Model save
Browse files
README.md
CHANGED
|
@@ -42,8 +42,8 @@ The following hyperparameters were used during training:
|
|
| 42 |
- gradient_accumulation_steps: 32
|
| 43 |
- total_train_batch_size: 32
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
-
- lr_scheduler_type:
|
| 46 |
-
- lr_scheduler_warmup_ratio: 0.
|
| 47 |
- num_epochs: 3
|
| 48 |
|
| 49 |
### Training results
|
|
|
|
| 42 |
- gradient_accumulation_steps: 32
|
| 43 |
- total_train_batch_size: 32
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
+
- lr_scheduler_type: inverse_sqrt
|
| 46 |
+
- lr_scheduler_warmup_ratio: 0.05
|
| 47 |
- num_epochs: 3
|
| 48 |
|
| 49 |
### Training results
|