| --- |
| library_name: peft |
| tags: |
| - trl |
| - sft |
| - generated_from_trainer |
| base_model: meta-llama/Llama-2-70b-chat-hf |
| model-index: |
| - name: output |
| 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. --> |
|
|
| # output |
|
|
| This model is a fine-tuned version of [meta-llama/Llama-2-70b-chat-hf](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf) on an unknown dataset. |
| It achieves the following results on the evaluation set: |
| - Loss: 1.3387 |
|
|
| ## 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: 0.0002 |
| - train_batch_size: 8 |
| - eval_batch_size: 8 |
| - seed: 42 |
| - gradient_accumulation_steps: 2 |
| - total_train_batch_size: 16 |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| - lr_scheduler_type: constant |
| - lr_scheduler_warmup_ratio: 0.3 |
| - num_epochs: 4 |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | |
| |:-------------:|:-----:|:----:|:---------------:| |
| | 1.7334 | 0.29 | 1 | 1.5084 | |
| | 1.7705 | 0.57 | 2 | 1.4977 | |
| | 1.7433 | 0.86 | 3 | 1.4736 | |
| | 1.6862 | 1.14 | 4 | 1.4434 | |
| | 1.6562 | 1.43 | 5 | 1.4161 | |
| | 1.615 | 1.71 | 6 | 1.3948 | |
| | 1.6227 | 2.0 | 7 | 1.3813 | |
| | 1.5609 | 2.29 | 8 | 1.3706 | |
| | 1.619 | 2.57 | 9 | 1.3603 | |
| | 1.5298 | 2.86 | 10 | 1.3511 | |
| | 1.4428 | 3.14 | 11 | 1.3437 | |
| | 1.5641 | 3.43 | 12 | 1.3387 | |
|
|
|
|
| ### Framework versions |
|
|
| - PEFT 0.8.2 |
| - Transformers 4.38.1 |
| - Pytorch 2.2.1 |
| - Datasets 2.17.1 |
| - Tokenizers 0.15.2 |
