metadata
license: gemma
datasets:
- changpt/ko-lima-vicuna
- GAIR/lima
language:
- en
- ko
from transformers import pipeline
messages = [
{"role": "user", "content": "Who are you?"},
]
pipe = pipeline("text-generation", model="heegyu/gemma-2-9b-lima", device_map="auto", torch_dtype="auto")
print(pipe(messages, max_new_tokens=128, eos_token_id=107))
output: I am an AI assistant, how can I help you today?