---
language:
- en
- zh
license: apache-2.0
pipeline_tag: text-generation
library_name: transformers
---

# BlockFFN-3B-SFT-EAGLE

This is the 3B BlockFFN model used in the paper *BlockFFN: Towards End-Side Acceleration-Friendly Mixture-of-Experts with Chunk-Level Activation Sparsity* for acceleration tests.

**BlockFFN** introduces a novel Mixture-of-Experts (MoE) architecture designed for efficient inference, particularly on end-side devices. It aims to achieve high token-level and chunk-level sparsity, making it acceleration-friendly and compatible with techniques like speculative decoding. This model is based on the [paper](https://arxiv.org/pdf/2507.08771).

For the full codebase and more details, visit the official [GitHub repository](https://github.com/thunlp/BlockFFN).

### Usage

You can easily load and use this model with the Hugging Face `transformers` library:

```python
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
import torch

model_name = "SparseLLM/BlockFFN-3B-SFT-EAGLE"

# Load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16, # or torch.float16 if bfloat16 is not supported
    device_map="auto",
    trust_remote_code=True,
)

# Create a text generation pipeline
pipe = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
)

# Generate text
prompt = "The quick brown fox jumps over the lazy"
output = pipe(prompt, max_new_tokens=50, do_sample=True, temperature=0.7)

print(output[0]['generated_text'])
```

### Citation

If you find our work useful for your research, please kindly cite our paper as follows:

```
@article{song2025blockffn,
      title={{BlockFFN}: Towards End-Side Acceleration-Friendly Mixture-of-Experts with Chunk-Level Activation Sparsity},
      author={Chenyang Song and Weilin Zhao and Xu Han and Chaojun Xiao and Yingfa Chen and Yuxuan Li and Zhiyuan Liu and Maosong Sun},
      journal={arXiv preprint arXiv:2507.08771},
      year={2025},
      url={https://arxiv.org/pdf/2507.08771},
}