|
|
--- |
|
|
license: mit |
|
|
base_model: microsoft/phi-2 |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: V0320MP4 |
|
|
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. --> |
|
|
|
|
|
# V0320MP4 |
|
|
|
|
|
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: 0.1222 |
|
|
|
|
|
## 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 | |
|
|
|:-------------:|:-----:|:----:|:---------------:| |
|
|
| 2.5867 | 0.09 | 10 | 2.4416 | |
|
|
| 2.2588 | 0.18 | 20 | 1.9249 | |
|
|
| 1.7112 | 0.27 | 30 | 1.3582 | |
|
|
| 1.2358 | 0.36 | 40 | 0.9648 | |
|
|
| 0.8554 | 0.45 | 50 | 0.5703 | |
|
|
| 0.4962 | 0.54 | 60 | 0.2624 | |
|
|
| 0.2702 | 0.63 | 70 | 0.1760 | |
|
|
| 0.2066 | 0.73 | 80 | 0.1524 | |
|
|
| 0.1811 | 0.82 | 90 | 0.1424 | |
|
|
| 0.1714 | 0.91 | 100 | 0.1379 | |
|
|
| 0.1628 | 1.0 | 110 | 0.1350 | |
|
|
| 0.1565 | 1.09 | 120 | 0.1331 | |
|
|
| 0.155 | 1.18 | 130 | 0.1316 | |
|
|
| 0.1554 | 1.27 | 140 | 0.1302 | |
|
|
| 0.1558 | 1.36 | 150 | 0.1289 | |
|
|
| 0.148 | 1.45 | 160 | 0.1276 | |
|
|
| 0.1467 | 1.54 | 170 | 0.1266 | |
|
|
| 0.1511 | 1.63 | 180 | 0.1257 | |
|
|
| 0.1448 | 1.72 | 190 | 0.1250 | |
|
|
| 0.1391 | 1.81 | 200 | 0.1244 | |
|
|
| 0.1414 | 1.9 | 210 | 0.1240 | |
|
|
| 0.1421 | 1.99 | 220 | 0.1236 | |
|
|
| 0.1449 | 2.08 | 230 | 0.1232 | |
|
|
| 0.1412 | 2.18 | 240 | 0.1229 | |
|
|
| 0.1393 | 2.27 | 250 | 0.1227 | |
|
|
| 0.1401 | 2.36 | 260 | 0.1225 | |
|
|
| 0.1456 | 2.45 | 270 | 0.1223 | |
|
|
| 0.1376 | 2.54 | 280 | 0.1223 | |
|
|
| 0.1369 | 2.63 | 290 | 0.1222 | |
|
|
| 0.1422 | 2.72 | 300 | 0.1222 | |
|
|
| 0.1421 | 2.81 | 310 | 0.1221 | |
|
|
| 0.1442 | 2.9 | 320 | 0.1221 | |
|
|
| 0.1432 | 2.99 | 330 | 0.1222 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.36.0.dev0 |
|
|
- Pytorch 2.1.2+cu121 |
|
|
- Datasets 2.14.6 |
|
|
- Tokenizers 0.14.1 |
|
|
|
