Skip to content
Navigation Menu
Sign in
Products
Platform
GitHub
Repository hosting, code review, and collaboration
Copilot
AI coding agent for tasks, projects, and automated work
Build & Ship
Actions
CI/CD and automation
Code Quality
Quality analysis, fixes, and enforcement
Codespaces
Cloud development environments
Secure
Code Security
Vulnerability detection and remediation
Secret Protection
Secret scanning and leak prevention
All Features
Developers
Learn
Documentation
Training
Certifications
Build
Build with Copilot
Customize Copilot
Integrations
Explore
Trending repos
GitHub Sponsors
Developer articles
Changelog
Changelog
Your source for the latest features, improvements, and fixes across GitHub products, all in one place.
Businesses
By Stage
Enterprises
Team
Startups
Nonprofits
By Industry
Healthcare
Financial services
Manufacturing
Government
See all industries
Evaluate GitHub
Why GitHub
Customer Stories
Business insights
Trust Center
Add-ons
GitHub Advanced Security
Enterprise-grade code security
Copilot for business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
Resources
Learn & Connect
Documentation
Resource center
Blog
Events & webinars
Community Forum
Customer Stories
Integrations
Support
Customer support
Training
GitHub Status
Company
About
Partners
Changelog
GitHub Universe
GitHub Universe
The event for developers, by developers — talks, workshops, and what’s next for GitHub.
Pricing
Search
/
Sign in
Sign up
Appearance settings
You signed in with another tab or window.
Reload
to refresh your session.
You signed out in another tab or window.
Reload
to refresh your session.
You switched accounts on another tab or window.
Reload
to refresh your session.
Dismiss alert
{{ message }}
google-pytorch
torchtitan
Repository navigation
Code
Issues
21
(21)
Pull requests
77
(77)
Actions
Projects
Security and quality
Insights
More
items
All pull requests
Labels
10
(10)
Milestones
0
(0)
New pull request
Search pull requests
is
:
pr
state
:
open
is:pr state:open
Clear filter
Search
Pull requests
Open
77
(77)
Closed
109
(109)
Author
Label
Projects
Milestones
Reviews
Assignee
Sort by
Newest
descending
Sort…
Sort by Newest descending
More items
Comfortable display density
Compact display density
Register logical spmd_types mesh axes so multi-axis meshes work with enable_tpu_topo_mesh.
#219
·
copybara-service[bot]
opened
Nov 20, 2026
Bot
·
·
tpu: splash_tbwd for Muse Glimmer / Gemma4 shapes (streamed K / V, sliding window)
#218
·
pianpwk
opened
Nov 20, 2026
Contributor
·
·
tpu: splash_tbwd returns dq sequence-minor
#217
·
pianpwk
opened
Nov 20, 2026
Contributor
·
·
feat(tpu): Qwen3.5 CP speedups: cp_full_positions, per-batch GDN document resets, AG/RS GatedDeltaNet CP, fused GDN projections (stacked on #201)
#216
·
jialei777
opened
Nov 20, 2026
Contributor
·
·
Rebase 6: merge upstream e44a045e4 (2026-09-30: Python config loading, recipes reorg, torchcomms removal)
#215
·
jerryx99
opened
Nov 20, 2026
Contributor
·
·
1
[tpu] Support per-row backward for linear_softmax_cross_entropy_loss(reduction="none")
#214
·
wenjung2007
opened
Nov 19, 2026
Contributor
·
·
1
tpu: splash_tbwd matches the forward's two q roundings for a non-default scale
#213
·
pianpwk
opened
Nov 19, 2026
Contributor
·
·
[TPU] Support FSDP2 CPU offload with device-resident param cache and grad accumulation
#212
·
syangx39
opened
Nov 19, 2026
·
4 tasks done
·
·
[TPU] Use row-gather index_select backward for deterministic_scatter_add
#211
·
syangx39
opened
Nov 19, 2026
·
3 tasks done
·
·
fix(tpu): skip get_proxy_mode() checks while torch.compile traces the Qwen3.5 EP dispatcher
#210
·
jialei777
opened
Nov 19, 2026
Contributor
·
·
tpu: share pallas_vmem.tpu_generation() with causal_conv1d and tokamax_gated_delta
#209
·
tanvibajpai
opened
Nov 18, 2026
Contributor
·
·
[EP 9/9] tpu qwen3.5: expert-parallel recipes for 35B-A3B and 397B-A17B
#208
·
tanvibajpai
opened
Nov 18, 2026
Contributor
·
·
[EP 8/9] tpu moe: sequence token chunking, FP32 cross-chunk dW, and chunk barriers
pull ready
#207
·
tanvibajpai
opened
Nov 18, 2026
Contributor
·
·
6
[TPU] Qwen3.5 EP on the regular trainer with per-layer torch.compile
#206
·
jialei777
opened
Nov 18, 2026
Contributor
·
·
feat(tpu): Qwen3.5 context parallelism on top of fast dropless EP (stacked on #142)
#201
·
jialei777
opened
Nov 17, 2026
Contributor
·
·
feat(tpu): context parallelism for Qwen3.5 (head-parallel GatedDeltaNet + CP full attention)
#200
·
jialei777
opened
Nov 17, 2026
Contributor
·
·
tpu graph trainer: donate the in-graph optimizer's parameter and AdamW-moment inputs
#199
·
vlasenkoalexey
opened
Nov 17, 2026
Contributor
·
·
[tpu][graph_trainer] Return the loss function's metrics from the traced step
#198
·
wenjung2007
opened
Nov 17, 2026
Contributor
·
·
5
[rebase 5/10] Merge upstream 5dc97a3e7 (09/28: GroupedLinear MoE, BaseOptimizer.Config, ParallelismContext)
#197
·
jerryx99
opened
Nov 17, 2026
Contributor
·
·
1
[rebase-4] Merge upstream pytorch/torchtitan@b9e1d1f8 (2026-09-21, #4810 Model.Config & parallelize ownership)
#196
·
jerryx99
opened
Nov 16, 2026
Contributor
·
·
1
[rebase] Merge upstream pytorch/torchtitan@e9956a39 (2026-09-21, TrainingEngine)
#195
·
jerryx99
opened
Nov 16, 2026
Contributor
·
8/8
·
·
1
Carry upstream docs/ and pyproject.toml; drop attn-gym from the TPU image
#194
·
jerryx99
opened
Nov 16, 2026
Contributor
·
7/8
·
·
[rebase] Merge upstream pytorch/torchtitan@9c530254 (9/17/2026)
#193
·
jerryx99
opened
Nov 16, 2026
Contributor
·
6/8
·
·
1
[TPU] Add fused QK RMSNorm + CosSinRoPE Pallas kernel for Qwen3 and Gemma 4
#191
·
Perseus14
opened
Nov 14, 2026
·
·
2
Switch TorchTitan TPU Docker scripts to the public TorchTPU Artifact Registry
#190
·
copybara-service[bot]
opened
Nov 13, 2026
Bot
·
·
Previous
1
2
3
4
Next
You can’t perform that action at this time.