Unofficial utility scripts for Microsoft Custom Vision Service
Project description
cvsutils
Unofficial utility scripts for Microsoft Custom Vision Service.
Setup
pip install cvsutils
export CVS_ENDPOINT='<endpoint>'
export CVS_TRAINING_KEY='<training_key>'
export CVS_PREDICTION_KEY='<prediction_key>'
export CVS_PREDICTION_RESOURCE_ID='<resource_id>'
Those keys and endpoint information can be found in the Custom Vision's settings page.
Available commands
# Show a list of projects
cvs_list_projects [--verbose]
# Create a new project
cvs_create_project <dataset_filepath> [--project_name <name>] [--domain_id <domain_id>]
# Download dataset from a project
cvs_download_project <project_id> <output_dir> [--filter_tag <tag_id> [<tag_id> ...]]
# Train a model
cvs_train_project <project_id> [--domain_id <domain_id>] [--type {multiclass,multilabel}] [--force]
# Export a model
cvs_export_model <project_id> <iteration_id> {tensorflow,coreml,onnx} [--output_filepath <filepath>]
And
- cvs_add_images
- cvs_download_predictions
- cvs_evaluate_project
- cvs_get_domains
- cvs_predict_dataset
- cvs_predict_image
- cvs_remove_iteration
- cvs_show_project
- cvs_validate_dataset
To see the detailed help, please run the command with "-h" option.
Dataset file format
This tool uses the SIMPLE dataset format to upload/download datasets from Custom Vision Service.
For details, please see the simpledataset repository.
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
cvsutils-0.1.11.tar.gz
(18.1 kB
view hashes)
Built Distribution
cvsutils-0.1.11-py3-none-any.whl
(26.8 kB
view hashes)
Close
Hashes for cvsutils-0.1.11-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | e89826b9d9cc9fa4840625b74061f7f38b4d25ef643303a2df1bb41378d23298 |
|
MD5 | 45d5a1cf07ee1162cf18a1c1e53ae8bc |
|
BLAKE2b-256 | 9149f14e9c1a981e34a7a922811970176b7abfd8fc7dd538916bdbb1aaaa9233 |