vm_backup/code/model_fp32.py

"""FP32 baseline of matched parameter count. Standard transformer with RMSNorm + SwiGLU + RoPE-free.
Used only as a reference; this is not binary."""
import math
import torch
import torch.nn as nn
import torch.nn.functional as F


class RMSNorm(nn.Module):
    def __init__(self, d, eps=1e-6):
        super().__init__()
        self.weight = nn.Parameter(torch.ones(d))
        self.eps = eps

    def forward(self, x):
        n = torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
        return x * n * self.weight


class MHA(nn.Module):
    def __init__(self, d, h):
        super().__init__()
        self.d = d
        self.h = h
        self.dh = d // h
        self.qkv = nn.Linear(d, 3 * d, bias=False)
        self.o = nn.Linear(d, d, bias=False)

    def forward(self, x):
        B, T, D = x.shape
        qkv = self.qkv(x).reshape(B, T, 3, self.h, self.dh).permute(2, 0, 3, 1, 4)
        q, k, v = qkv[0], qkv[1], qkv[2]
        y = F.scaled_dot_product_attention(q, k, v, is_causal=True)
        y = y.transpose(1, 2).contiguous().view(B, T, D)
        return self.o(y)


class SwiGLU(nn.Module):
    def __init__(self, d, d_ff):
        super().__init__()
        self.g = nn.Linear(d, d_ff, bias=False)
        self.u = nn.Linear(d, d_ff, bias=False)
        self.d = nn.Linear(d_ff, d, bias=False)

    def forward(self, x):
        return self.d(F.silu(self.g(x)) * self.u(x))


class Block(nn.Module):
    def __init__(self, d, h, d_ff):
        super().__init__()
        self.n1 = RMSNorm(d)
        self.a = MHA(d, h)
        self.n2 = RMSNorm(d)
        self.f = SwiGLU(d, d_ff)

    def forward(self, x):
        x = x + self.a(self.n1(x))
        x = x + self.f(self.n2(x))
        return x


class FP32LM(nn.Module):
    def __init__(self, vocab_size=128, d_model=256, n_layers=8, n_heads=8, d_ff=512, max_seq_len=256):
        super().__init__()
        self.vocab_size = vocab_size
        self.d_model = d_model
        self.max_seq_len = max_seq_len
        self.embed = nn.Embedding(vocab_size, d_model)
        self.pos = nn.Embedding(max_seq_len, d_model)
        self.blocks = nn.ModuleList([Block(d_model, n_heads, d_ff) for _ in range(n_layers)])
        self.norm_f = RMSNorm(d_model)
        self.head = nn.Linear(d_model, vocab_size, bias=False)
        self.head.weight = self.embed.weight  # tie

    def forward(self, idx, targets=None):
        B, T = idx.shape
        pos = torch.arange(T, device=idx.device)
        x = self.embed(idx) + self.pos(pos)
        for b in self.blocks:
            x = b(x)
        x = self.norm_f(x)
        logits = self.head(x)
        loss = None
        if targets is not None:
            loss = F.cross_entropy(logits.view(-1, self.vocab_size), targets.view(-1))
        return logits, loss

    @torch.no_grad()
    def generate(self, idx, max_new_tokens=200, temperature=1.0, top_k=None):
        self.eval()
        for _ in range(max_new_tokens):
            idx_cond = idx[:, -self.max_seq_len:]
            logits, _ = self(idx_cond)
            logits = logits[:, -1, :] / max(temperature, 1e-5)
            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


if __name__ == '__main__':
    m = FP32LM()
    n = sum(p.numel() for p in m.parameters())
    print(f"fp32 params: {n:,} ({n/1e6:.2f}M)")