StableDiffusion/KohyaSS
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix typos

Browse Source
This commit is contained in:
bmaltais 2023-03-01 19:20:05 -05:00
parent 078bd8d768
commit 5498539fda
5 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
library/convert_model_gui.py
View File

@ -217,7 +217,7 @@ def gradio_convert_model_tab():
],
)
target_save_precision_type = gr.Dropdown(
label='Target model precison',
label='Target model precision',
choices=['unspecified', 'fp16', 'bf16', 'float'],
value='unspecified',
)

2
library/extract_lora_gui.py
View File

@ -115,7 +115,7 @@ def gradio_extract_lora_tab():
outputs=save_to,
)
save_precision = gr.Dropdown(
label='Save precison',
label='Save precision',
choices=['fp16', 'bf16', 'float'],
value='float',
interactive=True,

4
library/merge_lora_gui.py
View File

@ -121,13 +121,13 @@ def gradio_merge_lora_tab():
outputs=save_to,
)
precision = gr.Dropdown(
label='Merge precison',
label='Merge precision',
choices=['fp16', 'bf16', 'float'],
value='float',
interactive=True,
)
save_precision = gr.Dropdown(
label='Save precison',
label='Save precision',
choices=['fp16', 'bf16', 'float'],
value='float',
interactive=True,

2
library/resize_lora_gui.py
View File

@ -94,7 +94,7 @@ def gradio_resize_lora_tab():
outputs=save_to,
)
save_precision = gr.Dropdown(
label='Save precison',
label='Save precision',
choices=['fp16', 'bf16', 'float'],
value='fp16',
interactive=True,

2
tools/convert_diffusers20_original_sd.md
View File

@ -24,7 +24,7 @@ python convert_diffusers20_original_sd.py ..\models\sd.ckpt
Specify the .ckpt file and the destination folder as arguments.
Model judgment is not possible, so please use the `--v1` option or the `--v2` option depending on the model.
Also, since `.ckpt` does not contain schduler and tokenizer information, you need to copy them from some existing Diffusers model. Please specify with `--reference_model`. You can specify the HuggingFace id or a local model directory.
Also, since `.ckpt` does not contain scheduler and tokenizer information, you need to copy them from some existing Diffusers model. Please specify with `--reference_model`. You can specify the HuggingFace id or a local model directory.
If you don't have a local model, you can specify "stabilityai/stable-diffusion-2" or "stabilityai/stable-diffusion-2-base" for v2.
For v1.4/1.5, "CompVis/stable-diffusion-v1-4" is fine (v1.4 and v1.5 seem to be the same).