README.md

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

# VanillaKD-Pretrain-Qwen-200M

[paper](https://arxiv.org/abs/2410.17215) | [code](https://github.com/thu-coai/MiniPLM)

**VanillaKD-Pretrain-Qwen-200M** is a 200M model with Qwen achitecture pre-trained with vanilla token-level knowledge distillation on [the Pile](https://huggingface.co/datasets/monology/pile-uncopyrighted) for 50B tokens. The teacher model is [Qwen1.5-1.8B](https://huggingface.co/Qwen/Qwen1.5-1.8B).

We also open-source the tokenized [pre-training corpus](https://huggingface.co/datasets/MiniLLM/pile-tokenized) for reproducibility.

**It is used as the baseline for [MiniLLM-Qwen-200M](https://huggingface.co/MiniLLM/MiniPLM-Qwen-200M)**

## Evaluation

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

<p align='left'>
    <img src="https://cdn-uploads.huggingface.co/production/uploads/624ac662102fcdff87be51b9/EOYzajQcwQFT5PobqL3j0.png" width="1000">
</p>

## Other Baselines
+ [Conventional Pre-Training](https://huggingface.co/MiniLLM/Pretrain-Qwen-200M)

## 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}
}
```