Create layers.py

model/layers.py

@@ -0,0 +1,356 @@
+from __future__ import annotations
+
+import math
+from dataclasses import dataclass
+from torch import Tensor, nn
+import torch
+from einops import rearrange, repeat
+from torch import Tensor
+from torch.nn.utils.rnn import pad_sequence
+
+try:
+    from flash_attn import (
+        flash_attn_varlen_func
+    )
+    FLASHATTN_IS_AVAILABLE = True
+except ImportError:
+    FLASHATTN_IS_AVAILABLE = False
+    flash_attn_varlen_func = None
+
+def attention(q: Tensor, k: Tensor, v: Tensor, pe: Tensor, mask: Tensor | None = None, backend = 'pytorch') -> Tensor:
+    q, k = apply_rope(q, k, pe)
+    if backend == 'pytorch':
+        if mask is not None and mask.dtype == torch.bool:
+            mask = torch.zeros_like(mask).to(q).masked_fill_(mask.logical_not(), -1e20)
+        x = torch.nn.functional.scaled_dot_product_attention(q, k, v, attn_mask=mask)
+        # x = torch.nan_to_num(x, nan=0.0, posinf=1e10, neginf=-1e10)
+        x = rearrange(x, "B H L D -> B L (H D)")
+    elif backend == 'flash_attn':
+        # q: (B, H, L, D)
+        # k: (B, H, S, D) now L = S
+        # v: (B, H, S, D)
+        b, h, lq, d = q.shape
+        _, _, lk, _ = k.shape
+        q = rearrange(q, "B H L D -> B L H D")
+        k = rearrange(k, "B H S D -> B S H D")
+        v = rearrange(v, "B H S D -> B S H D")
+        if mask is None:
+            q_lens = torch.tensor([lq] * b, dtype=torch.int32).to(q.device, non_blocking=True)
+            k_lens = torch.tensor([lk] * b, dtype=torch.int32).to(k.device, non_blocking=True)
+        else:
+            q_lens = torch.sum(mask[:, 0, :, 0], dim=1).int()
+            k_lens = torch.sum(mask[:, 0, 0, :], dim=1).int()
+        q = torch.cat([q_v[:q_l] for q_v, q_l in zip(q, q_lens)])
+        k = torch.cat([k_v[:k_l] for k_v, k_l in zip(k, k_lens)])
+        v = torch.cat([v_v[:v_l] for v_v, v_l in zip(v, k_lens)])
+        cu_seqlens_q = torch.cat([q_lens.new_zeros([1]), q_lens]).cumsum(0, dtype=torch.int32)
+        cu_seqlens_k = torch.cat([k_lens.new_zeros([1]), k_lens]).cumsum(0, dtype=torch.int32)
+        max_seqlen_q = q_lens.max()
+        max_seqlen_k = k_lens.max()
+
+        x = flash_attn_varlen_func(
+            q,
+            k,
+            v,
+            cu_seqlens_q=cu_seqlens_q,
+            cu_seqlens_k=cu_seqlens_k,
+            max_seqlen_q=max_seqlen_q,
+            max_seqlen_k=max_seqlen_k
+        )
+        x_list = [x[cu_seqlens_q[i]:cu_seqlens_q[i+1]] for i in range(b)]
+        x = pad_sequence(tuple(x_list), batch_first=True)
+        x = rearrange(x, "B L H D -> B L (H D)")
+    else:
+        raise NotImplementedError
+    return x
+
+
+def rope(pos: Tensor, dim: int, theta: int) -> Tensor:
+    assert dim % 2 == 0
+    scale = torch.arange(0, dim, 2, dtype=torch.float64, device=pos.device) / dim
+    omega = 1.0 / (theta**scale)
+    out = torch.einsum("...n,d->...nd", pos, omega)
+    out = torch.stack([torch.cos(out), -torch.sin(out), torch.sin(out), torch.cos(out)], dim=-1)
+    out = rearrange(out, "b n d (i j) -> b n d i j", i=2, j=2)
+    return out.float()
+
+
+def apply_rope(xq: Tensor, xk: Tensor, freqs_cis: Tensor) -> tuple[Tensor, Tensor]:
+    xq_ = xq.float().reshape(*xq.shape[:-1], -1, 1, 2)
+    xk_ = xk.float().reshape(*xk.shape[:-1], -1, 1, 2)
+    xq_out = freqs_cis[..., 0] * xq_[..., 0] + freqs_cis[..., 1] * xq_[..., 1]
+    xk_out = freqs_cis[..., 0] * xk_[..., 0] + freqs_cis[..., 1] * xk_[..., 1]
+    return xq_out.reshape(*xq.shape).type_as(xq), xk_out.reshape(*xk.shape).type_as(xk)
+
+class EmbedND(nn.Module):
+    def __init__(self, dim: int, theta: int, axes_dim: list[int]):
+        super().__init__()
+        self.dim = dim
+        self.theta = theta
+        self.axes_dim = axes_dim
+
+    def forward(self, ids: Tensor) -> Tensor:
+        n_axes = ids.shape[-1]
+        emb = torch.cat(
+            [rope(ids[..., i], self.axes_dim[i], self.theta) for i in range(n_axes)],
+            dim=-3,
+        )
+
+        return emb.unsqueeze(1)
+
+
+def timestep_embedding(t: Tensor, dim, max_period=10000, time_factor: float = 1000.0):
+    """
+    Create sinusoidal timestep embeddings.
+    :param t: a 1-D Tensor of N indices, one per batch element.
+                      These may be fractional.
+    :param dim: the dimension of the output.
+    :param max_period: controls the minimum frequency of the embeddings.
+    :return: an (N, D) Tensor of positional embeddings.
+    """
+    t = time_factor * t
+    half = dim // 2
+    freqs = torch.exp(-math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32) / half).to(
+        t.device
+    )
+
+    args = t[:, None].float() * freqs[None]
+    embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
+    if dim % 2:
+        embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
+    if torch.is_floating_point(t):
+        embedding = embedding.to(t)
+    return embedding
+
+
+class MLPEmbedder(nn.Module):
+    def __init__(self, in_dim: int, hidden_dim: int):
+        super().__init__()
+        self.in_layer = nn.Linear(in_dim, hidden_dim, bias=True)
+        self.silu = nn.SiLU()
+        self.out_layer = nn.Linear(hidden_dim, hidden_dim, bias=True)
+
+    def forward(self, x: Tensor) -> Tensor:
+        return self.out_layer(self.silu(self.in_layer(x)))
+
+
+class RMSNorm(torch.nn.Module):
+    def __init__(self, dim: int):
+        super().__init__()
+        self.scale = nn.Parameter(torch.ones(dim))
+
+    def forward(self, x: Tensor):
+        x_dtype = x.dtype
+        x = x.float()
+        rrms = torch.rsqrt(torch.mean(x**2, dim=-1, keepdim=True) + 1e-6)
+        return (x * rrms).to(dtype=x_dtype) * self.scale
+
+
+class QKNorm(torch.nn.Module):
+    def __init__(self, dim: int):
+        super().__init__()
+        self.query_norm = RMSNorm(dim)
+        self.key_norm = RMSNorm(dim)
+
+    def forward(self, q: Tensor, k: Tensor, v: Tensor) -> tuple[Tensor, Tensor]:
+        q = self.query_norm(q)
+        k = self.key_norm(k)
+        return q.to(v), k.to(v)
+
+
+class SelfAttention(nn.Module):
+    def __init__(self, dim: int, num_heads: int = 8, qkv_bias: bool = False):
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.norm = QKNorm(head_dim)
+        self.proj = nn.Linear(dim, dim)
+
+    def forward(self, x: Tensor, pe: Tensor, mask: Tensor | None = None) -> Tensor:
+        qkv = self.qkv(x)
+        q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        q, k = self.norm(q, k, v)
+        x = attention(q, k, v, pe=pe, mask=mask)
+        x = self.proj(x)
+        return x
+
+class CrossAttention(nn.Module):
+    def __init__(self, dim: int, context_dim: int, num_heads: int = 8, qkv_bias: bool = False):
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.q = nn.Linear(dim, dim, bias=qkv_bias)
+        self.kv = nn.Linear(dim, context_dim * 2, bias=qkv_bias)
+        self.norm = QKNorm(head_dim)
+        self.proj = nn.Linear(dim, dim)
+
+    def forward(self, x: Tensor, context: Tensor, pe: Tensor, mask: Tensor | None = None) -> Tensor:
+        qkv = self.qkv(x)
+        q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        q, k = self.norm(q, k, v)
+        x = attention(q, k, v, pe=pe, mask=mask)
+        x = self.proj(x)
+        return x
+
+
+@dataclass
+class ModulationOut:
+    shift: Tensor
+    scale: Tensor
+    gate: Tensor
+
+
+class Modulation(nn.Module):
+    def __init__(self, dim: int, double: bool):
+        super().__init__()
+        self.is_double = double
+        self.multiplier = 6 if double else 3
+        self.lin = nn.Linear(dim, self.multiplier * dim, bias=True)
+
+    def forward(self, vec: Tensor) -> tuple[ModulationOut, ModulationOut | None]:
+        out = self.lin(nn.functional.silu(vec))[:, None, :].chunk(self.multiplier, dim=-1)
+
+        return (
+            ModulationOut(*out[:3]),
+            ModulationOut(*out[3:]) if self.is_double else None,
+        )
+
+
+class DoubleStreamBlock(nn.Module):
+    def __init__(self, hidden_size: int, num_heads: int, mlp_ratio: float, qkv_bias: bool = False, backend = 'pytorch'):
+        super().__init__()
+
+        mlp_hidden_dim = int(hidden_size * mlp_ratio)
+        self.num_heads = num_heads
+        self.hidden_size = hidden_size
+        self.img_mod = Modulation(hidden_size, double=True)
+        self.img_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.img_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
+
+        self.img_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.img_mlp = nn.Sequential(
+            nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
+            nn.GELU(approximate="tanh"),
+            nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
+        )
+
+        self.backend = backend
+
+        self.txt_mod = Modulation(hidden_size, double=True)
+        self.txt_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.txt_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
+
+        self.txt_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.txt_mlp = nn.Sequential(
+            nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
+            nn.GELU(approximate="tanh"),
+            nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
+        )
+
+
+
+
+    def forward(self, x: Tensor, vec: Tensor, pe: Tensor, mask: Tensor = None, txt_length = None):
+        img_mod1, img_mod2 = self.img_mod(vec)
+        txt_mod1, txt_mod2 = self.txt_mod(vec)
+
+        txt, img = x[:, :txt_length], x[:, txt_length:]
+
+        # prepare image for attention
+        img_modulated = self.img_norm1(img)
+        img_modulated = (1 + img_mod1.scale) * img_modulated + img_mod1.shift
+        img_qkv = self.img_attn.qkv(img_modulated)
+        img_q, img_k, img_v = rearrange(img_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        img_q, img_k = self.img_attn.norm(img_q, img_k, img_v)
+        # prepare txt for attention
+        txt_modulated = self.txt_norm1(txt)
+        txt_modulated = (1 + txt_mod1.scale) * txt_modulated + txt_mod1.shift
+        txt_qkv = self.txt_attn.qkv(txt_modulated)
+        txt_q, txt_k, txt_v = rearrange(txt_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        txt_q, txt_k = self.txt_attn.norm(txt_q, txt_k, txt_v)
+
+        # run actual attention
+        q = torch.cat((txt_q, img_q), dim=2)
+        k = torch.cat((txt_k, img_k), dim=2)
+        v = torch.cat((txt_v, img_v), dim=2)
+        if mask is not None:
+            mask = repeat(mask, 'B L S->  B H L S', H=self.num_heads)
+        attn = attention(q, k, v, pe=pe, mask = mask, backend = self.backend)
+        txt_attn, img_attn = attn[:, : txt.shape[1]], attn[:, txt.shape[1] :]
+
+        # calculate the img bloks
+        img = img + img_mod1.gate * self.img_attn.proj(img_attn)
+        img = img + img_mod2.gate * self.img_mlp((1 + img_mod2.scale) * self.img_norm2(img) + img_mod2.shift)
+
+        # calculate the txt bloks
+        txt = txt + txt_mod1.gate * self.txt_attn.proj(txt_attn)
+        txt = txt + txt_mod2.gate * self.txt_mlp((1 + txt_mod2.scale) * self.txt_norm2(txt) + txt_mod2.shift)
+        x = torch.cat((txt, img), 1)
+        return x
+
+
+class SingleStreamBlock(nn.Module):
+    """
+    A DiT block with parallel linear layers as described in
+    https://arxiv.org/abs/2302.05442 and adapted modulation interface.
+    """
+
+    def __init__(
+        self,
+        hidden_size: int,
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        qk_scale: float | None = None,
+        backend='pytorch'
+    ):
+        super().__init__()
+        self.hidden_dim = hidden_size
+        self.num_heads = num_heads
+        head_dim = hidden_size // num_heads
+        self.scale = qk_scale or head_dim**-0.5
+
+        self.mlp_hidden_dim = int(hidden_size * mlp_ratio)
+        # qkv and mlp_in
+        self.linear1 = nn.Linear(hidden_size, hidden_size * 3 + self.mlp_hidden_dim)
+        # proj and mlp_out
+        self.linear2 = nn.Linear(hidden_size + self.mlp_hidden_dim, hidden_size)
+
+        self.norm = QKNorm(head_dim)
+
+        self.hidden_size = hidden_size
+        self.pre_norm = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+
+        self.mlp_act = nn.GELU(approximate="tanh")
+        self.modulation = Modulation(hidden_size, double=False)
+        self.backend = backend
+
+    def forward(self, x: Tensor, vec: Tensor, pe: Tensor, mask: Tensor = None) -> Tensor:
+        mod, _ = self.modulation(vec)
+        x_mod = (1 + mod.scale) * self.pre_norm(x) + mod.shift
+        qkv, mlp = torch.split(self.linear1(x_mod), [3 * self.hidden_size, self.mlp_hidden_dim], dim=-1)
+
+        q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        q, k = self.norm(q, k, v)
+        if mask is not None:
+            mask = repeat(mask, 'B L S->  B H L S', H=self.num_heads)
+        # compute attention
+        attn = attention(q, k, v, pe=pe, mask = mask, backend=self.backend)
+        # compute activation in mlp stream, cat again and run second linear layer
+        output = self.linear2(torch.cat((attn, self.mlp_act(mlp)), 2))
+        return x + mod.gate * output
+
+
+class LastLayer(nn.Module):
+    def __init__(self, hidden_size: int, patch_size: int, out_channels: int):
+        super().__init__()
+        self.norm_final = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.linear = nn.Linear(hidden_size, patch_size * patch_size * out_channels, bias=True)
+        self.adaLN_modulation = nn.Sequential(nn.SiLU(), nn.Linear(hidden_size, 2 * hidden_size, bias=True))
+
+    def forward(self, x: Tensor, vec: Tensor) -> Tensor:
+        shift, scale = self.adaLN_modulation(vec).chunk(2, dim=1)
+        x = (1 + scale[:, None, :]) * self.norm_final(x) + shift[:, None, :]
+        x = self.linear(x)
+        return x