Matthew Johnson b3b4ffbc21 make slicing compilation a little faster ...

For the common special case of int or slice-of-int/None indexing, we can
generate a lax.slice rather than a lax.gather. That makes compilation a little
faster, and makes the generated jaxpr a bit more wieldy too, to process in
transformations and to read when pretty-printed.

2022-08-19 09:14:37 -07:00

..

api_benchmark.py

make slicing compilation a little faster

2022-08-19 09:14:37 -07:00

benchmark.py

feat: refactor code using pyupgrade

2022-05-17 22:14:05 +01:00

gda_benchmark.py

Fix the gda_xla_sharding_match benchmark which was regressing. This was happening because that function was executed from top to bottom a couple of times and each time a new mesh object was created violating the already created cache which doesn't happen in real life.

2022-07-27 23:08:55 -07:00

pmap_benchmark.py

Revert import path changes to examples/ and benchmarks/

2021-01-20 17:35:55 -08:00