OPEA/Meta-Llama-3.1-70B-Instruct-int4-sym-inc

Model Card Details

This model is an int4 model with group_size 128 and symmetric quantization of meta-llama/Meta-Llama-3.1-70B-Instruct generated by intel/auto-round. Load the model with revision 90c15db to use AutoGPTQ format

Inference on CPU/HPU/CUDA

HPU: docker image with Gaudi Software Stack is recommended, please refer to following script for environment setup. More details can be found in Gaudi Guide.

from auto_round import AutoHfQuantizer ##must import for auto-round format
import torch
from transformers import AutoModelForCausalLM,AutoTokenizer
quantized_model_dir = "OPEA/Meta-Llama-3.1-70B-Instruct-int4-sym-inc"
tokenizer = AutoTokenizer.from_pretrained(quantized_model_dir)

model = AutoModelForCausalLM.from_pretrained(
    quantized_model_dir,
    torch_dtype='auto',
    device_map="auto",
    ##revision="90c15db", ##AutoGPTQ format
)

##import habana_frameworks.torch.core as htcore ## uncommnet it for HPU
##import habana_frameworks.torch.hpu as hthpu ## uncommnet it for HPU
##model = model.to(torch.bfloat16).to("hpu") ## uncommnet it for HPU

prompt = "There is a girl who likes adventure,"
messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": prompt}
]

tokenizer = AutoTokenizer.from_pretrained(quantized_model_dir)
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

generated_ids = model.generate(
    model_inputs.input_ids,
    max_new_tokens=200,  ##change this to align with the official usage
    do_sample=False  ##change this to align with the official usage
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]

response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)

##prompt = "There is a girl who likes adventure,"
##BF16 
"""That sounds exciting. What kind of adventures is she interested in? Is she more into outdoor activities like hiking, rock climbing, or exploring new places, or does she enjoy indoor adventures like solving puzzles, playing escape rooms, or reading fantasy novels?
"""
##INT4
"""That sounds exciting. What kind of adventures is she interested in? Is she more into outdoor activities like hiking, camping, or exploring new places, or is she drawn to thrilling experiences like skydiving, bungee jumping, or trying new extreme sports?
"""

##prompt = "Which one is larger, 9.11 or 9.8"
## INT4
"""9.11 is larger than 9.8."""

## BF16
"""9.11 is larger than 9.8."""

prompt = "How many r in strawberry."
## INT4
"""There are 2 R's in the word "strawberry""
## BF16
"""There are 2 R's in the word "strawberry"."""

##prompt = "Once upon a time,"
## INT4
"""It sounds like you're starting a story. Would you like me to continue it, or would you like to tell me the rest of the story yourself?
"""
## BF16 
"""it seems like we're about to start a classic fairy tale. Would you like to continue the story, or would you like me to take over and spin a yarn for you?
"""

Evaluate the model

pip3 install lm-eval==0.4.5

 auto-round --eval --model "OPEA/Meta-Llama-3.1-70B-Instruct-int4-sym-inc" --eval_bs 16  --tasks leaderboard_mmlu_pro,leaderboard_ifeval,lambada_openai,hellaswag,piqa,winogrande,truthfulqa_mc1,openbookqa,boolq,arc_easy,arc_challenge,mmlu,gsm8k

Metric	BF16	INT4
avg	0.69565	0.6945
leaderboard_mmlu_pro 5shot	0.5309	0.5226
leaderboard_ifeval	0.7582=(0.8010+0.7153)/2	0.75725=(0.8010+0.7135)/2
lambada_openai	0.7557	0.7572
hellaswag	0.6516	0.6467
winogrande	0.7861	0.8098
piqa	0.8313	0.8243
truthfulqa_mc1	0.4064	0.4027
openbookqa	0.3700	0.3620
boolq	0.8783	0.8761
arc_easy	0.8670	0.8590
arc_challenge	0.6237	0.6101
gsm8k(5shot) strict match	0.8886	0.9067

Reproduce the model

Here is the sample command to reproduce the model. We found auto-round is not stable for this model, please do not use --model_dtype "fp16" and symmetric quantization.

auto-round  \
--model  meta-llama/Meta-Llama-3.1-70B-Instruct \
--device 0 \
--group_size 128 \
--nsamples 512 \
--bits 4 \
--iter 1000 \
--disable_eval \
--low_gpu_mem_usage \
--format 'auto_round' \
--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

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