app.py

import os
from dotenv import load_dotenv
from transformers import TFBertForSequenceClassification, BertTokenizerFast
import tensorflow as tf

# Load environment variables
load_dotenv()

def load_model(model_name):
    try:
        # Try loading the model as a TensorFlow model
        model = TFBertForSequenceClassification.from_pretrained(model_name, use_auth_token=os.getenv('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'))
    except OSError:
        # If loading fails, assume it's a PyTorch model and use from_pt=True
        model = TFBertForSequenceClassification.from_pretrained(model_name, use_auth_token=os.getenv('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'), from_pt=True)
    return model

def load_tokenizer(model_name):
    tokenizer = BertTokenizerFast.from_pretrained(model_name, use_auth_token=os.getenv('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'))
    return tokenizer

def predict(text, model, tokenizer):
    inputs = tokenizer(text, return_tensors="tf")
    outputs = model(**inputs)
    return outputs

def main():
    model_name = os.getenv('Erfan11/Neuracraft')
    model = load_model(model_name)
    tokenizer = load_tokenizer(model_name)
    # Example usage
    text = "Sample input text"
    result = predict(text, model, tokenizer)
    print(result)

if __name__ == "__main__":
    main()