| | --- |
| | license: mit |
| | base_model: microsoft/phi-2 |
| | tags: |
| | - generated_from_trainer |
| | model-index: |
| | - name: V0413TUNE |
| | 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. --> |
| |
|
| | # V0413TUNE |
| |
|
| | 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: 0.0419 |
| |
|
| | ## 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: 0.003 |
| | - train_batch_size: 8 |
| | - eval_batch_size: 8 |
| | - seed: 42 |
| | - gradient_accumulation_steps: 8 |
| | - total_train_batch_size: 64 |
| | - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| | - lr_scheduler_type: cosine_with_restarts |
| | - lr_scheduler_warmup_steps: 100 |
| | - num_epochs: 3 |
| | - mixed_precision_training: Native AMP |
| |
|
| | ### Training results |
| |
|
| | | Training Loss | Epoch | Step | Validation Loss | |
| | |:-------------:|:-----:|:----:|:---------------:| |
| | | 0.6884 | 0.09 | 20 | 0.1584 | |
| | | 0.1153 | 0.18 | 40 | 0.0993 | |
| | | 0.096 | 0.27 | 60 | 0.0854 | |
| | | 0.1014 | 0.36 | 80 | 0.0820 | |
| | | 0.0813 | 0.45 | 100 | 0.0795 | |
| | | 0.0869 | 0.54 | 120 | 0.0707 | |
| | | 0.0858 | 0.63 | 140 | 0.0831 | |
| | | 0.0841 | 0.73 | 160 | 0.0780 | |
| | | 0.0895 | 0.82 | 180 | 0.0732 | |
| | | 0.0908 | 0.91 | 200 | 0.0808 | |
| | | 0.0872 | 1.0 | 220 | 0.0807 | |
| | | 0.0726 | 1.09 | 240 | 0.0720 | |
| | | 0.0644 | 1.18 | 260 | 0.0740 | |
| | | 0.216 | 1.27 | 280 | 0.2003 | |
| | | 0.0945 | 1.36 | 300 | 0.0814 | |
| | | 0.0937 | 1.45 | 320 | 0.0842 | |
| | | 0.0868 | 1.54 | 340 | 0.0801 | |
| | | 0.0714 | 1.63 | 360 | 0.0709 | |
| | | 0.0632 | 1.72 | 380 | 0.0639 | |
| | | 0.0626 | 1.81 | 400 | 0.0518 | |
| | | 0.0467 | 1.9 | 420 | 0.0510 | |
| | | 0.0541 | 1.99 | 440 | 0.0475 | |
| | | 0.0486 | 2.08 | 460 | 0.0580 | |
| | | 0.046 | 2.18 | 480 | 0.0484 | |
| | | 0.0385 | 2.27 | 500 | 0.0493 | |
| | | 0.0446 | 2.36 | 520 | 0.0470 | |
| | | 0.037 | 2.45 | 540 | 0.0424 | |
| | | 0.0446 | 2.54 | 560 | 0.0433 | |
| | | 0.0297 | 2.63 | 580 | 0.0441 | |
| | | 0.0317 | 2.72 | 600 | 0.0426 | |
| | | 0.0481 | 2.81 | 620 | 0.0425 | |
| | | 0.0318 | 2.9 | 640 | 0.0421 | |
| | | 0.0332 | 2.99 | 660 | 0.0419 | |
| |
|
| |
|
| | ### Framework versions |
| |
|
| | - Transformers 4.36.0.dev0 |
| | - Pytorch 2.1.2+cu121 |
| | - Datasets 2.14.6 |
| | - Tokenizers 0.14.1 |
| |
|
