This is an automated email from the ASF dual-hosted git repository.
jin pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/incubator-hugegraph-ai.git
The following commit(s) were added to refs/heads/main by this push:
new 900a465 fix(llm): replace getenv usage to settings (#133)
900a465 is described below
commit 900a465dae9c24f0987c2a081fa3f6c002712835
Author: chenzihong <[email protected]>
AuthorDate: Mon Dec 16 17:01:58 2024 +0800
fix(llm): replace getenv usage to settings (#133)
---
hugegraph-llm/src/hugegraph_llm/api/admin_api.py | 4 +++-
hugegraph-llm/src/hugegraph_llm/demo/rag_demo/admin_block.py | 4 ++--
hugegraph-llm/src/hugegraph_llm/demo/rag_demo/app.py | 9 ++++-----
hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py | 3 +--
hugegraph-llm/src/hugegraph_llm/demo/rag_demo/rag_block.py | 4 ++--
hugegraph-llm/src/hugegraph_llm/models/embeddings/qianfan.py | 7 ++++---
hugegraph-llm/src/hugegraph_llm/models/llms/qianfan.py | 6 +++---
7 files changed, 19 insertions(+), 18 deletions(-)
diff --git a/hugegraph-llm/src/hugegraph_llm/api/admin_api.py
b/hugegraph-llm/src/hugegraph_llm/api/admin_api.py
index 21fbf5f..a234c50 100644
--- a/hugegraph-llm/src/hugegraph_llm/api/admin_api.py
+++ b/hugegraph-llm/src/hugegraph_llm/api/admin_api.py
@@ -22,12 +22,14 @@ from fastapi.responses import StreamingResponse
from hugegraph_llm.api.exceptions.rag_exceptions import generate_response
from hugegraph_llm.api.models.rag_requests import LogStreamRequest
from hugegraph_llm.api.models.rag_response import RAGResponse
+from hugegraph_llm.config import admin_settings
+
# FIXME: line 31: E0702: Raising dict while only classes or instances are
allowed (raising-bad-type)
def admin_http_api(router: APIRouter, log_stream):
@router.post("/logs", status_code=status.HTTP_200_OK)
async def log_stream_api(req: LogStreamRequest):
- if os.getenv('ADMIN_TOKEN') != req.admin_token:
+ if admin_settings.admin_token != req.admin_token:
raise
generate_response(RAGResponse(status_code=status.HTTP_403_FORBIDDEN,
message="Invalid admin_token")) #pylint: disable=E0702
log_path = os.path.join("logs", req.log_file)
diff --git a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/admin_block.py
b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/admin_block.py
index 02b90dd..b8c1852 100644
--- a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/admin_block.py
+++ b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/admin_block.py
@@ -16,12 +16,12 @@
# under the License.
import asyncio
-import os
from collections import deque
import gradio as gr
from gradio import Request
+from hugegraph_llm.config import admin_settings
from hugegraph_llm.utils.log import log
@@ -72,7 +72,7 @@ def clear_llm_server_log():
# Function to validate password and control access to logs
def check_password(password, request: Request = None):
client_ip = request.client.host if request else "Unknown IP"
- admin_token = os.getenv('ADMIN_TOKEN')
+ admin_token = admin_settings.admin_token
if password == admin_token:
# Return logs and update visibility
diff --git a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/app.py
b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/app.py
index a247fb2..db6e2f2 100644
--- a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/app.py
+++ b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/app.py
@@ -17,7 +17,6 @@
import argparse
-import os
import gradio as gr
import uvicorn
@@ -26,7 +25,7 @@ from fastapi.security import HTTPBearer,
HTTPAuthorizationCredentials
from hugegraph_llm.api.admin_api import admin_http_api
from hugegraph_llm.api.rag_api import rag_http_api
-from hugegraph_llm.config import huge_settings, prompt
+from hugegraph_llm.config import admin_settings, huge_settings, prompt
from hugegraph_llm.demo.rag_demo.admin_block import create_admin_block,
log_stream
from hugegraph_llm.demo.rag_demo.configs_block import (
create_configs_block,
@@ -46,7 +45,7 @@ sec = HTTPBearer()
def authenticate(credentials: HTTPAuthorizationCredentials = Depends(sec)):
- correct_token = os.getenv("USER_TOKEN")
+ correct_token = admin_settings.user_token
if credentials.credentials != correct_token:
from fastapi import HTTPException
@@ -148,7 +147,7 @@ if __name__ == "__main__":
# settings.check_env()
prompt.update_yaml_file()
- auth_enabled = os.getenv("ENABLE_LOGIN", "False").lower() == "true"
+ auth_enabled = admin_settings.enable_login.lower() == "true"
log.info("(Status) Authentication is %s now.", "enabled" if auth_enabled
else "disabled")
api_auth = APIRouter(dependencies=[Depends(authenticate)] if auth_enabled
else [])
@@ -162,7 +161,7 @@ if __name__ == "__main__":
# TODO: support multi-user login when need
app = gr.mount_gradio_app(app, hugegraph_llm, path="/",
- auth=("rag", os.getenv("USER_TOKEN")) if
auth_enabled else None)
+ auth=("rag", admin_settings.user_token) if
auth_enabled else None)
# TODO: we can't use reload now due to the config 'app' of uvicorn.run
# ❎:f'{__name__}:app' / rag_web_demo:app /
hugegraph_llm.demo.rag_web_demo:app
diff --git a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
index b7ace5f..b45e589 100644
--- a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
+++ b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
@@ -16,7 +16,6 @@
# under the License.
import json
-import os
from functools import partial
from typing import Optional
@@ -376,7 +375,7 @@ def create_configs_block() -> list:
with gr.Accordion("4. Set up the Reranker.", open=False):
reranker_dropdown = gr.Dropdown(
choices=["cohere", "siliconflow", ("default/offline", "None")],
- value=os.getenv("reranker_type") or "None",
+ value=llm_settings.reranker_type or "None",
label="Reranker",
)
diff --git a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/rag_block.py
b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/rag_block.py
index 4b7449f..3edaaf5 100644
--- a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/rag_block.py
+++ b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/rag_block.py
@@ -24,7 +24,7 @@ import gradio as gr
import pandas as pd
from gradio.utils import NamedString
-from hugegraph_llm.config import resource_path, prompt, huge_settings
+from hugegraph_llm.config import resource_path, prompt, huge_settings,
llm_settings
from hugegraph_llm.operators.graph_rag_task import RAGPipeline
from hugegraph_llm.utils.log import log
@@ -128,7 +128,7 @@ def create_rag_block():
with gr.Column():
with gr.Row():
- online_rerank = os.getenv("reranker_type")
+ online_rerank = llm_settings.reranker_type
rerank_method = gr.Dropdown(
choices=["bleu", ("rerank (online)", "reranker")] if
online_rerank else ["bleu"],
value="reranker" if online_rerank else "bleu",
diff --git a/hugegraph-llm/src/hugegraph_llm/models/embeddings/qianfan.py
b/hugegraph-llm/src/hugegraph_llm/models/embeddings/qianfan.py
index 2f41fe5..1745eb2 100644
--- a/hugegraph-llm/src/hugegraph_llm/models/embeddings/qianfan.py
+++ b/hugegraph-llm/src/hugegraph_llm/models/embeddings/qianfan.py
@@ -18,9 +18,10 @@
from typing import Optional, List
-import os
import qianfan
+from hugegraph_llm.config import llm_settings
+
"""
"QianFan" platform can be understood as a unified LLM platform that
encompasses the
WenXin large model along with other
@@ -37,8 +38,8 @@ class QianFanEmbedding:
api_key: Optional[str] = None,
secret_key: Optional[str] = None
):
- qianfan.get_config().AK = api_key or os.getenv("QIANFAN_ACCESS_KEY")
- qianfan.get_config().SK = secret_key or os.getenv("QIANFAN_SECRET_KEY")
+ qianfan.get_config().AK = api_key or
llm_settings.qianfan_embedding_api_key
+ qianfan.get_config().SK = secret_key or
llm_settings.qianfan_embedding_secret_key
self.embedding_model_name = model_name
self.client = qianfan.Embedding()
diff --git a/hugegraph-llm/src/hugegraph_llm/models/llms/qianfan.py
b/hugegraph-llm/src/hugegraph_llm/models/llms/qianfan.py
index eeaf938..967c391 100644
--- a/hugegraph-llm/src/hugegraph_llm/models/llms/qianfan.py
+++ b/hugegraph-llm/src/hugegraph_llm/models/llms/qianfan.py
@@ -15,13 +15,13 @@
# specific language governing permissions and limitations
# under the License.
-import os
import json
from typing import Optional, List, Dict, Any, Callable
import qianfan
from retry import retry
+from hugegraph_llm.config import llm_settings
from hugegraph_llm.models.llms.base import BaseLLM
from hugegraph_llm.utils.log import log
@@ -29,8 +29,8 @@ from hugegraph_llm.utils.log import log
class QianfanClient(BaseLLM):
def __init__(self, model_name: Optional[str] = "ERNIE-4.0-Turbo-8K",
api_key: Optional[str] = None, secret_key: Optional[str] =
None):
- qianfan.get_config().AK = api_key or os.getenv("QIANFAN_ACCESS_KEY")
- qianfan.get_config().SK = secret_key or os.getenv("QIANFAN_SECRET_KEY")
+ qianfan.get_config().AK = api_key or llm_settings.qianfan_chat_api_key
+ qianfan.get_config().SK = secret_key or
llm_settings.qianfan_chat_secret_key
self.chat_model = model_name
self.chat_comp = qianfan.ChatCompletion()
