[TORCH] Added flex_attention hop function#4366

Open

keshavvinayak01 wants to merge 39 commits intollvm:mainfrom

keshavvinayak01:keshavvinayak01/torch-aten-flex_attention

Commits on Oct 22, 2025

Modified fx_importer to support hop_while_loop
keshavvinayak01
committed

Commits on Oct 23, 2025

Addressed Comments | Simplified unique child_func_name creation
keshavvinayak01
committed

Commits on Oct 27, 2025

Added Control flow test
keshavvinayak01
committed

Commits on Oct 29, 2025

Added AtenFluxAttentionOp
keshavvinayak01
committed

Commits on Oct 30, 2025

Added changes for correct functional references
keshavvinayak01
committed

Commits on Nov 6, 2025

return_lse is part of the kernel options
keshavvinayak01
committed

Commits on Nov 10, 2025

Modified arg extraction
keshavvinayak01
committed

Commits on Nov 12, 2025

Commits on Nov 13, 2025

Commits on Nov 14, 2025

Added basic_test for flex_attention
keshavvinayak01
committed

Commits on Nov 16, 2025

Formatting and allowed unused unpacked vals
keshavvinayak01
committed

Commits on Nov 19, 2025

Commits on Nov 20, 2025

Renamed aten.flex_attention -> hop_flex_attention; Added more lit tests
keshavvinayak01
committed

Commits on Nov 24, 2025

Commits on Nov 25, 2025

Added Verifier to HigherOrderFlexAttention operation
keshavvinayak01
committed

Commits on Nov 27, 2025

Removed Dynamic head check (backend specific)
keshavvinayak01
committed