mirrors/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2025-09-10 11:35:40 +00:00
Code Issues Packages Projects Releases Wiki Activity

Implement Cosmos Image/Video to World (Video) diffusion models.

Browse Source 
Use CosmosImageToVideoLatent to set the input image/video.
This commit is contained in:
comfyanonymous
2025-01-14 05:14:10 -05:00
parent 1f1c7b7b56
commit 3aaabb12d4
6 changed files with 84 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
comfy/supported_models.py
View File

@@ -824,9 +824,10 @@ class HunyuanVideo(supported_models_base.BASE):
hunyuan_detect = comfy.text_encoders.hunyuan_video.llama_detect(state_dict, "{}llama.transformer.".format(pref))
return supported_models_base.ClipTarget(comfy.text_encoders.hunyuan_video.HunyuanVideoTokenizer, comfy.text_encoders.hunyuan_video.hunyuan_video_clip(**hunyuan_detect))
class Cosmos(supported_models_base.BASE):
class CosmosT2V(supported_models_base.BASE):
unet_config = {
"image_model": "cosmos",
"in_channels": 16,
}
sampling_settings = {
@@ -854,7 +855,16 @@ class Cosmos(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.cosmos.CosmosT5Tokenizer, comfy.text_encoders.cosmos.te(**t5_detect))
class CosmosI2V(CosmosT2V):
unet_config = {
"image_model": "cosmos",
"in_channels": 17,
}
models = [Stable_Zero123, SD15_instructpix2pix, SD15, SD20, SD21UnclipL, SD21UnclipH, SDXL_instructpix2pix, SDXLRefiner, SDXL, SSD1B, KOALA_700M, KOALA_1B, Segmind_Vega, SD_X4Upscaler, Stable_Cascade_C, Stable_Cascade_B, SV3D_u, SV3D_p, SD3, StableAudio, AuraFlow, PixArtAlpha, PixArtSigma, HunyuanDiT, HunyuanDiT1, FluxInpaint, Flux, FluxSchnell, GenmoMochi, LTXV, HunyuanVideo, Cosmos]
def get_model(self, state_dict, prefix="", device=None):
out = model_base.CosmosVideo(self, image_to_video=True, device=device)
return out
models = [Stable_Zero123, SD15_instructpix2pix, SD15, SD20, SD21UnclipL, SD21UnclipH, SDXL_instructpix2pix, SDXLRefiner, SDXL, SSD1B, KOALA_700M, KOALA_1B, Segmind_Vega, SD_X4Upscaler, Stable_Cascade_C, Stable_Cascade_B, SV3D_u, SV3D_p, SD3, StableAudio, AuraFlow, PixArtAlpha, PixArtSigma, HunyuanDiT, HunyuanDiT1, FluxInpaint, Flux, FluxSchnell, GenmoMochi, LTXV, HunyuanVideo, CosmosT2V, CosmosI2V]
models += [SVD_img2vid]