George HotzandGitHub b824490e3f allocate generates a call (#14958) ...

* allocate generates a call

* symbolic works too

* DEFINE_VAR is param

* replace param later

* apply buffers

* name

* upd

* this was a bug...

2026-02-23 15:59:20 +08:00

..

apps

correct llm.py mem bw benchmark for moe (#14626)

2026-02-20 16:11:22 -05:00

codegen

test IMAGE=1 driving_vision in mac pytest (#14921)

2026-02-20 18:28:10 -05:00

engine

allocate generates a call (#14958)

2026-02-23 15:59:20 +08:00

mixin

move more to mixins (#14780)

2026-02-16 11:35:00 +08:00

nn

fix: optim device is never none here (#14963)

2026-02-22 23:34:57 -08:00

renderer

ptx: use setp.neu for float CMPNE (#14805)

2026-02-20 16:11:04 -05:00

runtime

sqtt: fix jitted program deduping, failing test for graphed kernels (#14951)

2026-02-22 15:22:31 +09:00

schedule

allocate generates a call (#14958)

2026-02-23 15:59:20 +08:00

uop

allocate generates a call (#14958)

2026-02-23 15:59:20 +08:00

viz

viz: smaller view for repeated asm instructions in cfg (#14954)

2026-02-23 10:41:43 +09:00

__init__.py

move files into uop dir (#10399)

2025-05-18 11:38:28 -07:00

device.py

remove handle_allreduce_multirank and group_id [pr] (#14869)

2026-02-18 16:13:54 -05:00

dtype.py

cleanup fp8 conversion helpers and fp8 edge-case tests (#14953)

2026-02-22 09:16:42 -05:00

gradient.py

remove all the outerworld stuff, it was too complex (#14852)

2026-02-18 17:44:11 +08:00

helpers.py

print schedule caller with DEBUG=1 (#14933)

2026-02-21 16:22:45 +08:00

py.typed

add a single py.typed (#6083)

2024-08-14 17:31:46 -07:00

tensor.py

break Tensor.pad into methods (#14922)

2026-02-20 20:10:09 -05:00