Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/curegit/stylegan-fastapi
Simplified Web REST API of the StyleGAN using FastAPI
https://github.com/curegit/stylegan-fastapi
chainer fastapi generative-adversarial-network image-synthesis openapi stylegan2
Last synced: about 5 hours ago
JSON representation
Simplified Web REST API of the StyleGAN using FastAPI
- Host: GitHub
- URL: https://github.com/curegit/stylegan-fastapi
- Owner: curegit
- License: other
- Created: 2022-09-05T05:21:21.000Z (about 2 years ago)
- Default Branch: master
- Last Pushed: 2024-10-29T08:23:06.000Z (20 days ago)
- Last Synced: 2024-10-29T09:43:21.257Z (20 days ago)
- Topics: chainer, fastapi, generative-adversarial-network, image-synthesis, openapi, stylegan2
- Language: Python
- Homepage: https://stylegan.curegit.jp
- Size: 201 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# StyleGAN FastAPI
Simplified Web REST API of the StyleGAN using FastAPI
This is a web backend application to make the generator models trained by [Precure StyleGAN ADA](https://github.com/curegit/precure-stylegan-ada) available on the web in a simplified way.
[StyleGAN Vue](https://github.com/curegit/stylegan-vue) is a sample implementation of its frontend client and is [deployed here](https://stylegan.curegit.jp/) connected to [a demo backend](https://stylegan.mermaid.blue/docs) with some models for immediate trial.
[Sample Docker configurations](https://github.com/curegit/stylegan-web) for deploying StyleGAN models and the web-based user interface are also available.
## API Specifications
Refer to the [Swagger](https://stylegan.mermaid.blue/docs) or [ReDoc](https://stylegan.mermaid.blue/redoc) documentation on the sample backend or on your deployed instance at the paths `/docs` or `/redoc`.
## Application Requirements
To run this application, you need a Unix-like system with Python >= 3.12 installed.
Use the `requirements.txt` to install the minimal dependencies for serving and inferencing, including the dependencies of [Precure StyleGAN ADA](https://github.com/curegit/precure-stylegan-ada).
```sh
pip3 install -r requirements.txt
```
While Uvicorn is the recommended as ASGI middleware, other alternatives such as Hypercorn may also work.
However, in a production environment, it is common to use process replication with a combination of Gunicorn and Uvicorn.
## Quick Start: Running the default server
This repository contains the default server configuration using Uvicorn with some trained models.
You can quickly try it out as follows:
```sh
git clone --recursive https://github.com/curegit/stylegan-fastapi.git
cd stylegan-fastapi
pip3 install -r requirements.txt
python3 -m uvicorn main:app
```
Please note that this setup is intended for trial or debugging purposes only and is not suitable for production use as it does not support process replication.
## Application Configuration
StyleGAN FastAPI can be customized by using a configuration file.
StyleGAN FastAPI uses the TOML format for configuration, and the default server's TOML file is located in `default/config.toml`.
You can specify a different configuration TOML file to use by setting the `STYLEGAN_TOML` environment variable.
`STYLEGAN_TOML` must be an absolute TOML file path or a TOML file path relative to the working directory.
Note that the `STYLEGAN_TOML` environment variable must be set before importing the `api/` Python package.
```sh
export STYLEGAN_TOML="./custom_config.toml"
uvicorn main:app
```
### Configuration TOML Specifications
The configuration file follows the standard TOML format.
A complete raw schema definition and its default values can be found in the `api/conf.py` Python file.
All properties are optional except the `file` field in `ModelConfig` to specify model files, and you must specify at least one model.
#### General Settings (Root)
| Key | Type | Description |
| ------------- | --------------------------------------- | -------------------------------------------------------------------- |
| `title` | string | The title of the API service. Used in the OpenAPI spec. |
| `version` | string | The version number of the API. Used in the OpenAPI spec. |
| `description` | string | A brief description of the API service. Used in the OpenAPI spec. |
| `docs` | boolean | Whether the Swagger page is enabled. The default is `true`. |
| `redoc` | boolean | Whether the ReDoc page is enabled. The default is `true`. |
| `lossy` | boolean | Whether to use lossy compression for output. The default is `false`. |
| `server` | [ServerConfig](#serverconfig) | Internal server configuration settings. |
| `models` | { string: [ModelConfig](#modelconfig) } | A dictionary of model configurations to use. |
#### ServerConfig
| Key | Type | Description |
| --------- | --------------------------- | ------------------------------------------------------------------------------------------------------------------------ |
| `gpu` | boolean or int | A boolean indicating whether GPU is enabled or an integer value specifying which GPU device is being used. |
| `logger` | string | The name of the parent logger from which this application's logger is derived. |
| `tmp_dir` | string | The path to the temporary directory where the runtime files are located. Volatile locations such as `tmpfs` are allowed. |
| `http` | [HTTPConfig](#httpconfig) | HTTP configuration settings. |
| `limit` | [LimitConfig](#limitconfig) | Limit configuration settings. |
#### HTTPConfig
| Key | Type | Description |
| ------------------- | ------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| `forwarded` | boolean | Indicates whether reverse proxy servers are involved in the path of the requests. The default is `false`. You can leave this `false` if your middleware such as Uvicorn is configured to fill in real remote address info. |
| `forwarded_headers` | [string] | A list of headers to check for the original client address, such as "Forwarded" or "X-Forwarded-For", which can be multiple and is left in order. This is used for identifying clients when `forwarded` is true. Regardless of the header type, the entire string equivalence of the header value is used to identify clients. |
| `cors` | [CORSConfig](#corsconfig) | Cross-Origin Resource Sharing (CORS) configuration settings. |
#### CORSConfig
| Key | Type | Description |
| --------- | -------- | ----------------------------------------------------------------------------------------- |
| `enabled` | boolean | A boolean value indicating whether CORS configuration is enabled. The default is `false`. |
| `origins` | [string] | A list of allowed origins. Set `["*"]` for wildcard. |
#### LimitConfig
| Key | Type | Description |
| ------------- | ------------------------------------------------- | ------------------------------------------------------- |
| `min_delay` | float | The minimum response delay time for CPU-bound requests. |
| `block` | [SignallingBlockConfig](#signallingblockconfig) | A block of settings for the signalling block. |
| `concurrency` | [ConcurrencyLimitConfig](#concurrencylimitconfig) | A block of settings for the concurrency limit. |
| `rate` | [RateLimitConfig](#ratelimitconfig) | A block of settings for rate limiting. |
#### SignallingBlockConfig
| Key | Type | Description |
| --------- | ------- | --------------------------------------------------------------------------------------------------------------------- |
| `enabled` | boolean | Whether the signalling block for CPU-bound requests (delaying simultaneous requests from the same client) is enabled. |
| `timeout` | float | The timeout duration in seconds for the signalling block. |
| `poll` | float | The polling interval in seconds for the signalling block. |
#### ConcurrencyLimitConfig
| Key | Type | Description |
| ----------------- | ------- | ---------------------------------------------------------------- |
| `enabled` | boolean | Whether the concurrency limit for CPU-bound requests is enabled. |
| `max_concurrency` | int | The maximum number of concurrent requests. |
| `max_queue` | int | The maximum number of requests in the queue. |
| `timeout` | float | The timeout duration in seconds for the concurrency limit. |
| `poll` | float | The polling interval in seconds for the concurrency limit. |
#### RateLimitConfig
| Key | Type | Description |
| ------------- | ------- | ----------------------------------------------------------------------------------- |
| `enabled` | boolean | A boolean value indicating whether rate limiting for CPU-bound requests is enabled. |
| `window` | float | The time window for rate limiting in seconds. |
| `max_request` | int | The maximum number of requests allowed in the time window. |
#### ModelConfig
| Key | Type | Description |
| ------------- | -------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| `file` | string | The path to the model file. |
| `relative` | boolean | Indicates whether the model file path is relative to this TOML file rather than the working directory. The default value is `false`. |
| `name` | string | A display name of the model. |
| `description` | string | A short description of the model. |
| `lossy` | boolean | Whether lossy compression is used. This overrides the `lossy` in the [General Settings](#general-settings-root). |
| `gpu` | boolean or int | A boolean indicating whether GPU is enabled or an integer value specifying which GPU device is being used. This overrides the [ServerConfig](#serverconfig) `gpu`. |
## Gunicorn Tips
### Merge Application Log into Gunicorn Log
To merge the application log into the Gunicorn log, set `logger = "gunicorn.error"` in the StyleGAN FastAPI TOML configuration file.
### Derive Gunicorn Settings from `guni/conf.py`
You can customize how Gunicorn runs the application by overriding various options derived from sample Gunicorn settings.
```py
from guni.conf import *
wsgi_app = "main:app"
workers = 3
raw_env = ["STYLEGAN_TOML=myconfig.toml"]
```
### Using CUDA/CuPy with Gunicorn
There are certain limitations when using CUDA/CuPy:
CUDA has a restriction that prevents processes from being forked after CUDA has been initialized.
To address this, set `preload = False` in Gunicorn settings so that CUDA is initialized in each forked process after Gunicorn creates child processes.
Note that this approach will consume more RAM/VRAM.
If RAM/VRAM usage is a concern, it may be advisable to use Uvicorn alone even in production environments.
## Service Configuration Example
### `stylegan.service`
```ini
[Unit]
Description=Gunicorn StyleGAN Daemon
Requires=stylegan.socket
After=network.target
After=syslog.target
[Service]
Type=notify
User=spam
Group=spam
WorkingDirectory=/home/spam/stylegan
ExecStart=/usr/local/bin/python3.12 -m gunicorn -c gunicorn.conf.py \
--log-level info \
--log-file /var/log/stylegan/system.log \
--access-logfile /var/log/stylegan/access.log \
--capture-output
RuntimeDirectory=stylegan
LogsDirectory=stylegan
PrivateTmp=true
ExecReload=/bin/kill -s HUP $MAINPID
KillMode=mixed
TimeoutStartSec=240
TimeoutStopSec=30
Restart=always
[Install]
WantedBy=multi-user.target
```
If the volatile `/run` location is available via `RuntimeDirectory`, it is recommended to set as `tmp_dir = "/run/stylegan"` in the StyleGAN FastAPI TOML for improved disk IO performance.
#### `stylegan.socket`
```ini
[Unit]
Description=Gunicorn StyleGAN Socket
[Socket]
ListenStream=/run/stylegan.sock
SocketUser=www-data
SocketMode=600
[Install]
WantedBy=sockets.target
```
#### `logrotate.d/stylegan`
```txt
/var/log/stylegan/*.log {
monthly
dateformat -%Y-%m-%d
rotate 12
compress
ifempty
missingok
copytruncate
maxsize 100M
}
```
## Notes
- Ensure that proxy headers, if used, are set correctly (e.g., Forwarded, X-Forwarded-For) for production environments with a reverse proxy layer.
- It is also recommended to set header and body size limits via Gunicorn or a reverse proxy.
## License
[CC BY-NC 4.0](LICENSE)