Skip to content

Pull requests: Lightning-AI/lightning-thunder

New pull request New
50 Open 1,113 Closed
50 Open 1,113 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

remove unused to_torch_translator
#1830 opened Mar 3, 2025 by crcrpar • Draft
improve imports and type annotation of thunder.executors.passes
#1829 opened Mar 3, 2025 by crcrpar Loading…
Remove FutureTensorLike and use FutureTensorProxy thoroughly
#1828 opened Mar 3, 2025 by crcrpar Loading…
[TE] get the number of saved tensor for meta function with an actual execution TransformerEngine
#1826 opened Mar 3, 2025 by kshitij12345 Loading…
2
Tensor Wrapper Subclasses support through trace transforms documentation Improvements or additions to documentation
#1824 opened Mar 2, 2025 by crcrpar • Draft
Update hypothesis requirement from ~=6.124.7 to ~=6.127.3 dependencies python Pull requests that update Python code
#1816 opened Mar 1, 2025 by dependabot bot Loading…
Update ipython[all] requirement from ~=8.30.0 to ~=8.33.0 dependencies python Pull requests that update Python code
#1815 opened Mar 1, 2025 by dependabot bot Loading…
[ThunderFX report] Adds TimerWithCUDAMemoryUsage to track CUDA memory usage alongside timing measurements
#1813 opened Feb 28, 2025 by kiya00 • Draft
4 tasks
3
[TEST] specify device_id in init_process_group
#1802 opened Feb 25, 2025 by Aidyn-A Loading…
1
Update cuDNN SDPA execution functions to use the strides of attn_mask cudnn executors strides
#1798 opened Feb 25, 2025 by IvanYashchuk Loading…
3
allow embedding without renorm to be lowered as nvfuser index_select
#1796 opened Feb 24, 2025 by jjsjann123 • Draft
restore nvfuser take_along_axis support
#1795 opened Feb 24, 2025 by jjsjann123 • Draft
Split get_computation_and_inputs
#1784 opened Feb 20, 2025 by t-vi Loading…
1
Introduce update_aliases in-place operators
#1780 opened Feb 18, 2025 by beverlylytle Loading…
42
benchmark_litgpt - update low_precision_mode accepted values
#1779 opened Feb 18, 2025 by kshitij12345 Loading…
1
Avoid name collisions in _convert_pytorchfunc_to_thundertrace
#1777 opened Feb 18, 2025 by crcrpar • Draft
1
Remove requires_grad specification from meta functions autograd
#1769 opened Feb 17, 2025 by IvanYashchuk Loading…
benchmark functional optimizers
#1764 opened Feb 13, 2025 by khushi-411 Loading…
2 of 4 tasks
49
Revert updates to saved_tensors info on fw_extrace, no need to remove duplicate variables autograd
#1756 opened Feb 7, 2025 by IvanYashchuk Loading…
1
Update vjp implementation to attach residuals to all outputs autograd
#1755 opened Feb 7, 2025 by IvanYashchuk • Draft
2
Simplify and fix match_fw_and_bw_saved_for_bw_proxies implementation tracing architecture
#1754 opened Feb 7, 2025 by IvanYashchuk • Draft
2
[WIP - CI test] Perform rms_norm in fp32
#1751 opened Feb 6, 2025 by riccardofelluga • Draft
[nvfuserex] Decomposed torch._scaled_mm
#1749 opened Feb 6, 2025 by crcrpar • Draft
Relax test_te_with_retain_graph xfail strictness in test_transformer_engine_executor.py
#1718 opened Jan 30, 2025 by nWEIdia Loading…
3 tasks
Grad transform patch
#1712 opened Jan 29, 2025 by jjsjann123 • Draft
10
ProTip! What’s not been updated in a month: updated:<2025-02-03.