app.py

import torch
import torch.nn as nn
from torch.nn import functional as F
import torch.optim as optim

# Hyperparameters
batch_size = 64
block_size = 512
max_iters = 5000
eval_interval = 100
learning_rate = 1e-3
device = 'cuda' if torch.cuda.is_available() else 'cpu'
eval_iters = 200
n_embd = 192
n_head = 8
n_layer = 8
dropout = 0.2

# Seed for reproducibility
torch.manual_seed(1337)

# Load data
with open('tiny_shakespear.txt', 'r', encoding='utf-8') as f:
    text = f.read()

# Tokenization
chars = sorted(list(set(text)))
vocab_size = len(chars)
stoi = {ch: i for i, ch in enumerate(chars)}
itos = {i: ch for i, ch in enumerate(chars)}
encode = lambda s: [stoi[c] for c in s]
decode = lambda l: ''.join([itos[i] for i in l])

data = torch.tensor(encode(text), dtype=torch.long)
n = int(0.9 * len(data))
train_data, val_data = data[:n], data[n:]


def get_batch(split):
    data_ = train_data if split == 'train' else val_data
    ix = torch.randint(len(data_) - block_size, (batch_size,))
    x = torch.stack([data_[i:i + block_size] for i in ix])
    y = torch.stack([data_[i + 1:i + block_size + 1] for i in ix])
    return x.to(device), y.to(device)


@torch.no_grad()
def estimate_loss():
    out = {}
    model.eval()
    for split in ['train', 'val']:
        losses = torch.zeros(eval_iters)
        for k in range(eval_iters):
            X, Y = get_batch(split)
            _, loss = model(X, Y)
            losses[k] = loss.item()
        out[split] = losses.mean().item()
    model.train()
    return out


class Head(nn.Module):
    def __init__(self, head_size):
        super().__init__()
        self.key = nn.Linear(n_embd, head_size, bias=True)
        self.query = nn.Linear(n_embd, head_size, bias=True)
        self.value = nn.Linear(n_embd, head_size, bias=True)
        self.register_buffer('mask', torch.tril(torch.ones(block_size, block_size)))
        self.dropout = nn.Dropout(dropout)

    def forward(self, x):
        B, T, C = x.shape
        k = self.key(x)
        q = self.query(x)
        v = self.value(x)
        wei = q @ k.transpose(-2, -1) * (C // n_head)**-0.5
        wei = wei.masked_fill(self.mask[:T, :T] == 0, float('-inf'))
        wei = F.softmax(wei, dim=-1)
        wei = self.dropout(wei)
        return wei @ v


class MultiHeadAttention(nn.Module):
    def __init__(self, num_heads, head_size):
        super().__init__()
        self.heads = nn.ModuleList([Head(head_size) for _ in range(num_heads)])
        self.proj = nn.Linear(n_embd, n_embd)
        self.dropout = nn.Dropout(dropout)

    def forward(self, x):
        out = torch.cat([h(x) for h in self.heads], dim=-1)
        return self.dropout(self.proj(out))


class FeedForward(nn.Module):
    def __init__(self, n_embd):
        super().__init__()
        self.net = nn.Sequential(
            nn.Linear(n_embd, 4 * n_embd),
            nn.GELU(),
            nn.Linear(4 * n_embd, n_embd),
            nn.Dropout(dropout)
        )

    def forward(self, x):
        return self.net(x)


class Block(nn.Module):
    def __init__(self, n_embd, n_head):
        super().__init__()
        head_size = n_embd // n_head
        self.sa = MultiHeadAttention(n_head, head_size)
        self.ffwd = FeedForward(n_embd)
        self.ln1 = nn.LayerNorm(n_embd)
        self.ln2 = nn.LayerNorm(n_embd)

    def forward(self, x):
        x = x + self.sa(self.ln1(x))
        return x + self.ffwd(self.ln2(x))


class BigramLanguageModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.token_embedding_table = nn.Embedding(vocab_size, n_embd)
        self.position_embedding_table = nn.Embedding(block_size, n_embd)
        self.blocks = nn.Sequential(*[Block(n_embd, n_head) for _ in range(n_layer)])
        self.ln_f = nn.LayerNorm(n_embd)
        self.lm_head = nn.Linear(n_embd, vocab_size)
        self.apply(self._init_weights)

    def _init_weights(self, m):
        if isinstance(m, nn.Linear):
            nn.init.xavier_uniform_(m.weight)
            if m.bias is not None:
                nn.init.zeros_(m.bias)

    def forward(self, idx, targets=None):
        B, T = idx.shape
        tok_emb = self.token_embedding_table(idx)
        pos_emb = self.position_embedding_table(torch.arange(T, device=device))
        x = self.blocks(tok_emb + pos_emb)
        logits = self.lm_head(self.ln_f(x))
        loss = F.cross_entropy(logits.view(B * T, -1), targets.view(B * T)) if targets is not None else None
        return logits, loss

    def generate(self, idx, max_new_tokens, temperature=1.0):
        for _ in range(max_new_tokens):
            idx_cond = idx if idx.size(1) <= block_size else idx[:, -block_size:]
            logits, _ = self(idx_cond)
            logits = logits[:, -1, :] / temperature
            probs = F.softmax(logits, dim=-1)
            idx_next = torch.multinomial(probs, num_samples=1)
            idx = torch.cat((idx, idx_next), dim=1)
        return idx


import gradio as gr
import torch

# Load the model
model = BigramLanguageModel().to(device)
checkpoint_path = "model_checkpoint_iter2500.pt"
model.load_state_dict(torch.load(checkpoint_path, map_location=device))
model.eval()

def generate_text(prompt, max_tokens=200, temperature=0.8):
    encoded_input = torch.tensor([encode(prompt)], dtype=torch.long).to(device)
    generated_output = model.generate(encoded_input, max_new_tokens=max_tokens, temperature=temperature)
    return decode(generated_output.tolist()[0])

# Set up the Gradio interface
interface = gr.Interface(
    fn=generate_text,
    inputs=[
        gr.Textbox(lines=5, label="Enter your prompt"),
        gr.Slider(50, 1000, step=50, value=200, label="Max tokens"),
        gr.Slider(0.1, 2.0, step=0.1, value=0.8, label="Temperature")
    ],
    outputs=gr.Textbox(lines=10, label="Generated Text"),
    title="Tiny Shakespeare Text Generator",
    description="Enter a prompt and generate Shakespeare-like text using a GPT model trained on the Tiny Shakespeare dataset!"
)

interface.launch()