| --- |
| license: mit |
| base_model: microsoft/phi-2 |
| tags: |
| - generated_from_trainer |
| model-index: |
| - name: V0415MA2plus |
| 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. --> |
|
|
| # V0415MA2plus |
|
|
| 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: 0.0731 |
|
|
| ## 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: 8 |
| - eval_batch_size: 8 |
| - seed: 42 |
| - gradient_accumulation_steps: 16 |
| - 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: 60 |
| - num_epochs: 3 |
| - mixed_precision_training: Native AMP |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | |
| |:-------------:|:-----:|:----:|:---------------:| |
| | 1.7881 | 0.09 | 10 | 0.1798 | |
| | 0.1395 | 0.18 | 20 | 0.1056 | |
| | 0.0994 | 0.27 | 30 | 0.0796 | |
| | 0.0852 | 0.36 | 40 | 0.0695 | |
| | 0.0718 | 0.45 | 50 | 0.0714 | |
| | 0.083 | 0.54 | 60 | 0.0731 | |
| | 0.0743 | 0.63 | 70 | 0.0669 | |
| | 0.0713 | 0.73 | 80 | 0.0658 | |
| | 0.0709 | 0.82 | 90 | 0.0618 | |
| | 0.074 | 0.91 | 100 | 0.0721 | |
| | 0.0697 | 1.0 | 110 | 0.0656 | |
| | 0.0515 | 1.09 | 120 | 0.0624 | |
| | 0.057 | 1.18 | 130 | 0.0756 | |
| | 0.0553 | 1.27 | 140 | 0.0623 | |
| | 0.0517 | 1.36 | 150 | 0.0667 | |
| | 0.0564 | 1.45 | 160 | 0.0582 | |
| | 0.0517 | 1.54 | 170 | 0.0671 | |
| | 0.0541 | 1.63 | 180 | 0.0586 | |
| | 0.0475 | 1.72 | 190 | 0.0587 | |
| | 0.0545 | 1.81 | 200 | 0.0582 | |
| | 0.0437 | 1.9 | 210 | 0.0627 | |
| | 0.0422 | 1.99 | 220 | 0.0654 | |
| | 0.0278 | 2.08 | 230 | 0.0590 | |
| | 0.0238 | 2.18 | 240 | 0.0699 | |
| | 0.0167 | 2.27 | 250 | 0.0802 | |
| | 0.0186 | 2.36 | 260 | 0.0847 | |
| | 0.024 | 2.45 | 270 | 0.0802 | |
| | 0.0213 | 2.54 | 280 | 0.0780 | |
| | 0.0217 | 2.63 | 290 | 0.0754 | |
| | 0.0237 | 2.72 | 300 | 0.0742 | |
| | 0.0235 | 2.81 | 310 | 0.0732 | |
| | 0.02 | 2.9 | 320 | 0.0732 | |
| | 0.0251 | 2.99 | 330 | 0.0731 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 4.36.0.dev0 |
| - Pytorch 2.1.2+cu121 |
| - Datasets 2.14.6 |
| - Tokenizers 0.14.1 |
|
|
