|
|
--- |
|
|
library_name: transformers |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
metrics: |
|
|
- accuracy |
|
|
model-index: |
|
|
- name: BABYLM_5000_41 |
|
|
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. --> |
|
|
|
|
|
# BABYLM_5000_41 |
|
|
|
|
|
This model was trained from scratch on an unknown dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 3.3799 |
|
|
- Accuracy: 0.3821 |
|
|
|
|
|
## 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: 0.0006 |
|
|
- train_batch_size: 32 |
|
|
- eval_batch_size: 32 |
|
|
- seed: 41 |
|
|
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
|
|
- lr_scheduler_type: linear |
|
|
- lr_scheduler_warmup_steps: 1000 |
|
|
- num_epochs: 10.0 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|
|:-------------:|:-----:|:-----:|:---------------:|:--------:| |
|
|
| 1.1021 | 1.0 | 1135 | 4.1783 | 0.2970 | |
|
|
| 0.9898 | 2.0 | 2270 | 4.0657 | 0.3065 | |
|
|
| 0.9617 | 3.0 | 3405 | 3.9151 | 0.3191 | |
|
|
| 0.9291 | 4.0 | 4540 | 3.7639 | 0.3347 | |
|
|
| 0.8903 | 5.0 | 5675 | 3.6331 | 0.3492 | |
|
|
| 0.8546 | 6.0 | 6810 | 3.5462 | 0.3599 | |
|
|
| 0.8321 | 7.0 | 7945 | 3.4829 | 0.3680 | |
|
|
| 0.798 | 8.0 | 9080 | 3.4338 | 0.3742 | |
|
|
| 0.7836 | 9.0 | 10215 | 3.4001 | 0.3786 | |
|
|
| 0.7747 | 10.0 | 11350 | 3.3799 | 0.3821 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.51.3 |
|
|
- Pytorch 2.5.1+cu121 |
|
|
- Datasets 3.6.0 |
|
|
- Tokenizers 0.21.1 |
|
|
|
