|
--- |
|
license: mit |
|
library_name: peft |
|
tags: |
|
- generated_from_trainer |
|
base_model: microsoft/phi-2 |
|
model-index: |
|
- name: fine-tuning-Phi2-with-webglm-qa-with-lora_4 |
|
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. --> |
|
|
|
# fine-tuning-Phi2-with-webglm-qa-with-lora_4 |
|
|
|
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: 0.1147 |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 2 |
|
- eval_batch_size: 2 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 5 |
|
- total_train_batch_size: 10 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 50 |
|
- training_steps: 500 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 8.2444 | 0.2 | 10 | 7.8267 | |
|
| 7.4754 | 0.4 | 20 | 6.3605 | |
|
| 4.8314 | 0.6 | 30 | 3.1457 | |
|
| 1.7327 | 0.8 | 40 | 0.6363 | |
|
| 0.5438 | 1.0 | 50 | 0.5673 | |
|
| 0.4569 | 1.2 | 60 | 0.4906 | |
|
| 0.4491 | 1.39 | 70 | 0.4269 | |
|
| 0.367 | 1.59 | 80 | 0.3729 | |
|
| 0.2821 | 1.79 | 90 | 0.3323 | |
|
| 0.2414 | 1.99 | 100 | 0.3013 | |
|
| 0.2521 | 2.19 | 110 | 0.2772 | |
|
| 0.2135 | 2.39 | 120 | 0.2603 | |
|
| 0.1982 | 2.59 | 130 | 0.2446 | |
|
| 0.2186 | 2.79 | 140 | 0.2278 | |
|
| 0.1741 | 2.99 | 150 | 0.2144 | |
|
| 0.1781 | 3.19 | 160 | 0.2062 | |
|
| 0.1702 | 3.39 | 170 | 0.1928 | |
|
| 0.157 | 3.59 | 180 | 0.1846 | |
|
| 0.1469 | 3.78 | 190 | 0.1770 | |
|
| 0.1644 | 3.98 | 200 | 0.1705 | |
|
| 0.1458 | 4.18 | 210 | 0.1654 | |
|
| 0.1282 | 4.38 | 220 | 0.1623 | |
|
| 0.1537 | 4.58 | 230 | 0.1568 | |
|
| 0.1197 | 4.78 | 240 | 0.1509 | |
|
| 0.1327 | 4.98 | 250 | 0.1464 | |
|
| 0.1349 | 5.18 | 260 | 0.1436 | |
|
| 0.1052 | 5.38 | 270 | 0.1409 | |
|
| 0.127 | 5.58 | 280 | 0.1381 | |
|
| 0.1303 | 5.78 | 290 | 0.1365 | |
|
| 0.1063 | 5.98 | 300 | 0.1338 | |
|
| 0.1145 | 6.18 | 310 | 0.1300 | |
|
| 0.1101 | 6.37 | 320 | 0.1287 | |
|
| 0.1088 | 6.57 | 330 | 0.1280 | |
|
| 0.1062 | 6.77 | 340 | 0.1254 | |
|
| 0.1016 | 6.97 | 350 | 0.1238 | |
|
| 0.1005 | 7.17 | 360 | 0.1232 | |
|
| 0.1084 | 7.37 | 370 | 0.1220 | |
|
| 0.101 | 7.57 | 380 | 0.1204 | |
|
| 0.1065 | 7.77 | 390 | 0.1200 | |
|
| 0.0943 | 7.97 | 400 | 0.1191 | |
|
| 0.0848 | 8.17 | 410 | 0.1184 | |
|
| 0.0913 | 8.37 | 420 | 0.1175 | |
|
| 0.1115 | 8.57 | 430 | 0.1169 | |
|
| 0.091 | 8.76 | 440 | 0.1161 | |
|
| 0.1009 | 8.96 | 450 | 0.1154 | |
|
| 0.0966 | 9.16 | 460 | 0.1150 | |
|
| 0.0931 | 9.36 | 470 | 0.1147 | |
|
| 0.0922 | 9.56 | 480 | 0.1150 | |
|
| 0.0912 | 9.76 | 490 | 0.1148 | |
|
| 0.0915 | 9.96 | 500 | 0.1147 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.1 |
|
- Transformers 4.36.2 |
|
- Pytorch 2.0.0 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
