nimlgenandGitHub 08ef77c721 hsa multigpu graph (#3403) ...

* init hsa multigraph

* better handling of accesses to buffers

* revert sdma0 only when copies from fd

2024-02-28 09:40:53 -08:00

..

codegen

Revert "wmma: add CUDA tensor core (#3464)" (#3474)

2024-02-22 11:58:16 +01:00

features

hsa multigpu graph (#3403)

2024-02-28 09:40:53 -08:00

nn

remove cpu and torch backends (#3399)

2024-02-15 16:55:39 +01:00

renderer

float64 function support for HIP (#3492)

2024-02-24 09:46:20 -05:00

runtime

hsa multigpu graph (#3403)

2024-02-28 09:40:53 -08:00

shape

fix Node.max can be symbolic (#3514)

2024-02-27 17:21:31 -05:00

__init__.py

move graph.py and jit.py into features (#3376)

2024-02-12 17:34:34 +01:00

device.py

hsa multigpu graph (#3403)

2024-02-28 09:40:53 -08:00

dtype.py

Rename .sz to .count on DType (#3413)

2024-02-15 15:03:49 +01:00

helpers.py

no cuda compile helper (#3512)

2024-02-28 01:50:10 +01:00

lazy.py

cleanup lazy reduce (#3517)

2024-02-28 08:15:01 -05:00

mlops.py

change frontend reduce API to use axis (#3460)

2024-02-21 12:26:17 +01:00

ops.py

_reduce_op is axis based now (#3462)

2024-02-21 16:36:31 +01:00

realize.py

disk runner (prereq for interpreted removal) (#3421)

2024-02-15 18:14:05 +01:00

tensor.py

fix Tensor.split not passing dim to Tensor.chunk (#3490)

2024-02-24 07:53:11 -05:00