|
--- |
|
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: 0.1176 |
|
|
|
## 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 | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 8.1178 | 0.2 | 10 | 7.7550 | |
|
| 7.3762 | 0.4 | 20 | 6.3827 | |
|
| 4.9217 | 0.6 | 30 | 3.2172 | |
|
| 1.7792 | 0.8 | 40 | 0.6700 | |
|
| 0.5779 | 1.0 | 50 | 0.5969 | |
|
| 0.4824 | 1.2 | 60 | 0.5149 | |
|
| 0.4689 | 1.39 | 70 | 0.4440 | |
|
| 0.3833 | 1.59 | 80 | 0.3862 | |
|
| 0.2916 | 1.79 | 90 | 0.3364 | |
|
| 0.2435 | 1.99 | 100 | 0.3013 | |
|
| 0.2538 | 2.19 | 110 | 0.2779 | |
|
| 0.2147 | 2.39 | 120 | 0.2619 | |
|
| 0.1982 | 2.59 | 130 | 0.2453 | |
|
| 0.2183 | 2.79 | 140 | 0.2275 | |
|
| 0.1737 | 2.99 | 150 | 0.2148 | |
|
| 0.1794 | 3.19 | 160 | 0.2068 | |
|
| 0.1692 | 3.39 | 170 | 0.1949 | |
|
| 0.1573 | 3.59 | 180 | 0.1864 | |
|
| 0.1478 | 3.78 | 190 | 0.1788 | |
|
| 0.164 | 3.98 | 200 | 0.1732 | |
|
| 0.1454 | 4.18 | 210 | 0.1676 | |
|
| 0.1279 | 4.38 | 220 | 0.1653 | |
|
| 0.1544 | 4.58 | 230 | 0.1595 | |
|
| 0.1206 | 4.78 | 240 | 0.1524 | |
|
| 0.1334 | 4.98 | 250 | 0.1486 | |
|
| 0.1342 | 5.18 | 260 | 0.1472 | |
|
| 0.1061 | 5.38 | 270 | 0.1442 | |
|
| 0.1265 | 5.58 | 280 | 0.1427 | |
|
| 0.131 | 5.78 | 290 | 0.1389 | |
|
| 0.1067 | 5.98 | 300 | 0.1374 | |
|
| 0.1158 | 6.18 | 310 | 0.1331 | |
|
| 0.1114 | 6.37 | 320 | 0.1323 | |
|
| 0.1104 | 6.57 | 330 | 0.1311 | |
|
| 0.108 | 6.77 | 340 | 0.1281 | |
|
| 0.1015 | 6.97 | 350 | 0.1271 | |
|
| 0.1 | 7.17 | 360 | 0.1262 | |
|
| 0.1091 | 7.37 | 370 | 0.1242 | |
|
| 0.1013 | 7.57 | 380 | 0.1230 | |
|
| 0.1074 | 7.77 | 390 | 0.1233 | |
|
| 0.0946 | 7.97 | 400 | 0.1226 | |
|
| 0.0854 | 8.17 | 410 | 0.1222 | |
|
| 0.0914 | 8.37 | 420 | 0.1205 | |
|
| 0.1117 | 8.57 | 430 | 0.1198 | |
|
| 0.0922 | 8.76 | 440 | 0.1194 | |
|
| 0.1012 | 8.96 | 450 | 0.1185 | |
|
| 0.0964 | 9.16 | 460 | 0.1185 | |
|
| 0.0948 | 9.36 | 470 | 0.1181 | |
|
| 0.0943 | 9.56 | 480 | 0.1178 | |
|
| 0.0915 | 9.76 | 490 | 0.1176 | |
|
| 0.0924 | 9.96 | 500 | 0.1176 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.1 |
|
- Transformers 4.36.2 |
|
- Pytorch 2.0.0 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
