|
--- |
|
license: mit |
|
base_model: microsoft/Phi-3-mini-4k-instruct |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: Phi0503HMA6 |
|
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. --> |
|
|
|
# Phi0503HMA6 |
|
|
|
This model is a fine-tuned version of [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.1670 |
|
|
|
## 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: 80 |
|
- num_epochs: 3 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 4.2469 | 0.09 | 10 | 0.9109 | |
|
| 0.4285 | 0.18 | 20 | 0.2506 | |
|
| 0.266 | 0.27 | 30 | 0.2427 | |
|
| 0.2313 | 0.36 | 40 | 0.2118 | |
|
| 0.1808 | 0.45 | 50 | 0.1604 | |
|
| 0.163 | 0.54 | 60 | 0.1760 | |
|
| 0.2571 | 0.63 | 70 | 0.1448 | |
|
| 0.2789 | 0.73 | 80 | 0.1488 | |
|
| 0.7096 | 0.82 | 90 | 1.2197 | |
|
| 1.051 | 0.91 | 100 | 1.2133 | |
|
| 0.4623 | 1.0 | 110 | 4.9980 | |
|
| 4.8479 | 1.09 | 120 | 2.3085 | |
|
| 1.6873 | 1.18 | 130 | 0.7471 | |
|
| 0.5896 | 1.27 | 140 | 0.3693 | |
|
| 0.334 | 1.36 | 150 | 0.2707 | |
|
| 0.2556 | 1.45 | 160 | 0.2347 | |
|
| 0.2087 | 1.54 | 170 | 0.1840 | |
|
| 0.187 | 1.63 | 180 | 0.1858 | |
|
| 0.1833 | 1.72 | 190 | 0.1842 | |
|
| 0.1755 | 1.81 | 200 | 0.1787 | |
|
| 0.1772 | 1.9 | 210 | 0.1708 | |
|
| 0.1698 | 1.99 | 220 | 0.1714 | |
|
| 0.1723 | 2.08 | 230 | 0.1691 | |
|
| 0.1674 | 2.18 | 240 | 0.1693 | |
|
| 0.1682 | 2.27 | 250 | 0.1709 | |
|
| 0.1684 | 2.36 | 260 | 0.1702 | |
|
| 0.166 | 2.45 | 270 | 0.1681 | |
|
| 0.1651 | 2.54 | 280 | 0.1683 | |
|
| 0.1689 | 2.63 | 290 | 0.1688 | |
|
| 0.17 | 2.72 | 300 | 0.1675 | |
|
| 0.1696 | 2.81 | 310 | 0.1674 | |
|
| 0.1663 | 2.9 | 320 | 0.1670 | |
|
| 0.1712 | 2.99 | 330 | 0.1670 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.0.dev0 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.14.6 |
|
- Tokenizers 0.14.0 |
|
|
