| --- |
| base_model: meta-llama/Llama-2-7b-hf |
| tags: |
| - generated_from_trainer |
| model-index: |
| - name: qlora-out |
| 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. --> |
|
|
| [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl) |
| # qlora-out |
|
|
| This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on the None dataset. |
| It achieves the following results on the evaluation set: |
| - Loss: 0.5703 |
|
|
| ## 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: 0.0004 |
| - train_batch_size: 1 |
| - eval_batch_size: 1 |
| - seed: 42 |
| - gradient_accumulation_steps: 16 |
| - total_train_batch_size: 16 |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| - lr_scheduler_type: cosine |
| - lr_scheduler_warmup_steps: 100 |
| - num_epochs: 2 |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | |
| |:-------------:|:-----:|:----:|:---------------:| |
| | 0.8756 | 0.06 | 20 | 0.7111 | |
| | 0.9058 | 0.11 | 40 | 0.6764 | |
| | 0.7526 | 0.17 | 60 | 0.6669 | |
| | 0.6926 | 0.23 | 80 | 0.6363 | |
| | 0.6731 | 0.28 | 100 | 0.6187 | |
| | 0.647 | 0.34 | 120 | 0.6162 | |
| | 0.6219 | 0.4 | 140 | 0.6041 | |
| | 0.5781 | 0.45 | 160 | 0.5937 | |
| | 0.6346 | 0.51 | 180 | 0.6006 | |
| | 0.7663 | 0.57 | 200 | 0.5926 | |
| | 0.5864 | 0.62 | 220 | 0.5866 | |
| | 0.5943 | 0.68 | 240 | 0.5756 | |
| | 0.5029 | 0.74 | 260 | 0.5733 | |
| | 0.5482 | 0.79 | 280 | 0.5712 | |
| | 0.5413 | 0.85 | 300 | 0.5820 | |
| | 0.657 | 0.91 | 320 | 0.5696 | |
| | 0.506 | 0.96 | 340 | 0.5839 | |
| | 0.4804 | 1.02 | 360 | 0.5803 | |
| | 0.5095 | 1.08 | 380 | 0.5974 | |
| | 0.4404 | 1.13 | 400 | 0.5746 | |
| | 0.3869 | 1.19 | 420 | 0.5740 | |
| | 0.4129 | 1.25 | 440 | 0.5777 | |
| | 0.4209 | 1.3 | 460 | 0.5825 | |
| | 0.4014 | 1.36 | 480 | 0.5742 | |
| | 0.3333 | 1.42 | 500 | 0.5851 | |
| | 0.5041 | 1.47 | 520 | 0.5798 | |
| | 0.5528 | 1.53 | 540 | 0.5631 | |
| | 0.4372 | 1.59 | 560 | 0.5747 | |
| | 0.3901 | 1.64 | 580 | 0.5625 | |
| | 0.5271 | 1.7 | 600 | 0.5746 | |
| | 0.4283 | 1.76 | 620 | 0.5662 | |
| | 0.4336 | 1.81 | 640 | 0.5652 | |
| | 0.3534 | 1.87 | 660 | 0.5697 | |
| | 0.4728 | 1.93 | 680 | 0.5713 | |
| | 0.5159 | 1.98 | 700 | 0.5703 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 4.34.1 |
| - Pytorch 2.0.1+cu118 |
| - Datasets 2.14.6 |
| - Tokenizers 0.14.1 |
|
|
