Baby-Llama-95M-Seq-2
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.9835
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: 128
- 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: 10
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
8.5851 | 1.0 | 1033 | 8.5318 |
3.4405 | 2.0 | 2066 | 3.9014 |
2.6473 | 3.0 | 3099 | 3.1378 |
2.2658 | 4.0 | 4132 | 2.6508 |
2.0133 | 5.0 | 5165 | 2.3921 |
1.8829 | 6.0 | 6198 | 2.2239 |
1.7884 | 7.0 | 7231 | 2.1093 |
1.7038 | 8.0 | 8264 | 2.0265 |
1.6884 | 9.0 | 9297 | 1.9905 |
1.6599 | 10.0 | 10330 | 1.9835 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Tokenizers 0.19.1
- Downloads last month
- 19
Inference API (serverless) is not available, repository is disabled.