Skip to content

Conversation

@avik-pal
Copy link
Collaborator

@avik-pal avik-pal commented Jan 8, 2026

No description provided.

@avik-pal avik-pal force-pushed the ap/compare_op_non_negative_simplify branch 2 times, most recently from 536d134 to e29d921 Compare January 9, 2026 00:42
@avik-pal avik-pal requested a review from wsmoses January 9, 2026 00:42
@avik-pal avik-pal marked this pull request as ready for review January 9, 2026 00:42
@avik-pal avik-pal force-pushed the ap/compare_op_non_negative_simplify branch from e29d921 to c2ea08a Compare January 9, 2026 00:42
@avik-pal avik-pal force-pushed the ap/compare_op_non_negative_simplify branch from c2ea08a to c9032fc Compare January 9, 2026 01:23
@avik-pal avik-pal merged commit 0aabb97 into main Jan 9, 2026
20 of 23 checks passed
@avik-pal avik-pal deleted the ap/compare_op_non_negative_simplify branch January 9, 2026 04:03
Copy link
Contributor

@github-actions github-actions bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

EnzymeJAX Benchmarks

Details
Benchmark suite Current: c9032fc Previous: 6fa5904 Ratio
actmtch / JaXPipe / cpu / Primal 0.000007201319931482431 s 0.000013693 s 0.53
actmtch / Jax / cpu / Primal 0.000006899879936099751 s 0.000013585 s 0.51
actmtch / HLOOpt / cpu / Primal 0.000008180900003935676 s 0.000014202 s 0.58
actmtch / PartOpt / cpu / Primal 0.00000693867999871145 s 0.000013269 s 0.52
actmtch / IPartOpt / cpu / Primal 0.000007204520061350195 s 0.000013422 s 0.54
actmtch / DefOpt / cpu / Primal 0.00000776888005930232 s 0.000013765 s 0.56
actmtch / IDefOpt / cpu / Primal 0.000007733199981885264 s 0.000014063 s 0.55
actmtch / JaXPipe / cpu / Forward 0.000011701580078806728 s 0.000019581 s 0.60
actmtch / Jax / cpu / Forward 0.000010763479931483744 s 0.000017748 s 0.61
actmtch / HLOOpt / cpu / Forward 0.000011524779947649222 s 0.000018994 s 0.61
actmtch / PartOpt / cpu / Forward 0.000011748980014090194 s 0.000018623 s 0.63
actmtch / IPartOpt / cpu / Forward 0.000012245100006111896 s 0.000019281 s 0.64
actmtch / DefOpt / cpu / Forward 0.000011359860091033624 s 0.000018901 s 0.60
actmtch / IDefOpt / cpu / Forward 0.000011715440032276091 s 0.000019099 s 0.61
actmtch / JaXPipe / cpu / PreRev 0.000011614999875746434 s 0.000019544 s 0.59
actmtch / JaXPipe / cpu / PostRev 0.000011189559954800642 s 0.000017476 s 0.64
actmtch / JaXPipe / cpu / BothRev 0.000011909900058526544 s 0.000019317 s 0.62
actmtch / Jax / cpu / BothRev 0.000010502559998712968 s 0.000017624 s 0.60
actmtch / HLOOpt / cpu / PreRev 0.000012520419932116057 s 0.000019388 s 0.65
actmtch / HLOOpt / cpu / PostRev 0.000013937340081611185 s 0.000019539 s 0.71
actmtch / HLOOpt / cpu / BothRev 0.000012083559995517135 s 0.000019324 s 0.63
actmtch / PartOpt / cpu / PreRev 0.000011736840024241246 s 0.000019265 s 0.61
actmtch / PartOpt / cpu / PostRev 0.000011038480006391182 s 0.000017578 s 0.63
actmtch / PartOpt / cpu / BothRev 0.00001202295998155023 s 0.000019093 s 0.63
actmtch / IPartOpt / cpu / PreRev 0.00001199831991471001 s 0.000019073 s 0.63
actmtch / IPartOpt / cpu / PostRev 0.000010725960055424369 s 0.000017405 s 0.62
actmtch / IPartOpt / cpu / BothRev 0.00001221437998538022 s 0.000019417 s 0.63
actmtch / DefOpt / cpu / PreRev 0.000011741720027202972 s 0.000019016 s 0.62
actmtch / DefOpt / cpu / PostRev 0.000012539139970613175 s 0.000019018 s 0.66
actmtch / DefOpt / cpu / BothRev 0.000011934440008189995 s 0.000019349 s 0.62
actmtch / IDefOpt / cpu / PreRev 0.00001198208001369494 s 0.000019193 s 0.62
actmtch / IDefOpt / cpu / PostRev 0.00001256468001884059 s 0.000018937 s 0.66
actmtch / IDefOpt / cpu / BothRev 0.00001217254000948742 s 0.000019236 s 0.63
actmtch / JaXPipe / cuda / Primal 0.000002016 s 0.0000024 s 0.84
actmtch / Jax / cuda / Primal 0.000002016 s 0.0000024 s 0.84
actmtch / HLOOpt / cuda / Primal 0.000002015 s 0.0000024 s 0.84
actmtch / PartOpt / cuda / Primal 0.000002015 s 0.0000024 s 0.84
actmtch / IPartOpt / cuda / Primal 0.000002015 s 0.0000024 s 0.84
actmtch / DefOpt / cuda / Primal 0.000002015 s 0.0000024 s 0.84
actmtch / IDefOpt / cuda / Primal 0.000002016 s 0.0000024 s 0.84
actmtch / JaXPipe / cuda / Forward 0.000010112 s 0.000010687 s 0.95
actmtch / Jax / cuda / Forward 0.000009792 s 0.00001104 s 0.89
actmtch / HLOOpt / cuda / Forward 0.000009793 s 0.0000104 s 0.94
actmtch / PartOpt / cuda / Forward 0.00001008 s 0.000010848 s 0.93
actmtch / IPartOpt / cuda / Forward 0.00001008 s 0.000010528 s 0.96
actmtch / DefOpt / cuda / Forward 0.000010112 s 0.00001088 s 0.93
actmtch / IDefOpt / cuda / Forward 0.000009888 s 0.000011071 s 0.89
actmtch / JaXPipe / cuda / PreRev 0.000009664 s 0.000010432 s 0.93
actmtch / JaXPipe / cuda / PostRev 0.000010048 s 0.00001072 s 0.94
actmtch / JaXPipe / cuda / BothRev 0.000009824 s 0.000010496 s 0.94
actmtch / Jax / cuda / BothRev 0.000010304 s 0.000010849 s 0.95
actmtch / HLOOpt / cuda / PreRev 0.000010048 s 0.000010592 s 0.95
actmtch / HLOOpt / cuda / PostRev 0.000010144 s 0.000010783 s 0.94
actmtch / HLOOpt / cuda / BothRev 0.000010272 s 0.000010528 s 0.98
actmtch / PartOpt / cuda / PreRev 0.00001008 s 0.00001072 s 0.94
actmtch / PartOpt / cuda / PostRev 0.000010048 s 0.000010752 s 0.93
actmtch / PartOpt / cuda / BothRev 0.000010049 s 0.000010047 s 1.00
actmtch / IPartOpt / cuda / PreRev 0.000010016 s 0.000010463 s 0.96
actmtch / IPartOpt / cuda / PostRev 0.000010208 s 0.000011424 s 0.89
actmtch / IPartOpt / cuda / BothRev 0.000010112 s 0.000010656 s 0.95
actmtch / DefOpt / cuda / PreRev 0.00001136 s 0.000010688 s 1.06
actmtch / DefOpt / cuda / PostRev 0.000009951 s 0.000010624 s 0.94
actmtch / DefOpt / cuda / BothRev 0.000010144 s 0.000010624 s 0.95
actmtch / IDefOpt / cuda / PreRev 0.00000976 s 0.00001072 s 0.91
actmtch / IDefOpt / cuda / PostRev 0.000010176 s 0.000010272 s 0.99
actmtch / IDefOpt / cuda / BothRev 0.000010112 s 0.00001088 s 0.93
actmtch / JaXPipe / cpu / Primal 0.000017187 s 0.000013693 s 1.26
actmtch / Jax / cpu / Primal 0.000017111000000000003 s 0.000013585 s 1.26
actmtch / HLOOpt / cpu / Primal 0.000017604000000000003 s 0.000014202 s 1.24
actmtch / PartOpt / cpu / Primal 0.000016953 s 0.000013269 s 1.28
actmtch / IPartOpt / cpu / Primal 0.000017122 s 0.000013422 s 1.28
actmtch / DefOpt / cpu / Primal 0.00001799 s 0.000013765 s 1.31
actmtch / IDefOpt / cpu / Primal 0.000017565000000000002 s 0.000014063 s 1.25
actmtch / JaXPipe / cpu / Forward 0.000024319 s 0.000019581 s 1.24
actmtch / Jax / cpu / Forward 0.000023146 s 0.000017748 s 1.30
actmtch / HLOOpt / cpu / Forward 0.000024225 s 0.000018994 s 1.28
actmtch / PartOpt / cpu / Forward 0.000025067 s 0.000018623 s 1.35
actmtch / IPartOpt / cpu / Forward 0.000023899 s 0.000019281 s 1.24
actmtch / DefOpt / cpu / Forward 0.000024835 s 0.000018901 s 1.31
actmtch / IDefOpt / cpu / Forward 0.000024257 s 0.000019099 s 1.27
actmtch / JaXPipe / cpu / PreRev 0.000024999 s 0.000019544 s 1.28
actmtch / JaXPipe / cpu / PostRev 0.000022427 s 0.000017476 s 1.28
actmtch / JaXPipe / cpu / BothRev 0.000031608 s 0.000019317 s 1.64
actmtch / Jax / cpu / BothRev 0.000022738 s 0.000017624 s 1.29
actmtch / HLOOpt / cpu / PreRev 0.000024907 s 0.000019388 s 1.28
actmtch / HLOOpt / cpu / PostRev 0.000025351 s 0.000019539 s 1.30
actmtch / HLOOpt / cpu / BothRev 0.000024748 s 0.000019324 s 1.28
actmtch / PartOpt / cpu / PreRev 0.000024524 s 0.000019265 s 1.27
actmtch / PartOpt / cpu / PostRev 0.000021846 s 0.000017578 s 1.24
actmtch / PartOpt / cpu / BothRev 0.00002493 s 0.000019093 s 1.31
actmtch / IPartOpt / cpu / PreRev 0.00002437 s 0.000019073 s 1.28
actmtch / IPartOpt / cpu / PostRev 0.000022499000000000003 s 0.000017405 s 1.29
actmtch / IPartOpt / cpu / BothRev 0.000031097 s 0.000019417 s 1.60
actmtch / DefOpt / cpu / PreRev 0.000024822 s 0.000019016 s 1.31
actmtch / DefOpt / cpu / PostRev 0.00002538 s 0.000019018 s 1.33
actmtch / DefOpt / cpu / BothRev 0.000025226 s 0.000019349 s 1.30
actmtch / IDefOpt / cpu / PreRev 0.00002453 s 0.000019193 s 1.28
actmtch / IDefOpt / cpu / PostRev 0.000024517 s 0.000018937 s 1.29
actmtch / IDefOpt / cpu / BothRev 0.000024662 s 0.000019236 s 1.28
add_one / JaXPipe / cpu / Primal 0.000007034380050754408 s 0.000012845 s 0.55
add_one / Jax / cpu / Primal 0.000007319819906115299 s 0.000013048 s 0.56
add_one / HLOOpt / cpu / Primal 0.000007308860094781267 s 0.000012855 s 0.57
add_one / PartOpt / cpu / Primal 0.000006844860017736209 s 0.000013045 s 0.52
add_one / IPartOpt / cpu / Primal 0.000007604719994560582 s 0.000012816 s 0.59
add_one / DefOpt / cpu / Primal 0.000007214479974209098 s 0.000012893 s 0.56
add_one / IDefOpt / cpu / Primal 0.0000072192400330095555 s 0.000012853 s 0.56
add_one / JaXPipe / cpu / Forward 0.000011213860125280916 s 0.000017968 s 0.62
add_one / Jax / cpu / Forward 0.000011201679990335834 s 0.000017565000000000002 s 0.64
add_one / HLOOpt / cpu / Forward 0.000011516659960761898 s 0.000017607999999999998 s 0.65
add_one / PartOpt / cpu / Forward 0.000011175659965374509 s 0.00001764 s 0.63
add_one / IPartOpt / cpu / Forward 0.000010871740068978396 s 0.000017749000000000002 s 0.61
add_one / DefOpt / cpu / Forward 0.000011006819931935752 s 0.000017621000000000003 s 0.62
add_one / IDefOpt / cpu / Forward 0.000011178679997101428 s 0.000017704000000000002 s 0.63
add_one / JaXPipe / cpu / PreRev 0.00001332221996563021 s 0.000020014 s 0.67
add_one / JaXPipe / cpu / PostRev 0.000012538280043372653 s 0.00001934 s 0.65
add_one / JaXPipe / cpu / BothRev 0.000013508799984265352 s 0.000019994 s 0.68
add_one / Jax / cpu / BothRev 0.000013378680014284328 s 0.000019969 s 0.67
add_one / HLOOpt / cpu / PreRev 0.00001322725998761598 s 0.000019379 s 0.68
add_one / HLOOpt / cpu / PostRev 0.000014914419989509042 s 0.000019822 s 0.75
add_one / HLOOpt / cpu / BothRev 0.00001320842005952727 s 0.000030914 s 0.43
add_one / PartOpt / cpu / PreRev 0.0000127738600349403 s 0.000019639 s 0.65
add_one / PartOpt / cpu / PostRev 0.000013462459955917438 s 0.000019533 s 0.69
add_one / PartOpt / cpu / BothRev 0.000013185060015530326 s 0.000019776 s 0.67
add_one / IPartOpt / cpu / PreRev 0.000013336700012587244 s 0.000019921 s 0.67
add_one / IPartOpt / cpu / PostRev 0.000012693620010395537 s 0.000019436 s 0.65
add_one / IPartOpt / cpu / BothRev 0.000012813619996450144 s 0.000019761 s 0.65
add_one / DefOpt / cpu / PreRev 0.000012694759971054736 s 0.000019927 s 0.64
add_one / DefOpt / cpu / PostRev 0.000012853939970227658 s 0.000019651 s 0.65
add_one / DefOpt / cpu / BothRev 0.000012963379849679769 s 0.000019777 s 0.66
add_one / IDefOpt / cpu / PreRev 0.000013084940019325584 s 0.00001971 s 0.66
add_one / IDefOpt / cpu / PostRev 0.000013241919969004812 s 0.000019798 s 0.67
add_one / IDefOpt / cpu / BothRev 0.000012539740055217408 s 0.000019814 s 0.63
add_one / JaXPipe / cuda / Primal 0.0000019200000000000003 s 0.000002304 s 0.83
add_one / Jax / cuda / Primal 0.0000019200000000000003 s 0.000002335 s 0.82
add_one / HLOOpt / cuda / Primal 0.0000019200000000000003 s 0.000002304 s 0.83
add_one / PartOpt / cuda / Primal 0.0000019200000000000003 s 0.000002304 s 0.83
add_one / IPartOpt / cuda / Primal 0.0000019200000000000003 s 0.000002304 s 0.83
add_one / DefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002303 s 0.83
add_one / IDefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002335 s 0.82
add_one / JaXPipe / cuda / Forward 0.00001136 s 0.000010783 s 1.05
add_one / Jax / cuda / Forward 0.000010016 s 0.000010784 s 0.93
add_one / HLOOpt / cuda / Forward 0.00000976 s 0.000010752 s 0.91
add_one / PartOpt / cuda / Forward 0.00001024 s 0.000010496 s 0.98
add_one / IPartOpt / cuda / Forward 0.00000992 s 0.000010464 s 0.95
add_one / DefOpt / cuda / Forward 0.000010016 s 0.000010752 s 0.93
add_one / IDefOpt / cuda / Forward 0.000009664 s 0.000010464 s 0.92
add_one / JaXPipe / cuda / PreRev 0.0000248 s 0.000025248 s 0.98
add_one / JaXPipe / cuda / PostRev 0.000024928 s 0.000025952 s 0.96
add_one / JaXPipe / cuda / BothRev 0.000025216 s 0.000025376 s 0.99
add_one / Jax / cuda / BothRev 0.000024704 s 0.000025824 s 0.96
add_one / HLOOpt / cuda / PreRev 0.000025024 s 0.00002592 s 0.97
add_one / HLOOpt / cuda / PostRev 0.00002496 s 0.0000256 s 0.98
add_one / HLOOpt / cuda / BothRev 0.000025056 s 0.000025856 s 0.97
add_one / PartOpt / cuda / PreRev 0.000025376 s 0.000026048 s 0.97
add_one / PartOpt / cuda / PostRev 0.000024863 s 0.000025536 s 0.97
add_one / PartOpt / cuda / BothRev 0.000025184 s 0.000025984 s 0.97
add_one / IPartOpt / cuda / PreRev 0.000025024 s 0.000026208 s 0.95
add_one / IPartOpt / cuda / PostRev 0.00002512 s 0.00002544 s 0.99
add_one / IPartOpt / cuda / BothRev 0.000025152 s 0.000026304 s 0.96
add_one / DefOpt / cuda / PreRev 0.000025536 s 0.000026976 s 0.95
add_one / DefOpt / cuda / PostRev 0.000025088 s 0.000025568 s 0.98
add_one / DefOpt / cuda / BothRev 0.000025344 s 0.000025983 s 0.98
add_one / IDefOpt / cuda / PreRev 0.000025632 s 0.000025792 s 0.99
add_one / IDefOpt / cuda / PostRev 0.0000248 s 0.000025504 s 0.97
add_one / IDefOpt / cuda / BothRev 0.000024607 s 0.000025728 s 0.96
add_one / JaXPipe / cpu / Primal 0.000016910999999999998 s 0.000012845 s 1.32
add_one / Jax / cpu / Primal 0.000016255 s 0.000013048 s 1.25
add_one / HLOOpt / cpu / Primal 0.000016346 s 0.000012855 s 1.27
add_one / PartOpt / cpu / Primal 0.000016702 s 0.000013045 s 1.28
add_one / IPartOpt / cpu / Primal 0.000017168 s 0.000012816 s 1.34
add_one / DefOpt / cpu / Primal 0.000016122000000000003 s 0.000012893 s 1.25
add_one / IDefOpt / cpu / Primal 0.000016597 s 0.000012853 s 1.29
add_one / JaXPipe / cpu / Forward 0.000022809 s 0.000017968 s 1.27
add_one / Jax / cpu / Forward 0.00002325 s 0.000017565000000000002 s 1.32
add_one / HLOOpt / cpu / Forward 0.000023168 s 0.000017607999999999998 s 1.32
add_one / PartOpt / cpu / Forward 0.000022062 s 0.00001764 s 1.25
add_one / IPartOpt / cpu / Forward 0.000022365 s 0.000017749000000000002 s 1.26
add_one / DefOpt / cpu / Forward 0.00002254 s 0.000017621000000000003 s 1.28
add_one / IDefOpt / cpu / Forward 0.00002353 s 0.000017704000000000002 s 1.33
add_one / JaXPipe / cpu / PreRev 0.000024604 s 0.000020014 s 1.23
add_one / JaXPipe / cpu / PostRev 0.000024662 s 0.00001934 s 1.28
add_one / JaXPipe / cpu / BothRev 0.000024958 s 0.000019994 s 1.25
add_one / Jax / cpu / BothRev 0.00002419 s 0.000019969 s 1.21
add_one / HLOOpt / cpu / PreRev 0.000025334 s 0.000019379 s 1.31
add_one / HLOOpt / cpu / PostRev 0.000024793 s 0.000019822 s 1.25
add_one / HLOOpt / cpu / BothRev 0.000024742 s 0.000030914 s 0.80
add_one / PartOpt / cpu / PreRev 0.000024883 s 0.000019639 s 1.27
add_one / PartOpt / cpu / PostRev 0.000025087 s 0.000019533 s 1.28
add_one / PartOpt / cpu / BothRev 0.000030182 s 0.000019776 s 1.53
add_one / IPartOpt / cpu / PreRev 0.00002498 s 0.000019921 s 1.25
add_one / IPartOpt / cpu / PostRev 0.000024899 s 0.000019436 s 1.28
add_one / IPartOpt / cpu / BothRev 0.000024874 s 0.000019761 s 1.26
add_one / DefOpt / cpu / PreRev 0.000024266 s 0.000019927 s 1.22
add_one / DefOpt / cpu / PostRev 0.000024573 s 0.000019651 s 1.25
add_one / DefOpt / cpu / BothRev 0.000024245 s 0.000019777 s 1.23
add_one / IDefOpt / cpu / PreRev 0.000025175 s 0.00001971 s 1.28
add_one / IDefOpt / cpu / PostRev 0.000024673 s 0.000019798 s 1.25
add_one / IDefOpt / cpu / BothRev 0.000024585 s 0.000019814 s 1.24
add_two / JaXPipe / cpu / Primal 0.000008259959977294783 s 0.000013174 s 0.63
add_two / Jax / cpu / Primal 0.000007852259950595907 s 0.000013197 s 0.60
add_two / HLOOpt / cpu / Primal 0.000007710800036875299 s 0.000013221 s 0.58
add_two / PartOpt / cpu / Primal 0.000007940079995023552 s 0.000013127 s 0.60
add_two / IPartOpt / cpu / Primal 0.000007903200021246448 s 0.000013694 s 0.58
add_two / DefOpt / cpu / Primal 0.000007476479913748335 s 0.000013493 s 0.55
add_two / IDefOpt / cpu / Primal 0.00000761823996072053 s 0.000013285 s 0.57
add_two / JaXPipe / cpu / Forward 0.000011124900011054706 s 0.00001841 s 0.60
add_two / Jax / cpu / Forward 0.00001174106002508779 s 0.000018237 s 0.64
add_two / HLOOpt / cpu / Forward 0.000011820599956990918 s 0.000018121 s 0.65
add_two / PartOpt / cpu / Forward 0.00001198922000185121 s 0.000017817 s 0.67
add_two / IPartOpt / cpu / Forward 0.00001154243993369164 s 0.000018347 s 0.63
add_two / DefOpt / cpu / Forward 0.000011452540038590086 s 0.000017979 s 0.64
add_two / IDefOpt / cpu / Forward 0.000011464599992905278 s 0.000017652 s 0.65
add_two / JaXPipe / cpu / PreRev 0.000015792379963386337 s 0.000023426 s 0.67
add_two / JaXPipe / cpu / PostRev 0.000015142579941311852 s 0.000023189 s 0.65
add_two / JaXPipe / cpu / BothRev 0.000015681359873269684 s 0.000023003 s 0.68
add_two / Jax / cpu / BothRev 0.000016002479987946572 s 0.000022819 s 0.70
add_two / HLOOpt / cpu / PreRev 0.000014953359932405872 s 0.000023062 s 0.65
add_two / HLOOpt / cpu / PostRev 0.000017233040034625445 s 0.000023305 s 0.74
add_two / HLOOpt / cpu / BothRev 0.00001474040001994581 s 0.000022702 s 0.65
add_two / PartOpt / cpu / PreRev 0.000015172719977272208 s 0.000023151 s 0.66
add_two / PartOpt / cpu / PostRev 0.000015370119999715826 s 0.000023277 s 0.66
add_two / PartOpt / cpu / BothRev 0.000015777060016262113 s 0.000023147 s 0.68
add_two / IPartOpt / cpu / PreRev 0.000016025779987103307 s 0.00002274 s 0.70
add_two / IPartOpt / cpu / PostRev 0.000014813960060564568 s 0.000023172 s 0.64
add_two / IPartOpt / cpu / BothRev 0.000015196239965007408 s 0.00002328 s 0.65
add_two / DefOpt / cpu / PreRev 0.000015920499972708057 s 0.000022870000000000003 s 0.70
add_two / DefOpt / cpu / PostRev 0.000015671059973101365 s 0.000023133 s 0.68
add_two / DefOpt / cpu / BothRev 0.00001569124000525335 s 0.000023381 s 0.67
add_two / IDefOpt / cpu / PreRev 0.000015320500006055225 s 0.00002299 s 0.67
add_two / IDefOpt / cpu / PostRev 0.00001476474015362328 s 0.000022887 s 0.65
add_two / IDefOpt / cpu / BothRev 0.00001537245991130476 s 0.000023289 s 0.66
add_two / JaXPipe / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
add_two / Jax / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
add_two / HLOOpt / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
add_two / PartOpt / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
add_two / IPartOpt / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
add_two / DefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
add_two / IDefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
add_two / JaXPipe / cuda / Forward 0.000009536 s 0.000010529 s 0.91
add_two / Jax / cuda / Forward 0.00000992 s 0.000010272 s 0.97
add_two / HLOOpt / cuda / Forward 0.00000912 s 0.0000104 s 0.88
add_two / PartOpt / cuda / Forward 0.0000096 s 0.000010591 s 0.91
add_two / IPartOpt / cuda / Forward 0.000009632 s 0.00001088 s 0.89
add_two / DefOpt / cuda / Forward 0.00000992 s 0.000010208 s 0.97
add_two / IDefOpt / cuda / Forward 0.00000976 s 0.00001072 s 0.91
add_two / JaXPipe / cuda / PreRev 0.000032256 s 0.0000352 s 0.92
add_two / JaXPipe / cuda / PostRev 0.000031583 s 0.000033472 s 0.94
add_two / JaXPipe / cuda / BothRev 0.000032544 s 0.000033472 s 0.97
add_two / Jax / cuda / BothRev 0.000032800000000000004 s 0.00003296 s 1.00
add_two / HLOOpt / cuda / PreRev 0.000032256 s 0.000033183 s 0.97
add_two / HLOOpt / cuda / PostRev 0.000032256 s 0.000032671 s 0.99
add_two / HLOOpt / cuda / BothRev 0.000032352 s 0.000033119999999999995 s 0.98
add_two / PartOpt / cuda / PreRev 0.000050592 s 0.000033696 s 1.50
add_two / PartOpt / cuda / PostRev 0.00003168 s 0.000032864 s 0.96
add_two / PartOpt / cuda / BothRev 0.000032447 s 0.000033408 s 0.97
add_two / IPartOpt / cuda / PreRev 0.000032735 s 0.000033695 s 0.97
add_two / IPartOpt / cuda / PostRev 0.000032608 s 0.000033248 s 0.98
add_two / IPartOpt / cuda / BothRev 0.000032160000000000004 s 0.000033216 s 0.97
add_two / DefOpt / cuda / PreRev 0.000032672 s 0.000033184 s 0.98
add_two / DefOpt / cuda / PostRev 0.000032 s 0.000033217000000000004 s 0.96
add_two / DefOpt / cuda / BothRev 0.000032608 s 0.000033248 s 0.98
add_two / IDefOpt / cuda / PreRev 0.000032608 s 0.000033407 s 0.98
add_two / IDefOpt / cuda / PostRev 0.000032448 s 0.000033344 s 0.97
add_two / IDefOpt / cuda / BothRev 0.000032 s 0.000033375000000000005 s 0.96
add_two / JaXPipe / cpu / Primal 0.000017527 s 0.000013174 s 1.33
add_two / Jax / cpu / Primal 0.000016643000000000003 s 0.000013197 s 1.26
add_two / HLOOpt / cpu / Primal 0.000017032 s 0.000013221 s 1.29
add_two / PartOpt / cpu / Primal 0.000017135 s 0.000013127 s 1.31
add_two / IPartOpt / cpu / Primal 0.000017109 s 0.000013694 s 1.25
add_two / DefOpt / cpu / Primal 0.000017247999999999998 s 0.000013493 s 1.28
add_two / IDefOpt / cpu / Primal 0.000017021999999999997 s 0.000013285 s 1.28
add_two / JaXPipe / cpu / Forward 0.000022628 s 0.00001841 s 1.23
add_two / Jax / cpu / Forward 0.000022445 s 0.000018237 s 1.23
add_two / HLOOpt / cpu / Forward 0.000022991 s 0.000018121 s 1.27
add_two / PartOpt / cpu / Forward 0.000022732 s 0.000017817 s 1.28
add_two / IPartOpt / cpu / Forward 0.000022375 s 0.000018347 s 1.22
add_two / DefOpt / cpu / Forward 0.00002273 s 0.000017979 s 1.26
add_two / IDefOpt / cpu / Forward 0.000022687 s 0.000017652 s 1.29
add_two / JaXPipe / cpu / PreRev 0.000029078 s 0.000023426 s 1.24
add_two / JaXPipe / cpu / PostRev 0.00002851 s 0.000023189 s 1.23
add_two / JaXPipe / cpu / BothRev 0.000028740000000000003 s 0.000023003 s 1.25
add_two / Jax / cpu / BothRev 0.000028589 s 0.000022819 s 1.25
add_two / HLOOpt / cpu / PreRev 0.000027803 s 0.000023062 s 1.21
add_two / HLOOpt / cpu / PostRev 0.000029359 s 0.000023305 s 1.26
add_two / HLOOpt / cpu / BothRev 0.000028242 s 0.000022702 s 1.24
add_two / PartOpt / cpu / PreRev 0.000028773 s 0.000023151 s 1.24
add_two / PartOpt / cpu / PostRev 0.000029281000000000003 s 0.000023277 s 1.26
add_two / PartOpt / cpu / BothRev 0.000029223 s 0.000023147 s 1.26
add_two / IPartOpt / cpu / PreRev 0.000034102 s 0.00002274 s 1.50
add_two / IPartOpt / cpu / PostRev 0.000028833 s 0.000023172 s 1.24
add_two / IPartOpt / cpu / BothRev 0.000028773 s 0.00002328 s 1.24
add_two / DefOpt / cpu / PreRev 0.000029006 s 0.000022870000000000003 s 1.27
add_two / DefOpt / cpu / PostRev 0.000028359 s 0.000023133 s 1.23
add_two / DefOpt / cpu / BothRev 0.00002806 s 0.000023381 s 1.20
add_two / IDefOpt / cpu / PreRev 0.000028431000000000003 s 0.00002299 s 1.24
add_two / IDefOpt / cpu / PostRev 0.000035423 s 0.000022887 s 1.55
add_two / IDefOpt / cpu / BothRev 0.00002943 s 0.000023289 s 1.26
cache / JaXPipe / cpu / Primal 0.000007788720013195417 s 0.000012915 s 0.60
cache / Jax / cpu / Primal 0.000006901620054122759 s 0.000013084 s 0.53
cache / HLOOpt / cpu / Primal 0.000007173459998739417 s 0.00001278 s 0.56
cache / PartOpt / cpu / Primal 0.000006703539984300733 s 0.000012674 s 0.53
cache / IPartOpt / cpu / Primal 0.000007330079915845999 s 0.000012571 s 0.58
cache / DefOpt / cpu / Primal 0.00000720321993867401 s 0.000013024 s 0.55
cache / IDefOpt / cpu / Primal 0.000007182459921750706 s 0.000012491 s 0.58
cache / JaXPipe / cpu / Forward 0.000016098240048449953 s 0.000023874 s 0.67
cache / Jax / cpu / Forward 0.000015890279992163415 s 0.000017412000000000002 s 0.91
cache / HLOOpt / cpu / Forward 0.00001702454002952436 s 0.000016971 s 1.00
cache / PartOpt / cpu / Forward 0.000016424240020569413 s 0.000017065 s 0.96
cache / IPartOpt / cpu / Forward 0.000016462699913972756 s 0.000017104 s 0.96
cache / DefOpt / cpu / Forward 0.000016849940020620125 s 0.000017124 s 0.98
cache / IDefOpt / cpu / Forward 0.00001571123995745438 s 0.000017069999999999998 s 0.92
cache / JaXPipe / cpu / PreRev 0.000017700279986456735 s 0.000017882999999999998 s 0.99
cache / JaXPipe / cpu / PostRev 0.000021089599977131 s 0.000020973 s 1.01
cache / JaXPipe / cpu / BothRev 0.00001802619995942223 s 0.00001835 s 0.98
cache / Jax / cpu / BothRev 0.00002078466006423696 s 0.000021256 s 0.98
cache / HLOOpt / cpu / PreRev 0.000017231199890375138 s 0.000017772 s 0.97
cache / HLOOpt / cpu / PostRev 0.000019978979998995783 s 0.000017704000000000002 s 1.13
cache / HLOOpt / cpu / BothRev 0.000017123080087912968 s 0.000018288 s 0.94
cache / PartOpt / cpu / PreRev 0.000015736880031909093 s 0.000017628 s 0.89
cache / PartOpt / cpu / PostRev 0.00002113714002916822 s 0.000019973 s 1.06
cache / PartOpt / cpu / BothRev 0.000017007519982144004 s 0.000017819 s 0.95
cache / IPartOpt / cpu / PreRev 0.000017031740062520838 s 0.000017434999999999998 s 0.98
cache / IPartOpt / cpu / PostRev 0.000019669999965117316 s 0.00002052 s 0.96
cache / IPartOpt / cpu / BothRev 0.00001779625994458911 s 0.00001842 s 0.97
cache / DefOpt / cpu / PreRev 0.00001696633988103713 s 0.000017628 s 0.96
cache / DefOpt / cpu / PostRev 0.000016383759993914283 s 0.00001768 s 0.93
cache / DefOpt / cpu / BothRev 0.000016485000032844254 s 0.000017449 s 0.94
cache / IDefOpt / cpu / PreRev 0.00001667702001213911 s 0.000017389000000000002 s 0.96
cache / IDefOpt / cpu / PostRev 0.000015776520067447565 s 0.00001734 s 0.91
cache / IDefOpt / cpu / BothRev 0.00003045335990464082 s 0.000018119 s 1.68
cache / JaXPipe / cuda / Primal 0.000002303 s 0.000002335 s 0.99
cache / Jax / cuda / Primal 0.000002304 s 0.000002335 s 0.99
cache / HLOOpt / cuda / Primal 0.00000224 s 0.000002335 s 0.96
cache / PartOpt / cuda / Primal 0.00000224 s 0.000002335 s 0.96
cache / IPartOpt / cuda / Primal 0.000002304 s 0.000002335 s 0.99
cache / DefOpt / cuda / Primal 0.00000224 s 0.000002335 s 0.96
cache / IDefOpt / cuda / Primal 0.000002209 s 0.000002336 s 0.95
cache / JaXPipe / cuda / Forward 0.000002335 s 0.000002336 s 1.00
cache / Jax / cuda / Forward 0.000002335 s 0.0000023670000000000004 s 0.99
cache / HLOOpt / cuda / Forward 0.000002335 s 0.000002335 s 1
cache / PartOpt / cuda / Forward 0.000002335 s 0.0000023670000000000004 s 0.99
cache / IPartOpt / cuda / Forward 0.000002335 s 0.000002336 s 1.00
cache / DefOpt / cuda / Forward 0.000002272 s 0.0000023670000000000004 s 0.96
cache / IDefOpt / cuda / Forward 0.000002335 s 0.000002336 s 1.00
cache / JaXPipe / cuda / PreRev 0.000010496 s 0.000010591 s 0.99
cache / JaXPipe / cuda / PostRev 0.000010304 s 0.000010752 s 0.96
cache / JaXPipe / cuda / BothRev 0.000010304 s 0.000010816 s 0.95
cache / Jax / cuda / BothRev 0.000010944 s 0.00001072 s 1.02
cache / HLOOpt / cuda / PreRev 0.000013536 s 0.0000136 s 1.00
cache / HLOOpt / cuda / PostRev 0.000013472 s 0.000013568 s 0.99
cache / HLOOpt / cuda / BothRev 0.000013536 s 0.0000136 s 1.00
cache / PartOpt / cuda / PreRev 0.000010784 s 0.000010688 s 1.01
cache / PartOpt / cuda / PostRev 0.000010784 s 0.000010688 s 1.01
cache / PartOpt / cuda / BothRev 0.000010495 s 0.000010752 s 0.98
cache / IPartOpt / cuda / PreRev 0.00001088 s 0.00001088 s 1
cache / IPartOpt / cuda / PostRev 0.0000104 s 0.000010592 s 0.98
cache / IPartOpt / cuda / BothRev 0.000010752 s 0.000010592 s 1.02
cache / DefOpt / cuda / PreRev 0.000011103 s 0.000010688 s 1.04
cache / DefOpt / cuda / PostRev 0.00001088 s 0.000010784 s 1.01
cache / DefOpt / cuda / BothRev 0.00001056 s 0.000010815 s 0.98
cache / IDefOpt / cuda / PreRev 0.000010656 s 0.000010592 s 1.01
cache / IDefOpt / cuda / PostRev 0.000010816 s 0.000010432 s 1.04
cache / IDefOpt / cuda / BothRev 0.00001072 s 0.000010944 s 0.98
cache / JaXPipe / cpu / Primal 0.000019204 s 0.000012915 s 1.49
cache / Jax / cpu / Primal 0.000019196 s 0.000013084 s 1.47
cache / HLOOpt / cpu / Primal 0.000019121 s 0.00001278 s 1.50
cache / PartOpt / cpu / Primal 0.000019375 s 0.000012674 s 1.53
cache / IPartOpt / cpu / Primal 0.000019127 s 0.000012571 s 1.52
cache / DefOpt / cpu / Primal 0.000019011 s 0.000013024 s 1.46
cache / IDefOpt / cpu / Primal 0.00001919 s 0.000012491 s 1.54
cache / JaXPipe / cpu / Forward 0.000027992 s 0.000023874 s 1.17
cache / Jax / cpu / Forward 0.000029526 s 0.000017412000000000002 s 1.70
cache / HLOOpt / cpu / Forward 0.000028862 s 0.000016971 s 1.70
cache / PartOpt / cpu / Forward 0.0000287 s 0.000017065 s 1.68
cache / IPartOpt / cpu / Forward 0.000023351 s 0.000017104 s 1.37
cache / DefOpt / cpu / Forward 0.000022127 s 0.000017124 s 1.29
cache / IDefOpt / cpu / Forward 0.000022363000000000003 s 0.000017069999999999998 s 1.31
cache / JaXPipe / cpu / PreRev 0.000023223 s 0.000017882999999999998 s 1.30
cache / JaXPipe / cpu / PostRev 0.000027043 s 0.000020973 s 1.29
cache / JaXPipe / cpu / BothRev 0.000023516 s 0.00001835 s 1.28
cache / Jax / cpu / BothRev 0.00002702 s 0.000021256 s 1.27
cache / HLOOpt / cpu / PreRev 0.000031205 s 0.000017772 s 1.76
cache / HLOOpt / cpu / PostRev 0.000022456 s 0.000017704000000000002 s 1.27
cache / HLOOpt / cpu / BothRev 0.000023465 s 0.000018288 s 1.28
cache / PartOpt / cpu / PreRev 0.000022647 s 0.000017628 s 1.28
cache / PartOpt / cpu / PostRev 0.000026537 s 0.000019973 s 1.33
cache / PartOpt / cpu / BothRev 0.000027707 s 0.000017819 s 1.55
cache / IPartOpt / cpu / PreRev 0.000024641 s 0.000017434999999999998 s 1.41
cache / IPartOpt / cpu / PostRev 0.000025745 s 0.00002052 s 1.25
cache / IPartOpt / cpu / BothRev 0.000033968 s 0.00001842 s 1.84
cache / DefOpt / cpu / PreRev 0.000035927000000000003 s 0.000017628 s 2.04
cache / DefOpt / cpu / PostRev 0.000034671 s 0.00001768 s 1.96
cache / DefOpt / cpu / BothRev 0.000022249 s 0.000017449 s 1.28
cache / IDefOpt / cpu / PreRev 0.000022741 s 0.000017389000000000002 s 1.31
cache / IDefOpt / cpu / PostRev 0.000022222 s 0.00001734 s 1.28
cache / IDefOpt / cpu / BothRev 0.00002324 s 0.000018119 s 1.28
Concat / JaXPipe / cpu / Primal 0.000007456739949702751 s 0.000012834 s 0.58
Concat / Jax / cpu / Primal 0.000007775160083838273 s 0.000012809 s 0.61
Concat / HLOOpt / cpu / Primal 0.000007000319947110256 s 0.000013168 s 0.53
Concat / PartOpt / cpu / Primal 0.000007046319951768964 s 0.000013103 s 0.54
Concat / IPartOpt / cpu / Primal 0.000007259900030476274 s 0.000012935 s 0.56
Concat / DefOpt / cpu / Primal 0.000007452939953509485 s 0.000013538 s 0.55
Concat / IDefOpt / cpu / Primal 0.000007563439976365771 s 0.000012679 s 0.60
Concat / JaXPipe / cpu / Forward 0.000011245020068599844 s 0.000018003 s 0.62
Concat / Jax / cpu / Forward 0.000011482760037324624 s 0.000017769 s 0.65
Concat / HLOOpt / cpu / Forward 0.000011660739983199164 s 0.000017542999999999998 s 0.66
Concat / PartOpt / cpu / Forward 0.000010989980019076027 s 0.000017561 s 0.63
Concat / IPartOpt / cpu / Forward 0.0000111760000254435 s 0.000017515 s 0.64
Concat / DefOpt / cpu / Forward 0.00001143011993917753 s 0.0000176 s 0.65
Concat / IDefOpt / cpu / Forward 0.000010942740009340924 s 0.000017825 s 0.61
Concat / JaXPipe / cpu / PreRev 0.000013116779937263346 s 0.00002049 s 0.64
Concat / JaXPipe / cpu / PostRev 0.000012138660040363902 s 0.000019962 s 0.61
Concat / JaXPipe / cpu / BothRev 0.00001262137991943746 s 0.000019878 s 0.63
Concat / Jax / cpu / BothRev 0.000012852839990955544 s 0.000020003 s 0.64
Concat / HLOOpt / cpu / PreRev 0.000012378939991322114 s 0.000020007 s 0.62
Concat / HLOOpt / cpu / PostRev 0.000014624119994550713 s 0.000020165 s 0.73
Concat / HLOOpt / cpu / BothRev 0.00001253108004675596 s 0.000019592 s 0.64
Concat / PartOpt / cpu / PreRev 0.000012716259971057295 s 0.000020409 s 0.62
Concat / PartOpt / cpu / PostRev 0.000012448319903342054 s 0.00002033 s 0.61
Concat / PartOpt / cpu / BothRev 0.000013392720065894536 s 0.000019275 s 0.69
Concat / IPartOpt / cpu / PreRev 0.000012670139931287847 s 0.000019956 s 0.63
Concat / IPartOpt / cpu / PostRev 0.00001253014004760189 s 0.000019605 s 0.64
Concat / IPartOpt / cpu / BothRev 0.000012867919976997655 s 0.000019625 s 0.66
Concat / DefOpt / cpu / PreRev 0.000012563260024762713 s 0.000020967 s 0.60
Concat / DefOpt / cpu / PostRev 0.000012539779945655029 s 0.000019458 s 0.64
Concat / DefOpt / cpu / BothRev 0.000013063999995210909 s 0.00001977 s 0.66
Concat / IDefOpt / cpu / PreRev 0.00001310702005866915 s 0.000020008 s 0.66
Concat / IDefOpt / cpu / PostRev 0.00001235310000993195 s 0.000019651 s 0.63
Concat / IDefOpt / cpu / BothRev 0.000012811820033675758 s 0.000019793 s 0.65
Concat / JaXPipe / cuda / Primal 0.0000019200000000000003 s 0.000002463 s 0.78
Concat / Jax / cuda / Primal 0.0000019200000000000003 s 0.000002463 s 0.78
Concat / HLOOpt / cuda / Primal 0.0000019200000000000003 s 0.000002432 s 0.79
Concat / PartOpt / cuda / Primal 0.0000019200000000000003 s 0.000002463 s 0.78
Concat / IPartOpt / cuda / Primal 0.0000019200000000000003 s 0.000002431 s 0.79
Concat / DefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002463 s 0.78
Concat / IDefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002432 s 0.79
Concat / JaXPipe / cuda / Forward 0.00000992 s 0.000010656 s 0.93
Concat / Jax / cuda / Forward 0.000010112 s 0.000010527 s 0.96
Concat / HLOOpt / cuda / Forward 0.000009952 s 0.000010688 s 0.93
Concat / PartOpt / cuda / Forward 0.000009728 s 0.000010848 s 0.90
Concat / IPartOpt / cuda / Forward 0.000009984 s 0.00001136 s 0.88
Concat / DefOpt / cuda / Forward 0.000010015 s 0.000010944 s 0.92
Concat / IDefOpt / cuda / Forward 0.00000992 s 0.000011712 s 0.85
Concat / JaXPipe / cuda / PreRev 0.000016416 s 0.000018752000000000003 s 0.88
Concat / JaXPipe / cuda / PostRev 0.000016545 s 0.000018208 s 0.91
Concat / JaXPipe / cuda / BothRev 0.000016255999999999998 s 0.000018688 s 0.87
Concat / Jax / cuda / BothRev 0.00001664 s 0.00001856 s 0.90
Concat / HLOOpt / cuda / PreRev 0.000016255 s 0.000016704 s 0.97
Concat / HLOOpt / cuda / PostRev 0.000016255999999999998 s 0.000016927999999999998 s 0.96
Concat / HLOOpt / cuda / BothRev 0.000016128 s 0.000016864 s 0.96
Concat / PartOpt / cuda / PreRev 0.0000168 s 0.00001712 s 0.98
Concat / PartOpt / cuda / PostRev 0.000017632 s 0.000017151 s 1.03
Concat / PartOpt / cuda / BothRev 0.000016 s 0.000017216 s 0.93
Concat / IPartOpt / cuda / PreRev 0.000016127 s 0.000017024 s 0.95
Concat / IPartOpt / cuda / PostRev 0.00001616 s 0.00001712 s 0.94
Concat / IPartOpt / cuda / BothRev 0.000015872 s 0.000017056 s 0.93
Concat / DefOpt / cuda / PreRev 0.000016096 s 0.000017151 s 0.94
Concat / DefOpt / cuda / PostRev 0.000016224 s 0.000016927000000000002 s 0.96
Concat / DefOpt / cuda / BothRev 0.000016416 s 0.0000168 s 0.98
Concat / IDefOpt / cuda / PreRev 0.000016352 s 0.000016736 s 0.98
Concat / IDefOpt / cuda / PostRev 0.000015904000000000002 s 0.00001712 s 0.93
Concat / IDefOpt / cuda / BothRev 0.000016032 s 0.000016896000000000002 s 0.95
Concat / JaXPipe / cpu / Primal 0.00001601 s 0.000012834 s 1.25
Concat / Jax / cpu / Primal 0.00001617 s 0.000012809 s 1.26
Concat / HLOOpt / cpu / Primal 0.0000163 s 0.000013168 s 1.24
Concat / PartOpt / cpu / Primal 0.000016381 s 0.000013103 s 1.25
Concat / IPartOpt / cpu / Primal 0.000016062 s 0.000012935 s 1.24
Concat / DefOpt / cpu / Primal 0.00001619 s 0.000013538 s 1.20
Concat / IDefOpt / cpu / Primal 0.000016722 s 0.000012679 s 1.32
Concat / JaXPipe / cpu / Forward 0.000022536 s 0.000018003 s 1.25
Concat / Jax / cpu / Forward 0.000022376 s 0.000017769 s 1.26
Concat / HLOOpt / cpu / Forward 0.000023705 s 0.000017542999999999998 s 1.35
Concat / PartOpt / cpu / Forward 0.000021783 s 0.000017561 s 1.24
Concat / IPartOpt / cpu / Forward 0.000022378 s 0.000017515 s 1.28
Concat / DefOpt / cpu / Forward 0.00002213 s 0.0000176 s 1.26
Concat / IDefOpt / cpu / Forward 0.000022386 s 0.000017825 s 1.26
Concat / JaXPipe / cpu / PreRev 0.000025438 s 0.00002049 s 1.24
Concat / JaXPipe / cpu / PostRev 0.000025675 s 0.000019962 s 1.29
Concat / JaXPipe / cpu / BothRev 0.000025249 s 0.000019878 s 1.27
Concat / Jax / cpu / BothRev 0.000025144 s 0.000020003 s 1.26
Concat / HLOOpt / cpu / PreRev 0.000025239 s 0.000020007 s 1.26
Concat / HLOOpt / cpu / PostRev 0.000025772 s 0.000020165 s 1.28
Concat / HLOOpt / cpu / BothRev 0.000025274 s 0.000019592 s 1.29
Concat / PartOpt / cpu / PreRev 0.000024645 s 0.000020409 s 1.21
Concat / PartOpt / cpu / PostRev 0.000025331 s 0.00002033 s 1.25
Concat / PartOpt / cpu / BothRev 0.000025811 s 0.000019275 s 1.34
Concat / IPartOpt / cpu / PreRev 0.000025042 s 0.000019956 s 1.25
Concat / IPartOpt / cpu / PostRev 0.000024779 s 0.000019605 s 1.26
Concat / IPartOpt / cpu / BothRev 0.000024942 s 0.000019625 s 1.27
Concat / DefOpt / cpu / PreRev 0.00002553 s 0.000020967 s 1.22
Concat / DefOpt / cpu / PostRev 0.000025127 s 0.000019458 s 1.29
Concat / DefOpt / cpu / BothRev 0.000024911 s 0.00001977 s 1.26
Concat / IDefOpt / cpu / PreRev 0.000024969 s 0.000020008 s 1.25
Concat / IDefOpt / cpu / PostRev 0.00002478 s 0.000019651 s 1.26
Concat / IDefOpt / cpu / BothRev 0.000025152 s 0.000019793 s 1.27
const_scatter / JaXPipe / cpu / Primal 0.000007154420090955682 s 0.000012852 s 0.56
const_scatter / Jax / cpu / Primal 0.000006759899952157866 s 0.000012697 s 0.53
const_scatter / HLOOpt / cpu / Primal 0.000007399020050797844 s 0.000013442 s 0.55
const_scatter / PartOpt / cpu / Primal 0.000007071519958117278 s 0.000012651 s 0.56
const_scatter / IPartOpt / cpu / Primal 0.000007502479929826222 s 0.000012527 s 0.60
const_scatter / DefOpt / cpu / Primal 0.000007541679933638079 s 0.00001329 s 0.57
const_scatter / IDefOpt / cpu / Primal 0.000007414639949274715 s 0.000013248 s 0.56
const_scatter / JaXPipe / cpu / Forward 0.000011879740068252433 s 0.000018184 s 0.65
const_scatter / Jax / cpu / Forward 0.000010172320035053416 s 0.000016712999999999997 s 0.61
const_scatter / HLOOpt / cpu / Forward 0.000012081260010745609 s 0.000017864 s 0.68
const_scatter / PartOpt / cpu / Forward 0.000011645359936665044 s 0.000017628 s 0.66
const_scatter / IPartOpt / cpu / Forward 0.00001197507997858338 s 0.000017791 s 0.67
const_scatter / DefOpt / cpu / Forward 0.000011461919948487775 s 0.000018052 s 0.63
const_scatter / IDefOpt / cpu / Forward 0.000011299039961158997 s 0.000018058 s 0.63
const_scatter / JaXPipe / cpu / PreRev 0.0002902136800003 s 0.000494193 s 0.59
const_scatter / JaXPipe / cpu / PostRev 0.0002845714599243 s 0.000490383 s 0.58
const_scatter / JaXPipe / cpu / BothRev 0.0002870615399842 s 0.000515952 s 0.56
const_scatter / Jax / cpu / BothRev 0.000284687319945 s 0.000499809 s 0.57
const_scatter / HLOOpt / cpu / PreRev 0.0002851120000013 s 0.000506807 s 0.56
const_scatter / HLOOpt / cpu / PostRev 0.0002875938599936 s 0.000532947 s 0.54
const_scatter / HLOOpt / cpu / BothRev 0.0002868951599702 s 0.000492198 s 0.58
const_scatter / PartOpt / cpu / PreRev 0.0002844207199996 s 0.000513277 s 0.55
const_scatter / PartOpt / cpu / PostRev 0.0002857115599726 s 0.000495613 s 0.58
const_scatter / PartOpt / cpu / BothRev 0.0002854904398373 s 0.000528017 s 0.54
const_scatter / IPartOpt / cpu / PreRev 0.0002865611399101 s 0.000518134 s 0.55
const_scatter / IPartOpt / cpu / PostRev 0.0002861602600023 s 0.000499348 s 0.57
const_scatter / IPartOpt / cpu / BothRev 0.0002903588000481 s 0.000509812 s 0.57
const_scatter / DefOpt / cpu / PreRev 0.0002857440000661 s 0.000512476 s 0.56
const_scatter / DefOpt / cpu / PostRev 0.0002870006400189 s 0.000525034 s 0.55
const_scatter / DefOpt / cpu / BothRev 0.0002847741400182 s 0.000517393 s 0.55
const_scatter / IDefOpt / cpu / PreRev 0.0002877239400368 s 0.000538514 s 0.53
const_scatter / IDefOpt / cpu / PostRev 0.0002858229000412 s 0.000505837 s 0.57
const_scatter / IDefOpt / cpu / BothRev 0.0002864221199524 s 0.000512877 s 0.56
const_scatter / JaXPipe / cuda / Primal 0.000001887 s 0.000002463 s 0.77
const_scatter / Jax / cuda / Primal 0.000001887 s 0.000002433 s 0.78
const_scatter / HLOOpt / cuda / Primal 0.000001887 s 0.000002463 s 0.77
const_scatter / PartOpt / cuda / Primal 0.000001888 s 0.000002463 s 0.77
const_scatter / IPartOpt / cuda / Primal 0.000001887 s 0.000002463 s 0.77
const_scatter / DefOpt / cuda / Primal 0.000001887 s 0.000002463 s 0.77
const_scatter / IDefOpt / cuda / Primal 0.000001887 s 0.000002463 s 0.77
const_scatter / JaXPipe / cuda / Forward 0.000009888 s 0.00001104 s 0.90
const_scatter / Jax / cuda / Forward 0.00000992 s 0.000010272 s 0.97
const_scatter / HLOOpt / cuda / Forward 0.000009664 s 0.000010497 s 0.92
const_scatter / PartOpt / cuda / Forward 0.00000976 s 0.000011137 s 0.88
const_scatter / IPartOpt / cuda / Forward 0.0000096 s 0.000010656 s 0.90
const_scatter / DefOpt / cuda / Forward 0.00000992 s 0.000010656 s 0.93
const_scatter / IDefOpt / cuda / Forward 0.00001008 s 0.000011009 s 0.92
const_scatter / JaXPipe / cuda / PreRev 0.000015937 s 0.000017472 s 0.91
const_scatter / JaXPipe / cuda / PostRev 0.00001632 s 0.00001696 s 0.96
const_scatter / JaXPipe / cuda / BothRev 0.000015809 s 0.000016672 s 0.95
const_scatter / Jax / cuda / BothRev 0.00001632 s 0.00001712 s 0.95
const_scatter / HLOOpt / cuda / PreRev 0.000015935999999999998 s 0.000017184 s 0.93
const_scatter / HLOOpt / cuda / PostRev 0.000016351 s 0.00001696 s 0.96
const_scatter / HLOOpt / cuda / BothRev 0.00001584 s 0.000017408 s 0.91
const_scatter / PartOpt / cuda / PreRev 0.000016542999999999997 s 0.00001664 s 0.99
const_scatter / PartOpt / cuda / PostRev 0.000016383999999999998 s 0.0000168 s 0.98
const_scatter / PartOpt / cuda / BothRev 0.000016416 s 0.000016864 s 0.97
const_scatter / IPartOpt / cuda / PreRev 0.000016096 s 0.000016927999999999998 s 0.95
const_scatter / IPartOpt / cuda / PostRev 0.000016096 s 0.000016512 s 0.97
const_scatter / IPartOpt / cuda / BothRev 0.000016128 s 0.00001744 s 0.92
const_scatter / DefOpt / cuda / PreRev 0.000016768000000000003 s 0.000016958999999999998 s 0.99
const_scatter / DefOpt / cuda / PostRev 0.000015711 s 0.000017056 s 0.92
const_scatter / DefOpt / cuda / BothRev 0.000016705 s 0.000017312 s 0.96
const_scatter / IDefOpt / cuda / PreRev 0.000016511 s 0.000017152 s 0.96
const_scatter / IDefOpt / cuda / PostRev 0.000015584000000000002 s 0.000017344 s 0.90
const_scatter / IDefOpt / cuda / BothRev 0.00001616 s 0.000017088 s 0.95
const_scatter / JaXPipe / cpu / Primal 0.000016519 s 0.000012852 s 1.29
const_scatter / Jax / cpu / Primal 0.000016459000000000003 s 0.000012697 s 1.30
const_scatter / HLOOpt / cpu / Primal 0.000017311 s 0.000013442 s 1.29
const_scatter / PartOpt / cpu / Primal 0.000016421 s 0.000012651 s 1.30
const_scatter / IPartOpt / cpu / Primal 0.000016572 s 0.000012527 s 1.32
const_scatter / DefOpt / cpu / Primal 0.000017155 s 0.00001329 s 1.29
const_scatter / IDefOpt / cpu / Primal 0.000017225000000000002 s 0.000013248 s 1.30
const_scatter / JaXPipe / cpu / Forward 0.000022848 s 0.000018184 s 1.26
const_scatter / Jax / cpu / Forward 0.000021435 s 0.000016712999999999997 s 1.28
const_scatter / HLOOpt / cpu / Forward 0.000022498 s 0.000017864 s 1.26
const_scatter / PartOpt / cpu / Forward 0.000022697 s 0.000017628 s 1.29
const_scatter / IPartOpt / cpu / Forward 0.000022812 s 0.000017791 s 1.28
const_scatter / DefOpt / cpu / Forward 0.000023017 s 0.000018052 s 1.28
const_scatter / IDefOpt / cpu / Forward 0.000022727 s 0.000018058 s 1.26
const_scatter / JaXPipe / cpu / PreRev 0.000574196 s 0.000494193 s 1.16
const_scatter / JaXPipe / cpu / PostRev 0.000551594 s 0.000490383 s 1.12
const_scatter / JaXPipe / cpu / BothRev 0.00054812 s 0.000515952 s 1.06
const_scatter / Jax / cpu / BothRev 0.00055762 s 0.000499809 s 1.12
const_scatter / HLOOpt / cpu / PreRev 0.000563702 s 0.000506807 s 1.11
const_scatter / HLOOpt / cpu / PostRev 0.000553084 s 0.000532947 s 1.04
const_scatter / HLOOpt / cpu / BothRev 0.000566906 s 0.000492198 s 1.15
const_scatter / PartOpt / cpu / PreRev 0.000577253 s 0.000513277 s 1.12
const_scatter / PartOpt / cpu / PostRev 0.000541583 s 0.000495613 s 1.09
const_scatter / PartOpt / cpu / BothRev 0.000554452 s 0.000528017 s 1.05
const_scatter / IPartOpt / cpu / PreRev 0.0005387769999999 s 0.000518134 s 1.04
const_scatter / IPartOpt / cpu / PostRev 0.000540141 s 0.000499348 s 1.08
const_scatter / IPartOpt / cpu / BothRev 0.000546308 s 0.000509812 s 1.07
const_scatter / DefOpt / cpu / PreRev 0.0005743559999999 s 0.000512476 s 1.12
const_scatter / DefOpt / cpu / PostRev 0.000568489 s 0.000525034 s 1.08
const_scatter / DefOpt / cpu / BothRev 0.000555437 s 0.000517393 s 1.07
const_scatter / IDefOpt / cpu / PreRev 0.000562543 s 0.000538514 s 1.04
const_scatter / IDefOpt / cpu / PostRev 0.0005744529999999 s 0.000505837 s 1.14
const_scatter / IDefOpt / cpu / BothRev 0.000563271 s 0.000512877 s 1.10
GenDot / JaXPipe / cpu / Primal 0.000008655499950691592 s 0.000014693 s 0.59
GenDot / Jax / cpu / Primal 0.00000805137997303973 s 0.000014872 s 0.54
GenDot / HLOOpt / cpu / Primal 0.00000810146000731038 s 0.000013814 s 0.59
GenDot / PartOpt / cpu / Primal 0.000007599399978062138 s 0.00001455 s 0.52
GenDot / IPartOpt / cpu / Primal 0.000008302819987875409 s 0.000015255 s 0.54
GenDot / DefOpt / cpu / Primal 0.00000757250003516674 s 0.000014114 s 0.54
GenDot / IDefOpt / cpu / Primal 0.000007959999984450405 s 0.000014076 s 0.57
GenDot / JaXPipe / cpu / Forward 0.000011477960088086547 s 0.000019737 s 0.58
GenDot / Jax / cpu / Forward 0.000011042220048693709 s 0.000020213 s 0.55
GenDot / HLOOpt / cpu / Forward 0.00001179348004370695 s 0.000019072 s 0.62
GenDot / PartOpt / cpu / Forward 0.000011551900061022024 s 0.000018608 s 0.62
GenDot / IPartOpt / cpu / Forward 0.000012953839996043825 s 0.000019455 s 0.67
GenDot / DefOpt / cpu / Forward 0.000011589319983613676 s 0.000019162 s 0.60
GenDot / IDefOpt / cpu / Forward 0.000012002180028503062 s 0.000018916 s 0.63
GenDot / JaXPipe / cpu / PreRev 0.000011949440013268031 s 0.000019669 s 0.61
GenDot / JaXPipe / cpu / PostRev 0.000011416520028433296 s 0.000020344 s 0.56
GenDot / JaXPipe / cpu / BothRev 0.000012764299935952294 s 0.000019259 s 0.66
GenDot / Jax / cpu / BothRev 0.000011430320009822026 s 0.000020241 s 0.56
GenDot / HLOOpt / cpu / PreRev 0.000012257779999345076 s 0.000019136 s 0.64
GenDot / HLOOpt / cpu / PostRev 0.000014491279998765094 s 0.000019546 s 0.74
GenDot / HLOOpt / cpu / BothRev 0.000012335060109762707 s 0.000019856 s 0.62
GenDot / PartOpt / cpu / PreRev 0.000012051580106344772 s 0.000019271 s 0.63
GenDot / PartOpt / cpu / PostRev 0.000011537380032677902 s 0.000020379000000000003 s 0.57
GenDot / PartOpt / cpu / BothRev 0.00001303833998463233 s 0.000019498 s 0.67
GenDot / IPartOpt / cpu / PreRev 0.000012118520025978796 s 0.000018723 s 0.65
GenDot / IPartOpt / cpu / PostRev 0.00001158631997896009 s 0.000020888 s 0.55
GenDot / IPartOpt / cpu / BothRev 0.0000120522000179335 s 0.000019431 s 0.62
GenDot / DefOpt / cpu / PreRev 0.000012296419954509477 s 0.000019354 s 0.64
GenDot / DefOpt / cpu / PostRev 0.000012982240059500328 s 0.000019336 s 0.67
GenDot / DefOpt / cpu / BothRev 0.000012908679855172522 s 0.000019724 s 0.65
GenDot / IDefOpt / cpu / PreRev 0.000011804280020442092 s 0.000019382 s 0.61
GenDot / IDefOpt / cpu / PostRev 0.00001281500004552072 s 0.000019436 s 0.66
GenDot / IDefOpt / cpu / BothRev 0.000011647059945971703 s 0.000019075000000000003 s 0.61
GenDot / JaXPipe / cuda / Primal 0.000002016 s 0.000002527 s 0.80
GenDot / Jax / cuda / Primal 0.000002015 s 0.000002528 s 0.80
GenDot / HLOOpt / cuda / Primal 0.000002015 s 0.000002527 s 0.80
GenDot / PartOpt / cuda / Primal 0.000002016 s 0.000002528 s 0.80
GenDot / IPartOpt / cuda / Primal 0.000002015 s 0.00000256 s 0.79
GenDot / DefOpt / cuda / Primal 0.000001983 s 0.000002527 s 0.78
GenDot / IDefOpt / cuda / Primal 0.000002015 s 0.000002528 s 0.80
GenDot / JaXPipe / cuda / Forward 0.000010016 s 0.00001072 s 0.93
GenDot / Jax / cuda / Forward 0.000009632 s 0.000010624 s 0.91
GenDot / HLOOpt / cuda / Forward 0.000009856 s 0.000010591 s 0.93
GenDot / PartOpt / cuda / Forward 0.000009856 s 0.000010464 s 0.94
GenDot / IPartOpt / cuda / Forward 0.000009856 s 0.000010911 s 0.90
GenDot / DefOpt / cuda / Forward 0.000009632 s 0.000010816 s 0.89
GenDot / IDefOpt / cuda / Forward 0.00000976 s 0.000010816 s 0.90
GenDot / JaXPipe / cuda / PreRev 0.000010144 s 0.000010816 s 0.94
GenDot / JaXPipe / cuda / PostRev 0.000009569 s 0.000010752 s 0.89
GenDot / JaXPipe / cuda / BothRev 0.000009728 s 0.000010528 s 0.92
GenDot / Jax / cuda / BothRev 0.000009632 s 0.000015007 s 0.64
GenDot / HLOOpt / cuda / PreRev 0.000009664 s 0.000010528 s 0.92
GenDot / HLOOpt / cuda / PostRev 0.000010048 s 0.000010656 s 0.94
GenDot / HLOOpt / cuda / BothRev 0.00000976 s 0.000010336 s 0.94
GenDot / PartOpt / cuda / PreRev 0.000009855 s 0.000010624 s 0.93
GenDot / PartOpt / cuda / PostRev 0.000009696 s 0.000010496 s 0.92
GenDot / PartOpt / cuda / BothRev 0.000009984 s 0.000010945 s 0.91
GenDot / IPartOpt / cuda / PreRev 0.000009728 s 0.000010495 s 0.93
GenDot / IPartOpt / cuda / PostRev 0.000009568 s 0.000010368 s 0.92
GenDot / IPartOpt / cuda / BothRev 0.000009889 s 0.00001056 s 0.94
GenDot / DefOpt / cuda / PreRev 0.000009663 s 0.000011232 s 0.86
GenDot / DefOpt / cuda / PostRev 0.000009952 s 0.00001056 s 0.94
GenDot / DefOpt / cuda / BothRev 0.00001008 s 0.000010688 s 0.94
GenDot / IDefOpt / cuda / PreRev 0.000010112 s 0.000010752 s 0.94
GenDot / IDefOpt / cuda / PostRev 0.0000096 s 0.000011168 s 0.86
GenDot / IDefOpt / cuda / BothRev 0.000010016 s 0.000011327 s 0.88
GenDot / JaXPipe / cpu / Primal 0.000019408 s 0.000014693 s 1.32
GenDot / Jax / cpu / Primal 0.000019218 s 0.000014872 s 1.29
GenDot / HLOOpt / cpu / Primal 0.000017813 s 0.000013814 s 1.29
GenDot / PartOpt / cpu / Primal 0.000019042 s 0.00001455 s 1.31
GenDot / IPartOpt / cpu / Primal 0.000018721000000000003 s 0.000015255 s 1.23
GenDot / DefOpt / cpu / Primal 0.000029979 s 0.000014114 s 2.12
GenDot / IDefOpt / cpu / Primal 0.00001834 s 0.000014076 s 1.30
GenDot / JaXPipe / cpu / Forward 0.000024669 s 0.000019737 s 1.25
GenDot / Jax / cpu / Forward 0.000026454 s 0.000020213 s 1.31
GenDot / HLOOpt / cpu / Forward 0.000024443 s 0.000019072 s 1.28
GenDot / PartOpt / cpu / Forward 0.000024813 s 0.000018608 s 1.33
GenDot / IPartOpt / cpu / Forward 0.000024456 s 0.000019455 s 1.26
GenDot / DefOpt / cpu / Forward 0.000024638 s 0.000019162 s 1.29
GenDot / IDefOpt / cpu / Forward 0.00002376 s 0.000018916 s 1.26
GenDot / JaXPipe / cpu / PreRev 0.000024978 s 0.000019669 s 1.27
GenDot / JaXPipe / cpu / PostRev 0.000025587 s 0.000020344 s 1.26
GenDot / JaXPipe / cpu / BothRev 0.000024689 s 0.000019259 s 1.28
GenDot / Jax / cpu / BothRev 0.000026965 s 0.000020241 s 1.33
GenDot / HLOOpt / cpu / PreRev 0.000025357 s 0.000019136 s 1.33
GenDot / HLOOpt / cpu / PostRev 0.000024891 s 0.000019546 s 1.27
GenDot / HLOOpt / cpu / BothRev 0.00002541 s 0.000019856 s 1.28
GenDot / PartOpt / cpu / PreRev 0.000026054 s 0.000019271 s 1.35
GenDot / PartOpt / cpu / PostRev 0.000026357 s 0.000020379000000000003 s 1.29
GenDot / PartOpt / cpu / BothRev 0.000026232 s 0.000019498 s 1.35
GenDot / IPartOpt / cpu / PreRev 0.000024981 s 0.000018723 s 1.33
GenDot / IPartOpt / cpu / PostRev 0.000025918 s 0.000020888 s 1.24
GenDot / IPartOpt / cpu / BothRev 0.000030182 s 0.000019431 s 1.55
GenDot / DefOpt / cpu / PreRev 0.000024657 s 0.000019354 s 1.27
GenDot / DefOpt / cpu / PostRev 0.00002571 s 0.000019336 s 1.33
GenDot / DefOpt / cpu / BothRev 0.000025181 s 0.000019724 s 1.28
GenDot / IDefOpt / cpu / PreRev 0.000024852 s 0.000019382 s 1.28
GenDot / IDefOpt / cpu / PostRev 0.000024769 s 0.000019436 s 1.27
GenDot / IDefOpt / cpu / BothRev 0.000024708 s 0.000019075000000000003 s 1.30
hlo_ffi / JaXPipe / cpu / Primal 0.00001026426003591041 s 0.000017462 s 0.59
hlo_ffi / Jax / cpu / Primal 0.000009718339988467053 s 0.000017542999999999998 s 0.55
hlo_ffi / HLOOpt / cpu / Primal 0.000009967559944925596 s 0.000017298 s 0.58
hlo_ffi / PartOpt / cpu / Primal 0.0000092871000742889 s 0.000017353 s 0.54
hlo_ffi / IPartOpt / cpu / Primal 0.000009971979961846957 s 0.000017631 s 0.57
hlo_ffi / DefOpt / cpu / Primal 0.000009495319973211736 s 0.000017361999999999997 s 0.55
hlo_ffi / IDefOpt / cpu / Primal 0.00000944432005780982 s 0.000017690999999999997 s 0.53
hlo_ffi / JaXPipe / cpu / Forward 0.0000139294599830464 s 0.000024507 s 0.57
hlo_ffi / Jax / cpu / Forward 0.0000135791999855428 s 0.00002347 s 0.58
hlo_ffi / HLOOpt / cpu / Forward 0.000013622679925902049 s 0.000023905 s 0.57
hlo_ffi / PartOpt / cpu / Forward 0.000013768179960607086 s 0.000023943 s 0.58
hlo_ffi / IPartOpt / cpu / Forward 0.000013878459976695012 s 0.000023503 s 0.59
hlo_ffi / DefOpt / cpu / Forward 0.00001381172005494591 s 0.000023983 s 0.58
hlo_ffi / IDefOpt / cpu / Forward 0.000013870379989384674 s 0.000023725 s 0.58
hlo_ffi / JaXPipe / cpu / PreRev 0.000014068299969949294 s 0.000024384 s 0.58
hlo_ffi / JaXPipe / cpu / PostRev 0.000014098520096013087 s 0.000023532 s 0.60
hlo_ffi / JaXPipe / cpu / BothRev 0.000014235319977160545 s 0.000023683 s 0.60
hlo_ffi / Jax / cpu / BothRev 0.000013988200025778497 s 0.000024125 s 0.58
hlo_ffi / HLOOpt / cpu / PreRev 0.000014236480037652654 s 0.000023644 s 0.60
hlo_ffi / HLOOpt / cpu / PostRev 0.000016109660027723293 s 0.000023644 s 0.68
hlo_ffi / HLOOpt / cpu / BothRev 0.000014328660054161446 s 0.00002379 s 0.60
hlo_ffi / PartOpt / cpu / PreRev 0.000013927000036346728 s 0.000023865 s 0.58
hlo_ffi / PartOpt / cpu / PostRev 0.000014315100033854832 s 0.0000235 s 0.61
hlo_ffi / PartOpt / cpu / BothRev 0.00001412080011505168 s 0.000024603 s 0.57
hlo_ffi / IPartOpt / cpu / PreRev 0.000014175540090946017 s 0.000023755 s 0.60
hlo_ffi / IPartOpt / cpu / PostRev 0.000013966919996164506 s 0.000023591 s 0.59
hlo_ffi / IPartOpt / cpu / BothRev 0.000013886939977965084 s 0.000023426 s 0.59
hlo_ffi / DefOpt / cpu / PreRev 0.000013985340083308984 s 0.00002395 s 0.58
hlo_ffi / DefOpt / cpu / PostRev 0.000013873640018573496 s 0.000023763 s 0.58
hlo_ffi / DefOpt / cpu / BothRev 0.0000140299399572541 s 0.000023529 s 0.60
hlo_ffi / IDefOpt / cpu / PreRev 0.000014588140038540586 s 0.000023718 s 0.62
hlo_ffi / IDefOpt / cpu / PostRev 0.000014058259985176848 s 0.000023959 s 0.59
hlo_ffi / IDefOpt / cpu / BothRev 0.000014363659975060728 s 0.000024330000000000003 s 0.59
hlo_ffi / JaXPipe / cuda / Primal 0.000001984 s 0.0000023670000000000004 s 0.84
hlo_ffi / Jax / cuda / Primal 0.000001984 s 0.000002368 s 0.84
hlo_ffi / HLOOpt / cuda / Primal 0.000001984 s 0.0000023670000000000004 s 0.84
hlo_ffi / PartOpt / cuda / Primal 0.000001984 s 0.000002368 s 0.84
hlo_ffi / IPartOpt / cuda / Primal 0.000001983 s 0.0000023670000000000004 s 0.84
hlo_ffi / DefOpt / cuda / Primal 0.000001984 s 0.000002368 s 0.84
hlo_ffi / IDefOpt / cuda / Primal 0.000001983 s 0.0000023670000000000004 s 0.84
hlo_ffi / JaXPipe / cuda / Forward 0.00000208 s 0.000002463 s 0.84
hlo_ffi / Jax / cuda / Forward 0.00000208 s 0.000002464 s 0.84
hlo_ffi / HLOOpt / cuda / Forward 0.00000208 s 0.000002463 s 0.84
hlo_ffi / PartOpt / cuda / Forward 0.00000208 s 0.000002463 s 0.84
hlo_ffi / IPartOpt / cuda / Forward 0.00000208 s 0.000002463 s 0.84
hlo_ffi / DefOpt / cuda / Forward 0.00000208 s 0.000002464 s 0.84
hlo_ffi / IDefOpt / cuda / Forward 0.00000208 s 0.000002463 s 0.84
hlo_ffi / JaXPipe / cuda / PreRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / JaXPipe / cuda / PostRev 0.000002047 s 0.000002432 s 0.84
hlo_ffi / JaXPipe / cuda / BothRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / Jax / cuda / BothRev 0.000002048 s 0.000002463 s 0.83
hlo_ffi / HLOOpt / cuda / PreRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / HLOOpt / cuda / PostRev 0.000002047 s 0.000002433 s 0.84
hlo_ffi / HLOOpt / cuda / BothRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / PartOpt / cuda / PreRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / PartOpt / cuda / PostRev 0.000002048 s 0.000002431 s 0.84
hlo_ffi / PartOpt / cuda / BothRev 0.000002047 s 0.000002433 s 0.84
hlo_ffi / IPartOpt / cuda / PreRev 0.000002048 s 0.000002431 s 0.84
hlo_ffi / IPartOpt / cuda / PostRev 0.000002048 s 0.000002463 s 0.83
hlo_ffi / IPartOpt / cuda / BothRev 0.000002047 s 0.000002432 s 0.84
hlo_ffi / DefOpt / cuda / PreRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / DefOpt / cuda / PostRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / DefOpt / cuda / BothRev 0.000002047 s 0.000002432 s 0.84
hlo_ffi / IDefOpt / cuda / PreRev 0.000002048 s 0.000002432 s 0.84
hlo_ffi / IDefOpt / cuda / PostRev 0.000002047 s 0.000002432 s 0.84
hlo_ffi / IDefOpt / cuda / BothRev 0.000002047 s 0.000002463 s 0.83
hlo_ffi / JaXPipe / cpu / Primal 0.000022758 s 0.000017462 s 1.30
hlo_ffi / Jax / cpu / Primal 0.000021705 s 0.000017542999999999998 s 1.24
hlo_ffi / HLOOpt / cpu / Primal 0.000021358 s 0.000017298 s 1.23
hlo_ffi / PartOpt / cpu / Primal 0.000021739 s 0.000017353 s 1.25
hlo_ffi / IPartOpt / cpu / Primal 0.000022104 s 0.000017631 s 1.25
hlo_ffi / DefOpt / cpu / Primal 0.000021974 s 0.000017361999999999997 s 1.27
hlo_ffi / IDefOpt / cpu / Primal 0.000021558 s 0.000017690999999999997 s 1.22
hlo_ffi / JaXPipe / cpu / Forward 0.00003027 s 0.000024507 s 1.24
hlo_ffi / Jax / cpu / Forward 0.000029797 s 0.00002347 s 1.27
hlo_ffi / HLOOpt / cpu / Forward 0.000029703 s 0.000023905 s 1.24
hlo_ffi / PartOpt / cpu / Forward 0.000029986 s 0.000023943 s 1.25
hlo_ffi / IPartOpt / cpu / Forward 0.000030212 s 0.000023503 s 1.29
hlo_ffi / DefOpt / cpu / Forward 0.000028995 s 0.000023983 s 1.21
hlo_ffi / IDefOpt / cpu / Forward 0.000029699 s 0.000023725 s 1.25
hlo_ffi / JaXPipe / cpu / PreRev 0.000029948 s 0.000024384 s 1.23
hlo_ffi / JaXPipe / cpu / PostRev 0.000029682 s 0.000023532 s 1.26
hlo_ffi / JaXPipe / cpu / BothRev 0.000029163 s 0.000023683 s 1.23
hlo_ffi / Jax / cpu / BothRev 0.000030212 s 0.000024125 s 1.25
hlo_ffi / HLOOpt / cpu / PreRev 0.000030195 s 0.000023644 s 1.28
hlo_ffi / HLOOpt / cpu / PostRev 0.000029494 s 0.000023644 s 1.25
hlo_ffi / HLOOpt / cpu / BothRev 0.000030975 s 0.00002379 s 1.30
hlo_ffi / PartOpt / cpu / PreRev 0.000030471 s 0.000023865 s 1.28
hlo_ffi / PartOpt / cpu / PostRev 0.000036754 s 0.0000235 s 1.56
hlo_ffi / PartOpt / cpu / BothRev 0.00003115 s 0.000024603 s 1.27
hlo_ffi / IPartOpt / cpu / PreRev 0.000030224 s 0.000023755 s 1.27
hlo_ffi / IPartOpt / cpu / PostRev 0.000030987 s 0.000023591 s 1.31
hlo_ffi / IPartOpt / cpu / BothRev 0.000029766 s 0.000023426 s 1.27
hlo_ffi / DefOpt / cpu / PreRev 0.000029992 s 0.00002395 s 1.25
hlo_ffi / DefOpt / cpu / PostRev 0.00003044 s 0.000023763 s 1.28
hlo_ffi / DefOpt / cpu / BothRev 0.000030658000000000004 s 0.000023529 s 1.30
hlo_ffi / IDefOpt / cpu / PreRev 0.000031735 s 0.000023718 s 1.34
hlo_ffi / IDefOpt / cpu / PostRev 0.000031658 s 0.000023959 s 1.32
hlo_ffi / IDefOpt / cpu / BothRev 0.000030994 s 0.000024330000000000003 s 1.27
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.0008948130001954 s 0.001763568 s 0.51
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Primal 0.0009404604001247 s 0.001485087 s 0.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Primal 0.0009444160001294 s 0.001712441 s 0.55
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.0008793189999778 s 0.0015765599999999 s 0.56
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Primal 0.0008898544001567 s 0.0015197989999999 s 0.59
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Primal 0.0009432273998754 s 0.0017493749999999 s 0.54
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.0009511481999652 s 0.001603363 s 0.59
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Forward 0.00217659000009 s 0.00437569 s 0.50
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.0023137436000979 s 0.004520619 s 0.51
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Forward 0.0022392284001398 s 0.004573614 s 0.49
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.0021778880001875 s 0.004502652 s 0.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Forward 0.0022469533996627 s 0.004387929 s 0.51
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.0021796589997393 s 0.004582755 s 0.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Forward 0.002219416600019 s 0.005122053 s 0.43
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PreRev 0.0051047533999735 s 0.007972802 s 0.64
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.0061896753997643 s 0.007790199 s 0.79
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / BothRev 0.0061058772000251 s 0.007701489 s 0.79
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / BothRev 0.0038429115998951 s 0.008807338 s 0.44
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PreRev 0.0057168687999364 s 0.008201033 s 0.70
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.0053523091999522 s 0.006543085 s 0.82
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.0056836800000382 s 0.009350031 s 0.61
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PreRev 0.0035767370001849 s 0.0074474669999999 s 0.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PostRev 0.0058589523998307 s 0.009056619 s 0.65
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.0035170327999367 s 0.007933135 s 0.44
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.0056425257998853 s 0.008086585 s 0.70
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PostRev 0.0050290429999222 s 0.007406945 s 0.68
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / BothRev 0.0056632367997735 s 0.009752857 s 0.58
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PreRev 0.0036136127997451 s 0.007572427 s 0.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PostRev 0.0055444097999497 s 0.008630254 s 0.64
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / BothRev 0.0049569376000363 s 0.007841915 s 0.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PreRev 0.0057989885999631 s 0.008346618 s 0.69
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PostRev 0.0035721701999136 s 0.007180154 s 0.50
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.0058907759999783 s 0.007490605 s 0.79
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / Primal 0.000281664 s 0.0002993589999999 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / Primal 0.000280704 s 0.000299712 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Primal 0.000288737 s 0.000306464 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / Primal 0.000280768 s 0.00029968 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Primal 0.0002814079999999 s 0.000298464 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Primal 0.00028848 s 0.0003068159999999 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / Primal 0.000288671 s 0.000306144 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / Forward 0.000557888 s 0.000584607 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / Forward 0.0005401589999999 s 0.000567936 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Forward 0.000558048 s 0.000583904 s 0.96
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / Forward 0.0005573759999999 s 0.000583296 s 0.96
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Forward 0.000559104 s 0.000584128 s 0.96
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Forward 0.0005572169999999 s 0.000583488 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / Forward 0.000557408 s 0.00058448 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / PreRev 0.00102784 s 0.001060191 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / PostRev 0.000987744 s 0.001013759 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / BothRev 0.001027744 s 0.0010551669999999 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / BothRev 0.000990144 s 0.001011808 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PreRev 0.001016385 s 0.001040191 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PostRev 0.001041184 s 0.00106496 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / BothRev 0.001015041 s 0.001041535 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PreRev 0.001030913 s 0.001055135 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PostRev 0.00097824 s 0.001001472 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / BothRev 0.001029536 s 0.001056832 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PreRev 0.001028545 s 0.001054911 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PostRev 0.000979456 s 0.001002655 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / BothRev 0.001029376 s 0.001056224 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / PreRev 0.001024928 s 0.001055168 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / PostRev 0.000963008 s 0.000989984 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / BothRev 0.001026497 s 0.001056703 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PreRev 0.001024736 s 0.001056032 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PostRev 0.00102288 s 0.001059072 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / BothRev 0.001024737 s 0.001057055 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.003871581 s 0.001763568 s 2.20
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Primal 0.0039086929999999 s 0.001485087 s 2.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Primal 0.0039098579999999 s 0.001712441 s 2.28
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.0042190349999999 s 0.0015765599999999 s 2.68
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Primal 0.003862298 s 0.0015197989999999 s 2.54
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Primal 0.003975592 s 0.0017493749999999 s 2.27
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.004404758 s 0.001603363 s 2.75
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Forward 0.009305477 s 0.00437569 s 2.13
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.009031173 s 0.004520619 s 2.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Forward 0.009220028 s 0.004573614 s 2.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.008917317 s 0.004502652 s 1.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Forward 0.008532958 s 0.004387929 s 1.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.009110337 s 0.004582755 s 1.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Forward 0.009191176 s 0.005122053 s 1.79
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PreRev 0.014379832 s 0.007972802 s 1.80
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.012626419 s 0.007790199 s 1.62
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / BothRev 0.011883875 s 0.007701489 s 1.54
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / BothRev 0.013049146 s 0.008807338 s 1.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PreRev 0.0117612439999999 s 0.008201033 s 1.43
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.012597004 s 0.006543085 s 1.93
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.012250112 s 0.009350031 s 1.31
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PreRev 0.012735262 s 0.0074474669999999 s 1.71
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PostRev 0.0127782319999999 s 0.009056619 s 1.41
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.011520987 s 0.007933135 s 1.45
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.0122450069999999 s 0.008086585 s 1.51
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PostRev 0.012909061 s 0.007406945 s 1.74
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / BothRev 0.011932401 s 0.009752857 s 1.22
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PreRev 0.012230362 s 0.007572427 s 1.62
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PostRev 0.011278726 s 0.008630254 s 1.31
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / BothRev 0.013528509 s 0.007841915 s 1.73
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PreRev 0.013437691 s 0.008346618 s 1.61
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PostRev 0.012209785 s 0.007180154 s 1.70
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.011579109 s 0.007490605 s 1.55
scatter_sum / JaXPipe / cpu / Primal 0.000009154540002782596 s 0.00001584 s 0.58
scatter_sum / Jax / cpu / Primal 0.000009623459991416894 s 0.000015703000000000002 s 0.61
scatter_sum / HLOOpt / cpu / Primal 0.000008490639938827371 s 0.000015737999999999997 s 0.54
scatter_sum / PartOpt / cpu / Primal 0.000008398960089834872 s 0.000015625 s 0.54
scatter_sum / IPartOpt / cpu / Primal 0.00000821985991933616 s 0.000015793000000000003 s 0.52
scatter_sum / DefOpt / cpu / Primal 0.000008579280020057923 s 0.000015398 s 0.56
scatter_sum / IDefOpt / cpu / Primal 0.000008555859985790447 s 0.000015412 s 0.56
scatter_sum / JaXPipe / cpu / Forward 0.00001269419997697696 s 0.000022844 s 0.56
scatter_sum / Jax / cpu / Forward 0.000012817679980798855 s 0.000022334 s 0.57
scatter_sum / HLOOpt / cpu / Forward 0.000013275460059958277 s 0.000022719 s 0.58
scatter_sum / PartOpt / cpu / Forward 0.000012351099958323176 s 0.000022468 s 0.55
scatter_sum / IPartOpt / cpu / Forward 0.0000129965800078935 s 0.000022597 s 0.58
scatter_sum / DefOpt / cpu / Forward 0.000012573780113598331 s 0.000022084 s 0.57
scatter_sum / IDefOpt / cpu / Forward 0.000012421560004440837 s 0.000038063 s 0.33
scatter_sum / JaXPipe / cpu / PreRev 0.00001290754007641226 s 0.000023225 s 0.56
scatter_sum / JaXPipe / cpu / PostRev 0.000013031940034125 s 0.000022668 s 0.57
scatter_sum / JaXPipe / cpu / BothRev 0.00001383471997542074 s 0.000022202 s 0.62
scatter_sum / Jax / cpu / BothRev 0.000012943959973199526 s 0.000022582 s 0.57
scatter_sum / HLOOpt / cpu / PreRev 0.00001327399999354384 s 0.000022679 s 0.59
scatter_sum / HLOOpt / cpu / PostRev 0.000015600480037392116 s 0.000022522 s 0.69
scatter_sum / HLOOpt / cpu / BothRev 0.000013094819951220416 s 0.000022435 s 0.58
scatter_sum / PartOpt / cpu / PreRev 0.000013007339966861763 s 0.000022439 s 0.58
scatter_sum / PartOpt / cpu / PostRev 0.000013879980015190084 s 0.000022575 s 0.61
scatter_sum / PartOpt / cpu / BothRev 0.000014095720016484848 s 0.000022569 s 0.62
scatter_sum / IPartOpt / cpu / PreRev 0.000013099059942760504 s 0.00002262 s 0.58
scatter_sum / IPartOpt / cpu / PostRev 0.000013950880002084886 s 0.0000223 s 0.63
scatter_sum / IPartOpt / cpu / BothRev 0.00001307709997490747 s 0.00002268 s 0.58
scatter_sum / DefOpt / cpu / PreRev 0.000013517839961423306 s 0.000022839000000000003 s 0.59
scatter_sum / DefOpt / cpu / PostRev 0.00001339279995590914 s 0.000022353 s 0.60
scatter_sum / DefOpt / cpu / BothRev 0.00001310420006120694 s 0.000022472 s 0.58
scatter_sum / IDefOpt / cpu / PreRev 0.000013152019946574 s 0.000022360000000000003 s 0.59
scatter_sum / IDefOpt / cpu / PostRev 0.000013564219989348204 s 0.000021894 s 0.62
scatter_sum / IDefOpt / cpu / BothRev 0.000012663240067922742 s 0.000023007 s 0.55
scatter_sum / JaXPipe / cuda / Primal 0.000009953 s 0.000011648 s 0.85
scatter_sum / Jax / cuda / Primal 0.000009888 s 0.000010816 s 0.91
scatter_sum / HLOOpt / cuda / Primal 0.000009984 s 0.000010784 s 0.93
scatter_sum / PartOpt / cuda / Primal 0.000009792 s 0.00001088 s 0.90
scatter_sum / IPartOpt / cuda / Primal 0.000010303 s 0.000010913 s 0.94
scatter_sum / DefOpt / cuda / Primal 0.000010336 s 0.000011072 s 0.93
scatter_sum / IDefOpt / cuda / Primal 0.00001008 s 0.00001104 s 0.91
scatter_sum / JaXPipe / cuda / Forward 0.00001712 s 0.000018304 s 0.94
scatter_sum / Jax / cuda / Forward 0.000016288 s 0.000016414999999999998 s 0.99
scatter_sum / HLOOpt / cuda / Forward 0.00001712 s 0.000017505 s 0.98
scatter_sum / PartOpt / cuda / Forward 0.000017056 s 0.000017313 s 0.99
scatter_sum / IPartOpt / cuda / Forward 0.00001696 s 0.0000176 s 0.96
scatter_sum / DefOpt / cuda / Forward 0.000017503999999999997 s 0.000017696 s 0.99
scatter_sum / IDefOpt / cuda / Forward 0.000016896000000000002 s 0.000017183 s 0.98
scatter_sum / JaXPipe / cuda / PreRev 0.000016864 s 0.00001728 s 0.98
scatter_sum / JaXPipe / cuda / PostRev 0.000016383999999999998 s 0.000017184 s 0.95
scatter_sum / JaXPipe / cuda / BothRev 0.0000168 s 0.000017824 s 0.94
scatter_sum / Jax / cuda / BothRev 0.000016927999999999998 s 0.000017568000000000002 s 0.96
scatter_sum / HLOOpt / cuda / PreRev 0.000016992 s 0.000018176 s 0.93
scatter_sum / HLOOpt / cuda / PostRev 0.00001712 s 0.000017536 s 0.98
scatter_sum / HLOOpt / cuda / BothRev 0.000017375999999999998 s 0.000017664 s 0.98
scatter_sum / PartOpt / cuda / PreRev 0.000017056 s 0.000018016 s 0.95
scatter_sum / PartOpt / cuda / PostRev 0.000016768000000000003 s 0.000017536 s 0.96
scatter_sum / PartOpt / cuda / BothRev 0.000017184 s 0.000018207 s 0.94
scatter_sum / IPartOpt / cuda / PreRev 0.00001696 s 0.000017919999999999998 s 0.95
scatter_sum / IPartOpt / cuda / PostRev 0.000016576000000000002 s 0.000016927999999999998 s 0.98
scatter_sum / IPartOpt / cuda / BothRev 0.000016544 s 0.000017824 s 0.93
scatter_sum / DefOpt / cuda / PreRev 0.000016288 s 0.000017952 s 0.91
scatter_sum / DefOpt / cuda / PostRev 0.000017375999999999998 s 0.000016992 s 1.02
scatter_sum / DefOpt / cuda / BothRev 0.000016672 s 0.000017088 s 0.98
scatter_sum / IDefOpt / cuda / PreRev 0.000016545 s 0.0000176 s 0.94
scatter_sum / IDefOpt / cuda / PostRev 0.000016672 s 0.000017408 s 0.96
scatter_sum / IDefOpt / cuda / BothRev 0.000017024 s 0.000017728 s 0.96
scatter_sum / JaXPipe / cpu / Primal 0.00001984 s 0.00001584 s 1.25
scatter_sum / Jax / cpu / Primal 0.000032305 s 0.000015703000000000002 s 2.06
scatter_sum / HLOOpt / cpu / Primal 0.000019872 s 0.000015737999999999997 s 1.26
scatter_sum / PartOpt / cpu / Primal 0.000019809 s 0.000015625 s 1.27
scatter_sum / IPartOpt / cpu / Primal 0.000020168 s 0.000015793000000000003 s 1.28
scatter_sum / DefOpt / cpu / Primal 0.00002545 s 0.000015398 s 1.65
scatter_sum / IDefOpt / cpu / Primal 0.000019842 s 0.000015412 s 1.29
scatter_sum / JaXPipe / cpu / Forward 0.000029514 s 0.000022844 s 1.29
scatter_sum / Jax / cpu / Forward 0.000029559 s 0.000022334 s 1.32
scatter_sum / HLOOpt / cpu / Forward 0.000029877 s 0.000022719 s 1.32
scatter_sum / PartOpt / cpu / Forward 0.000028278000000000003 s 0.000022468 s 1.26
scatter_sum / IPartOpt / cpu / Forward 0.000028307 s 0.000022597 s 1.25
scatter_sum / DefOpt / cpu / Forward 0.000028855 s 0.000022084 s 1.31
scatter_sum / IDefOpt / cpu / Forward 0.000028536 s 0.000038063 s 0.75
scatter_sum / JaXPipe / cpu / PreRev 0.000029477 s 0.000023225 s 1.27
scatter_sum / JaXPipe / cpu / PostRev 0.000028519 s 0.000022668 s 1.26
scatter_sum / JaXPipe / cpu / BothRev 0.000029352 s 0.000022202 s 1.32
scatter_sum / Jax / cpu / BothRev 0.000028989 s 0.000022582 s 1.28
scatter_sum / HLOOpt / cpu / PreRev 0.000029257 s 0.000022679 s 1.29
scatter_sum / HLOOpt / cpu / PostRev 0.000029472 s 0.000022522 s 1.31
scatter_sum / HLOOpt / cpu / BothRev 0.000029047 s 0.000022435 s 1.29
scatter_sum / PartOpt / cpu / PreRev 0.000030898 s 0.000022439 s 1.38
scatter_sum / PartOpt / cpu / PostRev 0.000029493 s 0.000022575 s 1.31
scatter_sum / PartOpt / cpu / BothRev 0.000029687 s 0.000022569 s 1.32
scatter_sum / IPartOpt / cpu / PreRev 0.000030404 s 0.00002262 s 1.34
scatter_sum / IPartOpt / cpu / PostRev 0.000029498 s 0.0000223 s 1.32
scatter_sum / IPartOpt / cpu / BothRev 0.000029419 s 0.00002268 s 1.30
scatter_sum / DefOpt / cpu / PreRev 0.000035113 s 0.000022839000000000003 s 1.54
scatter_sum / DefOpt / cpu / PostRev 0.000029329 s 0.000022353 s 1.31
scatter_sum / DefOpt / cpu / BothRev 0.000028844 s 0.000022472 s 1.28
scatter_sum / IDefOpt / cpu / PreRev 0.000028152 s 0.000022360000000000003 s 1.26
scatter_sum / IDefOpt / cpu / PostRev 0.000029448 s 0.000021894 s 1.35
scatter_sum / IDefOpt / cpu / BothRev 0.000028675 s 0.000023007 s 1.25
slicing / JaXPipe / cpu / Primal 0.000006988280038058292 s 0.000013153 s 0.53
slicing / Jax / cpu / Primal 0.000007066620037221583 s 0.000012428 s 0.57
slicing / HLOOpt / cpu / Primal 0.000006935719939065166 s 0.000012639 s 0.55
slicing / PartOpt / cpu / Primal 0.000006532999923365423 s 0.000012471 s 0.52
slicing / IPartOpt / cpu / Primal 0.00000698045998433372 s 0.000012629 s 0.55
slicing / DefOpt / cpu / Primal 0.000007035759972495726 s 0.000012661 s 0.56
slicing / IDefOpt / cpu / Primal 0.000006859160075691761 s 0.000012617 s 0.54
slicing / JaXPipe / cpu / Forward 0.0000106280599902675 s 0.000016938999999999998 s 0.63
slicing / Jax / cpu / Forward 0.000010632459998305422 s 0.000016388 s 0.65
slicing / HLOOpt / cpu / Forward 0.000010340820044802968 s 0.000016751999999999998 s 0.62
slicing / PartOpt / cpu / Forward 0.000010007799955928933 s 0.000016729000000000002 s 0.60
slicing / IPartOpt / cpu / Forward 0.000010841939983947668 s 0.000016976 s 0.64
slicing / DefOpt / cpu / Forward 0.00001065325992385624 s 0.000016904 s 0.63
slicing / IDefOpt / cpu / Forward 0.000010414560019853525 s 0.000017279 s 0.60
slicing / JaXPipe / cpu / PreRev 0.000011049479981011246 s 0.000017972 s 0.61
slicing / JaXPipe / cpu / PostRev 0.000011307579879940022 s 0.000017929 s 0.63
slicing / JaXPipe / cpu / BothRev 0.000011088060036854583 s 0.000017579000000000002 s 0.63
slicing / Jax / cpu / BothRev 0.000011041820071113762 s 0.000017622 s 0.63
slicing / HLOOpt / cpu / PreRev 0.000011057679930672748 s 0.000017704999999999997 s 0.62
slicing / HLOOpt / cpu / PostRev 0.000012811960023100256 s 0.000017486 s 0.73
slicing / HLOOpt / cpu / BothRev 0.000010890100093092767 s 0.000017561 s 0.62
slicing / PartOpt / cpu / PreRev 0.00001082938002582523 s 0.000017477 s 0.62
slicing / PartOpt / cpu / PostRev 0.000011123199965368255 s 0.000017547 s 0.63
slicing / PartOpt / cpu / BothRev 0.000011550819999683882 s 0.000017319 s 0.67
slicing / IPartOpt / cpu / PreRev 0.000010511539985600393 s 0.000017587 s 0.60
slicing / IPartOpt / cpu / PostRev 0.000011013359944627157 s 0.000017902000000000002 s 0.62
slicing / IPartOpt / cpu / BothRev 0.00001140196009146166 s 0.000017478 s 0.65
slicing / DefOpt / cpu / PreRev 0.000010362459906900767 s 0.00001729 s 0.60
slicing / DefOpt / cpu / PostRev 0.00001125102000514744 s 0.000017531000000000002 s 0.64
slicing / DefOpt / cpu / BothRev 0.000011375419999239968 s 0.000017117 s 0.66
slicing / IDefOpt / cpu / PreRev 0.000010499399959371658 s 0.000017479 s 0.60
slicing / IDefOpt / cpu / PostRev 0.000011229100091441068 s 0.000017400999999999998 s 0.65
slicing / IDefOpt / cpu / BothRev 0.000010706280099839206 s 0.000017239999999999998 s 0.62
slicing / JaXPipe / cuda / Primal 0.000001887 s 0.000002303 s 0.82
slicing / Jax / cuda / Primal 0.000001887 s 0.000002303 s 0.82
slicing / HLOOpt / cuda / Primal 0.000001887 s 0.000002303 s 0.82
slicing / PartOpt / cuda / Primal 0.000001887 s 0.000002272 s 0.83
slicing / IPartOpt / cuda / Primal 0.000001887 s 0.000002272 s 0.83
slicing / DefOpt / cuda / Primal 0.000001887 s 0.000002303 s 0.82
slicing / IDefOpt / cuda / Primal 0.000001887 s 0.000002303 s 0.82
slicing / JaXPipe / cuda / Forward 0.000009984 s 0.000010336 s 0.97
slicing / Jax / cuda / Forward 0.00001008 s 0.00001024 s 0.98
slicing / HLOOpt / cuda / Forward 0.000010016 s 0.000009536 s 1.05
slicing / PartOpt / cuda / Forward 0.000009696 s 0.000010655 s 0.91
slicing / IPartOpt / cuda / Forward 0.000010017 s 0.000010112 s 0.99
slicing / DefOpt / cuda / Forward 0.000009984 s 0.000010464 s 0.95
slicing / IDefOpt / cuda / Forward 0.00001008 s 0.000010176 s 0.99
slicing / JaXPipe / cuda / PreRev 0.000010143 s 0.000010272 s 0.99
slicing / JaXPipe / cuda / PostRev 0.000009824 s 0.000010496 s 0.94
slicing / JaXPipe / cuda / BothRev 0.000009951 s 0.000010336 s 0.96
slicing / Jax / cuda / BothRev 0.000010944 s 0.0000104 s 1.05
slicing / HLOOpt / cuda / PreRev 0.000010177 s 0.000010368 s 0.98
slicing / HLOOpt / cuda / PostRev 0.000010816 s 0.000010559 s 1.02
slicing / HLOOpt / cuda / BothRev 0.00001088 s 0.000010208 s 1.07
slicing / PartOpt / cuda / PreRev 0.000011136 s 0.000010336 s 1.08
slicing / PartOpt / cuda / PostRev 0.000009728 s 0.000010144 s 0.96
slicing / PartOpt / cuda / BothRev 0.00000992 s 0.000010273 s 0.97
slicing / IPartOpt / cuda / PreRev 0.000009664 s 0.000010496 s 0.92
slicing / IPartOpt / cuda / PostRev 0.00001072 s 0.000010304 s 1.04
slicing / IPartOpt / cuda / BothRev 0.000009024 s 0.000010529 s 0.86
slicing / DefOpt / cuda / PreRev 0.000009536 s 0.000010624 s 0.90
slicing / DefOpt / cuda / PostRev 0.00000976 s 0.000010273 s 0.95
slicing / DefOpt / cuda / BothRev 0.000012672 s 0.0000104 s 1.22
slicing / IDefOpt / cuda / PreRev 0.000010048 s 0.000010304 s 0.98
slicing / IDefOpt / cuda / PostRev 0.000009856 s 0.000010432 s 0.94
slicing / IDefOpt / cuda / BothRev 0.00000976 s 0.000010752 s 0.91
slicing / JaXPipe / cpu / Primal 0.000016136999999999998 s 0.000013153 s 1.23
slicing / Jax / cpu / Primal 0.000015879999999999997 s 0.000012428 s 1.28
slicing / HLOOpt / cpu / Primal 0.000016332 s 0.000012639 s 1.29
slicing / PartOpt / cpu / Primal 0.000016057 s 0.000012471 s 1.29
slicing / IPartOpt / cpu / Primal 0.000016131 s 0.000012629 s 1.28
slicing / DefOpt / cpu / Primal 0.000016325 s 0.000012661 s 1.29
slicing / IDefOpt / cpu / Primal 0.00001594 s 0.000012617 s 1.26
slicing / JaXPipe / cpu / Forward 0.000021582 s 0.000016938999999999998 s 1.27
slicing / Jax / cpu / Forward 0.000020663 s 0.000016388 s 1.26
slicing / HLOOpt / cpu / Forward 0.000021306 s 0.000016751999999999998 s 1.27
slicing / PartOpt / cpu / Forward 0.000021174 s 0.000016729000000000002 s 1.27
slicing / IPartOpt / cpu / Forward 0.00002111 s 0.000016976 s 1.24
slicing / DefOpt / cpu / Forward 0.000021703 s 0.000016904 s 1.28
slicing / IDefOpt / cpu / Forward 0.000021648 s 0.000017279 s 1.25
slicing / JaXPipe / cpu / PreRev 0.000022414 s 0.000017972 s 1.25
slicing / JaXPipe / cpu / PostRev 0.000027144000000000003 s 0.000017929 s 1.51
slicing / JaXPipe / cpu / BothRev 0.000022022 s 0.000017579000000000002 s 1.25
slicing / Jax / cpu / BothRev 0.000021541000000000003 s 0.000017622 s 1.22
slicing / HLOOpt / cpu / PreRev 0.000022528 s 0.000017704999999999997 s 1.27
slicing / HLOOpt / cpu / PostRev 0.000027577 s 0.000017486 s 1.58
slicing / HLOOpt / cpu / BothRev 0.000022011 s 0.000017561 s 1.25
slicing / PartOpt / cpu / PreRev 0.000022279 s 0.000017477 s 1.27
slicing / PartOpt / cpu / PostRev 0.000022178 s 0.000017547 s 1.26
slicing / PartOpt / cpu / BothRev 0.000021563 s 0.000017319 s 1.25
slicing / IPartOpt / cpu / PreRev 0.000021747 s 0.000017587 s 1.24
slicing / IPartOpt / cpu / PostRev 0.000022020000000000003 s 0.000017902000000000002 s 1.23
slicing / IPartOpt / cpu / BothRev 0.000022323 s 0.000017478 s 1.28
slicing / DefOpt / cpu / PreRev 0.000022468 s 0.00001729 s 1.30
slicing / DefOpt / cpu / PostRev 0.000027606 s 0.000017531000000000002 s 1.57
slicing / DefOpt / cpu / BothRev 0.000021705 s 0.000017117 s 1.27
slicing / IDefOpt / cpu / PreRev 0.000022063 s 0.000017479 s 1.26
slicing / IDefOpt / cpu / PostRev 0.000022101 s 0.000017400999999999998 s 1.27
slicing / IDefOpt / cpu / BothRev 0.000021656 s 0.000017239999999999998 s 1.26
sum / JaXPipe / cpu / Primal 0.000008094679960777285 s 0.000014912 s 0.54
sum / Jax / cpu / Primal 0.000008412319984927308 s 0.000014552 s 0.58
sum / HLOOpt / cpu / Primal 0.000008335219954460626 s 0.000014439 s 0.58
sum / PartOpt / cpu / Primal 0.00000827539995952975 s 0.000014603 s 0.57
sum / IPartOpt / cpu / Primal 0.000008628000014141435 s 0.000014616 s 0.59
sum / DefOpt / cpu / Primal 0.000008524339955329197 s 0.000014516 s 0.59
sum / IDefOpt / cpu / Primal 0.000008062200031417887 s 0.000014567 s 0.55
sum / JaXPipe / cpu / Forward 0.000012649779946514171 s 0.000020212 s 0.63
sum / Jax / cpu / Forward 0.000012850500024796929 s 0.000019509 s 0.66
sum / HLOOpt / cpu / Forward 0.000012065720020473235 s 0.000019181 s 0.63
sum / PartOpt / cpu / Forward 0.000012285040029382798 s 0.000019969 s 0.62
sum / IPartOpt / cpu / Forward 0.000012768180022249 s 0.000020112 s 0.63
sum / DefOpt / cpu / Forward 0.000012387380065774778 s 0.000019514 s 0.63
sum / IDefOpt / cpu / Forward 0.00001203039999381872 s 0.000020122 s 0.60
sum / JaXPipe / cpu / PreRev 0.00001196961997266044 s 0.000019419 s 0.62
sum / JaXPipe / cpu / PostRev 0.00001230307996593183 s 0.000018786 s 0.65
sum / JaXPipe / cpu / BothRev 0.000011749379937100456 s 0.000018807 s 0.62
sum / Jax / cpu / BothRev 0.000011738560006051558 s 0.000018584 s 0.63
sum / HLOOpt / cpu / PreRev 0.000011904619987035404 s 0.000019231000000000003 s 0.62
sum / HLOOpt / cpu / PostRev 0.000013870320035493933 s 0.000018961 s 0.73
sum / HLOOpt / cpu / BothRev 0.000011738459961634364 s 0.000018822 s 0.62
sum / PartOpt / cpu / PreRev 0.000011421019971749048 s 0.000018878 s 0.60
sum / PartOpt / cpu / PostRev 0.000011838479986181485 s 0.000018693 s 0.63
sum / PartOpt / cpu / BothRev 0.000012038359982398103 s 0.000018933 s 0.64
sum / IPartOpt / cpu / PreRev 0.000012066260114806938 s 0.000018947 s 0.64
sum / IPartOpt / cpu / PostRev 0.00001161054002295714 s 0.000018804 s 0.62
sum / IPartOpt / cpu / BothRev 0.000012356760089460297 s 0.000019245000000000003 s 0.64
sum / DefOpt / cpu / PreRev 0.000012004619966319297 s 0.000019159 s 0.63
sum / DefOpt / cpu / PostRev 0.000012117299975216156 s 0.000018694 s 0.65
sum / DefOpt / cpu / BothRev 0.000012032840022584425 s 0.0000187 s 0.64
sum / IDefOpt / cpu / PreRev 0.000011628960073721828 s 0.000018622 s 0.62
sum / IDefOpt / cpu / PostRev 0.00001172898006188916 s 0.000018793 s 0.62
sum / IDefOpt / cpu / BothRev 0.0000119376400107285 s 0.000018927 s 0.63
sum / JaXPipe / cuda / Primal 0.000002047 s 0.000002495 s 0.82
sum / Jax / cuda / Primal 0.000002048 s 0.000002496 s 0.82
sum / HLOOpt / cuda / Primal 0.000002048 s 0.000002496 s 0.82
sum / PartOpt / cuda / Primal 0.000002047 s 0.000002496 s 0.82
sum / IPartOpt / cuda / Primal 0.000002048 s 0.000002496 s 0.82
sum / DefOpt / cuda / Primal 0.000002047 s 0.000002496 s 0.82
sum / IDefOpt / cuda / Primal 0.000002047 s 0.000002495 s 0.82
sum / JaXPipe / cuda / Forward 0.000010208 s 0.000010624 s 0.96
sum / Jax / cuda / Forward 0.000010144 s 0.00001104 s 0.92
sum / HLOOpt / cuda / Forward 0.0000104 s 0.00001072 s 0.97
sum / PartOpt / cuda / Forward 0.000010016 s 0.000010496 s 0.95
sum / IPartOpt / cuda / Forward 0.000010016 s 0.000010912 s 0.92
sum / DefOpt / cuda / Forward 0.000010911 s 0.000010912 s 1.00
sum / IDefOpt / cuda / Forward 0.000010336 s 0.00001056 s 0.98
sum / JaXPipe / cuda / PreRev 0.000010625 s 0.000010176 s 1.04
sum / JaXPipe / cuda / PostRev 0.000010752 s 0.000010432 s 1.03
sum / JaXPipe / cuda / BothRev 0.000009887 s 0.000010271 s 0.96
sum / Jax / cuda / BothRev 0.00001008 s 0.000010112 s 1.00
sum / HLOOpt / cuda / PreRev 0.000009792 s 0.00001024 s 0.96
sum / HLOOpt / cuda / PostRev 0.00000944 s 0.0000104 s 0.91
sum / HLOOpt / cuda / BothRev 0.000009856 s 0.000010464 s 0.94
sum / PartOpt / cuda / PreRev 0.000010176 s 0.000010592 s 0.96
sum / PartOpt / cuda / PostRev 0.000010112 s 0.000010336 s 0.98
sum / PartOpt / cuda / BothRev 0.000010016 s 0.0000104 s 0.96
sum / IPartOpt / cuda / PreRev 0.000009984 s 0.000010592 s 0.94
sum / IPartOpt / cuda / PostRev 0.000009344 s 0.0000104 s 0.90
sum / IPartOpt / cuda / BothRev 0.000009568 s 0.00000992 s 0.96
sum / DefOpt / cuda / PreRev 0.000009792 s 0.000010816 s 0.91
sum / DefOpt / cuda / PostRev 0.000009376 s 0.000010432 s 0.90
sum / DefOpt / cuda / BothRev 0.000009663 s 0.000010304 s 0.94
sum / IDefOpt / cuda / PreRev 0.000009983 s 0.000010752 s 0.93
sum / IDefOpt / cuda / PostRev 0.000010176 s 0.000010688 s 0.95
sum / IDefOpt / cuda / BothRev 0.000009792 s 0.000010496 s 0.93
sum / JaXPipe / cpu / Primal 0.000018885 s 0.000014912 s 1.27
sum / Jax / cpu / Primal 0.000018780000000000003 s 0.000014552 s 1.29
sum / HLOOpt / cpu / Primal 0.000019347 s 0.000014439 s 1.34
sum / PartOpt / cpu / Primal 0.000019156 s 0.000014603 s 1.31
sum / IPartOpt / cpu / Primal 0.000018077 s 0.000014616 s 1.24
sum / DefOpt / cpu / Primal 0.000018712 s 0.000014516 s 1.29
sum / IDefOpt / cpu / Primal 0.000023731 s 0.000014567 s 1.63
sum / JaXPipe / cpu / Forward 0.000025582 s 0.000020212 s 1.27
sum / Jax / cpu / Forward 0.000025029 s 0.000019509 s 1.28
sum / HLOOpt / cpu / Forward 0.000024867 s 0.000019181 s 1.30
sum / PartOpt / cpu / Forward 0.000024642 s 0.000019969 s 1.23
sum / IPartOpt / cpu / Forward 0.000025543 s 0.000020112 s 1.27
sum / DefOpt / cpu / Forward 0.000025109 s 0.000019514 s 1.29
sum / IDefOpt / cpu / Forward 0.000030989 s 0.000020122 s 1.54
sum / JaXPipe / cpu / PreRev 0.000024745 s 0.000019419 s 1.27
sum / JaXPipe / cpu / PostRev 0.000024067 s 0.000018786 s 1.28
sum / JaXPipe / cpu / BothRev 0.000022797 s 0.000018807 s 1.21
sum / Jax / cpu / BothRev 0.000023684 s 0.000018584 s 1.27
sum / HLOOpt / cpu / PreRev 0.000023799 s 0.000019231000000000003 s 1.24
sum / HLOOpt / cpu / PostRev 0.000024567 s 0.000018961 s 1.30
sum / HLOOpt / cpu / BothRev 0.000024352 s 0.000018822 s 1.29
sum / PartOpt / cpu / PreRev 0.00002404 s 0.000018878 s 1.27
sum / PartOpt / cpu / PostRev 0.000024541 s 0.000018693 s 1.31
sum / PartOpt / cpu / BothRev 0.000024032 s 0.000018933 s 1.27
sum / IPartOpt / cpu / PreRev 0.000024361 s 0.000018947 s 1.29
sum / IPartOpt / cpu / PostRev 0.0000247 s 0.000018804 s 1.31
sum / IPartOpt / cpu / BothRev 0.000024032 s 0.000019245000000000003 s 1.25
sum / DefOpt / cpu / PreRev 0.000024139 s 0.000019159 s 1.26
sum / DefOpt / cpu / PostRev 0.00002442 s 0.000018694 s 1.31
sum / DefOpt / cpu / BothRev 0.000023737 s 0.0000187 s 1.27
sum / IDefOpt / cpu / PreRev 0.000024027 s 0.000018622 s 1.29
sum / IDefOpt / cpu / PostRev 0.000023509 s 0.000018793 s 1.25
sum / IDefOpt / cpu / BothRev 0.000024619 s 0.000018927 s 1.30
value_and_grad / JaXPipe / cpu / Primal 0.000015381460052594775 s 0.000023072 s 0.67
value_and_grad / Jax / cpu / Primal 0.000015481699902011313 s 0.000022463 s 0.69
value_and_grad / HLOOpt / cpu / Primal 0.000015754139931232202 s 0.000022896 s 0.69
value_and_grad / PartOpt / cpu / Primal 0.000015026540004328129 s 0.000022911 s 0.66
value_and_grad / IPartOpt / cpu / Primal 0.0000145659999907366 s 0.000022942 s 0.63
value_and_grad / DefOpt / cpu / Primal 0.0000151333398935094 s 0.000022862 s 0.66
value_and_grad / IDefOpt / cpu / Primal 0.00001472878000640776 s 0.000022936 s 0.64
value_and_grad / JaXPipe / cuda / Primal 0.000033535 s 0.000034944 s 0.96
value_and_grad / Jax / cuda / Primal 0.000033184 s 0.000036448 s 0.91
value_and_grad / HLOOpt / cuda / Primal 0.000033664 s 0.000035904 s 0.94
value_and_grad / PartOpt / cuda / Primal 0.000033472 s 0.000036352 s 0.92
value_and_grad / IPartOpt / cuda / Primal 0.000033472 s 0.000036671 s 0.91
value_and_grad / DefOpt / cuda / Primal 0.000033057000000000006 s 0.000036384 s 0.91
value_and_grad / IDefOpt / cuda / Primal 0.000033504 s 0.000035392 s 0.95
value_and_grad / JaXPipe / cpu / Primal 0.000028601000000000003 s 0.000023072 s 1.24
value_and_grad / Jax / cpu / Primal 0.000028157 s 0.000022463 s 1.25
value_and_grad / HLOOpt / cpu / Primal 0.000028392 s 0.000022896 s 1.24
value_and_grad / PartOpt / cpu / Primal 0.000030939 s 0.000022911 s 1.35
value_and_grad / IPartOpt / cpu / Primal 0.000028996 s 0.000022942 s 1.26
value_and_grad / DefOpt / cpu / Primal 0.000029375 s 0.000022862 s 1.28
value_and_grad / IDefOpt / cpu / Primal 0.000028282 s 0.000022936 s 1.23
jaxmd20 / JaXPipe / cuda / Primal 0.001417056 s 0.001456319 s 0.97
jaxmd20 / Jax / cuda / Primal 0.001478624 s 0.00147408 s 1.00
jaxmd20 / HLOOpt / cuda / Primal 0.001346143 s 0.001342494 s 1.00
jaxmd20 / PartOpt / cuda / Primal 0.001303138 s 0.001357439 s 0.96
jaxmd20 / IPartOpt / cuda / Primal 0.001306336 s 0.001378975 s 0.95
jaxmd20 / DefOpt / cuda / Primal 0.000926559 s 0.000939903 s 0.99
jaxmd20 / IDefOpt / cuda / Primal 0.0009497609999999 s 0.000970495 s 0.98
jaxmd20 / JaXPipe / cuda / Forward 0.001552993 s 0.001624415 s 0.96
jaxmd20 / Jax / cuda / Forward 0.001814432 s 0.001870656 s 0.97
jaxmd20 / HLOOpt / cuda / Forward 0.00164416 s 0.001714815 s 0.96
jaxmd20 / PartOpt / cuda / Forward 0.001636512 s 0.001707679 s 0.96
jaxmd20 / IPartOpt / cuda / Forward 0.001636385 s 0.001716833 s 0.95
jaxmd20 / DefOpt / cuda / Forward 0.00165936 s 0.001706431 s 0.97
jaxmd20 / IDefOpt / cuda / Forward 0.001626848 s 0.001709119 s 0.95
jaxmd20 / JaXPipe / cuda / PreRev 0.002656319 s 0.002762878 s 0.96
jaxmd20 / JaXPipe / cuda / PostRev 0.005333183 s 0.005460187 s 0.98
jaxmd20 / JaXPipe / cuda / BothRev 0.002693312 s 0.002752605 s 0.98
jaxmd20 / Jax / cuda / BothRev 0.005326783 s 0.005413761 s 0.98
jaxmd20 / HLOOpt / cuda / PreRev 0.00275213 s 0.002928446 s 0.94
jaxmd20 / HLOOpt / cuda / PostRev 0.005331266 s 0.005459772 s 0.98
jaxmd20 / HLOOpt / cuda / BothRev 0.002721218 s 0.00282531 s 0.96
jaxmd20 / PartOpt / cuda / PreRev 0.0028088329999999 s 0.002893278 s 0.97
jaxmd20 / PartOpt / cuda / PostRev 0.005564352 s 0.00555782 s 1.00
jaxmd20 / PartOpt / cuda / BothRev 0.002761762 s 0.002812062 s 0.98
jaxmd20 / IPartOpt / cuda / PreRev 0.002804225 s 0.00290387 s 0.97
jaxmd20 / IPartOpt / cuda / PostRev 0.005539298 s 0.005551674 s 1.00
jaxmd20 / IPartOpt / cuda / BothRev 0.00276352 s 0.00283331 s 0.98
jaxmd20 / DefOpt / cuda / PreRev 0.002830913 s 0.002903646 s 0.97
jaxmd20 / DefOpt / cuda / PostRev 0.002784509 s 0.002856703 s 0.97
jaxmd20 / DefOpt / cuda / BothRev 0.002754179 s 0.002813854 s 0.98
jaxmd20 / IDefOpt / cuda / PreRev 0.0028197139999999 s 0.002919998 s 0.97
jaxmd20 / IDefOpt / cuda / PostRev 0.002319455 s 0.0023530549999999 s 0.99
jaxmd20 / IDefOpt / cuda / BothRev 0.0027590089999999 s 0.0028736929999999 s 0.96
jaxmd40 / JaXPipe / cpu / Primal 0.080871744 s 0.066913492 s 1.21
jaxmd40 / Jax / cpu / Primal 0.086542735 s 0.072603204 s 1.19
jaxmd40 / HLOOpt / cpu / Primal 0.107738907 s 0.08960644 s 1.20
jaxmd40 / PartOpt / cpu / Primal 0.0875070699999999 s 0.068425087 s 1.28
jaxmd40 / IPartOpt / cpu / Primal 0.090285724 s 0.071339947 s 1.27
jaxmd40 / DefOpt / cpu / Primal 0.111931281 s 0.082478036 s 1.36
jaxmd40 / IDefOpt / cpu / Primal 0.111787245 s 0.0873339509999999 s 1.28
jaxmd40 / JaXPipe / cpu / Forward 0.2082339299999999 s 0.162652715 s 1.28
jaxmd40 / Jax / cpu / Forward 0.11438239 s 0.083123315 s 1.38
jaxmd40 / HLOOpt / cpu / Forward 0.202549076 s 0.1585048329999999 s 1.28
jaxmd40 / PartOpt / cpu / Forward 0.206394582 s 0.156655901 s 1.32
jaxmd40 / IPartOpt / cpu / Forward 0.2166409219999999 s 0.158348678 s 1.37
jaxmd40 / DefOpt / cpu / Forward 0.215566341 s 0.160751004 s 1.34
jaxmd40 / IDefOpt / cpu / Forward 0.219232094 s 0.156729236 s 1.40
jaxmd40 / JaXPipe / cpu / PreRev 0.285179328 s 0.236199343 s 1.21
jaxmd40 / JaXPipe / cpu / PostRev 0.1827301579999999 s 0.139126863 s 1.31
jaxmd40 / JaXPipe / cpu / BothRev 0.279620878 s 0.222342406 s 1.26
jaxmd40 / Jax / cpu / BothRev 0.175687619 s 0.139139926 s 1.26
jaxmd40 / HLOOpt / cpu / PreRev 0.271401804 s 0.229555762 s 1.18
jaxmd40 / HLOOpt / cpu / PostRev 0.225985165 s 0.166843464 s 1.35
jaxmd40 / HLOOpt / cpu / BothRev 0.296669074 s 0.227585488 s 1.30
jaxmd40 / PartOpt / cpu / PreRev 0.268977918 s 0.2264403519999999 s 1.19
jaxmd40 / PartOpt / cpu / PostRev 0.158362114 s 0.1348537139999999 s 1.17
jaxmd40 / PartOpt / cpu / BothRev 0.308250832 s 0.243010073 s 1.27
jaxmd40 / IPartOpt / cpu / PreRev 0.278988759 s 0.212828879 s 1.31
jaxmd40 / IPartOpt / cpu / PostRev 0.171958436 s 0.1201934069999999 s 1.43
jaxmd40 / IPartOpt / cpu / BothRev 0.307117154 s 0.241794462 s 1.27
jaxmd40 / DefOpt / cpu / PreRev 0.265080799 s 0.235562985 s 1.13
jaxmd40 / DefOpt / cpu / PostRev 0.224490867 s 0.163910763 s 1.37
jaxmd40 / DefOpt / cpu / BothRev 0.328720387 s 0.260948006 s 1.26
jaxmd40 / IDefOpt / cpu / PreRev 0.270359511 s 0.2254909209999999 s 1.20
jaxmd40 / IDefOpt / cpu / PostRev 0.218760151 s 0.174102821 s 1.26
jaxmd40 / IDefOpt / cpu / BothRev 0.2854744329999999 s 0.249471538 s 1.14
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / cuda / Primal 1.702636371 s 1.704837354 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / cuda / Primal 1.7054310700000002 s 1.708882085 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / cuda / Primal 1.7146967709999998 s 1.7198347470000002 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / cuda / Primal 1.696679978 s 1.696605072 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / cuda / Primal 1.694667207 s 1.693524776 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / cuda / Primal 1.666413894 s 1.668373454 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / cuda / Primal 1.912049832 s 1.910824517 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / JaXPipe / cpu / Primal 7.425411464 s 5.977225408000001 s 1.24
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / Jax / cpu / Primal 7.481438081999999 s 5.948982886 s 1.26
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / HLOOpt / cpu / Primal 7.255880417 s 5.901809903999999 s 1.23
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / PartOpt / cpu / Primal 7.565334965 s 5.981963038 s 1.26
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IPartOpt / cpu / Primal 7.679269209 s 6.013644241 s 1.28
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / DefOpt / cpu / Primal 3.254772846 s 2.365326266 s 1.38
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IDefOpt / cpu / Primal 8.417110942999999 s 6.395372049 s 1.32

This comment was automatically generated by workflow using github-action-benchmark.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants