Lightning-AI/lightning-thunder 0.2.5

Thunder 0.2.5 - Summer Harvest

Repository: Lightning-AI/lightning-thunder

Tag: 0.2.5

Published: 2025-09-10T12:27:27Z

Prerelease: no

Release notes:

What's Changed

• bump version to 0.2.5.dev0 by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2274
• [pre-commit.ci] pre-commit suggestions by @pre-commit-ci[bot] in https://github.com/Lightning-AI/lightning-thunder/pull/2271
• Add support for torch.argsort by @protonu in https://github.com/Lightning-AI/lightning-thunder/pull/2246
• Add test for Phi-3-vision-128k-instruct by @kshitij12345 in https://github.com/Lightning-AI/lightning-thunder/pull/1850
• DTensor: NVFuser Integration by @kshitij12345 in https://github.com/Lightning-AI/lightning-thunder/pull/2177
• fix lint from merge by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2277
• Remove F842 from ignore rules by @crcrpar in https://github.com/Lightning-AI/lightning-thunder/pull/2270
• E2E Coverage Test for Thunder by @tejapulagam in https://github.com/Lightning-AI/lightning-thunder/pull/2086
• convert pow gradient to new style by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2283
• add Windows xfail/skipif to tests not working on windows by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2284
• fix activation checkpointing in the joint trace by @beverlylytle in https://github.com/Lightning-AI/lightning-thunder/pull/2203
• Remove F811 from ignore rules by @crcrpar in https://github.com/Lightning-AI/lightning-thunder/pull/2268
• fix lint by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2286
• Return saved-for-backward objects as tuples by @beverlylytle in https://github.com/Lightning-AI/lightning-thunder/pull/2279
• [pre-commit.ci] pre-commit suggestions by @pre-commit-ci[bot] in https://github.com/Lightning-AI/lightning-thunder/pull/2287
• Update coverage requirement from ~=7.8.2 to ~=7.9.1 by @dependabot[bot] in https://github.com/Lightning-AI/lightning-thunder/pull/2292
• Add "transformer_engine_v2" to expected all executors set by @crcrpar in https://github.com/Lightning-AI/lightning-thunder/pull/2226
• Fix CI benchmarks by @KaelanDt in https://github.com/Lightning-AI/lightning-thunder/pull/2303
• Bump pytest-random-order from 1.1.1 to 1.2.0 by @dependabot[bot] in https://github.com/Lightning-AI/lightning-thunder/pull/2289
• prune redundant ifs in ci workflows by @Borda in https://github.com/Lightning-AI/lightning-thunder/pull/2296
• Bump pytest-cov from 6.1.1 to 6.2.1 by @dependabot[bot] in https://github.com/Lightning-AI/lightning-thunder/pull/2291
• Add hardsigmoid op by @beverlylytle in https://github.com/Lightning-AI/lightning-thunder/pull/2304
• Update hypothesis requirement from ~=6.133.0 to ~=6.135.20 by @dependabot[bot] in https://github.com/Lightning-AI/lightning-thunder/pull/2290
• skip complex dtype tensor from aminmax by @crcrpar in https://github.com/Lightning-AI/lightning-thunder/pull/2276
• Extend thunder.jit coverage on HF models by @lantiga in https://github.com/Lightning-AI/lightning-thunder/pull/2281
• Add PEFT benchmarking script in thunder/benchmarks by @riccardofelluga in https://github.com/Lightning-AI/lightning-thunder/pull/2254
• support tuples as replaces arg for operator registration by @KaelanDt in https://github.com/Lightning-AI/lightning-thunder/pull/2308
• Improve reporting from thunder.jit coverage CI job by @lantiga in https://github.com/Lightning-AI/lightning-thunder/pull/2309
• Decrease SKIPPED by adding dependencies by @lantiga in https://github.com/Lightning-AI/lightning-thunder/pull/2312
• Add scalar tensor input to full_sample_generator by @IvanYashchuk in https://github.com/Lightning-AI/lightning-thunder/pull/2318
• Update requirements/test.txt bitsandbytes to cover aarch64 platform_machine by @nWEIdia in https://github.com/Lightning-AI/lightning-thunder/pull/2321
• update TE test by @kshitij12345 in https://github.com/Lightning-AI/lightning-thunder/pull/2319
• [TE] catch different error for xfail test by @kshitij12345 in https://github.com/Lightning-AI/lightning-thunder/pull/2322
• Run TE tests in CI by @kshitij12345 in https://github.com/Lightning-AI/lightning-thunder/pull/2320
• Add ops for HF transformers by @kiya00 in https://github.com/Lightning-AI/lightning-thunder/pull/2217
• Move PEFT model materialization by @riccardofelluga in https://github.com/Lightning-AI/lightning-thunder/pull/2334
• Fix dataflow ordering for recomputed symbols by @riccardofelluga in https://github.com/Lightning-AI/lightning-thunder/pull/2317
• Update LoRA config for mamba models in PEFT by @riccardofelluga in https://github.com/Lightning-AI/lightning-thunder/pull/2333
• Remove external logger dependency by @riccardofelluga in https://github.com/Lightning-AI/lightning-thunder/pull/2327
• Relax inplace sanity check by @beverlylytle in https://github.com/Lightning-AI/lightning-thunder/pull/2314
• Make alias updating the default in-place operator approach by @beverlylytle in https://github.com/Lightning-AI/lightning-thunder/pull/2052
• Propagate backward tags more consistently by @beverlylytle in https://github.com/Lightning-AI/lightning-thunder/pull/2336
• Make inplace flags defaults consistent by @beverlylytle in https://github.com/Lightning-AI/lightning-thunder/pull/2349
• add/debug Lit CI by @Borda in https://github.com/Lightning-AI/lightning-thunder/pull/2339
• [pre-commit.ci] pre-commit suggestions by @pre-commit-ci[bot] in https://github.com/Lightning-AI/lightning-thunder/pull/2311
• ci: run notebooks with lit CI by @Borda in https://github.com/Lightning-AI/lightning-thunder/pull/2351
• fix docker sanity check by @Borda in https://github.com/Lightning-AI/lightning-thunder/pull/2352
• empty nvfuser.FusionCache after each test by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2354
• bump bitsandbytes from 0.42.0 to 0.46.1 by @lianakoleva in https://github.com/Lightning-AI/lightning-thunder/pull/2238
• remove spurious print by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2357
• limit gpu mem usage by @t-vi in https://github.com/Lightning-AI/lightning-thunder/pull/2358
• lit CI: switch to L4_X_2 by @Borda in https://github.com/Lightning-AI/lightning-thunder/pull/2360
• [grad test] relax tolerance by @kshitij12345 in https://github.com/Lightning-AI/lightning-thunder/pull/2364
• [dtensor] don't rely on __repr__ of DTenorSpec by @kshitij12345 in https://github.com/Lightning-AI/lightning-thunder/pull/2359
• [thunderfx]…