use conversion dest path when applying additional nets
This commit is contained in:
parent
1f6105a8fe
commit
f465120cad
|
@ -248,13 +248,13 @@ def convert_models(ctx: ConversionContext, args, models: Models):
|
|||
|
||||
converted = False
|
||||
if model_format in model_formats_original:
|
||||
converted, _dest = convert_diffusion_original(
|
||||
converted, dest = convert_diffusion_original(
|
||||
ctx,
|
||||
model,
|
||||
source,
|
||||
)
|
||||
else:
|
||||
converted, _dest = convert_diffusion_diffusers(
|
||||
converted, dest = convert_diffusion_diffusers(
|
||||
ctx,
|
||||
model,
|
||||
source,
|
||||
|
@ -272,8 +272,7 @@ def convert_models(ctx: ConversionContext, args, models: Models):
|
|||
if "text_encoder" not in blend_models:
|
||||
blend_models["text_encoder"] = load_model(
|
||||
path.join(
|
||||
ctx.model_path,
|
||||
model,
|
||||
dest,
|
||||
"text_encoder",
|
||||
"model.onnx",
|
||||
)
|
||||
|
@ -283,7 +282,7 @@ def convert_models(ctx: ConversionContext, args, models: Models):
|
|||
blend_models[
|
||||
"tokenizer"
|
||||
] = CLIPTokenizer.from_pretrained(
|
||||
path.join(ctx.model_path, model),
|
||||
dest,
|
||||
subfolder="tokenizer",
|
||||
)
|
||||
|
||||
|
@ -292,7 +291,7 @@ def convert_models(ctx: ConversionContext, args, models: Models):
|
|||
inversion_format = inversion.get("format", None)
|
||||
inversion_source = fetch_model(
|
||||
ctx,
|
||||
f"{name}-inversion-{inversion_name}",
|
||||
inversion_name,
|
||||
inversion_source,
|
||||
dest=inversion_dest,
|
||||
)
|
||||
|
@ -317,8 +316,7 @@ def convert_models(ctx: ConversionContext, args, models: Models):
|
|||
if "text_encoder" not in blend_models:
|
||||
blend_models["text_encoder"] = load_model(
|
||||
path.join(
|
||||
ctx.model_path,
|
||||
model,
|
||||
dest,
|
||||
"text_encoder",
|
||||
"model.onnx",
|
||||
)
|
||||
|
@ -326,9 +324,7 @@ def convert_models(ctx: ConversionContext, args, models: Models):
|
|||
|
||||
if "unet" not in blend_models:
|
||||
blend_models["text_encoder"] = load_model(
|
||||
path.join(
|
||||
ctx.model_path, model, "unet", "model.onnx"
|
||||
)
|
||||
path.join(dest, "unet", "model.onnx")
|
||||
)
|
||||
|
||||
# load models if not loaded yet
|
||||
|
|
|
@ -62,10 +62,7 @@ def blend_loras(
|
|||
model_type: Literal["text_encoder", "unet"],
|
||||
):
|
||||
base_model = base_name if isinstance(base_name, ModelProto) else load(base_name)
|
||||
|
||||
lora_count = len(loras)
|
||||
lora_models = [load_file(name) for name, _weight in loras]
|
||||
lora_weights = lora_weights or (np.ones((lora_count)) / lora_count)
|
||||
|
||||
if model_type == "text_encoder":
|
||||
lora_prefix = "lora_te_"
|
||||
|
|
Loading…
Reference in New Issue