| --- |
| license: mit |
| base_model: microsoft/phi-2 |
| tags: |
| - generated_from_trainer |
| model-index: |
| - name: V0309O2 |
| 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. --> |
|
|
| # V0309O2 |
|
|
| This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset. |
| It achieves the following results on the evaluation set: |
| - Loss: 0.0716 |
|
|
| ## 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: 0.0003 |
| - train_batch_size: 4 |
| - eval_batch_size: 8 |
| - seed: 42 |
| - gradient_accumulation_steps: 32 |
| - total_train_batch_size: 128 |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| - lr_scheduler_type: cosine_with_restarts |
| - lr_scheduler_warmup_steps: 20 |
| - num_epochs: 3 |
| - mixed_precision_training: Native AMP |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | |
| |:-------------:|:-----:|:----:|:---------------:| |
| | 1.6792 | 0.09 | 10 | 0.1456 | |
| | 0.164 | 0.17 | 20 | 0.1075 | |
| | 0.1211 | 0.26 | 30 | 0.0749 | |
| | 0.1029 | 0.34 | 40 | 0.0726 | |
| | 0.099 | 0.43 | 50 | 0.0684 | |
| | 0.0915 | 0.51 | 60 | 0.0691 | |
| | 0.0824 | 0.6 | 70 | 0.0664 | |
| | 0.0898 | 0.68 | 80 | 0.0716 | |
| | 0.0815 | 0.77 | 90 | 0.0759 | |
| | 0.0806 | 0.85 | 100 | 0.0762 | |
| | 0.0789 | 0.94 | 110 | 0.0664 | |
| | 0.0775 | 1.02 | 120 | 0.0641 | |
| | 0.073 | 1.11 | 130 | 0.0737 | |
| | 0.0668 | 1.19 | 140 | 0.0677 | |
| | 0.0642 | 1.28 | 150 | 0.0684 | |
| | 0.0646 | 1.37 | 160 | 0.0724 | |
| | 0.062 | 1.45 | 170 | 0.0695 | |
| | 0.0601 | 1.54 | 180 | 0.0689 | |
| | 0.0651 | 1.62 | 190 | 0.0652 | |
| | 0.0604 | 1.71 | 200 | 0.0684 | |
| | 0.0635 | 1.79 | 210 | 0.0679 | |
| | 0.0567 | 1.88 | 220 | 0.0703 | |
| | 0.057 | 1.96 | 230 | 0.0690 | |
| | 0.0557 | 2.05 | 240 | 0.0711 | |
| | 0.0447 | 2.13 | 250 | 0.0707 | |
| | 0.0479 | 2.22 | 260 | 0.0735 | |
| | 0.0434 | 2.3 | 270 | 0.0753 | |
| | 0.0493 | 2.39 | 280 | 0.0721 | |
| | 0.0496 | 2.47 | 290 | 0.0708 | |
| | 0.0468 | 2.56 | 300 | 0.0709 | |
| | 0.0525 | 2.65 | 310 | 0.0709 | |
| | 0.0419 | 2.73 | 320 | 0.0713 | |
| | 0.047 | 2.82 | 330 | 0.0715 | |
| | 0.0436 | 2.9 | 340 | 0.0715 | |
| | 0.0474 | 2.99 | 350 | 0.0716 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 4.36.0.dev0 |
| - Pytorch 2.1.2+cu121 |
| - Datasets 2.14.6 |
| - Tokenizers 0.14.1 |
|
|
