nodes/kijai-ComfyUI-WanVideoWrapper
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Context windows with camera control

Browse Source
This commit is contained in:
kijai
2025-05-01 18:58:44 +03:00
parent 77f78ceded
commit 03edd8cbf2
1 changed files with 7 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
nodes.py
+7 -3
View File
@@ -2823,7 +2823,7 @@ class WanVideoSampler:
#region model pred
def predict_with_cfg(z, cfg_scale, positive_embeds, negative_embeds, timestep, idx, image_cond=None, clip_fea=None,
control_latents=None, vace_data=None, unianim_data=None, audio_proj=None, teacache_state=None):
control_latents=None, vace_data=None, unianim_data=None, audio_proj=None, control_camera_latents=None, teacache_state=None):
z = z.to(dtype)
with torch.autocast(device_type=mm.get_autocast_device(device), dtype=dtype, enabled=("fp8" in model["quantization"])):
@@ -3232,6 +3232,10 @@ class WanVideoSampler:
if control_latents is not None:
partial_control_latents = control_latents[:, c]
partial_control_camera_latents = None
if control_camera_latents is not None:
partial_control_camera_latents = control_camera_latents[:, :, c]
partial_vace_context = None
if vace_data is not None:
partial_vace_context = vace_data[0]["context"][0][:, c]
@@ -3261,7 +3265,7 @@ class WanVideoSampler:
partial_latent_model_input,
cfg[idx], positive,
text_embeds["negative_prompt_embeds"],
timestep, idx, partial_img_emb, clip_fea, partial_control_latents, partial_vace_context, partial_unianim_data,partial_audio_proj,
timestep, idx, partial_img_emb, clip_fea, partial_control_latents, partial_vace_context, partial_unianim_data,partial_audio_proj,partial_control_camera_latents,
current_teacache)
if teacache_args is not None:
@@ -3278,7 +3282,7 @@ class WanVideoSampler:
cfg[idx],
text_embeds["prompt_embeds"],
text_embeds["negative_prompt_embeds"],
timestep, idx, image_cond, clip_fea, control_latents, vace_data, unianim_data, audio_proj,
timestep, idx, image_cond, clip_fea, control_latents, vace_data, unianim_data, audio_proj, control_camera_latents,
teacache_state=self.teacache_state)
if latent_shift_loop: