Skip to content

Conversation

@xys-syx
Copy link
Collaborator

@xys-syx xys-syx commented Dec 30, 2025

CUDA_HOME=$HOME/cuda-12.1 /home/xys/llvm-project/build/bin/clang -fplugin=/home/xys/release/lib/ClangEnzyme-22.so -mllvm -raising-plugin-path=/home/xys/release/lib/libRaise.so --cuda-path=/home/xys/cuda-12.1 -fno-exceptions -O1 /home/xys/Enzyme-JAX/test/backend_test.cu -o output -mllvm -reactant-backend=rocm -L/opt/rocm/lib -lamdhip64

I have use this on amd gpu

@xys-syx xys-syx marked this pull request as draft December 30, 2025 09:40
@wsmoses wsmoses marked this pull request as ready for review December 30, 2025 09:57
@wsmoses wsmoses merged commit 426a717 into main Dec 30, 2025
26 checks passed
@wsmoses wsmoses deleted the add-rocdl-dialect-translation-support branch December 30, 2025 11:33
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: a4ea7f6 Previous: f9c8af4 Ratio
actmtch / JaXPipe / cpu / Primal 0.000006757719947927399 s 0.0000065985600031126525 s 1.02
actmtch / Jax / cpu / Primal 0.000006450480004787095 s 0.000006702579994453117 s 0.96
actmtch / HLOOpt / cpu / Primal 0.000007274599993252196 s 0.000008150699986799737 s 0.89
actmtch / PartOpt / cpu / Primal 0.00000656491999507125 s 0.00000670475998049369 s 0.98
actmtch / IPartOpt / cpu / Primal 0.0000066709599832393 s 0.000006933339973329566 s 0.96
actmtch / DefOpt / cpu / Primal 0.0000070877600046515 s 0.00000715312002284918 s 0.99
actmtch / IDefOpt / cpu / Primal 0.000006900880043758661 s 0.000007291700030691572 s 0.95
actmtch / JaXPipe / cpu / Forward 0.00001057744001627725 s 0.000011272979982095422 s 0.94
actmtch / Jax / cpu / Forward 0.00000967279997894366 s 0.000010052439965875238 s 0.96
actmtch / HLOOpt / cpu / Forward 0.000010663700013537891 s 0.000013041160000284436 s 0.82
actmtch / PartOpt / cpu / Forward 0.00001069995994839701 s 0.000010845240003618528 s 0.99
actmtch / IPartOpt / cpu / Forward 0.00001078471998880559 s 0.000011831660021925928 s 0.91
actmtch / DefOpt / cpu / Forward 0.000010660800007826764 s 0.000010729999958130066 s 0.99
actmtch / IDefOpt / cpu / Forward 0.000010620299999573037 s 0.000010879839974222704 s 0.98
actmtch / JaXPipe / cpu / PreRev 0.000010701219998736632 s 0.00001105902000745118 s 0.97
actmtch / JaXPipe / cpu / PostRev 0.000009913919993778107 s 0.000010372240021752075 s 0.96
actmtch / JaXPipe / cpu / BothRev 0.000011367900006007403 s 0.000011834380002255784 s 0.96
actmtch / Jax / cpu / BothRev 0.000009272819979742054 s 0.000009667100039223442 s 0.96
actmtch / HLOOpt / cpu / PreRev 0.000011280020007689018 s 0.000011229740048293024 s 1.00
actmtch / HLOOpt / cpu / PostRev 0.00001297557999350829 s 0.000013224839995018555 s 0.98
actmtch / HLOOpt / cpu / BothRev 0.00001066507996256405 s 0.000011197019994142463 s 0.95
actmtch / PartOpt / cpu / PreRev 0.000011013639996235725 s 0.00001090343999749166 s 1.01
actmtch / PartOpt / cpu / PostRev 0.000009390740005983387 s 0.000010236600010102848 s 0.92
actmtch / PartOpt / cpu / BothRev 0.00001103204001992708 s 0.00001140428006692673 s 0.97
actmtch / IPartOpt / cpu / PreRev 0.000010940320071313182 s 0.00001053990002219507 s 1.04
actmtch / IPartOpt / cpu / PostRev 0.000009874960041997838 s 0.000009621200024412249 s 1.03
actmtch / IPartOpt / cpu / BothRev 0.000011172779986736714 s 0.000011268559992458904 s 0.99
actmtch / DefOpt / cpu / PreRev 0.00001061310004843108 s 0.000010980640008710906 s 0.97
actmtch / DefOpt / cpu / PostRev 0.00001145797997196496 s 0.000010973840016959002 s 1.04
actmtch / DefOpt / cpu / BothRev 0.000010816619987963347 s 0.000010829479988387902 s 1.00
actmtch / IDefOpt / cpu / PreRev 0.000010647740027707189 s 0.000010740020024968544 s 0.99
actmtch / IDefOpt / cpu / PostRev 0.000011365899927113788 s 0.000011196139985258924 s 1.02
actmtch / IDefOpt / cpu / BothRev 0.00001094387998819002 s 0.000010987959994963602 s 1.00
actmtch / JaXPipe / cuda / Primal 0.000002016 s 0.0000024 s 0.84
actmtch / Jax / cuda / Primal 0.000002015 s 0.0000024 s 0.84
actmtch / HLOOpt / cuda / Primal 0.000002016 s 0.0000024 s 0.84
actmtch / PartOpt / cuda / Primal 0.000002015 s 0.0000024 s 0.84
actmtch / IPartOpt / cuda / Primal 0.000002015 s 0.0000024 s 0.84
actmtch / DefOpt / cuda / Primal 0.000002016 s 0.0000024 s 0.84
actmtch / IDefOpt / cuda / Primal 0.000002016 s 0.0000024 s 0.84
actmtch / JaXPipe / cuda / Forward 0.000009664 s 0.000010528 s 0.92
actmtch / Jax / cuda / Forward 0.000009855 s 0.000010336 s 0.95
actmtch / HLOOpt / cuda / Forward 0.00000992 s 0.000010849 s 0.91
actmtch / PartOpt / cuda / Forward 0.000009888 s 0.00001056 s 0.94
actmtch / IPartOpt / cuda / Forward 0.000009696 s 0.00001136 s 0.85
actmtch / DefOpt / cuda / Forward 0.000009633 s 0.000011424 s 0.84
actmtch / IDefOpt / cuda / Forward 0.000010112 s 0.000011552 s 0.88
actmtch / JaXPipe / cuda / PreRev 0.000009568 s 0.000010528 s 0.91
actmtch / JaXPipe / cuda / PostRev 0.00000976 s 0.000010528 s 0.93
actmtch / JaXPipe / cuda / BothRev 0.000009984 s 0.000010848 s 0.92
actmtch / Jax / cuda / BothRev 0.00001024 s 0.000010944 s 0.94
actmtch / HLOOpt / cuda / PreRev 0.000010176 s 0.00001072 s 0.95
actmtch / HLOOpt / cuda / PostRev 0.000010272 s 0.000010848 s 0.95
actmtch / HLOOpt / cuda / BothRev 0.000010239 s 0.000010497 s 0.98
actmtch / PartOpt / cuda / PreRev 0.000010272 s 0.000010497 s 0.98
actmtch / PartOpt / cuda / PostRev 0.000010208 s 0.00001088 s 0.94
actmtch / PartOpt / cuda / BothRev 0.0000104 s 0.000010816 s 0.96
actmtch / IPartOpt / cuda / PreRev 0.000010272 s 0.000010944 s 0.94
actmtch / IPartOpt / cuda / PostRev 0.000010016 s 0.0000112 s 0.89
actmtch / IPartOpt / cuda / BothRev 0.0000104 s 0.000009792 s 1.06
actmtch / DefOpt / cuda / PreRev 0.000011392 s 0.000011073 s 1.03
actmtch / DefOpt / cuda / PostRev 0.00001104 s 0.000011296 s 0.98
actmtch / DefOpt / cuda / BothRev 0.000010656 s 0.000010559 s 1.01
actmtch / IDefOpt / cuda / PreRev 0.000010464 s 0.00001088 s 0.96
actmtch / IDefOpt / cuda / PostRev 0.000009919 s 0.000010465 s 0.95
actmtch / IDefOpt / cuda / BothRev 0.000010176 s 0.000010496 s 0.97
actmtch / JaXPipe / tpu / Primal 5.637e-7 s 5.63e-7 s 1.00
actmtch / Jax / tpu / Primal 5.9695e-7 s 5.96675e-7 s 1.00
actmtch / HLOOpt / tpu / Primal 0.0000021006 s 0.0000021012500000000004 s 1.00
actmtch / PartOpt / tpu / Primal 5.96575e-7 s 5.96825e-7 s 1.00
actmtch / IPartOpt / tpu / Primal 5.527750000000001e-7 s 5.522e-7 s 1.00
actmtch / DefOpt / tpu / Primal 0.00000216095 s 0.0000021646 s 1.00
actmtch / IDefOpt / tpu / Primal 0.00000209695 s 0.00000209885 s 1.00
actmtch / JaXPipe / tpu / Forward 0.0000038255 s 0.000003823274999999999 s 1.00
actmtch / Jax / tpu / Forward 0.00000120735 s 0.000001211375 s 1.00
actmtch / HLOOpt / tpu / Forward 0.000003929575 s 0.000003927575 s 1.00
actmtch / PartOpt / tpu / Forward 0.000003911 s 0.000003914475 s 1.00
actmtch / IPartOpt / tpu / Forward 0.0000039371 s 0.0000039378 s 1.00
actmtch / DefOpt / tpu / Forward 0.00000391635 s 0.00000391215 s 1.00
actmtch / IDefOpt / tpu / Forward 0.000003950175 s 0.000003938875 s 1.00
actmtch / JaXPipe / tpu / PreRev 0.0000034804750000000003 s 0.000003479625 s 1.00
actmtch / JaXPipe / tpu / PostRev 0.0000016411 s 0.000001636325 s 1.00
actmtch / JaXPipe / tpu / BothRev 0.0000034785 s 0.000003476425 s 1.00
actmtch / Jax / tpu / BothRev 0.00000163525 s 0.000001631775 s 1.00
actmtch / HLOOpt / tpu / PreRev 0.000003477675 s 0.000003472925 s 1.00
actmtch / HLOOpt / tpu / PostRev 0.0000033989 s 0.000003407575 s 1.00
actmtch / HLOOpt / tpu / BothRev 0.0000034966750000000003 s 0.0000034714 s 1.01
actmtch / PartOpt / tpu / PreRev 0.0000034109250000000003 s 0.000003412975 s 1.00
actmtch / PartOpt / tpu / PostRev 0.00000159505 s 0.0000015937 s 1.00
actmtch / PartOpt / tpu / BothRev 0.000003415275 s 0.000003415075 s 1.00
actmtch / IPartOpt / tpu / PreRev 0.000003467825 s 0.0000034804 s 1.00
actmtch / IPartOpt / tpu / PostRev 0.0000016334 s 0.00000163695 s 1.00
actmtch / IPartOpt / tpu / BothRev 0.000003468525 s 0.000003474675 s 1.00
actmtch / DefOpt / tpu / PreRev 0.000003416225 s 0.0000034107500000000003 s 1.00
actmtch / DefOpt / tpu / PostRev 0.00000341415 s 0.0000034171000000000003 s 1.00
actmtch / DefOpt / tpu / BothRev 0.0000034099 s 0.0000034122 s 1.00
actmtch / IDefOpt / tpu / PreRev 0.000003478 s 0.00000347625 s 1.00
actmtch / IDefOpt / tpu / PostRev 0.00000340405 s 0.000003407825 s 1.00
actmtch / IDefOpt / tpu / BothRev 0.00000347035 s 0.000003473225 s 1.00
actmtch / JaXPipe / cpu / Primal 0.000013022 s 0.0000065985600031126525 s 1.97
actmtch / Jax / cpu / Primal 0.000013563 s 0.000006702579994453117 s 2.02
actmtch / HLOOpt / cpu / Primal 0.000014249 s 0.000008150699986799737 s 1.75
actmtch / PartOpt / cpu / Primal 0.000013546 s 0.00000670475998049369 s 2.02
actmtch / IPartOpt / cpu / Primal 0.000013609 s 0.000006933339973329566 s 1.96
actmtch / DefOpt / cpu / Primal 0.000014339 s 0.00000715312002284918 s 2.00
actmtch / IDefOpt / cpu / Primal 0.000013987 s 0.000007291700030691572 s 1.92
actmtch / JaXPipe / cpu / Forward 0.000019302 s 0.000011272979982095422 s 1.71
actmtch / Jax / cpu / Forward 0.000018161 s 0.000010052439965875238 s 1.81
actmtch / HLOOpt / cpu / Forward 0.00001923 s 0.000013041160000284436 s 1.47
actmtch / PartOpt / cpu / Forward 0.000019289 s 0.000010845240003618528 s 1.78
actmtch / IPartOpt / cpu / Forward 0.00001903 s 0.000011831660021925928 s 1.61
actmtch / DefOpt / cpu / Forward 0.000019178 s 0.000010729999958130066 s 1.79
actmtch / IDefOpt / cpu / Forward 0.000019262000000000003 s 0.000010879839974222704 s 1.77
actmtch / JaXPipe / cpu / PreRev 0.000019617 s 0.00001105902000745118 s 1.77
actmtch / JaXPipe / cpu / PostRev 0.000017514 s 0.000010372240021752075 s 1.69
actmtch / JaXPipe / cpu / BothRev 0.000019363 s 0.000011834380002255784 s 1.64
actmtch / Jax / cpu / BothRev 0.000017742 s 0.000009667100039223442 s 1.84
actmtch / HLOOpt / cpu / PreRev 0.000019182 s 0.000011229740048293024 s 1.71
actmtch / HLOOpt / cpu / PostRev 0.000019379 s 0.000013224839995018555 s 1.47
actmtch / HLOOpt / cpu / BothRev 0.000019265 s 0.000011197019994142463 s 1.72
actmtch / PartOpt / cpu / PreRev 0.000019277 s 0.00001090343999749166 s 1.77
actmtch / PartOpt / cpu / PostRev 0.000017596 s 0.000010236600010102848 s 1.72
actmtch / PartOpt / cpu / BothRev 0.000019632 s 0.00001140428006692673 s 1.72
actmtch / IPartOpt / cpu / PreRev 0.000019359 s 0.00001053990002219507 s 1.84
actmtch / IPartOpt / cpu / PostRev 0.000017840000000000002 s 0.000009621200024412249 s 1.85
actmtch / IPartOpt / cpu / BothRev 0.000019497 s 0.000011268559992458904 s 1.73
actmtch / DefOpt / cpu / PreRev 0.000019408 s 0.000010980640008710906 s 1.77
actmtch / DefOpt / cpu / PostRev 0.000019521 s 0.000010973840016959002 s 1.78
actmtch / DefOpt / cpu / BothRev 0.00002016 s 0.000010829479988387902 s 1.86
actmtch / IDefOpt / cpu / PreRev 0.00002051 s 0.000010740020024968544 s 1.91
actmtch / IDefOpt / cpu / PostRev 0.000020513 s 0.000011196139985258924 s 1.83
actmtch / IDefOpt / cpu / BothRev 0.000019769 s 0.000010987959994963602 s 1.80
actmtch / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.0000065985600031126525 s 1.36
actmtch / Jax / cpu / Primal 0.000008999999999999999 s 0.000006702579994453117 s 1.34
actmtch / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000008150699986799737 s 1.10
actmtch / PartOpt / cpu / Primal 0.000008999999999999999 s 0.00000670475998049369 s 1.34
actmtch / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000006933339973329566 s 1.30
actmtch / DefOpt / cpu / Primal 0.00001 s 0.00000715312002284918 s 1.40
actmtch / IDefOpt / cpu / Primal 0.00001 s 0.000007291700030691572 s 1.37
actmtch / JaXPipe / cpu / Forward 0.000015 s 0.000011272979982095422 s 1.33
actmtch / Jax / cpu / Forward 0.000013 s 0.000010052439965875238 s 1.29
actmtch / HLOOpt / cpu / Forward 0.000013 s 0.000013041160000284436 s 1.00
actmtch / PartOpt / cpu / Forward 0.000013 s 0.000010845240003618528 s 1.20
actmtch / IPartOpt / cpu / Forward 0.000013 s 0.000011831660021925928 s 1.10
actmtch / DefOpt / cpu / Forward 0.000014 s 0.000010729999958130066 s 1.30
actmtch / IDefOpt / cpu / Forward 0.000013 s 0.000010879839974222704 s 1.19
actmtch / JaXPipe / cpu / PreRev 0.000013 s 0.00001105902000745118 s 1.18
actmtch / JaXPipe / cpu / PostRev 0.000013 s 0.000010372240021752075 s 1.25
actmtch / JaXPipe / cpu / BothRev 0.000013 s 0.000011834380002255784 s 1.10
actmtch / Jax / cpu / BothRev 0.000012 s 0.000009667100039223442 s 1.24
actmtch / HLOOpt / cpu / PreRev 0.000014 s 0.000011229740048293024 s 1.25
actmtch / HLOOpt / cpu / PostRev 0.000013 s 0.000013224839995018555 s 0.98
actmtch / HLOOpt / cpu / BothRev 0.000014 s 0.000011197019994142463 s 1.25
actmtch / PartOpt / cpu / PreRev 0.000014 s 0.00001090343999749166 s 1.28
actmtch / PartOpt / cpu / PostRev 0.000012 s 0.000010236600010102848 s 1.17
actmtch / PartOpt / cpu / BothRev 0.000014 s 0.00001140428006692673 s 1.23
actmtch / IPartOpt / cpu / PreRev 0.000013 s 0.00001053990002219507 s 1.23
actmtch / IPartOpt / cpu / PostRev 0.000012 s 0.000009621200024412249 s 1.25
actmtch / IPartOpt / cpu / BothRev 0.000014 s 0.000011268559992458904 s 1.24
actmtch / DefOpt / cpu / PreRev 0.000014 s 0.000010980640008710906 s 1.27
actmtch / DefOpt / cpu / PostRev 0.000015 s 0.000010973840016959002 s 1.37
actmtch / DefOpt / cpu / BothRev 0.000014 s 0.000010829479988387902 s 1.29
actmtch / IDefOpt / cpu / PreRev 0.000014 s 0.000010740020024968544 s 1.30
actmtch / IDefOpt / cpu / PostRev 0.000013 s 0.000011196139985258924 s 1.16
actmtch / IDefOpt / cpu / BothRev 0.000014 s 0.000010987959994963602 s 1.27
add_one / JaXPipe / cpu / Primal 0.000006754019996151328 s 0.000006855099973108736 s 0.99
add_one / Jax / cpu / Primal 0.000006715199997415767 s 0.000006435440000132076 s 1.04
add_one / HLOOpt / cpu / Primal 0.000006684820036753081 s 0.00000694028000907565 s 0.96
add_one / PartOpt / cpu / Primal 0.00000645683999209723 s 0.00000638672001514351 s 1.01
add_one / IPartOpt / cpu / Primal 0.000006681220047539682 s 0.0000071436000052926825 s 0.94
add_one / DefOpt / cpu / Primal 0.000006489159959528479 s 0.00000677024003380211 s 0.96
add_one / IDefOpt / cpu / Primal 0.000006709000017508515 s 0.000006581240022569545 s 1.02
add_one / JaXPipe / cpu / Forward 0.000010054060057882452 s 0.000010301519969289077 s 0.98
add_one / Jax / cpu / Forward 0.00000968522000221128 s 0.000009633440031393548 s 1.01
add_one / HLOOpt / cpu / Forward 0.000009920919956130092 s 0.000010201359973507351 s 0.97
add_one / PartOpt / cpu / Forward 0.000009765299992068322 s 0.000009867159988061758 s 0.99
add_one / IPartOpt / cpu / Forward 0.000009807599926716648 s 0.000010082579992740647 s 0.97
add_one / DefOpt / cpu / Forward 0.00000998275998426834 s 0.00000990603997706785 s 1.01
add_one / IDefOpt / cpu / Forward 0.000009796580015972725 s 0.000009821160019782838 s 1.00
add_one / JaXPipe / cpu / PreRev 0.000011570500009838723 s 0.000012167639997642256 s 0.95
add_one / JaXPipe / cpu / PostRev 0.000011568279960556538 s 0.000011753200051316523 s 0.98
add_one / JaXPipe / cpu / BothRev 0.00001179761997263995 s 0.000012313679999351734 s 0.96
add_one / Jax / cpu / BothRev 0.000011012319964720518 s 0.000011284179990980192 s 0.98
add_one / HLOOpt / cpu / PreRev 0.000012138300016886204 s 0.000012106219974157285 s 1.00
add_one / HLOOpt / cpu / PostRev 0.000013493159976860624 s 0.000013971079943075892 s 0.97
add_one / HLOOpt / cpu / BothRev 0.000011263740043432336 s 0.000011830500034193392 s 0.95
add_one / PartOpt / cpu / PreRev 0.00001116956000259961 s 0.000012122320003982168 s 0.92
add_one / PartOpt / cpu / PostRev 0.000011480319990369026 s 0.000012044100012644776 s 0.95
add_one / PartOpt / cpu / BothRev 0.00001171201998658944 s 0.00001201416000185418 s 0.97
add_one / IPartOpt / cpu / PreRev 0.0000111229799585999 s 0.00001151484000729397 s 0.97
add_one / IPartOpt / cpu / PostRev 0.00001240751999830536 s 0.000011937199988096836 s 1.04
add_one / IPartOpt / cpu / BothRev 0.000011488219961393042 s 0.00001157921998128586 s 0.99
add_one / DefOpt / cpu / PreRev 0.000011002960018231534 s 0.00001229476002663432 s 0.89
add_one / DefOpt / cpu / PostRev 0.000011300379992462694 s 0.000011425299999245907 s 0.99
add_one / DefOpt / cpu / BothRev 0.000011184379991391324 s 0.00001220008000927919 s 0.92
add_one / IDefOpt / cpu / PreRev 0.000011057259962399258 s 0.000011894500012203937 s 0.93
add_one / IDefOpt / cpu / PostRev 0.000011434639955041348 s 0.000011855460024889908 s 0.96
add_one / IDefOpt / cpu / BothRev 0.000011545000061232714 s 0.000011839600065286504 s 0.98
add_one / JaXPipe / cuda / Primal 0.0000019200000000000003 s 0.000002304 s 0.83
add_one / Jax / cuda / Primal 0.0000019200000000000003 s 0.000002303 s 0.83
add_one / HLOOpt / cuda / Primal 0.000001919 s 0.000002303 s 0.83
add_one / PartOpt / cuda / Primal 0.0000019200000000000003 s 0.000002303 s 0.83
add_one / IPartOpt / cuda / Primal 0.000001919 s 0.000002303 s 0.83
add_one / DefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002304 s 0.83
add_one / IDefOpt / cuda / Primal 0.0000019200000000000003 s 0.000002303 s 0.83
add_one / JaXPipe / cuda / Forward 0.000009952 s 0.00001072 s 0.93
add_one / Jax / cuda / Forward 0.000010144 s 0.000009792 s 1.04
add_one / HLOOpt / cuda / Forward 0.000010144 s 0.0000104 s 0.98
add_one / PartOpt / cuda / Forward 0.000010112 s 0.000010112 s 1
add_one / IPartOpt / cuda / Forward 0.000009856 s 0.00001104 s 0.89
add_one / DefOpt / cuda / Forward 0.000009856 s 0.000011072 s 0.89
add_one / IDefOpt / cuda / Forward 0.000009985 s 0.000010816 s 0.92
add_one / JaXPipe / cuda / PreRev 0.000024704 s 0.000026433000000000003 s 0.93
add_one / JaXPipe / cuda / PostRev 0.000024545 s 0.000026208 s 0.94
add_one / JaXPipe / cuda / BothRev 0.000024704 s 0.000025824 s 0.96
add_one / Jax / cuda / BothRev 0.000024832 s 0.000025632 s 0.97
add_one / HLOOpt / cuda / PreRev 0.000025407 s 0.000025856 s 0.98
add_one / HLOOpt / cuda / PostRev 0.000024736 s 0.000026016 s 0.95
add_one / HLOOpt / cuda / BothRev 0.000024832 s 0.000026272 s 0.95
add_one / PartOpt / cuda / PreRev 0.000025184 s 0.0000264 s 0.95
add_one / PartOpt / cuda / PostRev 0.000024991 s 0.00003008 s 0.83
add_one / PartOpt / cuda / BothRev 0.000024607 s 0.000030273 s 0.81
add_one / IPartOpt / cuda / PreRev 0.000025728 s 0.000030784 s 0.84
add_one / IPartOpt / cuda / PostRev 0.000029184 s 0.000030367 s 0.96
add_one / IPartOpt / cuda / BothRev 0.00002464 s 0.00002592 s 0.95
add_one / DefOpt / cuda / PreRev 0.000025312 s 0.000026016 s 0.97
add_one / DefOpt / cuda / PostRev 0.000024928 s 0.000026048 s 0.96
add_one / DefOpt / cuda / BothRev 0.000024992 s 0.000026304 s 0.95
add_one / IDefOpt / cuda / PreRev 0.00002512 s 0.000030849 s 0.81
add_one / IDefOpt / cuda / PostRev 0.0000248 s 0.000030592 s 0.81
add_one / IDefOpt / cuda / BothRev 0.000024928 s 0.000030752 s 0.81
add_one / JaXPipe / tpu / Primal 0.000001424525 s 0.0000014289250000000005 s 1.00
add_one / Jax / tpu / Primal 0.000001400675 s 0.000001408075 s 0.99
add_one / HLOOpt / tpu / Primal 0.000001429775 s 0.0000014253750000000003 s 1.00
add_one / PartOpt / tpu / Primal 0.0000014086000000000002 s 0.00000141215 s 1.00
add_one / IPartOpt / tpu / Primal 0.0000014301749999999998 s 0.000001434125 s 1.00
add_one / DefOpt / tpu / Primal 0.000001403925 s 0.00000140095 s 1.00
add_one / IDefOpt / tpu / Primal 0.0000014201 s 0.000001426525 s 1.00
add_one / JaXPipe / tpu / Forward 0.00000184865 s 0.0000018554 s 1.00
add_one / Jax / tpu / Forward 0.000001844225 s 0.000001839225 s 1.00
add_one / HLOOpt / tpu / Forward 0.000001849725 s 0.0000018481 s 1.00
add_one / PartOpt / tpu / Forward 0.0000018421 s 0.0000018365 s 1.00
add_one / IPartOpt / tpu / Forward 0.00000184335 s 0.0000018448 s 1.00
add_one / DefOpt / tpu / Forward 0.00000183725 s 0.000001842525 s 1.00
add_one / IDefOpt / tpu / Forward 0.000001845175 s 0.000001854225 s 1.00
add_one / JaXPipe / tpu / PreRev 0.000002237025 s 0.000002245275 s 1.00
add_one / JaXPipe / tpu / PostRev 0.000002240275 s 0.0000022363750000000003 s 1.00
add_one / JaXPipe / tpu / BothRev 0.0000022461 s 0.0000022528500000000004 s 1.00
add_one / Jax / tpu / BothRev 0.000002244475 s 0.000002235175 s 1.00
add_one / HLOOpt / tpu / PreRev 0.000002237525 s 0.0000022327 s 1.00
add_one / HLOOpt / tpu / PostRev 0.000002241425 s 0.00000223495 s 1.00
add_one / HLOOpt / tpu / BothRev 0.00000224755 s 0.0000022383 s 1.00
add_one / PartOpt / tpu / PreRev 0.000002243625 s 0.0000022333500000000004 s 1.00
add_one / PartOpt / tpu / PostRev 0.00000223575 s 0.000002231125 s 1.00
add_one / PartOpt / tpu / BothRev 0.0000022364 s 0.00000224385 s 1.00
add_one / IPartOpt / tpu / PreRev 0.0000022343 s 0.000002240875 s 1.00
add_one / IPartOpt / tpu / PostRev 0.000002233125 s 0.000002237275 s 1.00
add_one / IPartOpt / tpu / BothRev 0.000002242525 s 0.0000022348 s 1.00
add_one / DefOpt / tpu / PreRev 0.0000022466 s 0.000002238025 s 1.00
add_one / DefOpt / tpu / PostRev 0.0000022501 s 0.0000022386 s 1.01
add_one / DefOpt / tpu / BothRev 0.00000224495 s 0.00000224895 s 1.00
add_one / IDefOpt / tpu / PreRev 0.000002235725 s 0.000002235325 s 1.00
add_one / IDefOpt / tpu / PostRev 0.000002238025 s 0.00000223955 s 1.00
add_one / IDefOpt / tpu / BothRev 0.00000224415 s 0.00000223275 s 1.01
add_one / JaXPipe / cpu / Primal 0.000013332 s 0.000006855099973108736 s 1.94
add_one / Jax / cpu / Primal 0.000013296 s 0.000006435440000132076 s 2.07
add_one / HLOOpt / cpu / Primal 0.000013448 s 0.00000694028000907565 s 1.94
add_one / PartOpt / cpu / Primal 0.000013683 s 0.00000638672001514351 s 2.14
add_one / IPartOpt / cpu / Primal 0.000013046 s 0.0000071436000052926825 s 1.83
add_one / DefOpt / cpu / Primal 0.000013267 s 0.00000677024003380211 s 1.96
add_one / IDefOpt / cpu / Primal 0.00001306 s 0.000006581240022569545 s 1.98
add_one / JaXPipe / cpu / Forward 0.000018707 s 0.000010301519969289077 s 1.82
add_one / Jax / cpu / Forward 0.000018045 s 0.000009633440031393548 s 1.87
add_one / HLOOpt / cpu / Forward 0.00001745 s 0.000010201359973507351 s 1.71
add_one / PartOpt / cpu / Forward 0.000018010000000000002 s 0.000009867159988061758 s 1.83
add_one / IPartOpt / cpu / Forward 0.000017523 s 0.000010082579992740647 s 1.74
add_one / DefOpt / cpu / Forward 0.000017706 s 0.00000990603997706785 s 1.79
add_one / IDefOpt / cpu / Forward 0.000017534999999999997 s 0.000009821160019782838 s 1.79
add_one / JaXPipe / cpu / PreRev 0.00002005 s 0.000012167639997642256 s 1.65
add_one / JaXPipe / cpu / PostRev 0.000019935 s 0.000011753200051316523 s 1.70
add_one / JaXPipe / cpu / BothRev 0.000019763 s 0.000012313679999351734 s 1.60
add_one / Jax / cpu / BothRev 0.000019495 s 0.000011284179990980192 s 1.73
add_one / HLOOpt / cpu / PreRev 0.000019429 s 0.000012106219974157285 s 1.60
add_one / HLOOpt / cpu / PostRev 0.000020223000000000003 s 0.000013971079943075892 s 1.45
add_one / HLOOpt / cpu / BothRev 0.000020688000000000003 s 0.000011830500034193392 s 1.75
add_one / PartOpt / cpu / PreRev 0.00001972 s 0.000012122320003982168 s 1.63
add_one / PartOpt / cpu / PostRev 0.000020557 s 0.000012044100012644776 s 1.71
add_one / PartOpt / cpu / BothRev 0.000019706 s 0.00001201416000185418 s 1.64
add_one / IPartOpt / cpu / PreRev 0.000020214 s 0.00001151484000729397 s 1.76
add_one / IPartOpt / cpu / PostRev 0.000020318 s 0.000011937199988096836 s 1.70
add_one / IPartOpt / cpu / BothRev 0.000019895 s 0.00001157921998128586 s 1.72
add_one / DefOpt / cpu / PreRev 0.000020177 s 0.00001229476002663432 s 1.64
add_one / DefOpt / cpu / PostRev 0.000020332 s 0.000011425299999245907 s 1.78
add_one / DefOpt / cpu / BothRev 0.000019751 s 0.00001220008000927919 s 1.62
add_one / IDefOpt / cpu / PreRev 0.0000202 s 0.000011894500012203937 s 1.70
add_one / IDefOpt / cpu / PostRev 0.000020349 s 0.000011855460024889908 s 1.72
add_one / IDefOpt / cpu / BothRev 0.000020174 s 0.000011839600065286504 s 1.70
add_one / JaXPipe / cpu / Primal 0.000008 s 0.000006855099973108736 s 1.17
add_one / Jax / cpu / Primal 0.000008999999999999999 s 0.000006435440000132076 s 1.40
add_one / HLOOpt / cpu / Primal 0.000008 s 0.00000694028000907565 s 1.15
add_one / PartOpt / cpu / Primal 0.000008999999999999999 s 0.00000638672001514351 s 1.41
add_one / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.0000071436000052926825 s 1.26
add_one / DefOpt / cpu / Primal 0.000008999999999999999 s 0.00000677024003380211 s 1.33
add_one / IDefOpt / cpu / Primal 0.000008 s 0.000006581240022569545 s 1.22
add_one / JaXPipe / cpu / Forward 0.000011 s 0.000010301519969289077 s 1.07
add_one / Jax / cpu / Forward 0.000012 s 0.000009633440031393548 s 1.25
add_one / HLOOpt / cpu / Forward 0.000012 s 0.000010201359973507351 s 1.18
add_one / PartOpt / cpu / Forward 0.000012 s 0.000009867159988061758 s 1.22
add_one / IPartOpt / cpu / Forward 0.000012 s 0.000010082579992740647 s 1.19
add_one / DefOpt / cpu / Forward 0.000012 s 0.00000990603997706785 s 1.21
add_one / IDefOpt / cpu / Forward 0.000012 s 0.000009821160019782838 s 1.22
add_one / JaXPipe / cpu / PreRev 0.000014 s 0.000012167639997642256 s 1.15
add_one / JaXPipe / cpu / PostRev 0.000039 s 0.000011753200051316523 s 3.32
add_one / JaXPipe / cpu / BothRev 0.000014 s 0.000012313679999351734 s 1.14
add_one / Jax / cpu / BothRev 0.000014 s 0.000011284179990980192 s 1.24
add_one / HLOOpt / cpu / PreRev 0.000013 s 0.000012106219974157285 s 1.07
add_one / HLOOpt / cpu / PostRev 0.000013 s 0.000013971079943075892 s 0.93
add_one / HLOOpt / cpu / BothRev 0.000014 s 0.000011830500034193392 s 1.18
add_one / PartOpt / cpu / PreRev 0.000014 s 0.000012122320003982168 s 1.15
add_one / PartOpt / cpu / PostRev 0.000013 s 0.000012044100012644776 s 1.08
add_one / PartOpt / cpu / BothRev 0.000014 s 0.00001201416000185418 s 1.17
add_one / IPartOpt / cpu / PreRev 0.000013 s 0.00001151484000729397 s 1.13
add_one / IPartOpt / cpu / PostRev 0.000013 s 0.000011937199988096836 s 1.09
add_one / IPartOpt / cpu / BothRev 0.000013 s 0.00001157921998128586 s 1.12
add_one / DefOpt / cpu / PreRev 0.000014 s 0.00001229476002663432 s 1.14
add_one / DefOpt / cpu / PostRev 0.000014 s 0.000011425299999245907 s 1.23
add_one / DefOpt / cpu / BothRev 0.000014 s 0.00001220008000927919 s 1.15
add_one / IDefOpt / cpu / PreRev 0.000013 s 0.000011894500012203937 s 1.09
add_one / IDefOpt / cpu / PostRev 0.000014 s 0.000011855460024889908 s 1.18
add_one / IDefOpt / cpu / BothRev 0.000013 s 0.000011839600065286504 s 1.10
add_two / JaXPipe / cpu / Primal 0.000006807359968661331 s 0.000007165639999584528 s 0.95
add_two / Jax / cpu / Primal 0.000006671760011158767 s 0.00000704464002410532 s 0.95
add_two / HLOOpt / cpu / Primal 0.000007710379977652337 s 0.000006832419985585147 s 1.13
add_two / PartOpt / cpu / Primal 0.000006900080015839194 s 0.000006856040008642594 s 1.01
add_two / IPartOpt / cpu / Primal 0.000007284399962372845 s 0.000006973339968681103 s 1.04
add_two / DefOpt / cpu / Primal 0.000006812940009695012 s 0.000007349379984589177 s 0.93
add_two / IDefOpt / cpu / Primal 0.000007005940024100709 s 0.000007407479979519849 s 0.95
add_two / JaXPipe / cpu / Forward 0.000010401239969723974 s 0.000010194559954470606 s 1.02
add_two / Jax / cpu / Forward 0.0000103138199847308 s 0.000010586979988147504 s 0.97
add_two / HLOOpt / cpu / Forward 0.000010637659979693126 s 0.000010633860038069542 s 1.00
add_two / PartOpt / cpu / Forward 0.000010222699984296924 s 0.000010492540004634066 s 0.97
add_two / IPartOpt / cpu / Forward 0.000010368400025981828 s 0.000010203220008406787 s 1.02
add_two / DefOpt / cpu / Forward 0.000010171299982175696 s 0.000010392699996373269 s 0.98
add_two / IDefOpt / cpu / Forward 0.00001012417998026649 s 0.000010373139975854428 s 0.98
add_two / JaXPipe / cpu / PreRev 0.000014018559968462795 s 0.000014746380002179648 s 0.95
add_two / JaXPipe / cpu / PostRev 0.000013910560019212426 s 0.00001389832003042102 s 1.00
add_two / JaXPipe / cpu / BothRev 0.000013881140057492305 s 0.000014604879988837635 s 0.95
add_two / Jax / cpu / BothRev 0.00001364481999189593 s 0.00001416313998561236 s 0.96
add_two / HLOOpt / cpu / PreRev 0.000013855959987267853 s 0.000014704379982504178 s 0.94
add_two / HLOOpt / cpu / PostRev 0.000015312019977500314 s 0.000016084700037026777 s 0.95
add_two / HLOOpt / cpu / BothRev 0.00001356517997010087 s 0.000014112899989413564 s 0.96
add_two / PartOpt / cpu / PreRev 0.000014016080012879685 s 0.000014245980009945924 s 0.98
add_two / PartOpt / cpu / PostRev 0.000014083840014791348 s 0.000014353100023072329 s 0.98
add_two / PartOpt / cpu / BothRev 0.000013243140101621976 s 0.000015038659985293634 s 0.88
add_two / IPartOpt / cpu / PreRev 0.000013818679935866384 s 0.00001430959999197512 s 0.97
add_two / IPartOpt / cpu / PostRev 0.00001358825998067914 s 0.000014052940023248085 s 0.97
add_two / IPartOpt / cpu / BothRev 0.00001364914000077988 s 0.000014151780014799442 s 0.96
add_two / DefOpt / cpu / PreRev 0.000014626740039602735 s 0.000014432859998123604 s 1.01
add_two / DefOpt / cpu / PostRev 0.000013308620027601136 s 0.000014115400017544745 s 0.94
add_two / DefOpt / cpu / BothRev 0.00001347944000372081 s 0.000014274999957706311 s 0.94
add_two / IDefOpt / cpu / PreRev 0.000014210699964678497 s 0.000014062919981370214 s 1.01
add_two / IDefOpt / cpu / PostRev 0.00001362825999422057 s 0.000014034200003152364 s 0.97
add_two / IDefOpt / cpu / BothRev 0.000013545299989345947 s 0.000017536460027258728 s 0.77
add_two / JaXPipe / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
add_two / Jax / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
add_two / HLOOpt / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
add_two / PartOpt / cuda / Primal 0.000001919 s 0.0000024 s 0.80
add_two / IPartOpt / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
add_two / DefOpt / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
add_two / IDefOpt / cuda / Primal 0.000001919 s 0.0000024 s 0.80
add_two / JaXPipe / cuda / Forward 0.000009919 s 0.00001056 s 0.94
add_two / Jax / cuda / Forward 0.000009599 s 0.000010431 s 0.92
add_two / HLOOpt / cuda / Forward 0.000009599 s 0.000008255 s 1.16
add_two / PartOpt / cuda / Forward 0.000010016 s 0.000010464 s 0.96
add_two / IPartOpt / cuda / Forward 0.000009695 s 0.000010336 s 0.94
add_two / DefOpt / cuda / Forward 0.000009312000000000002 s 0.0000104 s 0.90
add_two / IDefOpt / cuda / Forward 0.0000096 s 0.000010208 s 0.94
add_two / JaXPipe / cuda / PreRev 0.000032096 s 0.000032704 s 0.98
add_two / JaXPipe / cuda / PostRev 0.000031264 s 0.000032801 s 0.95
add_two / JaXPipe / cuda / BothRev 0.000032096 s 0.000032769 s 0.98
add_two / Jax / cuda / BothRev 0.00003264 s 0.000033759999999999995 s 0.97
add_two / HLOOpt / cuda / PreRev 0.000031808000000000004 s 0.000033759999999999995 s 0.94
add_two / HLOOpt / cuda / PostRev 0.000032606999999999995 s 0.00003392 s 0.96
add_two / HLOOpt / cuda / BothRev 0.000031776 s 0.00003328 s 0.95
add_two / PartOpt / cuda / PreRev 0.000032383000000000005 s 0.00003408 s 0.95
add_two / PartOpt / cuda / PostRev 0.000032159 s 0.000033759999999999995 s 0.95
add_two / PartOpt / cuda / BothRev 0.000032671 s 0.000033951 s 0.96
add_two / IPartOpt / cuda / PreRev 0.0000312 s 0.000033792000000000004 s 0.92
add_two / IPartOpt / cuda / PostRev 0.000031744 s 0.000033408 s 0.95
add_two / IPartOpt / cuda / BothRev 0.000048671 s 0.000032672 s 1.49
add_two / DefOpt / cuda / PreRev 0.000031905 s 0.000033408 s 0.96
add_two / DefOpt / cuda / PostRev 0.000031424 s 0.000033728 s 0.93
add_two / DefOpt / cuda / BothRev 0.000031136 s 0.00003328 s 0.94
add_two / IDefOpt / cuda / PreRev 0.000032800000000000004 s 0.00003424 s 0.96
add_two / IDefOpt / cuda / PostRev 0.00003248 s 0.000033248 s 0.98
add_two / IDefOpt / cuda / BothRev 0.000031744 s 0.0000336 s 0.94
add_two / JaXPipe / tpu / Primal 0.00000144255 s 0.00000142855 s 1.01
add_two / Jax / tpu / Primal 0.00000148255 s 0.000001474375 s 1.01
add_two / HLOOpt / tpu / Primal 0.000001438625 s 0.0000014344000000000002 s 1.00
add_two / PartOpt / tpu / Primal 0.000001470425 s 0.000001471725 s 1.00
add_two / IPartOpt / tpu / Primal 0.0000014388 s 0.000001433875 s 1.00
add_two / DefOpt / tpu / Primal 0.0000014771999999999998 s 0.00000147285 s 1.00
add_two / IDefOpt / tpu / Primal 0.000001431675 s 0.000001431725 s 1.00
add_two / JaXPipe / tpu / Forward 0.000001823475 s 0.0000018264 s 1.00
add_two / Jax / tpu / Forward 0.000001826375 s 0.000001841275 s 0.99
add_two / HLOOpt / tpu / Forward 0.000001834 s 0.000001844525 s 0.99
add_two / PartOpt / tpu / Forward 0.0000018313 s 0.000001829775 s 1.00
add_two / IPartOpt / tpu / Forward 0.000001827925 s 0.0000018327 s 1.00
add_two / DefOpt / tpu / Forward 0.000001826475 s 0.000001821925 s 1.00
add_two / IDefOpt / tpu / Forward 0.00000183435 s 0.000001828125 s 1.00
add_two / JaXPipe / tpu / PreRev 0.0000028381 s 0.000002833125 s 1.00
add_two / JaXPipe / tpu / PostRev 0.0000027452 s 0.000002747575 s 1.00
add_two / JaXPipe / tpu / BothRev 0.000002834925 s 0.000002840275 s 1.00
add_two / Jax / tpu / BothRev 0.000002750275 s 0.00000275465 s 1.00
add_two / HLOOpt / tpu / PreRev 0.0000028365 s 0.0000028352 s 1.00
add_two / HLOOpt / tpu / PostRev 0.0000027511750000000004 s 0.0000027451750000000004 s 1.00
add_two / HLOOpt / tpu / BothRev 0.0000028307 s 0.000002824175 s 1.00
add_two / PartOpt / tpu / PreRev 0.000002753975 s 0.000002747775 s 1.00
add_two / PartOpt / tpu / PostRev 0.0000028371000000000003 s 0.000002833 s 1.00
add_two / PartOpt / tpu / BothRev 0.00000275775 s 0.0000027551750000000003 s 1.00
add_two / IPartOpt / tpu / PreRev 0.000002837725 s 0.000002833675 s 1.00
add_two / IPartOpt / tpu / PostRev 0.000002751575 s 0.000002750325 s 1.00
add_two / IPartOpt / tpu / BothRev 0.00000283235 s 0.0000028393000000000005 s 1.00
add_two / DefOpt / tpu / PreRev 0.00000274515 s 0.0000027476 s 1.00
add_two / DefOpt / tpu / PostRev 0.000002826825 s 0.000002846125 s 0.99
add_two / DefOpt / tpu / BothRev 0.0000027563 s 0.00000274815 s 1.00
add_two / IDefOpt / tpu / PreRev 0.0000028370250000000006 s 0.000002833225 s 1.00
add_two / IDefOpt / tpu / PostRev 0.0000027521 s 0.000002747675 s 1.00
add_two / IDefOpt / tpu / BothRev 0.000002839325 s 0.0000028406 s 1.00
add_two / JaXPipe / cpu / Primal 0.000013749 s 0.000007165639999584528 s 1.92
add_two / Jax / cpu / Primal 0.000013453 s 0.00000704464002410532 s 1.91
add_two / HLOOpt / cpu / Primal 0.000014421 s 0.000006832419985585147 s 2.11
add_two / PartOpt / cpu / Primal 0.000013617 s 0.000006856040008642594 s 1.99
add_two / IPartOpt / cpu / Primal 0.000014362 s 0.000006973339968681103 s 2.06
add_two / DefOpt / cpu / Primal 0.000013298 s 0.000007349379984589177 s 1.81
add_two / IDefOpt / cpu / Primal 0.000013875 s 0.000007407479979519849 s 1.87
add_two / JaXPipe / cpu / Forward 0.000018905000000000003 s 0.000010194559954470606 s 1.85
add_two / Jax / cpu / Forward 0.000018651 s 0.000010586979988147504 s 1.76
add_two / HLOOpt / cpu / Forward 0.000018991 s 0.000010633860038069542 s 1.79
add_two / PartOpt / cpu / Forward 0.000018585 s 0.000010492540004634066 s 1.77
add_two / IPartOpt / cpu / Forward 0.00001864 s 0.000010203220008406787 s 1.83
add_two / DefOpt / cpu / Forward 0.0000192 s 0.000010392699996373269 s 1.85
add_two / IDefOpt / cpu / Forward 0.00001807 s 0.000010373139975854428 s 1.74
add_two / JaXPipe / cpu / PreRev 0.000023571 s 0.000014746380002179648 s 1.60
add_two / JaXPipe / cpu / PostRev 0.000023483 s 0.00001389832003042102 s 1.69
add_two / JaXPipe / cpu / BothRev 0.000023564 s 0.000014604879988837635 s 1.61
add_two / Jax / cpu / BothRev 0.000023897 s 0.00001416313998561236 s 1.69
add_two / HLOOpt / cpu / PreRev 0.000023573 s 0.000014704379982504178 s 1.60
add_two / HLOOpt / cpu / PostRev 0.00002403 s 0.000016084700037026777 s 1.49
add_two / HLOOpt / cpu / BothRev 0.00002379 s 0.000014112899989413564 s 1.69
add_two / PartOpt / cpu / PreRev 0.000023103 s 0.000014245980009945924 s 1.62
add_two / PartOpt / cpu / PostRev 0.000023781 s 0.000014353100023072329 s 1.66
add_two / PartOpt / cpu / BothRev 0.000023207 s 0.000015038659985293634 s 1.54
add_two / IPartOpt / cpu / PreRev 0.000024585 s 0.00001430959999197512 s 1.72
add_two / IPartOpt / cpu / PostRev 0.000024643 s 0.000014052940023248085 s 1.75
add_two / IPartOpt / cpu / BothRev 0.000023869 s 0.000014151780014799442 s 1.69
add_two / DefOpt / cpu / PreRev 0.000023026000000000003 s 0.000014432859998123604 s 1.60
add_two / DefOpt / cpu / PostRev 0.000023451 s 0.000014115400017544745 s 1.66
add_two / DefOpt / cpu / BothRev 0.000023122 s 0.000014274999957706311 s 1.62
add_two / IDefOpt / cpu / PreRev 0.000022929 s 0.000014062919981370214 s 1.63
add_two / IDefOpt / cpu / PostRev 0.000023332 s 0.000014034200003152364 s 1.66
add_two / IDefOpt / cpu / BothRev 0.000023273 s 0.000017536460027258728 s 1.33
add_two / JaXPipe / cpu / Primal 0.000008 s 0.000007165639999584528 s 1.12
add_two / Jax / cpu / Primal 0.000008999999999999999 s 0.00000704464002410532 s 1.28
add_two / HLOOpt / cpu / Primal 0.000008 s 0.000006832419985585147 s 1.17
add_two / PartOpt / cpu / Primal 0.000008 s 0.000006856040008642594 s 1.17
add_two / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000006973339968681103 s 1.29
add_two / DefOpt / cpu / Primal 0.000008 s 0.000007349379984589177 s 1.09
add_two / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000007407479979519849 s 1.21
add_two / JaXPipe / cpu / Forward 0.000012 s 0.000010194559954470606 s 1.18
add_two / Jax / cpu / Forward 0.000011 s 0.000010586979988147504 s 1.04
add_two / HLOOpt / cpu / Forward 0.000011 s 0.000010633860038069542 s 1.03
add_two / PartOpt / cpu / Forward 0.000012 s 0.000010492540004634066 s 1.14
add_two / IPartOpt / cpu / Forward 0.000011 s 0.000010203220008406787 s 1.08
add_two / DefOpt / cpu / Forward 0.000011 s 0.000010392699996373269 s 1.06
add_two / IDefOpt / cpu / Forward 0.000012 s 0.000010373139975854428 s 1.16
add_two / JaXPipe / cpu / PreRev 0.000016 s 0.000014746380002179648 s 1.09
add_two / JaXPipe / cpu / PostRev 0.000016 s 0.00001389832003042102 s 1.15
add_two / JaXPipe / cpu / BothRev 0.000015 s 0.000014604879988837635 s 1.03
add_two / Jax / cpu / BothRev 0.000015 s 0.00001416313998561236 s 1.06
add_two / HLOOpt / cpu / PreRev 0.000016 s 0.000014704379982504178 s 1.09
add_two / HLOOpt / cpu / PostRev 0.000016 s 0.000016084700037026777 s 0.99
add_two / HLOOpt / cpu / BothRev 0.000016 s 0.000014112899989413564 s 1.13
add_two / PartOpt / cpu / PreRev 0.000016 s 0.000014245980009945924 s 1.12
add_two / PartOpt / cpu / PostRev 0.000016 s 0.000014353100023072329 s 1.11
add_two / PartOpt / cpu / BothRev 0.000016 s 0.000015038659985293634 s 1.06
add_two / IPartOpt / cpu / PreRev 0.000015 s 0.00001430959999197512 s 1.05
add_two / IPartOpt / cpu / PostRev 0.000016 s 0.000014052940023248085 s 1.14
add_two / IPartOpt / cpu / BothRev 0.000016 s 0.000014151780014799442 s 1.13
add_two / DefOpt / cpu / PreRev 0.000015 s 0.000014432859998123604 s 1.04
add_two / DefOpt / cpu / PostRev 0.000016 s 0.000014115400017544745 s 1.13
add_two / DefOpt / cpu / BothRev 0.000015 s 0.000014274999957706311 s 1.05
add_two / IDefOpt / cpu / PreRev 0.000015 s 0.000014062919981370214 s 1.07
add_two / IDefOpt / cpu / PostRev 0.000016 s 0.000014034200003152364 s 1.14
add_two / IDefOpt / cpu / BothRev 0.000015 s 0.000017536460027258728 s 0.86
cache / JaXPipe / cpu / Primal 0.000006157619991427055 s 0.000006194919978952384 s 0.99
cache / Jax / cpu / Primal 0.0000064763200134621 s 0.0000061645400364795934 s 1.05
cache / HLOOpt / cpu / Primal 0.000006083940043026815 s 0.000006119599993326119 s 0.99
cache / PartOpt / cpu / Primal 0.00000625949999630393 s 0.000006567199980054284 s 0.95
cache / IPartOpt / cpu / Primal 0.000006977979992370819 s 0.000006136799966043327 s 1.14
cache / DefOpt / cpu / Primal 0.000006520900005853036 s 0.000006693300010738313 s 0.97
cache / IDefOpt / cpu / Primal 0.000006280239986153901 s 0.0000060444800419645615 s 1.04
cache / JaXPipe / cpu / Forward 0.000014433740034291986 s 0.00001394723999510461 s 1.03
cache / Jax / cpu / Forward 0.000015279179979188485 s 0.00001463153999793576 s 1.04
cache / HLOOpt / cpu / Forward 0.000014856439966024482 s 0.0000155708400234289 s 0.95
cache / PartOpt / cpu / Forward 0.00001437333999092516 s 0.000014294260026872508 s 1.01
cache / IPartOpt / cpu / Forward 0.000015829019976081326 s 0.00001517994001005718 s 1.04
cache / DefOpt / cpu / Forward 0.000013944619922767744 s 0.000014456180006163777 s 0.96
cache / IDefOpt / cpu / Forward 0.000014098579995334147 s 0.000014219840040823327 s 0.99
cache / JaXPipe / cpu / PreRev 0.000015993319993867772 s 0.00001582649999363639 s 1.01
cache / JaXPipe / cpu / PostRev 0.000020504459980656977 s 0.000019728639990717056 s 1.04
cache / JaXPipe / cpu / BothRev 0.000015874979944783262 s 0.000016025759978219868 s 0.99
cache / Jax / cpu / BothRev 0.00001981506002266542 s 0.00002008118000958348 s 0.99
cache / HLOOpt / cpu / PreRev 0.000016655719955451788 s 0.000016810760025691708 s 0.99
cache / HLOOpt / cpu / PostRev 0.0000187399999867921 s 0.000017934899988176765 s 1.04
cache / HLOOpt / cpu / BothRev 0.000015330319984059314 s 0.000015689379970353912 s 0.98
cache / PartOpt / cpu / PreRev 0.000015098059957381338 s 0.00001536010001473187 s 0.98
cache / PartOpt / cpu / PostRev 0.00001992513994991896 s 0.000019944439936807613 s 1.00
cache / PartOpt / cpu / BothRev 0.000015282839985957252 s 0.000016913900026338524 s 0.90
cache / IPartOpt / cpu / PreRev 0.000015822860004846005 s 0.000015653360014766805 s 1.01
cache / IPartOpt / cpu / PostRev 0.0000198083000304905 s 0.000021137560015631606 s 0.94
cache / IPartOpt / cpu / BothRev 0.000015077859980010543 s 0.000015237540010275552 s 0.99
cache / DefOpt / cpu / PreRev 0.000014576699986719178 s 0.000016161280009328038 s 0.90
cache / DefOpt / cpu / PostRev 0.000015034180023576482 s 0.00001578093999341945 s 0.95
cache / DefOpt / cpu / BothRev 0.000014591819990528164 s 0.000015761919976284843 s 0.93
cache / IDefOpt / cpu / PreRev 0.000015308119964174692 s 0.000016224380005951388 s 0.94
cache / IDefOpt / cpu / PostRev 0.000015102840025065236 s 0.00001620706002540828 s 0.93
cache / IDefOpt / cpu / BothRev 0.000014755299989701598 s 0.000016699499974492936 s 0.88
cache / JaXPipe / cuda / Primal 0.000002303 s 0.000002335 s 0.99
cache / Jax / cuda / Primal 0.000002303 s 0.000002336 s 0.99
cache / HLOOpt / cuda / Primal 0.000002208 s 0.000002336 s 0.95
cache / PartOpt / cuda / Primal 0.00000224 s 0.000002335 s 0.96
cache / IPartOpt / cuda / Primal 0.000002303 s 0.000002335 s 0.99
cache / DefOpt / cuda / Primal 0.00000224 s 0.000002336 s 0.96
cache / IDefOpt / cuda / Primal 0.00000224 s 0.000002335 s 0.96
cache / JaXPipe / cuda / Forward 0.000002335 s 0.0000023670000000000004 s 0.99
cache / Jax / cuda / Forward 0.000002335 s 0.0000023670000000000004 s 0.99
cache / HLOOpt / cuda / Forward 0.000002335 s 0.000002335 s 1
cache / PartOpt / cuda / Forward 0.000002335 s 0.0000023670000000000004 s 0.99
cache / IPartOpt / cuda / Forward 0.000002335 s 0.000002335 s 1
cache / DefOpt / cuda / Forward 0.000002304 s 0.0000023670000000000004 s 0.97
cache / IDefOpt / cuda / Forward 0.000002335 s 0.000002336 s 1.00
cache / JaXPipe / cuda / PreRev 0.000010463 s 0.000010752 s 0.97
cache / JaXPipe / cuda / PostRev 0.00001056 s 0.000010591 s 1.00
cache / JaXPipe / cuda / BothRev 0.000010688 s 0.000010848 s 0.99
cache / Jax / cuda / BothRev 0.000010528 s 0.00001088 s 0.97
cache / HLOOpt / cuda / PreRev 0.000013504 s 0.000013696 s 0.99
cache / HLOOpt / cuda / PostRev 0.000013503 s 0.000013663 s 0.99
cache / HLOOpt / cuda / BothRev 0.000013504 s 0.000013664 s 0.99
cache / PartOpt / cuda / PreRev 0.000010847 s 0.000010656 s 1.02
cache / PartOpt / cuda / PostRev 0.000010912 s 0.000010464 s 1.04
cache / PartOpt / cuda / BothRev 0.00001072 s 0.000010688 s 1.00
cache / IPartOpt / cuda / PreRev 0.000010944 s 0.000011008 s 0.99
cache / IPartOpt / cuda / PostRev 0.000010752 s 0.000010848 s 0.99
cache / IPartOpt / cuda / BothRev 0.00001088 s 0.000010464 s 1.04
cache / DefOpt / cuda / PreRev 0.000010944 s 0.000011233 s 0.97
cache / DefOpt / cuda / PostRev 0.000010943 s 0.000010752 s 1.02
cache / DefOpt / cuda / BothRev 0.000010688 s 0.000010784 s 0.99
cache / IDefOpt / cuda / PreRev 0.000010336 s 0.000010944 s 0.94
cache / IDefOpt / cuda / PostRev 0.000010592 s 0.000010688 s 0.99
cache / IDefOpt / cuda / BothRev 0.000010719 s 0.000011296 s 0.95
cache / JaXPipe / tpu / Primal 0.000002468725 s 0.000002473675 s 1.00
cache / Jax / tpu / Primal 0.0000024691 s 0.000002475675 s 1.00
cache / HLOOpt / tpu / Primal 0.0000024706 s 0.0000024941250000000005 s 0.99
cache / PartOpt / tpu / Primal 0.000002463875 s 0.000002472675 s 1.00
cache / IPartOpt / tpu / Primal 0.00000245065 s 0.000002465625 s 0.99
cache / DefOpt / tpu / Primal 0.000002469975 s 0.0000024604750000000004 s 1.00
cache / IDefOpt / tpu / Primal 0.000002471975 s 0.0000024796000000000004 s 1.00
cache / JaXPipe / tpu / Forward 0.0000035627750000000004 s 0.000003569925 s 1.00
cache / Jax / tpu / Forward 0.000003543725 s 0.0000035356 s 1.00
cache / HLOOpt / tpu / Forward 0.000003556125 s 0.00000356165 s 1.00
cache / PartOpt / tpu / Forward 0.000003542625 s 0.000003522025 s 1.01
cache / IPartOpt / tpu / Forward 0.0000035458 s 0.0000035693500000000004 s 0.99
cache / DefOpt / tpu / Forward 0.00000353615 s 0.00000354475 s 1.00
cache / IDefOpt / tpu / Forward 0.0000035519 s 0.000003559 s 1.00
cache / JaXPipe / tpu / PreRev 0.000005011175 s 0.0000049973 s 1.00
cache / JaXPipe / tpu / PostRev 0.00000501085 s 0.0000049926 s 1.00
cache / JaXPipe / tpu / BothRev 0.000005037275 s 0.000005030324999999999 s 1.00
cache / Jax / tpu / BothRev 0.000005008125 s 0.000005001025 s 1.00
cache / HLOOpt / tpu / PreRev 0.0000039651750000000005 s 0.0000039548 s 1.00
cache / HLOOpt / tpu / PostRev 0.000004153625 s 0.000004139550000000001 s 1.00
cache / HLOOpt / tpu / BothRev 0.000003967 s 0.000003968524999999999 s 1.00
cache / PartOpt / tpu / PreRev 0.000005022299999999999 s 0.000005001125 s 1.00
cache / PartOpt / tpu / PostRev 0.0000050204 s 0.000005011575 s 1.00
cache / PartOpt / tpu / BothRev 0.00000502625 s 0.00000500935 s 1.00
cache / IPartOpt / tpu / PreRev 0.000005022675 s 0.000005030675 s 1.00
cache / IPartOpt / tpu / PostRev 0.000005018774999999999 s 0.000005022799999999999 s 1.00
cache / IPartOpt / tpu / BothRev 0.000005025175 s 0.0000050125 s 1.00
cache / DefOpt / tpu / PreRev 0.000005011775000000001 s 0.000005018225 s 1.00
cache / DefOpt / tpu / PostRev 0.000005015625 s 0.0000050143 s 1.00
cache / DefOpt / tpu / BothRev 0.000005007899999999999 s 0.0000050327 s 1.00
cache / IDefOpt / tpu / PreRev 0.00000503265 s 0.00000501565 s 1.00
cache / IDefOpt / tpu / PostRev 0.0000050247 s 0.0000050247 s 1
cache / IDefOpt / tpu / BothRev 0.000005009400000000001 s 0.00000499635 s 1.00
cache / JaXPipe / cpu / Primal 0.000013024 s 0.000006194919978952384 s 2.10
cache / Jax / cpu / Primal 0.000012666 s 0.0000061645400364795934 s 2.05
cache / HLOOpt / cpu / Primal 0.000012733 s 0.000006119599993326119 s 2.08
cache / PartOpt / cpu / Primal 0.000013443 s 0.000006567199980054284 s 2.05
cache / IPartOpt / cpu / Primal 0.00001316 s 0.000006136799966043327 s 2.14
cache / DefOpt / cpu / Primal 0.000013177 s 0.000006693300010738313 s 1.97
cache / IDefOpt / cpu / Primal 0.00001336 s 0.0000060444800419645615 s 2.21
cache / JaXPipe / cpu / Forward 0.000023825 s 0.00001394723999510461 s 1.71
cache / Jax / cpu / Forward 0.000017008 s 0.00001463153999793576 s 1.16
cache / HLOOpt / cpu / Forward 0.000024368 s 0.0000155708400234289 s 1.56
cache / PartOpt / cpu / Forward 0.00002337 s 0.000014294260026872508 s 1.63
cache / IPartOpt / cpu / Forward 0.000023142 s 0.00001517994001005718 s 1.52
cache / DefOpt / cpu / Forward 0.000025784000000000003 s 0.000014456180006163777 s 1.78
cache / IDefOpt / cpu / Forward 0.000024226 s 0.000014219840040823327 s 1.70
cache / JaXPipe / cpu / PreRev 0.000025697 s 0.00001582649999363639 s 1.62
cache / JaXPipe / cpu / PostRev 0.00003002 s 0.000019728639990717056 s 1.52
cache / JaXPipe / cpu / BothRev 0.000029037 s 0.000016025759978219868 s 1.81
cache / Jax / cpu / BothRev 0.000033883 s 0.00002008118000958348 s 1.69
cache / HLOOpt / cpu / PreRev 0.00002394 s 0.000016810760025691708 s 1.42
cache / HLOOpt / cpu / PostRev 0.000025119 s 0.000017934899988176765 s 1.40
cache / HLOOpt / cpu / BothRev 0.000027557 s 0.000015689379970353912 s 1.76
cache / PartOpt / cpu / PreRev 0.000028999 s 0.00001536010001473187 s 1.89
cache / PartOpt / cpu / PostRev 0.000019937 s 0.000019944439936807613 s 1.00
cache / PartOpt / cpu / BothRev 0.00002411 s 0.000016913900026338524 s 1.43
cache / IPartOpt / cpu / PreRev 0.000018099 s 0.000015653360014766805 s 1.16
cache / IPartOpt / cpu / PostRev 0.000020192000000000003 s 0.000021137560015631606 s 0.96
cache / IPartOpt / cpu / BothRev 0.000018007 s 0.000015237540010275552 s 1.18
cache / DefOpt / cpu / PreRev 0.000017760999999999998 s 0.000016161280009328038 s 1.10
cache / DefOpt / cpu / PostRev 0.000017464 s 0.00001578093999341945 s 1.11
cache / DefOpt / cpu / BothRev 0.00001745 s 0.000015761919976284843 s 1.11
cache / IDefOpt / cpu / PreRev 0.000017462 s 0.000016224380005951388 s 1.08
cache / IDefOpt / cpu / PostRev 0.000017590000000000003 s 0.00001620706002540828 s 1.09
cache / IDefOpt / cpu / BothRev 0.000024514 s 0.000016699499974492936 s 1.47
cache / JaXPipe / cpu / Primal 0.000008 s 0.000006194919978952384 s 1.29
cache / Jax / cpu / Primal 0.000008 s 0.0000061645400364795934 s 1.30
cache / HLOOpt / cpu / Primal 0.000008 s 0.000006119599993326119 s 1.31
cache / PartOpt / cpu / Primal 0.000008 s 0.000006567199980054284 s 1.22
cache / IPartOpt / cpu / Primal 0.000008 s 0.000006136799966043327 s 1.30
cache / DefOpt / cpu / Primal 0.000008 s 0.000006693300010738313 s 1.20
cache / IDefOpt / cpu / Primal 0.000008 s 0.0000060444800419645615 s 1.32
cache / JaXPipe / cpu / Forward 0.00001 s 0.00001394723999510461 s 0.72
cache / Jax / cpu / Forward 0.00001 s 0.00001463153999793576 s 0.68
cache / HLOOpt / cpu / Forward 0.000011 s 0.0000155708400234289 s 0.71
cache / PartOpt / cpu / Forward 0.00001 s 0.000014294260026872508 s 0.70
cache / IPartOpt / cpu / Forward 0.00001 s 0.00001517994001005718 s 0.66
cache / DefOpt / cpu / Forward 0.00001 s 0.000014456180006163777 s 0.69
cache / IDefOpt / cpu / Forward 0.00001 s 0.000014219840040823327 s 0.70
cache / JaXPipe / cpu / PreRev 0.00001 s 0.00001582649999363639 s 0.63
cache / JaXPipe / cpu / PostRev 0.000011 s 0.000019728639990717056 s 0.56
cache / JaXPipe / cpu / BothRev 0.00001 s 0.000016025759978219868 s 0.62
cache / Jax / cpu / BothRev 0.000011 s 0.00002008118000958348 s 0.55
cache / HLOOpt / cpu / PreRev 0.000011 s 0.000016810760025691708 s 0.65
cache / HLOOpt / cpu / PostRev 0.000011 s 0.000017934899988176765 s 0.61
cache / HLOOpt / cpu / BothRev 0.00001 s 0.000015689379970353912 s 0.64
cache / PartOpt / cpu / PreRev 0.000011 s 0.00001536010001473187 s 0.72
cache / PartOpt / cpu / PostRev 0.000011 s 0.000019944439936807613 s 0.55
cache / PartOpt / cpu / BothRev 0.000011 s 0.000016913900026338524 s 0.65
cache / IPartOpt / cpu / PreRev 0.000012 s 0.000015653360014766805 s 0.77
cache / IPartOpt / cpu / PostRev 0.000011 s 0.000021137560015631606 s 0.52
cache / IPartOpt / cpu / BothRev 0.000011 s 0.000015237540010275552 s 0.72
cache / DefOpt / cpu / PreRev 0.00001 s 0.000016161280009328038 s 0.62
cache / DefOpt / cpu / PostRev 0.00001 s 0.00001578093999341945 s 0.63
cache / DefOpt / cpu / BothRev 0.000035999999999999994 s 0.000015761919976284843 s 2.28
cache / IDefOpt / cpu / PreRev 0.000011 s 0.000016224380005951388 s 0.68
cache / IDefOpt / cpu / PostRev 0.00001 s 0.00001620706002540828 s 0.62
cache / IDefOpt / cpu / BothRev 0.00001 s 0.000016699499974492936 s 0.60
Concat / JaXPipe / cpu / Primal 0.000006711060041197925 s 0.000006462160035880515 s 1.04
Concat / Jax / cpu / Primal 0.000006429699997170246 s 0.000006423419954444398 s 1.00
Concat / HLOOpt / cpu / Primal 0.00000659505999465182 s 0.00000704545998814865 s 0.94
Concat / PartOpt / cpu / Primal 0.000006639019993599504 s 0.000006768500024918467 s 0.98
Concat / IPartOpt / cpu / Primal 0.000007015840001258766 s 0.00000666878001538862 s 1.05
Concat / DefOpt / cpu / Primal 0.000006272099963098299 s 0.000006222180008990108 s 1.01
Concat / IDefOpt / cpu / Primal 0.000006539119995068176 s 0.000006459919986809836 s 1.01
Concat / JaXPipe / cpu / Forward 0.000010166859992750689 s 0.000010299879977537784 s 0.99
Concat / Jax / cpu / Forward 0.000009702299985292484 s 0.000009709199985081797 s 1.00
Concat / HLOOpt / cpu / Forward 0.000009979499991459306 s 0.000010152300010304316 s 0.98
Concat / PartOpt / cpu / Forward 0.00000988301998404495 s 0.000009993960011343006 s 0.99
Concat / IPartOpt / cpu / Forward 0.000010338219972254592 s 0.000010250420000375016 s 1.01
Concat / DefOpt / cpu / Forward 0.000009914519987432868 s 0.000010049580014310776 s 0.99
Concat / IDefOpt / cpu / Forward 0.000009672020032667206 s 0.000009826820014495753 s 0.98
Concat / JaXPipe / cpu / PreRev 0.000010939860012513235 s 0.000011506600021675694 s 0.95
Concat / JaXPipe / cpu / PostRev 0.00001088297999558563 s 0.00001131426003667002 s 0.96
Concat / JaXPipe / cpu / BothRev 0.000011079500027335598 s 0.000011266000028626876 s 0.98
Concat / Jax / cpu / BothRev 0.000011378479975974186 s 0.000011426819974076352 s 1.00
Concat / HLOOpt / cpu / PreRev 0.000011753640037568405 s 0.000012031859987473582 s 0.98
Concat / HLOOpt / cpu / PostRev 0.00001295821998610336 s 0.000013319039990165038 s 0.97
Concat / HLOOpt / cpu / BothRev 0.000011045700002796366 s 0.000011193080008524702 s 0.99
Concat / PartOpt / cpu / PreRev 0.000011119979972136206 s 0.00001135953998527839 s 0.98
Concat / PartOpt / cpu / PostRev 0.000011111179956060367 s 0.000011545259958438692 s 0.96
Concat / PartOpt / cpu / BothRev 0.000011416400020607398 s 0.000012315920021137571 s 0.93
Concat / IPartOpt / cpu / PreRev 0.000011651059976429678 s 0.000011367440010872088 s 1.02
Concat / IPartOpt / cpu / PostRev 0.000011550199978955788 s 0.000012098559991500225 s 0.95
Concat / IPartOpt / cpu / BothRev 0.000011061699997299 s 0.00001150006001807924 s 0.96
Concat / DefOpt / cpu / PreRev 0.000011139560001538484 s 0.000011823339982584002 s 0.94
Concat / DefOpt / cpu / PostRev 0.000011023980032405234 s 0.000011176239977430667 s 0.99
Concat / DefOpt / cpu / BothRev 0.000011146139995616976 s 0.000011328839982525095 s 0.98
Concat / IDefOpt / cpu / PreRev 0.00001120174001698615 s 0.000011458319977464271 s 0.98
Concat / IDefOpt / cpu / PostRev 0.000011077640010626055 s 0.00001092478001737618 s 1.01
Concat / IDefOpt / cpu / BothRev 0.000011162840037286514 s 0.00001205603998641891 s 0.93
Concat / JaXPipe / cuda / Primal 0.000001919 s 0.0000024 s 0.80
Concat / Jax / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
Concat / HLOOpt / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
Concat / PartOpt / cuda / Primal 0.000001919 s 0.0000024 s 0.80
Concat / IPartOpt / cuda / Primal 0.000001919 s 0.0000024 s 0.80
Concat / DefOpt / cuda / Primal 0.000001919 s 0.000002431 s 0.79
Concat / IDefOpt / cuda / Primal 0.0000019200000000000003 s 0.0000024 s 0.80
Concat / JaXPipe / cuda / Forward 0.000009856 s 0.000010688 s 0.92
Concat / Jax / cuda / Forward 0.000009887 s 0.000010688 s 0.93
Concat / HLOOpt / cuda / Forward 0.000009824 s 0.000010176 s 0.97
Concat / PartOpt / cuda / Forward 0.000009792 s 0.000010752 s 0.91
Concat / IPartOpt / cuda / Forward 0.00001008 s 0.00001072 s 0.94
Concat / DefOpt / cuda / Forward 0.000009983 s 0.000010591 s 0.94
Concat / IDefOpt / cuda / Forward 0.000009792 s 0.000010496 s 0.93
Concat / JaXPipe / cuda / PreRev 0.000016255999999999998 s 0.000016864 s 0.96
Concat / JaXPipe / cuda / PostRev 0.000016352 s 0.0000168 s 0.97
Concat / JaXPipe / cuda / BothRev 0.000016255999999999998 s 0.000016736 s 0.97
Concat / Jax / cuda / BothRev 0.000016544 s 0.000016864 s 0.98
Concat / HLOOpt / cuda / PreRev 0.000015968 s 0.000016608 s 0.96
Concat / HLOOpt / cuda / PostRev 0.000016544 s 0.000016609 s 1.00
Concat / HLOOpt / cuda / BothRev 0.000016223 s 0.000016672 s 0.97
Concat / PartOpt / cuda / PreRev 0.00001648 s 0.000016992 s 0.97
Concat / PartOpt / cuda / PostRev 0.000016096 s 0.000016768000000000003 s 0.96
Concat / PartOpt / cuda / BothRev 0.000016352 s 0.000017152 s 0.95
Concat / IPartOpt / cuda / PreRev 0.000016832 s 0.000016704 s 1.01
Concat / IPartOpt / cuda / PostRev 0.000016416 s 0.000016832 s 0.98
Concat / IPartOpt / cuda / BothRev 0.000016703 s 0.000016768000000000003 s 1.00
Concat / DefOpt / cuda / PreRev 0.000016095 s 0.000016832 s 0.96
Concat / DefOpt / cuda / PostRev 0.000016352 s 0.000025377 s 0.64
Concat / DefOpt / cuda / BothRev 0.000015744 s 0.000016993 s 0.93
Concat / IDefOpt / cuda / PreRev 0.000016352 s 0.000016927000000000002 s 0.97
Concat / IDefOpt / cuda / PostRev 0.000016255999999999998 s 0.000016992 s 0.96
Concat / IDefOpt / cuda / BothRev 0.000016223 s 0.000016542999999999997 s 0.98
Concat / JaXPipe / tpu / Primal 0.0000015364749999999998 s 0.000001521575 s 1.01
Concat / Jax / tpu / Primal 0.000001522075 s 0.000001537225 s 0.99
Concat / HLOOpt / tpu / Primal 0.0000015394749999999998 s 0.000001529375 s 1.01
Concat / PartOpt / tpu / Primal 0.0000015211 s 0.0000015313750000000002 s 0.99
Concat / IPartOpt / tpu / Primal 0.00000153515 s 0.00000152085 s 1.01
Concat / DefOpt / tpu / Primal 0.0000015342 s 0.0000015367500000000002 s 1.00
Concat / IDefOpt / tpu / Primal 0.000001542975 s 0.00000153185 s 1.01
Concat / JaXPipe / tpu / Forward 0.000001569475 s 0.000001579675 s 0.99
Concat / Jax / tpu / Forward 0.00000155285 s 0.000001549625 s 1.00
Concat / HLOOpt / tpu / Forward 0.000001569325 s 0.000001567475 s 1.00
Concat / PartOpt / tpu / Forward 0.000001555425 s 0.000001558525 s 1.00
Concat / IPartOpt / tpu / Forward 0.0000015731 s 0.000001584675 s 0.99
Concat / DefOpt / tpu / Forward 0.0000015528 s 0.00000156225 s 0.99
Concat / IDefOpt / tpu / Forward 0.000001569475 s 0.0000015809500000000003 s 0.99
Concat / JaXPipe / tpu / PreRev 0.000002003625 s 0.000001999275 s 1.00
Concat / JaXPipe / tpu / PostRev 0.000002066775 s 0.0000020631500000000003 s 1.00
Concat / JaXPipe / tpu / BothRev 0.0000020024750000000003 s 0.000001994975 s 1.00
Concat / Jax / tpu / BothRev 0.00000206165 s 0.000002063925 s 1.00
Concat / HLOOpt / tpu / PreRev 0.000002000475 s 0.0000020009750000000004 s 1.00
Concat / HLOOpt / tpu / PostRev 0.000002064775 s 0.000002078475 s 0.99
Concat / HLOOpt / tpu / BothRev 0.00000201065 s 0.000001992725 s 1.01
Concat / PartOpt / tpu / PreRev 0.000002070675 s 0.00000206995 s 1.00
Concat / PartOpt / tpu / PostRev 0.000002006075 s 0.000002001825 s 1.00
Concat / PartOpt / tpu / BothRev 0.000002060175 s 0.00000206835 s 1.00
Concat / IPartOpt / tpu / PreRev 0.000002002075 s 0.000001999375 s 1.00
Concat / IPartOpt / tpu / PostRev 0.000002070925 s 0.000002072175 s 1.00
Concat / IPartOpt / tpu / BothRev 0.00000201405 s 0.0000020059750000000003 s 1.00
Concat / DefOpt / tpu / PreRev 0.00000206875 s 0.0000020715 s 1.00
Concat / DefOpt / tpu / PostRev 0.00000201585 s 0.000002000525 s 1.01
Concat / DefOpt / tpu / BothRev 0.000002070225 s 0.00000206345 s 1.00
Concat / IDefOpt / tpu / PreRev 0.000002007125 s 0.0000019996249999999995 s 1.00
Concat / IDefOpt / tpu / PostRev 0.00000206405 s 0.0000020799 s 0.99
Concat / IDefOpt / tpu / BothRev 0.000002015025 s 0.0000019935 s 1.01
Concat / JaXPipe / cpu / Primal 0.000012922 s 0.000006462160035880515 s 2.00
Concat / Jax / cpu / Primal 0.000012977 s 0.000006423419954444398 s 2.02
Concat / HLOOpt / cpu / Primal 0.000013237 s 0.00000704545998814865 s 1.88
Concat / PartOpt / cpu / Primal 0.00001319 s 0.000006768500024918467 s 1.95
Concat / IPartOpt / cpu / Primal 0.000013002 s 0.00000666878001538862 s 1.95
Concat / DefOpt / cpu / Primal 0.000013248 s 0.000006222180008990108 s 2.13
Concat / IDefOpt / cpu / Primal 0.000013046 s 0.000006459919986809836 s 2.02
Concat / JaXPipe / cpu / Forward 0.000018112 s 0.000010299879977537784 s 1.76
Concat / Jax / cpu / Forward 0.000017828 s 0.000009709199985081797 s 1.84
Concat / HLOOpt / cpu / Forward 0.000019174 s 0.000010152300010304316 s 1.89
Concat / PartOpt / cpu / Forward 0.00001747 s 0.000009993960011343006 s 1.75
Concat / IPartOpt / cpu / Forward 0.000018105 s 0.000010250420000375016 s 1.77
Concat / DefOpt / cpu / Forward 0.000018011 s 0.000010049580014310776 s 1.79
Concat / IDefOpt / cpu / Forward 0.000018868 s 0.000009826820014495753 s 1.92
Concat / JaXPipe / cpu / PreRev 0.000020445 s 0.000011506600021675694 s 1.78
Concat / JaXPipe / cpu / PostRev 0.000019842 s 0.00001131426003667002 s 1.75
Concat / JaXPipe / cpu / BothRev 0.000019846 s 0.000011266000028626876 s 1.76
Concat / Jax / cpu / BothRev 0.000020698 s 0.000011426819974076352 s 1.81
Concat / HLOOpt / cpu / PreRev 0.000019976 s 0.000012031859987473582 s 1.66
Concat / HLOOpt / cpu / PostRev 0.000020168 s 0.000013319039990165038 s 1.51
Concat / HLOOpt / cpu / BothRev 0.000019953 s 0.000011193080008524702 s 1.78
Concat / PartOpt / cpu / PreRev 0.000020070000000000003 s 0.00001135953998527839 s 1.77
Concat / PartOpt / cpu / PostRev 0.000020128 s 0.000011545259958438692 s 1.74
Concat / PartOpt / cpu / BothRev 0.000019952 s 0.000012315920021137571 s 1.62
Concat / IPartOpt / cpu / PreRev 0.000019713000000000003 s 0.000011367440010872088 s 1.73
Concat / IPartOpt / cpu / PostRev 0.000019501 s 0.000012098559991500225 s 1.61
Concat / IPartOpt / cpu / BothRev 0.000019581 s 0.00001150006001807924 s 1.70
Concat / DefOpt / cpu / PreRev 0.000019833 s 0.000011823339982584002 s 1.68
Concat / DefOpt / cpu / PostRev 0.000019939 s 0.000011176239977430667 s 1.78
Concat / DefOpt / cpu / BothRev 0.00002023 s 0.000011328839982525095 s 1.79
Concat / IDefOpt / cpu / PreRev 0.000020452 s 0.000011458319977464271 s 1.78
Concat / IDefOpt / cpu / PostRev 0.000020259 s 0.00001092478001737618 s 1.85
Concat / IDefOpt / cpu / BothRev 0.000019542 s 0.00001205603998641891 s 1.62
Concat / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000006462160035880515 s 1.39
Concat / Jax / cpu / Primal 0.000008999999999999999 s 0.000006423419954444398 s 1.40
Concat / HLOOpt / cpu / Primal 0.000008 s 0.00000704545998814865 s 1.14
Concat / PartOpt / cpu / Primal 0.000008 s 0.000006768500024918467 s 1.18
Concat / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.00000666878001538862 s 1.35
Concat / DefOpt / cpu / Primal 0.000008 s 0.000006222180008990108 s 1.29
Concat / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000006459919986809836 s 1.39
Concat / JaXPipe / cpu / Forward 0.000012 s 0.000010299879977537784 s 1.17
Concat / Jax / cpu / Forward 0.000012 s 0.000009709199985081797 s 1.24
Concat / HLOOpt / cpu / Forward 0.000012 s 0.000010152300010304316 s 1.18
Concat / PartOpt / cpu / Forward 0.000012 s 0.000009993960011343006 s 1.20
Concat / IPartOpt / cpu / Forward 0.000012 s 0.000010250420000375016 s 1.17
Concat / DefOpt / cpu / Forward 0.000012 s 0.000010049580014310776 s 1.19
Concat / IDefOpt / cpu / Forward 0.000012 s 0.000009826820014495753 s 1.22
Concat / JaXPipe / cpu / PreRev 0.000014 s 0.000011506600021675694 s 1.22
Concat / JaXPipe / cpu / PostRev 0.000013 s 0.00001131426003667002 s 1.15
Concat / JaXPipe / cpu / BothRev 0.000014 s 0.000011266000028626876 s 1.24
Concat / Jax / cpu / BothRev 0.000014 s 0.000011426819974076352 s 1.23
Concat / HLOOpt / cpu / PreRev 0.000014 s 0.000012031859987473582 s 1.16
Concat / HLOOpt / cpu / PostRev 0.000014 s 0.000013319039990165038 s 1.05
Concat / HLOOpt / cpu / BothRev 0.000014 s 0.000011193080008524702 s 1.25
Concat / PartOpt / cpu / PreRev 0.000014 s 0.00001135953998527839 s 1.23
Concat / PartOpt / cpu / PostRev 0.000014 s 0.000011545259958438692 s 1.21
Concat / PartOpt / cpu / BothRev 0.000014 s 0.000012315920021137571 s 1.14
Concat / IPartOpt / cpu / PreRev 0.000014 s 0.000011367440010872088 s 1.23
Concat / IPartOpt / cpu / PostRev 0.000013 s 0.000012098559991500225 s 1.07
Concat / IPartOpt / cpu / BothRev 0.000014 s 0.00001150006001807924 s 1.22
Concat / DefOpt / cpu / PreRev 0.000013 s 0.000011823339982584002 s 1.10
Concat / DefOpt / cpu / PostRev 0.000014 s 0.000011176239977430667 s 1.25
Concat / DefOpt / cpu / BothRev 0.000014 s 0.000011328839982525095 s 1.24
Concat / IDefOpt / cpu / PreRev 0.000013 s 0.000011458319977464271 s 1.13
Concat / IDefOpt / cpu / PostRev 0.000014 s 0.00001092478001737618 s 1.28
Concat / IDefOpt / cpu / BothRev 0.000014 s 0.00001205603998641891 s 1.16
const_scatter / JaXPipe / cpu / Primal 0.000006356639987643575 s 0.000006957139994483441 s 0.91
const_scatter / Jax / cpu / Primal 0.000006334580020848079 s 0.000006754500018359976 s 0.94
const_scatter / HLOOpt / cpu / Primal 0.000007195740026872954 s 0.000006998379976721481 s 1.03
const_scatter / PartOpt / cpu / Primal 0.000006383039990396356 s 0.000006252760031202343 s 1.02
const_scatter / IPartOpt / cpu / Primal 0.000006285660010689753 s 0.000006638139984715962 s 0.95
const_scatter / DefOpt / cpu / Primal 0.000006777520002287929 s 0.000007144199998947442 s 0.95
const_scatter / IDefOpt / cpu / Primal 0.000006809299984524841 s 0.000007046919999993407 s 0.97
const_scatter / JaXPipe / cpu / Forward 0.000010385459981989696 s 0.000010530240024309025 s 0.99
const_scatter / Jax / cpu / Forward 0.000009477980029259925 s 0.000009437960052309793 s 1.00
const_scatter / HLOOpt / cpu / Forward 0.000010588379973341944 s 0.000010992900033670594 s 0.96
const_scatter / PartOpt / cpu / Forward 0.000010278760000801411 s 0.000010934680003629185 s 0.94
const_scatter / IPartOpt / cpu / Forward 0.000011026880019926468 s 0.00001130093995925563 s 0.98
const_scatter / DefOpt / cpu / Forward 0.000010500659991521388 s 0.00001061286003277928 s 0.99
const_scatter / IDefOpt / cpu / Forward 0.00001045890000568761 s 0.000010548560003371675 s 0.99
const_scatter / JaXPipe / cpu / PreRev 0.0002855424799508 s 0.0002879407399541 s 0.99
const_scatter / JaXPipe / cpu / PostRev 0.0002816413599612 s 0.0002806210999824 s 1.00
const_scatter / JaXPipe / cpu / BothRev 0.0002820143400185 s 0.0002837718799764 s 0.99
const_scatter / Jax / cpu / BothRev 0.000281322419978 s 0.0002842210200014 s 0.99
const_scatter / HLOOpt / cpu / PreRev 0.0002950535199761 s 0.0002866692799489 s 1.03
const_scatter / HLOOpt / cpu / PostRev 0.0002844586400078 s 0.0002865757799918 s 0.99
const_scatter / HLOOpt / cpu / BothRev 0.000281799819968 s 0.00028323980001 s 0.99
const_scatter / PartOpt / cpu / PreRev 0.0002797417599595 s 0.0002816201200221 s 0.99
const_scatter / PartOpt / cpu / PostRev 0.0002778683200449 s 0.0002784744400105 s 1.00
const_scatter / PartOpt / cpu / BothRev 0.0002797910800018 s 0.0002816017400073 s 0.99
const_scatter / IPartOpt / cpu / PreRev 0.0002818124599889 s 0.0002837838400046 s 0.99
const_scatter / IPartOpt / cpu / PostRev 0.0002806165799938 s 0.0002813892200356 s 1.00
const_scatter / IPartOpt / cpu / BothRev 0.0002825130799737 s 0.0002830236400041 s 1.00
const_scatter / DefOpt / cpu / PreRev 0.000281767080005 s 0.0002838808800061 s 0.99
const_scatter / DefOpt / cpu / PostRev 0.0002816460400026 s 0.0002831627400064 s 0.99
const_scatter / DefOpt / cpu / BothRev 0.0002783594599623 s 0.0002837009600443 s 0.98
const_scatter / IDefOpt / cpu / PreRev 0.0002812273200197 s 0.0002835541400327 s 0.99
const_scatter / IDefOpt / cpu / PostRev 0.0002814970599956 s 0.0002814120199582 s 1.00
const_scatter / IDefOpt / cpu / BothRev 0.0002823395399718 s 0.0002812452999751 s 1.00
const_scatter / JaXPipe / cuda / Primal 0.000001887 s 0.0000024 s 0.79
const_scatter / Jax / cuda / Primal 0.000001887 s 0.0000024 s 0.79
const_scatter / HLOOpt / cuda / Primal 0.000001887 s 0.0000024 s 0.79
const_scatter / PartOpt / cuda / Primal 0.000001887 s 0.0000024 s 0.79
const_scatter / IPartOpt / cuda / Primal 0.000001887 s 0.0000024 s 0.79
const_scatter / DefOpt / cuda / Primal 0.000001887 s 0.000002432 s 0.78
const_scatter / IDefOpt / cuda / Primal 0.000001887 s 0.0000024 s 0.79
const_scatter / JaXPipe / cuda / Forward 0.000010208 s 0.000010592 s 0.96
const_scatter / Jax / cuda / Forward 0.000010048 s 0.00001072 s 0.94
const_scatter / HLOOpt / cuda / Forward 0.000009888 s 0.00001056 s 0.94
const_scatter / PartOpt / cuda / Forward 0.000009727 s 0.000010656 s 0.91
const_scatter / IPartOpt / cuda / Forward 0.00001008 s 0.000010591 s 0.95
const_scatter / DefOpt / cuda / Forward 0.000010336 s 0.00001056 s 0.98
const_scatter / IDefOpt / cuda / Forward 0.000009984 s 0.0000104 s 0.96
const_scatter / JaXPipe / cuda / PreRev 0.000015968 s 0.000017184 s 0.93
const_scatter / JaXPipe / cuda / PostRev 0.000015935000000000002 s 0.000016544 s 0.96
const_scatter / JaXPipe / cuda / BothRev 0.000016063999999999997 s 0.00001648 s 0.97
const_scatter / Jax / cuda / BothRev 0.000016448000000000002 s 0.00001696 s 0.97
const_scatter / HLOOpt / cuda / PreRev 0.000016352 s 0.000016608 s 0.98
const_scatter / HLOOpt / cuda / PostRev 0.00001568 s 0.000017375999999999998 s 0.90
const_scatter / HLOOpt / cuda / BothRev 0.000016191 s 0.00001648 s 0.98
const_scatter / PartOpt / cuda / PreRev 0.000016031 s 0.000017184 s 0.93
const_scatter / PartOpt / cuda / PostRev 0.000016032 s 0.000016544 s 0.97
const_scatter / PartOpt / cuda / BothRev 0.000016096 s 0.000017024999999999997 s 0.95
const_scatter / IPartOpt / cuda / PreRev 0.000016414999999999998 s 0.000016833 s 0.98
const_scatter / IPartOpt / cuda / PostRev 0.000016639 s 0.000016864 s 0.99
const_scatter / IPartOpt / cuda / BothRev 0.000016512 s 0.000017088 s 0.97
const_scatter / DefOpt / cuda / PreRev 0.000015935999999999998 s 0.000017408 s 0.92
const_scatter / DefOpt / cuda / PostRev 0.000016224 s 0.00001728 s 0.94
const_scatter / DefOpt / cuda / BothRev 0.00001648 s 0.000017152 s 0.96
const_scatter / IDefOpt / cuda / PreRev 0.000016128 s 0.000016832 s 0.96
const_scatter / IDefOpt / cuda / PostRev 0.000016096 s 0.000017056 s 0.94
const_scatter / IDefOpt / cuda / BothRev 0.00001632 s 0.000016673 s 0.98
const_scatter / JaXPipe / tpu / Primal 0.000003800075 s 0.000003787075 s 1.00
const_scatter / Jax / tpu / Primal 0.000003815075 s 0.000003811075 s 1.00
const_scatter / HLOOpt / tpu / Primal 0.0000038093 s 0.0000037736 s 1.01
const_scatter / PartOpt / tpu / Primal 0.000003819575 s 0.0000038083 s 1.00
const_scatter / IPartOpt / tpu / Primal 0.000003782975 s 0.000003805875 s 0.99
const_scatter / DefOpt / tpu / Primal 0.0000038136 s 0.000003811875 s 1.00
const_scatter / IDefOpt / tpu / Primal 0.00000377625 s 0.000003802625 s 0.99
const_scatter / JaXPipe / tpu / Forward 0.000006461175 s 0.000006502 s 0.99
const_scatter / Jax / tpu / Forward 0.00000648945 s 0.000006493175000000001 s 1.00
const_scatter / HLOOpt / tpu / Forward 0.0000064934 s 0.0000064633 s 1.00
const_scatter / PartOpt / tpu / Forward 0.000006475325 s 0.00000647815 s 1.00
const_scatter / IPartOpt / tpu / Forward 0.000006463775000000001 s 0.000006445625 s 1.00
const_scatter / DefOpt / tpu / Forward 0.000006497225 s 0.000006488575 s 1.00
const_scatter / IDefOpt / tpu / Forward 0.0000064615 s 0.0000064743 s 1.00
const_scatter / JaXPipe / tpu / PreRev 0.000006593649999999999 s 0.00000666135 s 0.99
const_scatter / JaXPipe / tpu / PostRev 0.000006603824999999999 s 0.000006666625000000001 s 0.99
const_scatter / JaXPipe / tpu / BothRev 0.000006630700000000001 s 0.000006674275 s 0.99
const_scatter / Jax / tpu / BothRev 0.00000662055 s 0.000006674675 s 0.99
const_scatter / HLOOpt / tpu / PreRev 0.0000066172 s 0.000006659374999999999 s 0.99
const_scatter / HLOOpt / tpu / PostRev 0.000006629350000000001 s 0.000006647675 s 1.00
const_scatter / HLOOpt / tpu / BothRev 0.0000065907 s 0.0000066701 s 0.99
const_scatter / PartOpt / tpu / PreRev 0.0000066098 s 0.000006665325 s 0.99
const_scatter / PartOpt / tpu / PostRev 0.000006618024999999999 s 0.000006675725 s 0.99
const_scatter / PartOpt / tpu / BothRev 0.000006618475 s 0.00000665525 s 0.99
const_scatter / IPartOpt / tpu / PreRev 0.0000066216 s 0.00000664535 s 1.00
const_scatter / IPartOpt / tpu / PostRev 0.0000066147750000000005 s 0.0000066836 s 0.99
const_scatter / IPartOpt / tpu / BothRev 0.00000659995 s 0.000006666674999999999 s 0.99
const_scatter / DefOpt / tpu / PreRev 0.000006613525 s 0.000006671650000000001 s 0.99
const_scatter / DefOpt / tpu / PostRev 0.000006604875 s 0.000006657475 s 0.99
const_scatter / DefOpt / tpu / BothRev 0.00000661515 s 0.000006662575 s 0.99
const_scatter / IDefOpt / tpu / PreRev 0.00000659475 s 0.0000066657500000000005 s 0.99
const_scatter / IDefOpt / tpu / PostRev 0.000006620325 s 0.000006667925 s 0.99
const_scatter / IDefOpt / tpu / BothRev 0.000006612575000000001 s 0.000006642375000000001 s 1.00
const_scatter / JaXPipe / cpu / Primal 0.000013406 s 0.000006957139994483441 s 1.93
const_scatter / Jax / cpu / Primal 0.000012626 s 0.000006754500018359976 s 1.87
const_scatter / HLOOpt / cpu / Primal 0.000013398 s 0.000006998379976721481 s 1.91
const_scatter / PartOpt / cpu / Primal 0.000013295 s 0.000006252760031202343 s 2.13
const_scatter / IPartOpt / cpu / Primal 0.000012791 s 0.000006638139984715962 s 1.93
const_scatter / DefOpt / cpu / Primal 0.00001374 s 0.000007144199998947442 s 1.92
const_scatter / IDefOpt / cpu / Primal 0.000013836 s 0.000007046919999993407 s 1.96
const_scatter / JaXPipe / cpu / Forward 0.000018924 s 0.000010530240024309025 s 1.80
const_scatter / Jax / cpu / Forward 0.000017962 s 0.000009437960052309793 s 1.90
const_scatter / HLOOpt / cpu / Forward 0.00001862 s 0.000010992900033670594 s 1.69
const_scatter / PartOpt / cpu / Forward 0.000019573 s 0.000010934680003629185 s 1.79
const_scatter / IPartOpt / cpu / Forward 0.000018361 s 0.00001130093995925563 s 1.62
const_scatter / DefOpt / cpu / Forward 0.000017855 s 0.00001061286003277928 s 1.68
const_scatter / IDefOpt / cpu / Forward 0.000017873 s 0.000010548560003371675 s 1.69
const_scatter / JaXPipe / cpu / PreRev 0.000523563 s 0.0002879407399541 s 1.82
const_scatter / JaXPipe / cpu / PostRev 0.000539954 s 0.0002806210999824 s 1.92
const_scatter / JaXPipe / cpu / BothRev 0.0005076939999999 s 0.0002837718799764 s 1.79
const_scatter / Jax / cpu / BothRev 0.000496962 s 0.0002842210200014 s 1.75
const_scatter / HLOOpt / cpu / PreRev 0.00052365 s 0.0002866692799489 s 1.83
const_scatter / HLOOpt / cpu / PostRev 0.000525946 s 0.0002865757799918 s 1.84
const_scatter / HLOOpt / cpu / BothRev 0.000512048 s 0.00028323980001 s 1.81
const_scatter / PartOpt / cpu / PreRev 0.000517721 s 0.0002816201200221 s 1.84
const_scatter / PartOpt / cpu / PostRev 0.000514347 s 0.0002784744400105 s 1.85
const_scatter / PartOpt / cpu / BothRev 0.0004954919999999 s 0.0002816017400073 s 1.76
const_scatter / IPartOpt / cpu / PreRev 0.000533065 s 0.0002837838400046 s 1.88
const_scatter / IPartOpt / cpu / PostRev 0.000529748 s 0.0002813892200356 s 1.88
const_scatter / IPartOpt / cpu / BothRev 0.000525575 s 0.0002830236400041 s 1.86
const_scatter / DefOpt / cpu / PreRev 0.000536782 s 0.0002838808800061 s 1.89
const_scatter / DefOpt / cpu / PostRev 0.000516642 s 0.0002831627400064 s 1.82
const_scatter / DefOpt / cpu / BothRev 0.000513734 s 0.0002837009600443 s 1.81
const_scatter / IDefOpt / cpu / PreRev 0.0005279589999999 s 0.0002835541400327 s 1.86
const_scatter / IDefOpt / cpu / PostRev 0.0005062799999999 s 0.0002814120199582 s 1.80
const_scatter / IDefOpt / cpu / BothRev 0.000522188 s 0.0002812452999751 s 1.86
const_scatter / JaXPipe / cpu / Primal 0.000008 s 0.000006957139994483441 s 1.15
const_scatter / Jax / cpu / Primal 0.000008 s 0.000006754500018359976 s 1.18
const_scatter / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000006998379976721481 s 1.29
const_scatter / PartOpt / cpu / Primal 0.000008 s 0.000006252760031202343 s 1.28
const_scatter / IPartOpt / cpu / Primal 0.000008 s 0.000006638139984715962 s 1.21
const_scatter / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000007144199998947442 s 1.26
const_scatter / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000007046919999993407 s 1.28
const_scatter / JaXPipe / cpu / Forward 0.000013 s 0.000010530240024309025 s 1.23
const_scatter / Jax / cpu / Forward 0.000011 s 0.000009437960052309793 s 1.17
const_scatter / HLOOpt / cpu / Forward 0.000013 s 0.000010992900033670594 s 1.18
const_scatter / PartOpt / cpu / Forward 0.000013 s 0.000010934680003629185 s 1.19
const_scatter / IPartOpt / cpu / Forward 0.000012 s 0.00001130093995925563 s 1.06
const_scatter / DefOpt / cpu / Forward 0.000013 s 0.00001061286003277928 s 1.22
const_scatter / IDefOpt / cpu / Forward 0.000012 s 0.000010548560003371675 s 1.14
const_scatter / JaXPipe / cpu / PreRev 0.000319 s 0.0002879407399541 s 1.11
const_scatter / JaXPipe / cpu / PostRev 0.000343 s 0.0002806210999824 s 1.22
const_scatter / JaXPipe / cpu / BothRev 0.000323 s 0.0002837718799764 s 1.14
const_scatter / Jax / cpu / BothRev 0.000327 s 0.0002842210200014 s 1.15
const_scatter / HLOOpt / cpu / PreRev 0.000322 s 0.0002866692799489 s 1.12
const_scatter / HLOOpt / cpu / PostRev 0.0003459999999999 s 0.0002865757799918 s 1.21
const_scatter / HLOOpt / cpu / BothRev 0.000322 s 0.00028323980001 s 1.14
const_scatter / PartOpt / cpu / PreRev 0.000328 s 0.0002816201200221 s 1.16
const_scatter / PartOpt / cpu / PostRev 0.000317 s 0.0002784744400105 s 1.14
const_scatter / PartOpt / cpu / BothRev 0.000324 s 0.0002816017400073 s 1.15
const_scatter / IPartOpt / cpu / PreRev 0.000343 s 0.0002837838400046 s 1.21
const_scatter / IPartOpt / cpu / PostRev 0.0003439999999999 s 0.0002813892200356 s 1.22
const_scatter / IPartOpt / cpu / BothRev 0.000335 s 0.0002830236400041 s 1.18
const_scatter / DefOpt / cpu / PreRev 0.000335 s 0.0002838808800061 s 1.18
const_scatter / DefOpt / cpu / PostRev 0.000316 s 0.0002831627400064 s 1.12
const_scatter / DefOpt / cpu / BothRev 0.00034 s 0.0002837009600443 s 1.20
const_scatter / IDefOpt / cpu / PreRev 0.000342 s 0.0002835541400327 s 1.21
const_scatter / IDefOpt / cpu / PostRev 0.000345 s 0.0002814120199582 s 1.23
const_scatter / IDefOpt / cpu / BothRev 0.000345 s 0.0002812452999751 s 1.23
GenDot / JaXPipe / cpu / Primal 0.000006965239990677219 s 0.000007272980037669186 s 0.96
GenDot / Jax / cpu / Primal 0.000006689760020890389 s 0.000006714679993820028 s 1.00
GenDot / HLOOpt / cpu / Primal 0.000007394379990728339 s 0.0000072533200091129405 s 1.02
GenDot / PartOpt / cpu / Primal 0.000007213359967863653 s 0.000007104000032995827 s 1.02
GenDot / IPartOpt / cpu / Primal 0.000007072100015648175 s 0.000006967859953874722 s 1.01
GenDot / DefOpt / cpu / Primal 0.000007107380024535815 s 0.000007702559996687341 s 0.92
GenDot / IDefOpt / cpu / Primal 0.000006915760013725958 s 0.000007013999993432662 s 0.99
GenDot / JaXPipe / cpu / Forward 0.000010921020020759896 s 0.000010708420004448272 s 1.02
GenDot / Jax / cpu / Forward 0.000010333359987271252 s 0.000010924940015684117 s 0.95
GenDot / HLOOpt / cpu / Forward 0.000011356320010236232 s 0.000011458319968369325 s 0.99
GenDot / PartOpt / cpu / Forward 0.000010568940051598476 s 0.000010547860019869404 s 1.00
GenDot / IPartOpt / cpu / Forward 0.000010403139976915552 s 0.000011269659953541124 s 0.92
GenDot / DefOpt / cpu / Forward 0.000010558639960436269 s 0.0000111043800097832 s 0.95
GenDot / IDefOpt / cpu / Forward 0.000010589199982860007 s 0.000010358019990235334 s 1.02
GenDot / JaXPipe / cpu / PreRev 0.000011312160022498576 s 0.000011568300005819764 s 0.98
GenDot / JaXPipe / cpu / PostRev 0.00001080494002053456 s 0.000009686379962658976 s 1.12
GenDot / JaXPipe / cpu / BothRev 0.000011591740021685836 s 0.00001107219999539666 s 1.05
GenDot / Jax / cpu / BothRev 0.000010161460013478064 s 0.000010564499998508836 s 0.96
GenDot / HLOOpt / cpu / PreRev 0.000011787439962063218 s 0.000011608579943640508 s 1.02
GenDot / HLOOpt / cpu / PostRev 0.00001304723996327084 s 0.000013093360003040289 s 1.00
GenDot / HLOOpt / cpu / BothRev 0.0000109259800046857 s 0.000010969279956043465 s 1.00
GenDot / PartOpt / cpu / PreRev 0.000010951879994536285 s 0.000011133559992231312 s 0.98
GenDot / PartOpt / cpu / PostRev 0.000010492200062799384 s 0.000010282300017934176 s 1.02
GenDot / PartOpt / cpu / BothRev 0.000011030919977201848 s 0.00001080339999134594 s 1.02
GenDot / IPartOpt / cpu / PreRev 0.000011121040006401018 s 0.000011101639984190116 s 1.00
GenDot / IPartOpt / cpu / PostRev 0.00001058195999576128 s 0.000010240600022370928 s 1.03
GenDot / IPartOpt / cpu / BothRev 0.000010865500053114376 s 0.000011099660005129408 s 0.98
GenDot / DefOpt / cpu / PreRev 0.000011271579978711088 s 0.00001056862000041292 s 1.07
GenDot / DefOpt / cpu / PostRev 0.00001067847994818294 s 0.000010853540006792172 s 0.98
GenDot / DefOpt / cpu / BothRev 0.00001162627998382959 s 0.00001106954000533733 s 1.05
GenDot / IDefOpt / cpu / PreRev 0.000010985220014845254 s 0.000010772000014185325 s 1.02
GenDot / IDefOpt / cpu / PostRev 0.000010590859956209895 s 0.000011088599994764082 s 0.96
GenDot / IDefOpt / cpu / BothRev 0.000010944239975287928 s 0.000010910359997069465 s 1.00
GenDot / JaXPipe / cuda / Primal 0.000002015 s 0.000002495 s 0.81
GenDot / Jax / cuda / Primal 0.000002015 s 0.000002496 s 0.81
GenDot / HLOOpt / cuda / Primal 0.000001984 s 0.000002496 s 0.79
GenDot / PartOpt / cuda / Primal 0.000002015 s 0.000002527 s 0.80
GenDot / IPartOpt / cuda / Primal 0.000002015 s 0.000002496 s 0.81
GenDot / DefOpt / cuda / Primal 0.000001983 s 0.000002495 s 0.79
GenDot / IDefOpt / cuda / Primal 0.000001984 s 0.000002495 s 0.80
GenDot / JaXPipe / cuda / Forward 0.00000976 s 0.000010528 s 0.93
GenDot / Jax / cuda / Forward 0.000009952 s 0.000011136 s 0.89
GenDot / HLOOpt / cuda / Forward 0.000009824 s 0.000010528 s 0.93
GenDot / PartOpt / cuda / Forward 0.000009792 s 0.000010688 s 0.92
GenDot / IPartOpt / cuda / Forward 0.000010111 s 0.000010496 s 0.96
GenDot / DefOpt / cuda / Forward 0.00000976 s 0.000010687 s 0.91
GenDot / IDefOpt / cuda / Forward 0.000009536 s 0.000010592 s 0.90
GenDot / JaXPipe / cuda / PreRev 0.000009535 s 0.000010849 s 0.88
GenDot / JaXPipe / cuda / PostRev 0.000010272 s 0.00001088 s 0.94
GenDot / JaXPipe / cuda / BothRev 0.000010048 s 0.000010496 s 0.96
GenDot / Jax / cuda / BothRev 0.00001008 s 0.000011295 s 0.89
GenDot / HLOOpt / cuda / PreRev 0.000009951 s 0.000010912 s 0.91
GenDot / HLOOpt / cuda / PostRev 0.000009824 s 0.000010624 s 0.92
GenDot / HLOOpt / cuda / BothRev 0.00000928 s 0.00001088 s 0.85
GenDot / PartOpt / cuda / PreRev 0.000010048 s 0.000010463 s 0.96
GenDot / PartOpt / cuda / PostRev 0.000009888 s 0.000010688 s 0.93
GenDot / PartOpt / cuda / BothRev 0.000010176 s 0.00001104 s 0.92
GenDot / IPartOpt / cuda / PreRev 0.000010015 s 0.00001088 s 0.92
GenDot / IPartOpt / cuda / PostRev 0.000010016 s 0.000010784 s 0.93
GenDot / IPartOpt / cuda / BothRev 0.000009407 s 0.000010912 s 0.86
GenDot / DefOpt / cuda / PreRev 0.000009984 s 0.000010624 s 0.94
GenDot / DefOpt / cuda / PostRev 0.000009888 s 0.00001056 s 0.94
GenDot / DefOpt / cuda / BothRev 0.000010016 s 0.000010912 s 0.92
GenDot / IDefOpt / cuda / PreRev 0.000009952 s 0.000011008 s 0.90
GenDot / IDefOpt / cuda / PostRev 0.000009952 s 0.000010656 s 0.93
GenDot / IDefOpt / cuda / BothRev 0.000010111 s 0.00001088 s 0.93
GenDot / JaXPipe / tpu / Primal 9.2605e-7 s 9.25975e-7 s 1.00
GenDot / Jax / tpu / Primal 9.25625e-7 s 9.26125e-7 s 1.00
GenDot / HLOOpt / tpu / Primal 0.00000156165 s 0.00000156145 s 1.00
GenDot / PartOpt / tpu / Primal 9.25575e-7 s 9.253e-7 s 1.00
GenDot / IPartOpt / tpu / Primal 9.26e-7 s 9.2585e-7 s 1.00
GenDot / DefOpt / tpu / Primal 0.000001482375 s 0.00000149165 s 0.99
GenDot / IDefOpt / tpu / Primal 0.0000015530250000000005 s 0.00000156785 s 0.99
GenDot / JaXPipe / tpu / Forward 0.00000315805 s 0.000003175 s 0.99
GenDot / Jax / tpu / Forward 0.00000231305 s 0.000002328925 s 0.99
GenDot / HLOOpt / tpu / Forward 0.000003109175 s 0.0000031329 s 0.99
GenDot / PartOpt / tpu / Forward 0.000003205475 s 0.000003214075 s 1.00
GenDot / IPartOpt / tpu / Forward 0.0000031187 s 0.000003122725 s 1.00
GenDot / DefOpt / tpu / Forward 0.0000032025500000000004 s 0.00000321805 s 1.00
GenDot / IDefOpt / tpu / Forward 0.0000031104500000000004 s 0.000003117475 s 1.00
GenDot / JaXPipe / tpu / PreRev 0.000002954275 s 0.0000029548 s 1.00
GenDot / JaXPipe / tpu / PostRev 0.0000024049 s 0.0000024061 s 1.00
GenDot / JaXPipe / tpu / BothRev 0.0000029537 s 0.000002953975 s 1.00
GenDot / Jax / tpu / BothRev 0.000002409025 s 0.0000024153250000000005 s 1.00
GenDot / HLOOpt / tpu / PreRev 0.00000295485 s 0.00000295205 s 1.00
GenDot / HLOOpt / tpu / PostRev 0.000002927175 s 0.000002932875 s 1.00
GenDot / HLOOpt / tpu / BothRev 0.000002948075 s 0.000002954175 s 1.00
GenDot / PartOpt / tpu / PreRev 0.000002924875 s 0.000002931125 s 1.00
GenDot / PartOpt / tpu / PostRev 0.0000023895749999999995 s 0.000002389 s 1.00
GenDot / PartOpt / tpu / BothRev 0.000002922575 s 0.000002931375 s 1.00
GenDot / IPartOpt / tpu / PreRev 0.000002953775 s 0.00000296195 s 1.00
GenDot / IPartOpt / tpu / PostRev 0.000002404825 s 0.000002408925 s 1.00
GenDot / IPartOpt / tpu / BothRev 0.00000294715 s 0.000002958225 s 1.00
GenDot / DefOpt / tpu / PreRev 0.0000029256 s 0.00000293445 s 1.00
GenDot / DefOpt / tpu / PostRev 0.000002947925 s 0.0000029609000000000003 s 1.00
GenDot / DefOpt / tpu / BothRev 0.00000293535 s 0.0000029365750000000003 s 1.00
GenDot / IDefOpt / tpu / PreRev 0.0000029406 s 0.0000029634 s 0.99
GenDot / IDefOpt / tpu / PostRev 0.000002941625 s 0.000002937875 s 1.00
GenDot / IDefOpt / tpu / BothRev 0.0000029468 s 0.000002956025 s 1.00
GenDot / JaXPipe / cpu / Primal 0.000015246 s 0.000007272980037669186 s 2.10
GenDot / Jax / cpu / Primal 0.000015073 s 0.000006714679993820028 s 2.24
GenDot / HLOOpt / cpu / Primal 0.000014105 s 0.0000072533200091129405 s 1.94
GenDot / PartOpt / cpu / Primal 0.000015156 s 0.000007104000032995827 s 2.13
GenDot / IPartOpt / cpu / Primal 0.000014286 s 0.000006967859953874722 s 2.05
GenDot / DefOpt / cpu / Primal 0.0000143 s 0.000007702559996687341 s 1.86
GenDot / IDefOpt / cpu / Primal 0.000014115 s 0.000007013999993432662 s 2.01
GenDot / JaXPipe / cpu / Forward 0.000019446 s 0.000010708420004448272 s 1.82
GenDot / Jax / cpu / Forward 0.000020775 s 0.000010924940015684117 s 1.90
GenDot / HLOOpt / cpu / Forward 0.000019183 s 0.000011458319968369325 s 1.67
GenDot / PartOpt / cpu / Forward 0.000019368 s 0.000010547860019869404 s 1.84
GenDot / IPartOpt / cpu / Forward 0.000019017 s 0.000011269659953541124 s 1.69
GenDot / DefOpt / cpu / Forward 0.000018906 s 0.0000111043800097832 s 1.70
GenDot / IDefOpt / cpu / Forward 0.000019574000000000003 s 0.000010358019990235334 s 1.89
GenDot / JaXPipe / cpu / PreRev 0.000020063 s 0.000011568300005819764 s 1.73
GenDot / JaXPipe / cpu / PostRev 0.000022087 s 0.000009686379962658976 s 2.28
GenDot / JaXPipe / cpu / BothRev 0.000019457 s 0.00001107219999539666 s 1.76
GenDot / Jax / cpu / BothRev 0.000020136 s 0.000010564499998508836 s 1.91
GenDot / HLOOpt / cpu / PreRev 0.00001913 s 0.000011608579943640508 s 1.65
GenDot / HLOOpt / cpu / PostRev 0.000019267 s 0.000013093360003040289 s 1.47
GenDot / HLOOpt / cpu / BothRev 0.000019105 s 0.000010969279956043465 s 1.74
GenDot / PartOpt / cpu / PreRev 0.000018961 s 0.000011133559992231312 s 1.70
GenDot / PartOpt / cpu / PostRev 0.000020178 s 0.000010282300017934176 s 1.96
GenDot / PartOpt / cpu / BothRev 0.000019366 s 0.00001080339999134594 s 1.79
GenDot / IPartOpt / cpu / PreRev 0.000019434 s 0.000011101639984190116 s 1.75
GenDot / IPartOpt / cpu / PostRev 0.000020335 s 0.000010240600022370928 s 1.99
GenDot / IPartOpt / cpu / BothRev 0.000019168 s 0.000011099660005129408 s 1.73
GenDot / DefOpt / cpu / PreRev 0.000020414 s 0.00001056862000041292 s 1.93
GenDot / DefOpt / cpu / PostRev 0.000020206 s 0.000010853540006792172 s 1.86
GenDot / DefOpt / cpu / BothRev 0.000019664 s 0.00001106954000533733 s 1.78
GenDot / IDefOpt / cpu / PreRev 0.000020464 s 0.000010772000014185325 s 1.90
GenDot / IDefOpt / cpu / PostRev 0.000019379 s 0.000011088599994764082 s 1.75
GenDot / IDefOpt / cpu / BothRev 0.000019207 s 0.000010910359997069465 s 1.76
GenDot / JaXPipe / cpu / Primal 0.00001 s 0.000007272980037669186 s 1.37
GenDot / Jax / cpu / Primal 0.00001 s 0.000006714679993820028 s 1.49
GenDot / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.0000072533200091129405 s 1.24
GenDot / PartOpt / cpu / Primal 0.00001 s 0.000007104000032995827 s 1.41
GenDot / IPartOpt / cpu / Primal 0.00001 s 0.000006967859953874722 s 1.44
GenDot / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000007702559996687341 s 1.17
GenDot / IDefOpt / cpu / Primal 0.00001 s 0.000007013999993432662 s 1.43
GenDot / JaXPipe / cpu / Forward 0.000013 s 0.000010708420004448272 s 1.21
GenDot / Jax / cpu / Forward 0.000014 s 0.000010924940015684117 s 1.28
GenDot / HLOOpt / cpu / Forward 0.000013 s 0.000011458319968369325 s 1.13
GenDot / PartOpt / cpu / Forward 0.000013 s 0.000010547860019869404 s 1.23
GenDot / IPartOpt / cpu / Forward 0.000013 s 0.000011269659953541124 s 1.15
GenDot / DefOpt / cpu / Forward 0.000013 s 0.0000111043800097832 s 1.17
GenDot / IDefOpt / cpu / Forward 0.000014 s 0.000010358019990235334 s 1.35
GenDot / JaXPipe / cpu / PreRev 0.000014 s 0.000011568300005819764 s 1.21
GenDot / JaXPipe / cpu / PostRev 0.000014 s 0.000009686379962658976 s 1.45
GenDot / JaXPipe / cpu / BothRev 0.000014 s 0.00001107219999539666 s 1.26
GenDot / Jax / cpu / BothRev 0.000013 s 0.000010564499998508836 s 1.23
GenDot / HLOOpt / cpu / PreRev 0.000014 s 0.000011608579943640508 s 1.21
GenDot / HLOOpt / cpu / PostRev 0.000014 s 0.000013093360003040289 s 1.07
GenDot / HLOOpt / cpu / BothRev 0.000014 s 0.000010969279956043465 s 1.28
GenDot / PartOpt / cpu / PreRev 0.000013 s 0.000011133559992231312 s 1.17
GenDot / PartOpt / cpu / PostRev 0.000013 s 0.000010282300017934176 s 1.26
GenDot / PartOpt / cpu / BothRev 0.000013 s 0.00001080339999134594 s 1.20
GenDot / IPartOpt / cpu / PreRev 0.000013 s 0.000011101639984190116 s 1.17
GenDot / IPartOpt / cpu / PostRev 0.000013 s 0.000010240600022370928 s 1.27
GenDot / IPartOpt / cpu / BothRev 0.000013 s 0.000011099660005129408 s 1.17
GenDot / DefOpt / cpu / PreRev 0.000013 s 0.00001056862000041292 s 1.23
GenDot / DefOpt / cpu / PostRev 0.000013 s 0.000010853540006792172 s 1.20
GenDot / DefOpt / cpu / BothRev 0.000014 s 0.00001106954000533733 s 1.26
GenDot / IDefOpt / cpu / PreRev 0.000013 s 0.000010772000014185325 s 1.21
GenDot / IDefOpt / cpu / PostRev 0.000013 s 0.000011088599994764082 s 1.17
GenDot / IDefOpt / cpu / BothRev 0.000014 s 0.000010910359997069465 s 1.28
hlo_ffi / JaXPipe / cpu / Primal 0.000012240140040375991 s 0.000012390339998091804 s 0.99
hlo_ffi / Jax / cpu / Primal 0.00001157638006588968 s 0.000012064980001014192 s 0.96
hlo_ffi / HLOOpt / cpu / Primal 0.000011721860000761809 s 0.000012583739935507766 s 0.93
hlo_ffi / PartOpt / cpu / Primal 0.000011699839997163507 s 0.00001154250004219648 s 1.01
hlo_ffi / IPartOpt / cpu / Primal 0.000011617480013228488 s 0.00001246871998773713 s 0.93
hlo_ffi / DefOpt / cpu / Primal 0.00001185231998533709 s 0.000011856080000143266 s 1.00
hlo_ffi / IDefOpt / cpu / Primal 0.000011716600056388415 s 0.000012595459984368063 s 0.93
hlo_ffi / JaXPipe / cpu / Forward 0.000015412739985549704 s 0.000016476139999213048 s 0.94
hlo_ffi / Jax / cpu / Forward 0.00001534925998385006 s 0.000016035579956223956 s 0.96
hlo_ffi / HLOOpt / cpu / Forward 0.000015859559998716576 s 0.000016663379983583582 s 0.95
hlo_ffi / PartOpt / cpu / Forward 0.000015761579988975426 s 0.000016540899996471125 s 0.95
hlo_ffi / IPartOpt / cpu / Forward 0.000016127080016303807 s 0.0000168811800176627 s 0.96
hlo_ffi / DefOpt / cpu / Forward 0.000015691139969931102 s 0.00001660346000790014 s 0.95
hlo_ffi / IDefOpt / cpu / Forward 0.000015982739969331307 s 0.000016710239988242394 s 0.96
hlo_ffi / JaXPipe / cpu / PreRev 0.00001624817997253558 s 0.000016964540018307163 s 0.96
hlo_ffi / JaXPipe / cpu / PostRev 0.000015988880022632657 s 0.00001678761998846312 s 0.95
hlo_ffi / JaXPipe / cpu / BothRev 0.000015965239990691772 s 0.000015760979995320667 s 1.01
hlo_ffi / Jax / cpu / BothRev 0.00001623366006242577 s 0.000016081939993455307 s 1.01
hlo_ffi / HLOOpt / cpu / PreRev 0.000016015119981602766 s 0.000016797040025267052 s 0.95
hlo_ffi / HLOOpt / cpu / PostRev 0.000018103880001945072 s 0.000017709400035528235 s 1.02
hlo_ffi / HLOOpt / cpu / BothRev 0.00001654747998145467 s 0.00001680888000919367 s 0.98
hlo_ffi / PartOpt / cpu / PreRev 0.000016460539991385304 s 0.00001630514003409189 s 1.01
hlo_ffi / PartOpt / cpu / PostRev 0.00001631067998459912 s 0.000015797839978404227 s 1.03
hlo_ffi / PartOpt / cpu / BothRev 0.000015923640012260877 s 0.000020467540052777623 s 0.78
hlo_ffi / IPartOpt / cpu / PreRev 0.000016011400011848308 s 0.000016567559987379356 s 0.97
hlo_ffi / IPartOpt / cpu / PostRev 0.000015601839986629785 s 0.00001622985998437798 s 0.96
hlo_ffi / IPartOpt / cpu / BothRev 0.000015575620000163327 s 0.000015737419980723645 s 0.99
hlo_ffi / DefOpt / cpu / PreRev 0.000016462519961351064 s 0.00001633539998692868 s 1.01
hlo_ffi / DefOpt / cpu / PostRev 0.000016163420032171416 s 0.000016016079953260487 s 1.01
hlo_ffi / DefOpt / cpu / BothRev 0.00001540840001325705 s 0.00001588574001289089 s 0.97
hlo_ffi / IDefOpt / cpu / PreRev 0.000016129840005305596 s 0.00001680954000221391 s 0.96
hlo_ffi / IDefOpt / cpu / PostRev 0.00001542936001897033 s 0.000015812420006113826 s 0.98
hlo_ffi / IDefOpt / cpu / BothRev 0.000015857279986448703 s 0.00001625476001208881 s 0.98
hlo_ffi / JaXPipe / cuda / Primal 0.000001983 s 0.0000023670000000000004 s 0.84
hlo_ffi / Jax / cuda / Primal 0.000001984 s 0.0000023670000000000004 s 0.84
hlo_ffi / HLOOpt / cuda / Primal 0.000001983 s 0.0000023670000000000004 s 0.84
hlo_ffi / PartOpt / cuda / Primal 0.000001984 s 0.0000023670000000000004 s 0.84
hlo_ffi / IPartOpt / cuda / Primal 0.000001984 s 0.0000023670000000000004 s 0.84
hlo_ffi / DefOpt / cuda / Primal 0.000001983 s 0.0000023670000000000004 s 0.84
hlo_ffi / IDefOpt / cuda / Primal 0.000001983 s 0.0000023670000000000004 s 0.84
hlo_ffi / JaXPipe / cuda / Forward 0.00000208 s 0.000002432 s 0.86
hlo_ffi / Jax / cuda / Forward 0.00000208 s 0.000002433 s 0.85
hlo_ffi / HLOOpt / cuda / Forward 0.000002048 s 0.000002432 s 0.84
hlo_ffi / PartOpt / cuda / Forward 0.00000208 s 0.000002432 s 0.86
hlo_ffi / IPartOpt / cuda / Forward 0.000002048 s 0.000002432 s 0.84
hlo_ffi / DefOpt / cuda / Forward 0.000002079 s 0.000002432 s 0.85
hlo_ffi / IDefOpt / cuda / Forward 0.000002079 s 0.000002432 s 0.85
hlo_ffi / JaXPipe / cuda / PreRev 0.000002048 s 0.000002431 s 0.84
hlo_ffi / JaXPipe / cuda / PostRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / JaXPipe / cuda / BothRev 0.000002048 s 0.000002431 s 0.84
hlo_ffi / Jax / cuda / BothRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / HLOOpt / cuda / PreRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / HLOOpt / cuda / PostRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / HLOOpt / cuda / BothRev 0.000002047 s 0.000002432 s 0.84
hlo_ffi / PartOpt / cuda / PreRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / PartOpt / cuda / PostRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / PartOpt / cuda / BothRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / IPartOpt / cuda / PreRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / IPartOpt / cuda / PostRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / IPartOpt / cuda / BothRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / DefOpt / cuda / PreRev 0.000002048 s 0.000002431 s 0.84
hlo_ffi / DefOpt / cuda / PostRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / DefOpt / cuda / BothRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / IDefOpt / cuda / PreRev 0.000002047 s 0.000002431 s 0.84
hlo_ffi / IDefOpt / cuda / PostRev 0.000002047 s 0.000002432 s 0.84
hlo_ffi / IDefOpt / cuda / BothRev 0.000002048 s 0.000002431 s 0.84
hlo_ffi / JaXPipe / tpu / Primal 9.21925e-7 s 9.18075e-7 s 1.00
hlo_ffi / Jax / tpu / Primal 9.5195e-7 s 9.489e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Primal 8.96025e-7 s 8.948749999999999e-7 s 1.00
hlo_ffi / PartOpt / tpu / Primal 9.509e-7 s 9.5065e-7 s 1.00
hlo_ffi / IPartOpt / tpu / Primal 8.984749999999999e-7 s 8.969749999999999e-7 s 1.00
hlo_ffi / DefOpt / tpu / Primal 9.499e-7 s 9.506e-7 s 1.00
hlo_ffi / IDefOpt / tpu / Primal 8.959749999999999e-7 s 8.94625e-7 s 1.00
hlo_ffi / JaXPipe / tpu / Forward 9.496e-7 s 9.49625e-7 s 1.00
hlo_ffi / Jax / tpu / Forward 9.82175e-7 s 9.818749999999998e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Forward 9.74525e-7 s 9.74075e-7 s 1.00
hlo_ffi / PartOpt / tpu / Forward 9.338e-7 s 9.341e-7 s 1.00
hlo_ffi / IPartOpt / tpu / Forward 9.745e-7 s 9.7415e-7 s 1.00
hlo_ffi / DefOpt / tpu / Forward 9.34575e-7 s 9.33925e-7 s 1.00
hlo_ffi / IDefOpt / tpu / Forward 9.74575e-7 s 9.74625e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PreRev 9.324e-7 s 9.32275e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PostRev 9.65225e-7 s 9.65375e-7 s 1.00
hlo_ffi / JaXPipe / tpu / BothRev 9.62375e-7 s 9.621e-7 s 1.00
hlo_ffi / Jax / tpu / BothRev 9.6525e-7 s 9.65525e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PreRev 9.623e-7 s 9.62875e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PostRev 9.6475e-7 s 9.65875e-7 s 1.00
hlo_ffi / HLOOpt / tpu / BothRev 9.624e-7 s 9.62425e-7 s 1.00
hlo_ffi / PartOpt / tpu / PreRev 9.652e-7 s 9.6535e-7 s 1.00
hlo_ffi / PartOpt / tpu / PostRev 9.622e-7 s 9.6265e-7 s 1.00
hlo_ffi / PartOpt / tpu / BothRev 9.64925e-7 s 9.655e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PreRev 9.6205e-7 s 9.6235e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PostRev 9.650749999999998e-7 s 9.657499999999998e-7 s 1.00
hlo_ffi / IPartOpt / tpu / BothRev 9.61925e-7 s 9.6255e-7 s 1.00
hlo_ffi / DefOpt / tpu / PreRev 9.656e-7 s 9.65375e-7 s 1.00
hlo_ffi / DefOpt / tpu / PostRev 9.621e-7 s 9.627e-7 s 1.00
hlo_ffi / DefOpt / tpu / BothRev 9.6545e-7 s 9.65e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PreRev 9.62375e-7 s 9.621e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PostRev 9.6565e-7 s 9.65675e-7 s 1.00
hlo_ffi / IDefOpt / tpu / BothRev 9.63075e-7 s 9.62675e-7 s 1.00
hlo_ffi / JaXPipe / cpu / Primal 0.000017537 s 0.000012390339998091804 s 1.42
hlo_ffi / Jax / cpu / Primal 0.000017506 s 0.000012064980001014192 s 1.45
hlo_ffi / HLOOpt / cpu / Primal 0.000017539 s 0.000012583739935507766 s 1.39
hlo_ffi / PartOpt / cpu / Primal 0.000017614 s 0.00001154250004219648 s 1.53
hlo_ffi / IPartOpt / cpu / Primal 0.000017468999999999998 s 0.00001246871998773713 s 1.40
hlo_ffi / DefOpt / cpu / Primal 0.000017661 s 0.000011856080000143266 s 1.49
hlo_ffi / IDefOpt / cpu / Primal 0.000017475 s 0.000012595459984368063 s 1.39
hlo_ffi / JaXPipe / cpu / Forward 0.000023934 s 0.000016476139999213048 s 1.45
hlo_ffi / Jax / cpu / Forward 0.000024525000000000003 s 0.000016035579956223956 s 1.53
hlo_ffi / HLOOpt / cpu / Forward 0.000025164 s 0.000016663379983583582 s 1.51
hlo_ffi / PartOpt / cpu / Forward 0.000024789 s 0.000016540899996471125 s 1.50
hlo_ffi / IPartOpt / cpu / Forward 0.000023626 s 0.0000168811800176627 s 1.40
hlo_ffi / DefOpt / cpu / Forward 0.00002402 s 0.00001660346000790014 s 1.45
hlo_ffi / IDefOpt / cpu / Forward 0.000023675000000000003 s 0.000016710239988242394 s 1.42
hlo_ffi / JaXPipe / cpu / PreRev 0.000024501 s 0.000016964540018307163 s 1.44
hlo_ffi / JaXPipe / cpu / PostRev 0.00002353 s 0.00001678761998846312 s 1.40
hlo_ffi / JaXPipe / cpu / BothRev 0.000023541 s 0.000015760979995320667 s 1.49
hlo_ffi / Jax / cpu / BothRev 0.000023936 s 0.000016081939993455307 s 1.49
hlo_ffi / HLOOpt / cpu / PreRev 0.000023608 s 0.000016797040025267052 s 1.41
hlo_ffi / HLOOpt / cpu / PostRev 0.00002421 s 0.000017709400035528235 s 1.37
hlo_ffi / HLOOpt / cpu / BothRev 0.000024639 s 0.00001680888000919367 s 1.47
hlo_ffi / PartOpt / cpu / PreRev 0.000023653 s 0.00001630514003409189 s 1.45
hlo_ffi / PartOpt / cpu / PostRev 0.000023578 s 0.000015797839978404227 s 1.49
hlo_ffi / PartOpt / cpu / BothRev 0.000023567 s 0.000020467540052777623 s 1.15
hlo_ffi / IPartOpt / cpu / PreRev 0.000023781 s 0.000016567559987379356 s 1.44
hlo_ffi / IPartOpt / cpu / PostRev 0.000023040000000000003 s 0.00001622985998437798 s 1.42
hlo_ffi / IPartOpt / cpu / BothRev 0.000023642 s 0.000015737419980723645 s 1.50
hlo_ffi / DefOpt / cpu / PreRev 0.00002367 s 0.00001633539998692868 s 1.45
hlo_ffi / DefOpt / cpu / PostRev 0.00002384 s 0.000016016079953260487 s 1.49
hlo_ffi / DefOpt / cpu / BothRev 0.000023982 s 0.00001588574001289089 s 1.51
hlo_ffi / IDefOpt / cpu / PreRev 0.000024887 s 0.00001680954000221391 s 1.48
hlo_ffi / IDefOpt / cpu / PostRev 0.000024646 s 0.000015812420006113826 s 1.56
hlo_ffi / IDefOpt / cpu / BothRev 0.000025731 s 0.00001625476001208881 s 1.58
hlo_ffi / JaXPipe / cpu / Primal 0.000012 s 0.000012390339998091804 s 0.97
hlo_ffi / Jax / cpu / Primal 0.000012 s 0.000012064980001014192 s 0.99
hlo_ffi / HLOOpt / cpu / Primal 0.000012 s 0.000012583739935507766 s 0.95
hlo_ffi / PartOpt / cpu / Primal 0.000012 s 0.00001154250004219648 s 1.04
hlo_ffi / IPartOpt / cpu / Primal 0.000012 s 0.00001246871998773713 s 0.96
hlo_ffi / DefOpt / cpu / Primal 0.000012 s 0.000011856080000143266 s 1.01
hlo_ffi / IDefOpt / cpu / Primal 0.000012 s 0.000012595459984368063 s 0.95
hlo_ffi / JaXPipe / cpu / Forward 0.000017999999999999997 s 0.000016476139999213048 s 1.09
hlo_ffi / Jax / cpu / Forward 0.000017999999999999997 s 0.000016035579956223956 s 1.12
hlo_ffi / HLOOpt / cpu / Forward 0.000017 s 0.000016663379983583582 s 1.02
hlo_ffi / PartOpt / cpu / Forward 0.000017999999999999997 s 0.000016540899996471125 s 1.09
hlo_ffi / IPartOpt / cpu / Forward 0.000017 s 0.0000168811800176627 s 1.01
hlo_ffi / DefOpt / cpu / Forward 0.000016 s 0.00001660346000790014 s 0.96
hlo_ffi / IDefOpt / cpu / Forward 0.000017 s 0.000016710239988242394 s 1.02
hlo_ffi / JaXPipe / cpu / PreRev 0.000017 s 0.000016964540018307163 s 1.00
hlo_ffi / JaXPipe / cpu / PostRev 0.000017999999999999997 s 0.00001678761998846312 s 1.07
hlo_ffi / JaXPipe / cpu / BothRev 0.000017999999999999997 s 0.000015760979995320667 s 1.14
hlo_ffi / Jax / cpu / BothRev 0.000017 s 0.000016081939993455307 s 1.06
hlo_ffi / HLOOpt / cpu / PreRev 0.000016 s 0.000016797040025267052 s 0.95
hlo_ffi / HLOOpt / cpu / PostRev 0.000017 s 0.000017709400035528235 s 0.96
hlo_ffi / HLOOpt / cpu / BothRev 0.000017 s 0.00001680888000919367 s 1.01
hlo_ffi / PartOpt / cpu / PreRev 0.000017 s 0.00001630514003409189 s 1.04
hlo_ffi / PartOpt / cpu / PostRev 0.000017999999999999997 s 0.000015797839978404227 s 1.14
hlo_ffi / PartOpt / cpu / BothRev 0.000017 s 0.000020467540052777623 s 0.83
hlo_ffi / IPartOpt / cpu / PreRev 0.000017999999999999997 s 0.000016567559987379356 s 1.09
hlo_ffi / IPartOpt / cpu / PostRev 0.000017999999999999997 s 0.00001622985998437798 s 1.11
hlo_ffi / IPartOpt / cpu / BothRev 0.000017 s 0.000015737419980723645 s 1.08
hlo_ffi / DefOpt / cpu / PreRev 0.000017999999999999997 s 0.00001633539998692868 s 1.10
hlo_ffi / DefOpt / cpu / PostRev 0.000017 s 0.000016016079953260487 s 1.06
hlo_ffi / DefOpt / cpu / BothRev 0.000017 s 0.00001588574001289089 s 1.07
hlo_ffi / IDefOpt / cpu / PreRev 0.000017 s 0.00001680954000221391 s 1.01
hlo_ffi / IDefOpt / cpu / PostRev 0.000017 s 0.000015812420006113826 s 1.08
hlo_ffi / IDefOpt / cpu / BothRev 0.000017 s 0.00001625476001208881 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 / cpu / Primal 0.0010088962002555 s 0.0010052658000859 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 / cpu / Primal 0.0009219673999723 s 0.0009495696001067 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Primal 0.0010422282000035 s 0.0009717128000374 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 / PartOpt / cpu / Primal 0.0010100028000124 s 0.0010850222000954 s 0.93
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 0.000961145599922 s 0.0009566142000949 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 / Primal 0.0010451174000991 s 0.0009814416000153 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 / cpu / Primal 0.0009948613998858 s 0.0010132667999641 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 / JaXPipe / cpu / Forward 0.002480371600086 s 0.002265069600071 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 / Jax / cpu / Forward 0.002328733000013 s 0.0024919319998844 s 0.93
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Forward 0.0023386279999613 s 0.0022515823999128 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 / cpu / Forward 0.0024689156001841 s 0.0022670832000585 s 1.09
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 0.0023543191999124 s 0.0023049661999721 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 / Forward 0.0022990140000729 s 0.002297788999931 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 / cpu / Forward 0.0022774761999244 s 0.0023780273999364 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 / JaXPipe / cpu / PreRev 0.0061234057999172 s 0.0058777129999725 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 / cpu / PostRev 0.0075263314000949 s 0.0059237931999632 s 1.27
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / BothRev 0.00399079660001 s 0.0058832524001445 s 0.68
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / BothRev 0.0062076285999864 s 0.0052997615999629 s 1.17
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 0.004074772199965 s 0.0062640774000101 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 / HLOOpt / cpu / PostRev 0.0069768165999448 s 0.0057815731999653 s 1.21
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 0.0042269330000635 s 0.0052943166000659 s 0.80
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 0.006580261199997 s 0.0062454572000206 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 / PostRev 0.0042616530000486 s 0.0064490196000406 s 0.66
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 0.0068245170000409 s 0.0064472054000361 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 / cpu / PreRev 0.0054450128001008 s 0.0062578827999459 s 0.87
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 0.0062739853999119 s 0.0061021370000162 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 / cpu / BothRev 0.00538210779996 s 0.0064745801998469 s 0.83
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 0.0069938533999447 s 0.0062355684001886 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 / DefOpt / cpu / PostRev 0.0040700307999941 s 0.0056638985999597 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 / DefOpt / cpu / BothRev 0.0063585260000763 s 0.0056925937999039 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 / IDefOpt / cpu / PreRev 0.0054348261998711 s 0.0060477895999611 s 0.90
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 0.0061171378001745 s 0.0065292453999973 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.0041775445999519 s 0.0050366497998766 s 0.83
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 0.00028019 s 0.0002965769999999 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / Primal 0.000279486 s 0.000297536 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Primal 0.000287166 s 0.00030368 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / Primal 0.00027955 s 0.000297153 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Primal 0.000279774 s 0.000296641 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Primal 0.000287133 s 0.000304224 s 0.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / Primal 0.000287742 s 0.00030384 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / Forward 0.000557693 s 0.000583713 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 / Jax / cuda / Forward 0.000538684 s 0.000567265 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Forward 0.000557467 s 0.000582465 s 0.96
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / Forward 0.000557308 s 0.000582817 s 0.96
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Forward 0.000558747 s 0.000583168 s 0.96
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Forward 0.000557756 s 0.000582561 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 / cuda / Forward 0.000557627 s 0.000582273 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 / JaXPipe / cuda / PreRev 0.00102764 s 0.001053409 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 / JaXPipe / cuda / PostRev 0.000987768 s 0.001012321 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 / JaXPipe / cuda / BothRev 0.001022197 s 0.001052257 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / BothRev 0.000990167 s 0.001006657 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PreRev 0.001010744 s 0.001039617 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PostRev 0.001038647 s 0.001060865 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / BothRev 0.001010585 s 0.001040353 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PreRev 0.001027224 s 0.00104845 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PostRev 0.000976761 s 0.000998721 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / BothRev 0.001027353 s 0.001049537 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PreRev 0.001025719 s 0.001050498 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PostRev 0.000975929 s 0.0009984969999999 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / BothRev 0.001025657 s 0.001050177 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 / DefOpt / cuda / PreRev 0.001021497 s 0.001052257 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / PostRev 0.000962746 s 0.000986753 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 / DefOpt / cuda / BothRev 0.0010229999999999 s 0.001053633 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PreRev 0.0010229999999999 s 0.001054049 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PostRev 0.001022616 s 0.001055809 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / BothRev 0.00102108 s 0.001053793 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / Primal 0.0001267035 s 0.000130254 s 0.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / Primal 0.000128645 s 0.00012370125 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 / HLOOpt / tpu / Primal 0.000155656 s 0.0001591159999999 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / tpu / Primal 0.000136319 s 0.0001310605 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 / IPartOpt / tpu / Primal 0.00013447375 s 0.0001376014999999 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 / DefOpt / tpu / Primal 0.0001496645 s 0.0001453375 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 / tpu / Primal 0.0001545392499999 s 0.00015721375 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 / JaXPipe / tpu / Forward 0.00021875375 s 0.0002137462499999 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 / Jax / tpu / Forward 0.0002610234999999 s 0.0002619775 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 / Forward 0.00021857825 s 0.00022068975 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 / Forward 0.0002165824999999 s 0.00021385625 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 / tpu / Forward 0.00021925425 s 0.0002156565 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 0.0002164332499999 s 0.000217844 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 / IDefOpt / tpu / Forward 0.00021864 s 0.0002160175 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 / JaXPipe / tpu / PreRev 0.0003567835 s 0.0003564565 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 0.0002562102499999 s 0.0002562292499999 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 / BothRev 0.00035678225 s 0.00035727825 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 0.000257847 s 0.00025751975 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 / PreRev 0.0003562055 s 0.00035678825 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 0.000291664 s 0.00029208975 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 0.00035688125 s 0.0003571864999999 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 0.00035812125 s 0.00035779925 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 / PostRev 0.0002742672499999 s 0.00027193675 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 / PartOpt / tpu / BothRev 0.0003581395 s 0.00035747775 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 / PreRev 0.0003566795 s 0.00035749825 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 0.00027288625 s 0.000272677 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 / BothRev 0.000356801 s 0.000357449 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 0.00035925625 s 0.000358871 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 0.0002842239999999 s 0.00028377875 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 0.00035955325 s 0.00035905675 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 0.0003582714999999 s 0.000358575 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 0.0003023117499999 s 0.000301865 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 0.000358022 s 0.0003584744999999 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 0.0024190749999999 s 0.0010052658000859 s 2.41
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 0.003384516 s 0.0009495696001067 s 3.56
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 0.002560232 s 0.0009717128000374 s 2.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.002300899 s 0.0010850222000954 s 2.12
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 0.002352575 s 0.0009566142000949 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 / DefOpt / cpu / Primal 0.002457024 s 0.0009814416000153 s 2.50
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.0025276459999999 s 0.0010132667999641 s 2.49
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 0.005801829 s 0.002265069600071 s 2.56
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 0.0067236859999999 s 0.0024919319998844 s 2.70
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 0.006386108 s 0.0022515823999128 s 2.84
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 0.006202174 s 0.0022670832000585 s 2.74
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Forward 0.005739278 s 0.0023049661999721 s 2.49
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 0.005849768 s 0.002297788999931 s 2.55
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 0.006001523 s 0.0023780273999364 s 2.52
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 0.009616455 s 0.0058777129999725 s 1.64
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.009480355 s 0.0059237931999632 s 1.60
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 0.010969809 s 0.0058832524001445 s 1.86
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 0.009346845 s 0.0052997615999629 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 / HLOOpt / cpu / PreRev 0.009255245 s 0.0062640774000101 s 1.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.0081781559999999 s 0.0057815731999653 s 1.41
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.008520356 s 0.0052943166000659 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 / PartOpt / cpu / PreRev 0.008282658 s 0.0062454572000206 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 / PartOpt / cpu / PostRev 0.010117756 s 0.0064490196000406 s 1.57
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 0.0085828349999999 s 0.0064472054000361 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 / IPartOpt / cpu / PreRev 0.010311741 s 0.0062578827999459 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 / IPartOpt / cpu / PostRev 0.008832471 s 0.0061021370000162 s 1.45
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / BothRev 0.01047643 s 0.0064745801998469 s 1.62
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PreRev 0.008464449 s 0.0062355684001886 s 1.36
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 0.009738257 s 0.0056638985999597 s 1.72
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 0.009228603 s 0.0056925937999039 s 1.62
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PreRev 0.010738707 s 0.0060477895999611 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 0.008882711 s 0.0065292453999973 s 1.36
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 0.008291957 s 0.0050366497998766 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 / JaXPipe / cpu / Primal 0.002319 s 0.0010052658000859 s 2.31
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 0.001846 s 0.0009495696001067 s 1.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Primal 0.00213 s 0.0009717128000374 s 2.19
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 0.001499 s 0.0010850222000954 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 / IPartOpt / cpu / Primal 0.001608 s 0.0009566142000949 s 1.68
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 0.002059 s 0.0009814416000153 s 2.10
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 0.001679 s 0.0010132667999641 s 1.66
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 0.004762 s 0.002265069600071 s 2.10
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 0.0048 s 0.0024919319998844 s 1.93
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Forward 0.004859 s 0.0022515823999128 s 2.16
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 0.004645 s 0.0022670832000585 s 2.05
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 0.0060069999999999 s 0.0023049661999721 s 2.61
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 0.00739 s 0.002297788999931 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 / IDefOpt / cpu / Forward 0.004429 s 0.0023780273999364 s 1.86
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 0.011866 s 0.0058777129999725 s 2.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.012379 s 0.0059237931999632 s 2.09
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 0.012907 s 0.0058832524001445 s 2.19
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 0.013781 s 0.0052997615999629 s 2.60
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 0.014175 s 0.0062640774000101 s 2.26
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 0.012619 s 0.0057815731999653 s 2.18
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 0.008666 s 0.0052943166000659 s 1.64
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 0.011783 s 0.0062454572000206 s 1.89
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 0.012855 s 0.0064490196000406 s 1.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.008316 s 0.0064472054000361 s 1.29
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 0.009982 s 0.0062578827999459 s 1.60
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 0.009832 s 0.0061021370000162 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 / IPartOpt / cpu / BothRev 0.008102 s 0.0064745801998469 s 1.25
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 0.008805 s 0.0062355684001886 s 1.41
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PostRev 0.009461 s 0.0056638985999597 s 1.67
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 0.008206 s 0.0056925937999039 s 1.44
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 0.007894 s 0.0060477895999611 s 1.31
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PostRev 0.009811 s 0.0065292453999973 s 1.50
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.009831 s 0.0050366497998766 s 1.95
scatter_sum / JaXPipe / cpu / Primal 0.000007789280007273191 s 0.000008018999978958164 s 0.97
scatter_sum / Jax / cpu / Primal 0.000007616299990331754 s 0.00000769448001847195 s 0.99
scatter_sum / HLOOpt / cpu / Primal 0.000007846079979572096 s 0.00000810788002127083 s 0.97
scatter_sum / PartOpt / cpu / Primal 0.000007591399999000714 s 0.000007375979985226877 s 1.03
scatter_sum / IPartOpt / cpu / Primal 0.00000832594000712561 s 0.000007563760000266484 s 1.10
scatter_sum / DefOpt / cpu / Primal 0.000007815040025889176 s 0.000007662380012334324 s 1.02
scatter_sum / IDefOpt / cpu / Primal 0.000007822699981261394 s 0.000007425240000884514 s 1.05
scatter_sum / JaXPipe / cpu / Forward 0.00001129542001763184 s 0.000012433319998308434 s 0.91
scatter_sum / Jax / cpu / Forward 0.000011392220012567122 s 0.00001258543996300432 s 0.91
scatter_sum / HLOOpt / cpu / Forward 0.00001185095998152974 s 0.000011732860002666712 s 1.01
scatter_sum / PartOpt / cpu / Forward 0.000011521120004545082 s 0.000011961240006712615 s 0.96
scatter_sum / IPartOpt / cpu / Forward 0.000011830499988718658 s 0.00001255495998520928 s 0.94
scatter_sum / DefOpt / cpu / Forward 0.000011214259984626551 s 0.000012197400028526316 s 0.92
scatter_sum / IDefOpt / cpu / Forward 0.00001120382002227416 s 0.000012331140014794071 s 0.91
scatter_sum / JaXPipe / cpu / PreRev 0.000012237460014148384 s 0.000012180780004200642 s 1.00
scatter_sum / JaXPipe / cpu / PostRev 0.000011556760018720524 s 0.00001229112003784394 s 0.94
scatter_sum / JaXPipe / cpu / BothRev 0.000011359500003891298 s 0.000013103359979140804 s 0.87
scatter_sum / Jax / cpu / BothRev 0.000011503380028443643 s 0.00001172898000731948 s 0.98
scatter_sum / HLOOpt / cpu / PreRev 0.000012443680006981594 s 0.000012049320021105814 s 1.03
scatter_sum / HLOOpt / cpu / PostRev 0.000013550839976232964 s 0.00001422884004568914 s 0.95
scatter_sum / HLOOpt / cpu / BothRev 0.00001109640002141532 s 0.000012154379992352916 s 0.91
scatter_sum / PartOpt / cpu / PreRev 0.000011458500039225327 s 0.00001176464003037836 s 0.97
scatter_sum / PartOpt / cpu / PostRev 0.000011841359992104117 s 0.000011788259989771177 s 1.00
scatter_sum / PartOpt / cpu / BothRev 0.00001200622005853802 s 0.000012802080000255956 s 0.94
scatter_sum / IPartOpt / cpu / PreRev 0.000011933680025322246 s 0.000011723319967131828 s 1.02
scatter_sum / IPartOpt / cpu / PostRev 0.000011097720016550738 s 0.000011865039996337143 s 0.94
scatter_sum / IPartOpt / cpu / BothRev 0.000011474660022940952 s 0.00001212475997817819 s 0.95
scatter_sum / DefOpt / cpu / PreRev 0.000011712100003933302 s 0.000011813040018751053 s 0.99
scatter_sum / DefOpt / cpu / PostRev 0.000011233039995204308 s 0.000012441740018402923 s 0.90
scatter_sum / DefOpt / cpu / BothRev 0.000011386259966457148 s 0.000012038839995511808 s 0.95
scatter_sum / IDefOpt / cpu / PreRev 0.0000117632000001322 s 0.00001186413993309543 s 0.99
scatter_sum / IDefOpt / cpu / PostRev 0.000011333800030115526 s 0.000012166299975433502 s 0.93
scatter_sum / IDefOpt / cpu / BothRev 0.00001147314001173072 s 0.000012370740032565663 s 0.93
scatter_sum / JaXPipe / cuda / Primal 0.000010048 s 0.000010432 s 0.96
scatter_sum / Jax / cuda / Primal 0.000009792 s 0.000010976 s 0.89
scatter_sum / HLOOpt / cuda / Primal 0.000009919 s 0.00001088 s 0.91
scatter_sum / PartOpt / cuda / Primal 0.000009728 s 0.000010752 s 0.90
scatter_sum / IPartOpt / cuda / Primal 0.000009728 s 0.000010623 s 0.92
scatter_sum / DefOpt / cuda / Primal 0.000010144 s 0.000010656 s 0.95
scatter_sum / IDefOpt / cuda / Primal 0.000009919 s 0.000010688 s 0.93
scatter_sum / JaXPipe / cuda / Forward 0.00001712 s 0.000017184 s 1.00
scatter_sum / Jax / cuda / Forward 0.000016927000000000002 s 0.000016833 s 1.01
scatter_sum / HLOOpt / cuda / Forward 0.000016896000000000002 s 0.000018016 s 0.94
scatter_sum / PartOpt / cuda / Forward 0.000016832 s 0.000017408 s 0.97
scatter_sum / IPartOpt / cuda / Forward 0.000016927999999999998 s 0.000017696 s 0.96
scatter_sum / DefOpt / cuda / Forward 0.000017088 s 0.000018112 s 0.94
scatter_sum / IDefOpt / cuda / Forward 0.000017216 s 0.000017344 s 0.99
scatter_sum / JaXPipe / cuda / PreRev 0.00001872 s 0.000017184 s 1.09
scatter_sum / JaXPipe / cuda / PostRev 0.000017152 s 0.000018016 s 0.95
scatter_sum / JaXPipe / cuda / BothRev 0.000019168 s 0.000017216 s 1.11
scatter_sum / Jax / cuda / BothRev 0.000017408 s 0.00001712 s 1.02
scatter_sum / HLOOpt / cuda / PreRev 0.000019103 s 0.000017152 s 1.11
scatter_sum / HLOOpt / cuda / PostRev 0.000016896000000000002 s 0.0000168 s 1.01
scatter_sum / HLOOpt / cuda / BothRev 0.000016993 s 0.000017312 s 0.98
scatter_sum / PartOpt / cuda / PreRev 0.000017375999999999998 s 0.000017632 s 0.99
scatter_sum / PartOpt / cuda / PostRev 0.000016768000000000003 s 0.000017216 s 0.97
scatter_sum / PartOpt / cuda / BothRev 0.000017567 s 0.000017695 s 0.99
scatter_sum / IPartOpt / cuda / PreRev 0.000017056 s 0.000017632 s 0.97
scatter_sum / IPartOpt / cuda / PostRev 0.00001664 s 0.000017184 s 0.97
scatter_sum / IPartOpt / cuda / BothRev 0.000016831 s 0.000017408 s 0.97
scatter_sum / DefOpt / cuda / PreRev 0.000017471 s 0.000017888000000000002 s 0.98
scatter_sum / DefOpt / cuda / PostRev 0.000016607 s 0.000016992 s 0.98
scatter_sum / DefOpt / cuda / BothRev 0.000016864 s 0.000026785 s 0.63
scatter_sum / IDefOpt / cuda / PreRev 0.000017023 s 0.000017344 s 0.98
scatter_sum / IDefOpt / cuda / PostRev 0.000017344 s 0.000017728 s 0.98
scatter_sum / IDefOpt / cuda / BothRev 0.000017152 s 0.00001744 s 0.98
scatter_sum / JaXPipe / tpu / Primal 0.000001350425 s 0.0000013505000000000002 s 1.00
scatter_sum / Jax / tpu / Primal 0.000001343075 s 0.000001342925 s 1.00
scatter_sum / HLOOpt / tpu / Primal 0.00000135055 s 0.0000013506000000000002 s 1.00
scatter_sum / PartOpt / tpu / Primal 0.000001342575 s 0.0000013431 s 1.00
scatter_sum / IPartOpt / tpu / Primal 0.00000135045 s 0.0000013503999999999998 s 1.00
scatter_sum / DefOpt / tpu / Primal 0.00000134315 s 0.0000013433999999999995 s 1.00
scatter_sum / IDefOpt / tpu / Primal 0.00000135 s 0.000001350375 s 1.00
scatter_sum / JaXPipe / tpu / Forward 0.00000268745 s 0.000002680175000000001 s 1.00
scatter_sum / Jax / tpu / Forward 0.000002732075 s 0.00000272515 s 1.00
scatter_sum / HLOOpt / tpu / Forward 0.0000026907 s 0.0000026813 s 1.00
scatter_sum / PartOpt / tpu / Forward 0.0000026926 s 0.0000026928000000000003 s 1.00
scatter_sum / IPartOpt / tpu / Forward 0.0000026798 s 0.000002680775 s 1.00
scatter_sum / DefOpt / tpu / Forward 0.0000026927250000000005 s 0.000002692225 s 1.00
scatter_sum / IDefOpt / tpu / Forward 0.000002683425 s 0.00000268125 s 1.00
scatter_sum / JaXPipe / tpu / PreRev 0.0000026924 s 0.00000268965 s 1.00
scatter_sum / JaXPipe / tpu / PostRev 0.000002681875 s 0.000002676675 s 1.00
scatter_sum / JaXPipe / tpu / BothRev 0.00000270325 s 0.0000027042250000000004 s 1.00
scatter_sum / Jax / tpu / BothRev 0.00000273905 s 0.0000027312500000000003 s 1.00
scatter_sum / HLOOpt / tpu / PreRev 0.0000027022 s 0.0000027101750000000003 s 1.00
scatter_sum / HLOOpt / tpu / PostRev 0.00000273205 s 0.0000027377 s 1.00
scatter_sum / HLOOpt / tpu / BothRev 0.0000027026 s 0.000002704725 s 1.00
scatter_sum / PartOpt / tpu / PreRev 0.0000027331 s 0.00000273555 s 1.00
scatter_sum / PartOpt / tpu / PostRev 0.0000027028 s 0.0000027078999999999995 s 1.00
scatter_sum / PartOpt / tpu / BothRev 0.00000273695 s 0.000002740675 s 1.00
scatter_sum / IPartOpt / tpu / PreRev 0.000002705675 s 0.00000270255 s 1.00
scatter_sum / IPartOpt / tpu / PostRev 0.0000027374 s 0.000002733025 s 1.00
scatter_sum / IPartOpt / tpu / BothRev 0.000002708025 s 0.00000270935 s 1.00
scatter_sum / DefOpt / tpu / PreRev 0.000002734375 s 0.000002740775 s 1.00
scatter_sum / DefOpt / tpu / PostRev 0.000002705725 s 0.000002711725 s 1.00
scatter_sum / DefOpt / tpu / BothRev 0.000002744425 s 0.0000027374500000000004 s 1.00
scatter_sum / IDefOpt / tpu / PreRev 0.00000270855 s 0.0000027077250000000004 s 1.00
scatter_sum / IDefOpt / tpu / PostRev 0.00000273935 s 0.0000027391500000000003 s 1.00
scatter_sum / IDefOpt / tpu / BothRev 0.000002707425 s 0.0000027092500000000004 s 1.00
scatter_sum / JaXPipe / cpu / Primal 0.000015841 s 0.000008018999978958164 s 1.98
scatter_sum / Jax / cpu / Primal 0.000015295 s 0.00000769448001847195 s 1.99
scatter_sum / HLOOpt / cpu / Primal 0.00001587 s 0.00000810788002127083 s 1.96
scatter_sum / PartOpt / cpu / Primal 0.000015587000000000002 s 0.000007375979985226877 s 2.11
scatter_sum / IPartOpt / cpu / Primal 0.000015339 s 0.000007563760000266484 s 2.03
scatter_sum / DefOpt / cpu / Primal 0.000015679 s 0.000007662380012334324 s 2.05
scatter_sum / IDefOpt / cpu / Primal 0.000015477 s 0.000007425240000884514 s 2.08
scatter_sum / JaXPipe / cpu / Forward 0.000022434 s 0.000012433319998308434 s 1.80
scatter_sum / Jax / cpu / Forward 0.000022327 s 0.00001258543996300432 s 1.77
scatter_sum / HLOOpt / cpu / Forward 0.000022259 s 0.000011732860002666712 s 1.90
scatter_sum / PartOpt / cpu / Forward 0.00002367 s 0.000011961240006712615 s 1.98
scatter_sum / IPartOpt / cpu / Forward 0.000023797 s 0.00001255495998520928 s 1.90
scatter_sum / DefOpt / cpu / Forward 0.000022549 s 0.000012197400028526316 s 1.85
scatter_sum / IDefOpt / cpu / Forward 0.000022885 s 0.000012331140014794071 s 1.86
scatter_sum / JaXPipe / cpu / PreRev 0.000022612 s 0.000012180780004200642 s 1.86
scatter_sum / JaXPipe / cpu / PostRev 0.000023674 s 0.00001229112003784394 s 1.93
scatter_sum / JaXPipe / cpu / BothRev 0.000023323 s 0.000013103359979140804 s 1.78
scatter_sum / Jax / cpu / BothRev 0.000024017 s 0.00001172898000731948 s 2.05
scatter_sum / HLOOpt / cpu / PreRev 0.000024544 s 0.000012049320021105814 s 2.04
scatter_sum / HLOOpt / cpu / PostRev 0.00002403 s 0.00001422884004568914 s 1.69
scatter_sum / HLOOpt / cpu / BothRev 0.000023798 s 0.000012154379992352916 s 1.96
scatter_sum / PartOpt / cpu / PreRev 0.000025194 s 0.00001176464003037836 s 2.14
scatter_sum / PartOpt / cpu / PostRev 0.000024095 s 0.000011788259989771177 s 2.04
scatter_sum / PartOpt / cpu / BothRev 0.000023641 s 0.000012802080000255956 s 1.85
scatter_sum / IPartOpt / cpu / PreRev 0.000024628 s 0.000011723319967131828 s 2.10
scatter_sum / IPartOpt / cpu / PostRev 0.000024405 s 0.000011865039996337143 s 2.06
scatter_sum / IPartOpt / cpu / BothRev 0.000024408 s 0.00001212475997817819 s 2.01
scatter_sum / DefOpt / cpu / PreRev 0.000023808 s 0.000011813040018751053 s 2.02
scatter_sum / DefOpt / cpu / PostRev 0.000022932 s 0.000012441740018402923 s 1.84
scatter_sum / DefOpt / cpu / BothRev 0.000022826 s 0.000012038839995511808 s 1.90
scatter_sum / IDefOpt / cpu / PreRev 0.000022334 s 0.00001186413993309543 s 1.88
scatter_sum / IDefOpt / cpu / PostRev 0.000022974 s 0.000012166299975433502 s 1.89
scatter_sum / IDefOpt / cpu / BothRev 0.000022403 s 0.000012370740032565663 s 1.81
scatter_sum / JaXPipe / cpu / Primal 0.00001 s 0.000008018999978958164 s 1.25
scatter_sum / Jax / cpu / Primal 0.000011 s 0.00000769448001847195 s 1.43
scatter_sum / HLOOpt / cpu / Primal 0.00001 s 0.00000810788002127083 s 1.23
scatter_sum / PartOpt / cpu / Primal 0.00001 s 0.000007375979985226877 s 1.36
scatter_sum / IPartOpt / cpu / Primal 0.000011 s 0.000007563760000266484 s 1.45
scatter_sum / DefOpt / cpu / Primal 0.00001 s 0.000007662380012334324 s 1.31
scatter_sum / IDefOpt / cpu / Primal 0.00001 s 0.000007425240000884514 s 1.35
scatter_sum / JaXPipe / cpu / Forward 0.000015 s 0.000012433319998308434 s 1.21
scatter_sum / Jax / cpu / Forward 0.000015 s 0.00001258543996300432 s 1.19
scatter_sum / HLOOpt / cpu / Forward 0.000015 s 0.000011732860002666712 s 1.28
scatter_sum / PartOpt / cpu / Forward 0.000015 s 0.000011961240006712615 s 1.25
scatter_sum / IPartOpt / cpu / Forward 0.000015 s 0.00001255495998520928 s 1.19
scatter_sum / DefOpt / cpu / Forward 0.000015 s 0.000012197400028526316 s 1.23
scatter_sum / IDefOpt / cpu / Forward 0.000016 s 0.000012331140014794071 s 1.30
scatter_sum / JaXPipe / cpu / PreRev 0.000016 s 0.000012180780004200642 s 1.31
scatter_sum / JaXPipe / cpu / PostRev 0.000016 s 0.00001229112003784394 s 1.30
scatter_sum / JaXPipe / cpu / BothRev 0.000016 s 0.000013103359979140804 s 1.22
scatter_sum / Jax / cpu / BothRev 0.000016 s 0.00001172898000731948 s 1.36
scatter_sum / HLOOpt / cpu / PreRev 0.000016 s 0.000012049320021105814 s 1.33
scatter_sum / HLOOpt / cpu / PostRev 0.000016 s 0.00001422884004568914 s 1.12
scatter_sum / HLOOpt / cpu / BothRev 0.000017 s 0.000012154379992352916 s 1.40
scatter_sum / PartOpt / cpu / PreRev 0.000016 s 0.00001176464003037836 s 1.36
scatter_sum / PartOpt / cpu / PostRev 0.000016 s 0.000011788259989771177 s 1.36
scatter_sum / PartOpt / cpu / BothRev 0.000016 s 0.000012802080000255956 s 1.25
scatter_sum / IPartOpt / cpu / PreRev 0.000016 s 0.000011723319967131828 s 1.36
scatter_sum / IPartOpt / cpu / PostRev 0.000016 s 0.000011865039996337143 s 1.35
scatter_sum / IPartOpt / cpu / BothRev 0.000016 s 0.00001212475997817819 s 1.32
scatter_sum / DefOpt / cpu / PreRev 0.000017 s 0.000011813040018751053 s 1.44
scatter_sum / DefOpt / cpu / PostRev 0.000016 s 0.000012441740018402923 s 1.29
scatter_sum / DefOpt / cpu / BothRev 0.000016 s 0.000012038839995511808 s 1.33
scatter_sum / IDefOpt / cpu / PreRev 0.000016 s 0.00001186413993309543 s 1.35
scatter_sum / IDefOpt / cpu / PostRev 0.000015 s 0.000012166299975433502 s 1.23
scatter_sum / IDefOpt / cpu / BothRev 0.000016 s 0.000012370740032565663 s 1.29
slicing / JaXPipe / cpu / Primal 0.000006185459997141152 s 0.000006647480022365926 s 0.93
slicing / Jax / cpu / Primal 0.0000066929799959325465 s 0.000006171579980218667 s 1.08
slicing / HLOOpt / cpu / Primal 0.000006423379973057308 s 0.000006497539980045985 s 0.99
slicing / PartOpt / cpu / Primal 0.000006232240029930836 s 0.000006176560018502641 s 1.01
slicing / IPartOpt / cpu / Primal 0.000006688540033792378 s 0.000006291220015555154 s 1.06
slicing / DefOpt / cpu / Primal 0.000006234780003069318 s 0.00000650018000669661 s 0.96
slicing / IDefOpt / cpu / Primal 0.0000062502600121661086 s 0.000006063379996703588 s 1.03
slicing / JaXPipe / cpu / Forward 0.000009421519953320967 s 0.000009319760056314409 s 1.01
slicing / Jax / cpu / Forward 0.000009333240041087264 s 0.000009332700028608088 s 1.00
slicing / HLOOpt / cpu / Forward 0.000009636639961172475 s 0.000009878240007310524 s 0.98
slicing / PartOpt / cpu / Forward 0.0000090460200135567 s 0.000009015859986902796 s 1.00
slicing / IPartOpt / cpu / Forward 0.000009868680008366937 s 0.000009798320033951314 s 1.01
slicing / DefOpt / cpu / Forward 0.000009215860009135211 s 0.000009734220002428629 s 0.95
slicing / IDefOpt / cpu / Forward 0.000009019780009111855 s 0.000009216180023940978 s 0.98
slicing / JaXPipe / cpu / PreRev 0.00001009476000945142 s 0.000010047279993159463 s 1.00
slicing / JaXPipe / cpu / PostRev 0.000010073319981529491 s 0.00001030494000588078 s 0.98
slicing / JaXPipe / cpu / BothRev 0.0000101426800210902 s 0.000010542600030021277 s 0.96
slicing / Jax / cpu / BothRev 0.0000101336200532387 s 0.000009942140031853342 s 1.02
slicing / HLOOpt / cpu / PreRev 0.000010461660031069186 s 0.00001010100000712555 s 1.04
slicing / HLOOpt / cpu / PostRev 0.000014899200014042435 s 0.000014954380012568436 s 1.00
slicing / HLOOpt / cpu / BothRev 0.000009948699962478712 s 0.000009744339995449992 s 1.02
slicing / PartOpt / cpu / PreRev 0.000009845639997365652 s 0.000009714440002426272 s 1.01
slicing / PartOpt / cpu / PostRev 0.000009985519982365076 s 0.000009923720026563388 s 1.01
slicing / PartOpt / cpu / BothRev 0.000010071300057461483 s 0.000010610080007609213 s 0.95
slicing / IPartOpt / cpu / PreRev 0.000009930240030371352 s 0.000010112460013260717 s 0.98
slicing / IPartOpt / cpu / PostRev 0.000009908920073939951 s 0.000009752199985086916 s 1.02
slicing / IPartOpt / cpu / BothRev 0.000009880699954010195 s 0.000009738460057633349 s 1.01
slicing / DefOpt / cpu / PreRev 0.000009563039984641364 s 0.000010649459991327603 s 0.90
slicing / DefOpt / cpu / PostRev 0.000010195860022577108 s 0.000010468679993209664 s 0.97
slicing / DefOpt / cpu / BothRev 0.000009879740018732265 s 0.000010292839997418925 s 0.96
slicing / IDefOpt / cpu / PreRev 0.000009764200012796209 s 0.000009613340007490478 s 1.02
slicing / IDefOpt / cpu / PostRev 0.0000099122199753765 s 0.000010384119996160734 s 0.95
slicing / IDefOpt / cpu / BothRev 0.000010055480015580542 s 0.000010011159984060214 s 1.00
slicing / JaXPipe / cuda / Primal 0.000001887 s 0.000002271 s 0.83
slicing / Jax / cuda / Primal 0.000001887 s 0.000002271 s 0.83
slicing / HLOOpt / cuda / Primal 0.000001887 s 0.000002271 s 0.83
slicing / PartOpt / cuda / Primal 0.000001888 s 0.000002271 s 0.83
slicing / IPartOpt / cuda / Primal 0.000001887 s 0.000002271 s 0.83
slicing / DefOpt / cuda / Primal 0.000001887 s 0.000002272 s 0.83
slicing / IDefOpt / cuda / Primal 0.000001887 s 0.000002271 s 0.83
slicing / JaXPipe / cuda / Forward 0.000009856 s 0.000010496 s 0.94
slicing / Jax / cuda / Forward 0.000010016 s 0.000010208 s 0.98
slicing / HLOOpt / cuda / Forward 0.000009984 s 0.000010592 s 0.94
slicing / PartOpt / cuda / Forward 0.000009888 s 0.000010272 s 0.96
slicing / IPartOpt / cuda / Forward 0.000009664 s 0.000010368 s 0.93
slicing / DefOpt / cuda / Forward 0.000009984 s 0.000010112 s 0.99
slicing / IDefOpt / cuda / Forward 0.00000976 s 0.0000104 s 0.94
slicing / JaXPipe / cuda / PreRev 0.0000096 s 0.000010208 s 0.94
slicing / JaXPipe / cuda / PostRev 0.000009408 s 0.00001024 s 0.92
slicing / JaXPipe / cuda / BothRev 0.00000976 s 0.000010592 s 0.92
slicing / Jax / cuda / BothRev 0.000009536 s 0.000010847 s 0.88
slicing / HLOOpt / cuda / PreRev 0.0000096 s 0.000010752 s 0.89
slicing / HLOOpt / cuda / PostRev 0.000009568 s 0.000010464 s 0.91
slicing / HLOOpt / cuda / BothRev 0.00000976 s 0.000010368 s 0.94
slicing / PartOpt / cuda / PreRev 0.000010209 s 0.000010304 s 0.99
slicing / PartOpt / cuda / PostRev 0.000009984 s 0.000010304 s 0.97
slicing / PartOpt / cuda / BothRev 0.000009952 s 0.000010913 s 0.91
slicing / IPartOpt / cuda / PreRev 0.000010752 s 0.000010561 s 1.02
slicing / IPartOpt / cuda / PostRev 0.000009856 s 0.000010495 s 0.94
slicing / IPartOpt / cuda / BothRev 0.00000992 s 0.000010112 s 0.98
slicing / DefOpt / cuda / PreRev 0.000009984 s 0.000010528 s 0.95
slicing / DefOpt / cuda / PostRev 0.00000992 s 0.000010497 s 0.95
slicing / DefOpt / cuda / BothRev 0.000009695 s 0.0000104 s 0.93
slicing / IDefOpt / cuda / PreRev 0.000009952 s 0.0000104 s 0.96
slicing / IDefOpt / cuda / PostRev 0.000009376 s 0.000010368 s 0.90
slicing / IDefOpt / cuda / BothRev 0.000009824 s 0.00001056 s 0.93
slicing / JaXPipe / tpu / Primal 9.63025e-7 s 9.66075e-7 s 1.00
slicing / Jax / tpu / Primal 9.63875e-7 s 9.6445e-7 s 1.00
slicing / HLOOpt / tpu / Primal 9.6485e-7 s 9.6335e-7 s 1.00
slicing / PartOpt / tpu / Primal 9.63275e-7 s 9.6055e-7 s 1.00
slicing / IPartOpt / tpu / Primal 9.77175e-7 s 9.6725e-7 s 1.01
slicing / DefOpt / tpu / Primal 9.68575e-7 s 9.62375e-7 s 1.01
slicing / IDefOpt / tpu / Primal 9.6435e-7 s 9.6735e-7 s 1.00
slicing / JaXPipe / tpu / Forward 0.0000014112250000000002 s 0.000001413625 s 1.00
slicing / Jax / tpu / Forward 0.00000142375 s 0.00000142315 s 1.00
slicing / HLOOpt / tpu / Forward 0.0000015221000000000002 s 0.000001520225 s 1.00
slicing / PartOpt / tpu / Forward 0.000001433725 s 0.000001432675 s 1.00
slicing / IPartOpt / tpu / Forward 0.0000015236 s 0.00000152765 s 1.00
slicing / DefOpt / tpu / Forward 0.000001441225 s 0.0000014381749999999998 s 1.00
slicing / IDefOpt / tpu / Forward 0.000001522025 s 0.0000015279 s 1.00
slicing / JaXPipe / tpu / PreRev 0.000002349725 s 0.000002346325 s 1.00
slicing / JaXPipe / tpu / PostRev 0.000002502825 s 0.000002544875 s 0.98
slicing / JaXPipe / tpu / BothRev 0.00000235865 s 0.00000236955 s 1.00
slicing / Jax / tpu / BothRev 0.0000025204 s 0.0000025438 s 0.99
slicing / HLOOpt / tpu / PreRev 0.0000023631 s 0.000002364175 s 1.00
slicing / HLOOpt / tpu / PostRev 0.000002530075 s 0.000002526 s 1.00
slicing / HLOOpt / tpu / BothRev 0.00000234825 s 0.0000023627000000000003 s 0.99
slicing / PartOpt / tpu / PreRev 0.000002534525 s 0.000002522925 s 1.00
slicing / PartOpt / tpu / PostRev 0.0000023715 s 0.0000023605 s 1.00
slicing / PartOpt / tpu / BothRev 0.000002531075 s 0.000002522375 s 1.00
slicing / IPartOpt / tpu / PreRev 0.00000236565 s 0.0000023649 s 1.00
slicing / IPartOpt / tpu / PostRev 0.000002531925 s 0.0000025277500000000005 s 1.00
slicing / IPartOpt / tpu / BothRev 0.0000023586 s 0.000002362175 s 1.00
slicing / DefOpt / tpu / PreRev 0.000002526025 s 0.00000252065 s 1.00
slicing / DefOpt / tpu / PostRev 0.00000235825 s 0.000002350225 s 1.00
slicing / DefOpt / tpu / BothRev 0.0000025251750000000003 s 0.00000252695 s 1.00
slicing / IDefOpt / tpu / PreRev 0.00000236095 s 0.000002355775 s 1.00
slicing / IDefOpt / tpu / PostRev 0.000002532725 s 0.0000025324 s 1.00
slicing / IDefOpt / tpu / BothRev 0.000002350125 s 0.000002359675 s 1.00
slicing / JaXPipe / cpu / Primal 0.000017386 s 0.000006647480022365926 s 2.62
slicing / Jax / cpu / Primal 0.000012611 s 0.000006171579980218667 s 2.04
slicing / HLOOpt / cpu / Primal 0.000012682 s 0.000006497539980045985 s 1.95
slicing / PartOpt / cpu / Primal 0.000012661 s 0.000006176560018502641 s 2.05
slicing / IPartOpt / cpu / Primal 0.000012707 s 0.000006291220015555154 s 2.02
slicing / DefOpt / cpu / Primal 0.000012687 s 0.00000650018000669661 s 1.95
slicing / IDefOpt / cpu / Primal 0.000012699 s 0.000006063379996703588 s 2.09
slicing / JaXPipe / cpu / Forward 0.000017007 s 0.000009319760056314409 s 1.82
slicing / Jax / cpu / Forward 0.000016859 s 0.000009332700028608088 s 1.81
slicing / HLOOpt / cpu / Forward 0.000017062 s 0.000009878240007310524 s 1.73
slicing / PartOpt / cpu / Forward 0.000016802 s 0.000009015859986902796 s 1.86
slicing / IPartOpt / cpu / Forward 0.000016845 s 0.000009798320033951314 s 1.72
slicing / DefOpt / cpu / Forward 0.000016759 s 0.000009734220002428629 s 1.72
slicing / IDefOpt / cpu / Forward 0.000016999 s 0.000009216180023940978 s 1.84
slicing / JaXPipe / cpu / PreRev 0.000018326 s 0.000010047279993159463 s 1.82
slicing / JaXPipe / cpu / PostRev 0.000017912 s 0.00001030494000588078 s 1.74
slicing / JaXPipe / cpu / BothRev 0.000017483 s 0.000010542600030021277 s 1.66
slicing / Jax / cpu / BothRev 0.00001771 s 0.000009942140031853342 s 1.78
slicing / HLOOpt / cpu / PreRev 0.000018010000000000002 s 0.00001010100000712555 s 1.78
slicing / HLOOpt / cpu / PostRev 0.000018039 s 0.000014954380012568436 s 1.21
slicing / HLOOpt / cpu / BothRev 0.000018484 s 0.000009744339995449992 s 1.90
slicing / PartOpt / cpu / PreRev 0.000018301 s 0.000009714440002426272 s 1.88
slicing / PartOpt / cpu / PostRev 0.00001768 s 0.000009923720026563388 s 1.78
slicing / PartOpt / cpu / BothRev 0.000017749000000000002 s 0.000010610080007609213 s 1.67
slicing / IPartOpt / cpu / PreRev 0.000018216 s 0.000010112460013260717 s 1.80
slicing / IPartOpt / cpu / PostRev 0.000017690000000000002 s 0.000009752199985086916 s 1.81
slicing / IPartOpt / cpu / BothRev 0.000018351 s 0.000009738460057633349 s 1.88
slicing / DefOpt / cpu / PreRev 0.0000182 s 0.000010649459991327603 s 1.71
slicing / DefOpt / cpu / PostRev 0.000017828 s 0.000010468679993209664 s 1.70
slicing / DefOpt / cpu / BothRev 0.00001738 s 0.000010292839997418925 s 1.69
slicing / IDefOpt / cpu / PreRev 0.000017517999999999997 s 0.000009613340007490478 s 1.82
slicing / IDefOpt / cpu / PostRev 0.000017743 s 0.000010384119996160734 s 1.71
slicing / IDefOpt / cpu / BothRev 0.000017416 s 0.000010011159984060214 s 1.74
slicing / JaXPipe / cpu / Primal 0.000008 s 0.000006647480022365926 s 1.20
slicing / Jax / cpu / Primal 0.000008 s 0.000006171579980218667 s 1.30
slicing / HLOOpt / cpu / Primal 0.000008 s 0.000006497539980045985 s 1.23
slicing / PartOpt / cpu / Primal 0.000008 s 0.000006176560018502641 s 1.30
slicing / IPartOpt / cpu / Primal 0.000008 s 0.000006291220015555154 s 1.27
slicing / DefOpt / cpu / Primal 0.000008 s 0.00000650018000669661 s 1.23
slicing / IDefOpt / cpu / Primal 0.000008 s 0.000006063379996703588 s 1.32
slicing / JaXPipe / cpu / Forward 0.000011 s 0.000009319760056314409 s 1.18
slicing / Jax / cpu / Forward 0.000011 s 0.000009332700028608088 s 1.18
slicing / HLOOpt / cpu / Forward 0.000011 s 0.000009878240007310524 s 1.11
slicing / PartOpt / cpu / Forward 0.000011 s 0.000009015859986902796 s 1.22
slicing / IPartOpt / cpu / Forward 0.000011 s 0.000009798320033951314 s 1.12
slicing / DefOpt / cpu / Forward 0.000011 s 0.000009734220002428629 s 1.13
slicing / IDefOpt / cpu / Forward 0.000011 s 0.000009216180023940978 s 1.19
slicing / JaXPipe / cpu / PreRev 0.000012 s 0.000010047279993159463 s 1.19
slicing / JaXPipe / cpu / PostRev 0.000012 s 0.00001030494000588078 s 1.16
slicing / JaXPipe / cpu / BothRev 0.000011 s 0.000010542600030021277 s 1.04
slicing / Jax / cpu / BothRev 0.000011 s 0.000009942140031853342 s 1.11
slicing / HLOOpt / cpu / PreRev 0.000012 s 0.00001010100000712555 s 1.19
slicing / HLOOpt / cpu / PostRev 0.000012 s 0.000014954380012568436 s 0.80
slicing / HLOOpt / cpu / BothRev 0.000011 s 0.000009744339995449992 s 1.13
slicing / PartOpt / cpu / PreRev 0.000012 s 0.000009714440002426272 s 1.24
slicing / PartOpt / cpu / PostRev 0.000012 s 0.000009923720026563388 s 1.21
slicing / PartOpt / cpu / BothRev 0.000012 s 0.000010610080007609213 s 1.13
slicing / IPartOpt / cpu / PreRev 0.000012 s 0.000010112460013260717 s 1.19
slicing / IPartOpt / cpu / PostRev 0.000012 s 0.000009752199985086916 s 1.23
slicing / IPartOpt / cpu / BothRev 0.000012 s 0.000009738460057633349 s 1.23
slicing / DefOpt / cpu / PreRev 0.000012 s 0.000010649459991327603 s 1.13
slicing / DefOpt / cpu / PostRev 0.000012 s 0.000010468679993209664 s 1.15
slicing / DefOpt / cpu / BothRev 0.000013 s 0.000010292839997418925 s 1.26
slicing / IDefOpt / cpu / PreRev 0.000012 s 0.000009613340007490478 s 1.25
slicing / IDefOpt / cpu / PostRev 0.000012 s 0.000010384119996160734 s 1.16
slicing / IDefOpt / cpu / BothRev 0.000012 s 0.000010011159984060214 s 1.20
sum / JaXPipe / cpu / Primal 0.000007472820025213877 s 0.000008205499962059547 s 0.91
sum / Jax / cpu / Primal 0.00000742518002880388 s 0.000007885579998401227 s 0.94
sum / HLOOpt / cpu / Primal 0.000007655639992663054 s 0.00000813256000583351 s 0.94
sum / PartOpt / cpu / Primal 0.00000797040001998539 s 0.00000766689997362846 s 1.04
sum / IPartOpt / cpu / Primal 0.000007883640009822556 s 0.000007831460034140036 s 1.01
sum / DefOpt / cpu / Primal 0.000008110500020848122 s 0.000007939400029499666 s 1.02
sum / IDefOpt / cpu / Primal 0.000007503620017814683 s 0.000007458860018232372 s 1.01
sum / JaXPipe / cpu / Forward 0.000011402840000300785 s 0.000011636680019364576 s 0.98
sum / Jax / cpu / Forward 0.00001114341998800228 s 0.000011493720003272755 s 0.97
sum / HLOOpt / cpu / Forward 0.000011387220010874444 s 0.000011685479976222268 s 0.97
sum / PartOpt / cpu / Forward 0.000011281780016361154 s 0.0000112998200529546 s 1.00
sum / IPartOpt / cpu / Forward 0.000011515359965414973 s 0.00001170238003396662 s 0.98
sum / DefOpt / cpu / Forward 0.000010784520036395406 s 0.000011018780005542794 s 0.98
sum / IDefOpt / cpu / Forward 0.000011007079992850776 s 0.000011090939979112591 s 0.99
sum / JaXPipe / cpu / PreRev 0.000010880899999392568 s 0.000011340399978507776 s 0.96
sum / JaXPipe / cpu / PostRev 0.000011065879980378669 s 0.000011195940041943686 s 0.99
sum / JaXPipe / cpu / BothRev 0.00001088690000869974 s 0.000011176199986948632 s 0.97
sum / Jax / cpu / BothRev 0.000010867880018849971 s 0.000011233220002395684 s 0.97
sum / HLOOpt / cpu / PreRev 0.000011239679961363436 s 0.000011018420027539832 s 1.02
sum / HLOOpt / cpu / PostRev 0.00001263182000911911 s 0.000012510959995779558 s 1.01
sum / HLOOpt / cpu / BothRev 0.000010829660022864116 s 0.000010940180018224055 s 0.99
sum / PartOpt / cpu / PreRev 0.000010897100009970018 s 0.000010965639994537923 s 0.99
sum / PartOpt / cpu / PostRev 0.000010608399970806205 s 0.000011055179975301143 s 0.96
sum / PartOpt / cpu / BothRev 0.000011078200013798778 s 0.00001108873997509363 s 1.00
sum / IPartOpt / cpu / PreRev 0.000010701920009523748 s 0.000010772719970191246 s 0.99
sum / IPartOpt / cpu / PostRev 0.00001098503997127409 s 0.00001088931998310727 s 1.01
sum / IPartOpt / cpu / BothRev 0.000010422779978398466 s 0.000010641720000421627 s 0.98
sum / DefOpt / cpu / PreRev 0.000010586239959593513 s 0.000011416340048526764 s 0.93
sum / DefOpt / cpu / PostRev 0.000010914000013144688 s 0.00001111735999074881 s 0.98
sum / DefOpt / cpu / BothRev 0.00001089163999495213 s 0.000011046720010199353 s 0.99
sum / IDefOpt / cpu / PreRev 0.000010979739990943926 s 0.000011240759968131896 s 0.98
sum / IDefOpt / cpu / PostRev 0.000010775500022646156 s 0.000011339099964970955 s 0.95
sum / IDefOpt / cpu / BothRev 0.000010523420014578733 s 0.000011133979987789645 s 0.95
sum / JaXPipe / cuda / Primal 0.000002047 s 0.000002432 s 0.84
sum / Jax / cuda / Primal 0.000002048 s 0.000002433 s 0.84
sum / HLOOpt / cuda / Primal 0.000002048 s 0.000002433 s 0.84
sum / PartOpt / cuda / Primal 0.000002048 s 0.000002432 s 0.84
sum / IPartOpt / cuda / Primal 0.000002048 s 0.000002463 s 0.83
sum / DefOpt / cuda / Primal 0.000002047 s 0.000002432 s 0.84
sum / IDefOpt / cuda / Primal 0.000002047 s 0.000002463 s 0.83
sum / JaXPipe / cuda / Forward 0.000010016 s 0.0000104 s 0.96
sum / Jax / cuda / Forward 0.000010016 s 0.000010496 s 0.95
sum / HLOOpt / cuda / Forward 0.0000096 s 0.00001056 s 0.91
sum / PartOpt / cuda / Forward 0.00001024 s 0.000010528 s 0.97
sum / IPartOpt / cuda / Forward 0.000009503 s 0.000010464 s 0.91
sum / DefOpt / cuda / Forward 0.000010143 s 0.000010464 s 0.97
sum / IDefOpt / cuda / Forward 0.00001008 s 0.00001056 s 0.95
sum / JaXPipe / cuda / PreRev 0.00000928 s 0.0000104 s 0.89
sum / JaXPipe / cuda / PostRev 0.000009536 s 0.000010464 s 0.91
sum / JaXPipe / cuda / BothRev 0.00000976 s 0.000010209 s 0.96
sum / Jax / cuda / BothRev 0.000009535 s 0.000010656 s 0.89
sum / HLOOpt / cuda / PreRev 0.000009856 s 0.000010368 s 0.95
sum / HLOOpt / cuda / PostRev 0.000009248 s 0.00001072 s 0.86
sum / HLOOpt / cuda / BothRev 0.000009536 s 0.000010368 s 0.92
sum / PartOpt / cuda / PreRev 0.000009856 s 0.000010432 s 0.94
sum / PartOpt / cuda / PostRev 0.000009408 s 0.000010528 s 0.89
sum / PartOpt / cuda / BothRev 0.000009503 s 0.000010368 s 0.92
sum / IPartOpt / cuda / PreRev 0.000009536 s 0.000010335 s 0.92
sum / IPartOpt / cuda / PostRev 0.000009856 s 0.000010528 s 0.94
sum / IPartOpt / cuda / BothRev 0.0000096 s 0.0000104 s 0.92
sum / DefOpt / cuda / PreRev 0.000009792 s 0.00001056 s 0.93
sum / DefOpt / cuda / PostRev 0.000009824 s 0.00001088 s 0.90
sum / DefOpt / cuda / BothRev 0.000009728 s 0.0000104 s 0.94
sum / IDefOpt / cuda / PreRev 0.000009888 s 0.000010849 s 0.91
sum / IDefOpt / cuda / PostRev 0.000009887 s 0.000010336 s 0.96
sum / IDefOpt / cuda / BothRev 0.000009184 s 0.00001088 s 0.84
sum / JaXPipe / tpu / Primal 5.0315e-7 s 5.028e-7 s 1.00
sum / Jax / tpu / Primal 5.4755e-7 s 5.47425e-7 s 1.00
sum / HLOOpt / tpu / Primal 5.029e-7 s 5.033750000000001e-7 s 1.00
sum / PartOpt / tpu / Primal 5.4725e-7 s 5.475e-7 s 1.00
sum / IPartOpt / tpu / Primal 5.03125e-7 s 5.0285e-7 s 1.00
sum / DefOpt / tpu / Primal 5.47275e-7 s 5.4715e-7 s 1.00
sum / IDefOpt / tpu / Primal 5.0295e-7 s 5.030749999999999e-7 s 1.00
sum / JaXPipe / tpu / Forward 0.000001549775 s 0.000001557475 s 1.00
sum / Jax / tpu / Forward 0.000001495925 s 0.00000149695 s 1.00
sum / HLOOpt / tpu / Forward 0.000001529275 s 0.00000153305 s 1.00
sum / PartOpt / tpu / Forward 0.000001490275 s 0.000001491625 s 1.00
sum / IPartOpt / tpu / Forward 0.00000153575 s 0.0000015359999999999998 s 1.00
sum / DefOpt / tpu / Forward 0.000001486375 s 0.0000014993500000000003 s 0.99
sum / IDefOpt / tpu / Forward 0.0000015284 s 0.0000015344749999999998 s 1.00
sum / JaXPipe / tpu / PreRev 9.8955e-7 s 9.958500000000002e-7 s 0.99
sum / JaXPipe / tpu / PostRev 0.000001037675 s 0.0000010341999999999998 s 1.00
sum / JaXPipe / tpu / BothRev 9.87575e-7 s 9.93325e-7 s 0.99
sum / Jax / tpu / BothRev 0.00000103135 s 0.0000010313 s 1.00
sum / HLOOpt / tpu / PreRev 9.917e-7 s 9.888e-7 s 1.00
sum / HLOOpt / tpu / PostRev 0.000001031175 s 0.0000010325 s 1.00
sum / HLOOpt / tpu / BothRev 9.8715e-7 s 9.90475e-7 s 1.00
sum / PartOpt / tpu / PreRev 0.000001031025 s 0.000001031675 s 1.00
sum / PartOpt / tpu / PostRev 9.927999999999998e-7 s 9.87275e-7 s 1.01
sum / PartOpt / tpu / BothRev 0.000001037625 s 0.0000010309 s 1.01
sum / IPartOpt / tpu / PreRev 9.86925e-7 s 9.99775e-7 s 0.99
sum / IPartOpt / tpu / PostRev 0.00000103155 s 0.00000103435 s 1.00
sum / IPartOpt / tpu / BothRev 9.879749999999998e-7 s 9.86825e-7 s 1.00
sum / DefOpt / tpu / PreRev 0.00000103085 s 0.000001036775 s 0.99
sum / DefOpt / tpu / PostRev 9.86625e-7 s 9.96775e-7 s 0.99
sum / DefOpt / tpu / BothRev 0.000001035325 s 0.0000010382500000000002 s 1.00
sum / IDefOpt / tpu / PreRev 9.9065e-7 s 9.93e-7 s 1.00
sum / IDefOpt / tpu / PostRev 0.00000103485 s 0.0000010374750000000002 s 1.00
sum / IDefOpt / tpu / BothRev 9.871e-7 s 9.92375e-7 s 0.99
sum / JaXPipe / cpu / Primal 0.000015478 s 0.000008205499962059547 s 1.89
sum / Jax / cpu / Primal 0.000015254 s 0.000007885579998401227 s 1.93
sum / HLOOpt / cpu / Primal 0.000015283 s 0.00000813256000583351 s 1.88
sum / PartOpt / cpu / Primal 0.00001516 s 0.00000766689997362846 s 1.98
sum / IPartOpt / cpu / Primal 0.000015483 s 0.000007831460034140036 s 1.98
sum / DefOpt / cpu / Primal 0.000015682 s 0.000007939400029499666 s 1.98
sum / IDefOpt / cpu / Primal 0.000015743 s 0.000007458860018232372 s 2.11
sum / JaXPipe / cpu / Forward 0.00002066 s 0.000011636680019364576 s 1.78
sum / Jax / cpu / Forward 0.000020273 s 0.000011493720003272755 s 1.76
sum / HLOOpt / cpu / Forward 0.000020852 s 0.000011685479976222268 s 1.78
sum / PartOpt / cpu / Forward 0.000021103 s 0.0000112998200529546 s 1.87
sum / IPartOpt / cpu / Forward 0.000020907 s 0.00001170238003396662 s 1.79
sum / DefOpt / cpu / Forward 0.000020471 s 0.000011018780005542794 s 1.86
sum / IDefOpt / cpu / Forward 0.000020070000000000003 s 0.000011090939979112591 s 1.81
sum / JaXPipe / cpu / PreRev 0.000018662 s 0.000011340399978507776 s 1.65
sum / JaXPipe / cpu / PostRev 0.000018646 s 0.000011195940041943686 s 1.67
sum / JaXPipe / cpu / BothRev 0.000019044 s 0.000011176199986948632 s 1.70
sum / Jax / cpu / BothRev 0.00001888 s 0.000011233220002395684 s 1.68
sum / HLOOpt / cpu / PreRev 0.000018554 s 0.000011018420027539832 s 1.68
sum / HLOOpt / cpu / PostRev 0.000019037 s 0.000012510959995779558 s 1.52
sum / HLOOpt / cpu / BothRev 0.000019074 s 0.000010940180018224055 s 1.74
sum / PartOpt / cpu / PreRev 0.000019335 s 0.000010965639994537923 s 1.76
sum / PartOpt / cpu / PostRev 0.000018641 s 0.000011055179975301143 s 1.69
sum / PartOpt / cpu / BothRev 0.000019636 s 0.00001108873997509363 s 1.77
sum / IPartOpt / cpu / PreRev 0.000019432000000000003 s 0.000010772719970191246 s 1.80
sum / IPartOpt / cpu / PostRev 0.000018657 s 0.00001088931998310727 s 1.71
sum / IPartOpt / cpu / BothRev 0.000019787 s 0.000010641720000421627 s 1.86
sum / DefOpt / cpu / PreRev 0.000019683 s 0.000011416340048526764 s 1.72
sum / DefOpt / cpu / PostRev 0.000019795 s 0.00001111735999074881 s 1.78
sum / DefOpt / cpu / BothRev 0.000019714 s 0.000011046720010199353 s 1.78
sum / IDefOpt / cpu / PreRev 0.000020124 s 0.000011240759968131896 s 1.79
sum / IDefOpt / cpu / PostRev 0.000019177 s 0.000011339099964970955 s 1.69
sum / IDefOpt / cpu / BothRev 0.000019491 s 0.000011133979987789645 s 1.75
sum / JaXPipe / cpu / Primal 0.00001 s 0.000008205499962059547 s 1.22
sum / Jax / cpu / Primal 0.00001 s 0.000007885579998401227 s 1.27
sum / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.00000813256000583351 s 1.11
sum / PartOpt / cpu / Primal 0.000011 s 0.00000766689997362846 s 1.43
sum / IPartOpt / cpu / Primal 0.00001 s 0.000007831460034140036 s 1.28
sum / DefOpt / cpu / Primal 0.00001 s 0.000007939400029499666 s 1.26
sum / IDefOpt / cpu / Primal 0.00001 s 0.000007458860018232372 s 1.34
sum / JaXPipe / cpu / Forward 0.000014 s 0.000011636680019364576 s 1.20
sum / Jax / cpu / Forward 0.000014 s 0.000011493720003272755 s 1.22
sum / HLOOpt / cpu / Forward 0.000014 s 0.000011685479976222268 s 1.20
sum / PartOpt / cpu / Forward 0.000013 s 0.0000112998200529546 s 1.15
sum / IPartOpt / cpu / Forward 0.000015 s 0.00001170238003396662 s 1.28
sum / DefOpt / cpu / Forward 0.000014 s 0.000011018780005542794 s 1.27
sum / IDefOpt / cpu / Forward 0.000013 s 0.000011090939979112591 s 1.17
sum / JaXPipe / cpu / PreRev 0.000013 s 0.000011340399978507776 s 1.15
sum / JaXPipe / cpu / PostRev 0.000013 s 0.000011195940041943686 s 1.16
sum / JaXPipe / cpu / BothRev 0.000013 s 0.000011176199986948632 s 1.16
sum / Jax / cpu / BothRev 0.000013 s 0.000011233220002395684 s 1.16
sum / HLOOpt / cpu / PreRev 0.000013 s 0.000011018420027539832 s 1.18
sum / HLOOpt / cpu / PostRev 0.000013 s 0.000012510959995779558 s 1.04
sum / HLOOpt / cpu / BothRev 0.000014 s 0.000010940180018224055 s 1.28
sum / PartOpt / cpu / PreRev 0.000013 s 0.000010965639994537923 s 1.19
sum / PartOpt / cpu / PostRev 0.000012 s 0.000011055179975301143 s 1.09
sum / PartOpt / cpu / BothRev 0.000015 s 0.00001108873997509363 s 1.35
sum / IPartOpt / cpu / PreRev 0.000013 s 0.000010772719970191246 s 1.21
sum / IPartOpt / cpu / PostRev 0.000013 s 0.00001088931998310727 s 1.19
sum / IPartOpt / cpu / BothRev 0.000013 s 0.000010641720000421627 s 1.22
sum / DefOpt / cpu / PreRev 0.000013 s 0.000011416340048526764 s 1.14
sum / DefOpt / cpu / PostRev 0.000014 s 0.00001111735999074881 s 1.26
sum / DefOpt / cpu / BothRev 0.000013 s 0.000011046720010199353 s 1.18
sum / IDefOpt / cpu / PreRev 0.000013 s 0.000011240759968131896 s 1.16
sum / IDefOpt / cpu / PostRev 0.000013 s 0.000011339099964970955 s 1.15
sum / IDefOpt / cpu / BothRev 0.000013 s 0.000011133979987789645 s 1.17
value_and_grad / JaXPipe / cpu / Primal 0.000013672579980266164 s 0.000014084919966990128 s 0.97
value_and_grad / Jax / cpu / Primal 0.000013729780011999537 s 0.000013931500006947315 s 0.99
value_and_grad / HLOOpt / cpu / Primal 0.00001340828000138572 s 0.000013500779987225542 s 0.99
value_and_grad / PartOpt / cpu / Primal 0.00001310235998062126 s 0.000013817400040352368 s 0.95
value_and_grad / IPartOpt / cpu / Primal 0.000013024139971093971 s 0.000013663420013472205 s 0.95
value_and_grad / DefOpt / cpu / Primal 0.00001341827999567613 s 0.000014199640008882851 s 0.94
value_and_grad / IDefOpt / cpu / Primal 0.00001336252000328386 s 0.000013077560006422572 s 1.02
value_and_grad / JaXPipe / cuda / Primal 0.000032448 s 0.000035007999999999994 s 0.93
value_and_grad / Jax / cuda / Primal 0.000032736 s 0.000036416 s 0.90
value_and_grad / HLOOpt / cuda / Primal 0.000032512 s 0.000034176 s 0.95
value_and_grad / PartOpt / cuda / Primal 0.000032576 s 0.000033952 s 0.96
value_and_grad / IPartOpt / cuda / Primal 0.000032736 s 0.00003472 s 0.94
value_and_grad / DefOpt / cuda / Primal 0.00003248 s 0.000034016 s 0.95
value_and_grad / IDefOpt / cuda / Primal 0.00003296 s 0.000034401 s 0.96
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 0.000023916 s 0.000014084919966990128 s 1.70
value_and_grad / Jax / cpu / Primal 0.000022856000000000003 s 0.000013931500006947315 s 1.64
value_and_grad / HLOOpt / cpu / Primal 0.000022937 s 0.000013500779987225542 s 1.70
value_and_grad / PartOpt / cpu / Primal 0.000023573 s 0.000013817400040352368 s 1.71
value_and_grad / IPartOpt / cpu / Primal 0.000023033 s 0.000013663420013472205 s 1.69
value_and_grad / DefOpt / cpu / Primal 0.000022979 s 0.000014199640008882851 s 1.62
value_and_grad / IDefOpt / cpu / Primal 0.00002279 s 0.000013077560006422572 s 1.74
value_and_grad / JaXPipe / cpu / Primal 0.000016 s 0.000014084919966990128 s 1.14
value_and_grad / Jax / cpu / Primal 0.000016 s 0.000013931500006947315 s 1.15
value_and_grad / HLOOpt / cpu / Primal 0.000016 s 0.000013500779987225542 s 1.19
value_and_grad / PartOpt / cpu / Primal 0.000016 s 0.000013817400040352368 s 1.16
value_and_grad / IPartOpt / cpu / Primal 0.000016 s 0.000013663420013472205 s 1.17
value_and_grad / DefOpt / cpu / Primal 0.000016 s 0.000014199640008882851 s 1.13
value_and_grad / IDefOpt / cpu / Primal 0.000016 s 0.000013077560006422572 s 1.22
jaxmd20 / JaXPipe / cuda / Primal 0.001433365 s 0.0015257609999999 s 0.94
jaxmd20 / Jax / cuda / Primal 0.001507284 s 0.001498498 s 1.01
jaxmd20 / HLOOpt / cuda / Primal 0.001324598 s 0.001346049 s 0.98
jaxmd20 / PartOpt / cuda / Primal 0.001325622 s 0.001359426 s 0.98
jaxmd20 / IPartOpt / cuda / Primal 0.001357045 s 0.00138253 s 0.98
jaxmd20 / DefOpt / cuda / Primal 0.00095708 s 0.000947521 s 1.01
jaxmd20 / IDefOpt / cuda / Primal 0.000968537 s 0.000964609 s 1.00
jaxmd20 / JaXPipe / cuda / Forward 0.0015710919999999 s 0.001626754 s 0.97
jaxmd20 / Jax / cuda / Forward 0.001796178 s 0.001856516 s 0.97
jaxmd20 / HLOOpt / cuda / Forward 0.001650771 s 0.001767171 s 0.93
jaxmd20 / PartOpt / cuda / Forward 0.001628083 s 0.00170061 s 0.96
jaxmd20 / IPartOpt / cuda / Forward 0.001614356 s 0.001720386 s 0.94
jaxmd20 / DefOpt / cuda / Forward 0.0016333 s 0.00170157 s 0.96
jaxmd20 / IDefOpt / cuda / Forward 0.001688818 s 0.0017180169999999 s 0.98
jaxmd20 / JaXPipe / cuda / PreRev 0.00267998 s 0.002760387 s 0.97
jaxmd20 / JaXPipe / cuda / PostRev 0.005330967 s 0.005442438 s 0.98
jaxmd20 / JaXPipe / cuda / BothRev 0.002678315 s 0.002774595 s 0.97
jaxmd20 / Jax / cuda / BothRev 0.005310807 s 0.005477414 s 0.97
jaxmd20 / HLOOpt / cuda / PreRev 0.002741259 s 0.002843169 s 0.96
jaxmd20 / HLOOpt / cuda / PostRev 0.005291159 s 0.0054804219999999 s 0.97
jaxmd20 / HLOOpt / cuda / BothRev 0.002708845 s 0.002806819 s 0.97
jaxmd20 / PartOpt / cuda / PreRev 0.002797802 s 0.002930979 s 0.95
jaxmd20 / PartOpt / cuda / PostRev 0.005416759 s 0.005581511 s 0.97
jaxmd20 / PartOpt / cuda / BothRev 0.002754251 s 0.002833571 s 0.97
jaxmd20 / IPartOpt / cuda / PreRev 0.002869194 s 0.002931043 s 0.98
jaxmd20 / IPartOpt / cuda / PostRev 0.005371031 s 0.005600198 s 0.96
jaxmd20 / IPartOpt / cuda / BothRev 0.002747435 s 0.002834115 s 0.97
jaxmd20 / DefOpt / cuda / PreRev 0.002839306 s 0.00292058 s 0.97
jaxmd20 / DefOpt / cuda / PostRev 0.002714315 s 0.002802275 s 0.97
jaxmd20 / DefOpt / cuda / BothRev 0.002751882 s 0.002825795 s 0.97
jaxmd20 / IDefOpt / cuda / PreRev 0.0028021549999999 s 0.002906115 s 0.96
jaxmd20 / IDefOpt / cuda / PostRev 0.002306254 s 0.0023555549999999 s 0.98
jaxmd20 / IDefOpt / cuda / BothRev 0.002739179 s 0.00283786 s 0.97
jaxmd20 / JaXPipe / tpu / Primal 0.0092657925 s 0.0092792274999999 s 1.00
jaxmd20 / Jax / tpu / Primal 0.009278880625 s 0.00926785375 s 1.00
jaxmd20 / HLOOpt / tpu / Primal 0.009165328125 s 0.009167618125 s 1.00
jaxmd20 / PartOpt / tpu / Primal 0.0092058225 s 0.009197459375 s 1.00
jaxmd20 / IPartOpt / tpu / Primal 0.0092033875 s 0.0092023018749999 s 1.00
jaxmd20 / DefOpt / tpu / Primal 0.008809266875 s 0.008798954375 s 1.00
jaxmd20 / IDefOpt / tpu / Primal 0.0086995924999999 s 0.008699993125 s 1.00
jaxmd20 / JaXPipe / tpu / Forward 0.017419693125 s 0.017414119375 s 1.00
jaxmd20 / Jax / tpu / Forward 0.018722006875 s 0.018728665625 s 1.00
jaxmd20 / HLOOpt / tpu / Forward 0.017398365 s 0.017394714375 s 1.00
jaxmd20 / PartOpt / tpu / Forward 0.017423145 s 0.017410929375 s 1.00
jaxmd20 / IPartOpt / tpu / Forward 0.017419463125 s 0.017412024375 s 1.00
jaxmd20 / DefOpt / tpu / Forward 0.01741421875 s 0.01741033625 s 1.00
jaxmd20 / IDefOpt / tpu / Forward 0.0174205425 s 0.0174184525 s 1.00
jaxmd20 / JaXPipe / tpu / PreRev 0.025465965 s 0.02544983 s 1.00
jaxmd20 / JaXPipe / tpu / PostRev 0.021851461875 s 0.021867141875 s 1.00
jaxmd20 / JaXPipe / tpu / BothRev 0.02546326375 s 0.025474003125 s 1.00
jaxmd20 / Jax / tpu / BothRev 0.021849925 s 0.021876434375 s 1.00
jaxmd20 / HLOOpt / tpu / PreRev 0.025581888125 s 0.025578508125 s 1.00
jaxmd20 / HLOOpt / tpu / PostRev 0.020804175625 s 0.020812361875 s 1.00
jaxmd20 / HLOOpt / tpu / BothRev 0.025687915625 s 0.025689636875 s 1.00
jaxmd20 / PartOpt / tpu / PreRev 0.02544238375 s 0.02549054125 s 1.00
jaxmd20 / PartOpt / tpu / PostRev 0.02150706125 s 0.0212508699999999 s 1.01
jaxmd20 / PartOpt / tpu / BothRev 0.025533224375 s 0.025562036875 s 1.00
jaxmd20 / IPartOpt / tpu / PreRev 0.02547254625 s 0.025474666875 s 1.00
jaxmd20 / IPartOpt / tpu / PostRev 0.021510494375 s 0.02152951375 s 1.00
jaxmd20 / IPartOpt / tpu / BothRev 0.025561353125 s 0.02556750125 s 1.00
jaxmd20 / DefOpt / tpu / PreRev 0.025441690625 s 0.025474653125 s 1.00
jaxmd20 / DefOpt / tpu / PostRev 0.0188212575 s 0.018803500625 s 1.00
jaxmd20 / DefOpt / tpu / BothRev 0.025536555625 s 0.025561023125 s 1.00
jaxmd20 / IDefOpt / tpu / PreRev 0.02547297 s 0.025476770625 s 1.00
jaxmd20 / IDefOpt / tpu / PostRev 0.018298068125 s 0.0183303137499999 s 1.00
jaxmd20 / IDefOpt / tpu / BothRev 0.025560403125 s 0.02554892875 s 1.00
jaxmd40 / JaXPipe / cpu / Primal 0.078738413 s 0.070117342 s 1.12
jaxmd40 / Jax / cpu / Primal 0.075014799 s 0.072817492 s 1.03
jaxmd40 / HLOOpt / cpu / Primal 0.096730552 s 0.093835609 s 1.03
jaxmd40 / PartOpt / cpu / Primal 0.074600263 s 0.071840139 s 1.04
jaxmd40 / IPartOpt / cpu / Primal 0.072025392 s 0.070842235 s 1.02
jaxmd40 / DefOpt / cpu / Primal 0.0853766899999999 s 0.089918741 s 0.95
jaxmd40 / IDefOpt / cpu / Primal 0.093626453 s 0.084506257 s 1.11
jaxmd40 / JaXPipe / cpu / Forward 0.171889989 s 0.165302914 s 1.04
jaxmd40 / Jax / cpu / Forward 0.090344886 s 0.089430703 s 1.01
jaxmd40 / HLOOpt / cpu / Forward 0.159450275 s 0.16514983 s 0.97
jaxmd40 / PartOpt / cpu / Forward 0.16519994 s 0.167051898 s 0.99
jaxmd40 / IPartOpt / cpu / Forward 0.169575869 s 0.163910112 s 1.03
jaxmd40 / DefOpt / cpu / Forward 0.15642214 s 0.16510477 s 0.95
jaxmd40 / IDefOpt / cpu / Forward 0.1676491269999999 s 0.164452005 s 1.02
jaxmd40 / JaXPipe / cpu / PreRev 0.2290246349999999 s 0.265458589 s 0.86
jaxmd40 / JaXPipe / cpu / PostRev 0.137916837 s 0.140804601 s 0.98
jaxmd40 / JaXPipe / cpu / BothRev 0.2376383879999999 s 0.255634347 s 0.93
jaxmd40 / Jax / cpu / BothRev 0.157535642 s 0.14182081 s 1.11
jaxmd40 / HLOOpt / cpu / PreRev 0.245390462 s 0.223691934 s 1.10
jaxmd40 / HLOOpt / cpu / PostRev 0.181269009 s 0.185176701 s 0.98
jaxmd40 / HLOOpt / cpu / BothRev 0.259055041 s 0.245576318 s 1.05
jaxmd40 / PartOpt / cpu / PreRev 0.2232584229999999 s 0.2292167729999999 s 0.97
jaxmd40 / PartOpt / cpu / PostRev 0.132015396 s 0.15071399 s 0.88
jaxmd40 / PartOpt / cpu / BothRev 0.265186233 s 0.240208136 s 1.10
jaxmd40 / IPartOpt / cpu / PreRev 0.234309584 s 0.224542864 s 1.04
jaxmd40 / IPartOpt / cpu / PostRev 0.125497865 s 0.137075042 s 0.92
jaxmd40 / IPartOpt / cpu / BothRev 0.2610459619999999 s 0.251974875 s 1.04
jaxmd40 / DefOpt / cpu / PreRev 0.219393167 s 0.220720154 s 0.99
jaxmd40 / DefOpt / cpu / PostRev 0.17803215 s 0.178141375 s 1.00
jaxmd40 / DefOpt / cpu / BothRev 0.24111409 s 0.269256488 s 0.90
jaxmd40 / IDefOpt / cpu / PreRev 0.226550349 s 0.245975364 s 0.92
jaxmd40 / IDefOpt / cpu / PostRev 0.18003828 s 0.18207069 s 0.99
jaxmd40 / IDefOpt / cpu / BothRev 0.2672159679999999 s 0.254276777 s 1.05
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / cuda / Primal 1.700705146 s 1.705607055 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / cuda / Primal 1.70364286 s 1.707929988 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / cuda / Primal 1.714447076 s 1.7179912039999998 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / cuda / Primal 1.695222633 s 1.6990359940000002 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / cuda / Primal 1.692677357 s 1.697318164 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / cuda / Primal 1.663852939 s 1.667815115 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / cuda / Primal 1.918772068 s 1.925575094 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / tpu / Primal 3.039171935625 s 3.0388671775 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / tpu / Primal 3.039670756875 s 3.039500189375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / tpu / Primal 3.121950946875 s 3.121960315 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / tpu / Primal 3.06047748875 s 3.060258405625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / tpu / Primal 3.060774455 s 3.060504856875 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / tpu / Primal 2.10250766875 s 2.102534495625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / tpu / Primal 2.944770424375 s 2.94487333 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / JaXPipe / cpu / Primal 6.275564322999999 s 6.120383601 s 1.03
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / Jax / cpu / Primal 6.419076309 s 6.174035622 s 1.04
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / HLOOpt / cpu / Primal 6.368401107 s 5.932616669 s 1.07
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / PartOpt / cpu / Primal 6.430310169999999 s 6.119237832 s 1.05
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IPartOpt / cpu / Primal 6.421519419 s 6.105743663 s 1.05
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / DefOpt / cpu / Primal 2.536757715 s 2.422323865 s 1.05
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IDefOpt / cpu / Primal 6.9571596190000005 s 6.601675161999999 s 1.05

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

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

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants