PyTorch 2.10.0 Release
AI Classified by Officially
# PyTorch 2.10.0 Release Notes
- [Highlights](#highlights)
- [Backwards Incompatible Changes](#backwards-incompatible-changes)
- [Deprecations](#deprecations)
- [New Features](#new-features)
- [Improvements](#improvements)
- [Bug fixes](#bug-fixes)
- [Performance](#performance)
- [Documentation](#documentation)
- [Developers](#developers)
- [Security](#security)
# Highlights
Python 3.14 support for
torch.compile(). Python 3.14t (freethreaded build) is experimentally supported as well.
Reduced kernel launch overhead with combo-kernels horizontal fusion in torchinductor
A new varlen_attn() op providing support for ragged and packed sequences
Efficient eigenvalue decompositions with DnXgeevtorch.compile() now respects use_deterministic_modeDebugMode for tracking dispatched calls and debugging numerical divergence - This makes it simpler to track down subtle numerical bugs.
Intel GPUs support:This is an extract. The publication continues at the source.
Read the original at the source: https://github.com/pytorch/pytorch/releases/tag/v2.10.0
Officially imported this from PyTorch’s own source and shows an extract. If you work there, claiming the profile and verifying the domain lets you choose to show the full text here.
This publication has changed since it was first published
2 versions recorded. The original is kept in full — nothing is overwritten.
- v2 imported change on current
- v1 as first published on
Provenance
- Organization
- PyTorch — imported from official source
- Official source
- https://api.github.com/repos/pytorch/pytorch/releases?per_page=25 API
- Imported
- September 15, 2026 19:08
- Versions
- 2 recorded
- Identity
278259050