jandres/GenAIExamples
1
0
Fork 0
Code Issues Pull Requests Actions 5 Packages Projects Releases Wiki Activity

move examples gateway (#992)

Browse Source 
Co-authored-by: root <root@idc708073.jf.intel.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Sihan Chen <39623753+Spycsh@users.noreply.github.com>
This commit is contained in:
lkk
2024-12-06 14:40:25 +08:00
committed by GitHub
parent f5c08d4fbb
commit bde285dfce
17 changed files with 1236 additions and 113 deletions
Download Patch File Download Diff File Expand all files Collapse all files

83
ChatQnA/chatqna.py
View File

@@ -6,7 +6,17 @@ import json
import os
import re
from comps import ChatQnAGateway, MicroService, ServiceOrchestrator, ServiceType
from comps import Gateway, MegaServiceEndpoint, MicroService, ServiceOrchestrator, ServiceType
from comps.cores.proto.api_protocol import (
ChatCompletionRequest,
ChatCompletionResponse,
ChatCompletionResponseChoice,
ChatMessage,
UsageInfo,
)
from comps.cores.proto.docarray import LLMParams, RerankerParms, RetrieverParms
from fastapi import Request
from fastapi.responses import StreamingResponse
from langchain_core.prompts import PromptTemplate
@@ -35,7 +45,6 @@ If you don't know the answer to a question, please don't share false information
return template.format(context=context_str, question=question)
MEGA_SERVICE_HOST_IP = os.getenv("MEGA_SERVICE_HOST_IP", "0.0.0.0")
MEGA_SERVICE_PORT = int(os.getenv("MEGA_SERVICE_PORT", 8888))
GUARDRAIL_SERVICE_HOST_IP = os.getenv("GUARDRAIL_SERVICE_HOST_IP", "0.0.0.0")
GUARDRAIL_SERVICE_PORT = int(os.getenv("GUARDRAIL_SERVICE_PORT", 80))
@@ -178,13 +187,14 @@ def align_generator(self, gen, **kwargs):
yield "data: [DONE]\n\n"
class ChatQnAService:
class ChatQnAService(Gateway):
def __init__(self, host="0.0.0.0", port=8000):
self.host = host
self.port = port
ServiceOrchestrator.align_inputs = align_inputs
ServiceOrchestrator.align_outputs = align_outputs
ServiceOrchestrator.align_generator = align_generator
self.megaservice = ServiceOrchestrator()
def add_remote_service(self):
@@ -228,7 +238,6 @@ class ChatQnAService:
self.megaservice.flow_to(embedding, retriever)
self.megaservice.flow_to(retriever, rerank)
self.megaservice.flow_to(rerank, llm)
self.gateway = ChatQnAGateway(megaservice=self.megaservice, host="0.0.0.0", port=self.port)
def add_remote_service_without_rerank(self):
@@ -261,7 +270,6 @@ class ChatQnAService:
self.megaservice.add(embedding).add(retriever).add(llm)
self.megaservice.flow_to(embedding, retriever)
self.megaservice.flow_to(retriever, llm)
self.gateway = ChatQnAGateway(megaservice=self.megaservice, host="0.0.0.0", port=self.port)
def add_remote_service_with_guardrails(self):
guardrail_in = MicroService(
@@ -319,7 +327,66 @@ class ChatQnAService:
self.megaservice.flow_to(retriever, rerank)
self.megaservice.flow_to(rerank, llm)
# self.megaservice.flow_to(llm, guardrail_out)
self.gateway = ChatQnAGateway(megaservice=self.megaservice, host="0.0.0.0", port=self.port)
async def handle_request(self, request: Request):
data = await request.json()
stream_opt = data.get("stream", True)
chat_request = ChatCompletionRequest.parse_obj(data)
prompt = self._handle_message(chat_request.messages)
parameters = LLMParams(
max_tokens=chat_request.max_tokens if chat_request.max_tokens else 1024,
top_k=chat_request.top_k if chat_request.top_k else 10,
top_p=chat_request.top_p if chat_request.top_p else 0.95,
temperature=chat_request.temperature if chat_request.temperature else 0.01,
frequency_penalty=chat_request.frequency_penalty if chat_request.frequency_penalty else 0.0,
presence_penalty=chat_request.presence_penalty if chat_request.presence_penalty else 0.0,
repetition_penalty=chat_request.repetition_penalty if chat_request.repetition_penalty else 1.03,
streaming=stream_opt,
chat_template=chat_request.chat_template if chat_request.chat_template else None,
)
retriever_parameters = RetrieverParms(
search_type=chat_request.search_type if chat_request.search_type else "similarity",
k=chat_request.k if chat_request.k else 4,
distance_threshold=chat_request.distance_threshold if chat_request.distance_threshold else None,
fetch_k=chat_request.fetch_k if chat_request.fetch_k else 20,
lambda_mult=chat_request.lambda_mult if chat_request.lambda_mult else 0.5,
score_threshold=chat_request.score_threshold if chat_request.score_threshold else 0.2,
)
reranker_parameters = RerankerParms(
top_n=chat_request.top_n if chat_request.top_n else 1,
)
result_dict, runtime_graph = await self.megaservice.schedule(
initial_inputs={"text": prompt},
llm_parameters=parameters,
retriever_parameters=retriever_parameters,
reranker_parameters=reranker_parameters,
)
for node, response in result_dict.items():
if isinstance(response, StreamingResponse):
return response
last_node = runtime_graph.all_leaves()[-1]
response = result_dict[last_node]["text"]
choices = []
usage = UsageInfo()
choices.append(
ChatCompletionResponseChoice(
index=0,
message=ChatMessage(role="assistant", content=response),
finish_reason="stop",
)
)
return ChatCompletionResponse(model="chatqna", choices=choices, usage=usage)
def start(self):
super().__init__(
megaservice=self.megaservice,
host=self.host,
port=self.port,
endpoint=str(MegaServiceEndpoint.CHAT_QNA),
input_datatype=ChatCompletionRequest,
output_datatype=ChatCompletionResponse,
)
if __name__ == "__main__":
@@ -329,10 +396,12 @@ if __name__ == "__main__":
args = parser.parse_args()
chatqna = ChatQnAService(host=MEGA_SERVICE_HOST_IP, port=MEGA_SERVICE_PORT)
chatqna = ChatQnAService(port=MEGA_SERVICE_PORT)
if args.without_rerank:
chatqna.add_remote_service_without_rerank()
elif args.with_guardrails:
chatqna.add_remote_service_with_guardrails()
else:
chatqna.add_remote_service()
chatqna.start()