Skip to content

Conversation

@avik-pal
Copy link
Collaborator

No description provided.

@avik-pal avik-pal force-pushed the ap/raising_kernels_nicely branch from 884a636 to 0366e8f Compare December 31, 2025 00:59
@avik-pal avik-pal changed the base branch from main to ap/more_reshape_simplify December 31, 2025 00:59
@avik-pal avik-pal marked this pull request as ready for review December 31, 2025 01:00
@avik-pal avik-pal requested a review from wsmoses December 31, 2025 01:12
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: 0366e8f Previous: 426a717 Ratio
actmtch / JaXPipe / cpu / Primal 0.000006534660069519305 s 0.0000070261399650917154 s 0.93
actmtch / Jax / cpu / Primal 0.000006411939957615687 s 0.000007681720026084804 s 0.83
actmtch / HLOOpt / cpu / Primal 0.000007368159949692199 s 0.000008456099985778564 s 0.87
actmtch / PartOpt / cpu / Primal 0.0000065999400158034405 s 0.000008354660039913142 s 0.79
actmtch / IPartOpt / cpu / Primal 0.000006515660015793401 s 0.000008498499983033981 s 0.77
actmtch / DefOpt / cpu / Primal 0.000007076279980537947 s 0.000008734280045246124 s 0.81
actmtch / IDefOpt / cpu / Primal 0.000007243339987326181 s 0.000008587439951952547 s 0.84
actmtch / JaXPipe / cpu / Forward 0.000010296739992554648 s 0.000012429179996615858 s 0.83
actmtch / Jax / cpu / Forward 0.00000896319997991668 s 0.000010656700023901068 s 0.84
actmtch / HLOOpt / cpu / Forward 0.000010796320011650096 s 0.000012589040015882348 s 0.86
actmtch / PartOpt / cpu / Forward 0.000009954679999282234 s 0.000011920439947061825 s 0.84
actmtch / IPartOpt / cpu / Forward 0.00001036839998960204 s 0.000012381419974190069 s 0.84
actmtch / DefOpt / cpu / Forward 0.00000991200000498793 s 0.000012271340056031477 s 0.81
actmtch / IDefOpt / cpu / Forward 0.000009930819987857831 s 0.000012316779984757888 s 0.81
actmtch / JaXPipe / cpu / PreRev 0.0000104997599919443 s 0.00001274931999432738 s 0.82
actmtch / JaXPipe / cpu / PostRev 0.00000958585996158945 s 0.000011261919989919989 s 0.85
actmtch / JaXPipe / cpu / BothRev 0.000010546300027272082 s 0.000012745980011459324 s 0.83
actmtch / Jax / cpu / BothRev 0.000009288999999625956 s 0.00001086691999262257 s 0.85
actmtch / HLOOpt / cpu / PreRev 0.000010674779987311924 s 0.000012499179983933571 s 0.85
actmtch / HLOOpt / cpu / PostRev 0.000012785860035364748 s 0.000014753979976376283 s 0.87
actmtch / HLOOpt / cpu / BothRev 0.000010241099980703438 s 0.00001177669998469355 s 0.87
actmtch / PartOpt / cpu / PreRev 0.000010303259987267666 s 0.000012723199997708434 s 0.81
actmtch / PartOpt / cpu / PostRev 0.000009160480003629345 s 0.000011315820029267342 s 0.81
actmtch / PartOpt / cpu / BothRev 0.000010901739988185 s 0.000012766059990099166 s 0.85
actmtch / IPartOpt / cpu / PreRev 0.000010015359994213213 s 0.000012880159993073904 s 0.78
actmtch / IPartOpt / cpu / PostRev 0.000009132400000453345 s 0.000010472339963598645 s 0.87
actmtch / IPartOpt / cpu / BothRev 0.00001041134003571642 s 0.000012367339950287716 s 0.84
actmtch / DefOpt / cpu / PreRev 0.00001013401999443886 s 0.000012020940012007486 s 0.84
actmtch / DefOpt / cpu / PostRev 0.00001063377999344084 s 0.000012939700009155783 s 0.82
actmtch / DefOpt / cpu / BothRev 0.000010669280027286733 s 0.000012443320019883686 s 0.86
actmtch / IDefOpt / cpu / PreRev 0.000010446419992149458 s 0.000012142540026616188 s 0.86
actmtch / IDefOpt / cpu / PostRev 0.000010785220028992626 s 0.00001286107997657382 s 0.84
actmtch / IDefOpt / cpu / BothRev 0.000010439199995744274 s 0.000012399619981806609 s 0.84
actmtch / JaXPipe / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / Jax / cuda / Primal 0.0000024 s 0.000002015 s 1.19
actmtch / HLOOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / PartOpt / cuda / Primal 0.0000024 s 0.000002015 s 1.19
actmtch / IPartOpt / cuda / Primal 0.0000024 s 0.000002015 s 1.19
actmtch / DefOpt / cuda / Primal 0.0000024 s 0.000002015 s 1.19
actmtch / IDefOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / JaXPipe / cuda / Forward 0.000011552 s 0.000009887 s 1.17
actmtch / Jax / cuda / Forward 0.00001136 s 0.000009888 s 1.15
actmtch / HLOOpt / cuda / Forward 0.0000112 s 0.000010048 s 1.11
actmtch / PartOpt / cuda / Forward 0.000009728 s 0.000010112 s 0.96
actmtch / IPartOpt / cuda / Forward 0.000010592 s 0.00000992 s 1.07
actmtch / DefOpt / cuda / Forward 0.000010815 s 0.000009696 s 1.12
actmtch / IDefOpt / cuda / Forward 0.000010432 s 0.000010272 s 1.02
actmtch / JaXPipe / cuda / PreRev 0.000010368 s 0.000009215 s 1.13
actmtch / JaXPipe / cuda / PostRev 0.000010368 s 0.000010624 s 0.98
actmtch / JaXPipe / cuda / BothRev 0.000010048 s 0.00000992 s 1.01
actmtch / Jax / cuda / BothRev 0.000011137 s 0.000010271 s 1.08
actmtch / HLOOpt / cuda / PreRev 0.00001072 s 0.000010144 s 1.06
actmtch / HLOOpt / cuda / PostRev 0.00001024 s 0.00001024 s 1
actmtch / HLOOpt / cuda / BothRev 0.000010208 s 0.000010304 s 0.99
actmtch / PartOpt / cuda / PreRev 0.000010464 s 0.00000992 s 1.05
actmtch / PartOpt / cuda / PostRev 0.000010336 s 0.00001008 s 1.03
actmtch / PartOpt / cuda / BothRev 0.0000104 s 0.000010015 s 1.04
actmtch / IPartOpt / cuda / PreRev 0.000010464 s 0.000010336 s 1.01
actmtch / IPartOpt / cuda / PostRev 0.000010591 s 0.000010272 s 1.03
actmtch / IPartOpt / cuda / BothRev 0.00001056 s 0.0000104 s 1.02
actmtch / DefOpt / cuda / PreRev 0.000010624 s 0.000010112 s 1.05
actmtch / DefOpt / cuda / PostRev 0.000010368 s 0.000011712 s 0.89
actmtch / DefOpt / cuda / BothRev 0.000010368 s 0.000010081 s 1.03
actmtch / IDefOpt / cuda / PreRev 0.000010368 s 0.000010177 s 1.02
actmtch / IDefOpt / cuda / PostRev 0.000010784 s 0.000009696 s 1.11
actmtch / IDefOpt / cuda / BothRev 0.00001312 s 0.000010208 s 1.29
actmtch / JaXPipe / tpu / Primal 5.83025e-7 s 5.63175e-7 s 1.04
actmtch / Jax / tpu / Primal 5.63725e-7 s 5.96825e-7 s 0.94
actmtch / HLOOpt / tpu / Primal 0.0000021734250000000003 s 0.000002097775 s 1.04
actmtch / PartOpt / tpu / Primal 5.63525e-7 s 5.9665e-7 s 0.94
actmtch / IPartOpt / tpu / Primal 5.756250000000001e-7 s 5.52675e-7 s 1.04
actmtch / DefOpt / tpu / Primal 0.0000020651 s 0.000002164025 s 0.95
actmtch / IDefOpt / tpu / Primal 0.0000021755 s 0.000002095375 s 1.04
actmtch / JaXPipe / tpu / Forward 0.000003863925 s 0.0000038255 s 1.01
actmtch / Jax / tpu / Forward 0.000001233575 s 0.0000012064999999999998 s 1.02
actmtch / HLOOpt / tpu / Forward 0.000003682125 s 0.0000039346750000000006 s 0.94
actmtch / PartOpt / tpu / Forward 0.000003898125 s 0.00000392895 s 0.99
actmtch / IPartOpt / tpu / Forward 0.0000036522 s 0.00000393315 s 0.93
actmtch / DefOpt / tpu / Forward 0.00000389455 s 0.00000391195 s 1.00
actmtch / IDefOpt / tpu / Forward 0.000003663625 s 0.0000039288250000000005 s 0.93
actmtch / JaXPipe / tpu / PreRev 0.00000376195 s 0.00000348385 s 1.08
actmtch / JaXPipe / tpu / PostRev 0.0000016236999999999998 s 0.000001634375 s 0.99
actmtch / JaXPipe / tpu / BothRev 0.0000037482 s 0.000003499325 s 1.07
actmtch / Jax / tpu / BothRev 0.000001625075 s 0.00000163605 s 0.99
actmtch / HLOOpt / tpu / PreRev 0.000003760925 s 0.00000348265 s 1.08
actmtch / HLOOpt / tpu / PostRev 0.0000034375 s 0.000003403275 s 1.01
actmtch / HLOOpt / tpu / BothRev 0.00000375995 s 0.00000347005 s 1.08
actmtch / PartOpt / tpu / PreRev 0.000003457025 s 0.00000340765 s 1.01
actmtch / PartOpt / tpu / PostRev 0.0000016749500000000002 s 0.0000015938749999999998 s 1.05
actmtch / PartOpt / tpu / BothRev 0.000003444925 s 0.000003406875 s 1.01
actmtch / IPartOpt / tpu / PreRev 0.0000037543 s 0.00000347205 s 1.08
actmtch / IPartOpt / tpu / PostRev 0.0000016164500000000002 s 0.00000163595 s 0.99
actmtch / IPartOpt / tpu / BothRev 0.000003741675 s 0.000003471625 s 1.08
actmtch / DefOpt / tpu / PreRev 0.000003450025 s 0.000003417425 s 1.01
actmtch / DefOpt / tpu / PostRev 0.00000367205 s 0.00000340655 s 1.08
actmtch / DefOpt / tpu / BothRev 0.0000034325 s 0.000003411675 s 1.01
actmtch / IDefOpt / tpu / PreRev 0.0000037473 s 0.00000347155 s 1.08
actmtch / IDefOpt / tpu / PostRev 0.00000344445 s 0.000003411 s 1.01
actmtch / IDefOpt / tpu / BothRev 0.0000037465 s 0.000003475375 s 1.08
actmtch / JaXPipe / cpu / Primal 0.00001347 s 0.0000070261399650917154 s 1.92
actmtch / Jax / cpu / Primal 0.00001312 s 0.000007681720026084804 s 1.71
actmtch / HLOOpt / cpu / Primal 0.000013994 s 0.000008456099985778564 s 1.65
actmtch / PartOpt / cpu / Primal 0.00001342 s 0.000008354660039913142 s 1.61
actmtch / IPartOpt / cpu / Primal 0.000013297 s 0.000008498499983033981 s 1.56
actmtch / DefOpt / cpu / Primal 0.000014596 s 0.000008734280045246124 s 1.67
actmtch / IDefOpt / cpu / Primal 0.000014213 s 0.000008587439951952547 s 1.66
actmtch / JaXPipe / cpu / Forward 0.000020177 s 0.000012429179996615858 s 1.62
actmtch / Jax / cpu / Forward 0.000018169 s 0.000010656700023901068 s 1.70
actmtch / HLOOpt / cpu / Forward 0.000019122 s 0.000012589040015882348 s 1.52
actmtch / PartOpt / cpu / Forward 0.000019259 s 0.000011920439947061825 s 1.62
actmtch / IPartOpt / cpu / Forward 0.000018894 s 0.000012381419974190069 s 1.53
actmtch / DefOpt / cpu / Forward 0.000020228 s 0.000012271340056031477 s 1.65
actmtch / IDefOpt / cpu / Forward 0.000019215 s 0.000012316779984757888 s 1.56
actmtch / JaXPipe / cpu / PreRev 0.000020015 s 0.00001274931999432738 s 1.57
actmtch / JaXPipe / cpu / PostRev 0.000017402 s 0.000011261919989919989 s 1.55
actmtch / JaXPipe / cpu / BothRev 0.000019791 s 0.000012745980011459324 s 1.55
actmtch / Jax / cpu / BothRev 0.000018457 s 0.00001086691999262257 s 1.70
actmtch / HLOOpt / cpu / PreRev 0.000019022 s 0.000012499179983933571 s 1.52
actmtch / HLOOpt / cpu / PostRev 0.000020343 s 0.000014753979976376283 s 1.38
actmtch / HLOOpt / cpu / BothRev 0.000019569 s 0.00001177669998469355 s 1.66
actmtch / PartOpt / cpu / PreRev 0.000019724 s 0.000012723199997708434 s 1.55
actmtch / PartOpt / cpu / PostRev 0.000018013999999999997 s 0.000011315820029267342 s 1.59
actmtch / PartOpt / cpu / BothRev 0.000019519 s 0.000012766059990099166 s 1.53
actmtch / IPartOpt / cpu / PreRev 0.000020114 s 0.000012880159993073904 s 1.56
actmtch / IPartOpt / cpu / PostRev 0.000017654 s 0.000010472339963598645 s 1.69
actmtch / IPartOpt / cpu / BothRev 0.000020042 s 0.000012367339950287716 s 1.62
actmtch / DefOpt / cpu / PreRev 0.000019936 s 0.000012020940012007486 s 1.66
actmtch / DefOpt / cpu / PostRev 0.000019593 s 0.000012939700009155783 s 1.51
actmtch / DefOpt / cpu / BothRev 0.000019561 s 0.000012443320019883686 s 1.57
actmtch / IDefOpt / cpu / PreRev 0.000019727000000000003 s 0.000012142540026616188 s 1.62
actmtch / IDefOpt / cpu / PostRev 0.000019547 s 0.00001286107997657382 s 1.52
actmtch / IDefOpt / cpu / BothRev 0.000019808 s 0.000012399619981806609 s 1.60
actmtch / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.0000070261399650917154 s 1.28
actmtch / Jax / cpu / Primal 0.000008999999999999999 s 0.000007681720026084804 s 1.17
actmtch / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000008456099985778564 s 1.06
actmtch / PartOpt / cpu / Primal 0.000008 s 0.000008354660039913142 s 0.96
actmtch / IPartOpt / cpu / Primal 0.000008 s 0.000008498499983033981 s 0.94
actmtch / DefOpt / cpu / Primal 0.00001 s 0.000008734280045246124 s 1.14
actmtch / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000008587439951952547 s 1.05
actmtch / JaXPipe / cpu / Forward 0.000013 s 0.000012429179996615858 s 1.05
actmtch / Jax / cpu / Forward 0.000012 s 0.000010656700023901068 s 1.13
actmtch / HLOOpt / cpu / Forward 0.000012 s 0.000012589040015882348 s 0.95
actmtch / PartOpt / cpu / Forward 0.000013 s 0.000011920439947061825 s 1.09
actmtch / IPartOpt / cpu / Forward 0.000013 s 0.000012381419974190069 s 1.05
actmtch / DefOpt / cpu / Forward 0.000013 s 0.000012271340056031477 s 1.06
actmtch / IDefOpt / cpu / Forward 0.000012 s 0.000012316779984757888 s 0.97
actmtch / JaXPipe / cpu / PreRev 0.000013 s 0.00001274931999432738 s 1.02
actmtch / JaXPipe / cpu / PostRev 0.000012 s 0.000011261919989919989 s 1.07
actmtch / JaXPipe / cpu / BothRev 0.000014 s 0.000012745980011459324 s 1.10
actmtch / Jax / cpu / BothRev 0.000011 s 0.00001086691999262257 s 1.01
actmtch / HLOOpt / cpu / PreRev 0.000013 s 0.000012499179983933571 s 1.04
actmtch / HLOOpt / cpu / PostRev 0.000013 s 0.000014753979976376283 s 0.88
actmtch / HLOOpt / cpu / BothRev 0.000013 s 0.00001177669998469355 s 1.10
actmtch / PartOpt / cpu / PreRev 0.000014 s 0.000012723199997708434 s 1.10
actmtch / PartOpt / cpu / PostRev 0.000012 s 0.000011315820029267342 s 1.06
actmtch / PartOpt / cpu / BothRev 0.000013 s 0.000012766059990099166 s 1.02
actmtch / IPartOpt / cpu / PreRev 0.000013 s 0.000012880159993073904 s 1.01
actmtch / IPartOpt / cpu / PostRev 0.000012 s 0.000010472339963598645 s 1.15
actmtch / IPartOpt / cpu / BothRev 0.000014 s 0.000012367339950287716 s 1.13
actmtch / DefOpt / cpu / PreRev 0.000013 s 0.000012020940012007486 s 1.08
actmtch / DefOpt / cpu / PostRev 0.000013 s 0.000012939700009155783 s 1.00
actmtch / DefOpt / cpu / BothRev 0.000013 s 0.000012443320019883686 s 1.04
actmtch / IDefOpt / cpu / PreRev 0.000013 s 0.000012142540026616188 s 1.07
actmtch / IDefOpt / cpu / PostRev 0.000012 s 0.00001286107997657382 s 0.93
actmtch / IDefOpt / cpu / BothRev 0.000013 s 0.000012399619981806609 s 1.05
add_one / JaXPipe / cpu / Primal 0.000006814180023866356 s 0.00000767638000070292 s 0.89
add_one / Jax / cpu / Primal 0.000007181920009315945 s 0.000007881059946157621 s 0.91
add_one / HLOOpt / cpu / Primal 0.00000649835996227921 s 0.00000769876004596881 s 0.84
add_one / PartOpt / cpu / Primal 0.000006530419987029745 s 0.00000769573998695705 s 0.85
add_one / IPartOpt / cpu / Primal 0.000006788119981138152 s 0.000007311640019906917 s 0.93
add_one / DefOpt / cpu / Primal 0.000006439200024033198 s 0.000006927979984538979 s 0.93
add_one / IDefOpt / cpu / Primal 0.000006555560030392372 s 0.000007248140027513728 s 0.90
add_one / JaXPipe / cpu / Forward 0.00000957288000790868 s 0.000011312520000501535 s 0.85
add_one / Jax / cpu / Forward 0.000009551199973429902 s 0.000010935220007013414 s 0.87
add_one / HLOOpt / cpu / Forward 0.000009657499958848348 s 0.00001113865996558161 s 0.87
add_one / PartOpt / cpu / Forward 0.00000978715997916879 s 0.000011150479995194472 s 0.88
add_one / IPartOpt / cpu / Forward 0.000009739699989950168 s 0.00001149292000263813 s 0.85
add_one / DefOpt / cpu / Forward 0.000009689859953141422 s 0.000011125779992653409 s 0.87
add_one / IDefOpt / cpu / Forward 0.00000953753998146567 s 0.000010953519959002734 s 0.87
add_one / JaXPipe / cpu / PreRev 0.000011123319982289104 s 0.00001329155999883369 s 0.84
add_one / JaXPipe / cpu / PostRev 0.000010463459984748624 s 0.00001282565998735663 s 0.82
add_one / JaXPipe / cpu / BothRev 0.000011179760022059782 s 0.000013342939955691691 s 0.84
add_one / Jax / cpu / BothRev 0.00001090562001991202 s 0.000013148820007700124 s 0.83
add_one / HLOOpt / cpu / PreRev 0.000011056539997298386 s 0.000013094879996060629 s 0.84
add_one / HLOOpt / cpu / PostRev 0.000013031879980189842 s 0.000014685159976579598 s 0.89
add_one / HLOOpt / cpu / BothRev 0.000010786380007630214 s 0.000013228520047050553 s 0.82
add_one / PartOpt / cpu / PreRev 0.00001099202000659716 s 0.000012965020023329998 s 0.85
add_one / PartOpt / cpu / PostRev 0.000011035780044039712 s 0.000012663999968935967 s 0.87
add_one / PartOpt / cpu / BothRev 0.000011127599973406176 s 0.000014077939986236744 s 0.79
add_one / IPartOpt / cpu / PreRev 0.000010632259991325556 s 0.00001328809998994984 s 0.80
add_one / IPartOpt / cpu / PostRev 0.000010702100043999962 s 0.00001288855998609506 s 0.83
add_one / IPartOpt / cpu / BothRev 0.000010765379975055112 s 0.000013109620003888268 s 0.82
add_one / DefOpt / cpu / PreRev 0.00001092376004635298 s 0.000013311499997143982 s 0.82
add_one / DefOpt / cpu / PostRev 0.00001075169997420744 s 0.000012807700004486833 s 0.84
add_one / DefOpt / cpu / BothRev 0.000010675779967641577 s 0.00001317183998253313 s 0.81
add_one / IDefOpt / cpu / PreRev 0.000010985000044456682 s 0.00001243973999407899 s 0.88
add_one / IDefOpt / cpu / PostRev 0.00001074214001164364 s 0.00001230946002579003 s 0.87
add_one / IDefOpt / cpu / BothRev 0.000010660579973773566 s 0.00001255659997696057 s 0.85
add_one / JaXPipe / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / Jax / cuda / Primal 0.000002336 s 0.0000019200000000000003 s 1.22
add_one / HLOOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / PartOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / IPartOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / DefOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / IDefOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / JaXPipe / cuda / Forward 0.00001008 s 0.000010016 s 1.01
add_one / Jax / cuda / Forward 0.000010528 s 0.000009792 s 1.08
add_one / HLOOpt / cuda / Forward 0.00001024 s 0.000010145 s 1.01
add_one / PartOpt / cuda / Forward 0.000010433 s 0.00001008 s 1.04
add_one / IPartOpt / cuda / Forward 0.000010431 s 0.00000992 s 1.05
add_one / DefOpt / cuda / Forward 0.00001024 s 0.000010144 s 1.01
add_one / IDefOpt / cuda / Forward 0.000010208 s 0.000009761 s 1.05
add_one / JaXPipe / cuda / PreRev 0.000024768 s 0.000024704 s 1.00
add_one / JaXPipe / cuda / PostRev 0.000025184 s 0.000024256 s 1.04
add_one / JaXPipe / cuda / BothRev 0.000025184 s 0.000024577 s 1.02
add_one / Jax / cuda / BothRev 0.000025728 s 0.000025088 s 1.03
add_one / HLOOpt / cuda / PreRev 0.00002496 s 0.000025088 s 0.99
add_one / HLOOpt / cuda / PostRev 0.000024672 s 0.000024256 s 1.02
add_one / HLOOpt / cuda / BothRev 0.000024768 s 0.0000312 s 0.79
add_one / PartOpt / cuda / PreRev 0.000024864 s 0.000024896 s 1.00
add_one / PartOpt / cuda / PostRev 0.000025728 s 0.000024609 s 1.05
add_one / PartOpt / cuda / BothRev 0.000025536 s 0.000024768 s 1.03
add_one / IPartOpt / cuda / PreRev 0.000025184 s 0.000024768 s 1.02
add_one / IPartOpt / cuda / PostRev 0.000025056 s 0.000024704 s 1.01
add_one / IPartOpt / cuda / BothRev 0.000026304 s 0.000023808 s 1.10
add_one / DefOpt / cuda / PreRev 0.00002544 s 0.000025024 s 1.02
add_one / DefOpt / cuda / PostRev 0.000025504 s 0.000025217 s 1.01
add_one / DefOpt / cuda / BothRev 0.000025312 s 0.00002528 s 1.00
add_one / IDefOpt / cuda / PreRev 0.000025504 s 0.00002496 s 1.02
add_one / IDefOpt / cuda / PostRev 0.000025344 s 0.00002496 s 1.02
add_one / IDefOpt / cuda / BothRev 0.000025536 s 0.00002432 s 1.05
add_one / JaXPipe / tpu / Primal 0.0000014475250000000002 s 0.0000014234749999999995 s 1.02
add_one / Jax / tpu / Primal 0.0000014468 s 0.00000140335 s 1.03
add_one / HLOOpt / tpu / Primal 0.000001458525 s 0.000001423325 s 1.02
add_one / PartOpt / tpu / Primal 0.0000014682 s 0.0000014036 s 1.05
add_one / IPartOpt / tpu / Primal 0.000001446275 s 0.000001422475 s 1.02
add_one / DefOpt / tpu / Primal 0.0000014483999999999998 s 0.000001397825 s 1.04
add_one / IDefOpt / tpu / Primal 0.00000145035 s 0.000001422125 s 1.02
add_one / JaXPipe / tpu / Forward 0.000001910025 s 0.00000184455 s 1.04
add_one / Jax / tpu / Forward 0.0000018688250000000005 s 0.000001838475 s 1.02
add_one / HLOOpt / tpu / Forward 0.000001911175 s 0.000001845525 s 1.04
add_one / PartOpt / tpu / Forward 0.00000186745 s 0.00000183625 s 1.02
add_one / IPartOpt / tpu / Forward 0.0000019152250000000004 s 0.0000018463 s 1.04
add_one / DefOpt / tpu / Forward 0.000001863125 s 0.00000183405 s 1.02
add_one / IDefOpt / tpu / Forward 0.000001901275 s 0.00000185615 s 1.02
add_one / JaXPipe / tpu / PreRev 0.000002254575 s 0.0000022316 s 1.01
add_one / JaXPipe / tpu / PostRev 0.0000023022999999999995 s 0.000002234475 s 1.03
add_one / JaXPipe / tpu / BothRev 0.0000022623 s 0.000002245975 s 1.01
add_one / Jax / tpu / BothRev 0.0000022953000000000003 s 0.00000224375 s 1.02
add_one / HLOOpt / tpu / PreRev 0.0000022573 s 0.000002241625 s 1.01
add_one / HLOOpt / tpu / PostRev 0.00000229485 s 0.0000022375999999999995 s 1.03
add_one / HLOOpt / tpu / BothRev 0.0000022623 s 0.0000022452 s 1.01
add_one / PartOpt / tpu / PreRev 0.000002295625 s 0.00000223565 s 1.03
add_one / PartOpt / tpu / PostRev 0.00000226215 s 0.00000224485 s 1.01
add_one / PartOpt / tpu / BothRev 0.00000229955 s 0.0000022485 s 1.02
add_one / IPartOpt / tpu / PreRev 0.00000225265 s 0.0000022385 s 1.01
add_one / IPartOpt / tpu / PostRev 0.0000022912750000000004 s 0.000002240525 s 1.02
add_one / IPartOpt / tpu / BothRev 0.000002261975 s 0.0000022301 s 1.01
add_one / DefOpt / tpu / PreRev 0.00000228735 s 0.0000022470250000000003 s 1.02
add_one / DefOpt / tpu / PostRev 0.000002261325 s 0.00000223715 s 1.01
add_one / DefOpt / tpu / BothRev 0.0000022914250000000003 s 0.000002235075 s 1.03
add_one / IDefOpt / tpu / PreRev 0.0000022663 s 0.000002230725 s 1.02
add_one / IDefOpt / tpu / PostRev 0.0000023006 s 0.000002232775 s 1.03
add_one / IDefOpt / tpu / BothRev 0.0000022603 s 0.000002229975 s 1.01
add_one / JaXPipe / cpu / Primal 0.000013488 s 0.00000767638000070292 s 1.76
add_one / Jax / cpu / Primal 0.000013069 s 0.000007881059946157621 s 1.66
add_one / HLOOpt / cpu / Primal 0.000013112 s 0.00000769876004596881 s 1.70
add_one / PartOpt / cpu / Primal 0.000013026 s 0.00000769573998695705 s 1.69
add_one / IPartOpt / cpu / Primal 0.000013158 s 0.000007311640019906917 s 1.80
add_one / DefOpt / cpu / Primal 0.000012628 s 0.000006927979984538979 s 1.82
add_one / IDefOpt / cpu / Primal 0.000012754 s 0.000007248140027513728 s 1.76
add_one / JaXPipe / cpu / Forward 0.00001795 s 0.000011312520000501535 s 1.59
add_one / Jax / cpu / Forward 0.000017689 s 0.000010935220007013414 s 1.62
add_one / HLOOpt / cpu / Forward 0.000017843000000000002 s 0.00001113865996558161 s 1.60
add_one / PartOpt / cpu / Forward 0.000018407 s 0.000011150479995194472 s 1.65
add_one / IPartOpt / cpu / Forward 0.000018088 s 0.00001149292000263813 s 1.57
add_one / DefOpt / cpu / Forward 0.000017582000000000002 s 0.000011125779992653409 s 1.58
add_one / IDefOpt / cpu / Forward 0.000017715999999999998 s 0.000010953519959002734 s 1.62
add_one / JaXPipe / cpu / PreRev 0.000019851 s 0.00001329155999883369 s 1.49
add_one / JaXPipe / cpu / PostRev 0.000019397 s 0.00001282565998735663 s 1.51
add_one / JaXPipe / cpu / BothRev 0.000019771 s 0.000013342939955691691 s 1.48
add_one / Jax / cpu / BothRev 0.000019735 s 0.000013148820007700124 s 1.50
add_one / HLOOpt / cpu / PreRev 0.000019845 s 0.000013094879996060629 s 1.52
add_one / HLOOpt / cpu / PostRev 0.000019869 s 0.000014685159976579598 s 1.35
add_one / HLOOpt / cpu / BothRev 0.000020317 s 0.000013228520047050553 s 1.54
add_one / PartOpt / cpu / PreRev 0.000019942 s 0.000012965020023329998 s 1.54
add_one / PartOpt / cpu / PostRev 0.000020433 s 0.000012663999968935967 s 1.61
add_one / PartOpt / cpu / BothRev 0.000019902 s 0.000014077939986236744 s 1.41
add_one / IPartOpt / cpu / PreRev 0.000019784 s 0.00001328809998994984 s 1.49
add_one / IPartOpt / cpu / PostRev 0.000019795 s 0.00001288855998609506 s 1.54
add_one / IPartOpt / cpu / BothRev 0.000019469 s 0.000013109620003888268 s 1.49
add_one / DefOpt / cpu / PreRev 0.000019714 s 0.000013311499997143982 s 1.48
add_one / DefOpt / cpu / PostRev 0.000019504 s 0.000012807700004486833 s 1.52
add_one / DefOpt / cpu / BothRev 0.000020244 s 0.00001317183998253313 s 1.54
add_one / IDefOpt / cpu / PreRev 0.000019867 s 0.00001243973999407899 s 1.60
add_one / IDefOpt / cpu / PostRev 0.000019645 s 0.00001230946002579003 s 1.60
add_one / IDefOpt / cpu / BothRev 0.000019797 s 0.00001255659997696057 s 1.58
add_one / JaXPipe / cpu / Primal 0.000008 s 0.00000767638000070292 s 1.04
add_one / Jax / cpu / Primal 0.000008 s 0.000007881059946157621 s 1.02
add_one / HLOOpt / cpu / Primal 0.000008 s 0.00000769876004596881 s 1.04
add_one / PartOpt / cpu / Primal 0.000008999999999999999 s 0.00000769573998695705 s 1.17
add_one / IPartOpt / cpu / Primal 0.000008 s 0.000007311640019906917 s 1.09
add_one / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000006927979984538979 s 1.30
add_one / IDefOpt / cpu / Primal 0.000008 s 0.000007248140027513728 s 1.10
add_one / JaXPipe / cpu / Forward 0.000011 s 0.000011312520000501535 s 0.97
add_one / Jax / cpu / Forward 0.000011 s 0.000010935220007013414 s 1.01
add_one / HLOOpt / cpu / Forward 0.000011 s 0.00001113865996558161 s 0.99
add_one / PartOpt / cpu / Forward 0.000011 s 0.000011150479995194472 s 0.99
add_one / IPartOpt / cpu / Forward 0.000011 s 0.00001149292000263813 s 0.96
add_one / DefOpt / cpu / Forward 0.000012 s 0.000011125779992653409 s 1.08
add_one / IDefOpt / cpu / Forward 0.000011 s 0.000010953519959002734 s 1.00
add_one / JaXPipe / cpu / PreRev 0.000014 s 0.00001329155999883369 s 1.05
add_one / JaXPipe / cpu / PostRev 0.000013 s 0.00001282565998735663 s 1.01
add_one / JaXPipe / cpu / BothRev 0.000013 s 0.000013342939955691691 s 0.97
add_one / Jax / cpu / BothRev 0.000014 s 0.000013148820007700124 s 1.06
add_one / HLOOpt / cpu / PreRev 0.000013 s 0.000013094879996060629 s 0.99
add_one / HLOOpt / cpu / PostRev 0.000013 s 0.000014685159976579598 s 0.89
add_one / HLOOpt / cpu / BothRev 0.000013 s 0.000013228520047050553 s 0.98
add_one / PartOpt / cpu / PreRev 0.000012 s 0.000012965020023329998 s 0.93
add_one / PartOpt / cpu / PostRev 0.000013 s 0.000012663999968935967 s 1.03
add_one / PartOpt / cpu / BothRev 0.000013 s 0.000014077939986236744 s 0.92
add_one / IPartOpt / cpu / PreRev 0.000013 s 0.00001328809998994984 s 0.98
add_one / IPartOpt / cpu / PostRev 0.000012 s 0.00001288855998609506 s 0.93
add_one / IPartOpt / cpu / BothRev 0.000013 s 0.000013109620003888268 s 0.99
add_one / DefOpt / cpu / PreRev 0.000013 s 0.000013311499997143982 s 0.98
add_one / DefOpt / cpu / PostRev 0.000013 s 0.000012807700004486833 s 1.02
add_one / DefOpt / cpu / BothRev 0.000013 s 0.00001317183998253313 s 0.99
add_one / IDefOpt / cpu / PreRev 0.000013 s 0.00001243973999407899 s 1.05
add_one / IDefOpt / cpu / PostRev 0.000013 s 0.00001230946002579003 s 1.06
add_one / IDefOpt / cpu / BothRev 0.000013 s 0.00001255659997696057 s 1.04
add_two / JaXPipe / cpu / Primal 0.000006800599994676304 s 0.00000840646003780421 s 0.81
add_two / Jax / cpu / Primal 0.00000699705999068101 s 0.000007408079991364502 s 0.94
add_two / HLOOpt / cpu / Primal 0.000007040059999781079 s 0.0000074600599691621025 s 0.94
add_two / PartOpt / cpu / Primal 0.000006775159990866087 s 0.000007752780065857223 s 0.87
add_two / IPartOpt / cpu / Primal 0.000007121239987100125 s 0.000008030020026126295 s 0.89
add_two / DefOpt / cpu / Primal 0.000006773800023438525 s 0.00000800805999460863 s 0.85
add_two / IDefOpt / cpu / Primal 0.0000070801399670017415 s 0.000007063699977152282 s 1.00
add_two / JaXPipe / cpu / Forward 0.000009906560026138322 s 0.000011770780010920136 s 0.84
add_two / Jax / cpu / Forward 0.000010106239997185183 s 0.000011452759990788764 s 0.88
add_two / HLOOpt / cpu / Forward 0.000010121039995283354 s 0.00001166930002909794 s 0.87
add_two / PartOpt / cpu / Forward 0.000010154880010304624 s 0.000011410620018068583 s 0.89
add_two / IPartOpt / cpu / Forward 0.000010076919998027734 s 0.000011299139960101456 s 0.89
add_two / DefOpt / cpu / Forward 0.000009863439936452778 s 0.000011692319976646103 s 0.84
add_two / IDefOpt / cpu / Forward 0.000010056960018118844 s 0.000011732559996744386 s 0.86
add_two / JaXPipe / cpu / PreRev 0.00001303953997194185 s 0.000015161400006036274 s 0.86
add_two / JaXPipe / cpu / PostRev 0.000012801260054402518 s 0.00001595720002114831 s 0.80
add_two / JaXPipe / cpu / BothRev 0.000012974179999218904 s 0.00001547035995827173 s 0.84
add_two / Jax / cpu / BothRev 0.000012792259958587238 s 0.00001497119998020935 s 0.85
add_two / HLOOpt / cpu / PreRev 0.000013045019977653285 s 0.000015616819973729433 s 0.84
add_two / HLOOpt / cpu / PostRev 0.000014957879948269692 s 0.000016793039994809077 s 0.89
add_two / HLOOpt / cpu / BothRev 0.00001310303999161988 s 0.000015394640031445305 s 0.85
add_two / PartOpt / cpu / PreRev 0.00001264147999791021 s 0.00001552674003505672 s 0.81
add_two / PartOpt / cpu / PostRev 0.000013238620022093528 s 0.00001590923999174265 s 0.83
add_two / PartOpt / cpu / BothRev 0.000012895520012534688 s 0.0000157065999792394 s 0.82
add_two / IPartOpt / cpu / PreRev 0.000012749719971907325 s 0.00001480517998970754 s 0.86
add_two / IPartOpt / cpu / PostRev 0.000012961920001544058 s 0.000015164839996941735 s 0.85
add_two / IPartOpt / cpu / BothRev 0.000012855780023528497 s 0.000015474340025320998 s 0.83
add_two / DefOpt / cpu / PreRev 0.000012846979989262764 s 0.00001505929999439104 s 0.85
add_two / DefOpt / cpu / PostRev 0.000012838180027756608 s 0.00001657381996665208 s 0.77
add_two / DefOpt / cpu / BothRev 0.00001295441993534041 s 0.000015267580029103555 s 0.85
add_two / IDefOpt / cpu / PreRev 0.000013023120000070777 s 0.000015865159994064016 s 0.82
add_two / IDefOpt / cpu / PostRev 0.000012931779983773596 s 0.00001576899999236048 s 0.82
add_two / IDefOpt / cpu / BothRev 0.000012723939980787692 s 0.000015866280000409462 s 0.80
add_two / JaXPipe / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / Jax / cuda / Primal 0.000002431 s 0.000001888 s 1.29
add_two / HLOOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / PartOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / IPartOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / DefOpt / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / IDefOpt / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / JaXPipe / cuda / Forward 0.000010144 s 0.000009504 s 1.07
add_two / Jax / cuda / Forward 0.000010176 s 0.00000992 s 1.03
add_two / HLOOpt / cuda / Forward 0.00001056 s 0.000009888 s 1.07
add_two / PartOpt / cuda / Forward 0.000010304 s 0.000009153 s 1.13
add_two / IPartOpt / cuda / Forward 0.000010368 s 0.00000992 s 1.05
add_two / DefOpt / cuda / Forward 0.000010529 s 0.000009536 s 1.10
add_two / IDefOpt / cuda / Forward 0.0000104 s 0.000009568 s 1.09
add_two / JaXPipe / cuda / PreRev 0.000032416 s 0.000031585 s 1.03
add_two / JaXPipe / cuda / PostRev 0.000032224 s 0.000031264 s 1.03
add_two / JaXPipe / cuda / BothRev 0.000032896000000000005 s 0.000031744 s 1.04
add_two / Jax / cuda / BothRev 0.000034368 s 0.00003152 s 1.09
add_two / HLOOpt / cuda / PreRev 0.000033184 s 0.000032 s 1.04
add_two / HLOOpt / cuda / PostRev 0.000032127999999999995 s 0.000031424 s 1.02
add_two / HLOOpt / cuda / BothRev 0.000033152000000000004 s 0.000041184 s 0.80
add_two / PartOpt / cuda / PreRev 0.000033217000000000004 s 0.000031584 s 1.05
add_two / PartOpt / cuda / PostRev 0.00003248 s 0.000031392 s 1.03
add_two / PartOpt / cuda / BothRev 0.000032672 s 0.00003136 s 1.04
add_two / IPartOpt / cuda / PreRev 0.000032513 s 0.000032896000000000005 s 0.99
add_two / IPartOpt / cuda / PostRev 0.000033088 s 0.000031616 s 1.05
add_two / IPartOpt / cuda / BothRev 0.00003344 s 0.000032192 s 1.04
add_two / DefOpt / cuda / PreRev 0.000033248 s 0.000031776 s 1.05
add_two / DefOpt / cuda / PostRev 0.000033696 s 0.00003184 s 1.06
add_two / DefOpt / cuda / BothRev 0.000032641 s 0.000032608 s 1.00
add_two / IDefOpt / cuda / PreRev 0.00003248 s 0.000036225 s 0.90
add_two / IDefOpt / cuda / PostRev 0.000032608 s 0.000035904 s 0.91
add_two / IDefOpt / cuda / BothRev 0.000032864 s 0.000035327 s 0.93
add_two / JaXPipe / tpu / Primal 0.0000013928999999999998 s 0.0000014340249999999998 s 0.97
add_two / Jax / tpu / Primal 0.000001449075 s 0.000001485725 s 0.98
add_two / HLOOpt / tpu / Primal 0.000001398425 s 0.000001440225 s 0.97
add_two / PartOpt / tpu / Primal 0.0000014428 s 0.00000147715 s 0.98
add_two / IPartOpt / tpu / Primal 0.000001392675 s 0.000001432175 s 0.97
add_two / DefOpt / tpu / Primal 0.0000014501 s 0.00000147175 s 0.99
add_two / IDefOpt / tpu / Primal 0.0000013977999999999998 s 0.00000144305 s 0.97
add_two / JaXPipe / tpu / Forward 0.00000180475 s 0.000001820975 s 0.99
add_two / Jax / tpu / Forward 0.000001784525 s 0.000001826575 s 0.98
add_two / HLOOpt / tpu / Forward 0.000001797975 s 0.0000018237 s 0.99
add_two / PartOpt / tpu / Forward 0.000001798725 s 0.000001831525 s 0.98
add_two / IPartOpt / tpu / Forward 0.0000018036 s 0.0000018267 s 0.99
add_two / DefOpt / tpu / Forward 0.000001786525 s 0.000001830275 s 0.98
add_two / IDefOpt / tpu / Forward 0.00000180995 s 0.000001832075 s 0.99
add_two / JaXPipe / tpu / PreRev 0.00000280515 s 0.000002834025 s 0.99
add_two / JaXPipe / tpu / PostRev 0.000002729475 s 0.000002750375 s 0.99
add_two / JaXPipe / tpu / BothRev 0.0000027960000000000004 s 0.0000028336250000000003 s 0.99
add_two / Jax / tpu / BothRev 0.0000027284 s 0.0000027459250000000003 s 0.99
add_two / HLOOpt / tpu / PreRev 0.00000279215 s 0.0000028386 s 0.98
add_two / HLOOpt / tpu / PostRev 0.000002725525 s 0.000002751625 s 0.99
add_two / HLOOpt / tpu / BothRev 0.0000028001 s 0.0000028408 s 0.99
add_two / PartOpt / tpu / PreRev 0.000002717825 s 0.0000027445750000000003 s 0.99
add_two / PartOpt / tpu / PostRev 0.0000028081 s 0.000002841075 s 0.99
add_two / PartOpt / tpu / BothRev 0.0000027265 s 0.00000274485 s 0.99
add_two / IPartOpt / tpu / PreRev 0.0000028194 s 0.00000282995 s 1.00
add_two / IPartOpt / tpu / PostRev 0.0000027135500000000005 s 0.0000027511750000000004 s 0.99
add_two / IPartOpt / tpu / BothRev 0.000002790625 s 0.0000028333000000000005 s 0.98
add_two / DefOpt / tpu / PreRev 0.000002721825 s 0.000002745525 s 0.99
add_two / DefOpt / tpu / PostRev 0.0000027932 s 0.00000283105 s 0.99
add_two / DefOpt / tpu / BothRev 0.000002718775 s 0.000002757425 s 0.99
add_two / IDefOpt / tpu / PreRev 0.000002805375 s 0.000002837325 s 0.99
add_two / IDefOpt / tpu / PostRev 0.000002718975 s 0.000002747225 s 0.99
add_two / IDefOpt / tpu / BothRev 0.000002793175 s 0.00000283425 s 0.99
add_two / JaXPipe / cpu / Primal 0.000013597 s 0.00000840646003780421 s 1.62
add_two / Jax / cpu / Primal 0.000013326 s 0.000007408079991364502 s 1.80
add_two / HLOOpt / cpu / Primal 0.000013355 s 0.0000074600599691621025 s 1.79
add_two / PartOpt / cpu / Primal 0.000013372 s 0.000007752780065857223 s 1.72
add_two / IPartOpt / cpu / Primal 0.000013491 s 0.000008030020026126295 s 1.68
add_two / DefOpt / cpu / Primal 0.000013344999999999998 s 0.00000800805999460863 s 1.67
add_two / IDefOpt / cpu / Primal 0.000013477 s 0.000007063699977152282 s 1.91
add_two / JaXPipe / cpu / Forward 0.000018184 s 0.000011770780010920136 s 1.54
add_two / Jax / cpu / Forward 0.000018166 s 0.000011452759990788764 s 1.59
add_two / HLOOpt / cpu / Forward 0.000017753 s 0.00001166930002909794 s 1.52
add_two / PartOpt / cpu / Forward 0.000018666 s 0.000011410620018068583 s 1.64
add_two / IPartOpt / cpu / Forward 0.000018318 s 0.000011299139960101456 s 1.62
add_two / DefOpt / cpu / Forward 0.000018307 s 0.000011692319976646103 s 1.57
add_two / IDefOpt / cpu / Forward 0.000018499 s 0.000011732559996744386 s 1.58
add_two / JaXPipe / cpu / PreRev 0.000024296000000000003 s 0.000015161400006036274 s 1.60
add_two / JaXPipe / cpu / PostRev 0.000022818 s 0.00001595720002114831 s 1.43
add_two / JaXPipe / cpu / BothRev 0.000023748 s 0.00001547035995827173 s 1.54
add_two / Jax / cpu / BothRev 0.000023629 s 0.00001497119998020935 s 1.58
add_two / HLOOpt / cpu / PreRev 0.000023371 s 0.000015616819973729433 s 1.50
add_two / HLOOpt / cpu / PostRev 0.000023589 s 0.000016793039994809077 s 1.40
add_two / HLOOpt / cpu / BothRev 0.000023835 s 0.000015394640031445305 s 1.55
add_two / PartOpt / cpu / PreRev 0.00002343 s 0.00001552674003505672 s 1.51
add_two / PartOpt / cpu / PostRev 0.000023172 s 0.00001590923999174265 s 1.46
add_two / PartOpt / cpu / BothRev 0.000023075 s 0.0000157065999792394 s 1.47
add_two / IPartOpt / cpu / PreRev 0.000023531 s 0.00001480517998970754 s 1.59
add_two / IPartOpt / cpu / PostRev 0.000023549 s 0.000015164839996941735 s 1.55
add_two / IPartOpt / cpu / BothRev 0.000023607 s 0.000015474340025320998 s 1.53
add_two / DefOpt / cpu / PreRev 0.000023412 s 0.00001505929999439104 s 1.55
add_two / DefOpt / cpu / PostRev 0.000023508000000000003 s 0.00001657381996665208 s 1.42
add_two / DefOpt / cpu / BothRev 0.000023051 s 0.000015267580029103555 s 1.51
add_two / IDefOpt / cpu / PreRev 0.000023447 s 0.000015865159994064016 s 1.48
add_two / IDefOpt / cpu / PostRev 0.000023753 s 0.00001576899999236048 s 1.51
add_two / IDefOpt / cpu / BothRev 0.000023071 s 0.000015866280000409462 s 1.45
add_two / JaXPipe / cpu / Primal 0.000008 s 0.00000840646003780421 s 0.95
add_two / Jax / cpu / Primal 0.000008 s 0.000007408079991364502 s 1.08
add_two / HLOOpt / cpu / Primal 0.000008 s 0.0000074600599691621025 s 1.07
add_two / PartOpt / cpu / Primal 0.000008 s 0.000007752780065857223 s 1.03
add_two / IPartOpt / cpu / Primal 0.000008 s 0.000008030020026126295 s 1.00
add_two / DefOpt / cpu / Primal 0.000008 s 0.00000800805999460863 s 1.00
add_two / IDefOpt / cpu / Primal 0.000008 s 0.000007063699977152282 s 1.13
add_two / JaXPipe / cpu / Forward 0.000011 s 0.000011770780010920136 s 0.93
add_two / Jax / cpu / Forward 0.000011 s 0.000011452759990788764 s 0.96
add_two / HLOOpt / cpu / Forward 0.000011 s 0.00001166930002909794 s 0.94
add_two / PartOpt / cpu / Forward 0.000011 s 0.000011410620018068583 s 0.96
add_two / IPartOpt / cpu / Forward 0.000012 s 0.000011299139960101456 s 1.06
add_two / DefOpt / cpu / Forward 0.000011 s 0.000011692319976646103 s 0.94
add_two / IDefOpt / cpu / Forward 0.000011 s 0.000011732559996744386 s 0.94
add_two / JaXPipe / cpu / PreRev 0.000015 s 0.000015161400006036274 s 0.99
add_two / JaXPipe / cpu / PostRev 0.000015 s 0.00001595720002114831 s 0.94
add_two / JaXPipe / cpu / BothRev 0.000015 s 0.00001547035995827173 s 0.97
add_two / Jax / cpu / BothRev 0.000015 s 0.00001497119998020935 s 1.00
add_two / HLOOpt / cpu / PreRev 0.000015 s 0.000015616819973729433 s 0.96
add_two / HLOOpt / cpu / PostRev 0.000015 s 0.000016793039994809077 s 0.89
add_two / HLOOpt / cpu / BothRev 0.000016 s 0.000015394640031445305 s 1.04
add_two / PartOpt / cpu / PreRev 0.000015 s 0.00001552674003505672 s 0.97
add_two / PartOpt / cpu / PostRev 0.000016 s 0.00001590923999174265 s 1.01
add_two / PartOpt / cpu / BothRev 0.000015 s 0.0000157065999792394 s 0.96
add_two / IPartOpt / cpu / PreRev 0.000015 s 0.00001480517998970754 s 1.01
add_two / IPartOpt / cpu / PostRev 0.000016 s 0.000015164839996941735 s 1.06
add_two / IPartOpt / cpu / BothRev 0.000015 s 0.000015474340025320998 s 0.97
add_two / DefOpt / cpu / PreRev 0.000015 s 0.00001505929999439104 s 1.00
add_two / DefOpt / cpu / PostRev 0.000015 s 0.00001657381996665208 s 0.91
add_two / DefOpt / cpu / BothRev 0.000016 s 0.000015267580029103555 s 1.05
add_two / IDefOpt / cpu / PreRev 0.000014 s 0.000015865159994064016 s 0.88
add_two / IDefOpt / cpu / PostRev 0.000015 s 0.00001576899999236048 s 0.95
add_two / IDefOpt / cpu / BothRev 0.000015 s 0.000015866280000409462 s 0.95
cache / JaXPipe / cpu / Primal 0.000006256600008782698 s 0.000006617600029130699 s 0.95
cache / Jax / cpu / Primal 0.000006006480025462224 s 0.0000073614199845906115 s 0.82
cache / HLOOpt / cpu / Primal 0.0000063915600276232 s 0.000007327980010813917 s 0.87
cache / PartOpt / cpu / Primal 0.0000061406199984048725 s 0.000006596160001208773 s 0.93
cache / IPartOpt / cpu / Primal 0.000006156799991003936 s 0.000006951079985810793 s 0.89
cache / DefOpt / cpu / Primal 0.000005854619994352106 s 0.000006810800023231423 s 0.86
cache / IDefOpt / cpu / Primal 0.000005962479981462821 s 0.000006719999992128578 s 0.89
cache / JaXPipe / cpu / Forward 0.000013720919969273384 s 0.00001635375998375821 s 0.84
cache / Jax / cpu / Forward 0.00001534640001409571 s 0.00001622899997528293 s 0.95
cache / HLOOpt / cpu / Forward 0.000014300640004876188 s 0.00001638741993701842 s 0.87
cache / PartOpt / cpu / Forward 0.00001395569999658619 s 0.00001637195999137475 s 0.85
cache / IPartOpt / cpu / Forward 0.00001422688002094219 s 0.000016631440003038735 s 0.86
cache / DefOpt / cpu / Forward 0.000013628320020870888 s 0.00001649958000598417 s 0.83
cache / IDefOpt / cpu / Forward 0.000014044900008229887 s 0.00001611459998457576 s 0.87
cache / JaXPipe / cpu / PreRev 0.000014679859987154488 s 0.000017117220031650505 s 0.86
cache / JaXPipe / cpu / PostRev 0.00001929273998030112 s 0.00002205438004239113 s 0.87
cache / JaXPipe / cpu / BothRev 0.00001531138001155341 s 0.00001823537998461688 s 0.84
cache / Jax / cpu / BothRev 0.000018907179992311285 s 0.00002179658003115037 s 0.87
cache / HLOOpt / cpu / PreRev 0.000015244159994836082 s 0.00001804116001949296 s 0.84
cache / HLOOpt / cpu / PostRev 0.00001687624000624055 s 0.000020775400025740963 s 0.81
cache / HLOOpt / cpu / BothRev 0.00001534340000944212 s 0.000018207139974038 s 0.84
cache / PartOpt / cpu / PreRev 0.00001596848006556684 s 0.000017113039994001157 s 0.93
cache / PartOpt / cpu / PostRev 0.000019073059993388595 s 0.000022788399955970818 s 0.84
cache / PartOpt / cpu / BothRev 0.000015210919955279678 s 0.000017054319996532286 s 0.89
cache / IPartOpt / cpu / PreRev 0.000015050779993543985 s 0.00001702454002952436 s 0.88
cache / IPartOpt / cpu / PostRev 0.000019962739979746403 s 0.00002129747997969389 s 0.94
cache / IPartOpt / cpu / BothRev 0.000014616220014431746 s 0.00001820888001930143 s 0.80
cache / DefOpt / cpu / PreRev 0.00001507988003140781 s 0.000016552119950574708 s 0.91
cache / DefOpt / cpu / PostRev 0.000014382940007635624 s 0.00001614170001630555 s 0.89
cache / DefOpt / cpu / BothRev 0.000014729120011907071 s 0.000016149760012922344 s 0.91
cache / IDefOpt / cpu / PreRev 0.000014770259986107705 s 0.000016202219967453857 s 0.91
cache / IDefOpt / cpu / PostRev 0.00001471565998144797 s 0.000015889159994912917 s 0.93
cache / IDefOpt / cpu / BothRev 0.000015139400002226466 s 0.000016051540005719288 s 0.94
cache / JaXPipe / cuda / Primal 0.000002335 s 0.000002303 s 1.01
cache / Jax / cuda / Primal 0.000002335 s 0.000002304 s 1.01
cache / HLOOpt / cuda / Primal 0.000002335 s 0.000002208 s 1.06
cache / PartOpt / cuda / Primal 0.000002336 s 0.000002207 s 1.06
cache / IPartOpt / cuda / Primal 0.000002335 s 0.000002303 s 1.01
cache / DefOpt / cuda / Primal 0.000002336 s 0.000002272 s 1.03
cache / IDefOpt / cuda / Primal 0.000002335 s 0.00000224 s 1.04
cache / JaXPipe / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / Jax / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / HLOOpt / cuda / Forward 0.000002368 s 0.000002335 s 1.01
cache / PartOpt / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / IPartOpt / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / DefOpt / cuda / Forward 0.0000023670000000000004 s 0.000002272 s 1.04
cache / IDefOpt / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / JaXPipe / cuda / PreRev 0.000010784 s 0.00001088 s 0.99
cache / JaXPipe / cuda / PostRev 0.00001056 s 0.0000112 s 0.94
cache / JaXPipe / cuda / BothRev 0.000010784 s 0.0000112 s 0.96
cache / Jax / cuda / BothRev 0.00001072 s 0.000010784 s 0.99
cache / HLOOpt / cuda / PreRev 0.000013664 s 0.000013536 s 1.01
cache / HLOOpt / cuda / PostRev 0.000013632 s 0.000013505 s 1.01
cache / HLOOpt / cuda / BothRev 0.000013664 s 0.000013569 s 1.01
cache / PartOpt / cuda / PreRev 0.000010592 s 0.00001104 s 0.96
cache / PartOpt / cuda / PostRev 0.000010816 s 0.000010913 s 0.99
cache / PartOpt / cuda / BothRev 0.000010624 s 0.000010847 s 0.98
cache / IPartOpt / cuda / PreRev 0.00001056 s 0.000010976 s 0.96
cache / IPartOpt / cuda / PostRev 0.000010688 s 0.000011167 s 0.96
cache / IPartOpt / cuda / BothRev 0.000010656 s 0.000010657 s 1.00
cache / DefOpt / cuda / PreRev 0.000010624 s 0.00001104 s 0.96
cache / DefOpt / cuda / PostRev 0.000010591 s 0.000010753 s 0.98
cache / DefOpt / cuda / BothRev 0.000010688 s 0.000010815 s 0.99
cache / IDefOpt / cuda / PreRev 0.000010529 s 0.000010912 s 0.96
cache / IDefOpt / cuda / PostRev 0.00001056 s 0.000010656 s 0.99
cache / IDefOpt / cuda / BothRev 0.000010752 s 0.000010431 s 1.03
cache / JaXPipe / tpu / Primal 0.0000024768 s 0.000002477725 s 1.00
cache / Jax / tpu / Primal 0.00000245935 s 0.000002473575 s 0.99
cache / HLOOpt / tpu / Primal 0.0000024585000000000003 s 0.00000247285 s 0.99
cache / PartOpt / tpu / Primal 0.00000246345 s 0.0000024729 s 1.00
cache / IPartOpt / tpu / Primal 0.0000024641750000000004 s 0.0000024635500000000003 s 1.00
cache / DefOpt / tpu / Primal 0.0000024569750000000003 s 0.0000024609000000000004 s 1.00
cache / IDefOpt / tpu / Primal 0.000002465675 s 0.000002476775 s 1.00
cache / JaXPipe / tpu / Forward 0.000003553725 s 0.0000035618 s 1.00
cache / Jax / tpu / Forward 0.00000351975 s 0.00000355105 s 0.99
cache / HLOOpt / tpu / Forward 0.0000035464 s 0.000003544875 s 1.00
cache / PartOpt / tpu / Forward 0.000003526075 s 0.000003535125 s 1.00
cache / IPartOpt / tpu / Forward 0.000003549875 s 0.000003558 s 1.00
cache / DefOpt / tpu / Forward 0.000003526525 s 0.0000035295500000000004 s 1.00
cache / IDefOpt / tpu / Forward 0.0000035530750000000003 s 0.0000035562 s 1.00
cache / JaXPipe / tpu / PreRev 0.000004934375 s 0.000004968549999999999 s 0.99
cache / JaXPipe / tpu / PostRev 0.000005019175 s 0.000004967025 s 1.01
cache / JaXPipe / tpu / BothRev 0.000005005725 s 0.00000499125 s 1.00
cache / Jax / tpu / BothRev 0.0000049916 s 0.000004999 s 1.00
cache / HLOOpt / tpu / PreRev 0.0000041122 s 0.0000039558750000000005 s 1.04
cache / HLOOpt / tpu / PostRev 0.00000414705 s 0.000004117675 s 1.01
cache / HLOOpt / tpu / BothRev 0.0000041355250000000005 s 0.0000039434 s 1.05
cache / PartOpt / tpu / PreRev 0.0000049984 s 0.000004983625 s 1.00
cache / PartOpt / tpu / PostRev 0.0000049926 s 0.000004951725 s 1.01
cache / PartOpt / tpu / BothRev 0.000004992625 s 0.000004973375 s 1.00
cache / IPartOpt / tpu / PreRev 0.000004987025 s 0.000004962074999999999 s 1.01
cache / IPartOpt / tpu / PostRev 0.000005020175 s 0.000004954775 s 1.01
cache / IPartOpt / tpu / BothRev 0.000004968825 s 0.00000498675 s 1.00
cache / DefOpt / tpu / PreRev 0.000005005524999999999 s 0.000004984925 s 1.00
cache / DefOpt / tpu / PostRev 0.000004961674999999999 s 0.000004961 s 1.00
cache / DefOpt / tpu / BothRev 0.000004990474999999999 s 0.0000049668000000000005 s 1.00
cache / IDefOpt / tpu / PreRev 0.000004990675 s 0.00000496935 s 1.00
cache / IDefOpt / tpu / PostRev 0.0000049899 s 0.0000049644 s 1.01
cache / IDefOpt / tpu / BothRev 0.000004958675 s 0.000004968675 s 1.00
cache / JaXPipe / cpu / Primal 0.000012802 s 0.000006617600029130699 s 1.93
cache / Jax / cpu / Primal 0.000012442 s 0.0000073614199845906115 s 1.69
cache / HLOOpt / cpu / Primal 0.000012475 s 0.000007327980010813917 s 1.70
cache / PartOpt / cpu / Primal 0.000012713 s 0.000006596160001208773 s 1.93
cache / IPartOpt / cpu / Primal 0.000012672 s 0.000006951079985810793 s 1.82
cache / DefOpt / cpu / Primal 0.00001262 s 0.000006810800023231423 s 1.85
cache / IDefOpt / cpu / Primal 0.000012869 s 0.000006719999992128578 s 1.92
cache / JaXPipe / cpu / Forward 0.00001788 s 0.00001635375998375821 s 1.09
cache / Jax / cpu / Forward 0.000017552 s 0.00001622899997528293 s 1.08
cache / HLOOpt / cpu / Forward 0.000016849 s 0.00001638741993701842 s 1.03
cache / PartOpt / cpu / Forward 0.000016788 s 0.00001637195999137475 s 1.03
cache / IPartOpt / cpu / Forward 0.000024017 s 0.000016631440003038735 s 1.44
cache / DefOpt / cpu / Forward 0.000025408 s 0.00001649958000598417 s 1.54
cache / IDefOpt / cpu / Forward 0.00002265 s 0.00001611459998457576 s 1.41
cache / JaXPipe / cpu / PreRev 0.000026066 s 0.000017117220031650505 s 1.52
cache / JaXPipe / cpu / PostRev 0.0000331 s 0.00002205438004239113 s 1.50
cache / JaXPipe / cpu / BothRev 0.000026196 s 0.00001823537998461688 s 1.44
cache / Jax / cpu / BothRev 0.000027988 s 0.00002179658003115037 s 1.28
cache / HLOOpt / cpu / PreRev 0.000025922 s 0.00001804116001949296 s 1.44
cache / HLOOpt / cpu / PostRev 0.00002637 s 0.000020775400025740963 s 1.27
cache / HLOOpt / cpu / BothRev 0.000018539 s 0.000018207139974038 s 1.02
cache / PartOpt / cpu / PreRev 0.000018939 s 0.000017113039994001157 s 1.11
cache / PartOpt / cpu / PostRev 0.000020512 s 0.000022788399955970818 s 0.90
cache / PartOpt / cpu / BothRev 0.00001961 s 0.000017054319996532286 s 1.15
cache / IPartOpt / cpu / PreRev 0.0000176 s 0.00001702454002952436 s 1.03
cache / IPartOpt / cpu / PostRev 0.000020575 s 0.00002129747997969389 s 0.97
cache / IPartOpt / cpu / BothRev 0.000018437 s 0.00001820888001930143 s 1.01
cache / DefOpt / cpu / PreRev 0.000034781 s 0.000016552119950574708 s 2.10
cache / DefOpt / cpu / PostRev 0.000025774 s 0.00001614170001630555 s 1.60
cache / DefOpt / cpu / BothRev 0.000024765 s 0.000016149760012922344 s 1.53
cache / IDefOpt / cpu / PreRev 0.000026711 s 0.000016202219967453857 s 1.65
cache / IDefOpt / cpu / PostRev 0.000025156 s 0.000015889159994912917 s 1.58
cache / IDefOpt / cpu / BothRev 0.000028418 s 0.000016051540005719288 s 1.77
cache / JaXPipe / cpu / Primal 0.000008 s 0.000006617600029130699 s 1.21
cache / Jax / cpu / Primal 0.000007 s 0.0000073614199845906115 s 0.95
cache / HLOOpt / cpu / Primal 0.000008 s 0.000007327980010813917 s 1.09
cache / PartOpt / cpu / Primal 0.000008 s 0.000006596160001208773 s 1.21
cache / IPartOpt / cpu / Primal 0.000008 s 0.000006951079985810793 s 1.15
cache / DefOpt / cpu / Primal 0.000008 s 0.000006810800023231423 s 1.17
cache / IDefOpt / cpu / Primal 0.000008 s 0.000006719999992128578 s 1.19
cache / JaXPipe / cpu / Forward 0.000041 s 0.00001635375998375821 s 2.51
cache / Jax / cpu / Forward 0.000011 s 0.00001622899997528293 s 0.68
cache / HLOOpt / cpu / Forward 0.00001 s 0.00001638741993701842 s 0.61
cache / PartOpt / cpu / Forward 0.00001 s 0.00001637195999137475 s 0.61
cache / IPartOpt / cpu / Forward 0.00001 s 0.000016631440003038735 s 0.60
cache / DefOpt / cpu / Forward 0.000052 s 0.00001649958000598417 s 3.15
cache / IDefOpt / cpu / Forward 0.000043 s 0.00001611459998457576 s 2.67
cache / JaXPipe / cpu / PreRev 0.000011 s 0.000017117220031650505 s 0.64
cache / JaXPipe / cpu / PostRev 0.000011 s 0.00002205438004239113 s 0.50
cache / JaXPipe / cpu / BothRev 0.00005 s 0.00001823537998461688 s 2.74
cache / Jax / cpu / BothRev 0.000011 s 0.00002179658003115037 s 0.50
cache / HLOOpt / cpu / PreRev 0.00001 s 0.00001804116001949296 s 0.55
cache / HLOOpt / cpu / PostRev 0.00001 s 0.000020775400025740963 s 0.48
cache / HLOOpt / cpu / BothRev 0.00001 s 0.000018207139974038 s 0.55
cache / PartOpt / cpu / PreRev 0.00001 s 0.000017113039994001157 s 0.58
cache / PartOpt / cpu / PostRev 0.000011 s 0.000022788399955970818 s 0.48
cache / PartOpt / cpu / BothRev 0.000011 s 0.000017054319996532286 s 0.64
cache / IPartOpt / cpu / PreRev 0.000043 s 0.00001702454002952436 s 2.53
cache / IPartOpt / cpu / PostRev 0.00001 s 0.00002129747997969389 s 0.47
cache / IPartOpt / cpu / BothRev 0.00001 s 0.00001820888001930143 s 0.55
cache / DefOpt / cpu / PreRev 0.000043 s 0.000016552119950574708 s 2.60
cache / DefOpt / cpu / PostRev 0.00001 s 0.00001614170001630555 s 0.62
cache / DefOpt / cpu / BothRev 0.000011 s 0.000016149760012922344 s 0.68
cache / IDefOpt / cpu / PreRev 0.000011 s 0.000016202219967453857 s 0.68
cache / IDefOpt / cpu / PostRev 0.00001 s 0.000015889159994912917 s 0.63
cache / IDefOpt / cpu / BothRev 0.000011 s 0.000016051540005719288 s 0.69
Concat / JaXPipe / cpu / Primal 0.000006732519959768979 s 0.000007270800006153877 s 0.93
Concat / Jax / cpu / Primal 0.000006773779996365192 s 0.000006953919983061496 s 0.97
Concat / HLOOpt / cpu / Primal 0.000006829520025348756 s 0.000007605120017615263 s 0.90
Concat / PartOpt / cpu / Primal 0.000006251119993976317 s 0.000006844539975645603 s 0.91
Concat / IPartOpt / cpu / Primal 0.000006552219974764739 s 0.000007298220016309642 s 0.90
Concat / DefOpt / cpu / Primal 0.000006889799997225055 s 0.000007233900023493334 s 0.95
Concat / IDefOpt / cpu / Primal 0.000006726220026394003 s 0.000007258480000018608 s 0.93
Concat / JaXPipe / cpu / Forward 0.000009803659977478674 s 0.000011428860007072216 s 0.86
Concat / Jax / cpu / Forward 0.000009392400006618118 s 0.00001162386000942206 s 0.81
Concat / HLOOpt / cpu / Forward 0.000009620579994589209 s 0.00001077782001630112 s 0.89
Concat / PartOpt / cpu / Forward 0.000009582820030118456 s 0.000011412619987822835 s 0.84
Concat / IPartOpt / cpu / Forward 0.000009423180017620323 s 0.000011144519994559232 s 0.85
Concat / DefOpt / cpu / Forward 0.00000948505998167093 s 0.000011024940040442745 s 0.86
Concat / IDefOpt / cpu / Forward 0.000009463579963266968 s 0.000011167399989062688 s 0.85
Concat / JaXPipe / cpu / PreRev 0.00001085009995222208 s 0.000012688919996435288 s 0.86
Concat / JaXPipe / cpu / PostRev 0.000010751379995781465 s 0.00001220757997543842 s 0.88
Concat / JaXPipe / cpu / BothRev 0.000010727560011218885 s 0.00001241400000253634 s 0.86
Concat / Jax / cpu / BothRev 0.000010791539989440934 s 0.000012415119981596944 s 0.87
Concat / HLOOpt / cpu / PreRev 0.000011173459988640388 s 0.00001305383997532772 s 0.86
Concat / HLOOpt / cpu / PostRev 0.000012898599989057402 s 0.000014871699977447862 s 0.87
Concat / HLOOpt / cpu / BothRev 0.00001071063999006583 s 0.000012056320047122426 s 0.89
Concat / PartOpt / cpu / PreRev 0.00001112858000851702 s 0.00001218095996591728 s 0.91
Concat / PartOpt / cpu / PostRev 0.000010874960007640766 s 0.00001255646002391586 s 0.87
Concat / PartOpt / cpu / BothRev 0.000011371899972800747 s 0.000012368560010145304 s 0.92
Concat / IPartOpt / cpu / PreRev 0.000010977059955621372 s 0.000013087139996059704 s 0.84
Concat / IPartOpt / cpu / PostRev 0.00001090472003852483 s 0.00001286647993765655 s 0.85
Concat / IPartOpt / cpu / BothRev 0.000011145799962832824 s 0.000012555660050566076 s 0.89
Concat / DefOpt / cpu / PreRev 0.000010860520005735452 s 0.000012576600011016123 s 0.86
Concat / DefOpt / cpu / PostRev 0.00001120167997214594 s 0.000012553639990073862 s 0.89
Concat / DefOpt / cpu / BothRev 0.00001090605996978411 s 0.00001252969997949549 s 0.87
Concat / IDefOpt / cpu / PreRev 0.00001082606003365072 s 0.000012443119985618976 s 0.87
Concat / IDefOpt / cpu / PostRev 0.00001074871998753224 s 0.0000129725800070446 s 0.83
Concat / IDefOpt / cpu / BothRev 0.00001068490002580802 s 0.00001261953995708609 s 0.85
Concat / JaXPipe / cuda / Primal 0.000002464 s 0.0000019200000000000003 s 1.28
Concat / Jax / cuda / Primal 0.000002464 s 0.0000019200000000000003 s 1.28
Concat / HLOOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / PartOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / IPartOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / DefOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / IDefOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / JaXPipe / cuda / Forward 0.00001088 s 0.00000976 s 1.11
Concat / Jax / cuda / Forward 0.000010496 s 0.000010112 s 1.04
Concat / HLOOpt / cuda / Forward 0.000010752 s 0.000010368 s 1.04
Concat / PartOpt / cuda / Forward 0.00001056 s 0.000010176 s 1.04
Concat / IPartOpt / cuda / Forward 0.000012992 s 0.000010304 s 1.26
Concat / DefOpt / cuda / Forward 0.000010656 s 0.000009952 s 1.07
Concat / IDefOpt / cuda / Forward 0.0000104 s 0.000010432 s 1.00
Concat / JaXPipe / cuda / PreRev 0.000016704 s 0.00001584 s 1.05
Concat / JaXPipe / cuda / PostRev 0.000016544 s 0.000016 s 1.03
Concat / JaXPipe / cuda / BothRev 0.000016352 s 0.000016512 s 0.99
Concat / Jax / cuda / BothRev 0.000016417000000000002 s 0.00001616 s 1.02
Concat / HLOOpt / cuda / PreRev 0.000016576000000000002 s 0.000015808 s 1.05
Concat / HLOOpt / cuda / PostRev 0.00001664 s 0.000015776 s 1.05
Concat / HLOOpt / cuda / BothRev 0.000016608 s 0.000016225 s 1.02
Concat / PartOpt / cuda / PreRev 0.000016768000000000003 s 0.000016352 s 1.03
Concat / PartOpt / cuda / PostRev 0.00001696 s 0.000016 s 1.06
Concat / PartOpt / cuda / BothRev 0.000016768000000000003 s 0.000016096 s 1.04
Concat / IPartOpt / cuda / PreRev 0.000016864 s 0.000015968 s 1.06
Concat / IPartOpt / cuda / PostRev 0.000016096 s 0.000015935999999999998 s 1.01
Concat / IPartOpt / cuda / BothRev 0.00001648 s 0.000016128 s 1.02
Concat / DefOpt / cuda / PreRev 0.00001696 s 0.000016096 s 1.05
Concat / DefOpt / cuda / PostRev 0.000016704 s 0.000016224 s 1.03
Concat / DefOpt / cuda / BothRev 0.000016672 s 0.000015809 s 1.05
Concat / IDefOpt / cuda / PreRev 0.00001632 s 0.000016224 s 1.01
Concat / IDefOpt / cuda / PostRev 0.00001648 s 0.000016448999999999998 s 1.00
Concat / IDefOpt / cuda / BothRev 0.00001632 s 0.000016383999999999998 s 1.00
Concat / JaXPipe / tpu / Primal 0.000001522 s 0.00000152745 s 1.00
Concat / Jax / tpu / Primal 0.00000151885 s 0.0000015306 s 0.99
Concat / HLOOpt / tpu / Primal 0.00000152085 s 0.00000153055 s 0.99
Concat / PartOpt / tpu / Primal 0.00000151055 s 0.00000153385 s 0.98
Concat / IPartOpt / tpu / Primal 0.00000152195 s 0.0000015226500000000002 s 1.00
Concat / DefOpt / tpu / Primal 0.0000015233749999999998 s 0.00000152385 s 1.00
Concat / IDefOpt / tpu / Primal 0.0000015319250000000002 s 0.000001522975 s 1.01
Concat / JaXPipe / tpu / Forward 0.0000015598 s 0.000001585125 s 0.98
Concat / Jax / tpu / Forward 0.0000015626499999999998 s 0.00000155495 s 1.00
Concat / HLOOpt / tpu / Forward 0.0000015647 s 0.000001570125 s 1.00
Concat / PartOpt / tpu / Forward 0.00000154975 s 0.000001559625 s 0.99
Concat / IPartOpt / tpu / Forward 0.00000156385 s 0.0000015803249999999998 s 0.99
Concat / DefOpt / tpu / Forward 0.000001563375 s 0.00000156675 s 1.00
Concat / IDefOpt / tpu / Forward 0.000001577525 s 0.0000015851499999999998 s 1.00
Concat / JaXPipe / tpu / PreRev 0.0000020271 s 0.0000019909750000000004 s 1.02
Concat / JaXPipe / tpu / PostRev 0.0000019905000000000003 s 0.0000020868 s 0.95
Concat / JaXPipe / tpu / BothRev 0.0000020238 s 0.0000019944 s 1.01
Concat / Jax / tpu / BothRev 0.000001990675 s 0.000002074825 s 0.96
Concat / HLOOpt / tpu / PreRev 0.00000202365 s 0.00000199615 s 1.01
Concat / HLOOpt / tpu / PostRev 0.0000019963 s 0.0000020738000000000004 s 0.96
Concat / HLOOpt / tpu / BothRev 0.000002023075 s 0.000002012075 s 1.01
Concat / PartOpt / tpu / PreRev 0.000001998425 s 0.000002072 s 0.96
Concat / PartOpt / tpu / PostRev 0.000002020725 s 0.00000199355 s 1.01
Concat / PartOpt / tpu / BothRev 0.0000019928 s 0.0000020755000000000003 s 0.96
Concat / IPartOpt / tpu / PreRev 0.000002032575 s 0.000002003575 s 1.01
Concat / IPartOpt / tpu / PostRev 0.000001992875 s 0.000002073975 s 0.96
Concat / IPartOpt / tpu / BothRev 0.000002017025 s 0.0000020022 s 1.01
Concat / DefOpt / tpu / PreRev 0.00000199065 s 0.00000207405 s 0.96
Concat / DefOpt / tpu / PostRev 0.000002028425 s 0.0000019901250000000003 s 1.02
Concat / DefOpt / tpu / BothRev 0.000001995225 s 0.00000206865 s 0.96
Concat / IDefOpt / tpu / PreRev 0.00000202755 s 0.00000199775 s 1.01
Concat / IDefOpt / tpu / PostRev 0.000002007825 s 0.0000020728 s 0.97
Concat / IDefOpt / tpu / BothRev 0.000002028275 s 0.0000019916 s 1.02
Concat / JaXPipe / cpu / Primal 0.000012855 s 0.000007270800006153877 s 1.77
Concat / Jax / cpu / Primal 0.000012772 s 0.000006953919983061496 s 1.84
Concat / HLOOpt / cpu / Primal 0.000013078 s 0.000007605120017615263 s 1.72
Concat / PartOpt / cpu / Primal 0.000012957 s 0.000006844539975645603 s 1.89
Concat / IPartOpt / cpu / Primal 0.00001331 s 0.000007298220016309642 s 1.82
Concat / DefOpt / cpu / Primal 0.00001291 s 0.000007233900023493334 s 1.78
Concat / IDefOpt / cpu / Primal 0.000012852 s 0.000007258480000018608 s 1.77
Concat / JaXPipe / cpu / Forward 0.000017828 s 0.000011428860007072216 s 1.56
Concat / Jax / cpu / Forward 0.00001761 s 0.00001162386000942206 s 1.51
Concat / HLOOpt / cpu / Forward 0.000018387 s 0.00001077782001630112 s 1.71
Concat / PartOpt / cpu / Forward 0.000017728 s 0.000011412619987822835 s 1.55
Concat / IPartOpt / cpu / Forward 0.000018037 s 0.000011144519994559232 s 1.62
Concat / DefOpt / cpu / Forward 0.000017559000000000002 s 0.000011024940040442745 s 1.59
Concat / IDefOpt / cpu / Forward 0.000017485 s 0.000011167399989062688 s 1.57
Concat / JaXPipe / cpu / PreRev 0.000020272 s 0.000012688919996435288 s 1.60
Concat / JaXPipe / cpu / PostRev 0.000019884 s 0.00001220757997543842 s 1.63
Concat / JaXPipe / cpu / BothRev 0.000019394 s 0.00001241400000253634 s 1.56
Concat / Jax / cpu / BothRev 0.000020069 s 0.000012415119981596944 s 1.62
Concat / HLOOpt / cpu / PreRev 0.000020483 s 0.00001305383997532772 s 1.57
Concat / HLOOpt / cpu / PostRev 0.000019517 s 0.000014871699977447862 s 1.31
Concat / HLOOpt / cpu / BothRev 0.000019168 s 0.000012056320047122426 s 1.59
Concat / PartOpt / cpu / PreRev 0.000020607 s 0.00001218095996591728 s 1.69
Concat / PartOpt / cpu / PostRev 0.000019358 s 0.00001255646002391586 s 1.54
Concat / PartOpt / cpu / BothRev 0.000019615 s 0.000012368560010145304 s 1.59
Concat / IPartOpt / cpu / PreRev 0.000020122 s 0.000013087139996059704 s 1.54
Concat / IPartOpt / cpu / PostRev 0.000019549 s 0.00001286647993765655 s 1.52
Concat / IPartOpt / cpu / BothRev 0.00001932 s 0.000012555660050566076 s 1.54
Concat / DefOpt / cpu / PreRev 0.000019916 s 0.000012576600011016123 s 1.58
Concat / DefOpt / cpu / PostRev 0.000020108 s 0.000012553639990073862 s 1.60
Concat / DefOpt / cpu / BothRev 0.000020018 s 0.00001252969997949549 s 1.60
Concat / IDefOpt / cpu / PreRev 0.00001989 s 0.000012443119985618976 s 1.60
Concat / IDefOpt / cpu / PostRev 0.000019411 s 0.0000129725800070446 s 1.50
Concat / IDefOpt / cpu / BothRev 0.000019635 s 0.00001261953995708609 s 1.56
Concat / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000007270800006153877 s 1.24
Concat / Jax / cpu / Primal 0.000008 s 0.000006953919983061496 s 1.15
Concat / HLOOpt / cpu / Primal 0.000008 s 0.000007605120017615263 s 1.05
Concat / PartOpt / cpu / Primal 0.000008 s 0.000006844539975645603 s 1.17
Concat / IPartOpt / cpu / Primal 0.000008 s 0.000007298220016309642 s 1.10
Concat / DefOpt / cpu / Primal 0.000008 s 0.000007233900023493334 s 1.11
Concat / IDefOpt / cpu / Primal 0.000008 s 0.000007258480000018608 s 1.10
Concat / JaXPipe / cpu / Forward 0.00001 s 0.000011428860007072216 s 0.87
Concat / Jax / cpu / Forward 0.00001 s 0.00001162386000942206 s 0.86
Concat / HLOOpt / cpu / Forward 0.000011 s 0.00001077782001630112 s 1.02
Concat / PartOpt / cpu / Forward 0.000011 s 0.000011412619987822835 s 0.96
Concat / IPartOpt / cpu / Forward 0.000011 s 0.000011144519994559232 s 0.99
Concat / DefOpt / cpu / Forward 0.000011 s 0.000011024940040442745 s 1.00
Concat / IDefOpt / cpu / Forward 0.000011 s 0.000011167399989062688 s 0.99
Concat / JaXPipe / cpu / PreRev 0.000012 s 0.000012688919996435288 s 0.95
Concat / JaXPipe / cpu / PostRev 0.000013 s 0.00001220757997543842 s 1.06
Concat / JaXPipe / cpu / BothRev 0.000013 s 0.00001241400000253634 s 1.05
Concat / Jax / cpu / BothRev 0.000013 s 0.000012415119981596944 s 1.05
Concat / HLOOpt / cpu / PreRev 0.000012 s 0.00001305383997532772 s 0.92
Concat / HLOOpt / cpu / PostRev 0.000013 s 0.000014871699977447862 s 0.87
Concat / HLOOpt / cpu / BothRev 0.000013 s 0.000012056320047122426 s 1.08
Concat / PartOpt / cpu / PreRev 0.000013 s 0.00001218095996591728 s 1.07
Concat / PartOpt / cpu / PostRev 0.000013 s 0.00001255646002391586 s 1.04
Concat / PartOpt / cpu / BothRev 0.000013 s 0.000012368560010145304 s 1.05
Concat / IPartOpt / cpu / PreRev 0.000013 s 0.000013087139996059704 s 0.99
Concat / IPartOpt / cpu / PostRev 0.000014 s 0.00001286647993765655 s 1.09
Concat / IPartOpt / cpu / BothRev 0.000014 s 0.000012555660050566076 s 1.12
Concat / DefOpt / cpu / PreRev 0.000013 s 0.000012576600011016123 s 1.03
Concat / DefOpt / cpu / PostRev 0.000013 s 0.000012553639990073862 s 1.04
Concat / DefOpt / cpu / BothRev 0.000014 s 0.00001252969997949549 s 1.12
Concat / IDefOpt / cpu / PreRev 0.000014 s 0.000012443119985618976 s 1.13
Concat / IDefOpt / cpu / PostRev 0.000013 s 0.0000129725800070446 s 1.00
Concat / IDefOpt / cpu / BothRev 0.000013 s 0.00001261953995708609 s 1.03
const_scatter / JaXPipe / cpu / Primal 0.000006466699996963143 s 0.0000066701799732982185 s 0.97
const_scatter / Jax / cpu / Primal 0.000006187819981278153 s 0.000006743799976902665 s 0.92
const_scatter / HLOOpt / cpu / Primal 0.000006724380000378005 s 0.000007621859967912315 s 0.88
const_scatter / PartOpt / cpu / Primal 0.000006026619985277648 s 0.000006892500023241155 s 0.87
const_scatter / IPartOpt / cpu / Primal 0.0000063824200060480505 s 0.000006738659985785489 s 0.95
const_scatter / DefOpt / cpu / Primal 0.000006766760016034823 s 0.0000071026000114216 s 0.95
const_scatter / IDefOpt / cpu / Primal 0.000006864439974378911 s 0.000007858320032028132 s 0.87
const_scatter / JaXPipe / cpu / Forward 0.00001008954003737017 s 0.000011870759990415536 s 0.85
const_scatter / Jax / cpu / Forward 0.000009278879997509647 s 0.000010619959966788884 s 0.87
const_scatter / HLOOpt / cpu / Forward 0.00001022983998154814 s 0.000011783159970946145 s 0.87
const_scatter / PartOpt / cpu / Forward 0.000010146700014956878 s 0.000011229000037928926 s 0.90
const_scatter / IPartOpt / cpu / Forward 0.000010265080018143637 s 0.000012111659980291734 s 0.85
const_scatter / DefOpt / cpu / Forward 0.00001029491997542209 s 0.000011104419972980396 s 0.93
const_scatter / IDefOpt / cpu / Forward 0.000009952979989975577 s 0.000012228739997226512 s 0.81
const_scatter / JaXPipe / cpu / PreRev 0.0002955759400265 s 0.0002900659799706 s 1.02
const_scatter / JaXPipe / cpu / PostRev 0.0002818762199967 s 0.0002828032600064 s 1.00
const_scatter / JaXPipe / cpu / BothRev 0.000282674480004 s 0.0002831334199981 s 1.00
const_scatter / Jax / cpu / BothRev 0.0002788815200256 s 0.0002821311999741 s 0.99
const_scatter / HLOOpt / cpu / PreRev 0.0002824387000146 s 0.0002847428000131 s 0.99
const_scatter / HLOOpt / cpu / PostRev 0.0002839483199932 s 0.0002864551800121 s 0.99
const_scatter / HLOOpt / cpu / BothRev 0.0002828984399275 s 0.0002834857999732 s 1.00
const_scatter / PartOpt / cpu / PreRev 0.0002828365799814 s 0.000282958500029 s 1.00
const_scatter / PartOpt / cpu / PostRev 0.0002775305400245 s 0.0002846486800353 s 0.97
const_scatter / PartOpt / cpu / BothRev 0.0002834331600115 s 0.0002827322999928 s 1.00
const_scatter / IPartOpt / cpu / PreRev 0.0002797878800174 s 0.0002970088200254 s 0.94
const_scatter / IPartOpt / cpu / PostRev 0.0002785776800374 s 0.0002822358000321 s 0.99
const_scatter / IPartOpt / cpu / BothRev 0.0002815161399576 s 0.0002828837800097 s 1.00
const_scatter / DefOpt / cpu / PreRev 0.0002787176600031 s 0.0002843984800074 s 0.98
const_scatter / DefOpt / cpu / PostRev 0.00028175112001 s 0.0002837708200422 s 0.99
const_scatter / DefOpt / cpu / BothRev 0.0002803370999845 s 0.0002855213999737 s 0.98
const_scatter / IDefOpt / cpu / PreRev 0.0002805964199797 s 0.0003056161200311 s 0.92
const_scatter / IDefOpt / cpu / PostRev 0.0002806790000158 s 0.0002853217799838 s 0.98
const_scatter / IDefOpt / cpu / BothRev 0.0002838792599959 s 0.0002853424200202 s 0.99
const_scatter / JaXPipe / cuda / Primal 0.000002432 s 0.000001887 s 1.29
const_scatter / Jax / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / HLOOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / PartOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / IPartOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / DefOpt / cuda / Primal 0.000002432 s 0.000001887 s 1.29
const_scatter / IDefOpt / cuda / Primal 0.000002464 s 0.000001887 s 1.31
const_scatter / JaXPipe / cuda / Forward 0.00001056 s 0.00000992 s 1.06
const_scatter / Jax / cuda / Forward 0.000010496 s 0.0000096 s 1.09
const_scatter / HLOOpt / cuda / Forward 0.00001056 s 0.00001008 s 1.05
const_scatter / PartOpt / cuda / Forward 0.000010464 s 0.000009857 s 1.06
const_scatter / IPartOpt / cuda / Forward 0.000010784 s 0.000010048 s 1.07
const_scatter / DefOpt / cuda / Forward 0.00001024 s 0.00001008 s 1.02
const_scatter / IDefOpt / cuda / Forward 0.000010496 s 0.000009856 s 1.06
const_scatter / JaXPipe / cuda / PreRev 0.000016768000000000003 s 0.000015968 s 1.05
const_scatter / JaXPipe / cuda / PostRev 0.000017472 s 0.000016063000000000002 s 1.09
const_scatter / JaXPipe / cuda / BothRev 0.0000176 s 0.000017984 s 0.98
const_scatter / Jax / cuda / BothRev 0.000017536 s 0.0000184 s 0.95
const_scatter / HLOOpt / cuda / PreRev 0.000016352 s 0.000016096 s 1.02
const_scatter / HLOOpt / cuda / PostRev 0.000016992 s 0.000015744 s 1.08
const_scatter / HLOOpt / cuda / BothRev 0.000022208 s 0.000016255999999999998 s 1.37
const_scatter / PartOpt / cuda / PreRev 0.000016608 s 0.000017857000000000002 s 0.93
const_scatter / PartOpt / cuda / PostRev 0.000016448000000000002 s 0.000016417000000000002 s 1.00
const_scatter / PartOpt / cuda / BothRev 0.00001648 s 0.000016416 s 1.00
const_scatter / IPartOpt / cuda / PreRev 0.000016288 s 0.00001616 s 1.01
const_scatter / IPartOpt / cuda / PostRev 0.000016768000000000003 s 0.000016544 s 1.01
const_scatter / IPartOpt / cuda / BothRev 0.000016255999999999998 s 0.000015904000000000002 s 1.02
const_scatter / DefOpt / cuda / PreRev 0.000016544 s 0.000016352 s 1.01
const_scatter / DefOpt / cuda / PostRev 0.000016768000000000003 s 0.000015935999999999998 s 1.05
const_scatter / DefOpt / cuda / BothRev 0.00002432 s 0.00001584 s 1.54
const_scatter / IDefOpt / cuda / PreRev 0.000018208 s 0.000021888 s 0.83
const_scatter / IDefOpt / cuda / PostRev 0.000017537 s 0.000015935999999999998 s 1.10
const_scatter / IDefOpt / cuda / BothRev 0.000017728 s 0.000016255999999999998 s 1.09
const_scatter / JaXPipe / tpu / Primal 0.000003804825 s 0.000003818225 s 1.00
const_scatter / Jax / tpu / Primal 0.0000038127 s 0.000003825575 s 1.00
const_scatter / HLOOpt / tpu / Primal 0.0000037991 s 0.000003793475 s 1.00
const_scatter / PartOpt / tpu / Primal 0.0000038382 s 0.000003842975 s 1.00
const_scatter / IPartOpt / tpu / Primal 0.00000378775 s 0.000003810875 s 0.99
const_scatter / DefOpt / tpu / Primal 0.000003832175 s 0.000003816975 s 1.00
const_scatter / IDefOpt / tpu / Primal 0.0000037782 s 0.000003805875 s 0.99
const_scatter / JaXPipe / tpu / Forward 0.000006490574999999999 s 0.00000650045 s 1.00
const_scatter / Jax / tpu / Forward 0.00000646425 s 0.000006493925 s 1.00
const_scatter / HLOOpt / tpu / Forward 0.000006492149999999999 s 0.0000064629250000000005 s 1.00
const_scatter / PartOpt / tpu / Forward 0.000006450875 s 0.00000649415 s 0.99
const_scatter / IPartOpt / tpu / Forward 0.000006478525000000001 s 0.000006451675 s 1.00
const_scatter / DefOpt / tpu / Forward 0.000006472250000000001 s 0.00000650245 s 1.00
const_scatter / IDefOpt / tpu / Forward 0.0000064861500000000006 s 0.000006453625 s 1.01
const_scatter / JaXPipe / tpu / PreRev 0.000006639975 s 0.00000668875 s 0.99
const_scatter / JaXPipe / tpu / PostRev 0.00000665285 s 0.000006669874999999999 s 1.00
const_scatter / JaXPipe / tpu / BothRev 0.000006638849999999999 s 0.000006674075000000001 s 0.99
const_scatter / Jax / tpu / BothRev 0.000006634975 s 0.00000666065 s 1.00
const_scatter / HLOOpt / tpu / PreRev 0.000006654775000000001 s 0.000006668475 s 1.00
const_scatter / HLOOpt / tpu / PostRev 0.000006642275 s 0.000006640350000000001 s 1.00
const_scatter / HLOOpt / tpu / BothRev 0.000006631525 s 0.000006670825 s 0.99
const_scatter / PartOpt / tpu / PreRev 0.000006657175 s 0.000006679725 s 1.00
const_scatter / PartOpt / tpu / PostRev 0.000006634875 s 0.0000066626 s 1.00
const_scatter / PartOpt / tpu / BothRev 0.000006645775000000001 s 0.00000665535 s 1.00
const_scatter / IPartOpt / tpu / PreRev 0.00000664235 s 0.000006675325 s 1.00
const_scatter / IPartOpt / tpu / PostRev 0.000006641725 s 0.0000066698 s 1.00
const_scatter / IPartOpt / tpu / BothRev 0.00000664745 s 0.00000667635 s 1.00
const_scatter / DefOpt / tpu / PreRev 0.00000666955 s 0.0000066897 s 1.00
const_scatter / DefOpt / tpu / PostRev 0.000006623225 s 0.000006663175 s 0.99
const_scatter / DefOpt / tpu / BothRev 0.000006638150000000001 s 0.00000667775 s 0.99
const_scatter / IDefOpt / tpu / PreRev 0.000006630525000000001 s 0.000006667025 s 0.99
const_scatter / IDefOpt / tpu / PostRev 0.0000066329000000000005 s 0.000006663125 s 1.00
const_scatter / IDefOpt / tpu / BothRev 0.000006652525 s 0.000006661175 s 1.00
const_scatter / JaXPipe / cpu / Primal 0.00001317 s 0.0000066701799732982185 s 1.97
const_scatter / Jax / cpu / Primal 0.00001281 s 0.000006743799976902665 s 1.90
const_scatter / HLOOpt / cpu / Primal 0.000013461 s 0.000007621859967912315 s 1.77
const_scatter / PartOpt / cpu / Primal 0.000012747 s 0.000006892500023241155 s 1.85
const_scatter / IPartOpt / cpu / Primal 0.000012951 s 0.000006738659985785489 s 1.92
const_scatter / DefOpt / cpu / Primal 0.000013368 s 0.0000071026000114216 s 1.88
const_scatter / IDefOpt / cpu / Primal 0.00001348 s 0.000007858320032028132 s 1.72
const_scatter / JaXPipe / cpu / Forward 0.000018319 s 0.000011870759990415536 s 1.54
const_scatter / Jax / cpu / Forward 0.000016718 s 0.000010619959966788884 s 1.57
const_scatter / HLOOpt / cpu / Forward 0.00001852 s 0.000011783159970946145 s 1.57
const_scatter / PartOpt / cpu / Forward 0.000018029 s 0.000011229000037928926 s 1.61
const_scatter / IPartOpt / cpu / Forward 0.000018527 s 0.000012111659980291734 s 1.53
const_scatter / DefOpt / cpu / Forward 0.000017943 s 0.000011104419972980396 s 1.62
const_scatter / IDefOpt / cpu / Forward 0.000018061 s 0.000012228739997226512 s 1.48
const_scatter / JaXPipe / cpu / PreRev 0.000520437 s 0.0002900659799706 s 1.79
const_scatter / JaXPipe / cpu / PostRev 0.0005250169999999 s 0.0002828032600064 s 1.86
const_scatter / JaXPipe / cpu / BothRev 0.000538249 s 0.0002831334199981 s 1.90
const_scatter / Jax / cpu / BothRev 0.000544145 s 0.0002821311999741 s 1.93
const_scatter / HLOOpt / cpu / PreRev 0.000528589 s 0.0002847428000131 s 1.86
const_scatter / HLOOpt / cpu / PostRev 0.000500648 s 0.0002864551800121 s 1.75
const_scatter / HLOOpt / cpu / BothRev 0.000507856 s 0.0002834857999732 s 1.79
const_scatter / PartOpt / cpu / PreRev 0.000525286 s 0.000282958500029 s 1.86
const_scatter / PartOpt / cpu / PostRev 0.000527123 s 0.0002846486800353 s 1.85
const_scatter / PartOpt / cpu / BothRev 0.000518336 s 0.0002827322999928 s 1.83
const_scatter / IPartOpt / cpu / PreRev 0.000520841 s 0.0002970088200254 s 1.75
const_scatter / IPartOpt / cpu / PostRev 0.000521988 s 0.0002822358000321 s 1.85
const_scatter / IPartOpt / cpu / BothRev 0.000525766 s 0.0002828837800097 s 1.86
const_scatter / DefOpt / cpu / PreRev 0.000519819 s 0.0002843984800074 s 1.83
const_scatter / DefOpt / cpu / PostRev 0.000527051 s 0.0002837708200422 s 1.86
const_scatter / DefOpt / cpu / BothRev 0.000532401 s 0.0002855213999737 s 1.86
const_scatter / IDefOpt / cpu / PreRev 0.0005223179999999 s 0.0003056161200311 s 1.71
const_scatter / IDefOpt / cpu / PostRev 0.000541482 s 0.0002853217799838 s 1.90
const_scatter / IDefOpt / cpu / BothRev 0.00052007 s 0.0002853424200202 s 1.82
const_scatter / JaXPipe / cpu / Primal 0.000008 s 0.0000066701799732982185 s 1.20
const_scatter / Jax / cpu / Primal 0.000008 s 0.000006743799976902665 s 1.19
const_scatter / HLOOpt / cpu / Primal 0.000008 s 0.000007621859967912315 s 1.05
const_scatter / PartOpt / cpu / Primal 0.000008 s 0.000006892500023241155 s 1.16
const_scatter / IPartOpt / cpu / Primal 0.000008 s 0.000006738659985785489 s 1.19
const_scatter / DefOpt / cpu / Primal 0.000008 s 0.0000071026000114216 s 1.13
const_scatter / IDefOpt / cpu / Primal 0.000008 s 0.000007858320032028132 s 1.02
const_scatter / JaXPipe / cpu / Forward 0.000013 s 0.000011870759990415536 s 1.10
const_scatter / Jax / cpu / Forward 0.000012 s 0.000010619959966788884 s 1.13
const_scatter / HLOOpt / cpu / Forward 0.000012 s 0.000011783159970946145 s 1.02
const_scatter / PartOpt / cpu / Forward 0.000012 s 0.000011229000037928926 s 1.07
const_scatter / IPartOpt / cpu / Forward 0.000012 s 0.000012111659980291734 s 0.99
const_scatter / DefOpt / cpu / Forward 0.000012 s 0.000011104419972980396 s 1.08
const_scatter / IDefOpt / cpu / Forward 0.000012 s 0.000012228739997226512 s 0.98
const_scatter / JaXPipe / cpu / PreRev 0.0003439999999999 s 0.0002900659799706 s 1.19
const_scatter / JaXPipe / cpu / PostRev 0.000342 s 0.0002828032600064 s 1.21
const_scatter / JaXPipe / cpu / BothRev 0.00035 s 0.0002831334199981 s 1.24
const_scatter / Jax / cpu / BothRev 0.000348 s 0.0002821311999741 s 1.23
const_scatter / HLOOpt / cpu / PreRev 0.000387 s 0.0002847428000131 s 1.36
const_scatter / HLOOpt / cpu / PostRev 0.000378 s 0.0002864551800121 s 1.32
const_scatter / HLOOpt / cpu / BothRev 0.000365 s 0.0002834857999732 s 1.29
const_scatter / PartOpt / cpu / PreRev 0.00032 s 0.000282958500029 s 1.13
const_scatter / PartOpt / cpu / PostRev 0.00033 s 0.0002846486800353 s 1.16
const_scatter / PartOpt / cpu / BothRev 0.000363 s 0.0002827322999928 s 1.28
const_scatter / IPartOpt / cpu / PreRev 0.000358 s 0.0002970088200254 s 1.21
const_scatter / IPartOpt / cpu / PostRev 0.000321 s 0.0002822358000321 s 1.14
const_scatter / IPartOpt / cpu / BothRev 0.000332 s 0.0002828837800097 s 1.17
const_scatter / DefOpt / cpu / PreRev 0.000376 s 0.0002843984800074 s 1.32
const_scatter / DefOpt / cpu / PostRev 0.000336 s 0.0002837708200422 s 1.18
const_scatter / DefOpt / cpu / BothRev 0.000359 s 0.0002855213999737 s 1.26
const_scatter / IDefOpt / cpu / PreRev 0.000345 s 0.0003056161200311 s 1.13
const_scatter / IDefOpt / cpu / PostRev 0.000338 s 0.0002853217799838 s 1.18
const_scatter / IDefOpt / cpu / BothRev 0.000328 s 0.0002853424200202 s 1.15
GenDot / JaXPipe / cpu / Primal 0.000006651920029980829 s 0.000008296160003737896 s 0.80
GenDot / Jax / cpu / Primal 0.0000065009599620680095 s 0.000008259619971795473 s 0.79
GenDot / HLOOpt / cpu / Primal 0.000007147499982238514 s 0.000008371400035684928 s 0.85
GenDot / PartOpt / cpu / Primal 0.00000666674000058265 s 0.00000901005999367044 s 0.74
GenDot / IPartOpt / cpu / Primal 0.000006770139998479863 s 0.000008573640043323394 s 0.79
GenDot / DefOpt / cpu / Primal 0.000007250139960888191 s 0.000007987099979800406 s 0.91
GenDot / IDefOpt / cpu / Primal 0.000007028680029179668 s 0.000007670840022910853 s 0.92
GenDot / JaXPipe / cpu / Forward 0.000010201840004810948 s 0.000011564219994397715 s 0.88
GenDot / Jax / cpu / Forward 0.00000963145999776316 s 0.000011687760006680036 s 0.82
GenDot / HLOOpt / cpu / Forward 0.000010538880014792084 s 0.000012549280017992717 s 0.84
GenDot / PartOpt / cpu / Forward 0.00001054524000210222 s 0.000011167880002176387 s 0.94
GenDot / IPartOpt / cpu / Forward 0.000010787900046125286 s 0.000012273299962544117 s 0.88
GenDot / DefOpt / cpu / Forward 0.000010112219988513972 s 0.000011512759956531226 s 0.88
GenDot / IDefOpt / cpu / Forward 0.000010144239977307736 s 0.000012003320016447103 s 0.85
GenDot / JaXPipe / cpu / PreRev 0.00001083381997887045 s 0.000012277580008230873 s 0.88
GenDot / JaXPipe / cpu / PostRev 0.000009711479988254725 s 0.000011590839994823907 s 0.84
GenDot / JaXPipe / cpu / BothRev 0.000010529719993428444 s 0.000011976359992331707 s 0.88
GenDot / Jax / cpu / BothRev 0.000009719720019347733 s 0.000011309560022709776 s 0.86
GenDot / HLOOpt / cpu / PreRev 0.000010553560005064356 s 0.000012416999970810138 s 0.85
GenDot / HLOOpt / cpu / PostRev 0.000012213080008223188 s 0.00001339757995083346 s 0.91
GenDot / HLOOpt / cpu / BothRev 0.000010457799980940764 s 0.000011228120001760544 s 0.93
GenDot / PartOpt / cpu / PreRev 0.000010580940015643137 s 0.000011987920024694176 s 0.88
GenDot / PartOpt / cpu / PostRev 0.000010094020017277216 s 0.00001231975996233814 s 0.82
GenDot / PartOpt / cpu / BothRev 0.00001106197999433789 s 0.000012321479962338344 s 0.90
GenDot / IPartOpt / cpu / PreRev 0.000010256320028929624 s 0.000011272580013610423 s 0.91
GenDot / IPartOpt / cpu / PostRev 0.00000995344001239573 s 0.000012042980015394278 s 0.83
GenDot / IPartOpt / cpu / BothRev 0.000010253199970975402 s 0.000011842540016004931 s 0.87
GenDot / DefOpt / cpu / PreRev 0.000010731260026659583 s 0.000011467599997558865 s 0.94
GenDot / DefOpt / cpu / PostRev 0.000010164579998672707 s 0.000011891319991264026 s 0.85
GenDot / DefOpt / cpu / BothRev 0.000010537980015215 s 0.000012002500052403774 s 0.88
GenDot / IDefOpt / cpu / PreRev 0.000010817579977810966 s 0.000011844639984701644 s 0.91
GenDot / IDefOpt / cpu / PostRev 0.000010634140016918536 s 0.000012110100014979252 s 0.88
GenDot / IDefOpt / cpu / BothRev 0.000010440200021548664 s 0.000012075799977537829 s 0.86
GenDot / JaXPipe / cuda / Primal 0.000002528 s 0.000002015 s 1.25
GenDot / Jax / cuda / Primal 0.000002527 s 0.000002016 s 1.25
GenDot / HLOOpt / cuda / Primal 0.000002527 s 0.000001984 s 1.27
GenDot / PartOpt / cuda / Primal 0.000002529 s 0.000002015 s 1.26
GenDot / IPartOpt / cuda / Primal 0.00000256 s 0.000002015 s 1.27
GenDot / DefOpt / cuda / Primal 0.000002528 s 0.000001984 s 1.27
GenDot / IDefOpt / cuda / Primal 0.000002528 s 0.000001984 s 1.27
GenDot / JaXPipe / cuda / Forward 0.00001056 s 0.000009792 s 1.08
GenDot / Jax / cuda / Forward 0.000010785 s 0.000010016 s 1.08
GenDot / HLOOpt / cuda / Forward 0.000010624 s 0.000009952 s 1.07
GenDot / PartOpt / cuda / Forward 0.000010784 s 0.00000992 s 1.09
GenDot / IPartOpt / cuda / Forward 0.00001088 s 0.000009792 s 1.11
GenDot / DefOpt / cuda / Forward 0.000010912 s 0.000009632 s 1.13
GenDot / IDefOpt / cuda / Forward 0.000010592 s 0.00001008 s 1.05
GenDot / JaXPipe / cuda / PreRev 0.000009472 s 0.00000992 s 0.95
GenDot / JaXPipe / cuda / PostRev 0.00001072 s 0.000009984 s 1.07
GenDot / JaXPipe / cuda / BothRev 0.000010591 s 0.000009952 s 1.06
GenDot / Jax / cuda / BothRev 0.000010656 s 0.000009952 s 1.07
GenDot / HLOOpt / cuda / PreRev 0.00001056 s 0.000009792 s 1.08
GenDot / HLOOpt / cuda / PostRev 0.000010752 s 0.000010464 s 1.03
GenDot / HLOOpt / cuda / BothRev 0.000010944 s 0.00000992 s 1.10
GenDot / PartOpt / cuda / PreRev 0.000010656 s 0.000009855 s 1.08
GenDot / PartOpt / cuda / PostRev 0.000010624 s 0.000009984 s 1.06
GenDot / PartOpt / cuda / BothRev 0.000010751 s 0.000010016 s 1.07
GenDot / IPartOpt / cuda / PreRev 0.00001056 s 0.000009888 s 1.07
GenDot / IPartOpt / cuda / PostRev 0.000010272 s 0.000010145 s 1.01
GenDot / IPartOpt / cuda / BothRev 0.000010527 s 0.000009664 s 1.09
GenDot / DefOpt / cuda / PreRev 0.0000104 s 0.000010016 s 1.04
GenDot / DefOpt / cuda / PostRev 0.000010656 s 0.000009665 s 1.10
GenDot / DefOpt / cuda / BothRev 0.000010656 s 0.000009856 s 1.08
GenDot / IDefOpt / cuda / PreRev 0.000010592 s 0.000009984 s 1.06
GenDot / IDefOpt / cuda / PostRev 0.00001056 s 0.000010016 s 1.05
GenDot / IDefOpt / cuda / BothRev 0.00001008 s 0.000009792 s 1.03
GenDot / JaXPipe / tpu / Primal 9.42975e-7 s 9.30175e-7 s 1.01
GenDot / Jax / tpu / Primal 9.30325e-7 s 9.25725e-7 s 1.00
GenDot / HLOOpt / tpu / Primal 0.0000015984999999999998 s 0.00000157525 s 1.01
GenDot / PartOpt / tpu / Primal 9.30475e-7 s 9.25625e-7 s 1.01
GenDot / IPartOpt / tpu / Primal 9.43475e-7 s 9.297e-7 s 1.01
GenDot / DefOpt / tpu / Primal 0.00000151085 s 0.00000148965 s 1.01
GenDot / IDefOpt / tpu / Primal 0.0000016095249999999998 s 0.0000015781 s 1.02
GenDot / JaXPipe / tpu / Forward 0.000003054775 s 0.000003164375 s 0.97
GenDot / Jax / tpu / Forward 0.00000227725 s 0.000002315075 s 0.98
GenDot / HLOOpt / tpu / Forward 0.000003119825 s 0.000003115875 s 1.00
GenDot / PartOpt / tpu / Forward 0.0000031356750000000003 s 0.00000322785 s 0.97
GenDot / IPartOpt / tpu / Forward 0.000003119725 s 0.000003113875 s 1.00
GenDot / DefOpt / tpu / Forward 0.000003135375 s 0.0000032144750000000003 s 0.98
GenDot / IDefOpt / tpu / Forward 0.0000031125249999999995 s 0.00000311405 s 1.00
GenDot / JaXPipe / tpu / PreRev 0.000003038575 s 0.0000029543750000000004 s 1.03
GenDot / JaXPipe / tpu / PostRev 0.000002379325 s 0.00000240305 s 0.99
GenDot / JaXPipe / tpu / BothRev 0.00000302405 s 0.0000029627 s 1.02
GenDot / Jax / tpu / BothRev 0.00000237895 s 0.000002401275 s 0.99
GenDot / HLOOpt / tpu / PreRev 0.000003016475 s 0.0000029584250000000004 s 1.02
GenDot / HLOOpt / tpu / PostRev 0.0000029387249999999995 s 0.00000294135 s 1.00
GenDot / HLOOpt / tpu / BothRev 0.00000300865 s 0.00000295935 s 1.02
GenDot / PartOpt / tpu / PreRev 0.000002944425 s 0.000002942475 s 1.00
GenDot / PartOpt / tpu / PostRev 0.00000241545 s 0.0000023939 s 1.01
GenDot / PartOpt / tpu / BothRev 0.00000293935 s 0.000002942875 s 1.00
GenDot / IPartOpt / tpu / PreRev 0.00000300825 s 0.0000029719 s 1.01
GenDot / IPartOpt / tpu / PostRev 0.000002380775 s 0.000002410775 s 0.99
GenDot / IPartOpt / tpu / BothRev 0.000003019 s 0.000002961425 s 1.02
GenDot / DefOpt / tpu / PreRev 0.0000029474000000000004 s 0.0000029349 s 1.00
GenDot / DefOpt / tpu / PostRev 0.000003025125 s 0.000002974775 s 1.02
GenDot / DefOpt / tpu / BothRev 0.000002944025 s 0.0000029336750000000003 s 1.00
GenDot / IDefOpt / tpu / PreRev 0.0000030021 s 0.000002968575 s 1.01
GenDot / IDefOpt / tpu / PostRev 0.00000294525 s 0.0000029402250000000003 s 1.00
GenDot / IDefOpt / tpu / BothRev 0.00000302555 s 0.000002968075 s 1.02
GenDot / JaXPipe / cpu / Primal 0.00001524 s 0.000008296160003737896 s 1.84
GenDot / Jax / cpu / Primal 0.000015364 s 0.000008259619971795473 s 1.86
GenDot / HLOOpt / cpu / Primal 0.000014066 s 0.000008371400035684928 s 1.68
GenDot / PartOpt / cpu / Primal 0.000014816 s 0.00000901005999367044 s 1.64
GenDot / IPartOpt / cpu / Primal 0.000015493 s 0.000008573640043323394 s 1.81
GenDot / DefOpt / cpu / Primal 0.000014038 s 0.000007987099979800406 s 1.76
GenDot / IDefOpt / cpu / Primal 0.000014089 s 0.000007670840022910853 s 1.84
GenDot / JaXPipe / cpu / Forward 0.000020107 s 0.000011564219994397715 s 1.74
GenDot / Jax / cpu / Forward 0.000020803 s 0.000011687760006680036 s 1.78
GenDot / HLOOpt / cpu / Forward 0.000019572 s 0.000012549280017992717 s 1.56
GenDot / PartOpt / cpu / Forward 0.000019363 s 0.000011167880002176387 s 1.73
GenDot / IPartOpt / cpu / Forward 0.000019599 s 0.000012273299962544117 s 1.60
GenDot / DefOpt / cpu / Forward 0.000019556 s 0.000011512759956531226 s 1.70
GenDot / IDefOpt / cpu / Forward 0.000019547 s 0.000012003320016447103 s 1.63
GenDot / JaXPipe / cpu / PreRev 0.000019794 s 0.000012277580008230873 s 1.61
GenDot / JaXPipe / cpu / PostRev 0.000020922 s 0.000011590839994823907 s 1.81
GenDot / JaXPipe / cpu / BothRev 0.000019724 s 0.000011976359992331707 s 1.65
GenDot / Jax / cpu / BothRev 0.000020991 s 0.000011309560022709776 s 1.86
GenDot / HLOOpt / cpu / PreRev 0.000020158 s 0.000012416999970810138 s 1.62
GenDot / HLOOpt / cpu / PostRev 0.000020672 s 0.00001339757995083346 s 1.54
GenDot / HLOOpt / cpu / BothRev 0.000019939 s 0.000011228120001760544 s 1.78
GenDot / PartOpt / cpu / PreRev 0.000019417 s 0.000011987920024694176 s 1.62
GenDot / PartOpt / cpu / PostRev 0.000021776 s 0.00001231975996233814 s 1.77
GenDot / PartOpt / cpu / BothRev 0.000019175 s 0.000012321479962338344 s 1.56
GenDot / IPartOpt / cpu / PreRev 0.000021393 s 0.000011272580013610423 s 1.90
GenDot / IPartOpt / cpu / PostRev 0.000021035 s 0.000012042980015394278 s 1.75
GenDot / IPartOpt / cpu / BothRev 0.000020045 s 0.000011842540016004931 s 1.69
GenDot / DefOpt / cpu / PreRev 0.000020075 s 0.000011467599997558865 s 1.75
GenDot / DefOpt / cpu / PostRev 0.000019571000000000003 s 0.000011891319991264026 s 1.65
GenDot / DefOpt / cpu / BothRev 0.000020466 s 0.000012002500052403774 s 1.71
GenDot / IDefOpt / cpu / PreRev 0.000020057 s 0.000011844639984701644 s 1.69
GenDot / IDefOpt / cpu / PostRev 0.000020324 s 0.000012110100014979252 s 1.68
GenDot / IDefOpt / cpu / BothRev 0.000020281 s 0.000012075799977537829 s 1.68
GenDot / JaXPipe / cpu / Primal 0.00001 s 0.000008296160003737896 s 1.21
GenDot / Jax / cpu / Primal 0.000008999999999999999 s 0.000008259619971795473 s 1.09
GenDot / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000008371400035684928 s 1.08
GenDot / PartOpt / cpu / Primal 0.000008999999999999999 s 0.00000901005999367044 s 1.00
GenDot / IPartOpt / cpu / Primal 0.00001 s 0.000008573640043323394 s 1.17
GenDot / DefOpt / cpu / Primal 0.000008 s 0.000007987099979800406 s 1.00
GenDot / IDefOpt / cpu / Primal 0.000008 s 0.000007670840022910853 s 1.04
GenDot / JaXPipe / cpu / Forward 0.000012 s 0.000011564219994397715 s 1.04
GenDot / Jax / cpu / Forward 0.000012 s 0.000011687760006680036 s 1.03
GenDot / HLOOpt / cpu / Forward 0.000012 s 0.000012549280017992717 s 0.96
GenDot / PartOpt / cpu / Forward 0.000012 s 0.000011167880002176387 s 1.07
GenDot / IPartOpt / cpu / Forward 0.000012 s 0.000012273299962544117 s 0.98
GenDot / DefOpt / cpu / Forward 0.000013 s 0.000011512759956531226 s 1.13
GenDot / IDefOpt / cpu / Forward 0.000013 s 0.000012003320016447103 s 1.08
GenDot / JaXPipe / cpu / PreRev 0.000012 s 0.000012277580008230873 s 0.98
GenDot / JaXPipe / cpu / PostRev 0.000013 s 0.000011590839994823907 s 1.12
GenDot / JaXPipe / cpu / BothRev 0.000013 s 0.000011976359992331707 s 1.09
GenDot / Jax / cpu / BothRev 0.000013 s 0.000011309560022709776 s 1.15
GenDot / HLOOpt / cpu / PreRev 0.000013 s 0.000012416999970810138 s 1.05
GenDot / HLOOpt / cpu / PostRev 0.000012 s 0.00001339757995083346 s 0.90
GenDot / HLOOpt / cpu / BothRev 0.000013 s 0.000011228120001760544 s 1.16
GenDot / PartOpt / cpu / PreRev 0.000013 s 0.000011987920024694176 s 1.08
GenDot / PartOpt / cpu / PostRev 0.000013 s 0.00001231975996233814 s 1.06
GenDot / PartOpt / cpu / BothRev 0.000013 s 0.000012321479962338344 s 1.06
GenDot / IPartOpt / cpu / PreRev 0.000012 s 0.000011272580013610423 s 1.06
GenDot / IPartOpt / cpu / PostRev 0.000013 s 0.000012042980015394278 s 1.08
GenDot / IPartOpt / cpu / BothRev 0.000013 s 0.000011842540016004931 s 1.10
GenDot / DefOpt / cpu / PreRev 0.000012 s 0.000011467599997558865 s 1.05
GenDot / DefOpt / cpu / PostRev 0.000013 s 0.000011891319991264026 s 1.09
GenDot / DefOpt / cpu / BothRev 0.000013 s 0.000012002500052403774 s 1.08
GenDot / IDefOpt / cpu / PreRev 0.000012 s 0.000011844639984701644 s 1.01
GenDot / IDefOpt / cpu / PostRev 0.000013 s 0.000012110100014979252 s 1.07
GenDot / IDefOpt / cpu / BothRev 0.000013 s 0.000012075799977537829 s 1.08
hlo_ffi / JaXPipe / cpu / Primal 0.000010057419940494585 s 0.000010163059996557424 s 0.99
hlo_ffi / Jax / cpu / Primal 0.000009597700027370592 s 0.000009343899982923176 s 1.03
hlo_ffi / HLOOpt / cpu / Primal 0.00001007964004202222 s 0.000009920599950419272 s 1.02
hlo_ffi / PartOpt / cpu / Primal 0.000009184099999401951 s 0.000009339799971712637 s 0.98
hlo_ffi / IPartOpt / cpu / Primal 0.00000988384003903775 s 0.000009727540027597567 s 1.02
hlo_ffi / DefOpt / cpu / Primal 0.000009398479969604524 s 0.000009657739974500146 s 0.97
hlo_ffi / IDefOpt / cpu / Primal 0.000009200759996019767 s 0.000009740259984027945 s 0.94
hlo_ffi / JaXPipe / cpu / Forward 0.0000134085799982131 s 0.00001370602000861254 s 0.98
hlo_ffi / Jax / cpu / Forward 0.000013463280029100133 s 0.000013844219993188744 s 0.97
hlo_ffi / HLOOpt / cpu / Forward 0.00001351715999589942 s 0.00001383915999213059 s 0.98
hlo_ffi / PartOpt / cpu / Forward 0.000013447060018734191 s 0.000013616979977086883 s 0.99
hlo_ffi / IPartOpt / cpu / Forward 0.000013710160010305117 s 0.0000135127200428542 s 1.01
hlo_ffi / DefOpt / cpu / Forward 0.000013340959985725932 s 0.000013661500024682028 s 0.98
hlo_ffi / IDefOpt / cpu / Forward 0.000013192420037739794 s 0.000013947420038675773 s 0.95
hlo_ffi / JaXPipe / cpu / PreRev 0.000013574940003309168 s 0.000013623760041809874 s 1.00
hlo_ffi / JaXPipe / cpu / PostRev 0.000013455400012389873 s 0.00001384286003485613 s 0.97
hlo_ffi / JaXPipe / cpu / BothRev 0.000013688420031030546 s 0.000014081040008022683 s 0.97
hlo_ffi / Jax / cpu / BothRev 0.000013489140019373736 s 0.000013902360005886294 s 0.97
hlo_ffi / HLOOpt / cpu / PreRev 0.000013586340010078855 s 0.000013973720015201252 s 0.97
hlo_ffi / HLOOpt / cpu / PostRev 0.00001554592002321442 s 0.00001598522000676894 s 0.97
hlo_ffi / HLOOpt / cpu / BothRev 0.00001351910001176293 s 0.000013878419968023082 s 0.97
hlo_ffi / PartOpt / cpu / PreRev 0.000013509640039046645 s 0.000013816640012009884 s 0.98
hlo_ffi / PartOpt / cpu / PostRev 0.000013434899983622016 s 0.000013948579971838624 s 0.96
hlo_ffi / PartOpt / cpu / BothRev 0.000013497500031007803 s 0.000014386660022864815 s 0.94
hlo_ffi / IPartOpt / cpu / PreRev 0.000013556259964389029 s 0.000013554640017900966 s 1.00
hlo_ffi / IPartOpt / cpu / PostRev 0.000013567539981522714 s 0.000013573120013461448 s 1.00
hlo_ffi / IPartOpt / cpu / BothRev 0.000013514239981304854 s 0.000013850839995939167 s 0.98
hlo_ffi / DefOpt / cpu / PreRev 0.00001331516004029254 s 0.000013287640031194317 s 1.00
hlo_ffi / DefOpt / cpu / PostRev 0.000013310860022102133 s 0.000013696799987883425 s 0.97
hlo_ffi / DefOpt / cpu / BothRev 0.000013430640028673223 s 0.000013824719981130328 s 0.97
hlo_ffi / IDefOpt / cpu / PreRev 0.0000133494799501932 s 0.000013894240000809077 s 0.96
hlo_ffi / IDefOpt / cpu / PostRev 0.00001378451999698882 s 0.000013869060003344204 s 0.99
hlo_ffi / IDefOpt / cpu / BothRev 0.000013581499997599168 s 0.000013805299995510722 s 0.98
hlo_ffi / JaXPipe / cuda / Primal 0.0000023670000000000004 s 0.000001983 s 1.19
hlo_ffi / Jax / cuda / Primal 0.000002369 s 0.000001983 s 1.19
hlo_ffi / HLOOpt / cuda / Primal 0.000002368 s 0.000001983 s 1.19
hlo_ffi / PartOpt / cuda / Primal 0.000002368 s 0.000001984 s 1.19
hlo_ffi / IPartOpt / cuda / Primal 0.000002368 s 0.000001983 s 1.19
hlo_ffi / DefOpt / cuda / Primal 0.000002368 s 0.000001984 s 1.19
hlo_ffi / IDefOpt / cuda / Primal 0.000002368 s 0.000001984 s 1.19
hlo_ffi / JaXPipe / cuda / Forward 0.000002464 s 0.00000208 s 1.18
hlo_ffi / Jax / cuda / Forward 0.000002463 s 0.000002049 s 1.20
hlo_ffi / HLOOpt / cuda / Forward 0.000002463 s 0.000002048 s 1.20
hlo_ffi / PartOpt / cuda / Forward 0.000002464 s 0.000002048 s 1.20
hlo_ffi / IPartOpt / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / DefOpt / cuda / Forward 0.000002464 s 0.00000208 s 1.18
hlo_ffi / IDefOpt / cuda / Forward 0.000002463 s 0.00000208 s 1.18
hlo_ffi / JaXPipe / cuda / PreRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / JaXPipe / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / JaXPipe / cuda / BothRev 0.000002464 s 0.000002047 s 1.20
hlo_ffi / Jax / cuda / BothRev 0.000002463 s 0.000002048 s 1.20
hlo_ffi / HLOOpt / cuda / PreRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / HLOOpt / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / HLOOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / PartOpt / cuda / PreRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / PartOpt / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / PartOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IPartOpt / cuda / PreRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / IPartOpt / cuda / PostRev 0.000002463 s 0.000002048 s 1.20
hlo_ffi / IPartOpt / cuda / BothRev 0.000002432 s 0.000002048 s 1.19
hlo_ffi / DefOpt / cuda / PreRev 0.000002433 s 0.000002048 s 1.19
hlo_ffi / DefOpt / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / DefOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IDefOpt / cuda / PreRev 0.000002433 s 0.000002048 s 1.19
hlo_ffi / IDefOpt / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IDefOpt / cuda / BothRev 0.000002431 s 0.000002048 s 1.19
hlo_ffi / JaXPipe / tpu / Primal 9.07375e-7 s 9.3215e-7 s 0.97
hlo_ffi / Jax / tpu / Primal 9.713e-7 s 9.53125e-7 s 1.02
hlo_ffi / HLOOpt / tpu / Primal 9.411e-7 s 9.07375e-7 s 1.04
hlo_ffi / PartOpt / tpu / Primal 9.70525e-7 s 9.50025e-7 s 1.02
hlo_ffi / IPartOpt / tpu / Primal 9.41325e-7 s 9.1355e-7 s 1.03
hlo_ffi / DefOpt / tpu / Primal 9.71825e-7 s 9.58125e-7 s 1.01
hlo_ffi / IDefOpt / tpu / Primal 9.4065e-7 s 9.0525e-7 s 1.04
hlo_ffi / JaXPipe / tpu / Forward 9.4885e-7 s 9.49075e-7 s 1.00
hlo_ffi / Jax / tpu / Forward 9.81725e-7 s 9.81575e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Forward 9.7355e-7 s 9.73475e-7 s 1.00
hlo_ffi / PartOpt / tpu / Forward 9.5945e-7 s 9.33675e-7 s 1.03
hlo_ffi / IPartOpt / tpu / Forward 9.735e-7 s 9.736749999999998e-7 s 1.00
hlo_ffi / DefOpt / tpu / Forward 9.59e-7 s 9.33775e-7 s 1.03
hlo_ffi / IDefOpt / tpu / Forward 9.735e-7 s 9.73925e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PreRev 9.53825e-7 s 9.3785e-7 s 1.02
hlo_ffi / JaXPipe / tpu / PostRev 9.65e-7 s 9.64675e-7 s 1.00
hlo_ffi / JaXPipe / tpu / BothRev 9.9465e-7 s 9.62325e-7 s 1.03
hlo_ffi / Jax / tpu / BothRev 9.644e-7 s 9.64675e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PreRev 9.94525e-7 s 9.62175e-7 s 1.03
hlo_ffi / HLOOpt / tpu / PostRev 9.64475e-7 s 9.64525e-7 s 1.00
hlo_ffi / HLOOpt / tpu / BothRev 9.949e-7 s 9.62325e-7 s 1.03
hlo_ffi / PartOpt / tpu / PreRev 9.647e-7 s 9.647e-7 s 1
hlo_ffi / PartOpt / tpu / PostRev 9.94975e-7 s 9.6175e-7 s 1.03
hlo_ffi / PartOpt / tpu / BothRev 9.64825e-7 s 9.64425e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PreRev 9.94525e-7 s 9.61625e-7 s 1.03
hlo_ffi / IPartOpt / tpu / PostRev 9.64775e-7 s 9.64925e-7 s 1.00
hlo_ffi / IPartOpt / tpu / BothRev 9.94675e-7 s 9.618500000000002e-7 s 1.03
hlo_ffi / DefOpt / tpu / PreRev 9.647e-7 s 9.64575e-7 s 1.00
hlo_ffi / DefOpt / tpu / PostRev 9.94725e-7 s 9.6215e-7 s 1.03
hlo_ffi / DefOpt / tpu / BothRev 9.64975e-7 s 9.6475e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PreRev 9.943249999999998e-7 s 9.6165e-7 s 1.03
hlo_ffi / IDefOpt / tpu / PostRev 9.6455e-7 s 9.64825e-7 s 1.00
hlo_ffi / IDefOpt / tpu / BothRev 9.94825e-7 s 9.61925e-7 s 1.03
hlo_ffi / JaXPipe / cpu / Primal 0.000018081 s 0.000010163059996557424 s 1.78
hlo_ffi / Jax / cpu / Primal 0.000017586 s 0.000009343899982923176 s 1.88
hlo_ffi / HLOOpt / cpu / Primal 0.000017587999999999998 s 0.000009920599950419272 s 1.77
hlo_ffi / PartOpt / cpu / Primal 0.000017583 s 0.000009339799971712637 s 1.88
hlo_ffi / IPartOpt / cpu / Primal 0.000018028 s 0.000009727540027597567 s 1.85
hlo_ffi / DefOpt / cpu / Primal 0.000018361 s 0.000009657739974500146 s 1.90
hlo_ffi / IDefOpt / cpu / Primal 0.000018564 s 0.000009740259984027945 s 1.91
hlo_ffi / JaXPipe / cpu / Forward 0.00002515 s 0.00001370602000861254 s 1.83
hlo_ffi / Jax / cpu / Forward 0.000024209 s 0.000013844219993188744 s 1.75
hlo_ffi / HLOOpt / cpu / Forward 0.000024763 s 0.00001383915999213059 s 1.79
hlo_ffi / PartOpt / cpu / Forward 0.000025531 s 0.000013616979977086883 s 1.87
hlo_ffi / IPartOpt / cpu / Forward 0.000024967 s 0.0000135127200428542 s 1.85
hlo_ffi / DefOpt / cpu / Forward 0.000024357 s 0.000013661500024682028 s 1.78
hlo_ffi / IDefOpt / cpu / Forward 0.000025141 s 0.000013947420038675773 s 1.80
hlo_ffi / JaXPipe / cpu / PreRev 0.00002474 s 0.000013623760041809874 s 1.82
hlo_ffi / JaXPipe / cpu / PostRev 0.000023729 s 0.00001384286003485613 s 1.71
hlo_ffi / JaXPipe / cpu / BothRev 0.000024972 s 0.000014081040008022683 s 1.77
hlo_ffi / Jax / cpu / BothRev 0.000025607 s 0.000013902360005886294 s 1.84
hlo_ffi / HLOOpt / cpu / PreRev 0.000025655 s 0.000013973720015201252 s 1.84
hlo_ffi / HLOOpt / cpu / PostRev 0.000024846 s 0.00001598522000676894 s 1.55
hlo_ffi / HLOOpt / cpu / BothRev 0.000024836 s 0.000013878419968023082 s 1.79
hlo_ffi / PartOpt / cpu / PreRev 0.000025532 s 0.000013816640012009884 s 1.85
hlo_ffi / PartOpt / cpu / PostRev 0.000024827 s 0.000013948579971838624 s 1.78
hlo_ffi / PartOpt / cpu / BothRev 0.000025075 s 0.000014386660022864815 s 1.74
hlo_ffi / IPartOpt / cpu / PreRev 0.000024895 s 0.000013554640017900966 s 1.84
hlo_ffi / IPartOpt / cpu / PostRev 0.000025268 s 0.000013573120013461448 s 1.86
hlo_ffi / IPartOpt / cpu / BothRev 0.000024098 s 0.000013850839995939167 s 1.74
hlo_ffi / DefOpt / cpu / PreRev 0.00002425 s 0.000013287640031194317 s 1.83
hlo_ffi / DefOpt / cpu / PostRev 0.000023965 s 0.000013696799987883425 s 1.75
hlo_ffi / DefOpt / cpu / BothRev 0.000023848000000000003 s 0.000013824719981130328 s 1.73
hlo_ffi / IDefOpt / cpu / PreRev 0.000025046 s 0.000013894240000809077 s 1.80
hlo_ffi / IDefOpt / cpu / PostRev 0.000024637 s 0.000013869060003344204 s 1.78
hlo_ffi / IDefOpt / cpu / BothRev 0.00002447 s 0.000013805299995510722 s 1.77
hlo_ffi / JaXPipe / cpu / Primal 0.000011 s 0.000010163059996557424 s 1.08
hlo_ffi / Jax / cpu / Primal 0.000011 s 0.000009343899982923176 s 1.18
hlo_ffi / HLOOpt / cpu / Primal 0.000012 s 0.000009920599950419272 s 1.21
hlo_ffi / PartOpt / cpu / Primal 0.000011 s 0.000009339799971712637 s 1.18
hlo_ffi / IPartOpt / cpu / Primal 0.000012 s 0.000009727540027597567 s 1.23
hlo_ffi / DefOpt / cpu / Primal 0.000012 s 0.000009657739974500146 s 1.24
hlo_ffi / IDefOpt / cpu / Primal 0.000011 s 0.000009740259984027945 s 1.13
hlo_ffi / JaXPipe / cpu / Forward 0.000017 s 0.00001370602000861254 s 1.24
hlo_ffi / Jax / cpu / Forward 0.000016 s 0.000013844219993188744 s 1.16
hlo_ffi / HLOOpt / cpu / Forward 0.000017 s 0.00001383915999213059 s 1.23
hlo_ffi / PartOpt / cpu / Forward 0.000016 s 0.000013616979977086883 s 1.18
hlo_ffi / IPartOpt / cpu / Forward 0.000016 s 0.0000135127200428542 s 1.18
hlo_ffi / DefOpt / cpu / Forward 0.000016 s 0.000013661500024682028 s 1.17
hlo_ffi / IDefOpt / cpu / Forward 0.000016 s 0.000013947420038675773 s 1.15
hlo_ffi / JaXPipe / cpu / PreRev 0.000015 s 0.000013623760041809874 s 1.10
hlo_ffi / JaXPipe / cpu / PostRev 0.000016 s 0.00001384286003485613 s 1.16
hlo_ffi / JaXPipe / cpu / BothRev 0.000016 s 0.000014081040008022683 s 1.14
hlo_ffi / Jax / cpu / BothRev 0.000015 s 0.000013902360005886294 s 1.08
hlo_ffi / HLOOpt / cpu / PreRev 0.000017 s 0.000013973720015201252 s 1.22
hlo_ffi / HLOOpt / cpu / PostRev 0.000016 s 0.00001598522000676894 s 1.00
hlo_ffi / HLOOpt / cpu / BothRev 0.000016 s 0.000013878419968023082 s 1.15
hlo_ffi / PartOpt / cpu / PreRev 0.000017 s 0.000013816640012009884 s 1.23
hlo_ffi / PartOpt / cpu / PostRev 0.000016 s 0.000013948579971838624 s 1.15
hlo_ffi / PartOpt / cpu / BothRev 0.000016 s 0.000014386660022864815 s 1.11
hlo_ffi / IPartOpt / cpu / PreRev 0.000016 s 0.000013554640017900966 s 1.18
hlo_ffi / IPartOpt / cpu / PostRev 0.000016 s 0.000013573120013461448 s 1.18
hlo_ffi / IPartOpt / cpu / BothRev 0.000017 s 0.000013850839995939167 s 1.23
hlo_ffi / DefOpt / cpu / PreRev 0.000017 s 0.000013287640031194317 s 1.28
hlo_ffi / DefOpt / cpu / PostRev 0.000016 s 0.000013696799987883425 s 1.17
hlo_ffi / DefOpt / cpu / BothRev 0.000016 s 0.000013824719981130328 s 1.16
hlo_ffi / IDefOpt / cpu / PreRev 0.000015 s 0.000013894240000809077 s 1.08
hlo_ffi / IDefOpt / cpu / PostRev 0.000016 s 0.000013869060003344204 s 1.15
hlo_ffi / IDefOpt / cpu / BothRev 0.000017 s 0.000013805299995510722 s 1.23
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 0.0009187847999783 s 0.0008871002001797 s 1.04
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 0.0009269996000512 s 0.0008837956000206 s 1.05
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 0.0009733178000715 s 0.0009228236001035 s 1.05
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 0.0010169392001444 s 0.0008805019999272 s 1.15
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 0.0009170368000923 s 0.0008894220000001 s 1.03
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 0.0009781311998267 s 0.0009508617999017 s 1.03
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 0.000984045199948 s 0.0009291351999308 s 1.06
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 0.0023619201999281 s 0.0021042777999355 s 1.12
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 0.0024165683999854 s 0.0022587204000046 s 1.07
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 0.0022583170000871 s 0.0021500653998373 s 1.05
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 0.0022445555999183 s 0.0022119191999081 s 1.01
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 0.0023041448000185 s 0.0021429494000585 s 1.08
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 0.0022968314000536 s 0.0021400593999715 s 1.07
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 0.002262568200058 s 0.0021881026000301 s 1.03
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 0.0056173855999986 s 0.0050751822000165 s 1.11
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 0.0065307915999255 s 0.0056805405999512 s 1.15
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 0.005857640800059 s 0.0058089991998713 s 1.01
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 0.0040817025998876 s 0.0052328876001411 s 0.78
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 0.0062605933999293 s 0.0055911849999574 s 1.12
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 0.0063398268001037 s 0.0047908924000694 s 1.32
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 0.0050503892000051 s 0.0059099502001117 s 0.85
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 0.0063620181999795 s 0.0036509223998109 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 / PartOpt / cpu / PostRev 0.0060140849999697 s 0.0054702081999494 s 1.10
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 0.0033959379999032 s 0.0052289548000771 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 / IPartOpt / cpu / PreRev 0.0035569932000726 s 0.0050246252000761 s 0.71
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 0.0049243536001085 s 0.0065582606001044 s 0.75
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 0.0034775404000356 s 0.0034128457998122 s 1.02
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 0.005382198999996 s 0.0054061793999608 s 1.00
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 0.0047845456000686 s 0.004691940000066 s 1.02
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 0.0035333757999978 s 0.0050492209999902 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 / IDefOpt / cpu / PreRev 0.0051398689999587 s 0.0050608083999577 s 1.02
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 0.005425823799851 s 0.0056577018001007 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 / IDefOpt / cpu / BothRev 0.0034540200000265 s 0.0047789924000426 s 0.72
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 0.00030016 s 0.000284001 s 1.06
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 0.000301024 s 0.000284033 s 1.06
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 0.000305984 s 0.000291169 s 1.05
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 0.000299617 s 0.000282881 s 1.06
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 0.000300033 s 0.000283489 s 1.06
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 0.000307617 s 0.000290817 s 1.06
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 0.0003065919999999 s 0.00029104 s 1.05
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 0.000584545 s 0.000555745 s 1.05
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 0.000568385 s 0.000537986 s 1.06
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 0.000584193 s 0.0005573129999999 s 1.05
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 0.000584448 s 0.000556898 s 1.05
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 0.000584449 s 0.000556737 s 1.05
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 0.000584449 s 0.000557762 s 1.05
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 0.000584288 s 0.000557154 s 1.05
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 0.001061185 s 0.001026882 s 1.03
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 0.0010161929999999 s 0.000984706 s 1.03
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 0.001056192 s 0.001019711 s 1.04
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 0.001010401 s 0.000986369 s 1.02
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 0.001041569 s 0.0010103089999999 s 1.03
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 0.001064609 s 0.001035842 s 1.03
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 0.001039553 s 0.001006659 s 1.03
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 0.001055073 s 0.001025059 s 1.03
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 0.001003809 s 0.000975939 s 1.03
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 0.001055393 s 0.001024803 s 1.03
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 0.001056897 s 0.001022979 s 1.03
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 0.001001473 s 0.000974241 s 1.03
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 0.001054785 s 0.001024098 s 1.03
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 0.001054465 s 0.00101962 s 1.03
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 0.000989888 s 0.00095821 s 1.03
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 0.001056513 s 0.001023619 s 1.03
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 0.001056514 s 0.001021217 s 1.03
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 0.001057825 s 0.001017858 s 1.04
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 0.001057153 s 0.001020512 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / Primal 0.00013110725 s 0.00012352525 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / Primal 0.00012398 s 0.00012669775 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 / tpu / Primal 0.0001601117499999 s 0.00015251875 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / Primal 0.00013089325 s 0.000134193 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 / tpu / Primal 0.000138506 s 0.00013077025 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / Primal 0.00014552375 s 0.00014781275 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 / IDefOpt / tpu / Primal 0.000158217 s 0.00015106675 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / Forward 0.0002135645 s 0.000211938 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / Forward 0.000262684 s 0.0002610617499999 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / Forward 0.0002202342499999 s 0.0002121707499999 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / Forward 0.000215079 s 0.0002180217499999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / Forward 0.00021605 s 0.00021172775 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / Forward 0.00021803525 s 0.00021815225 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / Forward 0.0002160935 s 0.0002117155 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PreRev 0.0003561915 s 0.0003562175 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PostRev 0.000255838 s 0.0002588909999999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / BothRev 0.000356417 s 0.0003564275 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / BothRev 0.000257003 s 0.0002595994999999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / PreRev 0.0003566037499999 s 0.00035665025 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / PostRev 0.00029139925 s 0.00029234875 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / BothRev 0.0003560714999999 s 0.0003562995 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / PreRev 0.0003558517499999 s 0.00035881575 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / PostRev 0.00027236675 s 0.0002721429999999 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / BothRev 0.0003559614999999 s 0.000358587 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / PreRev 0.000356304 s 0.00035639775 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / PostRev 0.0002720652499999 s 0.0002747947499999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / BothRev 0.00035610225 s 0.00035658475 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / PreRev 0.0003582667499999 s 0.0003597825 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / PostRev 0.0002840794999999 s 0.0002841942499999 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / BothRev 0.00035831775 s 0.0003598795 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / PreRev 0.00035841575 s 0.000357839 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / PostRev 0.00030145325 s 0.000302267 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / BothRev 0.000358556 s 0.0003577502499999 s 1.00
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 0.0028831959999999 s 0.0008871002001797 s 3.25
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 0.002713818 s 0.0008837956000206 s 3.07
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 0.00321451 s 0.0009228236001035 s 3.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.00274036 s 0.0008805019999272 s 3.11
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 0.002824053 s 0.0008894220000001 s 3.18
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 0.002939613 s 0.0009508617999017 s 3.09
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 0.003052437 s 0.0009291351999308 s 3.29
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 0.006945267 s 0.0021042777999355 s 3.30
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 0.007027573 s 0.0022587204000046 s 3.11
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 0.007231585 s 0.0021500653998373 s 3.36
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 0.0065548859999999 s 0.0022119191999081 s 2.96
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 0.006821829 s 0.0021429494000585 s 3.18
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 0.006831938 s 0.0021400593999715 s 3.19
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 0.00703947 s 0.0021881026000301 s 3.22
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 0.010792557 s 0.0050751822000165 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 / JaXPipe / cpu / PostRev 0.010286442 s 0.0056805405999512 s 1.81
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 0.008816719 s 0.0058089991998713 s 1.52
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 0.010454945 s 0.0052328876001411 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 / PreRev 0.010466498 s 0.0055911849999574 s 1.87
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 0.009768714 s 0.0047908924000694 s 2.04
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 0.010496469 s 0.0059099502001117 s 1.78
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 0.008974524 s 0.0036509223998109 s 2.46
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 0.009790116 s 0.0054702081999494 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 / PartOpt / cpu / BothRev 0.009188827 s 0.0052289548000771 s 1.76
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 0.010081388 s 0.0050246252000761 s 2.01
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 0.010708134 s 0.0065582606001044 s 1.63
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 0.010330913 s 0.0034128457998122 s 3.03
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 0.009602437 s 0.0054061793999608 s 1.78
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 0.009692763 s 0.004691940000066 s 2.07
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 0.009130718 s 0.0050492209999902 s 1.81
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 0.009006486 s 0.0050608083999577 s 1.78
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 0.0097862449999999 s 0.0056577018001007 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 / BothRev 0.009545373 s 0.0047789924000426 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 / JaXPipe / cpu / Primal 0.001771 s 0.0008871002001797 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 / Jax / cpu / Primal 0.001577 s 0.0008837956000206 s 1.78
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 0.00156 s 0.0009228236001035 s 1.69
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 0.0020819999999999 s 0.0008805019999272 s 2.36
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 0.001464 s 0.0008894220000001 s 1.65
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 0.001873 s 0.0009508617999017 s 1.97
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 0.001896 s 0.0009291351999308 s 2.04
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 0.003995 s 0.0021042777999355 s 1.90
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 0.003839 s 0.0022587204000046 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 / HLOOpt / cpu / Forward 0.004354 s 0.0021500653998373 s 2.03
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 0.00509 s 0.0022119191999081 s 2.30
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 0.003816 s 0.0021429494000585 s 1.78
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 0.0037949999999999 s 0.0021400593999715 s 1.77
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 0.00373 s 0.0021881026000301 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 / JaXPipe / cpu / PreRev 0.007233 s 0.0050751822000165 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 / JaXPipe / cpu / PostRev 0.008792 s 0.0056805405999512 s 1.55
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 0.0077 s 0.0058089991998713 s 1.33
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 0.010794 s 0.0052328876001411 s 2.06
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 0.007373 s 0.0055911849999574 s 1.32
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 0.006711 s 0.0047908924000694 s 1.40
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 0.008154 s 0.0059099502001117 s 1.38
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 0.007759 s 0.0036509223998109 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 / PartOpt / cpu / PostRev 0.008933 s 0.0054702081999494 s 1.63
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 0.007058 s 0.0052289548000771 s 1.35
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 0.007741 s 0.0050246252000761 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 / IPartOpt / cpu / PostRev 0.010409 s 0.0065582606001044 s 1.59
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 0.0073019999999999 s 0.0034128457998122 s 2.14
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 0.007259 s 0.0054061793999608 s 1.34
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 0.007208 s 0.004691940000066 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 / DefOpt / cpu / BothRev 0.007307 s 0.0050492209999902 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 / IDefOpt / cpu / PreRev 0.0081479999999999 s 0.0050608083999577 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 0.0069169999999999 s 0.0056577018001007 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 / IDefOpt / cpu / BothRev 0.00893 s 0.0047789924000426 s 1.87
scatter_sum / JaXPipe / cpu / Primal 0.000007869600003687084 s 0.000008399699972869711 s 0.94
scatter_sum / Jax / cpu / Primal 0.000007655459967281786 s 0.00000856758005284064 s 0.89
scatter_sum / HLOOpt / cpu / Primal 0.000007622459988851915 s 0.000009258660020350362 s 0.82
scatter_sum / PartOpt / cpu / Primal 0.000007392239976979908 s 0.000009661359954407087 s 0.77
scatter_sum / IPartOpt / cpu / Primal 0.000007593800046379329 s 0.000009734099976412837 s 0.78
scatter_sum / DefOpt / cpu / Primal 0.000007478279985662084 s 0.000009198239977195045 s 0.81
scatter_sum / IDefOpt / cpu / Primal 0.000007268199997270131 s 0.000008779059990047244 s 0.83
scatter_sum / JaXPipe / cpu / Forward 0.000010821519990713567 s 0.000013061339986961684 s 0.83
scatter_sum / Jax / cpu / Forward 0.000010745359959400958 s 0.000012101459969926508 s 0.89
scatter_sum / HLOOpt / cpu / Forward 0.00001144124004895275 s 0.000012741700002152357 s 0.90
scatter_sum / PartOpt / cpu / Forward 0.000010545400000410154 s 0.000012532019991340348 s 0.84
scatter_sum / IPartOpt / cpu / Forward 0.000011321920001137186 s 0.0000133172999630915 s 0.85
scatter_sum / DefOpt / cpu / Forward 0.000011093259972767555 s 0.00001335968000603316 s 0.83
scatter_sum / IDefOpt / cpu / Forward 0.000010686560026442748 s 0.000013295580029080156 s 0.80
scatter_sum / JaXPipe / cpu / PreRev 0.000011304599975119344 s 0.000012938739982928384 s 0.87
scatter_sum / JaXPipe / cpu / PostRev 0.000011322120008117052 s 0.00001292862001719186 s 0.88
scatter_sum / JaXPipe / cpu / BothRev 0.000011462859984021634 s 0.000012804760017388616 s 0.90
scatter_sum / Jax / cpu / BothRev 0.000010823119973792929 s 0.000012629419989025336 s 0.86
scatter_sum / HLOOpt / cpu / PreRev 0.000011406100011299712 s 0.000013125519999448442 s 0.87
scatter_sum / HLOOpt / cpu / PostRev 0.000012933179987157927 s 0.000014698720024171053 s 0.88
scatter_sum / HLOOpt / cpu / BothRev 0.000010644159992807544 s 0.000013126020039635478 s 0.81
scatter_sum / PartOpt / cpu / PreRev 0.000010643800005709635 s 0.000013033599962000154 s 0.82
scatter_sum / PartOpt / cpu / PostRev 0.0000109837799664092 s 0.000012915639981656568 s 0.85
scatter_sum / PartOpt / cpu / BothRev 0.000011411540062908898 s 0.000012708979975286638 s 0.90
scatter_sum / IPartOpt / cpu / PreRev 0.000010801499993249308 s 0.000012781279965565772 s 0.85
scatter_sum / IPartOpt / cpu / PostRev 0.000010973380003633793 s 0.000013472300042849384 s 0.81
scatter_sum / IPartOpt / cpu / BothRev 0.000011085759979323484 s 0.000012834419976570643 s 0.86
scatter_sum / DefOpt / cpu / PreRev 0.000010736379999798374 s 0.000013145060020178787 s 0.82
scatter_sum / DefOpt / cpu / PostRev 0.000010675600033209777 s 0.00001275038002859219 s 0.84
scatter_sum / DefOpt / cpu / BothRev 0.000010966019990519271 s 0.000012998379997952724 s 0.84
scatter_sum / IDefOpt / cpu / PreRev 0.0000112115399861068 s 0.000013091199998598312 s 0.86
scatter_sum / IDefOpt / cpu / PostRev 0.000011269359974903637 s 0.00001284074000068358 s 0.88
scatter_sum / IDefOpt / cpu / BothRev 0.00001058306001141318 s 0.000012844779957958964 s 0.82
scatter_sum / JaXPipe / cuda / Primal 0.00001024 s 0.000009665 s 1.06
scatter_sum / Jax / cuda / Primal 0.000010367 s 0.000010208 s 1.02
scatter_sum / HLOOpt / cuda / Primal 0.000010496 s 0.00000976 s 1.08
scatter_sum / PartOpt / cuda / Primal 0.000010336 s 0.000009824 s 1.05
scatter_sum / IPartOpt / cuda / Primal 0.000010688 s 0.00000992 s 1.08
scatter_sum / DefOpt / cuda / Primal 0.000010304 s 0.000009888 s 1.04
scatter_sum / IDefOpt / cuda / Primal 0.000010463 s 0.000009952 s 1.05
scatter_sum / JaXPipe / cuda / Forward 0.000017375999999999998 s 0.000016768000000000003 s 1.04
scatter_sum / Jax / cuda / Forward 0.000016704 s 0.000016255999999999998 s 1.03
scatter_sum / HLOOpt / cuda / Forward 0.000017312 s 0.000016927999999999998 s 1.02
scatter_sum / PartOpt / cuda / Forward 0.000017760000000000003 s 0.000016576000000000002 s 1.07
scatter_sum / IPartOpt / cuda / Forward 0.000018015 s 0.00001696 s 1.06
scatter_sum / DefOpt / cuda / Forward 0.000017503999999999997 s 0.000016288 s 1.07
scatter_sum / IDefOpt / cuda / Forward 0.000017056 s 0.000018336 s 0.93
scatter_sum / JaXPipe / cuda / PreRev 0.000017152 s 0.000018624000000000003 s 0.92
scatter_sum / JaXPipe / cuda / PostRev 0.000017472 s 0.000016511 s 1.06
scatter_sum / JaXPipe / cuda / BothRev 0.0000168 s 0.00001616 s 1.04
scatter_sum / Jax / cuda / BothRev 0.0000168 s 0.00001712 s 0.98
scatter_sum / HLOOpt / cuda / PreRev 0.0000168 s 0.000016704 s 1.01
scatter_sum / HLOOpt / cuda / PostRev 0.000016576000000000002 s 0.000015935999999999998 s 1.04
scatter_sum / HLOOpt / cuda / BothRev 0.000016864 s 0.000021728 s 0.78
scatter_sum / PartOpt / cuda / PreRev 0.000017216 s 0.000016832 s 1.02
scatter_sum / PartOpt / cuda / PostRev 0.00001696 s 0.000015904000000000002 s 1.07
scatter_sum / PartOpt / cuda / BothRev 0.000017249 s 0.000017568000000000002 s 0.98
scatter_sum / IPartOpt / cuda / PreRev 0.000017344 s 0.000016832 s 1.03
scatter_sum / IPartOpt / cuda / PostRev 0.000016768000000000003 s 0.000016191 s 1.04
scatter_sum / IPartOpt / cuda / BothRev 0.000016672 s 0.00001696 s 0.98
scatter_sum / DefOpt / cuda / PreRev 0.000019936 s 0.000017311 s 1.15
scatter_sum / DefOpt / cuda / PostRev 0.000019649 s 0.000016417000000000002 s 1.20
scatter_sum / DefOpt / cuda / BothRev 0.00001712 s 0.000016257 s 1.05
scatter_sum / IDefOpt / cuda / PreRev 0.000017344 s 0.00001696 s 1.02
scatter_sum / IDefOpt / cuda / PostRev 0.000016927999999999998 s 0.000021728 s 0.78
scatter_sum / IDefOpt / cuda / BothRev 0.00001696 s 0.00001712 s 0.99
scatter_sum / JaXPipe / tpu / Primal 0.000001342975 s 0.000001342625 s 1.00
scatter_sum / Jax / tpu / Primal 0.00000134345 s 0.000001403475 s 0.96
scatter_sum / HLOOpt / tpu / Primal 0.0000013428 s 0.000001342625 s 1.00
scatter_sum / PartOpt / tpu / Primal 0.0000013429499999999998 s 0.0000014036 s 0.96
scatter_sum / IPartOpt / tpu / Primal 0.000001342775 s 0.000001342725 s 1.00
scatter_sum / DefOpt / tpu / Primal 0.000001343 s 0.00000140385 s 0.96
scatter_sum / IDefOpt / tpu / Primal 0.000001342625 s 0.0000013422999999999998 s 1.00
scatter_sum / JaXPipe / tpu / Forward 0.0000027349000000000003 s 0.0000027100000000000003 s 1.01
scatter_sum / Jax / tpu / Forward 0.0000027463500000000004 s 0.000002715175 s 1.01
scatter_sum / HLOOpt / tpu / Forward 0.000002737725 s 0.0000027005750000000004 s 1.01
scatter_sum / PartOpt / tpu / Forward 0.0000027187 s 0.0000026893499999999995 s 1.01
scatter_sum / IPartOpt / tpu / Forward 0.000002743675 s 0.0000027078 s 1.01
scatter_sum / DefOpt / tpu / Forward 0.000002719775 s 0.00000269455 s 1.01
scatter_sum / IDefOpt / tpu / Forward 0.000002736575 s 0.0000026998 s 1.01
scatter_sum / JaXPipe / tpu / PreRev 0.0000027117500000000003 s 0.0000026912750000000003 s 1.01
scatter_sum / JaXPipe / tpu / PostRev 0.00000273175 s 0.000002691025 s 1.02
scatter_sum / JaXPipe / tpu / BothRev 0.000002730225 s 0.00000269995 s 1.01
scatter_sum / Jax / tpu / BothRev 0.000002789825 s 0.0000027454000000000004 s 1.02
scatter_sum / HLOOpt / tpu / PreRev 0.0000027314 s 0.0000026961750000000005 s 1.01
scatter_sum / HLOOpt / tpu / PostRev 0.000002789425 s 0.0000027491 s 1.01
scatter_sum / HLOOpt / tpu / BothRev 0.00000274045 s 0.000002698375 s 1.02
scatter_sum / PartOpt / tpu / PreRev 0.0000027862749999999994 s 0.000002740025 s 1.02
scatter_sum / PartOpt / tpu / PostRev 0.000002721 s 0.0000026995500000000003 s 1.01
scatter_sum / PartOpt / tpu / BothRev 0.0000027895 s 0.000002742875 s 1.02
scatter_sum / IPartOpt / tpu / PreRev 0.00000272595 s 0.00000270345 s 1.01
scatter_sum / IPartOpt / tpu / PostRev 0.0000027846 s 0.000002738425 s 1.02
scatter_sum / IPartOpt / tpu / BothRev 0.0000027294000000000005 s 0.0000026955499999999995 s 1.01
scatter_sum / DefOpt / tpu / PreRev 0.000002789725 s 0.0000027395750000000004 s 1.02
scatter_sum / DefOpt / tpu / PostRev 0.00000272095 s 0.000002693525 s 1.01
scatter_sum / DefOpt / tpu / BothRev 0.000002791475 s 0.000002744675 s 1.02
scatter_sum / IDefOpt / tpu / PreRev 0.0000027256 s 0.00000269625 s 1.01
scatter_sum / IDefOpt / tpu / PostRev 0.0000027852 s 0.0000027459 s 1.01
scatter_sum / IDefOpt / tpu / BothRev 0.000002729725 s 0.0000026999 s 1.01
scatter_sum / JaXPipe / cpu / Primal 0.000015804 s 0.000008399699972869711 s 1.88
scatter_sum / Jax / cpu / Primal 0.000016211 s 0.00000856758005284064 s 1.89
scatter_sum / HLOOpt / cpu / Primal 0.000016408 s 0.000009258660020350362 s 1.77
scatter_sum / PartOpt / cpu / Primal 0.000015696 s 0.000009661359954407087 s 1.62
scatter_sum / IPartOpt / cpu / Primal 0.000016247000000000002 s 0.000009734099976412837 s 1.67
scatter_sum / DefOpt / cpu / Primal 0.00001587 s 0.000009198239977195045 s 1.73
scatter_sum / IDefOpt / cpu / Primal 0.00001577 s 0.000008779059990047244 s 1.80
scatter_sum / JaXPipe / cpu / Forward 0.000024595 s 0.000013061339986961684 s 1.88
scatter_sum / Jax / cpu / Forward 0.000023757 s 0.000012101459969926508 s 1.96
scatter_sum / HLOOpt / cpu / Forward 0.000024404 s 0.000012741700002152357 s 1.92
scatter_sum / PartOpt / cpu / Forward 0.000023865 s 0.000012532019991340348 s 1.90
scatter_sum / IPartOpt / cpu / Forward 0.0000241 s 0.0000133172999630915 s 1.81
scatter_sum / DefOpt / cpu / Forward 0.000023185 s 0.00001335968000603316 s 1.74
scatter_sum / IDefOpt / cpu / Forward 0.000024643 s 0.000013295580029080156 s 1.85
scatter_sum / JaXPipe / cpu / PreRev 0.000024087 s 0.000012938739982928384 s 1.86
scatter_sum / JaXPipe / cpu / PostRev 0.000023368 s 0.00001292862001719186 s 1.81
scatter_sum / JaXPipe / cpu / BothRev 0.000022826 s 0.000012804760017388616 s 1.78
scatter_sum / Jax / cpu / BothRev 0.00002402 s 0.000012629419989025336 s 1.90
scatter_sum / HLOOpt / cpu / PreRev 0.000023898 s 0.000013125519999448442 s 1.82
scatter_sum / HLOOpt / cpu / PostRev 0.000023243 s 0.000014698720024171053 s 1.58
scatter_sum / HLOOpt / cpu / BothRev 0.000023277 s 0.000013126020039635478 s 1.77
scatter_sum / PartOpt / cpu / PreRev 0.000024215 s 0.000013033599962000154 s 1.86
scatter_sum / PartOpt / cpu / PostRev 0.000023845000000000003 s 0.000012915639981656568 s 1.85
scatter_sum / PartOpt / cpu / BothRev 0.000023675000000000003 s 0.000012708979975286638 s 1.86
scatter_sum / IPartOpt / cpu / PreRev 0.000024224 s 0.000012781279965565772 s 1.90
scatter_sum / IPartOpt / cpu / PostRev 0.000022985 s 0.000013472300042849384 s 1.71
scatter_sum / IPartOpt / cpu / BothRev 0.000023699 s 0.000012834419976570643 s 1.85
scatter_sum / DefOpt / cpu / PreRev 0.000024222 s 0.000013145060020178787 s 1.84
scatter_sum / DefOpt / cpu / PostRev 0.000024346 s 0.00001275038002859219 s 1.91
scatter_sum / DefOpt / cpu / BothRev 0.000024176 s 0.000012998379997952724 s 1.86
scatter_sum / IDefOpt / cpu / PreRev 0.000025216 s 0.000013091199998598312 s 1.93
scatter_sum / IDefOpt / cpu / PostRev 0.000023977 s 0.00001284074000068358 s 1.87
scatter_sum / IDefOpt / cpu / BothRev 0.000024233 s 0.000012844779957958964 s 1.89
scatter_sum / JaXPipe / cpu / Primal 0.00001 s 0.000008399699972869711 s 1.19
scatter_sum / Jax / cpu / Primal 0.00001 s 0.00000856758005284064 s 1.17
scatter_sum / HLOOpt / cpu / Primal 0.00001 s 0.000009258660020350362 s 1.08
scatter_sum / PartOpt / cpu / Primal 0.000008999999999999999 s 0.000009661359954407087 s 0.93
scatter_sum / IPartOpt / cpu / Primal 0.00001 s 0.000009734099976412837 s 1.03
scatter_sum / DefOpt / cpu / Primal 0.00001 s 0.000009198239977195045 s 1.09
scatter_sum / IDefOpt / cpu / Primal 0.00001 s 0.000008779059990047244 s 1.14
scatter_sum / JaXPipe / cpu / Forward 0.000015 s 0.000013061339986961684 s 1.15
scatter_sum / Jax / cpu / Forward 0.000015 s 0.000012101459969926508 s 1.24
scatter_sum / HLOOpt / cpu / Forward 0.000014 s 0.000012741700002152357 s 1.10
scatter_sum / PartOpt / cpu / Forward 0.000016 s 0.000012532019991340348 s 1.28
scatter_sum / IPartOpt / cpu / Forward 0.000015 s 0.0000133172999630915 s 1.13
scatter_sum / DefOpt / cpu / Forward 0.000015 s 0.00001335968000603316 s 1.12
scatter_sum / IDefOpt / cpu / Forward 0.000014 s 0.000013295580029080156 s 1.05
scatter_sum / JaXPipe / cpu / PreRev 0.000015 s 0.000012938739982928384 s 1.16
scatter_sum / JaXPipe / cpu / PostRev 0.000015 s 0.00001292862001719186 s 1.16
scatter_sum / JaXPipe / cpu / BothRev 0.000015 s 0.000012804760017388616 s 1.17
scatter_sum / Jax / cpu / BothRev 0.000015 s 0.000012629419989025336 s 1.19
scatter_sum / HLOOpt / cpu / PreRev 0.000015 s 0.000013125519999448442 s 1.14
scatter_sum / HLOOpt / cpu / PostRev 0.000015 s 0.000014698720024171053 s 1.02
scatter_sum / HLOOpt / cpu / BothRev 0.000015 s 0.000013126020039635478 s 1.14
scatter_sum / PartOpt / cpu / PreRev 0.000016 s 0.000013033599962000154 s 1.23
scatter_sum / PartOpt / cpu / PostRev 0.000015 s 0.000012915639981656568 s 1.16
scatter_sum / PartOpt / cpu / BothRev 0.000015 s 0.000012708979975286638 s 1.18
scatter_sum / IPartOpt / cpu / PreRev 0.000015 s 0.000012781279965565772 s 1.17
scatter_sum / IPartOpt / cpu / PostRev 0.000015 s 0.000013472300042849384 s 1.11
scatter_sum / IPartOpt / cpu / BothRev 0.000015 s 0.000012834419976570643 s 1.17
scatter_sum / DefOpt / cpu / PreRev 0.000015 s 0.000013145060020178787 s 1.14
scatter_sum / DefOpt / cpu / PostRev 0.000016 s 0.00001275038002859219 s 1.25
scatter_sum / DefOpt / cpu / BothRev 0.000016 s 0.000012998379997952724 s 1.23
scatter_sum / IDefOpt / cpu / PreRev 0.000015 s 0.000013091199998598312 s 1.15
scatter_sum / IDefOpt / cpu / PostRev 0.000015 s 0.00001284074000068358 s 1.17
scatter_sum / IDefOpt / cpu / BothRev 0.000015 s 0.000012844779957958964 s 1.17
slicing / JaXPipe / cpu / Primal 0.000006264780004130444 s 0.000007288319993676851 s 0.86
slicing / Jax / cpu / Primal 0.000006255160014916328 s 0.000006467119974331581 s 0.97
slicing / HLOOpt / cpu / Primal 0.000006507100006274413 s 0.000007340799984376645 s 0.89
slicing / PartOpt / cpu / Primal 0.000005997479993311572 s 0.000006723080005031079 s 0.89
slicing / IPartOpt / cpu / Primal 0.000006148100010250346 s 0.00000703373997566814 s 0.87
slicing / DefOpt / cpu / Primal 0.000006263859995669919 s 0.000006649759980064118 s 0.94
slicing / IDefOpt / cpu / Primal 0.000006177559971547452 s 0.000006718660024489509 s 0.92
slicing / JaXPipe / cpu / Forward 0.0000093867799751024 s 0.000010664820019883336 s 0.88
slicing / Jax / cpu / Forward 0.000009037659992827685 s 0.00001085290004084527 s 0.83
slicing / HLOOpt / cpu / Forward 0.000009287919992857496 s 0.000010787459987113834 s 0.86
slicing / PartOpt / cpu / Forward 0.00000872779995916062 s 0.000009920160018737078 s 0.88
slicing / IPartOpt / cpu / Forward 0.000009227700011251728 s 0.00001104420003684936 s 0.84
slicing / DefOpt / cpu / Forward 0.000008869200000845012 s 0.00001050316004693741 s 0.84
slicing / IDefOpt / cpu / Forward 0.000008759960019233404 s 0.00001045673998305574 s 0.84
slicing / JaXPipe / cpu / PreRev 0.000009797879993129757 s 0.00001086409999516036 s 0.90
slicing / JaXPipe / cpu / PostRev 0.000009650920019339537 s 0.000011132019972137642 s 0.87
slicing / JaXPipe / cpu / BothRev 0.000009962200037989532 s 0.000010977259998981026 s 0.91
slicing / Jax / cpu / BothRev 0.000009631900011299875 s 0.00001058601999829989 s 0.91
slicing / HLOOpt / cpu / PreRev 0.000009825240040299832 s 0.000011480219991426563 s 0.86
slicing / HLOOpt / cpu / PostRev 0.000014128759994491702 s 0.000013056880015938076 s 1.08
slicing / HLOOpt / cpu / BothRev 0.000009629039968785946 s 0.000011306099986541084 s 0.85
slicing / PartOpt / cpu / PreRev 0.000009361080046801362 s 0.000011312520027786376 s 0.83
slicing / PartOpt / cpu / PostRev 0.00000995435994809668 s 0.000010987319992636912 s 0.91
slicing / PartOpt / cpu / BothRev 0.000009767200017449795 s 0.000011153960022056708 s 0.88
slicing / IPartOpt / cpu / PreRev 0.000009848460022112704 s 0.000010705020013119791 s 0.92
slicing / IPartOpt / cpu / PostRev 0.00000962890000664629 s 0.000010705700015023467 s 0.90
slicing / IPartOpt / cpu / BothRev 0.000009350960008305264 s 0.000010749940001915092 s 0.87
slicing / DefOpt / cpu / PreRev 0.000009219019984811894 s 0.000011468539933048303 s 0.80
slicing / DefOpt / cpu / PostRev 0.000009596220015737345 s 0.000010730119965955964 s 0.89
slicing / DefOpt / cpu / BothRev 0.000009382520029248553 s 0.000011325499990562091 s 0.83
slicing / IDefOpt / cpu / PreRev 0.000009618899966881144 s 0.000010818559994731911 s 0.89
slicing / IDefOpt / cpu / PostRev 0.000009779260008144774 s 0.000011040860026696464 s 0.89
slicing / IDefOpt / cpu / BothRev 0.000009401299994351575 s 0.00001058544001352857 s 0.89
slicing / JaXPipe / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / Jax / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / HLOOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / PartOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / IPartOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / DefOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / IDefOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / JaXPipe / cuda / Forward 0.00000992 s 0.000010112 s 0.98
slicing / Jax / cuda / Forward 0.000010369 s 0.000009664 s 1.07
slicing / HLOOpt / cuda / Forward 0.000010049 s 0.000009728 s 1.03
slicing / PartOpt / cuda / Forward 0.000010177 s 0.000009312000000000002 s 1.09
slicing / IPartOpt / cuda / Forward 0.000010463 s 0.000009792 s 1.07
slicing / DefOpt / cuda / Forward 0.000009888 s 0.000010145 s 0.97
slicing / IDefOpt / cuda / Forward 0.000009952 s 0.000010048 s 0.99
slicing / JaXPipe / cuda / PreRev 0.00001024 s 0.000011232 s 0.91
slicing / JaXPipe / cuda / PostRev 0.000010016 s 0.000010976 s 0.91
slicing / JaXPipe / cuda / BothRev 0.000010175 s 0.00001008 s 1.01
slicing / Jax / cuda / BothRev 0.000010624 s 0.000009952 s 1.07
slicing / HLOOpt / cuda / PreRev 0.000010113 s 0.000009696 s 1.04
slicing / HLOOpt / cuda / PostRev 0.000010208 s 0.000009568 s 1.07
slicing / HLOOpt / cuda / BothRev 0.00001024 s 0.000009888 s 1.04
slicing / PartOpt / cuda / PreRev 0.000010176 s 0.0000096 s 1.06
slicing / PartOpt / cuda / PostRev 0.000010112 s 0.000009728 s 1.04
slicing / PartOpt / cuda / BothRev 0.000010112 s 0.000009983 s 1.01
slicing / IPartOpt / cuda / PreRev 0.000010496 s 0.000011137 s 0.94
slicing / IPartOpt / cuda / PostRev 0.000010112 s 0.000009568 s 1.06
slicing / IPartOpt / cuda / BothRev 0.00001024 s 0.000009953 s 1.03
slicing / DefOpt / cuda / PreRev 0.000010304 s 0.00000976 s 1.06
slicing / DefOpt / cuda / PostRev 0.000010368 s 0.000009633 s 1.08
slicing / DefOpt / cuda / BothRev 0.000010369 s 0.000009824 s 1.06
slicing / IDefOpt / cuda / PreRev 0.00001056 s 0.000009569 s 1.10
slicing / IDefOpt / cuda / PostRev 0.00001024 s 0.000009856 s 1.04
slicing / IDefOpt / cuda / BothRev 0.000010624 s 0.000009568 s 1.11
slicing / JaXPipe / tpu / Primal 9.61525e-7 s 9.63975e-7 s 1.00
slicing / Jax / tpu / Primal 9.64525e-7 s 9.6105e-7 s 1.00
slicing / HLOOpt / tpu / Primal 9.596e-7 s 9.61225e-7 s 1.00
slicing / PartOpt / tpu / Primal 9.619e-7 s 9.65325e-7 s 1.00
slicing / IPartOpt / tpu / Primal 9.50825e-7 s 9.59025e-7 s 0.99
slicing / DefOpt / tpu / Primal 9.667e-7 s 9.6145e-7 s 1.01
slicing / IDefOpt / tpu / Primal 9.56525e-7 s 9.6245e-7 s 0.99
slicing / JaXPipe / tpu / Forward 0.000001410025 s 0.000001406125 s 1.00
slicing / Jax / tpu / Forward 0.000001405775 s 0.000001412075 s 1.00
slicing / HLOOpt / tpu / Forward 0.00000151345 s 0.00000151405 s 1.00
slicing / PartOpt / tpu / Forward 0.0000014316000000000002 s 0.0000014341999999999998 s 1.00
slicing / IPartOpt / tpu / Forward 0.0000015081 s 0.000001513225 s 1.00
slicing / DefOpt / tpu / Forward 0.000001426275 s 0.000001437075 s 0.99
slicing / IDefOpt / tpu / Forward 0.0000015097 s 0.00000151655 s 1.00
slicing / JaXPipe / tpu / PreRev 0.000002331675 s 0.00000238275 s 0.98
slicing / JaXPipe / tpu / PostRev 0.0000025052000000000006 s 0.0000025262 s 0.99
slicing / JaXPipe / tpu / BothRev 0.00000234635 s 0.0000023967250000000003 s 0.98
slicing / Jax / tpu / BothRev 0.0000025279 s 0.000002552125 s 0.99
slicing / HLOOpt / tpu / PreRev 0.000002350225 s 0.00000239915 s 0.98
slicing / HLOOpt / tpu / PostRev 0.00000252275 s 0.000002548575 s 0.99
slicing / HLOOpt / tpu / BothRev 0.0000023509 s 0.0000024025000000000003 s 0.98
slicing / PartOpt / tpu / PreRev 0.000002523525 s 0.00000254185 s 0.99
slicing / PartOpt / tpu / PostRev 0.000002348175 s 0.0000024003 s 0.98
slicing / PartOpt / tpu / BothRev 0.000002521075 s 0.0000025502 s 0.99
slicing / IPartOpt / tpu / PreRev 0.0000023581 s 0.0000023919000000000003 s 0.99
slicing / IPartOpt / tpu / PostRev 0.0000025241 s 0.0000025419750000000003 s 0.99
slicing / IPartOpt / tpu / BothRev 0.000002338825 s 0.000002392475 s 0.98
slicing / DefOpt / tpu / PreRev 0.0000025221500000000005 s 0.0000025409 s 0.99
slicing / DefOpt / tpu / PostRev 0.000002343475 s 0.00000239555 s 0.98
slicing / DefOpt / tpu / BothRev 0.000002520225 s 0.0000025384750000000003 s 0.99
slicing / IDefOpt / tpu / PreRev 0.0000023393 s 0.0000023991 s 0.98
slicing / IDefOpt / tpu / PostRev 0.000002541225 s 0.000002548875 s 1.00
slicing / IDefOpt / tpu / BothRev 0.0000023485 s 0.00000240065 s 0.98
slicing / JaXPipe / cpu / Primal 0.000012785 s 0.000007288319993676851 s 1.75
slicing / Jax / cpu / Primal 0.000012563 s 0.000006467119974331581 s 1.94
slicing / HLOOpt / cpu / Primal 0.000012386 s 0.000007340799984376645 s 1.69
slicing / PartOpt / cpu / Primal 0.000012925 s 0.000006723080005031079 s 1.92
slicing / IPartOpt / cpu / Primal 0.000012555 s 0.00000703373997566814 s 1.78
slicing / DefOpt / cpu / Primal 0.000012387 s 0.000006649759980064118 s 1.86
slicing / IDefOpt / cpu / Primal 0.00001275 s 0.000006718660024489509 s 1.90
slicing / JaXPipe / cpu / Forward 0.000016951 s 0.000010664820019883336 s 1.59
slicing / Jax / cpu / Forward 0.000016896000000000002 s 0.00001085290004084527 s 1.56
slicing / HLOOpt / cpu / Forward 0.000016986 s 0.000010787459987113834 s 1.57
slicing / PartOpt / cpu / Forward 0.000016698000000000002 s 0.000009920160018737078 s 1.68
slicing / IPartOpt / cpu / Forward 0.000017259 s 0.00001104420003684936 s 1.56
slicing / DefOpt / cpu / Forward 0.000017134 s 0.00001050316004693741 s 1.63
slicing / IDefOpt / cpu / Forward 0.000017233999999999998 s 0.00001045673998305574 s 1.65
slicing / JaXPipe / cpu / PreRev 0.000018054 s 0.00001086409999516036 s 1.66
slicing / JaXPipe / cpu / PostRev 0.000017728 s 0.000011132019972137642 s 1.59
slicing / JaXPipe / cpu / BothRev 0.000017221 s 0.000010977259998981026 s 1.57
slicing / Jax / cpu / BothRev 0.000019444 s 0.00001058601999829989 s 1.84
slicing / HLOOpt / cpu / PreRev 0.000017759 s 0.000011480219991426563 s 1.55
slicing / HLOOpt / cpu / PostRev 0.000017721000000000002 s 0.000013056880015938076 s 1.36
slicing / HLOOpt / cpu / BothRev 0.000017099 s 0.000011306099986541084 s 1.51
slicing / PartOpt / cpu / PreRev 0.00001806 s 0.000011312520027786376 s 1.60
slicing / PartOpt / cpu / PostRev 0.000017417999999999998 s 0.000010987319992636912 s 1.59
slicing / PartOpt / cpu / BothRev 0.00001747 s 0.000011153960022056708 s 1.57
slicing / IPartOpt / cpu / PreRev 0.000017887 s 0.000010705020013119791 s 1.67
slicing / IPartOpt / cpu / PostRev 0.000017102 s 0.000010705700015023467 s 1.60
slicing / IPartOpt / cpu / BothRev 0.000017226 s 0.000010749940001915092 s 1.60
slicing / DefOpt / cpu / PreRev 0.000017615 s 0.000011468539933048303 s 1.54
slicing / DefOpt / cpu / PostRev 0.000017451999999999998 s 0.000010730119965955964 s 1.63
slicing / DefOpt / cpu / BothRev 0.000017692 s 0.000011325499990562091 s 1.56
slicing / IDefOpt / cpu / PreRev 0.000018094 s 0.000010818559994731911 s 1.67
slicing / IDefOpt / cpu / PostRev 0.000017821 s 0.000011040860026696464 s 1.61
slicing / IDefOpt / cpu / BothRev 0.000016893999999999998 s 0.00001058544001352857 s 1.60
slicing / JaXPipe / cpu / Primal 0.000008 s 0.000007288319993676851 s 1.10
slicing / Jax / cpu / Primal 0.000008 s 0.000006467119974331581 s 1.24
slicing / HLOOpt / cpu / Primal 0.000008 s 0.000007340799984376645 s 1.09
slicing / PartOpt / cpu / Primal 0.000008 s 0.000006723080005031079 s 1.19
slicing / IPartOpt / cpu / Primal 0.000008 s 0.00000703373997566814 s 1.14
slicing / DefOpt / cpu / Primal 0.000008 s 0.000006649759980064118 s 1.20
slicing / IDefOpt / cpu / Primal 0.000008 s 0.000006718660024489509 s 1.19
slicing / JaXPipe / cpu / Forward 0.000011 s 0.000010664820019883336 s 1.03
slicing / Jax / cpu / Forward 0.000011 s 0.00001085290004084527 s 1.01
slicing / HLOOpt / cpu / Forward 0.000011 s 0.000010787459987113834 s 1.02
slicing / PartOpt / cpu / Forward 0.000011 s 0.000009920160018737078 s 1.11
slicing / IPartOpt / cpu / Forward 0.000011 s 0.00001104420003684936 s 1.00
slicing / DefOpt / cpu / Forward 0.000011 s 0.00001050316004693741 s 1.05
slicing / IDefOpt / cpu / Forward 0.000011 s 0.00001045673998305574 s 1.05
slicing / JaXPipe / cpu / PreRev 0.000012 s 0.00001086409999516036 s 1.10
slicing / JaXPipe / cpu / PostRev 0.000011 s 0.000011132019972137642 s 0.99
slicing / JaXPipe / cpu / BothRev 0.000011 s 0.000010977259998981026 s 1.00
slicing / Jax / cpu / BothRev 0.000011 s 0.00001058601999829989 s 1.04
slicing / HLOOpt / cpu / PreRev 0.000011 s 0.000011480219991426563 s 0.96
slicing / HLOOpt / cpu / PostRev 0.000011 s 0.000013056880015938076 s 0.84
slicing / HLOOpt / cpu / BothRev 0.000012 s 0.000011306099986541084 s 1.06
slicing / PartOpt / cpu / PreRev 0.000011 s 0.000011312520027786376 s 0.97
slicing / PartOpt / cpu / PostRev 0.000011 s 0.000010987319992636912 s 1.00
slicing / PartOpt / cpu / BothRev 0.000012 s 0.000011153960022056708 s 1.08
slicing / IPartOpt / cpu / PreRev 0.000012 s 0.000010705020013119791 s 1.12
slicing / IPartOpt / cpu / PostRev 0.000012 s 0.000010705700015023467 s 1.12
slicing / IPartOpt / cpu / BothRev 0.000011 s 0.000010749940001915092 s 1.02
slicing / DefOpt / cpu / PreRev 0.000012 s 0.000011468539933048303 s 1.05
slicing / DefOpt / cpu / PostRev 0.000012 s 0.000010730119965955964 s 1.12
slicing / DefOpt / cpu / BothRev 0.000012 s 0.000011325499990562091 s 1.06
slicing / IDefOpt / cpu / PreRev 0.000011 s 0.000010818559994731911 s 1.02
slicing / IDefOpt / cpu / PostRev 0.000012 s 0.000011040860026696464 s 1.09
slicing / IDefOpt / cpu / BothRev 0.000012 s 0.00001058544001352857 s 1.13
sum / JaXPipe / cpu / Primal 0.0000075548800214164655 s 0.000008742739983063074 s 0.86
sum / Jax / cpu / Primal 0.000007359840010394691 s 0.000008432739987256355 s 0.87
sum / HLOOpt / cpu / Primal 0.000007465280014002928 s 0.00000868413997523021 s 0.86
sum / PartOpt / cpu / Primal 0.000007360359968515695 s 0.000008593839993409347 s 0.86
sum / IPartOpt / cpu / Primal 0.000007606960025441367 s 0.000008873579981809598 s 0.86
sum / DefOpt / cpu / Primal 0.000007630460040672914 s 0.000008341580014530337 s 0.91
sum / IDefOpt / cpu / Primal 0.000007197960030680406 s 0.000008160259958458483 s 0.88
sum / JaXPipe / cpu / Forward 0.00001099707992580079 s 0.000012359000020296664 s 0.89
sum / Jax / cpu / Forward 0.000010767999992822296 s 0.000012516539954958715 s 0.86
sum / HLOOpt / cpu / Forward 0.000010978419995808508 s 0.000012630199980776524 s 0.87
sum / PartOpt / cpu / Forward 0.000010579819991107795 s 0.000012017819990433054 s 0.88
sum / IPartOpt / cpu / Forward 0.00001107310004044848 s 0.0000124326400236896 s 0.89
sum / DefOpt / cpu / Forward 0.000010513959996387711 s 0.000012678339962803876 s 0.83
sum / IDefOpt / cpu / Forward 0.000010771079960250065 s 0.00001217218003148446 s 0.88
sum / JaXPipe / cpu / PreRev 0.00001079610002307163 s 0.000011711759980244096 s 0.92
sum / JaXPipe / cpu / PostRev 0.000010292640017723898 s 0.0000120298000274488 s 0.86
sum / JaXPipe / cpu / BothRev 0.000010720399986894337 s 0.000012270700017324998 s 0.87
sum / Jax / cpu / BothRev 0.00001015813996673387 s 0.000012310340025578626 s 0.83
sum / HLOOpt / cpu / PreRev 0.000011108700009572203 s 0.000012216640025144444 s 0.91
sum / HLOOpt / cpu / PostRev 0.000012311740001678117 s 0.000013434619959298288 s 0.92
sum / HLOOpt / cpu / BothRev 0.000010325359990019933 s 0.000012282459956622916 s 0.84
sum / PartOpt / cpu / PreRev 0.000010431940027046948 s 0.000012226860008013318 s 0.85
sum / PartOpt / cpu / PostRev 0.000010556060015005642 s 0.000011789480022343924 s 0.90
sum / PartOpt / cpu / BothRev 0.000010523640003157198 s 0.000012224460024299332 s 0.86
sum / IPartOpt / cpu / PreRev 0.000010397119958724942 s 0.000011990420025540516 s 0.87
sum / IPartOpt / cpu / PostRev 0.000010359279995100224 s 0.000012199719985801494 s 0.85
sum / IPartOpt / cpu / BothRev 0.000010047400000985363 s 0.000011894060035047004 s 0.84
sum / DefOpt / cpu / PreRev 0.00001012886000353319 s 0.000012119699986214985 s 0.84
sum / DefOpt / cpu / PostRev 0.000010618719970807432 s 0.00001206306002131896 s 0.88
sum / DefOpt / cpu / BothRev 0.00001030170000376529 s 0.00001233915999364399 s 0.83
sum / IDefOpt / cpu / PreRev 0.000010315719964637538 s 0.000011822360029327683 s 0.87
sum / IDefOpt / cpu / PostRev 0.000010298959950887366 s 0.00001169123997897259 s 0.88
sum / IDefOpt / cpu / BothRev 0.000010380080038885352 s 0.000011358259980625007 s 0.91
sum / JaXPipe / cuda / Primal 0.000002464 s 0.000002047 s 1.20
sum / Jax / cuda / Primal 0.000002464 s 0.000002047 s 1.20
sum / HLOOpt / cuda / Primal 0.000002463 s 0.000002047 s 1.20
sum / PartOpt / cuda / Primal 0.000002463 s 0.000002047 s 1.20
sum / IPartOpt / cuda / Primal 0.000002463 s 0.000002047 s 1.20
sum / DefOpt / cuda / Primal 0.000002463 s 0.000002048 s 1.20
sum / IDefOpt / cuda / Primal 0.000002463 s 0.000002047 s 1.20
sum / JaXPipe / cuda / Forward 0.000010816 s 0.000010303 s 1.05
sum / Jax / cuda / Forward 0.000010592 s 0.000009952 s 1.06
sum / HLOOpt / cuda / Forward 0.000010368 s 0.000009728 s 1.07
sum / PartOpt / cuda / Forward 0.000010592 s 0.000009984 s 1.06
sum / IPartOpt / cuda / Forward 0.000010624 s 0.000009536 s 1.11
sum / DefOpt / cuda / Forward 0.000010592 s 0.000009824 s 1.08
sum / IDefOpt / cuda / Forward 0.000010144 s 0.00001008 s 1.01
sum / JaXPipe / cuda / PreRev 0.000010208 s 0.000009632 s 1.06
sum / JaXPipe / cuda / PostRev 0.000010016 s 0.000009951 s 1.01
sum / JaXPipe / cuda / BothRev 0.000010304 s 0.000009792 s 1.05
sum / Jax / cuda / BothRev 0.000010913 s 0.000010144 s 1.08
sum / HLOOpt / cuda / PreRev 0.00001008 s 0.000009696 s 1.04
sum / HLOOpt / cuda / PostRev 0.000010271 s 0.00000944 s 1.09
sum / HLOOpt / cuda / BothRev 0.000010336 s 0.00000976 s 1.06
sum / PartOpt / cuda / PreRev 0.000010336 s 0.000009824 s 1.05
sum / PartOpt / cuda / PostRev 0.000010784 s 0.00001024 s 1.05
sum / PartOpt / cuda / BothRev 0.000010368 s 0.000009696 s 1.07
sum / IPartOpt / cuda / PreRev 0.000010208 s 0.000009824 s 1.04
sum / IPartOpt / cuda / PostRev 0.00001024 s 0.000009984 s 1.03
sum / IPartOpt / cuda / BothRev 0.00001024 s 0.000010047 s 1.02
sum / DefOpt / cuda / PreRev 0.000010048 s 0.000010144 s 0.99
sum / DefOpt / cuda / PostRev 0.000010112 s 0.000009985 s 1.01
sum / DefOpt / cuda / BothRev 0.000010432 s 0.00000992 s 1.05
sum / IDefOpt / cuda / PreRev 0.000009824 s 0.000009824 s 1
sum / IDefOpt / cuda / PostRev 0.000011424 s 0.00000976 s 1.17
sum / IDefOpt / cuda / BothRev 0.000011424 s 0.000009856 s 1.16
sum / JaXPipe / tpu / Primal 5.1755e-7 s 5.1025e-7 s 1.01
sum / Jax / tpu / Primal 5.47325e-7 s 5.4695e-7 s 1.00
sum / HLOOpt / tpu / Primal 5.17425e-7 s 5.10625e-7 s 1.01
sum / PartOpt / tpu / Primal 5.47875e-7 s 5.471999999999999e-7 s 1.00
sum / IPartOpt / tpu / Primal 5.176e-7 s 5.102750000000001e-7 s 1.01
sum / DefOpt / tpu / Primal 5.4765e-7 s 5.470250000000001e-7 s 1.00
sum / IDefOpt / tpu / Primal 5.17525e-7 s 5.104e-7 s 1.01
sum / JaXPipe / tpu / Forward 0.000001554775 s 0.000001546225 s 1.01
sum / Jax / tpu / Forward 0.0000015114 s 0.000001498375 s 1.01
sum / HLOOpt / tpu / Forward 0.000001527825 s 0.000001529575 s 1.00
sum / PartOpt / tpu / Forward 0.000001500025 s 0.000001495975 s 1.00
sum / IPartOpt / tpu / Forward 0.000001527525 s 0.0000015346499999999998 s 1.00
sum / DefOpt / tpu / Forward 0.000001500925 s 0.0000014994 s 1.00
sum / IDefOpt / tpu / Forward 0.000001531525 s 0.000001527925 s 1.00
sum / JaXPipe / tpu / PreRev 0.00000100725 s 0.0000010069500000000002 s 1.00
sum / JaXPipe / tpu / PostRev 0.000001035675 s 0.00000103485 s 1.00
sum / JaXPipe / tpu / BothRev 0.0000010078 s 0.000001003175 s 1.00
sum / Jax / tpu / BothRev 0.00000103335 s 0.000001038925 s 0.99
sum / HLOOpt / tpu / PreRev 9.99525e-7 s 0.0000010023 s 1.00
sum / HLOOpt / tpu / PostRev 0.000001033225 s 0.000001040175 s 0.99
sum / HLOOpt / tpu / BothRev 0.000001008525 s 0.0000010055000000000002 s 1.00
sum / PartOpt / tpu / PreRev 0.00000103305 s 0.0000010352 s 1.00
sum / PartOpt / tpu / PostRev 0.000001003725 s 9.97725e-7 s 1.01
sum / PartOpt / tpu / BothRev 0.0000010358 s 0.000001040425 s 1.00
sum / IPartOpt / tpu / PreRev 0.000001 s 0.000001005175 s 0.99
sum / IPartOpt / tpu / PostRev 0.000001033375 s 0.00000103545 s 1.00
sum / IPartOpt / tpu / BothRev 0.000001006875 s 0.0000010101 s 1.00
sum / DefOpt / tpu / PreRev 0.0000010386 s 0.000001034625 s 1.00
sum / DefOpt / tpu / PostRev 0.00000100065 s 9.9925e-7 s 1.00
sum / DefOpt / tpu / BothRev 0.000001033375 s 0.000001041325 s 0.99
sum / IDefOpt / tpu / PreRev 0.0000010037 s 0.000001001975 s 1.00
sum / IDefOpt / tpu / PostRev 0.00000104105 s 0.0000010411 s 1.00
sum / IDefOpt / tpu / BothRev 0.0000010089249999999998 s 0.00000100035 s 1.01
sum / JaXPipe / cpu / Primal 0.000015131 s 0.000008742739983063074 s 1.73
sum / Jax / cpu / Primal 0.000014627 s 0.000008432739987256355 s 1.73
sum / HLOOpt / cpu / Primal 0.000014552 s 0.00000868413997523021 s 1.68
sum / PartOpt / cpu / Primal 0.000015164 s 0.000008593839993409347 s 1.76
sum / IPartOpt / cpu / Primal 0.000014755 s 0.000008873579981809598 s 1.66
sum / DefOpt / cpu / Primal 0.000014929 s 0.000008341580014530337 s 1.79
sum / IDefOpt / cpu / Primal 0.000014809 s 0.000008160259958458483 s 1.81
sum / JaXPipe / cpu / Forward 0.000020253 s 0.000012359000020296664 s 1.64
sum / Jax / cpu / Forward 0.000020444 s 0.000012516539954958715 s 1.63
sum / HLOOpt / cpu / Forward 0.00002041 s 0.000012630199980776524 s 1.62
sum / PartOpt / cpu / Forward 0.000020496 s 0.000012017819990433054 s 1.71
sum / IPartOpt / cpu / Forward 0.000020554 s 0.0000124326400236896 s 1.65
sum / DefOpt / cpu / Forward 0.000020088 s 0.000012678339962803876 s 1.58
sum / IDefOpt / cpu / Forward 0.000020186 s 0.00001217218003148446 s 1.66
sum / JaXPipe / cpu / PreRev 0.000019471 s 0.000011711759980244096 s 1.66
sum / JaXPipe / cpu / PostRev 0.000019102 s 0.0000120298000274488 s 1.59
sum / JaXPipe / cpu / BothRev 0.00001896 s 0.000012270700017324998 s 1.55
sum / Jax / cpu / BothRev 0.000019143 s 0.000012310340025578626 s 1.56
sum / HLOOpt / cpu / PreRev 0.000018984 s 0.000012216640025144444 s 1.55
sum / HLOOpt / cpu / PostRev 0.000031719 s 0.000013434619959298288 s 2.36
sum / HLOOpt / cpu / BothRev 0.000019527 s 0.000012282459956622916 s 1.59
sum / PartOpt / cpu / PreRev 0.000018932 s 0.000012226860008013318 s 1.55
sum / PartOpt / cpu / PostRev 0.000018443 s 0.000011789480022343924 s 1.56
sum / PartOpt / cpu / BothRev 0.000018815 s 0.000012224460024299332 s 1.54
sum / IPartOpt / cpu / PreRev 0.000019652 s 0.000011990420025540516 s 1.64
sum / IPartOpt / cpu / PostRev 0.000019227 s 0.000012199719985801494 s 1.58
sum / IPartOpt / cpu / BothRev 0.000019631 s 0.000011894060035047004 s 1.65
sum / DefOpt / cpu / PreRev 0.000019228 s 0.000012119699986214985 s 1.59
sum / DefOpt / cpu / PostRev 0.00001964 s 0.00001206306002131896 s 1.63
sum / DefOpt / cpu / BothRev 0.000019718 s 0.00001233915999364399 s 1.60
sum / IDefOpt / cpu / PreRev 0.000019828 s 0.000011822360029327683 s 1.68
sum / IDefOpt / cpu / PostRev 0.000020068 s 0.00001169123997897259 s 1.72
sum / IDefOpt / cpu / BothRev 0.000019179 s 0.000011358259980625007 s 1.69
sum / JaXPipe / cpu / Primal 0.00001 s 0.000008742739983063074 s 1.14
sum / Jax / cpu / Primal 0.000008999999999999999 s 0.000008432739987256355 s 1.07
sum / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.00000868413997523021 s 1.04
sum / PartOpt / cpu / Primal 0.00001 s 0.000008593839993409347 s 1.16
sum / IPartOpt / cpu / Primal 0.00001 s 0.000008873579981809598 s 1.13
sum / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000008341580014530337 s 1.08
sum / IDefOpt / cpu / Primal 0.00001 s 0.000008160259958458483 s 1.23
sum / JaXPipe / cpu / Forward 0.000013 s 0.000012359000020296664 s 1.05
sum / Jax / cpu / Forward 0.000014 s 0.000012516539954958715 s 1.12
sum / HLOOpt / cpu / Forward 0.000014 s 0.000012630199980776524 s 1.11
sum / PartOpt / cpu / Forward 0.000013 s 0.000012017819990433054 s 1.08
sum / IPartOpt / cpu / Forward 0.000014 s 0.0000124326400236896 s 1.13
sum / DefOpt / cpu / Forward 0.000013 s 0.000012678339962803876 s 1.03
sum / IDefOpt / cpu / Forward 0.000014 s 0.00001217218003148446 s 1.15
sum / JaXPipe / cpu / PreRev 0.000013 s 0.000011711759980244096 s 1.11
sum / JaXPipe / cpu / PostRev 0.000013 s 0.0000120298000274488 s 1.08
sum / JaXPipe / cpu / BothRev 0.000012 s 0.000012270700017324998 s 0.98
sum / Jax / cpu / BothRev 0.000013 s 0.000012310340025578626 s 1.06
sum / HLOOpt / cpu / PreRev 0.000013 s 0.000012216640025144444 s 1.06
sum / HLOOpt / cpu / PostRev 0.000013 s 0.000013434619959298288 s 0.97
sum / HLOOpt / cpu / BothRev 0.000013 s 0.000012282459956622916 s 1.06
sum / PartOpt / cpu / PreRev 0.000013 s 0.000012226860008013318 s 1.06
sum / PartOpt / cpu / PostRev 0.000013 s 0.000011789480022343924 s 1.10
sum / PartOpt / cpu / BothRev 0.000013 s 0.000012224460024299332 s 1.06
sum / IPartOpt / cpu / PreRev 0.000013 s 0.000011990420025540516 s 1.08
sum / IPartOpt / cpu / PostRev 0.000013 s 0.000012199719985801494 s 1.07
sum / IPartOpt / cpu / BothRev 0.000013 s 0.000011894060035047004 s 1.09
sum / DefOpt / cpu / PreRev 0.000014 s 0.000012119699986214985 s 1.16
sum / DefOpt / cpu / PostRev 0.000013 s 0.00001206306002131896 s 1.08
sum / DefOpt / cpu / BothRev 0.000013 s 0.00001233915999364399 s 1.05
sum / IDefOpt / cpu / PreRev 0.000013 s 0.000011822360029327683 s 1.10
sum / IDefOpt / cpu / PostRev 0.000013 s 0.00001169123997897259 s 1.11
sum / IDefOpt / cpu / BothRev 0.000014 s 0.000011358259980625007 s 1.23
value_and_grad / JaXPipe / cpu / Primal 0.000013130079996699352 s 0.000015538480001850986 s 0.85
value_and_grad / Jax / cpu / Primal 0.000013007359939365417 s 0.000015070379977260016 s 0.86
value_and_grad / HLOOpt / cpu / Primal 0.000012811719989258565 s 0.000014645099963672691 s 0.87
value_and_grad / PartOpt / cpu / Primal 0.000012402680031300406 s 0.00001437700002497877 s 0.86
value_and_grad / IPartOpt / cpu / Primal 0.000012669500001720737 s 0.000014697280030304682 s 0.86
value_and_grad / DefOpt / cpu / Primal 0.000012529639971035069 s 0.00001450112002203241 s 0.86
value_and_grad / IDefOpt / cpu / Primal 0.00001245735999873432 s 0.000014610480011469918 s 0.85
value_and_grad / JaXPipe / cuda / Primal 0.00003328 s 0.000032352 s 1.03
value_and_grad / Jax / cuda / Primal 0.00003776 s 0.000032896000000000005 s 1.15
value_and_grad / HLOOpt / cuda / Primal 0.000037473 s 0.000032225 s 1.16
value_and_grad / PartOpt / cuda / Primal 0.000038048 s 0.000032385 s 1.17
value_and_grad / IPartOpt / cuda / Primal 0.000037567 s 0.000032672 s 1.15
value_and_grad / DefOpt / cuda / Primal 0.000037152 s 0.000032608 s 1.14
value_and_grad / IDefOpt / cuda / Primal 0.000033248 s 0.00003264 s 1.02
value_and_grad / JaXPipe / tpu / Primal 0 s 0 s 1
value_and_grad / Jax / tpu / Primal 0 s 0 s 1
value_and_grad / HLOOpt / tpu / Primal 0 s 0 s 1
value_and_grad / PartOpt / tpu / Primal 0 s 0 s 1
value_and_grad / IPartOpt / tpu / Primal 0 s 0 s 1
value_and_grad / DefOpt / tpu / Primal 0 s 0 s 1
value_and_grad / IDefOpt / tpu / Primal 0 s 0 s 1
value_and_grad / JaXPipe / cpu / Primal 0.000022864 s 0.000015538480001850986 s 1.47
value_and_grad / Jax / cpu / Primal 0.000022756 s 0.000015070379977260016 s 1.51
value_and_grad / HLOOpt / cpu / Primal 0.00002315 s 0.000014645099963672691 s 1.58
value_and_grad / PartOpt / cpu / Primal 0.000023052 s 0.00001437700002497877 s 1.60
value_and_grad / IPartOpt / cpu / Primal 0.00002378 s 0.000014697280030304682 s 1.62
value_and_grad / DefOpt / cpu / Primal 0.000023167 s 0.00001450112002203241 s 1.60
value_and_grad / IDefOpt / cpu / Primal 0.000023343 s 0.000014610480011469918 s 1.60
value_and_grad / JaXPipe / cpu / Primal 0.000016 s 0.000015538480001850986 s 1.03
value_and_grad / Jax / cpu / Primal 0.000015 s 0.000015070379977260016 s 1.00
value_and_grad / HLOOpt / cpu / Primal 0.000016 s 0.000014645099963672691 s 1.09
value_and_grad / PartOpt / cpu / Primal 0.000015 s 0.00001437700002497877 s 1.04
value_and_grad / IPartOpt / cpu / Primal 0.000016 s 0.000014697280030304682 s 1.09
value_and_grad / DefOpt / cpu / Primal 0.000015 s 0.00001450112002203241 s 1.03
value_and_grad / IDefOpt / cpu / Primal 0.000015 s 0.000014610480011469918 s 1.03
jaxmd20 / JaXPipe / cuda / Primal 0.001479522 s 0.001938565 s 0.76
jaxmd20 / Jax / cuda / Primal 0.001493218 s 0.001496676 s 1.00
jaxmd20 / HLOOpt / cuda / Primal 0.0013433619999999 s 0.0013105629999999 s 1.03
jaxmd20 / PartOpt / cuda / Primal 0.00137072 s 0.001318019 s 1.04
jaxmd20 / IPartOpt / cuda / Primal 0.001371778 s 0.001323556 s 1.04
jaxmd20 / DefOpt / cuda / Primal 0.000956833 s 0.000916066 s 1.04
jaxmd20 / IDefOpt / cuda / Primal 0.0009704 s 0.000951714 s 1.02
jaxmd20 / JaXPipe / cuda / Forward 0.00163581 s 0.001571588 s 1.04
jaxmd20 / Jax / cuda / Forward 0.001863684 s 0.0017808059999999 s 1.05
jaxmd20 / HLOOpt / cuda / Forward 0.001719586 s 0.001623371 s 1.06
jaxmd20 / PartOpt / cuda / Forward 0.001714562 s 0.001664068 s 1.03
jaxmd20 / IPartOpt / cuda / Forward 0.001710818 s 0.001614468 s 1.06
jaxmd20 / DefOpt / cuda / Forward 0.001723971 s 0.001624164 s 1.06
jaxmd20 / IDefOpt / cuda / Forward 0.0017154899999999 s 0.0016119399999999 s 1.06
jaxmd20 / JaXPipe / cuda / PreRev 0.002766659 s 0.002690471 s 1.03
jaxmd20 / JaXPipe / cuda / PostRev 0.005493702 s 0.005677523 s 0.97
jaxmd20 / JaXPipe / cuda / BothRev 0.002776674 s 0.002690601 s 1.03
jaxmd20 / Jax / cuda / BothRev 0.005466853 s 0.005316715 s 1.03
jaxmd20 / HLOOpt / cuda / PreRev 0.002878659 s 0.002768517 s 1.04
jaxmd20 / HLOOpt / cuda / PostRev 0.005540934 s 0.005294383 s 1.05
jaxmd20 / HLOOpt / cuda / BothRev 0.002813571 s 0.002714063 s 1.04
jaxmd20 / PartOpt / cuda / PreRev 0.002923587 s 0.002801832 s 1.04
jaxmd20 / PartOpt / cuda / PostRev 0.005566277 s 0.005661502 s 0.98
jaxmd20 / PartOpt / cuda / BothRev 0.0028407379999999 s 0.002745029 s 1.03
jaxmd20 / IPartOpt / cuda / PreRev 0.002919235 s 0.002794438 s 1.04
jaxmd20 / IPartOpt / cuda / PostRev 0.005624038 s 0.0054231819999999 s 1.04
jaxmd20 / IPartOpt / cuda / BothRev 0.0028427549999999 s 0.002751239 s 1.03
jaxmd20 / DefOpt / cuda / PreRev 0.002946243 s 0.002836042 s 1.04
jaxmd20 / DefOpt / cuda / PostRev 0.002864355 s 0.002758054 s 1.04
jaxmd20 / DefOpt / cuda / BothRev 0.002868547 s 0.0027710789999999 s 1.04
jaxmd20 / IDefOpt / cuda / PreRev 0.002933442 s 0.002802917 s 1.05
jaxmd20 / IDefOpt / cuda / PostRev 0.002367074 s 0.002298309 s 1.03
jaxmd20 / IDefOpt / cuda / BothRev 0.0028565459999999 s 0.0027711109999999 s 1.03
jaxmd20 / JaXPipe / tpu / Primal 0.00927798375 s 0.009277693125 s 1.00
jaxmd20 / Jax / tpu / Primal 0.009275096875 s 0.0092693106249999 s 1.00
jaxmd20 / HLOOpt / tpu / Primal 0.0091606806249999 s 0.009168335 s 1.00
jaxmd20 / PartOpt / tpu / Primal 0.00919678875 s 0.0092010406249999 s 1.00
jaxmd20 / IPartOpt / tpu / Primal 0.009200415 s 0.009196949375 s 1.00
jaxmd20 / DefOpt / tpu / Primal 0.00880501375 s 0.008796551875 s 1.00
jaxmd20 / IDefOpt / tpu / Primal 0.00869844 s 0.008697765625 s 1.00
jaxmd20 / JaXPipe / tpu / Forward 0.01741313625 s 0.017415084375 s 1.00
jaxmd20 / Jax / tpu / Forward 0.01872609 s 0.018733859375 s 1.00
jaxmd20 / HLOOpt / tpu / Forward 0.01740404875 s 0.017391626875 s 1.00
jaxmd20 / PartOpt / tpu / Forward 0.017422063125 s 0.017410939375 s 1.00
jaxmd20 / IPartOpt / tpu / Forward 0.0174149199999999 s 0.017415836875 s 1.00
jaxmd20 / DefOpt / tpu / Forward 0.017416939375 s 0.017405771875 s 1.00
jaxmd20 / IDefOpt / tpu / Forward 0.017412120625 s 0.0174111487499999 s 1.00
jaxmd20 / JaXPipe / tpu / PreRev 0.0254453325 s 0.02547067625 s 1.00
jaxmd20 / JaXPipe / tpu / PostRev 0.021866235625 s 0.021862679375 s 1.00
jaxmd20 / JaXPipe / tpu / BothRev 0.025432741875 s 0.025455455 s 1.00
jaxmd20 / Jax / tpu / BothRev 0.021862765625 s 0.021865015625 s 1.00
jaxmd20 / HLOOpt / tpu / PreRev 0.02555354125 s 0.02556869625 s 1.00
jaxmd20 / HLOOpt / tpu / PostRev 0.020708548125 s 0.02080791125 s 1.00
jaxmd20 / HLOOpt / tpu / BothRev 0.0256574825 s 0.0256773987499999 s 1.00
jaxmd20 / PartOpt / tpu / PreRev 0.02547597875 s 0.025461623125 s 1.00
jaxmd20 / PartOpt / tpu / PostRev 0.021507865625 s 0.02152470625 s 1.00
jaxmd20 / PartOpt / tpu / BothRev 0.02557842125 s 0.0255368143749999 s 1.00
jaxmd20 / IPartOpt / tpu / PreRev 0.025448341875 s 0.025459676875 s 1.00
jaxmd20 / IPartOpt / tpu / PostRev 0.021521585625 s 0.021515244375 s 1.00
jaxmd20 / IPartOpt / tpu / BothRev 0.025543025 s 0.02555456625 s 1.00
jaxmd20 / DefOpt / tpu / PreRev 0.02548425625 s 0.02545990625 s 1.00
jaxmd20 / DefOpt / tpu / PostRev 0.018812498125 s 0.018808618125 s 1.00
jaxmd20 / DefOpt / tpu / BothRev 0.0255785025 s 0.0255338281249999 s 1.00
jaxmd20 / IDefOpt / tpu / PreRev 0.025456756875 s 0.0254602499999999 s 1.00
jaxmd20 / IDefOpt / tpu / PostRev 0.018338653125 s 0.018301281875 s 1.00
jaxmd20 / IDefOpt / tpu / BothRev 0.025547111875 s 0.02555139 s 1.00
jaxmd40 / JaXPipe / cpu / Primal 0.068202021 s 0.0715142149999999 s 0.95
jaxmd40 / Jax / cpu / Primal 0.0762889569999999 s 0.071793129 s 1.06
jaxmd40 / HLOOpt / cpu / Primal 0.088505221 s 0.096050322 s 0.92
jaxmd40 / PartOpt / cpu / Primal 0.075134664 s 0.075406883 s 1.00
jaxmd40 / IPartOpt / cpu / Primal 0.074176085 s 0.072501867 s 1.02
jaxmd40 / DefOpt / cpu / Primal 0.097068663 s 0.093348485 s 1.04
jaxmd40 / IDefOpt / cpu / Primal 0.0981026379999999 s 0.096174862 s 1.02
jaxmd40 / JaXPipe / cpu / Forward 0.169362397 s 0.167727442 s 1.01
jaxmd40 / Jax / cpu / Forward 0.092802536 s 0.089430882 s 1.04
jaxmd40 / HLOOpt / cpu / Forward 0.1826705049999999 s 0.16773789 s 1.09
jaxmd40 / PartOpt / cpu / Forward 0.1817904719999999 s 0.16891999 s 1.08
jaxmd40 / IPartOpt / cpu / Forward 0.168493108 s 0.170297303 s 0.99
jaxmd40 / DefOpt / cpu / Forward 0.177448555 s 0.168942012 s 1.05
jaxmd40 / IDefOpt / cpu / Forward 0.165426677 s 0.172370059 s 0.96
jaxmd40 / JaXPipe / cpu / PreRev 0.246396962 s 0.246402501 s 1.00
jaxmd40 / JaXPipe / cpu / PostRev 0.150599096 s 0.140108943 s 1.07
jaxmd40 / JaXPipe / cpu / BothRev 0.256474981 s 0.234342806 s 1.09
jaxmd40 / Jax / cpu / BothRev 0.14381951 s 0.137563376 s 1.05
jaxmd40 / HLOOpt / cpu / PreRev 0.2249384489999999 s 0.226273474 s 0.99
jaxmd40 / HLOOpt / cpu / PostRev 0.182010844 s 0.178276109 s 1.02
jaxmd40 / HLOOpt / cpu / BothRev 0.261037696 s 0.248443251 s 1.05
jaxmd40 / PartOpt / cpu / PreRev 0.237610548 s 0.227452227 s 1.04
jaxmd40 / PartOpt / cpu / PostRev 0.126075797 s 0.130284111 s 0.97
jaxmd40 / PartOpt / cpu / BothRev 0.25525005 s 0.2465847699999999 s 1.04
jaxmd40 / IPartOpt / cpu / PreRev 0.240450019 s 0.22577963 s 1.06
jaxmd40 / IPartOpt / cpu / PostRev 0.143312512 s 0.14343332 s 1.00
jaxmd40 / IPartOpt / cpu / BothRev 0.241585476 s 0.2539809779999999 s 0.95
jaxmd40 / DefOpt / cpu / PreRev 0.245041572 s 0.230823172 s 1.06
jaxmd40 / DefOpt / cpu / PostRev 0.188940886 s 0.1840009849999999 s 1.03
jaxmd40 / DefOpt / cpu / BothRev 0.254320856 s 0.251601763 s 1.01
jaxmd40 / IDefOpt / cpu / PreRev 0.221363948 s 0.239430548 s 0.92
jaxmd40 / IDefOpt / cpu / PostRev 0.170232941 s 0.1760787939999999 s 0.97
jaxmd40 / IDefOpt / cpu / BothRev 0.2472362829999999 s 0.260008438 s 0.95
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / cuda / Primal 1.704148124 s 1.701965816 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / cuda / Primal 1.7072552559999998 s 1.7047771029999998 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / cuda / Primal 1.7179808989999998 s 1.714594805 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / cuda / Primal 1.698520706 s 1.696450168 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / cuda / Primal 1.696167039 s 1.694809473 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / cuda / Primal 1.667414723 s 1.664932468 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / cuda / Primal 1.916473119 s 1.920949623 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / tpu / Primal 3.038961810625 s 3.038568994375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / tpu / Primal 3.039549049375 s 3.039189840625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / tpu / Primal 3.1215633825 s 3.1215918150000004 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / tpu / Primal 3.06038183625 s 3.05983537625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / tpu / Primal 3.060598040625 s 3.060004419375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / tpu / Primal 2.102515459375 s 2.10238759125 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / tpu / Primal 2.948400034375 s 2.944661759375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / JaXPipe / cpu / Primal 6.291374588 s 6.304237769 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / Jax / cpu / Primal 6.476884226 s 6.35860167 s 1.02
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / HLOOpt / cpu / Primal 6.347617171 s 6.390747236999999 s 0.99
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / PartOpt / cpu / Primal 6.438195516 s 6.4081710780000005 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IPartOpt / cpu / Primal 6.420867025 s 6.465016108 s 0.99
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / DefOpt / cpu / Primal 2.6741122570000004 s 2.635569419 s 1.01
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IDefOpt / cpu / Primal 6.993192194000001 s 7.045801032 s 0.99

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

Copy link
Member

@wsmoses wsmoses left a comment

Choose a reason for hiding this comment

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

this seems to share a lot of functionality with

struct SumToReduceWindow
, is it possible to share some code as well [or perhaps extend the latter to support dynamic slice?]

@avik-pal avik-pal force-pushed the ap/more_reshape_simplify branch from d1b5e26 to 19db57f Compare January 1, 2026 21:14
Base automatically changed from ap/more_reshape_simplify to main January 3, 2026 22:26
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