|
|
--- |
|
|
license: mit |
|
|
base_model: microsoft/phi-2 |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: V0320MP6 |
|
|
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. --> |
|
|
|
|
|
# V0320MP6 |
|
|
|
|
|
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: 0.1245 |
|
|
|
|
|
## 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: 20 |
|
|
- num_epochs: 3 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|
|:-------------:|:-----:|:----:|:---------------:| |
|
|
| 2.5658 | 0.09 | 10 | 2.4014 | |
|
|
| 2.2012 | 0.18 | 20 | 1.8503 | |
|
|
| 1.6313 | 0.27 | 30 | 1.2804 | |
|
|
| 1.1536 | 0.36 | 40 | 0.8569 | |
|
|
| 0.7302 | 0.45 | 50 | 0.3937 | |
|
|
| 0.3829 | 0.54 | 60 | 0.2028 | |
|
|
| 0.2398 | 0.63 | 70 | 0.1611 | |
|
|
| 0.2003 | 0.73 | 80 | 0.1476 | |
|
|
| 0.1817 | 0.82 | 90 | 0.1419 | |
|
|
| 0.1747 | 0.91 | 100 | 0.1384 | |
|
|
| 0.1706 | 1.0 | 110 | 0.1359 | |
|
|
| 0.1557 | 1.09 | 120 | 0.1340 | |
|
|
| 0.1601 | 1.18 | 130 | 0.1327 | |
|
|
| 0.1575 | 1.27 | 140 | 0.1314 | |
|
|
| 0.1625 | 1.36 | 150 | 0.1304 | |
|
|
| 0.1509 | 1.45 | 160 | 0.1293 | |
|
|
| 0.1481 | 1.54 | 170 | 0.1286 | |
|
|
| 0.1552 | 1.63 | 180 | 0.1275 | |
|
|
| 0.1465 | 1.72 | 190 | 0.1267 | |
|
|
| 0.1434 | 1.81 | 200 | 0.1268 | |
|
|
| 0.1475 | 1.9 | 210 | 0.1260 | |
|
|
| 0.1478 | 1.99 | 220 | 0.1253 | |
|
|
| 0.147 | 2.08 | 230 | 0.1254 | |
|
|
| 0.1438 | 2.18 | 240 | 0.1252 | |
|
|
| 0.1443 | 2.27 | 250 | 0.1250 | |
|
|
| 0.146 | 2.36 | 260 | 0.1249 | |
|
|
| 0.1446 | 2.45 | 270 | 0.1246 | |
|
|
| 0.1414 | 2.54 | 280 | 0.1247 | |
|
|
| 0.1414 | 2.63 | 290 | 0.1245 | |
|
|
| 0.1436 | 2.72 | 300 | 0.1247 | |
|
|
| 0.1469 | 2.81 | 310 | 0.1246 | |
|
|
| 0.143 | 2.9 | 320 | 0.1247 | |
|
|
| 0.1456 | 2.99 | 330 | 0.1245 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.36.0.dev0 |
|
|
- Pytorch 2.1.2+cu121 |
|
|
- Datasets 2.14.6 |
|
|
- Tokenizers 0.14.1 |
|
|
|
