ivanwu/jarvis-models
1
0
Fork 0
mirror of https://github.com/BoardWare-Genius/jarvis-models.git synced 2025-12-14 17:13:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

refactor: processing of blackbox chroma query and chat

Browse Source
This commit is contained in:
ACBBZ
2024-05-28 07:06:16 +00:00
parent 76971c87f5
commit 347eea2f9c
4 changed files with 44 additions and 49 deletions
Download Patch File Download Diff File Expand all files Collapse all files

69
src/blackbox/chroma_chat.py
View File

@ -31,64 +31,47 @@ class ChromaChat(Blackbox):
@logging_time(logger=logger)
def processing(self, question: str, context: list, settings: dict) -> str:
# chroma_chat settings
# {
# "chroma_embedding_model": "bge-large-zh-v1.5",
# "chroma_host": "10.6.82.192",
# "chroma_port": "8000",
# "chroma_collection_id": "123",
# "chroma_n_results": 3,
# "model_name": "Qwen1.5-14B-Chat",
# "context": [],
# "template": "",
# "temperature": 0.8,
# "top_p": 0.8,
# "n": 1,
# "max_tokens": 1024,
# "frequency_penalty": 0.5,
# "presence_penalty": 0.8,
# "stop": 100,
# "model_url": "http://120.196.116.194:48892/v1/chat/completions",
# "model_key": "YOUR_API_KEY"
# }
if settings is None:
settings = {}
# # chat setting
user_model_name = settings.get("model_name")
user_context = context
user_question = question
user_template = settings.get("template")
user_temperature = settings.get("temperature")
user_top_p = settings.get("top_p")
user_n = settings.get("n")
user_max_tokens = settings.get("max_tokens")
user_stop = settings.get("stop")
user_frequency_penalty = settings.get("frequency_penalty")
user_presence_penalty = settings.get("presence_penalty")
# # chroma_query settings
chroma_embedding_model = settings.get("chroma_embedding_model")
chroma_host = settings.get("chroma_host")
chroma_port = settings.get("chroma_port")
chroma_collection_id = settings.get("chroma_collection_id")
chroma_n_results = settings.get("chroma_n_results")
if context == None:
context = []
if user_context == None:
user_context = []
if user_question is None:
return JSONResponse(content={"error": "question is required"}, status_code=status.HTTP_400_BAD_REQUEST)
chroma_settings_json={
"chroma_embedding_model": chroma_embedding_model,
"chroma_host": chroma_host,
"chroma_port": chroma_port,
"chroma_collection_id": chroma_collection_id,
"chroma_n_results": chroma_n_results
}
# chroma answer
chroma_result = self.chroma_query(user_question, chroma_settings_json)
chroma_result = self.chroma_query(user_question, settings)
# chat prompt
fast_question = f"问题: {user_question}。根据问题，总结以下内容和来源：{chroma_result}"
chat_settings_json = {
"model_name": user_model_name,
"context": user_context,
"template": user_template,
"temperature": user_temperature,
"top_p": user_top_p,
"n": user_n,
"max_tokens": user_max_tokens,
"stop": user_stop,
"frequency_penalty": user_frequency_penalty,
"presence_penalty": user_presence_penalty
}
fast_question = f"问题: {user_question}。- 根据知识库内的检索结果，以清晰简洁的表达方式回答问题。只从检索的内容中选取与问题相关信息。- 不要编造答案，如果答案不在经核实的资料中或无法从经核实的资料中得出，请回答“我无法回答您的问题。”检索内容：{chroma_result}"
# chat answer
response = self.chat(fast_question, chat_settings_json)
response = self.chat(fast_question, user_context, settings)
return response