README.md

---
license: apache-2.0
datasets:
- NeelNanda/pile-10k
language:
- en
---


## Model Details

This model is an int4 model with group_size 128 of [Qwen/Qwen1.5-0.5B-Chat](https://huggingface.co/Qwen/Qwen1.5-0.5B-Chat)  generated by [intel/auto-round](https://github.com/intel/auto-round). 
Inference of this model is compatible with AutoGPTQ's Kernel.







### Reproduce the model

Here is the sample command to reproduce the model

```bash
git clone https://github.com/intel/auto-round
cd auto-round/examples/language-modeling
pip install -r requirements.txt
python3 main.py \
--model_name  Qwen/Qwen1.5-0.5B-Chat \
--device 0 \
--group_size 128 \
--bits 4 \
--iters 200 \
--nsamples 512 \
--deployment_device 'gpu' \
--minmax_lr 2e-3 \
--output_dir "./tmp_autoround" \

```




### Evaluate the model 

Install [lm-eval-harness 0.4.2](https://github.com/EleutherAI/lm-evaluation-harness.git) from source.

```bash
lm_eval --model hf --model_args pretrained="Intel/Qwen1.5-0.5B-Chat-int4-inc",autogptq=True,gptq_use_triton=True --device cuda:0 --tasks lambada_openai,hellaswag,piqa,winogrande,truthfulqa_mc1,openbookqa,boolq,arc_easy,arc_challenge,mmlu --batch_size 32
```



| Metric         | FP16   | INT4   |
| -------------- | ------ | ------ |
| Avg.           | 0.3896 | 0.3848 |
| mmlu           | 0.3162 | 0.2911 |
| lambada_openai | 0.4137 | 0.4002 |
| hellaswag      | 0.3630 | 0.3601 |
| winogrande     | 0.5509 | 0.5430 |
| piqa           | 0.6730 | 0.6632 |
| truthfulqa_mc1 | 0.2583 | 0.2485 |
| openbookqa     | 0.1880 | 0.1880 |
| boolq          | 0.3994 | 0.4379 |
| arc_easy       | 0.4848 | 0.4697 |
| arc_challenge  | 0.2491 | 0.2466 |





## Ethical Considerations and Limitations

The model can produce factually incorrect output, and should not be relied on to produce factually accurate information. Because of the limitations of the pretrained model and the finetuning datasets, it is possible that this model could generate lewd, biased or otherwise offensive outputs.

Therefore, before deploying any applications of the model, developers should perform safety testing.



## Caveats and Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.

Here are a couple of useful links to learn more about Intel's AI software:

* Intel Neural Compressor [link](https://github.com/intel/neural-compressor)
* Intel Extension for Transformers [link](https://github.com/intel/intel-extension-for-transformers)



## Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please consult an attorney before using this model for commercial purposes.



## Cite

@article{cheng2023optimize, title={Optimize weight rounding via signed gradient descent for the quantization of llms}, author={Cheng, Wenhua and Zhang, Weiwei and Shen, Haihao and Cai, Yiyang and He, Xin and Lv, Kaokao}, journal={arXiv preprint arXiv:2309.05516}, year={2023} }

[arxiv](https://arxiv.org/abs/2309.05516) [github](https://github.com/intel/auto-round)