apextoaster
/
onnx-web
1
0
Fork 0
Code Actions

fix(docs): describe how to install inpainting model

This commit is contained in:
Sean Sube 2023-01-08 21:14:47 -06:00
parent 1e2321f843
commit 2332c44cee
1 changed files with 17 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
README.md
View File

@ -33,6 +33,9 @@ This is still an early project and the instructions are a little rough, but it w
- txt2img mode - txt2img mode
- outputs are saved to file - outputs are saved to file
- show image history - show image history
- img2img mode
- mask painting
- source to mask conversion tools
## Contents ## Contents
@ -249,7 +252,11 @@ Download the conversion script from the `huggingface/diffusers` repository to th
Run the conversion script with your desired model(s): Run the conversion script with your desired model(s):
```shell ```shell
# on linux:
> python convert_stable_diffusion_checkpoint_to_onnx.py --model_path="runwayml/stable-diffusion-v1-5" --output_path="./models/stable-diffusion-onnx-v1-5" > python convert_stable_diffusion_checkpoint_to_onnx.py --model_path="runwayml/stable-diffusion-v1-5" --output_path="./models/stable-diffusion-onnx-v1-5"
# on windows:
> python convert_stable_diffusion_checkpoint_to_onnx.py --model_path="runwayml/stable-diffusion-v1-5" --output_path=".\models\stable-diffusion-onnx-v1-5"
``` ```
This will take a little while to convert each model. Stable diffusion v1.4 is about 6GB, v1.5 is at least 10GB or so. This will take a little while to convert each model. Stable diffusion v1.4 is about 6GB, v1.5 is at least 10GB or so.
@ -257,6 +264,16 @@ This will take a little while to convert each model. Stable diffusion v1.4 is ab
You should verify that all of the steps up to this point have worked correctly by attempting to run the You should verify that all of the steps up to this point have worked correctly by attempting to run the
`api/test-setup.py` script, which is a slight variation on the original txt2img script. `api/test-setup.py` script, which is a slight variation on the original txt2img script.
If you want to use inpainting, you will need a second model trained for that purpose:
```shell
# on linux:
> python convert_stable_diffusion_checkpoint_to_onnx.py --model_path="runwayml/stable-diffusion-inpainting" --output_path="./models/stable-diffusion-inpainting"
# on windows:
> python convert_stable_diffusion_checkpoint_to_onnx.py --model_path="runwayml/stable-diffusion-inpainting" --output_path=".\models\stable-diffusion-inpainting"
```
## Usage ## Usage
### Configuring and running the server ### Configuring and running the server