George Hotz 62dbf12655 call matmul

2026-03-11 21:50:31 +08:00

..

apps

llm: fix chunked prefill (#15182)

2026-03-07 22:08:31 +08:00

codegen

even more divmod recombine (#15163)

2026-03-08 23:52:26 -04:00

engine

rm ops.encdec (#15193)

2026-03-09 18:52:48 +03:00

mixin

move more to mixins (#14780)

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

nn

don't use intermediate dict in onnx parse (#15181)

2026-03-07 00:08:03 -05:00

renderer

viz: always detect sqtt packet overlaps, add timeline tests (#15211)

2026-03-11 05:32:38 +09:00

runtime

nv: nvdec as NVDEC:0 device (#15209)

2026-03-10 14:44:50 +03:00

schedule

fix TestPartialAssignToSharedBuffer (#15202)

2026-03-09 23:14:23 -04:00

uop

call matmul

2026-03-11 21:50:31 +08:00

viz

viz: always detect sqtt packet overlaps, add timeline tests (#15211)

2026-03-11 05:32:38 +09:00

__init__.py

start function and add walk rewrite (#14992)

2026-02-25 13:56:27 +08:00

device.py

compile QCOM without opening device (#15165)

2026-03-06 06:24:27 -05:00

dtype.py

Invalid in Tensor (#15154)

2026-03-10 02:49:54 -04:00

function.py

call matmul

2026-03-11 21:50:31 +08:00

gradient.py

functions for llama trainer (#15045)

2026-02-28 12:15:18 +08:00

helpers.py

compile QCOM without opening device (#15165)

2026-03-06 06:24:27 -05:00

py.typed

add a single py.typed (#6083)

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

tensor.py

start using Invalid in image_conv2d (#15208)

2026-03-10 07:11:06 -04:00