xinyun/ComfyUI-CogVideoXWrapper
1
0
Fork 0
mirror of https://git.datalinker.icu/kijai/ComfyUI-CogVideoXWrapper.git synced 2025-12-10 05:14:22 +08:00
Code Issues Packages Projects Releases Wiki Activity

fun text2vid

Browse Source
This commit is contained in:
kijai 2024-09-19 01:07:18 +03:00
parent 09f53ce0c0
commit a125f61aad
2 changed files with 17 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
cogvideox_fun/pipeline_cogvideox_inpaint.py
View File

@ -674,7 +674,7 @@ class CogVideoX_Fun_Pipeline_Inpaint(DiffusionPipeline):
height,
width,
video_length,
prompt_embeds.dtype,
self.vae.dtype,
device,
generator,
latents,
@ -721,7 +721,7 @@ class CogVideoX_Fun_Pipeline_Inpaint(DiffusionPipeline):
batch_size,
height,
width,
prompt_embeds.dtype,
self.vae.dtype,
device,
generator,
do_classifier_free_guidance,

13
nodes.py
View File

@ -525,6 +525,7 @@ class CogVideoXFunSampler:
"optional":{
"start_img": ("IMAGE",),
"end_img": ("IMAGE",),
"opt_empty_latent": ("LATENT",),
},
}
@ -533,7 +534,8 @@ class CogVideoXFunSampler:
FUNCTION = "process"
CATEGORY = "CogVideoWrapper"
def process(self, pipeline, positive, negative, video_length, base_resolution, seed, steps, cfg, scheduler, start_img=None, end_img=None):
def process(self, pipeline, positive, negative, video_length, base_resolution, seed, steps, cfg, scheduler,
start_img=None, end_img=None, opt_empty_latent=None):
device = mm.get_torch_device()
offload_device = mm.unet_offload_device()
pipe = pipeline["pipe"]
@ -543,13 +545,20 @@ class CogVideoXFunSampler:
mm.soft_empty_cache()
aspect_ratio_sample_size = {key : [x / 512 * base_resolution for x in ASPECT_RATIO_512[key]] for key in ASPECT_RATIO_512.keys()}
if start_img is not None:
start_img = [to_pil(_start_img) for _start_img in start_img] if start_img is not None else None
end_img = [to_pil(_end_img) for _end_img in end_img] if end_img is not None else None
# Count most suitable height and width
aspect_ratio_sample_size = {key : [x / 512 * base_resolution for x in ASPECT_RATIO_512[key]] for key in ASPECT_RATIO_512.keys()}
original_width, original_height = start_img[0].size if type(start_img) is list else Image.open(start_img).size
else:
original_width = opt_empty_latent["samples"][0].shape[-1] * 8
original_height = opt_empty_latent["samples"][0].shape[-2] * 8
closest_size, closest_ratio = get_closest_ratio(original_height, original_width, ratios=aspect_ratio_sample_size)
height, width = [int(x / 16) * 16 for x in closest_size]
print(f"Closest size: {width}:{height}")
base_path = pipeline["base_path"]