Spaces:
Running
Running
File size: 3,452 Bytes
5e9cd1d |
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 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 |
import sys
import os
from fastchat.conversation import Conversation
from server.model_workers.base import *
from server.utils import get_httpx_client
from fastchat import conversation as conv
import json
from typing import List, Dict
from configs import logger, log_verbose
class AzureWorker(ApiModelWorker):
def __init__(
self,
*,
controller_addr: str = None,
worker_addr: str = None,
model_names: List[str] = ["azure-api"],
version: str = "gpt-35-turbo",
**kwargs,
):
kwargs.update(model_names=model_names, controller_addr=controller_addr, worker_addr=worker_addr)
super().__init__(**kwargs)
self.version = version
def do_chat(self, params: ApiChatParams) -> Dict:
params.load_config(self.model_names[0])
data = dict(
messages=params.messages,
temperature=params.temperature,
max_tokens=params.max_tokens if params.max_tokens else None,
stream=True,
)
url = ("https://{}.openai.azure.com/openai/deployments/{}/chat/completions?api-version={}"
.format(params.resource_name, params.deployment_name, params.api_version))
headers = {
'Content-Type': 'application/json',
'Accept': 'application/json',
'api-key': params.api_key,
}
text = ""
if log_verbose:
logger.info(f'{self.__class__.__name__}:url: {url}')
logger.info(f'{self.__class__.__name__}:headers: {headers}')
logger.info(f'{self.__class__.__name__}:data: {data}')
with get_httpx_client() as client:
with client.stream("POST", url, headers=headers, json=data) as response:
print(data)
for line in response.iter_lines():
if not line.strip() or "[DONE]" in line:
continue
if line.startswith("data: "):
line = line[6:]
resp = json.loads(line)
if choices := resp["choices"]:
if chunk := choices[0].get("delta", {}).get("content"):
text += chunk
yield {
"error_code": 0,
"text": text
}
print(text)
else:
self.logger.error(f"请求 Azure API 时发生错误：{resp}")
def get_embeddings(self, params):
print("embedding")
print(params)
def make_conv_template(self, conv_template: str = None, model_path: str = None) -> Conversation:
return conv.Conversation(
name=self.model_names[0],
system_message="You are a helpful, respectful and honest assistant.",
messages=[],
roles=["user", "assistant"],
sep="\n### ",
stop_str="###",
)
if __name__ == "__main__":
import uvicorn
from server.utils import MakeFastAPIOffline
from fastchat.serve.base_model_worker import app
worker = AzureWorker(
controller_addr="http://127.0.0.1:20001",
worker_addr="http://127.0.0.1:21008",
)
sys.modules["fastchat.serve.model_worker"].worker = worker
MakeFastAPIOffline(app)
uvicorn.run(app, port=21008) |