app.py

import gradio as gr
import os
import json
import uuid
import torch
import datetime
import pandas as pd
from transformers import AutoTokenizer, AutoModelForSequenceClassification
from huggingface_hub import HfApi, create_repo, upload_file
from datasets import Dataset

# Configuration
MODEL_NAME = "distilbert-base-uncased-finetuned-sst-2-english"
HF_DATASET_REPO = "M2ai/mgtd-logs"
HF_TOKEN = os.getenv("Mgtd")
DATASET_CREATED = False

# Load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)

# Make directories
os.makedirs("logs", exist_ok=True)

def setup_hf_dataset():
    global DATASET_CREATED
    if not DATASET_CREATED and HF_TOKEN:
        try:
            api = HfApi()
            create_repo(HF_DATASET_REPO, repo_type="dataset", token=HF_TOKEN, exist_ok=True)
            DATASET_CREATED = True
            print(f"Dataset {HF_DATASET_REPO} is ready")
        except Exception as e:
            print(f"Error setting up dataset: {e}")
    elif not HF_TOKEN:
        print("Warning: HF_TOKEN not set. Logs will be saved locally only.")

def infer_and_log(text_input):
    inputs = tokenizer(text_input, return_tensors="pt", truncation=True)
    with torch.no_grad():
        outputs = model(**inputs)
        logits = outputs.logits.tolist()
        predicted = torch.argmax(outputs.logits, dim=-1).item()
        label = model.config.id2label[predicted]

    timestamp = datetime.datetime.now().isoformat()
    submission_id = str(uuid.uuid4())
    log_data = {
        "id": submission_id,
        "timestamp": timestamp,
        "input": text_input,
        "logits": logits
    }

    log_file = f"logs/{timestamp.replace(':', '_')}.json"
    with open(log_file, "w") as f:
        json.dump(log_data, f, indent=2)

    if HF_TOKEN and DATASET_CREATED:
        try:
            api = HfApi()
            api.upload_file(
                path_or_fileobj=log_file,
                path_in_repo=f"logs/{os.path.basename(log_file)}",
                repo_id=HF_DATASET_REPO,
                repo_type="dataset",
                token=HF_TOKEN
            )
            print(f"Uploaded log {submission_id} to {HF_DATASET_REPO}")
        except Exception as e:
            print(f"Error uploading to HF dataset: {e}")

    return label

def clear_fields():
    return "", ""

# Setup the dataset on startup
setup_hf_dataset()

with gr.Blocks() as app:
    gr.Markdown("## AI Text Detector")

    with gr.Row():
        input_box = gr.Textbox(label="Input Text", lines=10, interactive=True)
        output_box = gr.Textbox(label="Output", lines=2, interactive=False)

    with gr.Row():
        submit_btn = gr.Button("Submit")
        clear_btn = gr.Button("Clear")

    submit_btn.click(fn=infer_and_log, inputs=input_box, outputs=output_box)
    clear_btn.click(fn=clear_fields, outputs=[input_box, output_box])

if __name__ == "__main__":
    app.launch()