Merge: branch main from upstream

This commit is contained in:
ManishMadan2882
2024-09-11 23:12:15 +05:30
32 changed files with 1622 additions and 165 deletions

View File

@@ -24,6 +24,7 @@ conversations_collection = db["conversations"]
sources_collection = db["sources"]
prompts_collection = db["prompts"]
api_key_collection = db["api_keys"]
user_logs_collection = db["user_logs"]
answer = Blueprint("answer", __name__)
gpt_model = ""
@@ -37,7 +38,9 @@ if settings.MODEL_NAME: # in case there is particular model name configured
gpt_model = settings.MODEL_NAME
# load the prompts
current_dir = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
current_dir = os.path.dirname(
os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
)
with open(os.path.join(current_dir, "prompts", "chat_combine_default.txt"), "r") as f:
chat_combine_template = f.read()
@@ -98,9 +101,12 @@ def get_retriever(source_id: str):
return retriever_name
def is_azure_configured():
return settings.OPENAI_API_BASE and settings.OPENAI_API_VERSION and settings.AZURE_DEPLOYMENT_NAME
return (
settings.OPENAI_API_BASE
and settings.OPENAI_API_VERSION
and settings.AZURE_DEPLOYMENT_NAME
)
def save_conversation(conversation_id, question, response, source_log_docs, llm):
@@ -201,6 +207,21 @@ def complete_stream(
data = json.dumps({"type": "id", "id": str(conversation_id)})
yield f"data: {data}\n\n"
retriever_params = retriever.get_params()
user_logs_collection.insert_one(
{
"action": "stream_answer",
"level": "info",
"user": "local",
"api_key": user_api_key,
"question": question,
"response": response_full,
"sources": source_log_docs,
"retriever_params": retriever_params,
"timestamp": datetime.datetime.now(datetime.timezone.utc),
}
)
data = json.dumps({"type": "end"})
yield f"data: {data}\n\n"
except Exception as e:
@@ -258,7 +279,7 @@ def stream():
user_api_key = data["api_key"]
elif "active_docs" in data:
source = {"active_docs" : data["active_docs"]}
source = {"active_docs": data["active_docs"]}
retriever_name = get_retriever(data["active_docs"]) or retriever_name
user_api_key = None
@@ -266,12 +287,13 @@ def stream():
source = {}
user_api_key = None
current_app.logger.info(f"/stream - request_data: {data}, source: {source}",
extra={"data": json.dumps({"request_data": data, "source": source})}
current_app.logger.info(
f"/stream - request_data: {data}, source: {source}",
extra={"data": json.dumps({"request_data": data, "source": source})},
)
prompt = get_prompt(prompt_id)
retriever = RetrieverCreator.create_retriever(
retriever_name,
question=question,
@@ -304,8 +326,9 @@ def stream():
mimetype="text/event-stream",
)
except Exception as e:
current_app.logger.error(f"/stream - error: {str(e)} - traceback: {traceback.format_exc()}",
extra={"error": str(e), "traceback": traceback.format_exc()}
current_app.logger.error(
f"/stream - error: {str(e)} - traceback: {traceback.format_exc()}",
extra={"error": str(e), "traceback": traceback.format_exc()},
)
message = e.args[0]
status_code = 400
@@ -364,7 +387,7 @@ def api_answer():
retriever_name = data_key["retriever"] or retriever_name
user_api_key = data["api_key"]
elif "active_docs" in data:
source = {"active_docs":data["active_docs"]}
source = {"active_docs": data["active_docs"]}
retriever_name = get_retriever(data["active_docs"]) or retriever_name
user_api_key = None
else:
@@ -373,8 +396,9 @@ def api_answer():
prompt = get_prompt(prompt_id)
current_app.logger.info(f"/api/answer - request_data: {data}, source: {source}",
extra={"data": json.dumps({"request_data": data, "source": source})}
current_app.logger.info(
f"/api/answer - request_data: {data}, source: {source}",
extra={"data": json.dumps({"request_data": data, "source": source})},
)
retriever = RetrieverCreator.create_retriever(
@@ -406,13 +430,30 @@ def api_answer():
result = {"answer": response_full, "sources": source_log_docs}
result["conversation_id"] = str(
save_conversation(conversation_id, question, response_full, source_log_docs, llm)
save_conversation(
conversation_id, question, response_full, source_log_docs, llm
)
)
retriever_params = retriever.get_params()
user_logs_collection.insert_one(
{
"action": "api_answer",
"level": "info",
"user": "local",
"api_key": user_api_key,
"question": question,
"response": response_full,
"sources": source_log_docs,
"retriever_params": retriever_params,
"timestamp": datetime.datetime.now(datetime.timezone.utc),
}
)
return result
except Exception as e:
current_app.logger.error(f"/api/answer - error: {str(e)} - traceback: {traceback.format_exc()}",
extra={"error": str(e), "traceback": traceback.format_exc()}
current_app.logger.error(
f"/api/answer - error: {str(e)} - traceback: {traceback.format_exc()}",
extra={"error": str(e), "traceback": traceback.format_exc()},
)
return bad_request(500, str(e))
@@ -431,7 +472,7 @@ def api_search():
source = {"active_docs":data_key["source"]}
user_api_key = data["api_key"]
elif "active_docs" in data:
source = {"active_docs":data["active_docs"]}
source = {"active_docs": data["active_docs"]}
user_api_key = None
else:
source = {}
@@ -445,9 +486,10 @@ def api_search():
token_limit = data["token_limit"]
else:
token_limit = settings.DEFAULT_MAX_HISTORY
current_app.logger.info(f"/api/answer - request_data: {data}, source: {source}",
extra={"data": json.dumps({"request_data": data, "source": source})}
current_app.logger.info(
f"/api/answer - request_data: {data}, source: {source}",
extra={"data": json.dumps({"request_data": data, "source": source})},
)
retriever = RetrieverCreator.create_retriever(
@@ -463,6 +505,20 @@ def api_search():
)
docs = retriever.search()
retriever_params = retriever.get_params()
user_logs_collection.insert_one(
{
"action": "api_search",
"level": "info",
"user": "local",
"api_key": user_api_key,
"question": question,
"sources": docs,
"retriever_params": retriever_params,
"timestamp": datetime.datetime.now(datetime.timezone.utc),
}
)
if data.get("isNoneDoc"):
for doc in docs:
doc["source"] = "None"