|
--- |
|
license: mit |
|
base_model: microsoft/Phi-3-mini-4k-instruct |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: Phi0503HMA17 |
|
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. --> |
|
|
|
# Phi0503HMA17 |
|
|
|
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: 0.1875 |
|
|
|
## 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 | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 3.75 | 0.09 | 10 | 0.5453 | |
|
| 0.3378 | 0.18 | 20 | 0.2824 | |
|
| 0.3447 | 0.27 | 30 | 0.3174 | |
|
| 0.2287 | 0.36 | 40 | 0.1820 | |
|
| 0.1614 | 0.45 | 50 | 0.1595 | |
|
| 0.1573 | 0.54 | 60 | 0.1415 | |
|
| 0.1433 | 0.63 | 70 | 0.1354 | |
|
| 0.2011 | 0.73 | 80 | 1.2273 | |
|
| 0.311 | 0.82 | 90 | 0.1241 | |
|
| 0.1326 | 0.91 | 100 | 0.1242 | |
|
| 0.129 | 1.0 | 110 | 0.1278 | |
|
| 0.0812 | 1.09 | 120 | 0.0927 | |
|
| 0.1272 | 1.18 | 130 | 0.2229 | |
|
| 3.2223 | 1.27 | 140 | 2.3282 | |
|
| 1.9049 | 1.36 | 150 | 1.0767 | |
|
| 0.8589 | 1.45 | 160 | 0.6321 | |
|
| 0.5045 | 1.54 | 170 | 0.4553 | |
|
| 0.5118 | 1.63 | 180 | 0.3732 | |
|
| 0.3558 | 1.72 | 190 | 0.3349 | |
|
| 0.3136 | 1.81 | 200 | 0.3106 | |
|
| 0.3007 | 1.9 | 210 | 0.2946 | |
|
| 0.3073 | 1.99 | 220 | 0.2602 | |
|
| 0.3275 | 2.08 | 230 | 0.4014 | |
|
| 0.5342 | 2.18 | 240 | 0.4574 | |
|
| 0.3191 | 2.27 | 250 | 0.3423 | |
|
| 0.2647 | 2.36 | 260 | 0.2386 | |
|
| 0.23 | 2.45 | 270 | 0.2366 | |
|
| 0.2493 | 2.54 | 280 | 0.2132 | |
|
| 0.2129 | 2.63 | 290 | 0.2020 | |
|
| 0.2034 | 2.72 | 300 | 0.1892 | |
|
| 0.225 | 2.81 | 310 | 0.1930 | |
|
| 0.1989 | 2.9 | 320 | 0.1899 | |
|
| 0.1969 | 2.99 | 330 | 0.1875 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.0.dev0 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.14.6 |
|
- Tokenizers 0.14.0 |
|
|
