src/core/gpt.py

"""Mały GPT od zera (char-level) — architektura z lekcji L07.
Embedding -> [Blok: uwaga (Q/K/V + maska) + MLP/GELU + residual + LayerNorm] x N -> logity.
Czysty PyTorch, bez gotowych warstw transformera.
"""
import math
from dataclasses import dataclass
import torch
import torch.nn as nn
from torch.nn import functional as F


@dataclass
class GPTConfig:
    vocab_size: int = 53
    block_size: int = 256      # ile znaków kontekstu (uwaga sięga tak daleko)
    n_layer: int = 4
    n_head: int = 4
    n_embd: int = 128
    dropout: float = 0.1


class CausalSelfAttention(nn.Module):
    """Uwaga z maską przyczynową (L07): token patrzy tylko wstecz."""
    def __init__(self, cfg: GPTConfig):
        super().__init__()
        assert cfg.n_embd % cfg.n_head == 0
        self.qkv = nn.Linear(cfg.n_embd, 3 * cfg.n_embd)   # Q, K, V naraz
        self.proj = nn.Linear(cfg.n_embd, cfg.n_embd)
        self.attn_drop = nn.Dropout(cfg.dropout)
        self.resid_drop = nn.Dropout(cfg.dropout)
        self.n_head = cfg.n_head
        self.n_embd = cfg.n_embd
        # maska dolnotrójkątna: pozycja t widzi tylko <= t
        self.register_buffer("mask", torch.tril(torch.ones(cfg.block_size, cfg.block_size))
                             .view(1, 1, cfg.block_size, cfg.block_size))

    def forward(self, x):
        B, T, C = x.shape
        q, k, v = self.qkv(x).split(self.n_embd, dim=2)
        # rozbij na głowice: (B, nh, T, hs)
        hs = C // self.n_head
        q = q.view(B, T, self.n_head, hs).transpose(1, 2)
        k = k.view(B, T, self.n_head, hs).transpose(1, 2)
        v = v.view(B, T, self.n_head, hs).transpose(1, 2)
        # scores = Q·Kᵀ / √d  (L07: /√d normalizuje wariancję)
        att = (q @ k.transpose(-2, -1)) * (1.0 / math.sqrt(hs))
        att = att.masked_fill(self.mask[:, :, :T, :T] == 0, float("-inf"))
        att = self.attn_drop(F.softmax(att, dim=-1))
        y = att @ v                                  # ważona suma V
        y = y.transpose(1, 2).contiguous().view(B, T, C)
        return self.resid_drop(self.proj(y))


class MLP(nn.Module):
    """Przetwarzanie w tokenie (L07): 768->4x->768 z GELU."""
    def __init__(self, cfg: GPTConfig):
        super().__init__()
        self.fc = nn.Linear(cfg.n_embd, 4 * cfg.n_embd)
        self.proj = nn.Linear(4 * cfg.n_embd, cfg.n_embd)
        self.drop = nn.Dropout(cfg.dropout)

    def forward(self, x):
        return self.drop(self.proj(F.gelu(self.fc(x))))


class Block(nn.Module):
    """Residual + LayerNorm wokół uwagi i MLP (L07: autostrada x + f(x))."""
    def __init__(self, cfg: GPTConfig):
        super().__init__()
        self.ln1 = nn.LayerNorm(cfg.n_embd)
        self.attn = CausalSelfAttention(cfg)
        self.ln2 = nn.LayerNorm(cfg.n_embd)
        self.mlp = MLP(cfg)

    def forward(self, x):
        x = x + self.attn(self.ln1(x))      # komunikacja MIĘDZY tokenami
        x = x + self.mlp(self.ln2(x))       # przetwarzanie W tokenie
        return x


class GPT(nn.Module):
    def __init__(self, cfg: GPTConfig):
        super().__init__()
        self.cfg = cfg
        self.tok_emb = nn.Embedding(cfg.vocab_size, cfg.n_embd)   # ID -> wektor
        self.pos_emb = nn.Embedding(cfg.block_size, cfg.n_embd)   # pozycja
        self.drop = nn.Dropout(cfg.dropout)
        self.blocks = nn.ModuleList([Block(cfg) for _ in range(cfg.n_layer)])
        self.ln_f = nn.LayerNorm(cfg.n_embd)
        self.head = nn.Linear(cfg.n_embd, cfg.vocab_size, bias=False)
        self.head.weight = self.tok_emb.weight                   # weight tying
        self.apply(self._init)

    def _init(self, m):
        if isinstance(m, (nn.Linear, nn.Embedding)):
            nn.init.normal_(m.weight, mean=0.0, std=0.02)
            if isinstance(m, nn.Linear) and m.bias is not None:
                nn.init.zeros_(m.bias)

    def forward(self, idx, targets=None):
        B, T = idx.shape
        pos = torch.arange(T, device=idx.device)
        x = self.drop(self.tok_emb(idx) + self.pos_emb(pos))
        for blk in self.blocks:
            x = blk(x)
        logits = self.head(self.ln_f(x))
        loss = None
        if targets is not None:
            # cross-entropy (L03): -log p(prawdziwego następnego znaku)
            loss = F.cross_entropy(logits.view(-1, logits.size(-1)), targets.view(-1))
        return logits, loss

    @torch.no_grad()
    def generate(self, idx, max_new_tokens, temperature=0.8, top_k=None):
        for _ in range(max_new_tokens):
            idx_cond = idx[:, -self.cfg.block_size:]      # przytnij do okna
            logits, _ = self(idx_cond)
            logits = logits[:, -1, :] / temperature
            if top_k is not None:
                v, _ = torch.topk(logits, top_k)
                logits[logits < v[:, [-1]]] = float("-inf")
            probs = F.softmax(logits, dim=-1)
            nxt = torch.multinomial(probs, num_samples=1)
            idx = torch.cat((idx, nxt), dim=1)
        return idx

    def num_params(self):
        return sum(p.numel() for p in self.parameters())