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

fix text embed casting

Browse Source
This commit is contained in:
kijai
2025-03-02 17:16:29 +02:00
parent 588a7aa455
commit f41f10a4ee
1 changed files with 5 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
nodes.py
+5 -5
View File
@@ -756,10 +756,10 @@ class WanVideoTextEmbedBridge:
DESCRIPTION = "Bridge between ComfyUI native text embedding and WanVideoWrapper text embedding"
def process(self, positive, negative):
device=mm.get_torch_device()
prompt_embeds_dict = {
"prompt_embeds": positive[0][0],
"negative_prompt_embeds": negative[0][0],
"prompt_embeds": positive[0][0].to(device),
"negative_prompt_embeds": negative[0][0].to(device),
}
return (prompt_embeds_dict,)
@@ -1140,9 +1140,9 @@ class WanVideoSampler:
})
arg_c = base_args.copy()
arg_c.update({'context': [text_embeds["prompt_embeds"][0].to(device)]})
arg_c.update({'context': [text_embeds["prompt_embeds"][0]]})
arg_null = base_args.copy()
arg_null.update({'context': text_embeds["negative_prompt_embeds"].to(device)})
arg_null.update({'context': text_embeds["negative_prompt_embeds"]})
pbar = ProgressBar(steps)