##pip install auto-gptq==0.7.1
from transformers import AutoModelForCausalLM,AutoTokenizer
quantized_model_dir = "Intel/Qwen2-1.5B-Insturct-int4-inc"
tokenizer = AutoTokenizer.from_pretrained(quantized_model_dir)
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
text = "There is a girl who likes adventure,"
text = "Once upon a time,"
inputs = tokenizer(text, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(**inputs, max_new_tokens=50, do_sample=False)[0]))
##There is a girl who likes adventure, she is always looking for new things to do. She is very adventurous and loves to try new things. She is always looking for new things to do and is always looking for new things to try. She is always looking for new things to do and

##Once upon a time, there was a young boy named Jack. He lived in a small village in England. Jack was very curious and loved to explore the world around him. One day, he decided to go on a journey to find a magical tree that could grant him three

Evaluate the model

pip3 install lm-eval==0.4.2

lm_eval --model hf --model_args pretrained="Intel/Qwen2-1.5B-Instuct-int4-inc" --device cuda:0 --tasks lambada_openai,hellaswag,piqa,winogrande,truthfulqa_mc1,openbookqa,boolq,arc_easy,arc_challenge,mmlu --batch_size 16

Metric-1.5B-instruct	BF16	INT4
Avg	0.5518	0.5474
mmlu	0.5565	0.5499
lambada_openai	0.6484	0.6412
hellaswag	0.4935	0.4859
winogrande	0.6543	0.6606
piqa	0.7612	0.7573
truthfulqa_mc1	0.2827	0.2766
openbookqa	0.2780	0.274
boolq	0.7654	0.7639
arc_easy	0.7024	0.6949
arc_challenge	0.3754	0.3695

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/Qwen2-1.5B-Instruct \
--device 0 \
--group_size 32 \
--bits 4 \
--sym \
--iter 1000 \
--deployment_device 'gpu' \
--output_dir "./tmp_autoround" 

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
Intel Extension for Transformers link

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 and Liu, Yi}, journal={arXiv preprint arXiv:2309.05516}, year={2023} }

arxiv github