pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refa: cleanup synchronous functions in chat_model and implement synchronization for conversation and dialog chats (#11779)

Browse Source 
### What problem does this PR solve?

Cleanup synchronous functions in chat_model and implement
synchronization for conversation and dialog chats.

### Type of change

- [x] Refactoring
- [x] Performance Improvement
This commit is contained in:
Yongteng Lei
2025-12-08 09:43:03 +08:00
committed by GitHub
parent 9b8971a9de
commit 51ec708c58
10 changed files with 421 additions and 843 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
api/apps/conversation_app.py
View File

@ -23,7 +23,7 @@ from quart import Response, request
from api.apps import current_user, login_required
from api.db.db_models import APIToken
from api.db.services.conversation_service import ConversationService, structure_answer
from api.db.services.dialog_service import DialogService, ask, chat, gen_mindmap
from api.db.services.dialog_service import DialogService, async_ask, async_chat, gen_mindmap
from api.db.services.llm_service import LLMBundle
from api.db.services.search_service import SearchService
from api.db.services.tenant_llm_service import TenantLLMService
@ -218,10 +218,10 @@ async def completion():
dia.llm_setting = chat_model_config
is_embedded = bool(chat_model_id)
def stream():
async def stream():
nonlocal dia, msg, req, conv
try:
for ans in chat(dia, msg, True, **req):
async for ans in async_chat(dia, msg, True, **req):
ans = structure_answer(conv, ans, message_id, conv.id)
yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
if not is_embedded:
@ -241,7 +241,7 @@ async def completion():
else:
answer = None
for ans in chat(dia, msg, **req):
async for ans in async_chat(dia, msg, **req):
answer = structure_answer(conv, ans, message_id, conv.id)
if not is_embedded:
ConversationService.update_by_id(conv.id, conv.to_dict())
@ -406,10 +406,10 @@ async def ask_about():
if search_app:
search_config = search_app.get("search_config", {})
def stream():
async def stream():
nonlocal req, uid
try:
for ans in ask(req["question"], req["kb_ids"], uid, search_config=search_config):
async for ans in async_ask(req["question"], req["kb_ids"], uid, search_config=search_config):
yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
except Exception as e:
yield "data:" + json.dumps({"code": 500, "message": str(e), "data": {"answer": "**ERROR**: " + str(e), "reference": []}}, ensure_ascii=False) + "\n\n"

19
api/apps/langfuse_app.py
View File

@ -34,8 +34,9 @@ async def set_api_key():
if not all([secret_key, public_key, host]):
return get_error_data_result(message="Missing required fields")
current_user_id = current_user.id
langfuse_keys = dict(
tenant_id=current_user.id,
tenant_id=current_user_id,
secret_key=secret_key,
public_key=public_key,
host=host,
@ -45,23 +46,24 @@ async def set_api_key():
if not langfuse.auth_check():
return get_error_data_result(message="Invalid Langfuse keys")
langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user.id)
langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user_id)
with DB.atomic():
try:
if not langfuse_entry:
TenantLangfuseService.save(**langfuse_keys)
else:
TenantLangfuseService.update_by_tenant(tenant_id=current_user.id, langfuse_keys=langfuse_keys)
TenantLangfuseService.update_by_tenant(tenant_id=current_user_id, langfuse_keys=langfuse_keys)
return get_json_result(data=langfuse_keys)
except Exception as e:
server_error_response(e)
return server_error_response(e)
@manager.route("/api_key", methods=["GET"]) # noqa: F821
@login_required
@validate_request()
def get_api_key():
langfuse_entry = TenantLangfuseService.filter_by_tenant_with_info(tenant_id=current_user.id)
current_user_id = current_user.id
langfuse_entry = TenantLangfuseService.filter_by_tenant_with_info(tenant_id=current_user_id)
if not langfuse_entry:
return get_json_result(message="Have not record any Langfuse keys.")
@ -72,7 +74,7 @@ def get_api_key():
except langfuse.api.core.api_error.ApiError as api_err:
return get_json_result(message=f"Error from Langfuse: {api_err}")
except Exception as e:
server_error_response(e)
return server_error_response(e)
langfuse_entry["project_id"] = langfuse.api.projects.get().dict()["data"][0]["id"]
langfuse_entry["project_name"] = langfuse.api.projects.get().dict()["data"][0]["name"]
@ -84,7 +86,8 @@ def get_api_key():
@login_required
@validate_request()
def delete_api_key():
langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user.id)
current_user_id = current_user.id
langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user_id)
if not langfuse_entry:
return get_json_result(message="Have not record any Langfuse keys.")
@ -93,4 +96,4 @@ def delete_api_key():
TenantLangfuseService.delete_model(langfuse_entry)
return get_json_result(data=True)
except Exception as e:
server_error_response(e)
return server_error_response(e)

4
api/apps/llm_app.py
View File

@ -74,7 +74,7 @@ async def set_api_key():
assert factory in ChatModel, f"Chat model from {factory} is not supported yet."
mdl = ChatModel[factory](req["api_key"], llm.llm_name, base_url=req.get("base_url"), **extra)
try:
m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9, "max_tokens": 50})
m, tc = await mdl.async_chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9, "max_tokens": 50})
if m.find("**ERROR**") >= 0:
raise Exception(m)
chat_passed = True
@ -217,7 +217,7 @@ async def add_llm():
**extra,
)
try:
m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9})
m, tc = await mdl.async_chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9})
if not tc and m.find("**ERROR**:") >= 0:
raise Exception(m)
except Exception as e:

25
api/apps/sdk/session.py
View File

@ -26,9 +26,10 @@ from api.db.db_models import APIToken
from api.db.services.api_service import API4ConversationService
from api.db.services.canvas_service import UserCanvasService, completion_openai
from api.db.services.canvas_service import completion as agent_completion
from api.db.services.conversation_service import ConversationService, iframe_completion
from api.db.services.conversation_service import completion as rag_completion
from api.db.services.dialog_service import DialogService, ask, chat, gen_mindmap, meta_filter
from api.db.services.conversation_service import ConversationService
from api.db.services.conversation_service import async_iframe_completion as iframe_completion
from api.db.services.conversation_service import async_completion as rag_completion
from api.db.services.dialog_service import DialogService, async_ask, async_chat, gen_mindmap, meta_filter
from api.db.services.document_service import DocumentService
from api.db.services.knowledgebase_service import KnowledgebaseService
from api.db.services.llm_service import LLMBundle
@ -141,7 +142,7 @@ async def chat_completion(tenant_id, chat_id):
return resp
else:
answer = None
for ans in rag_completion(tenant_id, chat_id, **req):
async for ans in rag_completion(tenant_id, chat_id, **req):
answer = ans
break
return get_result(data=answer)
@ -245,7 +246,7 @@ async def chat_completion_openai_like(tenant_id, chat_id):
# The value for the usage field on all chunks except for the last one will be null.
# The usage field on the last chunk contains token usage statistics for the entire request.
# The choices field on the last chunk will always be an empty array [].
def streamed_response_generator(chat_id, dia, msg):
async def streamed_response_generator(chat_id, dia, msg):
token_used = 0
answer_cache = ""
reasoning_cache = ""
@ -274,7 +275,7 @@ async def chat_completion_openai_like(tenant_id, chat_id):
}
try:
for ans in chat(dia, msg, True, toolcall_session=toolcall_session, tools=tools, quote=need_reference):
async for ans in async_chat(dia, msg, True, toolcall_session=toolcall_session, tools=tools, quote=need_reference):
last_ans = ans
answer = ans["answer"]
@ -342,7 +343,7 @@ async def chat_completion_openai_like(tenant_id, chat_id):
return resp
else:
answer = None
for ans in chat(dia, msg, False, toolcall_session=toolcall_session, tools=tools, quote=need_reference):
async for ans in async_chat(dia, msg, False, toolcall_session=toolcall_session, tools=tools, quote=need_reference):
# focus answer content only
answer = ans
break
@ -733,10 +734,10 @@ async def ask_about(tenant_id):
return get_error_data_result(f"The dataset {kb_id} doesn't own parsed file")
uid = tenant_id
def stream():
async def stream():
nonlocal req, uid
try:
for ans in ask(req["question"], req["kb_ids"], uid):
async for ans in async_ask(req["question"], req["kb_ids"], uid):
yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
except Exception as e:
yield "data:" + json.dumps(
@ -827,7 +828,7 @@ async def chatbot_completions(dialog_id):
resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
return resp
for answer in iframe_completion(dialog_id, **req):
async for answer in iframe_completion(dialog_id, **req):
return get_result(data=answer)
@ -918,10 +919,10 @@ async def ask_about_embedded():
if search_app := SearchService.get_detail(search_id):
search_config = search_app.get("search_config", {})
def stream():
async def stream():
nonlocal req, uid
try:
for ans in ask(req["question"], req["kb_ids"], uid, search_config=search_config):
async for ans in async_ask(req["question"], req["kb_ids"], uid, search_config=search_config):
yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
except Exception as e:
yield "data:" + json.dumps(