---
pipeline_tag: text-generation
license: apache-2.0
language:
- zh
- en
---
# Model Card for MediaTek Research Breeze-7B-32k-Instruct-v1_0
MediaTek Research Breeze-7B (hereinafter referred to as Breeze-7B) is a language model family that builds on top of [Mistral-7B](https://huggingface.co/mistralai/Mistral-7B-v0.1), specifically intended for Traditional Chinese use.
[Breeze-7B-Base](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v1_0) is the base model for the Breeze-7B series.
It is suitable for use if you have substantial fine-tuning data to tune it for your specific use case.
[Breeze-7B-Instruct](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-v1_0) derives from the base model Breeze-7B-Base, making the resulting model amenable to be used as-is for commonly seen tasks.
[Breeze-7B-32k-Base](https://huggingface.co/MediaTek-Research/Breeze-7B-32k-Base-v1_0) is extended from the base model with more data, base change, and the disabling of the sliding window.
Roughly speaking, that is equivalent to 44k Traditional Chinese characters.
[Breeze-7B-32k-Instruct](https://huggingface.co/MediaTek-Research/Breeze-7B-32k-Instruct-v1_0) derives from the base model Breeze-7B-32k-Base, making the resulting model amenable to be used as-is for commonly seen tasks.
Practicality-wise:
- Breeze-7B-Base expands the original vocabulary with additional 30,000 Traditional Chinese tokens. With the expanded vocabulary, everything else being equal, Breeze-7B operates at twice the inference speed for Traditional Chinese to Mistral-7B and Llama 7B. [See [Inference Performance](#inference-performance).]
- Breeze-7B-Instruct can be used as is for common tasks such as Q&A, RAG, multi-round chat, and summarization.
- Breeze-7B-32k-Instruct can perform tasks at a document level (For Chinese, 20 ~ 40 pages).
*A project by the members (in alphabetical order): Chan-Jan Hsu 許湛然, Feng-Ting Liao 廖峰挺, Po-Chun Hsu 許博竣, Yi-Chang Chen 陳宜昌, and the supervisor Da-Shan Shiu 許大山.*
## Features
- Breeze-7B-32k-Base-v1_0
- Expanding the vocabulary dictionary size from 32k to 62k to better support Traditional Chinese
- 32k-token context length
- Breeze-7B-32k-Instruct-v1_0
- Expanding the vocabulary dictionary size from 32k to 62k to better support Traditional Chinese
- 32k-token context length
- Multi-turn dialogue (without special handling for harmfulness)
## Model Details
- Breeze-7B-32k-Base-v1_0
- Pretrained from: [Breeze-7B-Base](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v1_0)
- Model type: Causal decoder-only transformer language model
- Language: English and Traditional Chinese (zh-tw)
- Breeze-7B-32k-Instruct-v1_0
- Finetuned from: [Breeze-7B-32k-Base](https://huggingface.co/MediaTek-Research/Breeze-7B-32k-Base-v1_0)
- Model type: Causal decoder-only transformer language model
- Language: English and Traditional Chinese (zh-tw)
## Long-context Performance
#### Needle-in-a-haystack Performance
We use the passkey retrieval task to test the model's ability to attend to different various depths in a given sequence.
A key in placed within a long context distracting document for the model to retrieve.
The key position is binned into 16 bins, and there are 20 testcases for each bin.
Breeze-7B-32k-Base clears the tasks with 90+% accuracy, shown in the figure below.
#### Long-DRCD Performance
| **Model/Performance(EM)** | **DRCD** | **DRCD-16k** | **DRCD-32k** |
|---------------------------|----------|--------------|--------------|
| **Breeze-7B-32k-Instruct-v1\_0** | 76.9 | 54.82 | 44.26 |
| **Breeze-7B-32k-Base-v1\_0** | 79.73 | 69.68 | 61.55 |
| **Breeze-7B-Base-v1\_0** | 80.61 | 21.79 | 15.29 |
#### Short-Benchmark Performance
| **Model/Performance(EM)** | **TMMLU+** | **MMLU** | **TABLE** | **MT-Bench-tw** | **MT-Bench** |
|---------------------------|----------|--------------|--------------|-----|-----|
| **Breeze-7B-32k-Instruct-v1\_0** | 41.37 | 61.34 | 34 | 5.8 | 7.4 |
| **Breeze-7B-Instruct-v1\_0** | 42.67 | 62.73 | 39.58 | 6.0 | 7.4 |
## Use in Transformers
First, install direct dependencies:
```
pip install transformers torch accelerate
```
Flash-attention2 is strongly recommended for long context scenarios.
```bash
pip install packaging ninja
pip install flash-attn
```
Then load the model in transformers:
```python
>>> from transformers import AutoModelForCausalLM, AutoTokenizer
>>> tokenizer = AutoTokenizer.from_pretrained("MediaTek-Research/Breeze-7B-32k-Instruct-v1_0/")
>>> model = AutoModelForCausalLM.from_pretrained(
>>> "MediaTek-Research/Breeze-7B-32k-Instruct-v1_0",
... device_map="auto",
... torch_dtype=torch.bfloat16,
... attn_implementation="flash_attention_2"
... )
>>> chat = [
... {"role": "user", "content": "你好，請問你可以完成什麼任務？"},
... {"role": "assistant", "content": "你好，我可以幫助您解決各種問題、提供資訊和協助您完成許多不同的任務。例如：回答技術問題、提供建議、翻譯文字、尋找資料或協助您安排行程等。請告訴我如何能幫助您。"},
... {"role": "user", "content": "太棒了！"},
... ]
>>> tokenizer.apply_chat_template(chat, tokenize=False)
"You are a helpful AI assistant built by MediaTek Research. The user you are helping speaks Traditional Chinese and comes from Taiwan. [INST] 你好，請問你可以完成什麼任務？ [/INST] 你好，我可以幫助您解決各種問題、提供資訊和協助您完成許多不同的任務。例如：回答技術問題、提供建議、翻譯文字、尋找資料或協助您安排行程等。請告訴我如何能幫助您。 [INST] 太棒了！ [/INST] "
# Tokenized results
# ['▁', '你好', '，', '請問', '你', '可以', '完成', '什麼', '任務', '？']
# ['▁', '你好', '，', '我', '可以', '幫助', '您', '解決', '各種', '問題', '、', '提供', '資訊', '和', '協助', '您', '完成', '許多', '不同', '的', '任務', '。', '例如', '：', '回答', '技術', '問題', '、', '提供', '建議', '、', '翻譯', '文字', '、', '尋找', '資料', '或', '協助', '您', '安排', '行程', '等', '。', '請', '告訴', '我', '如何', '能', '幫助', '您', '。']
# ['▁', '太', '棒', '了', '！']
```
## Citation
```
@article{MediaTek-Research2024breeze7b,
title={Breeze-7B Technical Report},
author={Chan-Jan Hsu and Chang-Le Liu and Feng-Ting Liao and Po-Chun Hsu and Yi-Chang Chen and Da-Shan Shiu},
year={2024},
eprint={2403.02712},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```