 |
|
| 58 |
+
| 1.536 | 7.0 | 1183 | 1.4689 |
|
| 59 |
+
| 1.5254 | 8.0 | 1352 | 1.4612 |
|
| 60 |
+
| 1.5244 | 9.0 | 1521 | 1.4560 |
|
| 61 |
+
| 1.5263 | 10.0 | 1690 | 1.4580 |
|
| 62 |
+
| 1.5249 | 11.0 | 1859 | 1.4482 |
|
| 63 |
+
| 1.5238 | 12.0 | 2028 | 1.4565 |
|
| 64 |
+
| 1.526 | 13.0 | 2197 | 1.4568 |
|
| 65 |
+
| 1.5232 | 14.0 | 2366 | 1.4547 |
|
| 66 |
+
| 1.5232 | 15.0 | 2535 | 1.4558 |
|
| 67 |
+
|
| 68 |
+
|
| 69 |
### Framework versions
|
| 70 |
|
| 71 |
- Transformers 4.25.1
|