Alexandra Zapko-Willmes commited on
Commit
f081761
·
verified ·
1 Parent(s): 23cf7a9

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +29 -50
app.py CHANGED
@@ -1,64 +1,43 @@
1
  import gradio as gr
2
  from transformers import pipeline
3
  import pandas as pd
4
- import io
5
 
6
- # Load once
7
- classifier = pipeline("zero-shot-classification", model="facebook/bart-large-mnli")
 
 
 
8
 
9
- response_table = []
 
 
 
10
 
11
- def classify_items(questions_text, labels_text):
12
- questions = [q.strip() for q in questions_text.strip().split("\n") if q.strip()]
13
- labels = [l.strip() for l in labels_text.strip().split(",") if l.strip()]
14
-
15
- if not labels or not questions:
16
- return "Please provide both items and at least two response options.", ""
17
 
18
- global response_table
19
- response_table = []
20
- output_lines = []
21
 
22
- for i, question in enumerate(questions, 1):
23
- result = classifier(question, labels, multi_label=False)
24
- probs = dict(zip(result['labels'], result['scores']))
25
-
26
- output_lines.append(f"{i}. {question}")
27
- for label in labels:
28
- output_lines.append(f"→ {label}: {round(probs.get(label, 0.0), 3)}")
29
- output_lines.append("")
30
-
31
- row = {"Item #": i, "Item": question}
32
- row.update({label: round(probs.get(label, 0.0), 3) for label in labels})
33
- response_table.append(row)
34
-
35
- return "\n".join(output_lines), None
36
-
37
- def download_csv():
38
- global response_table
39
- if not response_table:
40
- return None
41
- df = pd.DataFrame(response_table)
42
- csv_buffer = io.StringIO()
43
- df.to_csv(csv_buffer, index=False)
44
- return csv_buffer.getvalue()
45
-
46
- # Gradio UI
47
  with gr.Blocks() as demo:
48
- gr.Markdown("# 🧠 Zero-Shot Classification for Questionnaire Responses")
49
- gr.Markdown("Paste questionnaire items (one per line), and provide your own response labels (comma-separated).")
50
 
51
  with gr.Row():
52
- with gr.Column():
53
- questions_input = gr.Textbox(label="Questionnaire Items", lines=10, placeholder="e.g.\nI feel in control of my life.\nI enjoy being around others.")
54
- labels_input = gr.Textbox(label="Response Options (comma-separated)", placeholder="Strongly disagree, Disagree, Neutral, Agree, Strongly agree")
55
- submit_btn = gr.Button("Classify Items")
56
- csv_btn = gr.Button("📥 Download CSV")
57
- with gr.Column():
58
- output_box = gr.Textbox(label="Classification Output", lines=20)
59
- file_output = gr.File(label="Download CSV", visible=False)
60
 
61
- submit_btn.click(fn=classify_items, inputs=[questions_input, labels_input], outputs=[output_box, file_output])
62
- csv_btn.click(fn=download_csv, inputs=[], outputs=file_output)
 
63
 
64
  demo.launch()
 
1
  import gradio as gr
2
  from transformers import pipeline
3
  import pandas as pd
 
4
 
5
+ MODEL_MAP = {
6
+ "MoritzLaurer/deberta-v3-large-zeroshot-v2.0": "MoritzLaurer/deberta-v3-large-zeroshot-v2.0",
7
+ "MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7": "MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7",
8
+ "joeddav/xlm-roberta-large-xnli": "joeddav/xlm-roberta-large-xnli"
9
+ }
10
 
11
+ def classify_items(model_name, items_text, labels_text):
12
+ classifier = pipeline("zero-shot-classification", model=MODEL_MAP[model_name])
13
+ items = [item.strip() for item in items_text.split("\n") if item.strip()]
14
+ labels = [label.strip() for label in labels_text.split(",") if label.strip()]
15
 
16
+ results = []
17
+ for item in items:
18
+ out = classifier(item, labels, multi_label=True)
19
+ scores = {label: prob for label, prob in zip(out["labels"], out["scores"])}
20
+ scores["item"] = item
21
+ results.append(scores)
22
 
23
+ df = pd.DataFrame(results).fillna(0)
24
+ return df, gr.File.update(value=df.to_csv(index=False), visible=True)
 
25
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
26
  with gr.Blocks() as demo:
27
+ gr.Markdown("## 🧠 Zero-Shot Questionnaire Classifier")
 
28
 
29
  with gr.Row():
30
+ model_choice = gr.Dropdown(choices=list(MODEL_MAP.keys()), label="Choose a zero-shot model")
31
+
32
+ item_input = gr.Textbox(label="Enter questionnaire items (one per line)", lines=6, placeholder="I enjoy social gatherings.\nI prefer planning over spontaneity.")
33
+ label_input = gr.Textbox(label="Enter response options (comma-separated)", placeholder="Strongly disagree, Disagree, Neutral, Agree, Strongly agree")
34
+
35
+ run_button = gr.Button("Classify")
36
+ output_table = gr.Dataframe(label="Classification Results")
37
+ download_csv = gr.File(label="Download CSV", visible=False)
38
 
39
+ run_button.click(fn=classify_items,
40
+ inputs=[model_choice, item_input, label_input],
41
+ outputs=[output_table, download_csv])
42
 
43
  demo.launch()