JoyImageEditPlus is the multi-image (1-6 reference images) variant of
JoyImageEdit, trained from the same base. Its diffusers transformer shares
byte-identical weight structure with the single-image variant (894 keys, zero
rename) but injects references differently: instead of the single-image
slot-stack (stack refs + noise into a 6D tensor and rotate on the frame dim,
which forces all items to share resolution), each reference is independently
patchified and concatenated on the sequence dim with per-image temporal-offset
3D RoPE, allowing references at different resolutions.
Since the single-image port is not yet upstream, this unifies both variants
onto the Plus-style forward rather than keeping two paths; single-image is now
the ref=1 special case. Verified numerically: at ref=1 with equal resolution
the new path's RoPE is bit-identical to the old slot-stack layout, and the
transformer output matches the diffusers Plus reference (fp32, incl. the
different-resolution case).
ComfyUI runs cond/uncond in one forward with a shared reference configuration,
so the diffusers Plus batched RoPE, padding attention_mask, and dedicated
attention processor are unnecessary here: the unified forward reuses the
existing unbatched _apply_rotary_emb and JoyImageAttention. Confirmed
equivalent to the diffusers batched+mask path for a single sample.
- comfy/ldm/joyimage/model.py: forward takes ref_latents and builds
components=[target, ref0, ...]; per-component patchify + temporal-offset
RoPE; output keeps only the target segment. Old single-grid RoPE removed.
- comfy/model_base.py: JoyImage drops the slot-stack / frame-rotation /
shape-equality path in _apply_model, passing ref_latents straight to the
transformer. Guidance-rescale and the reference_latents requirement are kept.
- comfy/text_encoders/joyimage.py: the image template emits one vision block
per reference (N = image count); N=1 is byte-for-byte the old template.
- comfy_extras/nodes_joyimage.py: add TextEncodeJoyImageEditPlus with optional
image1..image6 inputs, each bucket-resized and VAE-encoded into the
reference_latents list.
Detection, supported_models, and sd.py need no changes: the identical weight
structure routes both variants through image_model="joyimage".
LoadTrainingDataset was the only torch.load call in the codebase without
weights_only=True; comfy/utils.py and comfy/sd1_clip.py already pass it.
Recent PyTorch defaults to weights_only=True, so this is defense-in-depth
for installs pinned to older PyTorch. Verified a typical shard (latents +
standard conditioning) round-trips cleanly under weights_only=True.
JoyImageEdit is an image-edit diffusion transformer from JD (jd-opensource),
Apache 2.0. This adds native ComfyUI support so it loads and runs like other
edit models (load checkpoint -> TextEncode + ReferenceLatent -> KSampler ->
VAEDecode), with no diffusers dependency.
Architecture:
- Transformer (comfy/ldm/joyimage/model.py): dual-stream (img/txt) DiT with a
Conv3d patch embed (patch_size [1,2,2]), Wan-style learnable modulation,
and 3D RoPE (rope_dim_list [16,56,56]). All attention goes through
comfy.ldm.modules.attention.optimized_attention.
- Text encoder (comfy/text_encoders/{qwen3_vl,joyimage}.py): a reusable
Qwen3-VL multimodal stack (vision tower + LM) in qwen3_vl.py, plus a thin
JoyImage-specific layer (prompt templates, drop_idx, tokenizer, te() factory)
in joyimage.py that depends on it. text_dim 4096.
- VAE: reuses the existing Wan 2.1 latent format (AutoencoderKLWan), no new
latent format.
- Edit conditioning: reuses the reference_latents mechanism. Reference and
noise latents are stacked on a new n-slot dimension and rotated at the model
boundary (model_base.JoyImage), so the transformer stays 5D-in/5D-out.
Guidance-rescale is built into the CFG path.
Model wiring:
- model_base.JoyImage uses ModelType.FLOW with sampling_settings
multiplier=1000 (the time embedding is trained on t in [0,1000]) and
shift=1.5; FLOW's linear time_snr_shift matches the diffusers
FlowMatchEuler sigma schedule.
- model_detection sniffs the transformer state-dict (double_blocks.*,
condition_embedder.*, 5D img_in Conv3d) to route image_model="joyimage".
- supported_models.JoyImage and the CLIPLoader "joyimage" type register it.
User-facing node TextEncodeJoyImageEdit (comfy_extras/nodes_joyimage.py)
bucket-resizes the input image to the nearest 1024-base bucket, encodes the
prompt with the image, and emits both the conditioning and the bucketed image
so the same pixels feed VAEEncode and the negative encode (JoyImage requires
noise and reference latents to share spatial dims).
Create Video gets a bit_depth option (8-bit/10-bit); the selected depth is carried by the video and applied when it gets encoded. Save Video and Video Slice now keep the source bit depth instead of always quantizing to 8-bit, so 10-bit videos stay 10-bit. 10-bit uses h264 with the yuv420p10le pixel format,so there's no new codec or container.
Signed-off-by: bigcat88 <bigcat88@icloud.com>
huangfeice
Alexis Rolland
pythongosssss
Terry Jia
comfyanonymous
Yousef R. Gamaleldin
Barish Ozbay
Matt Miller
Jukka Seppänen
Maksim
Alexander Piskun
Kohaku-Blueleaf
Talmaj
Silver
John Pollock
person4268
vidigoat