FuriosaAI model registry
Project description
Furiosa Registry
Furiosa Registry defines Model
which allows client to communicate with several types of registries.
Getting started
Install
pip install furiosa-registry
List up available models
import furiosa.common.thread as synchronous
import furiosa.registry as registry
# Repository where to load models from.
repository = "https://github.com/furiosa-ai/furiosa-artifacts"
# List up the available artifacts.
models: List[registry.Model] = synchronous(registry.list)(repository)
for model in models:
print(model.name)
Load models
import furiosa.common.thread as synchronous
import furiosa.registry as registry
# Repository where to load models from.
repository = "https://github.com/furiosa-ai/furiosa-artifacts"
models: registry.Model = synchronous(registry.load)(uri=repository, name="MLCommonsResNet50")
# Access the model
print(model.name)
print(model.version)
print(model.metadata.description)
Print documentation about a model
import furiosa.common.thread as synchronous
import furiosa.registry as registry
# Repository where to load models from.
repository = "https://github.com/furiosa-ai/furiosa-artifacts"
print(synchronous(registry.help)(repository, "MLCommonsResNet50"))
Development
Generate artfiact JSON schema from pydantic model definition.
When you changed model schema, you can generate modified schema via
python -c 'from furiosa.registry import Model;\
print(Model.schema_json(indent=2), file=open("model_schema.json", "w"))'
See model_schema.json
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
furiosa-registry-0.8.2.tar.gz
(71.1 kB
view hashes)
Built Distribution
Close
Hashes for furiosa_registry-0.8.2-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 386855173ab0a4811ef39ba0c5e797b14d766d919509eb2e5d02a04f5bdf832f |
|
MD5 | 5840bed32608f4b2cbdd887b66f31ffa |
|
BLAKE2b-256 | 9c2646fef630ebf05e712c71501aad66df20d2c1d894aee5e2b85a208ea02ebd |