Text Generation
Transformers
PyTorch
Safetensors
Chinese
English
mistral
conversational
text-generation-inference
Instructions to use MediaTek-Research/Breeze-7B-Instruct-v1_0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use MediaTek-Research/Breeze-7B-Instruct-v1_0 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="MediaTek-Research/Breeze-7B-Instruct-v1_0") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("MediaTek-Research/Breeze-7B-Instruct-v1_0") model = AutoModelForCausalLM.from_pretrained("MediaTek-Research/Breeze-7B-Instruct-v1_0", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=256) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use MediaTek-Research/Breeze-7B-Instruct-v1_0 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "MediaTek-Research/Breeze-7B-Instruct-v1_0" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "MediaTek-Research/Breeze-7B-Instruct-v1_0", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/MediaTek-Research/Breeze-7B-Instruct-v1_0
- SGLang
How to use MediaTek-Research/Breeze-7B-Instruct-v1_0 with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "MediaTek-Research/Breeze-7B-Instruct-v1_0" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "MediaTek-Research/Breeze-7B-Instruct-v1_0", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "MediaTek-Research/Breeze-7B-Instruct-v1_0" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "MediaTek-Research/Breeze-7B-Instruct-v1_0", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use MediaTek-Research/Breeze-7B-Instruct-v1_0 with Docker Model Runner:
docker model run hf.co/MediaTek-Research/Breeze-7B-Instruct-v1_0
model()與model.generate()的output相同嗎?
#3
by cathat - opened
我只是一個路人,但根據我遠古以前(2 年前)的經驗,我記得 model() 跟 model.generate() 在考慮的 context 會不同。這是 Huggingface API 的問題
簡言之前者是 greedy decoding (類似 teacher-forcing 的概念,也因而比較快) 但後者是 auto-regressive (考慮自己生成的 token ,因此可能會比較慢),而且還支援各種 sampling 方法
這邊應該可以參考: https://discuss.huggingface.co/t/what-is-the-difference-between-forward-and-generate/10235
No description provided.
cathat changed discussion status to closed
cathat changed discussion status to open
cathat changed discussion status to closed
cathat changed discussion status to open
謝謝您抽空回答我的問題,這幫了我很大的忙。
cathat changed discussion status to closed


