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

voc_size = 10000
block_size = 128
n_emb = 384
n_head = 6
n_layer = 4
dropout_rate = 0.3

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

    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**-0.5
        wei = wei.masked_fill(self.tril[:T, :T] == 0, float('-inf'))
        wei = F.softmax(wei, dim=-1)
        wei = self.dropout(wei)
        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_heads * head_size, n_emb)
        self.dropout = nn.Dropout(dropout_rate)

    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, n_emb):
        super().__init__()
        self.net = nn.Sequential(
            nn.Linear(n_emb, 4 * n_emb),
            nn.ReLU(),
            nn.Linear(4 * n_emb, n_emb),
            nn.Dropout(dropout_rate)
        )

    def forward(self, x):
        return self.net(x)
    
class Block(nn.Module):
    def __init__(self, n_emb, n_head):
        super().__init__()
        head_size = n_emb // n_head
        self.sa_head = MultiHeadAttention(n_head, head_size)
        self.ffwd = FeedForward(n_emb)
        self.ln1 = nn.LayerNorm(n_emb)
        self.ln2 = nn.LayerNorm(n_emb)

    def forward(self, x):
        x = x + self.sa_head(self.ln1(x))
        x = x + self.ffwd(self.ln2(x))
        return x
    
class RapformerLangModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.token_embedding_table = nn.Embedding(voc_size, n_emb)
        self.position_embedding_table = nn.Embedding(block_size, n_emb)
        self.blocks = nn.Sequential(*[Block(n_emb, n_head=n_head) for _ in range(n_layer)])
        self.ln_f = nn.LayerNorm(n_emb)
        self.lm_head = nn.Linear(n_emb, voc_size)

    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_f(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=0.8, top_k=50):
        for _ in range(max_new_tokens):
            idx_trunc = idx[:, -block_size:]
            logits, _ = self(idx_trunc)
            logits = logits[:, -1, :]
            logits = logits / temperature
            if top_k is not None:
                v, _ = torch.topk(logits, min(top_k, logits.size(-1)))
                logits[logits < v[:, [-1]]] = -float('Inf')

            probs = F.softmax(logits, dim=-1)
            idx_next = torch.multinomial(probs, num_samples=1)
            idx = torch.cat((idx, idx_next), dim=1)
        return idx