Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,27 +1,30 @@
|
|
1 |
-
# app.py
|
2 |
import gradio as gr
|
3 |
-
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
|
4 |
import pandas as pd
|
|
|
|
|
5 |
|
6 |
-
# Load
|
7 |
df = pd.read_csv('synthetic_profit.csv')
|
8 |
|
9 |
-
#
|
10 |
-
MODEL_ID = "microsoft/tapex-
|
|
|
|
|
|
|
11 |
|
12 |
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
|
13 |
model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_ID)
|
14 |
|
15 |
-
# Build a table-QA pipeline
|
16 |
table_qa = pipeline(
|
17 |
"table-question-answering",
|
18 |
model=model,
|
19 |
tokenizer=tokenizer,
|
20 |
framework="pt",
|
21 |
-
device
|
22 |
)
|
23 |
|
24 |
-
|
|
|
25 |
table = df.to_dict(orient="records")
|
26 |
try:
|
27 |
out = table_qa(table=table, query=question)
|
@@ -29,16 +32,13 @@ def answer_profitability(question):
|
|
29 |
except Exception as e:
|
30 |
return f"Error: {e}"
|
31 |
|
32 |
-
# Gradio
|
33 |
iface = gr.Interface(
|
34 |
fn=answer_profitability,
|
35 |
inputs=gr.Textbox(lines=2, placeholder="Ask a question about profitability…"),
|
36 |
outputs="text",
|
37 |
-
title="SAP Profitability Q&A (TAPEX-
|
38 |
-
description=""
|
39 |
-
Ask free-form questions on the synthetic profitability dataset.
|
40 |
-
Powered end-to-end by microsoft/tapex-small-finetuned-wikisql.
|
41 |
-
"""
|
42 |
)
|
43 |
|
44 |
if __name__ == "__main__":
|
|
|
|
|
1 |
import gradio as gr
|
|
|
2 |
import pandas as pd
|
3 |
+
import torch
|
4 |
+
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
|
5 |
|
6 |
+
# 1) Load data
|
7 |
df = pd.read_csv('synthetic_profit.csv')
|
8 |
|
9 |
+
# 2) Use the publicly available TAPEX base WikiSQL model
|
10 |
+
MODEL_ID = "microsoft/tapex-base-finetuned-wikisql"
|
11 |
+
|
12 |
+
# 3) Ensure backend is available
|
13 |
+
device = 0 if torch.cuda.is_available() else -1
|
14 |
|
15 |
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
|
16 |
model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_ID)
|
17 |
|
|
|
18 |
table_qa = pipeline(
|
19 |
"table-question-answering",
|
20 |
model=model,
|
21 |
tokenizer=tokenizer,
|
22 |
framework="pt",
|
23 |
+
device=device,
|
24 |
)
|
25 |
|
26 |
+
# 4) QA function
|
27 |
+
def answer_profitability(question: str) -> str:
|
28 |
table = df.to_dict(orient="records")
|
29 |
try:
|
30 |
out = table_qa(table=table, query=question)
|
|
|
32 |
except Exception as e:
|
33 |
return f"Error: {e}"
|
34 |
|
35 |
+
# 5) Gradio UI
|
36 |
iface = gr.Interface(
|
37 |
fn=answer_profitability,
|
38 |
inputs=gr.Textbox(lines=2, placeholder="Ask a question about profitability…"),
|
39 |
outputs="text",
|
40 |
+
title="SAP Profitability Q&A (TAPEX-Base)",
|
41 |
+
description="Free-form questions on synthetic profitability data using microsoft/tapex-base-finetuned-wikisql."
|
|
|
|
|
|
|
42 |
)
|
43 |
|
44 |
if __name__ == "__main__":
|