storytold/FineTrainers-Conditioning
2
0
Fork 0
mirror of https://github.com/storytold/FineTrainers-Conditioning.git synced 2026-10-23 08:15:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

make style

Browse Source
This commit is contained in:
Aryan
2024-12-21 15:19:03 +01:00
parent 0db7bf28aa
commit b045037850
1 changed files with 18 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
finetrainers/hunyuan_video/hunyuan_video_lora.py
+18 -5
View File
@@ -25,10 +25,21 @@ def load_condition_models(
**kwargs, **kwargs,
) -> Dict[str, nn.Module]: ) -> Dict[str, nn.Module]:
tokenizer = AutoTokenizer.from_pretrained(model_id, subfolder="tokenizer", revision=revision, cache_dir=cache_dir) tokenizer = AutoTokenizer.from_pretrained(model_id, subfolder="tokenizer", revision=revision, cache_dir=cache_dir)
text_encoder = LlamaModel.from_pretrained(model_id, subfolder="text_encoder", torch_dtype=text_encoder_dtype, revision=revision, cache_dir=cache_dir) text_encoder = LlamaModel.from_pretrained(
tokenizer_2 = CLIPTokenizer.from_pretrained(model_id, subfolder="tokenizer_2", revision=revision, cache_dir=cache_dir) model_id, subfolder="text_encoder", torch_dtype=text_encoder_dtype, revision=revision, cache_dir=cache_dir
text_encoder_2 = CLIPTextModel.from_pretrained(model_id, subfolder="text_encoder_2", torch_dtype=text_encoder_2_dtype, revision=revision, cache_dir=cache_dir) )
return {"tokenizer": tokenizer, "text_encoder": text_encoder, "tokenizer_2": tokenizer_2, "text_encoder_2": text_encoder_2} tokenizer_2 = CLIPTokenizer.from_pretrained(
model_id, subfolder="tokenizer_2", revision=revision, cache_dir=cache_dir
)
text_encoder_2 = CLIPTextModel.from_pretrained(
model_id, subfolder="text_encoder_2", torch_dtype=text_encoder_2_dtype, revision=revision, cache_dir=cache_dir
)
return {
"tokenizer": tokenizer,
"text_encoder": text_encoder,
"tokenizer_2": tokenizer_2,
"text_encoder_2": text_encoder_2,
}
def load_latent_models( def load_latent_models(
@@ -38,7 +49,9 @@ def load_latent_models(
cache_dir: Optional[str] = None, cache_dir: Optional[str] = None,
**kwargs, **kwargs,
) -> Dict[str, nn.Module]: ) -> Dict[str, nn.Module]:
vae = AutoencoderKLHunyuanVideo.from_pretrained(model_id, subfolder="vae", torch_dtype=vae_dtype, revision=revision, cache_dir=cache_dir) vae = AutoencoderKLHunyuanVideo.from_pretrained(
model_id, subfolder="vae", torch_dtype=vae_dtype, revision=revision, cache_dir=cache_dir
)
return {"vae": vae} return {"vae": vae}