This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of FlameF0X/TinyMoE-100m-2x8-retrained generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	FlameF0X/TinyMoE-100m-2x8-retrained
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	84 MB

Evaluation Results

Task	Accuracy
hellaswag	0.2603
mmlu	0.2292
mmlu_abstract_algebra	0.2200
mmlu_anatomy	0.1926
mmlu_astronomy	0.1711
mmlu_business_ethics	0.2700
mmlu_clinical_knowledge	0.2113
mmlu_college_biology	0.2708
mmlu_college_chemistry	0.1600
mmlu_college_computer_science	0.2600
mmlu_college_mathematics	0.2200
mmlu_college_medicine	0.2081
mmlu_college_physics	0.2157
mmlu_computer_security	0.2700
mmlu_conceptual_physics	0.2596
mmlu_econometrics	0.2281
mmlu_electrical_engineering	0.2345
mmlu_elementary_mathematics	0.2143
mmlu_formal_logic	0.2778
mmlu_global_facts	0.1800
mmlu_high_school_biology	0.2000
mmlu_high_school_chemistry	0.1872
mmlu_high_school_computer_science	0.2900
mmlu_high_school_european_history	0.2242
mmlu_high_school_geography	0.1818
mmlu_high_school_government_and_politics	0.1917
mmlu_high_school_macroeconomics	0.2026
mmlu_high_school_mathematics	0.2111
mmlu_high_school_microeconomics	0.2101
mmlu_high_school_physics	0.2053
mmlu_high_school_psychology	0.1908
mmlu_high_school_statistics	0.1481
mmlu_high_school_us_history	0.2696
mmlu_high_school_world_history	0.2743
mmlu_human_aging	0.3094
mmlu_human_sexuality	0.2595
mmlu_humanities	0.2431
mmlu_international_law	0.2397
mmlu_jurisprudence	0.2685
mmlu_logical_fallacies	0.2270
mmlu_machine_learning	0.2946
mmlu_management	0.1845
mmlu_marketing	0.2949
mmlu_medical_genetics	0.2700
mmlu_miscellaneous	0.2350
mmlu_moral_disputes	0.2486
mmlu_moral_scenarios	0.2380
mmlu_nutrition	0.2026
mmlu_other	0.2330
mmlu_philosophy	0.1865
mmlu_prehistory	0.2068
mmlu_professional_accounting	0.2305
mmlu_professional_law	0.2458
mmlu_professional_medicine	0.1801
mmlu_professional_psychology	0.2516
mmlu_public_relations	0.2091
mmlu_security_studies	0.1878
mmlu_social_sciences	0.2164
mmlu_sociology	0.2438
mmlu_stem	0.2169
mmlu_us_foreign_policy	0.2800
mmlu_virology	0.2590
mmlu_world_religions	0.3275
piqa	0.5468

How to Use

HF Usage

Step 1: Install AutoRound

pip install auto-round

Step 2: Load and run the quantized model

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "TinyMoE-100m-2x8-retrained-AutoRound-W4A16-RTN"

# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")

# prepare the model input
prompt = "Write a quick sort algorithm."
messages = [{"role": "user", "content": prompt}]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

# conduct text completion
generated_ids = model.generate(**model_inputs, max_new_tokens=512)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]) :].tolist()

content = tokenizer.decode(output_ids, skip_special_tokens=True)
print("content:", content)

VLLM Usage

bash

vllm serve TinyMoE-100m-2x8-retrained-AutoRound-W4A16-RTN \
    --trust-remote-code \
    --dtype bfloat16 \
    --tensor_parallel_size 1

If you encounter any issues, feel free to open an issue on the AutoRound GitHub repo or provide feedback on the Low-Bit Open LLM Leaderboard.

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:

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

This model is part of the Intel Low-Bit Open LLM Leaderboard initiative.

LeaderboardModel1/TinyMoE-100m-2x8-retrained-AutoRound-W4A16-RTN

TinyMoE-100m-2x8-retrained-AutoRound-W4A16-RTN

Model Details