metadata
tags:
- generated_from_trainer
model-index:
- name: Baby-Llama-58M-ORIGINAL
results: []
Baby-Llama-58M-ORIGINAL
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 7.6445
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: 0.00025
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 200
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 19.8206 | 1.0 | 9 | 19.5231 |
| 17.2243 | 2.0 | 18 | 16.9614 |
| 13.4949 | 3.0 | 27 | 13.0683 |
| 10.354 | 4.0 | 36 | 10.1351 |
| 8.4304 | 5.0 | 45 | 8.3060 |
| 7.5971 | 6.0 | 54 | 7.6445 |
Framework versions
- Transformers 4.39.1
- Pytorch 2.1.2+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0