import torch
import torch.nn as nn
import torch.nn.functional as F

# Hyperparameters
block_size = 256
num_embed = 512  # Increased embedding size
num_heads = 8
num_layers = 12  # Increased number of layers
dropout = 0.3


class Head(nn.Module):
    def __init__(self, head_size):
        super().__init__()
        self.key = nn.Linear(num_embed, head_size)
        self.query = nn.Linear(num_embed, head_size)
        self.value = nn.Linear(num_embed, head_size)
        self.register_buffer(
            "tril", 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)
        wei = q @ k.transpose(-2, -1) * C**-0.5
        wei = wei.masked_fill(self.tril[:T, :T] == 0, float("-inf"))
        wei = F.softmax(wei, dim=-1)
        wei = self.dropout(wei)
        v = self.value(x)
        out = wei @ v
        return out


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(num_embed, num_embed)
        self.dropout = nn.Dropout(dropout)

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


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

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


class Block(nn.Module):
    def __init__(self, num_embed, num_head):
        super().__init__()
        head_size = num_embed // num_head
        self.sa = MultiHeadAttention(num_head, head_size)
        self.ff = FeedForward(num_embed)
        self.ln1 = nn.LayerNorm(num_embed)
        self.ln2 = nn.LayerNorm(num_embed)

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


class GPT(nn.Module):
    def __init__(self, vocab_size):
        super().__init__()
        self.token_embedding_table = nn.Embedding(vocab_size, num_embed)
        self.position_embedding_table = nn.Embedding(block_size, num_embed)
        self.blocks = nn.Sequential(
            *[Block(num_embed, num_heads) for _ in range(num_layers)]
        )
        self.ln = nn.LayerNorm(num_embed)
        self.lm_head = nn.Linear(num_embed, vocab_size)
        self.apply(self._init_weights)

    def _init_weights(self, module):
        if isinstance(module, nn.Linear):
            torch.nn.init.normal_(module.weight, mean=0.0, std=0.02)
            if module.bias is not None:
                torch.nn.init.zeros_(module.bias)
        elif isinstance(module, nn.Embedding):
            torch.nn.init.normal_(module.weight, mean=0.0, std=0.02)

    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=idx.device)
        )
        x = tok_emb + pos_emb
        x = self.blocks(x)
        x = self.ln(x)
        logits = self.lm_head(x)

        if targets is None:
            loss = None
        else:
            B, T, C = logits.shape
            logits = logits.view(B * T, C)
            targets = targets.view(B * T)
            loss = F.cross_entropy(logits, targets)
        return logits, loss

    def generate(self, idx, max_new_tokens, temperature):
        for _ in range(max_new_tokens):
            idx_cond = 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


def encode(s, string_to_int):
    return [string_to_int.get(c, string_to_int["<unk>"]) for c in s]


def decode(lst, int_to_string):
    return "".join([int_to_string[i] for i in lst])


def load_model(vocab_size, model_path=None):
    model = GPT(vocab_size)
    if model_path:
        try:
            model.load_state_dict(torch.load(model_path))
            print("Model loaded successfully.")
        except FileNotFoundError:
            print("No pre-trained model found. Initialized a new model.")
    return model