mirror of
https://github.com/RYDE-WORK/Langchain-Chatchat.git
synced 2026-01-19 21:37:20 +08:00
6cb1bdf623
* 添加LLM模型切换功能，需要在server_config中设置可切换的模型 * add tests for api.py/llm_model/* * - 支持模型切换 - 支持智普AI线上模型 - startup.py增加参数`--api-worker`，自动运行所有的线上API模型。使用`-a (--all-webui), --all-api`时默认开启该选项 * 修复被fastchat覆盖的标准输出 * 对fastchat日志进行更细致的控制，startup.py中增加-q(--quiet)开关，可以减少无用的fastchat日志输出 * 修正chatglm api的对话模板 Co-authored-by: liunux4odoo <liunu@qq.com>
72 lines
2.0 KiB
Python
72 lines
2.0 KiB
Python
from configs.model_config import LOG_PATH
|
||
import fastchat.constants
|
||
fastchat.constants.LOGDIR = LOG_PATH
|
||
from fastchat.serve.model_worker import BaseModelWorker
|
||
import uuid
|
||
import json
|
||
import sys
|
||
from pydantic import BaseModel
|
||
import fastchat
|
||
import threading
|
||
from typing import Dict, List
|
||
|
||
|
||
# 恢复被fastchat覆盖的标准输出
|
||
sys.stdout = sys.__stdout__
|
||
sys.stderr = sys.__stderr__
|
||
|
||
|
||
class ApiModelOutMsg(BaseModel):
|
||
error_code: int = 0
|
||
text: str
|
||
|
||
class ApiModelWorker(BaseModelWorker):
|
||
BASE_URL: str
|
||
SUPPORT_MODELS: List
|
||
|
||
def __init__(
|
||
self,
|
||
model_names: List[str],
|
||
controller_addr: str,
|
||
worker_addr: str,
|
||
context_len: int = 2048,
|
||
**kwargs,
|
||
):
|
||
kwargs.setdefault("worker_id", uuid.uuid4().hex[:8])
|
||
kwargs.setdefault("model_path", "")
|
||
kwargs.setdefault("limit_worker_concurrency", 5)
|
||
super().__init__(model_names=model_names,
|
||
controller_addr=controller_addr,
|
||
worker_addr=worker_addr,
|
||
**kwargs)
|
||
self.context_len = context_len
|
||
self.init_heart_beat()
|
||
|
||
def count_token(self, params):
|
||
# TODO：需要完善
|
||
print("count token")
|
||
print(params)
|
||
prompt = params["prompt"]
|
||
return {"count": len(str(prompt)), "error_code": 0}
|
||
|
||
def generate_stream_gate(self, params):
|
||
self.call_ct += 1
|
||
|
||
def generate_gate(self, params):
|
||
for x in self.generate_stream_gate(params):
|
||
pass
|
||
return json.loads(x[:-1].decode())
|
||
|
||
def get_embeddings(self, params):
|
||
print("embedding")
|
||
print(params)
|
||
|
||
# workaround to make program exit with Ctrl+c
|
||
# it should be deleted after pr is merged by fastchat
|
||
def init_heart_beat(self):
|
||
self.register_to_controller()
|
||
self.heart_beat_thread = threading.Thread(
|
||
target=fastchat.serve.model_worker.heart_beat_worker, args=(self,), daemon=True,
|
||
)
|
||
self.heart_beat_thread.start()
|