src/utils.py

import torch
import tiktoken

def load_model(model_path):
    """Load the trained model from the specified path."""
    from src.inference import GPT
    from src.utils import GPTConfig

    config = GPTConfig()
    model = GPT(config)
    model.load_state_dict(torch.load(model_path))
    model.eval()
    return model

def tokenize_input(text):
    """Tokenize the input text using the GPT-2 tokenizer."""
    enc = tiktoken.get_encoding('gpt2')
    tokens = enc.encode(text)
    return torch.tensor(tokens).unsqueeze(0)  # Add batch dimension

def decode_output(tokens):
    """Decode the generated tokens back to text."""
    enc = tiktoken.get_encoding('gpt2')
    return enc.decode(tokens.tolist())

def generate_text(model, input_text, max_length=30):
    """Generate text using the trained model based on the input text."""
    input_tokens = tokenize_input(input_text)
    generated_tokens = input_tokens

    while generated_tokens.size(1) < max_length:
        with torch.no_grad():
            logits = model(generated_tokens)[0]
            logits = logits[:, -1, :]
            probs = torch.softmax(logits, dim=-1)
            topk_probs, topk_indices = torch.topk(probs, 50, dim=-1)
            ix = torch.multinomial(topk_probs, 1)
            xcol = torch.gather(topk_indices, -1, ix)
            generated_tokens = torch.cat((generated_tokens, xcol), dim=1)

    return decode_output(generated_tokens[0])  # Return the decoded output for the first sequence