lm-sys / FastChat

An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
Apache License 2.0
36.55k stars 4.51k forks source link

qwen-72b-chat runs in Fastchat vllm , the input has weird message that is not mine. #2902

Open dongxu opened 8 months ago

dongxu commented 8 months ago

The input is blow, it has massive message that is not mine.

INFO 01-09 11:27:21 async_llm_engine.py:370] Received request 386a4fc3c7824a57a7c894a38db53934: prompt: "A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.\n### Human: Got any creative ideas for a 10 year old’s birthday?\n### Assistant: Of course! Here are some creative ideas for a 10-year-old's birthday party:\n1. Treasure Hunt: Organize a treasure hunt in your backyard or nearby park. Create clues and riddles for the kids to solve, leading them to hidden treasures and surprises.\n2. Science Party: Plan a science-themed party where kids can engage in fun and interactive experiments. You can set up different stations with activities like making slime, erupting volcanoes, or creating simple chemical reactions.\n3. Outdoor Movie Night: Set up a backyard movie night with a projector and a large screen or white sheet. Create a cozy seating area with blankets and pillows, and serve popcorn and snacks while the kids enjoy a favorite movie under the stars.\n4. DIY Crafts Party: Arrange a craft party where kids can unleash their creativity. Provide a variety of craft supplies like beads, paints, and fabrics, and let them create their own unique masterpieces to take home as party favors.\n5. Sports Olympics: Host a mini Olympics event with various sports and games. Set up different stations for activities like sack races, relay races, basketball shooting, and obstacle courses. Give out medals or certificates to the participants.\n6. Cooking Party: Have a cooking-themed party where the kids can prepare their own mini pizzas, cupcakes, or cookies. Provide toppings, frosting, and decorating supplies, and let them get hands-on in the kitchen.\n7. Superhero Training Camp: Create a superhero-themed party where the kids can engage in fun training activities. Set up an obstacle course, have them design their own superhero capes or masks, and organize superhero-themed games and challenges.\n8. Outdoor Adventure: Plan an outdoor adventure party at a local park or nature reserve. Arrange activities like hiking, nature scavenger hunts, or a picnic with games. Encourage exploration and appreciation for the outdoors.\nRemember to tailor the activities to the birthday child's interests and preferences. Have a great celebration!\n### Human: Hello! What is your name?\n### Assistant:", sampling params: SamplingParams(n=1, best_of=1, presence_penalty=0.0, frequency_penalty=0.0, repetition_penalty=1.0, temperature=0.7, top_p=1.0, top_k=-1, min_p=0.0, use_beam_search=False, length_penalty=1.0, early_stopping=False, stop=['###', '<|im_end|>'], stop_token_ids=[], ignore_eos=False, max_tokens=32315, logprobs=None, prompt_logprobs=None, skip_special_tokens=True, spaces_between_special_tokens=True), prompt token ids: None

dongxu commented 8 months ago

I have set the --conv-template qwen-7b-chat, but it still get the input like that.any one knows how to solve.

dongxu commented 8 months ago

The model path is qwen/qwen-72b-chat, but it still gets wrong conversation template. Any one knows?

infwinston commented 8 months ago

this is unexpected. can you try to add --debug

python3 -m fastchat.serve.cli --model-path qwen/qwen-72b-chat --debug

and check what exactly is the loaded chat template?

infwinston commented 8 months ago

what commands did you use to launch vllm worker?

dongxu commented 8 months ago

what commands did you use to launch vllm worker? this is my commands. And I will try it in debug model. thanks python -m fastchat.serve.vllm_worker --model-path qwen/qwen-72b-chat --tensor-parallel-size 4 --model-name qwen --trust-remote-code --dtype float16

dongxu commented 8 months ago

--debug

vllm worker can't run as debug? 2024-01-09 13:14:18 | ERROR | stderr | vllm_worker.py: error: unrecognized arguments: --debug

dongxu commented 8 months ago

Any one know that Problem?

dongxu commented 8 months ago

I see covn_template is "qwen-7b-chat" in debug model, but why in my input has weired message: prompt: "A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.\n### Human: Got any creative ideas for a 10 year old’s birthday?\n### Assistant: Of course! Here are some creative ideas for a 10-year-old's birthday party:\n1. Treasure Hunt: Organize a treasure hunt in your backyard or nearby park. Create clues and riddles for the kids to solve, leading them to hidden treasures and surprises.\n2. Science Party: Plan a science-themed party where kids can engage in fun and interactive experiments. You can set up different stations with activities like making slime, erupting volcanoes, or creating simple chemical reactions.\n3. Outdoor Movie Night: Set up a backyard movie night with a projector and a large screen or white sheet. Create a cozy seating area with blankets and pillows, and serve popcorn and snacks while the kids enjoy a favorite movie under the stars.\n4. DIY Crafts Party: Arrange a craft party where kids can unleash their creativity. Provide a variety of craft supplies like beads, paints, and fabrics, and let them create their own unique masterpieces to take home as party favors.\n5. Sports Olympics: Host a mini Olympics event with various sports and games. Set up different stations for activities like sack races, relay races, basketball shooting, and obstacle courses. Give out medals or certificates to the participants.\n6. Cooking Party: Have a cooking-themed party where the kids can prepare their own mini pizzas, cupcakes, or cookies. Provide toppings, frosting, and decorating supplies, and let them get hands-on in the kitchen.\n7. Superhero Training Camp: Create a superhero-themed party where the kids can engage in fun training activities. Set up an obstacle course, have them design their own superhero capes or masks, and organize superhero-themed games and challenges.\n8. Outdoor Adventure: Plan an outdoor adventure party at a local park or nature reserve. Arrange activities like hiking, nature scavenger hunts, or a picnic with games. Encourage exploration and appreciation for the outdoors.\nRemember to tailor the activities to the birthday child's interests and preferences. Have a great celebration!\n### Human: Hello! What is your name?\n### Assistant:", sampling params: SamplingParams(n=1, best_of=1, presence_penalty=0.0, frequency_penalty=0.0, repetition_penalty=1.0, temperature=0.7, top_p=1.0, top_k=-1, min_p=0.0, use_beam_search=False, length_penalty=1.0, early_stopping=False, stop=['<|im_end|>', '###'], stop_token_ids=[], ignore_eos=False, max_tokens=32315, logprobs=None, prompt_logprobs=None, skip_special_tokens=True, spaces_between_special_tokens=True), prompt token ids: None.

infwinston commented 8 months ago

I can't reproduce your issue. what's your fastchat version? does qwen-7b-chat work?

gjafachini commented 4 months ago

Apparently is the default one-shot chat template: https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py Need to change the conversation template