wehub-resource-sync
58 lines
2.3 KiB
Python
58 lines
2.3 KiB
Python
# Copyright (c) ModelScope Contributors. All rights reserved.
|
|
import os
|
|
from typing import List
|
|
|
|
os.environ['CUDA_VISIBLE_DEVICES'] = '0'
|
|
|
|
|
|
def infer_batch(engine: 'InferEngine', infer_requests: List['InferRequest']):
|
|
request_config = RequestConfig(max_tokens=512, temperature=0)
|
|
metric = InferStats()
|
|
resp_list = engine.infer(infer_requests, request_config, metrics=[metric])
|
|
query0 = infer_requests[0].messages[0]['content']
|
|
print(f'query0: {query0}')
|
|
print(f'response0: {resp_list[0].choices[0].message.content}')
|
|
print(f'metric: {metric.compute()}')
|
|
# metric.reset() # reuse
|
|
|
|
|
|
def infer_stream(engine: 'InferEngine', infer_request: 'InferRequest'):
|
|
request_config = RequestConfig(max_tokens=512, temperature=0, stream=True)
|
|
metric = InferStats()
|
|
gen_list = engine.infer([infer_request], request_config, metrics=[metric])
|
|
query = infer_request.messages[0]['content']
|
|
print(f'query: {query}\nresponse: ', end='')
|
|
for resp in gen_list[0]:
|
|
if resp is None:
|
|
continue
|
|
print(resp.choices[0].delta.content, end='', flush=True)
|
|
print()
|
|
print(f'metric: {metric.compute()}')
|
|
|
|
|
|
if __name__ == '__main__':
|
|
from swift import InferEngine, InferRequest, InferStats, RequestConfig, TransformersEngine, load_dataset
|
|
model = 'Qwen/Qwen2.5-1.5B-Instruct'
|
|
infer_backend = 'transformers'
|
|
|
|
if infer_backend == 'transformers':
|
|
engine = TransformersEngine(model, max_batch_size=64)
|
|
elif infer_backend == 'vllm':
|
|
from swift.infer_engine import VllmEngine
|
|
engine = VllmEngine(model, max_model_len=8192)
|
|
elif infer_backend == 'sglang':
|
|
from swift.infer_engine import SglangEngine
|
|
engine = SglangEngine(model)
|
|
elif infer_backend == 'lmdeploy':
|
|
from swift.infer_engine import LmdeployEngine
|
|
engine = LmdeployEngine(model)
|
|
|
|
# Here, `load_dataset` is used for convenience; `infer_batch` does not require creating a dataset.
|
|
dataset = load_dataset(['AI-ModelScope/alpaca-gpt4-data-zh#1000'], seed=42)[0]
|
|
print(f'dataset: {dataset}')
|
|
infer_requests = [InferRequest(**data) for data in dataset]
|
|
infer_batch(engine, infer_requests)
|
|
|
|
messages = [{'role': 'user', 'content': 'who are you?'}]
|
|
infer_stream(engine, InferRequest(messages=messages))
|