|
--- |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: longt5_xl_sfd_bp_40 |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# longt5_xl_sfd_bp_40 |
|
|
|
This model was trained from scratch on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 3.3318 |
|
|
|
## Model description |
|
|
|
More information needed |
|
|
|
## Intended uses & limitations |
|
|
|
More information needed |
|
|
|
## Training and evaluation data |
|
|
|
More information needed |
|
|
|
## Training procedure |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 0.001 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 32 |
|
- total_train_batch_size: 256 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: constant |
|
- num_epochs: 20.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 0.1033 | 0.97 | 14 | 3.1040 | |
|
| 0.0836 | 1.95 | 28 | 3.0540 | |
|
| 0.0717 | 2.99 | 43 | 2.9414 | |
|
| 0.0614 | 3.97 | 57 | 3.0238 | |
|
| 0.1275 | 4.94 | 71 | 2.8326 | |
|
| 0.0511 | 5.98 | 86 | 3.0479 | |
|
| 0.0666 | 6.96 | 100 | 3.1255 | |
|
| 0.0398 | 8.0 | 115 | 3.2240 | |
|
| 0.0396 | 8.97 | 129 | 3.1667 | |
|
| 0.0466 | 9.95 | 143 | 3.2775 | |
|
| 0.043 | 10.99 | 158 | 3.3289 | |
|
| 0.0538 | 11.97 | 172 | 2.8202 | |
|
| 0.028 | 12.94 | 186 | 3.4366 | |
|
| 0.1056 | 13.98 | 201 | 3.3447 | |
|
| 0.0303 | 14.96 | 215 | 3.0069 | |
|
| 0.0234 | 16.0 | 230 | 3.3524 | |
|
| 0.0263 | 16.97 | 244 | 3.2473 | |
|
| 0.0225 | 17.95 | 258 | 3.3365 | |
|
| 0.0225 | 18.99 | 273 | 3.4389 | |
|
| 0.0211 | 19.48 | 280 | 3.3318 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.2 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.16.1 |
|
- Tokenizers 0.15.0 |
|
|
