Add SeedVR2 support (CORE-6) (#14110)

This commit is contained in:
John Pollock
2026-06-08 05:15:05 -05:00
committed by GitHub
parent 739061dd4c
commit 7863cf0e53
26 changed files with 7383 additions and 40 deletions

View File

@@ -1672,6 +1672,35 @@ class Chroma(supported_models_base.BASE):
t5_detect = comfy.text_encoders.sd3_clip.t5_xxl_detect(state_dict, "{}t5xxl.transformer.".format(pref))
return supported_models_base.ClipTarget(comfy.text_encoders.pixart_t5.PixArtTokenizer, comfy.text_encoders.pixart_t5.pixart_te(**t5_detect))
class SeedVR2(supported_models_base.BASE):
unet_config = {
"image_model": "seedvr2"
}
latent_format = comfy.latent_formats.SeedVR2
vae_key_prefix = ["vae."]
text_encoder_key_prefix = ["text_encoders."]
supported_inference_dtypes = [torch.bfloat16, torch.float16, torch.float32]
sampling_settings = {
"shift": 1.0,
}
def set_inference_dtype(self, dtype, manual_cast_dtype, device=None):
if (
dtype == torch.float16
and manual_cast_dtype is None
and comfy.model_management.should_use_bf16(device)
):
manual_cast_dtype = torch.bfloat16
super().set_inference_dtype(dtype, manual_cast_dtype, device=device)
def get_model(self, state_dict, prefix="", device=None):
out = model_base.SeedVR2(self, device=device)
return out
def clip_target(self, state_dict={}):
return None
class ChromaRadiance(Chroma):
unet_config = {
"image_model": "chroma_radiance",
@@ -2029,7 +2058,6 @@ class LongCatImage(supported_models_base.BASE):
hunyuan_detect = comfy.text_encoders.hunyuan_video.llama_detect(state_dict, "{}qwen25_7b.transformer.".format(pref))
return supported_models_base.ClipTarget(comfy.text_encoders.longcat_image.LongCatImageTokenizer, comfy.text_encoders.longcat_image.te(**hunyuan_detect))
class RT_DETR_v4(supported_models_base.BASE):
unet_config = {
"image_model": "RT_DETR_v4",
@@ -2267,6 +2295,7 @@ models = [
HiDream,
HiDreamO1,
Chroma,
SeedVR2,
ChromaRadiance,
ACEStep,
ACEStep15,