| Name | art-training JSON |
| Version |
0.2.1
JSON |
| download |
| home_page | |
| Summary | ART - Actually Robust Training framework - a framework that teaches good practices when training deep neural networks. |
| upload_time | 2023-11-26 17:35:09 |
| maintainer | |
| docs_url | None |
| author | |
| requires_python | >=3.10 |
| license | |
| keywords |
deep learning
framework
pytorch
training
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
<p align="center"><img src="docs/art.png" alt="image" width="200" height="auto"></p>
# ART - Actually Robust Training framework
----
**ART** is a framework that teaches and keeps an eye on good practices when training deep neural networks. It is inspired by a [blog post by Andrej Karpathy “A Recipe for Training Neural Networks”](https://karpathy.github.io/2019/04/25/recipe/). The framework teaches the user how to properly train DNNs by encouraging the user to use built-in mechanisms that ensure the correctness and robustness of the pipeline using easily usable steps. It allows users not only to learn but also to use it in their future projects to speed up model development.
**Table of contents:**
- [ART - Actually Robust Training framework](#art---actually-robust-training-framework)
- [Installation](#installation)
- [Project creation](#project-creation)
- [Dashboard](#dashboard)
- [Tutorials](#tutorials)
- [Required knowledge](#required-knowledge)
- [Contributing](#contributing)
## Installation
To get started, install ART package using:
```sh
pip install art-training
```
## Project creation
To use most of art's features we encourage you to create a new folder for your project using the CLI tool:
```sh
python -m art.cli create-project my_project_name
```
This will create a new folder `my_project` with a basic structure for your project. To learn more about ART we encourage you to read our [documentation](https://actually-robust-training.readthedocs.io/en/latest/), and check our [tutorials](#tutorials)!
## Dashboard
After you run some steps you can see compare their execution in the dashboard. To use the dashboard, firstly install required dependencies:
```sh
pip install art-training[dashboard]
```
and run this command in the directory of your project (directory with folder called art_checkpoints).
```sh
python -m art.cli run-dashboard
```
Optionally you can use --experiment-folder switch to pass path to the folder. For more info, use --help switch.
## Tutorials
1. A showcase of ART's features. To check it out type:
```sh
python -m art.cli get-started
```
and launch tutorial.ipynb
After running all cells run dashboard with
```sh
python -m art.cli run-dashboard
```
2. A tutorial showing how to use ART for transfer learning in an NLP task.
```sh
python -m art.cli bert-transfer-learning-tutorial
```
## Required knowledge
In order to use ART, you should have a basic knowledge of:
- Python - you can find many tutorials online, e.g. [here](https://www.learnpython.org/)
- Basic knowledge of machine learning & neural networks - you can find many tutorials online, e.g. [here](https://www.coursera.org/learn/machine-learning)
- PyTorch - you can find many tutorials online, e.g. [here](https://pytorch.org/tutorials/)
- PyTorch Lightning - you can find many tutorials online, e.g. [here](https://lightning.ai/docs/pytorch/stable/levels/core_skills.html)
## Contributing
We welcome contributions to ART! Please check out our [contributing guide](https://github.com/SebChw/art/wiki/Contributing)
Raw data
{
"_id": null,
"home_page": "",
"name": "art-training",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.10",
"maintainer_email": "",
"keywords": "deep learning,framework,pytorch,training",
"author": "",
"author_email": "Sebastian Chwilczynski <sebastian.chwilczynski@student.put.poznan.pl>, Kacper Trebacz <kacper.trebacz@student.put.poznan.pl>, Mateusz Malecki <mateusz.malecki.1@student.put.poznan.pl>, Karol Cyganik <karol.cyganik@student.put.poznan.pl>",
"download_url": "https://files.pythonhosted.org/packages/fd/7f/7a9cb70b6ac3e9274672d4647233fe7f41ec6cc47eaa52dcbbad4e415049/art_training-0.2.1.tar.gz",
"platform": null,
"description": "<p align=\"center\"><img src=\"docs/art.png\" alt=\"image\" width=\"200\" height=\"auto\"></p>\n\n# ART - Actually Robust Training framework\n\n\n\n\n----\n\n**ART** is a framework that teaches and keeps an eye on good practices when training deep neural networks. It is inspired by a [blog post by Andrej Karpathy \u201cA Recipe for Training Neural Networks\u201d](https://karpathy.github.io/2019/04/25/recipe/). The framework teaches the user how to properly train DNNs by encouraging the user to use built-in mechanisms that ensure the correctness and robustness of the pipeline using easily usable steps. It allows users not only to learn but also to use it in their future projects to speed up model development.\n\n**Table of contents:**\n- [ART - Actually Robust Training framework](#art---actually-robust-training-framework)\n - [Installation](#installation)\n - [Project creation](#project-creation)\n - [Dashboard](#dashboard)\n - [Tutorials](#tutorials)\n - [Required knowledge](#required-knowledge)\n - [Contributing](#contributing)\n\n## Installation\nTo get started, install ART package using:\n```sh\npip install art-training\n```\n## Project creation\nTo use most of art's features we encourage you to create a new folder for your project using the CLI tool:\n```sh\npython -m art.cli create-project my_project_name\n```\n\nThis will create a new folder `my_project` with a basic structure for your project. To learn more about ART we encourage you to read our [documentation](https://actually-robust-training.readthedocs.io/en/latest/), and check our [tutorials](#tutorials)!\n\n## Dashboard\nAfter you run some steps you can see compare their execution in the dashboard. To use the dashboard, firstly install required dependencies:\n```sh\npip install art-training[dashboard]\n```\nand run this command in the directory of your project (directory with folder called art_checkpoints).\n```sh\npython -m art.cli run-dashboard\n```\nOptionally you can use --experiment-folder switch to pass path to the folder. For more info, use --help switch.\n\n## Tutorials\n1. A showcase of ART's features. To check it out type:\n```sh\npython -m art.cli get-started\n```\nand launch tutorial.ipynb\n\nAfter running all cells run dashboard with\n\n```sh\npython -m art.cli run-dashboard\n```\n\n2. A tutorial showing how to use ART for transfer learning in an NLP task.\n```sh\npython -m art.cli bert-transfer-learning-tutorial\n```\n\n\n## Required knowledge\nIn order to use ART, you should have a basic knowledge of:\n- Python - you can find many tutorials online, e.g. [here](https://www.learnpython.org/)\n- Basic knowledge of machine learning & neural networks - you can find many tutorials online, e.g. [here](https://www.coursera.org/learn/machine-learning)\n- PyTorch - you can find many tutorials online, e.g. [here](https://pytorch.org/tutorials/)\n- PyTorch Lightning - you can find many tutorials online, e.g. [here](https://lightning.ai/docs/pytorch/stable/levels/core_skills.html)\n\n## Contributing\nWe welcome contributions to ART! Please check out our [contributing guide](https://github.com/SebChw/art/wiki/Contributing)",
"bugtrack_url": null,
"license": "",
"summary": "ART - Actually Robust Training framework - a framework that teaches good practices when training deep neural networks.",
"version": "0.2.1",
"project_urls": {
"Bug Tracker": "https://github.com/sebchw/art/issues",
"Homepage": "https://github.com/sebchw/art"
},
"split_keywords": [
"deep learning",
"framework",
"pytorch",
"training"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "74a781fd074315740a38b24d3ae75042b806556ee6c6e652ea88f08d275ff1f8",
"md5": "f5c961492d256b463bca62994d5bec4f",
"sha256": "4309e9f204b3fb193e198994986345bfb0b6a9379ccb78cf34c8ce381beee0c1"
},
"downloads": -1,
"filename": "art_training-0.2.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "f5c961492d256b463bca62994d5bec4f",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10",
"size": 30032,
"upload_time": "2023-11-26T17:35:07",
"upload_time_iso_8601": "2023-11-26T17:35:07.911045Z",
"url": "https://files.pythonhosted.org/packages/74/a7/81fd074315740a38b24d3ae75042b806556ee6c6e652ea88f08d275ff1f8/art_training-0.2.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "fd7f7a9cb70b6ac3e9274672d4647233fe7f41ec6cc47eaa52dcbbad4e415049",
"md5": "49091943d754d2e7ad19798465538c20",
"sha256": "f3c2c886b6561fbb63ef96e960aedb59e555a7a94b1f05374b59c2bcd6c4cc43"
},
"downloads": -1,
"filename": "art_training-0.2.1.tar.gz",
"has_sig": false,
"md5_digest": "49091943d754d2e7ad19798465538c20",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10",
"size": 95961,
"upload_time": "2023-11-26T17:35:09",
"upload_time_iso_8601": "2023-11-26T17:35:09.999460Z",
"url": "https://files.pythonhosted.org/packages/fd/7f/7a9cb70b6ac3e9274672d4647233fe7f41ec6cc47eaa52dcbbad4e415049/art_training-0.2.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-11-26 17:35:09",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "sebchw",
"github_project": "art",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"tox": true,
"lcname": "art-training"
}
