Masterful AutoML Platform.
Project description
Masterful is the training platform for computer vision models.
Pass in your model architecture and data. Masterful figures out the right hyperparameters for its SOTA regularization, optimization, and semi-supervised learning engines. And then trains your model weights to the peak accuracy.
Features
- Built on Tensorflow 2.
- SOTA regularization implementation to improve accuracy from your existing labeled training data while pushing all computations to GPU.
- SOTA semi-supervised learning implementation to improve accuracy from your raw, unlabeled images.
- Automatic, high speed hyperparameter discovery algorithms, based on custom search algorithms and analytical / closed-form solutions, to reduce manual experimentation.
- Maximum GPU utilization and minimium GPU-hours used.
- Support for most computer vision tasks:
- Binary classification
- Single-label classification
- Multi-label classifcation
- Detection
- Semantic Segmentation
- Instance Segmentation (coming soon)
- Keypoint Detection (coming soon)
Key Links
Installation
For detailed instructions with a troubleshooting guide, see the Installation Guide.
The tldr instructions:
sudo apt install python3-pip
pip install --upgrade pip
pip install masterful
python -c "import masterful; print(masterful.__version__)"
Getting Help
Visit the documentation to see quick tutorials and in depth guides at www.masterfulai.com/docs.
Join the Slack community to speak to fellow users of Masterful and with Masterful AI's engineering team. You'll find a link to Slack from www.masterfulai.com.
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
Built Distribution
File details
Details for the file mful-dev-0.4.1.dev202204251650943961.tar.gz
.
File metadata
- Download URL: mful-dev-0.4.1.dev202204251650943961.tar.gz
- Upload date:
- Size: 25.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.8.0 pkginfo/1.8.2 readme-renderer/34.0 requests/2.27.1 requests-toolbelt/0.9.1 urllib3/1.26.9 tqdm/4.64.0 importlib-metadata/4.8.3 keyring/23.4.1 rfc3986/1.5.0 colorama/0.4.4 CPython/3.6.9
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | f50b35168d41add8e1c4c6204a4d174992e276af6af9843cbbc99951590f3370 |
|
MD5 | 6faaa0ad2b4326353f11fafddb9498af |
|
BLAKE2b-256 | e1fe5e86b77ea218af8b0c917d10d83406db0c4a1b5d55820bd6203d9ff1e931 |
File details
Details for the file mful_dev-0.4.1.dev202204251650943961-py3-none-any.whl
.
File metadata
- Download URL: mful_dev-0.4.1.dev202204251650943961-py3-none-any.whl
- Upload date:
- Size: 29.9 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.8.0 pkginfo/1.8.2 readme-renderer/34.0 requests/2.27.1 requests-toolbelt/0.9.1 urllib3/1.26.9 tqdm/4.64.0 importlib-metadata/4.8.3 keyring/23.4.1 rfc3986/1.5.0 colorama/0.4.4 CPython/3.6.9
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6a33fc646b01d5c87bb2ed903be4ed508c9b3585e94a6ec0f7323afaf10da387 |
|
MD5 | b5fa04f3b0bc4804aabf87c671a94ebb |
|
BLAKE2b-256 | 917a7ce0609f4fe1f46018e865c30a2f82de02ab35e6204ecfc2286cc91d11d8 |