|
--- |
|
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_6 |
|
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_6 |
|
|
|
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: 0.1212 |
|
|
|
## 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: 60 |
|
- training_steps: 500 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 7.3419 | 0.31 | 20 | 6.2616 | |
|
| 4.0421 | 0.63 | 40 | 0.8963 | |
|
| 0.6465 | 0.94 | 60 | 0.5726 | |
|
| 0.4575 | 1.26 | 80 | 0.3999 | |
|
| 0.309 | 1.57 | 100 | 0.3044 | |
|
| 0.2531 | 1.89 | 120 | 0.2605 | |
|
| 0.2235 | 2.2 | 140 | 0.2273 | |
|
| 0.1922 | 2.52 | 160 | 0.2091 | |
|
| 0.1793 | 2.83 | 180 | 0.1858 | |
|
| 0.1488 | 3.14 | 200 | 0.1734 | |
|
| 0.16 | 3.46 | 220 | 0.1646 | |
|
| 0.1497 | 3.77 | 240 | 0.1557 | |
|
| 0.1336 | 4.09 | 260 | 0.1489 | |
|
| 0.1278 | 4.4 | 280 | 0.1415 | |
|
| 0.1291 | 4.72 | 300 | 0.1392 | |
|
| 0.1244 | 5.03 | 320 | 0.1342 | |
|
| 0.1184 | 5.35 | 340 | 0.1319 | |
|
| 0.118 | 5.66 | 360 | 0.1289 | |
|
| 0.1153 | 5.97 | 380 | 0.1279 | |
|
| 0.1052 | 6.29 | 400 | 0.1250 | |
|
| 0.1058 | 6.6 | 420 | 0.1243 | |
|
| 0.1142 | 6.92 | 440 | 0.1226 | |
|
| 0.1026 | 7.23 | 460 | 0.1222 | |
|
| 0.1051 | 7.55 | 480 | 0.1214 | |
|
| 0.0977 | 7.86 | 500 | 0.1212 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.1 |
|
- Transformers 4.36.2 |
|
- Pytorch 2.0.0 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
