Use Flux RoPE

comfy/ldm/lens/model.py

@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from typing import Any, Dict, List, Optional, Tuple, Union
+from typing import Any, Dict, Optional, Tuple
 
 import torch
 import torch.nn as nn
@@ -10,6 +10,8 @@ import torch.nn.functional as F
 
 import comfy.ldm.flux.layers
 import comfy.patcher_extension
+from comfy.ldm.flux.layers import EmbedND
+from comfy.ldm.flux.math import apply_rope
 from comfy.ldm.modules.attention import optimized_attention
 
 
@@ -17,96 +19,39 @@ def _lens_time_proj(t: torch.Tensor, dim: int = 256) -> torch.Tensor:
     return comfy.ldm.flux.layers.timestep_embedding(t, dim)
 
 
-def apply_rotary_emb_lens(x: torch.Tensor, freqs_cis: torch.Tensor) -> torch.Tensor:
+def _lens_position_ids(
-    x_complex = torch.view_as_complex(x.float().reshape(*x.shape[:-1], -1, 2))
+    frame: int, height: int, width: int, text_seq_len: int,
-    freqs_cis = freqs_cis.unsqueeze(1)
+    scale_rope: bool = True, device=None,
-    x_out = torch.view_as_real(x_complex * freqs_cis).flatten(3)
+) -> torch.Tensor:
-    return x_out.type_as(x)
+    """Lens axial (frame, h, w) position ids for joint image + text sequence.
 
-
+    With ``scale_rope=True`` h/w are centered around 0 (negative + positive
-class LensEmbedRope(nn.Module):
+    halves) and text starts at ``max(h//2, w//2)``. Result shape ``[seq, 3]``;
-    """Frame/H/W axial RoPE shared between image and text streams."""
+    caller adds a batch dim for ``EmbedND``.
-
+    """
-    def __init__(self, theta: int = 10000, axes_dim=(8, 28, 28), scale_rope: bool = True) -> None:
+    if scale_rope:
-        super().__init__()
+        h_pos = torch.cat([torch.arange(-(height - height // 2), 0, device=device),
-        self.theta = theta
+                           torch.arange(0, height // 2, device=device)])
-        self.axes_dim = list(axes_dim)
+        w_pos = torch.cat([torch.arange(-(width - width // 2), 0, device=device),
-        self.scale_rope = scale_rope
+                           torch.arange(0, width // 2, device=device)])
-        pos_index = torch.arange(4096)
+        text_start = max(height // 2, width // 2)
-        neg_index = torch.arange(4096).flip(0) * -1 - 1
-        # Plain attributes (not buffers): register_buffer strips complex imag.
-        self.pos_freqs = torch.cat(
-            [self._rope_params(pos_index, d, theta) for d in self.axes_dim], dim=1
-        )
-        self.neg_freqs = torch.cat(
-            [self._rope_params(neg_index, d, theta) for d in self.axes_dim], dim=1
-        )
-        self.rope_cache: Dict[str, torch.Tensor] = {}
-
-    @staticmethod
-    def _rope_params(index: torch.Tensor, dim: int, theta: int = 10000) -> torch.Tensor:
-        assert dim % 2 == 0
-        freqs = torch.outer(
-            index, 1.0 / torch.pow(theta, torch.arange(0, dim, 2).float().div(dim))
-        )
-        return torch.polar(torch.ones_like(freqs), freqs)
-
-    def forward(
-        self,
-        video_fhw: Union[List[Tuple[int, int, int]], Tuple[int, int, int]],
-        txt_seq_lens: Union[List[int], int],
-        device: torch.device,
-    ) -> Tuple[torch.Tensor, torch.Tensor]:
-        if self.pos_freqs.device != device:
-            self.pos_freqs = self.pos_freqs.to(device)
-            self.neg_freqs = self.neg_freqs.to(device)
-
-        if isinstance(video_fhw, list):
-            video_fhw = video_fhw[0]
-        if not isinstance(video_fhw, list):
-            video_fhw = [video_fhw]
-        if not isinstance(txt_seq_lens, list):
-            txt_seq_lens = [txt_seq_lens]
-        assert len(video_fhw) == 1, "video_fhw must have length 1"
-
-        vid_freqs = []
-        max_vid_index = 0
-        for idx, fhw in enumerate(video_fhw):
-            frame, height, width = fhw
-            rope_key = f"{idx}_{height}_{width}_{device}"
-            video_freq = self.rope_cache.get(rope_key)
-            if video_freq is None:
-                video_freq = self._compute_video_freqs(frame, height, width, idx=0).to(device)
-                self.rope_cache[rope_key] = video_freq
-            if self.scale_rope:
-                max_vid_index = max(height // 2, width // 2, max_vid_index)
     else:
-                max_vid_index = max(height, width, max_vid_index)
+        h_pos = torch.arange(height, device=device)
-            vid_freqs.append(video_freq)
+        w_pos = torch.arange(width, device=device)
+        text_start = max(height, width)
 
-        max_len = max(txt_seq_lens)
+    f_pos = torch.arange(frame, device=device)
-        txt_freqs = self.pos_freqs[max_vid_index : max_vid_index + max_len, ...]
+    img_ids = torch.zeros(frame, height, width, 3, device=device)
-        return torch.cat(vid_freqs, dim=0), txt_freqs
+    img_ids[..., 0] = f_pos[:, None, None]
+    img_ids[..., 1] = h_pos[None, :, None]
+    img_ids[..., 2] = w_pos[None, None, :]
+    img_ids = img_ids.reshape(-1, 3)
 
-    def _compute_video_freqs(self, frame: int, height: int, width: int, idx: int = 0) -> torch.Tensor:
+    # Text positions replicate across all 3 axes (matches original packing).
-        seq_lens = frame * height * width
+    txt_pos = torch.arange(text_start, text_start + text_seq_len, device=device).float()
-        freqs_pos = self.pos_freqs.split([d // 2 for d in self.axes_dim], dim=1)
+    txt_ids = txt_pos[:, None].expand(text_seq_len, 3)
-        freqs_neg = self.neg_freqs.split([d // 2 for d in self.axes_dim], dim=1)
 
-        freqs_frame = freqs_pos[0][idx : idx + frame].view(frame, 1, 1, -1).expand(frame, height, width, -1)
+    return torch.cat([img_ids, txt_ids], dim=0)
-        if self.scale_rope:
-            freqs_height = torch.cat(
-                [freqs_neg[1][-(height - height // 2) :], freqs_pos[1][: height // 2]], dim=0
-            ).view(1, height, 1, -1).expand(frame, height, width, -1)
-            freqs_width = torch.cat(
-                [freqs_neg[2][-(width - width // 2) :], freqs_pos[2][: width // 2]], dim=0
-            ).view(1, 1, width, -1).expand(frame, height, width, -1)
-        else:
-            freqs_height = freqs_pos[1][:height].view(1, height, 1, -1).expand(frame, height, width, -1)
-            freqs_width = freqs_pos[2][:width].view(1, 1, width, -1).expand(frame, height, width, -1)
-
-        freqs = torch.cat([freqs_frame, freqs_height, freqs_width], dim=-1).reshape(seq_lens, -1)
-        return freqs.clone().contiguous()
 
 
 class _TimestepEmbedder(nn.Module):
@@ -184,7 +129,7 @@ class LensJointAttention(nn.Module):
         self,
         hidden_states: torch.Tensor,
         encoder_hidden_states: torch.Tensor,
-        image_rotary_emb: Tuple[torch.Tensor, torch.Tensor],
+        freqs_cis: torch.Tensor,
         attention_mask: Optional[torch.Tensor] = None,
         transformer_options: Optional[Dict[str, Any]] = None,
     ) -> Tuple[torch.Tensor, torch.Tensor]:
@@ -207,19 +152,6 @@ class LensJointAttention(nn.Module):
         txt_v = txt_v.contiguous()
         del txt_qkv
 
-        img_freqs, txt_freqs = image_rotary_emb
-        if img_freqs.shape[0] < seq_img:
-            raise ValueError(f"Image RoPE length {img_freqs.shape[0]} < {seq_img}")
-        img_freqs = img_freqs[:seq_img]
-        img_q = apply_rotary_emb_lens(img_q, img_freqs)
-        img_k = apply_rotary_emb_lens(img_k, img_freqs)
-        if seq_txt > 0:
-            if txt_freqs.shape[0] < seq_txt:
-                raise ValueError(f"Text RoPE length {txt_freqs.shape[0]} < {seq_txt}")
-            txt_freqs = txt_freqs[:seq_txt]
-            txt_q = apply_rotary_emb_lens(txt_q, txt_freqs)
-            txt_k = apply_rotary_emb_lens(txt_k, txt_freqs)
-
         # [B, S, H, D] → [B, H, S, D] for attention, dels to avoid VRAM peaks
         q = torch.cat([img_q, txt_q], dim=1).transpose(1, 2)
         del img_q, txt_q
@@ -228,6 +160,8 @@ class LensJointAttention(nn.Module):
         v = torch.cat([img_v, txt_v], dim=1).transpose(1, 2)
         del img_v, txt_v
 
+        q, k = apply_rope(q, k, freqs_cis)
+
         if attention_mask is not None:
             expected = (bsz, 1, 1, seq_img + seq_txt)
             if attention_mask.shape != expected:
@@ -308,7 +242,7 @@ class LensTransformerBlock(nn.Module):
         hidden_states: torch.Tensor,
         encoder_hidden_states: torch.Tensor,
         temb: torch.Tensor,
-        image_rotary_emb: Tuple[torch.Tensor, torch.Tensor],
+        freqs_cis: torch.Tensor,
         attention_mask: Optional[torch.Tensor] = None,
         transformer_options: Optional[Dict[str, Any]] = None,
     ) -> Tuple[torch.Tensor, torch.Tensor]:
@@ -321,7 +255,7 @@ class LensTransformerBlock(nn.Module):
         img_attn, txt_attn = self.attn(
             hidden_states=img_modulated,
             encoder_hidden_states=txt_modulated,
-            image_rotary_emb=image_rotary_emb,
+            freqs_cis=freqs_cis,
             attention_mask=attention_mask,
             transformer_options=transformer_options,
         )
@@ -391,7 +325,7 @@ class LensTransformer2DModel(nn.Module):
         self.selected_layer_index = list(selected_layer_index)
         self.dtype = dtype
 
-        self.pos_embed = LensEmbedRope(theta=10000, axes_dim=list(axes_dims_rope), scale_rope=True)
+        self.pos_embed = EmbedND(dim=attention_head_dim, theta=10000, axes_dim=list(axes_dims_rope))
         self.time_text_embed = LensTimestepProjEmbeddings(
             embedding_dim=self.inner_dim, dtype=dtype, device=device, operations=operations
         )
@@ -502,7 +436,8 @@ class LensTransformer2DModel(nn.Module):
                 encoder_hidden_states = out["txt"]
 
         temb = self.time_text_embed(timestep, hidden_states)
-        image_rotary_emb = self.pos_embed([(1, h, w)], [text_seq_len], device=hidden_states.device)
+        ids = _lens_position_ids(1, h, w, text_seq_len, device=hidden_states.device).unsqueeze(0)
+        freqs_cis = self.pos_embed(ids)
 
         transformer_options["total_blocks"] = len(self.transformer_blocks)
         transformer_options["block_type"] = "double"
@@ -515,7 +450,7 @@ class LensTransformer2DModel(nn.Module):
                         hidden_states=args["img"],
                         encoder_hidden_states=args["txt"],
                         temb=args["vec"],
-                        image_rotary_emb=args["pe"],
+                        freqs_cis=args["pe"],
                         attention_mask=args.get("attn_mask"),
                         transformer_options=args.get("transformer_options"),
                     )
@@ -525,7 +460,7 @@ class LensTransformer2DModel(nn.Module):
                         "img": hidden_states,
                         "txt": encoder_hidden_states,
                         "vec": temb,
-                        "pe": image_rotary_emb,
+                        "pe": freqs_cis,
                         "attn_mask": joint_mask,
                         "transformer_options": transformer_options,
                     },
@@ -538,7 +473,7 @@ class LensTransformer2DModel(nn.Module):
                     hidden_states=hidden_states,
                     encoder_hidden_states=encoder_hidden_states,
                     temb=temb,
-                    image_rotary_emb=image_rotary_emb,
+                    freqs_cis=freqs_cis,
                     attention_mask=joint_mask,
                     transformer_options=transformer_options,
                 )