Spaces:
Sleeping
Sleeping
Download app.py from TanujInsane/doc-class-fix: direct link, hf CLI and curl.
- Browser
- Download file 3.52 kB
-
https://huggingface.co/spaces/TanujInsane/doc-class-fix/resolve/main/app.py
- Command line
-
hf download hf://spaces/TanujInsane/doc-class-fix/app.py
-
curl -L -o app.py https://huggingface.co/spaces/TanujInsane/doc-class-fix/resolve/main/app.py
3.52 kB
| import gradio as gr | |
| import threading | |
| import os | |
| import pickle | |
| from flask import Flask, request, jsonify | |
| from environment import DocumentClassificationEnv | |
| from baseline_inference import run_task, load_or_train | |
| from agent import TicketAgent | |
| print("Pre-training models...") | |
| models = {} | |
| for d in ["easy", "medium", "hard"]: | |
| models[d] = load_or_train(d) | |
| print(f" {d} model ready!") | |
| print("All models ready!") | |
| agent = TicketAgent() | |
| print("Agent ready!") | |
| flask_app = Flask('app') | |
| def classify(): | |
| data = request.get_json() | |
| text = data.get('text', '') | |
| difficulty = data.get('difficulty', 'easy') | |
| model = models.get(difficulty) | |
| if model is None: | |
| return jsonify({'error': 'Invalid difficulty'}), 400 | |
| prediction = int(model.predict([text])[0]) | |
| return jsonify({'category': prediction}) | |
| def health(): | |
| return jsonify({'status': 'ok'}) | |
| def run_flask(): | |
| flask_app.run(host='0.0.0.0', port=7861, debug=False) | |
| flask_thread = threading.Thread(target=run_flask, daemon=True) | |
| flask_thread.start() | |
| def classify_document(text, difficulty): | |
| if not text.strip(): | |
| return "Please enter some text.", "", "" | |
| model = models.get(difficulty) | |
| if model is None: | |
| return "Invalid difficulty selected.", "", "" | |
| prediction = int(model.predict([text])[0]) | |
| env = DocumentClassificationEnv(task_difficulty=difficulty, seed=42) | |
| category_name = env.categories.get(prediction, str(prediction)) | |
| agent_response = agent.process(text, prediction, category_name) | |
| return f"Category ID: {prediction}", f"Category Name: {category_name}", agent_response | |
| def run_benchmark(difficulty): | |
| score, elapsed = run_task(difficulty) | |
| return f"Score: {score:.4f} | Time: {elapsed:.1f}s" | |
| with gr.Blocks(title="Document Classification") as demo: | |
| gr.Markdown("# Document Classification Environment") | |
| gr.Markdown("Classify documents using trained ML models across different difficulty levels.") | |
| with gr.Tab("Classify Document"): | |
| with gr.Row(): | |
| with gr.Column(): | |
| input_text = gr.Textbox( | |
| label="Document Text", | |
| placeholder="Enter document text here...", | |
| lines=5 | |
| ) | |
| difficulty_dropdown = gr.Dropdown( | |
| choices=["easy", "medium", "hard"], | |
| value="easy", | |
| label="Difficulty Level" | |
| ) | |
| classify_btn = gr.Button("Classify", variant="primary") | |
| with gr.Column(): | |
| output_id = gr.Textbox(label="Category ID") | |
| output_name = gr.Textbox(label="Category Name") | |
| output_agent = gr.Textbox(label="Agent Response", lines=3) | |
| classify_btn.click( | |
| fn=classify_document, | |
| inputs=[input_text, difficulty_dropdown], | |
| outputs=[output_id, output_name, output_agent] | |
| ) | |
| with gr.Tab("Benchmark"): | |
| bench_difficulty = gr.Dropdown( | |
| choices=["easy", "medium", "hard"], | |
| value="easy", | |
| label="Difficulty Level" | |
| ) | |
| bench_btn = gr.Button("Run Benchmark", variant="primary") | |
| bench_output = gr.Textbox(label="Result") | |
| bench_btn.click( | |
| fn=run_benchmark, | |
| inputs=[bench_difficulty], | |
| outputs=[bench_output] | |
| ) | |
| demo.launch(server_name="0.0.0.0", server_port=7860) |