1
0
Fork 0

fix(docs): note launch-extras script

This commit is contained in:
Sean Sube 2023-02-13 23:21:56 -06:00
parent 4b1857883f
commit cc7cafabfd
Signed by: ssube
GPG Key ID: 3EED7B957D362AF1
2 changed files with 16 additions and 6 deletions

View File

@ -286,17 +286,21 @@ Log into the HuggingFace CLI:
# on linux: # on linux:
> huggingface-cli login > huggingface-cli login
# on windows: # on Windows:
> huggingface-cli.exe login > huggingface-cli.exe login
``` ```
Issue an API token from https://huggingface.co/settings/tokens, naming it something memorable like `onnx-web`, and then Issue an API token from https://huggingface.co/settings/tokens, naming it something memorable like `onnx-web`, and then
paste it into the prompt. paste it into the prompt.
Run the provided conversion script from the `api/` directory: Running the launch script from the `api/` directory will convert the base models before starting the API server:
```shell ```shell
> python -m onnx_web.convert --diffusion --correction --upscaling # on Linux:
> ./launch.sh
# on Windows:
> launch.bat
``` ```
Models that have already been downloaded and converted will be skipped, so it should be safe to run this script after Models that have already been downloaded and converted will be skipped, so it should be safe to run this script after
@ -336,13 +340,16 @@ launch script:
```shell ```shell
# on Linux: # on Linux:
> export ONNX_WEB_EXTRA_MODELS="/home/ssube/onnx-web-extras.json" > export ONNX_WEB_EXTRA_MODELS="/home/ssube/onnx-web-extras.json"
> ./launch.sh > ./launch-extras.sh
# on Windows: # on Windows:
> set ONNX_WEB_EXTRA_MODELS=C:\Users\ssube\onnx-web-extras.json > set ONNX_WEB_EXTRA_MODELS=C:\Users\ssube\onnx-web-extras.json
> launch.bat > launch-extras.bat
``` ```
Make sure to use the `launch-extras.sh` or `.bat` script if you want to convert the extra models, especially if you
have added your own.
### Test the models ### Test the models
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

View File

@ -207,14 +207,17 @@ Models can be added from the directories used by `diffusers` as well as SafeTens
loading PickleTensors, as they may contain unsafe code which can be executed on your machine, and use SafeTensor instead loading PickleTensors, as they may contain unsafe code which can be executed on your machine, and use SafeTensor instead
whenever possible. whenever possible.
Set the `ONNX_WEB_EXTRA_MODELS` environment variable to the path to your file. For example: Set the `ONNX_WEB_EXTRA_MODELS` environment variable to the path to your file and make sure to use the `launch-extras`
script. For example:
```shell ```shell
# on Linux: # on Linux:
> export ONNX_WEB_EXTRA_MODELS=~/onnx-web-extras.json > export ONNX_WEB_EXTRA_MODELS=~/onnx-web-extras.json
> ./launch-extras.sh
# on Windows: # on Windows:
> set ONNX_WEB_EXTRA_MODELS=C:\Users\ssube\onnx-web-extras.json > set ONNX_WEB_EXTRA_MODELS=C:\Users\ssube\onnx-web-extras.json
> launch-extras.bat
``` ```
Extras using the older file format with nested arrays (`"diffusion": [[]]`) can be mixed with the newer format. You Extras using the older file format with nested arrays (`"diffusion": [[]]`) can be mixed with the newer format. You