This is part of the Gavin Family of Repos. A basic FastAPI to be able to make http requests to Gavin. Everything is returned in JSON.
git submodule initInitialise Submodule Git Repos.git submodule updateFetch all required repos.
git clone https://github.com/Scot-Survivor/GavinFastAPI.git --recurse-submodulesTo automatically run the above commands.
python -m virtualenv ./GavinAPI/Create Virtual Environment.- Win:
./GavinAPI/Scripts/activate.batLinux/Mac:source ./GavinAPI/bin/activate pip install -r requirements.txtTo install all required libraries inside venv.
python3 main.py
Please note some settings api_config.json will not be parsed; you must manually edit the start.sh file:
--host <str> --port <int> --workers <int>
chmod +x start.sh./start.sh
PREFORMER(bool) True if the model is a preformer, false otherwise.MODEL_DIR(str) Path to where all the models will be stored such as "./models"DEFAULT_MODEL_NAME(str) Name of the model to load.VERSION_OVERRIDE(bool) Whether to overrideDEFAULT_MODEL_NAMEand base it on versions instead.VERSION(str) If this is not blank, then load a model with this version.HOST(str) The host to bind to.PORT(int) The port to bind to.FILTERED_WORDS(list) Words to censor on response from GavinTF_CPP_MIN_LOG_LEVEL(str["0", "1", "2", "3"]) Tensorflow log level. 0 Most verbose, 3 Least verboseUVICORN_WORKERS(int) How many workers to use.MESSAGE_TIMEOUT(int) How many seconds till the API returns "Message Timed Out Error."CACHE_REQUEST_MAX(int) How many requests should a Cache addition last for.MAX_CACHE_STORE(int) Max amount of messages that should be stored in the cache.LOGGING_LEVEL(str["INFO", "WARN", "ERROR", "DEBUG"]) Log level of internal server.
- Implement In Memory Cache (Added: 19/05/21)
- Implement Database (SQLite) based Cache (Added: 19/05/21)
- Implement Serving the associated Model Image (Added: 19/05/21)
GNU GPLv3. Should have a copy with this software.