Update app.py
Browse files
app.py
CHANGED
@@ -1,79 +1,26 @@
|
|
1 |
-
import logging
|
2 |
import gradio as gr
|
3 |
from transformers import pipeline
|
4 |
-
import os
|
5 |
import torch
|
6 |
-
|
7 |
-
from fastapi import FastAPI, Request
|
8 |
-
from fastapi.responses import JSONResponse
|
9 |
-
|
10 |
-
# Cấu hình logging
|
11 |
-
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
|
12 |
|
13 |
-
|
14 |
-
try:
|
15 |
-
login(token=os.getenv("HF_TOKEN"))
|
16 |
-
logging.info("Logged in to Hugging Face Hub successfully")
|
17 |
-
except Exception as e:
|
18 |
-
logging.error(f"Failed to login to Hugging Face Hub: {e}")
|
19 |
-
raise
|
20 |
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
model="nguyenvulebinh/vi-mrc-base",
|
27 |
-
device=0 if torch.cuda.is_available() else -1
|
28 |
-
)
|
29 |
-
logging.info("Model loaded successfully")
|
30 |
-
except Exception as e:
|
31 |
-
logging.error(f"Failed to load model: {e}")
|
32 |
-
raise
|
33 |
|
34 |
-
|
35 |
-
def gradio_answer(question, context):
|
36 |
result = qa_pipeline(question=question, context=context)
|
37 |
return result["answer"]
|
38 |
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
data = await request.json()
|
46 |
-
question = data.get("question")
|
47 |
-
context = data.get("context")
|
48 |
-
logging.info(f"Received request - Question: {question}, Context: {context[:200]}...")
|
49 |
-
if not question or not context:
|
50 |
-
logging.error("Missing question or context")
|
51 |
-
return JSONResponse({"error": "Missing question or context"}, status_code=400)
|
52 |
-
result = qa_pipeline(question=question, context=context)
|
53 |
-
logging.info(f"Response - Answer: {result['answer']}")
|
54 |
-
return JSONResponse({"answer": result["answer"]})
|
55 |
-
except Exception as e:
|
56 |
-
logging.error(f"API error: {e}")
|
57 |
-
return JSONResponse({"error": str(e)}, status_code=500)
|
58 |
-
|
59 |
-
# Tạo Gradio Blocks
|
60 |
-
with gr.Blocks() as demo:
|
61 |
-
gr.Markdown("# AgriBot: Hỏi đáp nông nghiệp")
|
62 |
-
gr.Markdown("Nhập câu hỏi và ngữ cảnh để nhận câu trả lời về nông nghiệp.")
|
63 |
-
|
64 |
-
with gr.Row():
|
65 |
-
question_input = gr.Textbox(label="Câu hỏi", placeholder="Nhập câu hỏi của bạn...")
|
66 |
-
context_input = gr.Textbox(label="Ngữ cảnh", placeholder="Nhập ngữ cảnh liên quan...")
|
67 |
-
output = gr.Textbox(label="Câu trả lời")
|
68 |
-
submit_btn = gr.Button("Gửi")
|
69 |
-
submit_btn.click(fn=gradio_answer, inputs=[question_input, context_input], outputs=output)
|
70 |
|
71 |
-
# Chạy ứng dụng
|
72 |
if __name__ == "__main__":
|
73 |
-
|
74 |
-
demo.launch(
|
75 |
-
server_name="0.0.0.0",
|
76 |
-
server_port=7860,
|
77 |
-
inline=False,
|
78 |
-
app=app
|
79 |
-
)
|
|
|
|
|
1 |
import gradio as gr
|
2 |
from transformers import pipeline
|
|
|
3 |
import torch
|
4 |
+
import logging
|
|
|
|
|
|
|
|
|
|
|
5 |
|
6 |
+
logging.basicConfig(level=logging.INFO)
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
|
8 |
+
qa_pipeline = pipeline(
|
9 |
+
"question-answering",
|
10 |
+
model="nguyenvulebinh/vi-mrc-base",
|
11 |
+
device=0 if torch.cuda.is_available() else -1
|
12 |
+
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
|
14 |
+
def answer_fn(question, context):
|
|
|
15 |
result = qa_pipeline(question=question, context=context)
|
16 |
return result["answer"]
|
17 |
|
18 |
+
iface = gr.Interface(
|
19 |
+
fn=answer_fn,
|
20 |
+
inputs=["text", "text"],
|
21 |
+
outputs="text",
|
22 |
+
title="AgriBot: Hỏi đáp nông nghiệp"
|
23 |
+
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
24 |
|
|
|
25 |
if __name__ == "__main__":
|
26 |
+
iface.launch()
|
|
|
|
|
|
|
|
|
|
|
|