app.py

"""
HuggingFace Spaces App for GPT-2 124M Shakespeare Model
"""
import torch
import torch.nn as nn
from torch.nn import functional as F
import tiktoken
import gradio as gr
import math
from dataclasses import dataclass


class CausalSelfAttention(nn.Module):
    def __init__(self, config):
        super().__init__()
        assert config.n_embd % config.n_head == 0
        self.c_attn = nn.Linear(config.n_embd, 3 * config.n_embd)
        self.c_proj = nn.Linear(config.n_embd, config.n_embd)
        self.c_proj.NANOGPT_SCALE_INIT = 1
        self.n_head = config.n_head
        self.n_embd = config.n_embd
        self.register_buffer("bias", torch.tril(torch.ones(config.block_size, config.block_size)).view(1, 1, config.block_size, config.block_size))

    def forward(self, x):
        B, T, C = x.size()
        qkv = self.c_attn(x)
        q, k, v = qkv.split(self.n_embd, dim=2)
        k = k.view(B, T, self.n_head, C // self.n_head).transpose(1, 2)
        q = q.view(B, T, self.n_head, C // self.n_head).transpose(1, 2)
        v = v.view(B, T, self.n_head, C // self.n_head).transpose(1, 2)

        att = (q @ k.transpose(-2, -1)) * (1.0 / math.sqrt(k.size(-1)))
        att = att.masked_fill(self.bias[:, :, :T, :T] == 0, float('-inf'))
        att = F.softmax(att, dim=-1)
        y = att @ v

        y = y.transpose(1, 2).contiguous().view(B, T, C)
        y = self.c_proj(y)
        return y


class MLP(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.c_fc = nn.Linear(config.n_embd, 4 * config.n_embd)
        self.gelu = nn.GELU(approximate='tanh')
        self.c_proj = nn.Linear(4 * config.n_embd, config.n_embd)
        self.c_proj.NANOGPT_SCALE_INIT = 1

    def forward(self, x):
        x = self.c_fc(x)
        x = self.gelu(x)
        x = self.c_proj(x)
        return x


class Block(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.ln_1 = nn.LayerNorm(config.n_embd)
        self.attn = CausalSelfAttention(config)
        self.ln_2 = nn.LayerNorm(config.n_embd)
        self.mlp = MLP(config)

    def forward(self, x):
        x = x + self.attn(self.ln_1(x))
        x = x + self.mlp(self.ln_2(x))
        return x


@dataclass
class GPTConfig:
    block_size: int = 1024
    vocab_size: int = 50257
    n_layer: int = 12
    n_head: int = 12
    n_embd: int = 768


class GPT(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.config = config

        self.transformer = nn.ModuleDict(dict(
            wte=nn.Embedding(config.vocab_size, config.n_embd),
            wpe=nn.Embedding(config.block_size, config.n_embd),
            h=nn.ModuleList([Block(config) for _ in range(config.n_layer)]),
            ln_f=nn.LayerNorm(config.n_embd),
        ))
        self.lm_head = nn.Linear(config.n_embd, config.vocab_size, bias=False)
        self.transformer.wte.weight = self.lm_head.weight

    def forward(self, idx, targets=None):
        B, T = idx.size()
        assert T <= self.config.block_size, f"Cannot forward sequence of length {T}, block size is only {self.config.block_size}"
        pos = torch.arange(0, T, dtype=torch.long, device=idx.device)
        pos_emb = self.transformer.wpe(pos)
        tok_emb = self.transformer.wte(idx)
        x = tok_emb + pos_emb
        for block in self.transformer.h:
            x = block(x)
        x = self.transformer.ln_f(x)
        logits = self.lm_head(x)
        loss = None
        if targets is not None:
            loss = F.cross_entropy(logits.view(-1, logits.size(-1)), targets.view(-1))
        return logits, loss


# Load model
print("Loading model...")
device = 'cuda' if torch.cuda.is_available() else 'cpu'
config = GPTConfig()
model = GPT(config)

# Try to load model (works both locally and on HuggingFace)
try:
    checkpoint = torch.load('model_checkpoint_final.pt', map_location=device)
    model.load_state_dict(checkpoint['model_state_dict'])
    print("Model loaded from checkpoint")
except FileNotFoundError:
    print("Warning: Model checkpoint not found. Using untrained model.")
    # Model will be randomly initialized - not ideal but won't crash

model.to(device)
model.eval()
print(f"Model ready on {device}")

enc = tiktoken.get_encoding('gpt2')


def generate_text(prompt, max_new_tokens=100, temperature=0.8, top_k=50):
    """Generate text from prompt"""
    try:
        # Encode prompt
        tokens = enc.encode(prompt)
        tokens = torch.tensor(tokens, dtype=torch.long, device=device).unsqueeze(0)
        
        # Generate
        with torch.no_grad():
            for _ in range(max_new_tokens):
                # Forward pass
                logits, _ = model(tokens)
                logits = logits[:, -1, :] / temperature
                
                # Top-k sampling
                topk_probs, topk_indices = torch.topk(F.softmax(logits, dim=-1), top_k, dim=-1)
                ix = torch.multinomial(topk_probs, 1)
                next_token = torch.gather(topk_indices, -1, ix)
                
                # Append to sequence
                tokens = torch.cat([tokens, next_token], dim=1)
                
                # Stop if we hit max length
                if tokens.size(1) >= config.block_size:
                    break
        
        # Decode
        generated_text = enc.decode(tokens[0].tolist())
        return generated_text
    except Exception as e:
        return f"Error: {str(e)}"


# Create Gradio interface
with gr.Blocks(title="GPT-2 124M Shakespeare Model") as demo:
    gr.Markdown("""
    # 🎭 GPT-2 124M Shakespeare Language Model
    
    This is a 124M parameter decoder-only transformer model trained on Shakespeare's complete works.
    
    **Training Results:**
    - Final Loss: 0.095127 (Target: < 0.099999) ✅
    - Model Parameters: 124.44M
    - Training Steps: 1,637
    
    Enter a prompt below to generate Shakespeare-style text!
    """)
    
    with gr.Row():
        with gr.Column():
            prompt_input = gr.Textbox(
                label="Prompt",
                placeholder="Enter your prompt here (e.g., 'First Citizen:', 'ROMEO:', 'To be or not')",
                value="First Citizen:",
                lines=3
            )
            max_tokens = gr.Slider(
                label="Max Tokens",
                minimum=50,
                maximum=200,
                value=100,
                step=10
            )
            temperature = gr.Slider(
                label="Temperature",
                minimum=0.1,
                maximum=2.0,
                value=0.8,
                step=0.1
            )
            top_k = gr.Slider(
                label="Top-K",
                minimum=10,
                maximum=100,
                value=50,
                step=10
            )
            generate_btn = gr.Button("Generate", variant="primary")
        
        with gr.Column():
            output = gr.Textbox(
                label="Generated Text",
                lines=10,
                interactive=False
            )
    
    # Example prompts
    gr.Markdown("### Example Prompts:")
    examples = gr.Examples(
        examples=[
            ["First Citizen:"],
            ["ROMEO:"],
            ["To be or not"],
            ["HAMLET:"],
            ["MACBETH:"],
        ],
        inputs=prompt_input
    )
    
    generate_btn.click(
        fn=generate_text,
        inputs=[prompt_input, max_tokens, temperature, top_k],
        outputs=output
    )
    
    gr.Markdown("""
    ---
    **Note:** The model was trained on Shakespeare text and generates text in that style.
    Generated text may not always be coherent but should follow Shakespearean patterns.
    """)

if __name__ == "__main__":
    demo.launch(share=True)