| Name | synapx JSON |
| Version |
0.1.0a0
JSON |
| download |
| home_page | None |
| Summary | A deep learning library with an autograd engine and core operations implemented in C++, featuring a Torch-like API. |
| upload_time | 2025-01-06 00:02:25 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | <3.13,>=3.9 |
| license | MIT |
| keywords |
autograd
tensor
machine learning
deep learning
c++
python
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# SynapX
**SynapX** is a Python Deep Learning library that implements its core functionality in C++ with Python bindings, designed to outperform my previous project, [SynapGrad](https://github.com/pgmesa/synapgrad). SynapGrad was built entirely in Python with NumPy as its tensor backend, while SynapX takes advantage of C++ for critical operations, offering higher performance and scalability.
This project combines the raw computational power of C++ with Python's ease of use, leveraging **libtorch** as the backend for CPU tensor operations, and aims to support GPU acceleration in the near future.
## Backend Exploration
The main branch uses libtorch as the tensor backend, selected for its efficiency and robust support for operations like broadcasting and batched matrix multiplications. For insights into the backend evaluation process and comparisons with alternatives like Xtensor, check out the [backend exploration branch](https://github.com/pgmesa/synapx/tree/xtensor-openblas).
## Project Status
SynapX is still under development...
Raw data
{
"_id": null,
"home_page": null,
"name": "synapx",
"maintainer": null,
"docs_url": null,
"requires_python": "<3.13,>=3.9",
"maintainer_email": null,
"keywords": "Autograd, Tensor, Machine Learning, Deep Learning, C++, Python",
"author": null,
"author_email": "Pablo Garc\u00eda Mesa <pgmesa.sm@gmail.com>",
"download_url": null,
"platform": null,
"description": "# SynapX\r\n\r\n**SynapX** is a Python Deep Learning library that implements its core functionality in C++ with Python bindings, designed to outperform my previous project, [SynapGrad](https://github.com/pgmesa/synapgrad). SynapGrad was built entirely in Python with NumPy as its tensor backend, while SynapX takes advantage of C++ for critical operations, offering higher performance and scalability. \r\n\r\nThis project combines the raw computational power of C++ with Python's ease of use, leveraging **libtorch** as the backend for CPU tensor operations, and aims to support GPU acceleration in the near future.\r\n\r\n## Backend Exploration \r\n\r\nThe main branch uses libtorch as the tensor backend, selected for its efficiency and robust support for operations like broadcasting and batched matrix multiplications. For insights into the backend evaluation process and comparisons with alternatives like Xtensor, check out the [backend exploration branch](https://github.com/pgmesa/synapx/tree/xtensor-openblas). \r\n\r\n## Project Status \r\n\r\nSynapX is still under development...\r\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "A deep learning library with an autograd engine and core operations implemented in C++, featuring a Torch-like API.",
"version": "0.1.0a0",
"project_urls": {
"Homepage": "https://github.com/pgmesa/synapx"
},
"split_keywords": [
"autograd",
" tensor",
" machine learning",
" deep learning",
" c++",
" python"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "9d6ab80098328267b98a925dac33df4b653ef5e87e56efa05420f69aa328f0c6",
"md5": "83e1edd4cdd29b8e4f864f80f603d4a3",
"sha256": "5ee1884c67f027b3eee289f57e84f5f9c5d63510c7ff70f4d61bf97fdddab2c1"
},
"downloads": -1,
"filename": "synapx-0.1.0a0-cp310-cp310-win_amd64.whl",
"has_sig": false,
"md5_digest": "83e1edd4cdd29b8e4f864f80f603d4a3",
"packagetype": "bdist_wheel",
"python_version": "cp310",
"requires_python": "<3.13,>=3.9",
"size": 894424,
"upload_time": "2025-01-06T00:02:25",
"upload_time_iso_8601": "2025-01-06T00:02:25.851896Z",
"url": "https://files.pythonhosted.org/packages/9d/6a/b80098328267b98a925dac33df4b653ef5e87e56efa05420f69aa328f0c6/synapx-0.1.0a0-cp310-cp310-win_amd64.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "b676c48d21978a13a07cc4e9ab918f5262a832cc156683b3be852aff01373419",
"md5": "de052c314f636a24da5a4cac15e06196",
"sha256": "20128bd847c2e7a1c160603c1765cc45611bac7f4cb918182feb7c2c25a0c32d"
},
"downloads": -1,
"filename": "synapx-0.1.0a0-cp311-cp311-win_amd64.whl",
"has_sig": false,
"md5_digest": "de052c314f636a24da5a4cac15e06196",
"packagetype": "bdist_wheel",
"python_version": "cp311",
"requires_python": "<3.13,>=3.9",
"size": 901146,
"upload_time": "2025-01-06T00:02:29",
"upload_time_iso_8601": "2025-01-06T00:02:29.567105Z",
"url": "https://files.pythonhosted.org/packages/b6/76/c48d21978a13a07cc4e9ab918f5262a832cc156683b3be852aff01373419/synapx-0.1.0a0-cp311-cp311-win_amd64.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "8bd63a2c5f8cf5c328306db6217446b49337cee821b4ec5aa84b648b01e7997e",
"md5": "8fafea2402ed758bbca384d5cbc09ed5",
"sha256": "be446fcf08803aa2c2fdd0c20ebccd69d35a0d350ee768eabb222bef57604101"
},
"downloads": -1,
"filename": "synapx-0.1.0a0-cp312-cp312-win_amd64.whl",
"has_sig": false,
"md5_digest": "8fafea2402ed758bbca384d5cbc09ed5",
"packagetype": "bdist_wheel",
"python_version": "cp312",
"requires_python": "<3.13,>=3.9",
"size": 600840,
"upload_time": "2025-01-06T00:02:33",
"upload_time_iso_8601": "2025-01-06T00:02:33.848578Z",
"url": "https://files.pythonhosted.org/packages/8b/d6/3a2c5f8cf5c328306db6217446b49337cee821b4ec5aa84b648b01e7997e/synapx-0.1.0a0-cp312-cp312-win_amd64.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "fddbe6a61cff242f927f19111c4573e9bc4dfb8294ba11548c2f486a14855953",
"md5": "0a5dcca6c39fbdf6f72179edcd8925f8",
"sha256": "23e4499093f0859ddd7a6baf374a4e5b5a28d2ef634eb3ae7722f5c24285531d"
},
"downloads": -1,
"filename": "synapx-0.1.0a0-cp39-cp39-win_amd64.whl",
"has_sig": false,
"md5_digest": "0a5dcca6c39fbdf6f72179edcd8925f8",
"packagetype": "bdist_wheel",
"python_version": "cp39",
"requires_python": "<3.13,>=3.9",
"size": 889551,
"upload_time": "2025-01-06T00:02:37",
"upload_time_iso_8601": "2025-01-06T00:02:37.602654Z",
"url": "https://files.pythonhosted.org/packages/fd/db/e6a61cff242f927f19111c4573e9bc4dfb8294ba11548c2f486a14855953/synapx-0.1.0a0-cp39-cp39-win_amd64.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-01-06 00:02:25",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "pgmesa",
"github_project": "synapx",
"github_not_found": true,
"lcname": "synapx"
}
