You can deploy this model and start earning money today!

RichardErkhov/
MiniLLM_-_Pretrain-Qwen-200M-gguf

Name: MiniLLM_-_Pretrain-Qwen-200M-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Pretrain-Qwen-200M - GGUF

Model creator: https://huggingface.co/MiniLLM/
Original model: https://huggingface.co/MiniLLM/Pretrain-Qwen-200M/

Name	Quant method	Size
Pretrain-Qwen-200M.Q2_K.gguf	Q2_K	0.13GB
Pretrain-Qwen-200M.IQ3_XS.gguf	IQ3_XS	0.13GB
Pretrain-Qwen-200M.IQ3_S.gguf	IQ3_S	0.13GB
Pretrain-Qwen-200M.Q3_K_S.gguf	Q3_K_S	0.13GB
Pretrain-Qwen-200M.IQ3_M.gguf	IQ3_M	0.13GB
Pretrain-Qwen-200M.Q3_K.gguf	Q3_K	0.14GB
Pretrain-Qwen-200M.Q3_K_M.gguf	Q3_K_M	0.14GB
Pretrain-Qwen-200M.Q3_K_L.gguf	Q3_K_L	0.14GB
Pretrain-Qwen-200M.IQ4_XS.gguf	IQ4_XS	0.14GB
Pretrain-Qwen-200M.Q4_0.gguf	Q4_0	0.14GB
Pretrain-Qwen-200M.IQ4_NL.gguf	IQ4_NL	0.14GB
Pretrain-Qwen-200M.Q4_K_S.gguf	Q4_K_S	0.14GB
Pretrain-Qwen-200M.Q4_K.gguf	Q4_K	0.15GB
Pretrain-Qwen-200M.Q4_K_M.gguf	Q4_K_M	0.15GB
Pretrain-Qwen-200M.Q4_1.gguf	Q4_1	0.15GB
Pretrain-Qwen-200M.Q5_0.gguf	Q5_0	0.15GB
Pretrain-Qwen-200M.Q5_K_S.gguf	Q5_K_S	0.15GB
Pretrain-Qwen-200M.Q5_K.gguf	Q5_K	0.15GB
Pretrain-Qwen-200M.Q5_K_M.gguf	Q5_K_M	0.15GB
Pretrain-Qwen-200M.Q5_1.gguf	Q5_1	0.16GB
Pretrain-Qwen-200M.Q6_K.gguf	Q6_K	0.17GB
Pretrain-Qwen-200M.Q8_0.gguf	Q8_0	0.21GB

Original model description:

datasets:

monology/pile-uncopyrighted
MiniLLM/pile-tokenized language:
en library_name: transformers license: apache-2.0 metrics:
accuracy pipeline_tag: text-generation

Pretrain-Qwen-200M

paper | code

Pretrain-Qwen-200M is a 200M model with QWen achitecture conventionally pre-trained from scratch on the Pile for 50B tokens.

We also open-source the tokenized pre-training corpus for reproducibility.

It is used as the baseline for MiniLLM-Qwen-200M

Evaluation

MiniPLM models achieves better performance given the same computation and scales well across model sizes:

Other Baselines

VanillaKD

Citation

bibtext

@article{miniplm,
    title={MiniPLM: Knowledge Distillation for Pre-Training Language Models}, 
    author={Yuxian Gu and Hao Zhou and Fandong Meng and Jie Zhou and Minlie Huang},
    journal={arXiv preprint arXiv:2410.17215},
    year={2024}
}

RichardErkhov/MiniLLM_-_Pretrain-Qwen-200M-gguf