torch

PyPI v2.11.0

18,324,039 weekly downloads · BSD-3-Clause · 14 Dependencies

Recently published

Source data as of:

Summary

Tensors and Dynamic neural networks in Python with strong GPU acceleration

Install pip install torch

Registry values

Reproduced verbatim from the official registry, with the source named on each value.

Weekly downloads 18,324,039 Source: PyPI (Python Package Index)
Releases 46 Last release: 2026-03-23 Source: PyPI (Python Package Index)
Dependencies 14 Source: PyPI (Python Package Index)

Description

Tensors and Dynamic neural networks in Python with strong GPU acceleration

Registry-supplied description, cleaned to plain text. Source: PyPI (Python Package Index).

Package details

Package
torch
Registry
PyPI
Version
2.11.0
Weekly downloads
18,324,039 (weekly)
License
BSD-3-Clause
Dependencies
14: filelock, typing-extensions, setuptools, sympy, networkx, jinja2, fsspec, cuda-toolkit, cuda-bindings, nvidia-cudnn-cu13, nvidia-cusparselt-cu13, nvidia-nccl-cu13, nvidia-nvshmem-cu13, triton
Releases
46
Last release
2026-03-23
First published
2018-12-07
Keywords
pytorch, machine learning
Repository
https://github.com/pytorch/pytorch

Why a field shows "—": A dash means the value is not published by the source registry for that package. Quanteta never fills in a guessed value — a missing field stays blank.

Derived indices (computation method published)

Quanteta-computed from the registry values below. This is a derived index, not a measured registry metric. See the formula on the Data & Sources page.

Q-Vitality Quanteta 64.9 / 100 Maintenance activity index (release cadence + download level).
Q-Trust Quanteta 10.1 / 100 Adoption / stability index (community size + download stability + age).
Q-Risk Quanteta 59.3 / 100 Dependency-surface index. Higher = more risk factors detected.

Data & Sources