Spaces:
Running
Running
File size: 1,085 Bytes
354d4c2 f1a0050 84a6c36 55535ff f1a0050 84a6c36 e862792 84a6c36 f1a0050 5d41cba f1a0050 84a6c36 f1a0050 84a6c36 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 |
import spaces
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "O1-OPEN/OpenO1-LLama-8B-v0.1"
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
@spaces.GPU
def api_call(messages):
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=8192
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
return response
def call_gpt(history, prompt):
return api_call(history+[{"role":"user", "content":prompt}])
if __name__ == "__main__":
messages = [{"role":"user", "content":"你是谁？"}]
print(api_call(messages))
breakpoint()
|