Open-source projects · Established
pytorch/pytorch
Python tensor and neural-network framework
- ★ 105,062
- 7 days +241
- 30 days +1,905
- Forks 31,869
- Language:Python
- License:No license
Description on GitHub: Tensors and Dynamic neural networks in Python with strong GPU acceleration
Highlights
- Provides Python tensors that run on CPUs or GPUs, with operations for indexing, mathematics, and linear algebra.
- Tracks differentiable tensor operations with automatic differentiation, and provides torch.nn for building neural networks.
- Can be used with Python packages such as NumPy and SciPy; the README lists pip and Conda binaries as well as source builds.
- TechSeeker measures 105,062 stars and 1,905 new stars in the last 30 days; the project dates to 2016 and has multiple recent releases.
Our take
PyTorch spans tensor computation and neural-network construction, so teams should check CPU/GPU support and installation options against their target environments. TechSeeker detected no license file; confirm the applicable terms before incorporating it into a product or redistributing packages.
Star history
Data updated: Oct 25, 2026
Releases
- PyTorch 2.14.1 Release · Sep 30, 2026
- PyTorch 2.14.0 Release · Sep 2, 2026
- PyTorch 2.13.0 Release · Jul 8, 2026
- PyTorch 2.12.1 Release, bug fix release · Jun 18, 2026
- PyTorch 2.12.0 Release · May 13, 2026
The summary and analysis were generated by AI. The original sources are authoritative. AI content policy