|
--- |
|
license: mit |
|
base_model: microsoft/Phi-3-mini-4k-instruct |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: Phi0503HMA16 |
|
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. --> |
|
|
|
# Phi0503HMA16 |
|
|
|
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: 0.0775 |
|
|
|
## 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: 100 |
|
- num_epochs: 3 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 4.3598 | 0.09 | 10 | 0.9925 | |
|
| 0.4108 | 0.18 | 20 | 0.2307 | |
|
| 0.2372 | 0.27 | 30 | 0.2352 | |
|
| 0.2131 | 0.36 | 40 | 0.2189 | |
|
| 0.1969 | 0.45 | 50 | 0.1518 | |
|
| 0.1415 | 0.54 | 60 | 0.0999 | |
|
| 0.0976 | 0.63 | 70 | 0.1068 | |
|
| 0.0853 | 0.73 | 80 | 0.0846 | |
|
| 0.0864 | 0.82 | 90 | 0.0784 | |
|
| 0.0782 | 0.91 | 100 | 0.0734 | |
|
| 0.0866 | 1.0 | 110 | 0.0806 | |
|
| 0.0649 | 1.09 | 120 | 0.0712 | |
|
| 0.0663 | 1.18 | 130 | 0.0769 | |
|
| 0.0704 | 1.27 | 140 | 0.0729 | |
|
| 0.0634 | 1.36 | 150 | 0.0740 | |
|
| 0.068 | 1.45 | 160 | 0.0709 | |
|
| 0.0645 | 1.54 | 170 | 0.0687 | |
|
| 0.063 | 1.63 | 180 | 0.0689 | |
|
| 0.0584 | 1.72 | 190 | 0.0604 | |
|
| 0.065 | 1.81 | 200 | 0.0608 | |
|
| 0.0532 | 1.9 | 210 | 0.0681 | |
|
| 0.0539 | 1.99 | 220 | 0.0694 | |
|
| 0.0313 | 2.08 | 230 | 0.0816 | |
|
| 0.0356 | 2.18 | 240 | 0.0880 | |
|
| 0.0296 | 2.27 | 250 | 0.0834 | |
|
| 0.0287 | 2.36 | 260 | 0.0780 | |
|
| 0.0336 | 2.45 | 270 | 0.0801 | |
|
| 0.0236 | 2.54 | 280 | 0.0827 | |
|
| 0.0263 | 2.63 | 290 | 0.0828 | |
|
| 0.0335 | 2.72 | 300 | 0.0794 | |
|
| 0.0317 | 2.81 | 310 | 0.0780 | |
|
| 0.0296 | 2.9 | 320 | 0.0773 | |
|
| 0.0289 | 2.99 | 330 | 0.0775 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.0.dev0 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.14.6 |
|
- Tokenizers 0.14.0 |
|
|
