Skip to content

Conversation

@ftynse
Copy link
Collaborator

@ftynse ftynse commented Jan 12, 2026

No description provided.

@ftynse ftynse mentioned this pull request Jan 12, 2026
Base automatically changed from users/ftynse/idempotent-dus to main January 12, 2026 20:56
@ftynse ftynse force-pushed the users/ftynse/wip-dus branch from d1b9215 to 3b9cc2c Compare January 28, 2026 21:51
@ftynse ftynse force-pushed the users/ftynse/wip-dus branch from 3b9cc2c to 55c389a Compare January 28, 2026 21:53
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: 55c389a Previous: ec8818c Ratio
actmtch / JaXPipe / cpu / Primal 0.000007300540009964607 s 0.00000692249999701744 s 1.05
actmtch / Jax / cpu / Primal 0.000007287299958989024 s 0.000007192939992819447 s 1.01
actmtch / HLOOpt / cpu / Primal 0.000011170899988428571 s 0.000010044379996543284 s 1.11
actmtch / PartOpt / cpu / Primal 0.000006461359989771154 s 0.0000064412800475111 s 1.00
actmtch / IPartOpt / cpu / Primal 0.000006793140000809217 s 0.000006818480023866869 s 1.00
actmtch / DefOpt / cpu / Primal 0.000011109440029031248 s 0.000011361360047885685 s 0.98
actmtch / IDefOpt / cpu / Primal 0.000007494020001104218 s 0.00000708371998371149 s 1.06
actmtch / JaXPipe / cpu / Forward 0.00001139960000728024 s 0.00001113240002268867 s 1.02
actmtch / Jax / cpu / Forward 0.000009947280004780624 s 0.000009812340013013454 s 1.01
actmtch / HLOOpt / cpu / Forward 0.000015465339965885506 s 0.00001489797998146969 s 1.04
actmtch / PartOpt / cpu / Forward 0.00001542135996714933 s 0.000015003800044723902 s 1.03
actmtch / IPartOpt / cpu / Forward 0.000011232480001126532 s 0.000009955099958460778 s 1.13
actmtch / DefOpt / cpu / Forward 0.000016258940022453316 s 0.00001583584001309646 s 1.03
actmtch / IDefOpt / cpu / Forward 0.000010977879965139436 s 0.000010704540018195984 s 1.03
actmtch / JaXPipe / cpu / PreRev 0.000012802679984815768 s 0.000011638940004559117 s 1.10
actmtch / JaXPipe / cpu / PostRev 0.000011503040022944333 s 0.000012065279997841572 s 0.95
actmtch / JaXPipe / cpu / BothRev 0.000012007199993604445 s 0.0000117583600240323 s 1.02
actmtch / Jax / cpu / BothRev 0.00001049929998771404 s 0.000010410939994471846 s 1.01
actmtch / HLOOpt / cpu / PreRev 0.000011763180018533604 s 0.00001159960002041771 s 1.01
actmtch / HLOOpt / cpu / PostRev 0.00001586659997883544 s 0.000015379520009446422 s 1.03
actmtch / HLOOpt / cpu / BothRev 0.000013963060036985552 s 0.000012474900013330625 s 1.12
actmtch / PartOpt / cpu / PreRev 0.000011584619987843326 s 0.000011061339973821305 s 1.05
actmtch / PartOpt / cpu / PostRev 0.000010994800004482383 s 0.000010003460010921117 s 1.10
actmtch / PartOpt / cpu / BothRev 0.000012488780012063216 s 0.000011381120011719758 s 1.10
actmtch / IPartOpt / cpu / PreRev 0.000011905239980478654 s 0.000011578979974729008 s 1.03
actmtch / IPartOpt / cpu / PostRev 0.000010612479982228252 s 0.000010558000030869151 s 1.01
actmtch / IPartOpt / cpu / BothRev 0.000012353099982647107 s 0.00001147782000771258 s 1.08
actmtch / DefOpt / cpu / PreRev 0.000012026760005028335 s 0.000010500259968466707 s 1.15
actmtch / DefOpt / cpu / PostRev 0.00001208511997901951 s 0.000010752300040621776 s 1.12
actmtch / DefOpt / cpu / BothRev 0.000011682939939419156 s 0.000011039860000892076 s 1.06
actmtch / IDefOpt / cpu / PreRev 0.00001155577999270463 s 0.000011022119952031064 s 1.05
actmtch / IDefOpt / cpu / PostRev 0.000012456840013328474 s 0.000011094300007243871 s 1.12
actmtch / IDefOpt / cpu / BothRev 0.000012218960018799408 s 0.000011542900010681478 s 1.06
actmtch / JaXPipe / tpu / Primal 5.6315e-7 s 5.63325e-7 s 1.00
actmtch / Jax / tpu / Primal 6.06625e-7 s 6.063499999999999e-7 s 1.00
actmtch / HLOOpt / tpu / Primal 0.0000021008 s 0.000002106275 s 1.00
actmtch / PartOpt / tpu / Primal 6.06475e-7 s 6.070249999999999e-7 s 1.00
actmtch / IPartOpt / tpu / Primal 5.6235e-7 s 5.62425e-7 s 1.00
actmtch / DefOpt / tpu / Primal 0.000002155825 s 0.0000021686000000000003 s 0.99
actmtch / IDefOpt / tpu / Primal 0.000002089975 s 0.000002093 s 1.00
actmtch / JaXPipe / tpu / Forward 0.000003826075 s 0.00000383005 s 1.00
actmtch / Jax / tpu / Forward 0.00000122085 s 0.000001219325 s 1.00
actmtch / HLOOpt / tpu / Forward 0.000003932925 s 0.000003926374999999999 s 1.00
actmtch / PartOpt / tpu / Forward 0.000003917075 s 0.0000039193 s 1.00
actmtch / IPartOpt / tpu / Forward 0.0000039381500000000006 s 0.000003935075 s 1.00
actmtch / DefOpt / tpu / Forward 0.0000039195 s 0.00000390945 s 1.00
actmtch / IDefOpt / tpu / Forward 0.000003936724999999999 s 0.00000393845 s 1.00
actmtch / JaXPipe / tpu / PreRev 0.000003484 s 0.000003469025 s 1.00
actmtch / JaXPipe / tpu / PostRev 0.000001637875 s 0.0000016335 s 1.00
actmtch / JaXPipe / tpu / BothRev 0.00000347505 s 0.0000034643 s 1.00
actmtch / Jax / tpu / BothRev 0.0000016327 s 0.000001636025 s 1.00
actmtch / HLOOpt / tpu / PreRev 0.0000034955750000000004 s 0.0000034855249999999995 s 1.00
actmtch / HLOOpt / tpu / PostRev 0.000003427825 s 0.00000341325 s 1.00
actmtch / HLOOpt / tpu / BothRev 0.0000034659 s 0.000003471125 s 1.00
actmtch / PartOpt / tpu / PreRev 0.000003404125 s 0.000003408125 s 1.00
actmtch / PartOpt / tpu / PostRev 0.000001596725 s 0.000001592625 s 1.00
actmtch / PartOpt / tpu / BothRev 0.00000341365 s 0.00000341455 s 1.00
actmtch / IPartOpt / tpu / PreRev 0.000003471975 s 0.0000034695 s 1.00
actmtch / IPartOpt / tpu / PostRev 0.000001631975 s 0.00000163085 s 1.00
actmtch / IPartOpt / tpu / BothRev 0.0000034738000000000005 s 0.00000347485 s 1.00
actmtch / DefOpt / tpu / PreRev 0.000003415825 s 0.00000342005 s 1.00
actmtch / DefOpt / tpu / PostRev 0.00000342475 s 0.000003431825 s 1.00
actmtch / DefOpt / tpu / BothRev 0.0000034329500000000003 s 0.0000034074749999999995 s 1.01
actmtch / IDefOpt / tpu / PreRev 0.000003478475 s 0.0000034717750000000003 s 1.00
actmtch / IDefOpt / tpu / PostRev 0.00000343865 s 0.0000034117 s 1.01
actmtch / IDefOpt / tpu / BothRev 0.00000348995 s 0.0000034676 s 1.01
actmtch / JaXPipe / cpu / Primal 0.000013051 s 0.00000692249999701744 s 1.89
actmtch / Jax / cpu / Primal 0.00001321 s 0.000007192939992819447 s 1.84
actmtch / HLOOpt / cpu / Primal 0.000013502 s 0.000010044379996543284 s 1.34
actmtch / PartOpt / cpu / Primal 0.000012972 s 0.0000064412800475111 s 2.01
actmtch / IPartOpt / cpu / Primal 0.000013266 s 0.000006818480023866869 s 1.95
actmtch / DefOpt / cpu / Primal 0.00001366 s 0.000011361360047885685 s 1.20
actmtch / IDefOpt / cpu / Primal 0.000014141 s 0.00000708371998371149 s 2.00
actmtch / JaXPipe / cpu / Forward 0.000018747 s 0.00001113240002268867 s 1.68
actmtch / Jax / cpu / Forward 0.000018076 s 0.000009812340013013454 s 1.84
actmtch / HLOOpt / cpu / Forward 0.000019021 s 0.00001489797998146969 s 1.28
actmtch / PartOpt / cpu / Forward 0.000018723 s 0.000015003800044723902 s 1.25
actmtch / IPartOpt / cpu / Forward 0.000018864 s 0.000009955099958460778 s 1.89
actmtch / DefOpt / cpu / Forward 0.000018714 s 0.00001583584001309646 s 1.18
actmtch / IDefOpt / cpu / Forward 0.000018936000000000003 s 0.000010704540018195984 s 1.77
actmtch / JaXPipe / cpu / PreRev 0.000019109 s 0.000011638940004559117 s 1.64
actmtch / JaXPipe / cpu / PostRev 0.000017463 s 0.000012065279997841572 s 1.45
actmtch / JaXPipe / cpu / BothRev 0.000019151 s 0.0000117583600240323 s 1.63
actmtch / Jax / cpu / BothRev 0.000017183 s 0.000010410939994471846 s 1.65
actmtch / HLOOpt / cpu / PreRev 0.000018571 s 0.00001159960002041771 s 1.60
actmtch / HLOOpt / cpu / PostRev 0.0000192 s 0.000015379520009446422 s 1.25
actmtch / HLOOpt / cpu / BothRev 0.000019124 s 0.000012474900013330625 s 1.53
actmtch / PartOpt / cpu / PreRev 0.000019201 s 0.000011061339973821305 s 1.74
actmtch / PartOpt / cpu / PostRev 0.000016898 s 0.000010003460010921117 s 1.69
actmtch / PartOpt / cpu / BothRev 0.000019331 s 0.000011381120011719758 s 1.70
actmtch / IPartOpt / cpu / PreRev 0.000018803 s 0.000011578979974729008 s 1.62
actmtch / IPartOpt / cpu / PostRev 0.000017092 s 0.000010558000030869151 s 1.62
actmtch / IPartOpt / cpu / BothRev 0.00001906 s 0.00001147782000771258 s 1.66
actmtch / DefOpt / cpu / PreRev 0.000019408 s 0.000010500259968466707 s 1.85
actmtch / DefOpt / cpu / PostRev 0.000019413 s 0.000010752300040621776 s 1.81
actmtch / DefOpt / cpu / BothRev 0.00002008 s 0.000011039860000892076 s 1.82
actmtch / IDefOpt / cpu / PreRev 0.000019107 s 0.000011022119952031064 s 1.73
actmtch / IDefOpt / cpu / PostRev 0.000020788 s 0.000011094300007243871 s 1.87
actmtch / IDefOpt / cpu / BothRev 0.000019747 s 0.000011542900010681478 s 1.71
actmtch / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.00000692249999701744 s 1.30
actmtch / Jax / cpu / Primal 0.000008 s 0.000007192939992819447 s 1.11
actmtch / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000010044379996543284 s 0.90
actmtch / PartOpt / cpu / Primal 0.000008999999999999999 s 0.0000064412800475111 s 1.40
actmtch / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000006818480023866869 s 1.32
actmtch / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000011361360047885685 s 0.79
actmtch / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.00000708371998371149 s 1.27
actmtch / JaXPipe / cpu / Forward 0.000013 s 0.00001113240002268867 s 1.17
actmtch / Jax / cpu / Forward 0.000012 s 0.000009812340013013454 s 1.22
actmtch / HLOOpt / cpu / Forward 0.000012 s 0.00001489797998146969 s 0.81
actmtch / PartOpt / cpu / Forward 0.000013 s 0.000015003800044723902 s 0.87
actmtch / IPartOpt / cpu / Forward 0.000013 s 0.000009955099958460778 s 1.31
actmtch / DefOpt / cpu / Forward 0.000014 s 0.00001583584001309646 s 0.88
actmtch / IDefOpt / cpu / Forward 0.000013 s 0.000010704540018195984 s 1.21
actmtch / JaXPipe / cpu / PreRev 0.000012 s 0.000011638940004559117 s 1.03
actmtch / JaXPipe / cpu / PostRev 0.000011 s 0.000012065279997841572 s 0.91
actmtch / JaXPipe / cpu / BothRev 0.000013 s 0.0000117583600240323 s 1.11
actmtch / Jax / cpu / BothRev 0.000011 s 0.000010410939994471846 s 1.06
actmtch / HLOOpt / cpu / PreRev 0.000014 s 0.00001159960002041771 s 1.21
actmtch / HLOOpt / cpu / PostRev 0.000013 s 0.000015379520009446422 s 0.85
actmtch / HLOOpt / cpu / BothRev 0.000014 s 0.000012474900013330625 s 1.12
actmtch / PartOpt / cpu / PreRev 0.000013 s 0.000011061339973821305 s 1.18
actmtch / PartOpt / cpu / PostRev 0.000012 s 0.000010003460010921117 s 1.20
actmtch / PartOpt / cpu / BothRev 0.000013 s 0.000011381120011719758 s 1.14
actmtch / IPartOpt / cpu / PreRev 0.000013 s 0.000011578979974729008 s 1.12
actmtch / IPartOpt / cpu / PostRev 0.000012 s 0.000010558000030869151 s 1.14
actmtch / IPartOpt / cpu / BothRev 0.000013 s 0.00001147782000771258 s 1.13
actmtch / DefOpt / cpu / PreRev 0.000013 s 0.000010500259968466707 s 1.24
actmtch / DefOpt / cpu / PostRev 0.000013 s 0.000010752300040621776 s 1.21
actmtch / DefOpt / cpu / BothRev 0.000013 s 0.000011039860000892076 s 1.18
actmtch / IDefOpt / cpu / PreRev 0.000013 s 0.000011022119952031064 s 1.18
actmtch / IDefOpt / cpu / PostRev 0.000014 s 0.000011094300007243871 s 1.26
actmtch / IDefOpt / cpu / BothRev 0.000013 s 0.000011542900010681478 s 1.13
add_one / JaXPipe / cpu / Primal 0.000007342220014834311 s 0.0000073595999401732115 s 1.00
add_one / Jax / cpu / Primal 0.000007308700050998596 s 0.000007034759973976179 s 1.04
add_one / HLOOpt / cpu / Primal 0.000010920079994320986 s 0.00000991676000012376 s 1.10
add_one / PartOpt / cpu / Primal 0.000007385780027107102 s 0.000006248380041142809 s 1.18
add_one / IPartOpt / cpu / Primal 0.000007316780056498828 s 0.000007288220003829337 s 1.00
add_one / DefOpt / cpu / Primal 0.00001124008000260801 s 0.00001052379998327524 s 1.07
add_one / IDefOpt / cpu / Primal 0.0000072742200336506355 s 0.00000706286001332046 s 1.03
add_one / JaXPipe / cpu / Forward 0.000011336879997543291 s 0.000010157779997825856 s 1.12
add_one / Jax / cpu / Forward 0.000011273220006842168 s 0.000010256300001856288 s 1.10
add_one / HLOOpt / cpu / Forward 0.000011372239951015218 s 0.000014874860007694225 s 0.76
add_one / PartOpt / cpu / Forward 0.000016031880004447885 s 0.000015278780010703487 s 1.05
add_one / IPartOpt / cpu / Forward 0.000011278179999862916 s 0.00001031749999128806 s 1.09
add_one / DefOpt / cpu / Forward 0.00001547474003928073 s 0.0000152864399842656 s 1.01
add_one / IDefOpt / cpu / Forward 0.000011440139987826114 s 0.000010301780048393991 s 1.11
add_one / JaXPipe / cpu / PreRev 0.000012894139981654009 s 0.000011906379986612592 s 1.08
add_one / JaXPipe / cpu / PostRev 0.000012720859949695295 s 0.000011541499998202198 s 1.10
add_one / JaXPipe / cpu / BothRev 0.000017315999984930384 s 0.00001620450001610152 s 1.07
add_one / Jax / cpu / BothRev 0.000012695939994955552 s 0.000011179479997736053 s 1.14
add_one / HLOOpt / cpu / PreRev 0.000013111619991832412 s 0.000012173239983894746 s 1.08
add_one / HLOOpt / cpu / PostRev 0.000017873520009743514 s 0.00001608182004019909 s 1.11
add_one / HLOOpt / cpu / BothRev 0.000014331059946925962 s 0.00001294726001106028 s 1.11
add_one / PartOpt / cpu / PreRev 0.000013333520028027124 s 0.00001129604002017004 s 1.18
add_one / PartOpt / cpu / PostRev 0.000013619099945572088 s 0.000011593139970500487 s 1.17
add_one / PartOpt / cpu / BothRev 0.000013152480014468892 s 0.000011710300022969022 s 1.12
add_one / IPartOpt / cpu / PreRev 0.000016360719964723103 s 0.00001609911997547897 s 1.02
add_one / IPartOpt / cpu / PostRev 0.000013051979994997964 s 0.000011518619985508848 s 1.13
add_one / IPartOpt / cpu / BothRev 0.000013017719948038576 s 0.000011486419980428764 s 1.13
add_one / DefOpt / cpu / PreRev 0.000012923600006615743 s 0.000011135619997730828 s 1.16
add_one / DefOpt / cpu / PostRev 0.000012992440006200922 s 0.000011578360044950388 s 1.12
add_one / DefOpt / cpu / BothRev 0.00001305291997596214 s 0.00001155000003564055 s 1.13
add_one / IDefOpt / cpu / PreRev 0.000013260859977890505 s 0.000011922440025955438 s 1.11
add_one / IDefOpt / cpu / PostRev 0.000012767420003001462 s 0.00001165133999165846 s 1.10
add_one / IDefOpt / cpu / BothRev 0.00001302536000366672 s 0.00001182244000119681 s 1.10
add_one / JaXPipe / tpu / Primal 0.0000014312 s 0.00000142875 s 1.00
add_one / Jax / tpu / Primal 0.00000140045 s 0.000001409425 s 0.99
add_one / HLOOpt / tpu / Primal 0.00000142765 s 0.00000143245 s 1.00
add_one / PartOpt / tpu / Primal 0.0000014053 s 0.0000013988 s 1.00
add_one / IPartOpt / tpu / Primal 0.000001426625 s 0.000001435375 s 0.99
add_one / DefOpt / tpu / Primal 0.0000014094750000000002 s 0.00000139965 s 1.01
add_one / IDefOpt / tpu / Primal 0.00000143115 s 0.0000014273 s 1.00
add_one / JaXPipe / tpu / Forward 0.00000185405 s 0.000001852525 s 1.00
add_one / Jax / tpu / Forward 0.000001846425 s 0.00000184345 s 1.00
add_one / HLOOpt / tpu / Forward 0.0000018519 s 0.0000018489 s 1.00
add_one / PartOpt / tpu / Forward 0.0000018474 s 0.000001853725 s 1.00
add_one / IPartOpt / tpu / Forward 0.0000018519 s 0.0000018641 s 0.99
add_one / DefOpt / tpu / Forward 0.0000018487 s 0.00000185065 s 1.00
add_one / IDefOpt / tpu / Forward 0.000001849475 s 0.00000184625 s 1.00
add_one / JaXPipe / tpu / PreRev 0.000002251975 s 0.000002229975 s 1.01
add_one / JaXPipe / tpu / PostRev 0.0000022364 s 0.000002232975 s 1.00
add_one / JaXPipe / tpu / BothRev 0.00000223545 s 0.0000022385 s 1.00
add_one / Jax / tpu / BothRev 0.000002237225 s 0.000002242025 s 1.00
add_one / HLOOpt / tpu / PreRev 0.000002241525 s 0.000002239425 s 1.00
add_one / HLOOpt / tpu / PostRev 0.000002246075 s 0.000002238575 s 1.00
add_one / HLOOpt / tpu / BothRev 0.000002229975 s 0.0000022336750000000003 s 1.00
add_one / PartOpt / tpu / PreRev 0.000002241925 s 0.00000224675 s 1.00
add_one / PartOpt / tpu / PostRev 0.000002242175 s 0.0000022308 s 1.01
add_one / PartOpt / tpu / BothRev 0.000002248625 s 0.00000224115 s 1.00
add_one / IPartOpt / tpu / PreRev 0.000002235325 s 0.0000022415 s 1.00
add_one / IPartOpt / tpu / PostRev 0.000002236725 s 0.000002238925 s 1.00
add_one / IPartOpt / tpu / BothRev 0.0000022339 s 0.0000022433 s 1.00
add_one / DefOpt / tpu / PreRev 0.0000022372 s 0.000002241425 s 1.00
add_one / DefOpt / tpu / PostRev 0.000002236775 s 0.00000223545 s 1.00
add_one / DefOpt / tpu / BothRev 0.0000022412 s 0.000002242525 s 1.00
add_one / IDefOpt / tpu / PreRev 0.000002232075 s 0.0000022373 s 1.00
add_one / IDefOpt / tpu / PostRev 0.00000223295 s 0.0000022454750000000003 s 0.99
add_one / IDefOpt / tpu / BothRev 0.000002244575 s 0.0000022413 s 1.00
add_one / JaXPipe / cpu / Primal 0.000013174999999999998 s 0.0000073595999401732115 s 1.79
add_one / Jax / cpu / Primal 0.00001331 s 0.000007034759973976179 s 1.89
add_one / HLOOpt / cpu / Primal 0.000012717 s 0.00000991676000012376 s 1.28
add_one / PartOpt / cpu / Primal 0.000013168 s 0.000006248380041142809 s 2.11
add_one / IPartOpt / cpu / Primal 0.000012703 s 0.000007288220003829337 s 1.74
add_one / DefOpt / cpu / Primal 0.000012662 s 0.00001052379998327524 s 1.20
add_one / IDefOpt / cpu / Primal 0.000012869 s 0.00000706286001332046 s 1.82
add_one / JaXPipe / cpu / Forward 0.000017262 s 0.000010157779997825856 s 1.70
add_one / Jax / cpu / Forward 0.000016698000000000002 s 0.000010256300001856288 s 1.63
add_one / HLOOpt / cpu / Forward 0.000017205000000000002 s 0.000014874860007694225 s 1.16
add_one / PartOpt / cpu / Forward 0.000017570999999999998 s 0.000015278780010703487 s 1.15
add_one / IPartOpt / cpu / Forward 0.000017489 s 0.00001031749999128806 s 1.70
add_one / DefOpt / cpu / Forward 0.00001704 s 0.0000152864399842656 s 1.11
add_one / IDefOpt / cpu / Forward 0.000017565000000000002 s 0.000010301780048393991 s 1.71
add_one / JaXPipe / cpu / PreRev 0.00002006 s 0.000011906379986612592 s 1.68
add_one / JaXPipe / cpu / PostRev 0.000019488 s 0.000011541499998202198 s 1.69
add_one / JaXPipe / cpu / BothRev 0.000020398 s 0.00001620450001610152 s 1.26
add_one / Jax / cpu / BothRev 0.000019424 s 0.000011179479997736053 s 1.74
add_one / HLOOpt / cpu / PreRev 0.000019231000000000003 s 0.000012173239983894746 s 1.58
add_one / HLOOpt / cpu / PostRev 0.000020065 s 0.00001608182004019909 s 1.25
add_one / HLOOpt / cpu / BothRev 0.000019755000000000003 s 0.00001294726001106028 s 1.53
add_one / PartOpt / cpu / PreRev 0.000019482 s 0.00001129604002017004 s 1.72
add_one / PartOpt / cpu / PostRev 0.000020454 s 0.000011593139970500487 s 1.76
add_one / PartOpt / cpu / BothRev 0.000020213 s 0.000011710300022969022 s 1.73
add_one / IPartOpt / cpu / PreRev 0.000019644 s 0.00001609911997547897 s 1.22
add_one / IPartOpt / cpu / PostRev 0.000020736 s 0.000011518619985508848 s 1.80
add_one / IPartOpt / cpu / BothRev 0.0000203 s 0.000011486419980428764 s 1.77
add_one / DefOpt / cpu / PreRev 0.000020314 s 0.000011135619997730828 s 1.82
add_one / DefOpt / cpu / PostRev 0.000020409 s 0.000011578360044950388 s 1.76
add_one / DefOpt / cpu / BothRev 0.000018885 s 0.00001155000003564055 s 1.64
add_one / IDefOpt / cpu / PreRev 0.000019061000000000003 s 0.000011922440025955438 s 1.60
add_one / IDefOpt / cpu / PostRev 0.000020055 s 0.00001165133999165846 s 1.72
add_one / IDefOpt / cpu / BothRev 0.000020525 s 0.00001182244000119681 s 1.74
add_one / JaXPipe / cpu / Primal 0.000008 s 0.0000073595999401732115 s 1.09
add_one / Jax / cpu / Primal 0.000008 s 0.000007034759973976179 s 1.14
add_one / HLOOpt / cpu / Primal 0.000008 s 0.00000991676000012376 s 0.81
add_one / PartOpt / cpu / Primal 0.000008 s 0.000006248380041142809 s 1.28
add_one / IPartOpt / cpu / Primal 0.000008 s 0.000007288220003829337 s 1.10
add_one / DefOpt / cpu / Primal 0.000008 s 0.00001052379998327524 s 0.76
add_one / IDefOpt / cpu / Primal 0.000008 s 0.00000706286001332046 s 1.13
add_one / JaXPipe / cpu / Forward 0.000012 s 0.000010157779997825856 s 1.18
add_one / Jax / cpu / Forward 0.000011 s 0.000010256300001856288 s 1.07
add_one / HLOOpt / cpu / Forward 0.000011 s 0.000014874860007694225 s 0.74
add_one / PartOpt / cpu / Forward 0.000011 s 0.000015278780010703487 s 0.72
add_one / IPartOpt / cpu / Forward 0.000011 s 0.00001031749999128806 s 1.07
add_one / DefOpt / cpu / Forward 0.000011 s 0.0000152864399842656 s 0.72
add_one / IDefOpt / cpu / Forward 0.000011 s 0.000010301780048393991 s 1.07
add_one / JaXPipe / cpu / PreRev 0.000013 s 0.000011906379986612592 s 1.09
add_one / JaXPipe / cpu / PostRev 0.000013 s 0.000011541499998202198 s 1.13
add_one / JaXPipe / cpu / BothRev 0.000013 s 0.00001620450001610152 s 0.80
add_one / Jax / cpu / BothRev 0.000013 s 0.000011179479997736053 s 1.16
add_one / HLOOpt / cpu / PreRev 0.000013 s 0.000012173239983894746 s 1.07
add_one / HLOOpt / cpu / PostRev 0.000013 s 0.00001608182004019909 s 0.81
add_one / HLOOpt / cpu / BothRev 0.000014 s 0.00001294726001106028 s 1.08
add_one / PartOpt / cpu / PreRev 0.000013 s 0.00001129604002017004 s 1.15
add_one / PartOpt / cpu / PostRev 0.000013 s 0.000011593139970500487 s 1.12
add_one / PartOpt / cpu / BothRev 0.000014 s 0.000011710300022969022 s 1.20
add_one / IPartOpt / cpu / PreRev 0.000013 s 0.00001609911997547897 s 0.81
add_one / IPartOpt / cpu / PostRev 0.000013 s 0.000011518619985508848 s 1.13
add_one / IPartOpt / cpu / BothRev 0.000014 s 0.000011486419980428764 s 1.22
add_one / DefOpt / cpu / PreRev 0.000013 s 0.000011135619997730828 s 1.17
add_one / DefOpt / cpu / PostRev 0.000013 s 0.000011578360044950388 s 1.12
add_one / DefOpt / cpu / BothRev 0.000013 s 0.00001155000003564055 s 1.13
add_one / IDefOpt / cpu / PreRev 0.000012 s 0.000011922440025955438 s 1.01
add_one / IDefOpt / cpu / PostRev 0.000013 s 0.00001165133999165846 s 1.12
add_one / IDefOpt / cpu / BothRev 0.000013 s 0.00001182244000119681 s 1.10
add_two / JaXPipe / cpu / Primal 0.000008801879994280171 s 0.000008058220000748406 s 1.09
add_two / Jax / cpu / Primal 0.000007770960037305485 s 0.000006939919994692901 s 1.12
add_two / HLOOpt / cpu / Primal 0.000011369999992894008 s 0.00001038918002450373 s 1.09
add_two / PartOpt / cpu / Primal 0.000008214099998440361 s 0.0000068256200029281896 s 1.20
add_two / IPartOpt / cpu / Primal 0.000007630360032635508 s 0.000007170440030677128 s 1.06
add_two / DefOpt / cpu / Primal 0.00001179445999696327 s 0.000011246620015299414 s 1.05
add_two / IDefOpt / cpu / Primal 0.000007342380013142247 s 0.0000069606599936378185 s 1.05
add_two / JaXPipe / cpu / Forward 0.00001180423998448532 s 0.000010616780009513605 s 1.11
add_two / Jax / cpu / Forward 0.000011212179979338543 s 0.00001036977996591304 s 1.08
add_two / HLOOpt / cpu / Forward 0.00001155329999164678 s 0.000015078660026119903 s 0.77
add_two / PartOpt / cpu / Forward 0.00001668850000896782 s 0.000015557280021312182 s 1.07
add_two / IPartOpt / cpu / Forward 0.000011495479993754998 s 0.000010255780007355496 s 1.12
add_two / DefOpt / cpu / Forward 0.000016344079967893775 s 0.00001550278000649996 s 1.05
add_two / IDefOpt / cpu / Forward 0.000011738919993149464 s 0.000010459999939484988 s 1.12
add_two / JaXPipe / cpu / PreRev 0.000016086079986052935 s 0.000013837839987900223 s 1.16
add_two / JaXPipe / cpu / PostRev 0.00001591484001437493 s 0.000013908720056861057 s 1.14
add_two / JaXPipe / cpu / BothRev 0.00001526151998405112 s 0.000013828220016876002 s 1.10
add_two / Jax / cpu / BothRev 0.00001533804002974648 s 0.000014264560040828656 s 1.08
add_two / HLOOpt / cpu / PreRev 0.00001594121998095943 s 0.000014075380013309769 s 1.13
add_two / HLOOpt / cpu / PostRev 0.00001549643995531369 s 0.000013675300060640438 s 1.13
add_two / HLOOpt / cpu / BothRev 0.000016837119992487715 s 0.000016088479987956817 s 1.05
add_two / PartOpt / cpu / PreRev 0.000015427339985762955 s 0.000013798339987260988 s 1.12
add_two / PartOpt / cpu / PostRev 0.000015323559964599553 s 0.000013859300015610642 s 1.11
add_two / PartOpt / cpu / BothRev 0.000015072619989950908 s 0.000013659700052812696 s 1.10
add_two / IPartOpt / cpu / PreRev 0.000015451759982170187 s 0.00001460233999750926 s 1.06
add_two / IPartOpt / cpu / PostRev 0.00001547687999845948 s 0.000014303160014605965 s 1.08
add_two / IPartOpt / cpu / BothRev 0.000015735740007585263 s 0.00001426799999535433 s 1.10
add_two / DefOpt / cpu / PreRev 0.000015865879986449728 s 0.000013933239979451172 s 1.14
add_two / DefOpt / cpu / PostRev 0.000016083759983303026 s 0.000014049399933355744 s 1.14
add_two / DefOpt / cpu / BothRev 0.00001518266001767188 s 0.000014438600001085431 s 1.05
add_two / IDefOpt / cpu / PreRev 0.00001598312001078739 s 0.000014167980016281944 s 1.13
add_two / IDefOpt / cpu / PostRev 0.00001651402005336422 s 0.000014254260022426025 s 1.16
add_two / IDefOpt / cpu / BothRev 0.00001533012000436429 s 0.000014405700057977813 s 1.06
add_two / JaXPipe / tpu / Primal 0.0000014303250000000002 s 0.0000014328 s 1.00
add_two / Jax / tpu / Primal 0.00000147575 s 0.0000014742 s 1.00
add_two / HLOOpt / tpu / Primal 0.000001428725 s 0.00000143695 s 0.99
add_two / PartOpt / tpu / Primal 0.0000014712 s 0.0000014734249999999998 s 1.00
add_two / IPartOpt / tpu / Primal 0.0000014240500000000002 s 0.0000014325750000000002 s 0.99
add_two / DefOpt / tpu / Primal 0.0000014727499999999998 s 0.00000147375 s 1.00
add_two / IDefOpt / tpu / Primal 0.00000142905 s 0.0000014436 s 0.99
add_two / JaXPipe / tpu / Forward 0.000001830325 s 0.0000018271 s 1.00
add_two / Jax / tpu / Forward 0.000001832425 s 0.00000182135 s 1.01
add_two / HLOOpt / tpu / Forward 0.00000182685 s 0.000001835825 s 1.00
add_two / PartOpt / tpu / Forward 0.000001832425 s 0.000001837925 s 1.00
add_two / IPartOpt / tpu / Forward 0.00000182405 s 0.000001823675 s 1.00
add_two / DefOpt / tpu / Forward 0.000001826725 s 0.000001829975 s 1.00
add_two / IDefOpt / tpu / Forward 0.0000018309 s 0.00000183595 s 1.00
add_two / JaXPipe / tpu / PreRev 0.00000284205 s 0.000002844175 s 1.00
add_two / JaXPipe / tpu / PostRev 0.0000027467750000000005 s 0.00000275715 s 1.00
add_two / JaXPipe / tpu / BothRev 0.000002834675 s 0.000002839675 s 1.00
add_two / Jax / tpu / BothRev 0.000002761725 s 0.000002757175 s 1.00
add_two / HLOOpt / tpu / PreRev 0.0000028428750000000003 s 0.0000028357 s 1.00
add_two / HLOOpt / tpu / PostRev 0.000002746375 s 0.000002749775 s 1.00
add_two / HLOOpt / tpu / BothRev 0.0000028374 s 0.000002840225 s 1.00
add_two / PartOpt / tpu / PreRev 0.0000027419500000000005 s 0.000002751025 s 1.00
add_two / PartOpt / tpu / PostRev 0.000002838025 s 0.000002834025 s 1.00
add_two / PartOpt / tpu / BothRev 0.0000027608250000000004 s 0.0000027528 s 1.00
add_two / IPartOpt / tpu / PreRev 0.00000283725 s 0.00000284095 s 1.00
add_two / IPartOpt / tpu / PostRev 0.0000027548 s 0.00000275005 s 1.00
add_two / IPartOpt / tpu / BothRev 0.000002832375 s 0.0000028408 s 1.00
add_two / DefOpt / tpu / PreRev 0.00000275115 s 0.0000027644999999999995 s 1.00
add_two / DefOpt / tpu / PostRev 0.0000028431 s 0.000002837725 s 1.00
add_two / DefOpt / tpu / BothRev 0.000002741925 s 0.000002755775 s 0.99
add_two / IDefOpt / tpu / PreRev 0.0000028510000000000003 s 0.000002834225 s 1.01
add_two / IDefOpt / tpu / PostRev 0.00000275385 s 0.000002741775 s 1.00
add_two / IDefOpt / tpu / BothRev 0.000002844625 s 0.00000283465 s 1.00
add_two / JaXPipe / cpu / Primal 0.000013558 s 0.000008058220000748406 s 1.68
add_two / Jax / cpu / Primal 0.000013607 s 0.000006939919994692901 s 1.96
add_two / HLOOpt / cpu / Primal 0.000013204 s 0.00001038918002450373 s 1.27
add_two / PartOpt / cpu / Primal 0.000012962 s 0.0000068256200029281896 s 1.90
add_two / IPartOpt / cpu / Primal 0.000013292 s 0.000007170440030677128 s 1.85
add_two / DefOpt / cpu / Primal 0.000013158 s 0.000011246620015299414 s 1.17
add_two / IDefOpt / cpu / Primal 0.000012915 s 0.0000069606599936378185 s 1.86
add_two / JaXPipe / cpu / Forward 0.000017392000000000002 s 0.000010616780009513605 s 1.64
add_two / Jax / cpu / Forward 0.000017291 s 0.00001036977996591304 s 1.67
add_two / HLOOpt / cpu / Forward 0.000017663 s 0.000015078660026119903 s 1.17
add_two / PartOpt / cpu / Forward 0.000017645 s 0.000015557280021312182 s 1.13
add_two / IPartOpt / cpu / Forward 0.000017264000000000003 s 0.000010255780007355496 s 1.68
add_two / DefOpt / cpu / Forward 0.000017590000000000003 s 0.00001550278000649996 s 1.13
add_two / IDefOpt / cpu / Forward 0.000017642 s 0.000010459999939484988 s 1.69
add_two / JaXPipe / cpu / PreRev 0.000023521 s 0.000013837839987900223 s 1.70
add_two / JaXPipe / cpu / PostRev 0.000022713 s 0.000013908720056861057 s 1.63
add_two / JaXPipe / cpu / BothRev 0.000023228 s 0.000013828220016876002 s 1.68
add_two / Jax / cpu / BothRev 0.000022885 s 0.000014264560040828656 s 1.60
add_two / HLOOpt / cpu / PreRev 0.000022633 s 0.000014075380013309769 s 1.61
add_two / HLOOpt / cpu / PostRev 0.000022687 s 0.000013675300060640438 s 1.66
add_two / HLOOpt / cpu / BothRev 0.000023034 s 0.000016088479987956817 s 1.43
add_two / PartOpt / cpu / PreRev 0.000023188 s 0.000013798339987260988 s 1.68
add_two / PartOpt / cpu / PostRev 0.000022985 s 0.000013859300015610642 s 1.66
add_two / PartOpt / cpu / BothRev 0.000023175 s 0.000013659700052812696 s 1.70
add_two / IPartOpt / cpu / PreRev 0.000022805 s 0.00001460233999750926 s 1.56
add_two / IPartOpt / cpu / PostRev 0.00002353 s 0.000014303160014605965 s 1.65
add_two / IPartOpt / cpu / BothRev 0.000023506 s 0.00001426799999535433 s 1.65
add_two / DefOpt / cpu / PreRev 0.000023351 s 0.000013933239979451172 s 1.68
add_two / DefOpt / cpu / PostRev 0.000022943 s 0.000014049399933355744 s 1.63
add_two / DefOpt / cpu / BothRev 0.000023435 s 0.000014438600001085431 s 1.62
add_two / IDefOpt / cpu / PreRev 0.000023302 s 0.000014167980016281944 s 1.64
add_two / IDefOpt / cpu / PostRev 0.000023138 s 0.000014254260022426025 s 1.62
add_two / IDefOpt / cpu / BothRev 0.000023508000000000003 s 0.000014405700057977813 s 1.63
add_two / JaXPipe / cpu / Primal 0.000008 s 0.000008058220000748406 s 0.99
add_two / Jax / cpu / Primal 0.000008 s 0.000006939919994692901 s 1.15
add_two / HLOOpt / cpu / Primal 0.000008 s 0.00001038918002450373 s 0.77
add_two / PartOpt / cpu / Primal 0.000008 s 0.0000068256200029281896 s 1.17
add_two / IPartOpt / cpu / Primal 0.000008 s 0.000007170440030677128 s 1.12
add_two / DefOpt / cpu / Primal 0.000008 s 0.000011246620015299414 s 0.71
add_two / IDefOpt / cpu / Primal 0.000008 s 0.0000069606599936378185 s 1.15
add_two / JaXPipe / cpu / Forward 0.000012 s 0.000010616780009513605 s 1.13
add_two / Jax / cpu / Forward 0.000011 s 0.00001036977996591304 s 1.06
add_two / HLOOpt / cpu / Forward 0.000012 s 0.000015078660026119903 s 0.80
add_two / PartOpt / cpu / Forward 0.000011 s 0.000015557280021312182 s 0.71
add_two / IPartOpt / cpu / Forward 0.000012 s 0.000010255780007355496 s 1.17
add_two / DefOpt / cpu / Forward 0.000012 s 0.00001550278000649996 s 0.77
add_two / IDefOpt / cpu / Forward 0.000012 s 0.000010459999939484988 s 1.15
add_two / JaXPipe / cpu / PreRev 0.000015 s 0.000013837839987900223 s 1.08
add_two / JaXPipe / cpu / PostRev 0.000016 s 0.000013908720056861057 s 1.15
add_two / JaXPipe / cpu / BothRev 0.000015 s 0.000013828220016876002 s 1.08
add_two / Jax / cpu / BothRev 0.000015 s 0.000014264560040828656 s 1.05
add_two / HLOOpt / cpu / PreRev 0.000015 s 0.000014075380013309769 s 1.07
add_two / HLOOpt / cpu / PostRev 0.000015 s 0.000013675300060640438 s 1.10
add_two / HLOOpt / cpu / BothRev 0.000017 s 0.000016088479987956817 s 1.06
add_two / PartOpt / cpu / PreRev 0.000015 s 0.000013798339987260988 s 1.09
add_two / PartOpt / cpu / PostRev 0.000015 s 0.000013859300015610642 s 1.08
add_two / PartOpt / cpu / BothRev 0.000016 s 0.000013659700052812696 s 1.17
add_two / IPartOpt / cpu / PreRev 0.000015 s 0.00001460233999750926 s 1.03
add_two / IPartOpt / cpu / PostRev 0.000015 s 0.000014303160014605965 s 1.05
add_two / IPartOpt / cpu / BothRev 0.000017 s 0.00001426799999535433 s 1.19
add_two / DefOpt / cpu / PreRev 0.000015 s 0.000013933239979451172 s 1.08
add_two / DefOpt / cpu / PostRev 0.000015 s 0.000014049399933355744 s 1.07
add_two / DefOpt / cpu / BothRev 0.000015 s 0.000014438600001085431 s 1.04
add_two / IDefOpt / cpu / PreRev 0.000015 s 0.000014167980016281944 s 1.06
add_two / IDefOpt / cpu / PostRev 0.000015 s 0.000014254260022426025 s 1.05
add_two / IDefOpt / cpu / BothRev 0.000016 s 0.000014405700057977813 s 1.11
cache / JaXPipe / cpu / Primal 0.000007187820001490763 s 0.000006895420019645826 s 1.04
cache / Jax / cpu / Primal 0.000008135559992297204 s 0.000007318139987546601 s 1.11
cache / HLOOpt / cpu / Primal 0.000007411020014842506 s 0.000006648640010098461 s 1.11
cache / PartOpt / cpu / Primal 0.000006961499984754482 s 0.000006451139988712384 s 1.08
cache / IPartOpt / cpu / Primal 0.000006934580014785752 s 0.00000616145996900741 s 1.13
cache / DefOpt / cpu / Primal 0.000007759980007904232 s 0.000006204340015756316 s 1.25
cache / IDefOpt / cpu / Primal 0.000006615940010306076 s 0.000006574280014319811 s 1.01
cache / JaXPipe / cpu / Forward 0.000016196240012504858 s 0.000014936860034140408 s 1.08
cache / Jax / cpu / Forward 0.000014648099986516172 s 0.000015530239970757975 s 0.94
cache / HLOOpt / cpu / Forward 0.000020189340011711467 s 0.00001966315998288337 s 1.03
cache / PartOpt / cpu / Forward 0.000019482580009935192 s 0.0000200468400180398 s 0.97
cache / IPartOpt / cpu / Forward 0.000015469039944946416 s 0.000015144639974096208 s 1.02
cache / DefOpt / cpu / Forward 0.00002072656000564166 s 0.000019512480021148805 s 1.06
cache / IDefOpt / cpu / Forward 0.000014792160018259892 s 0.000014492459995381068 s 1.02
cache / JaXPipe / cpu / PreRev 0.000016726479998396826 s 0.00001606625997737865 s 1.04
cache / JaXPipe / cpu / PostRev 0.00002074416001960344 s 0.00002136004000021785 s 0.97
cache / JaXPipe / cpu / BothRev 0.00002057165997939592 s 0.000016573820003031868 s 1.24
cache / Jax / cpu / BothRev 0.000028148500014140155 s 0.00002134235994162737 s 1.32
cache / HLOOpt / cpu / PreRev 0.000015712280010120594 s 0.000015764980034873587 s 1.00
cache / HLOOpt / cpu / PostRev 0.000019704319993252283 s 0.00001578153998707421 s 1.25
cache / HLOOpt / cpu / BothRev 0.00001850266001383716 s 0.000018151820013372346 s 1.02
cache / PartOpt / cpu / PreRev 0.000016094819957288565 s 0.000015171119976002956 s 1.06
cache / PartOpt / cpu / PostRev 0.00002121802000147 s 0.00002224828001089918 s 0.95
cache / PartOpt / cpu / BothRev 0.00001608001996828534 s 0.000015535459942839226 s 1.04
cache / IPartOpt / cpu / PreRev 0.000017044460018951213 s 0.000015548999990642186 s 1.10
cache / IPartOpt / cpu / PostRev 0.000022518000023410426 s 0.000021092159995532716 s 1.07
cache / IPartOpt / cpu / BothRev 0.00001627611996809719 s 0.0000211883000247326 s 0.77
cache / DefOpt / cpu / PreRev 0.00001670541998464614 s 0.000015965940010573833 s 1.05
cache / DefOpt / cpu / PostRev 0.000017998159983108053 s 0.00001623538001695124 s 1.11
cache / DefOpt / cpu / BothRev 0.00001621139998860599 s 0.00001580041998749948 s 1.03
cache / IDefOpt / cpu / PreRev 0.000016785060015536146 s 0.00001638905999243434 s 1.02
cache / IDefOpt / cpu / PostRev 0.00001637185996514745 s 0.000017023899972627986 s 0.96
cache / IDefOpt / cpu / BothRev 0.000016457280044051003 s 0.000015551980022792123 s 1.06
cache / JaXPipe / tpu / Primal 0.000002459975 s 0.00000247365 s 0.99
cache / Jax / tpu / Primal 0.000002466125 s 0.0000024525250000000003 s 1.01
cache / HLOOpt / tpu / Primal 0.00000245335 s 0.000002493575 s 0.98
cache / PartOpt / tpu / Primal 0.00000247065 s 0.000002479275 s 1.00
cache / IPartOpt / tpu / Primal 0.0000024598 s 0.000002469825 s 1.00
cache / DefOpt / tpu / Primal 0.0000024475 s 0.0000024529 s 1.00
cache / IDefOpt / tpu / Primal 0.0000024547 s 0.000002479875 s 0.99
cache / JaXPipe / tpu / Forward 0.0000035674250000000004 s 0.0000035343750000000004 s 1.01
cache / Jax / tpu / Forward 0.000003540125 s 0.0000035359249999999995 s 1.00
cache / HLOOpt / tpu / Forward 0.00000356165 s 0.000003560325 s 1.00
cache / PartOpt / tpu / Forward 0.00000353175 s 0.000003533125 s 1.00
cache / IPartOpt / tpu / Forward 0.00000356985 s 0.0000035517749999999995 s 1.01
cache / DefOpt / tpu / Forward 0.0000035425 s 0.000003531575 s 1.00
cache / IDefOpt / tpu / Forward 0.0000035428 s 0.000003548275 s 1.00
cache / JaXPipe / tpu / PreRev 0.000003938175 s 0.00000497815 s 0.79
cache / JaXPipe / tpu / PostRev 0.0000049548 s 0.0000049616 s 1.00
cache / JaXPipe / tpu / BothRev 0.000003948 s 0.0000049991 s 0.79
cache / Jax / tpu / BothRev 0.000004981525 s 0.00000498965 s 1.00
cache / HLOOpt / tpu / PreRev 0.0000039439000000000005 s 0.00000395315 s 1.00
cache / HLOOpt / tpu / PostRev 0.0000041357 s 0.000004137525000000001 s 1.00
cache / HLOOpt / tpu / BothRev 0.000003937375 s 0.000003955225 s 1.00
cache / PartOpt / tpu / PreRev 0.000004135825 s 0.0000049973 s 0.83
cache / PartOpt / tpu / PostRev 0.000004965775000000001 s 0.0000049899 s 1.00
cache / PartOpt / tpu / BothRev 0.000004120175 s 0.00000498855 s 0.83
cache / IPartOpt / tpu / PreRev 0.00000393525 s 0.0000049985250000000006 s 0.79
cache / IPartOpt / tpu / PostRev 0.000004958675 s 0.000004992775 s 0.99
cache / IPartOpt / tpu / BothRev 0.0000039413 s 0.0000049979 s 0.79
cache / DefOpt / tpu / PreRev 0.00000410795 s 0.000004976525 s 0.83
cache / DefOpt / tpu / PostRev 0.000003939975 s 0.000005009625 s 0.79
cache / DefOpt / tpu / BothRev 0.0000041134750000000005 s 0.000004992949999999999 s 0.82
cache / IDefOpt / tpu / PreRev 0.00000393345 s 0.00000500705 s 0.79
cache / IDefOpt / tpu / PostRev 0.000004137575 s 0.000004990474999999999 s 0.83
cache / IDefOpt / tpu / BothRev 0.0000039373 s 0.0000050122 s 0.79
cache / JaXPipe / cpu / Primal 0.000012557 s 0.000006895420019645826 s 1.82
cache / Jax / cpu / Primal 0.000012962 s 0.000007318139987546601 s 1.77
cache / HLOOpt / cpu / Primal 0.000012475 s 0.000006648640010098461 s 1.88
cache / PartOpt / cpu / Primal 0.000012821 s 0.000006451139988712384 s 1.99
cache / IPartOpt / cpu / Primal 0.000012621 s 0.00000616145996900741 s 2.05
cache / DefOpt / cpu / Primal 0.00001239 s 0.000006204340015756316 s 2.00
cache / IDefOpt / cpu / Primal 0.00001272 s 0.000006574280014319811 s 1.93
cache / JaXPipe / cpu / Forward 0.000016527 s 0.000014936860034140408 s 1.11
cache / Jax / cpu / Forward 0.000016747 s 0.000015530239970757975 s 1.08
cache / HLOOpt / cpu / Forward 0.000016500999999999997 s 0.00001966315998288337 s 0.84
cache / PartOpt / cpu / Forward 0.000016774000000000002 s 0.0000200468400180398 s 0.84
cache / IPartOpt / cpu / Forward 0.000016831 s 0.000015144639974096208 s 1.11
cache / DefOpt / cpu / Forward 0.00001681 s 0.000019512480021148805 s 0.86
cache / IDefOpt / cpu / Forward 0.000016904 s 0.000014492459995381068 s 1.17
cache / JaXPipe / cpu / PreRev 0.000027741 s 0.00001606625997737865 s 1.73
cache / JaXPipe / cpu / PostRev 0.00004194 s 0.00002136004000021785 s 1.96
cache / JaXPipe / cpu / BothRev 0.000026432 s 0.000016573820003031868 s 1.59
cache / Jax / cpu / BothRev 0.000032272000000000005 s 0.00002134235994162737 s 1.51
cache / HLOOpt / cpu / PreRev 0.000023161 s 0.000015764980034873587 s 1.47
cache / HLOOpt / cpu / PostRev 0.000027202 s 0.00001578153998707421 s 1.72
cache / HLOOpt / cpu / BothRev 0.000021119 s 0.000018151820013372346 s 1.16
cache / PartOpt / cpu / PreRev 0.000026392 s 0.000015171119976002956 s 1.74
cache / PartOpt / cpu / PostRev 0.000030378 s 0.00002224828001089918 s 1.37
cache / PartOpt / cpu / BothRev 0.000024806000000000003 s 0.000015535459942839226 s 1.60
cache / IPartOpt / cpu / PreRev 0.000025086 s 0.000015548999990642186 s 1.61
cache / IPartOpt / cpu / PostRev 0.000027394 s 0.000021092159995532716 s 1.30
cache / IPartOpt / cpu / BothRev 0.000024278 s 0.0000211883000247326 s 1.15
cache / DefOpt / cpu / PreRev 0.000025918 s 0.000015965940010573833 s 1.62
cache / DefOpt / cpu / PostRev 0.000035470000000000004 s 0.00001623538001695124 s 2.18
cache / DefOpt / cpu / BothRev 0.000025059 s 0.00001580041998749948 s 1.59
cache / IDefOpt / cpu / PreRev 0.000030263 s 0.00001638905999243434 s 1.85
cache / IDefOpt / cpu / PostRev 0.000024674 s 0.000017023899972627986 s 1.45
cache / IDefOpt / cpu / BothRev 0.000026762 s 0.000015551980022792123 s 1.72
cache / JaXPipe / cpu / Primal 0.000008 s 0.000006895420019645826 s 1.16
cache / Jax / cpu / Primal 0.000008 s 0.000007318139987546601 s 1.09
cache / HLOOpt / cpu / Primal 0.000008 s 0.000006648640010098461 s 1.20
cache / PartOpt / cpu / Primal 0.000008 s 0.000006451139988712384 s 1.24
cache / IPartOpt / cpu / Primal 0.000008 s 0.00000616145996900741 s 1.30
cache / DefOpt / cpu / Primal 0.000008 s 0.000006204340015756316 s 1.29
cache / IDefOpt / cpu / Primal 0.000008 s 0.000006574280014319811 s 1.22
cache / JaXPipe / cpu / Forward 0.000011 s 0.000014936860034140408 s 0.74
cache / Jax / cpu / Forward 0.000011 s 0.000015530239970757975 s 0.71
cache / HLOOpt / cpu / Forward 0.000035999999999999994 s 0.00001966315998288337 s 1.83
cache / PartOpt / cpu / Forward 0.00001 s 0.0000200468400180398 s 0.50
cache / IPartOpt / cpu / Forward 0.000052 s 0.000015144639974096208 s 3.43
cache / DefOpt / cpu / Forward 0.000011 s 0.000019512480021148805 s 0.56
cache / IDefOpt / cpu / Forward 0.00001 s 0.000014492459995381068 s 0.69
cache / JaXPipe / cpu / PreRev 0.00001 s 0.00001606625997737865 s 0.62
cache / JaXPipe / cpu / PostRev 0.000052 s 0.00002136004000021785 s 2.43
cache / JaXPipe / cpu / BothRev 0.000011 s 0.000016573820003031868 s 0.66
cache / Jax / cpu / BothRev 0.000038 s 0.00002134235994162737 s 1.78
cache / HLOOpt / cpu / PreRev 0.000039 s 0.000015764980034873587 s 2.47
cache / HLOOpt / cpu / PostRev 0.000011 s 0.00001578153998707421 s 0.70
cache / HLOOpt / cpu / BothRev 0.00001 s 0.000018151820013372346 s 0.55
cache / PartOpt / cpu / PreRev 0.000011 s 0.000015171119976002956 s 0.73
cache / PartOpt / cpu / PostRev 0.000012 s 0.00002224828001089918 s 0.54
cache / PartOpt / cpu / BothRev 0.00001 s 0.000015535459942839226 s 0.64
cache / IPartOpt / cpu / PreRev 0.00003 s 0.000015548999990642186 s 1.93
cache / IPartOpt / cpu / PostRev 0.000014 s 0.000021092159995532716 s 0.66
cache / IPartOpt / cpu / BothRev 0.000011 s 0.0000211883000247326 s 0.52
cache / DefOpt / cpu / PreRev 0.000011 s 0.000015965940010573833 s 0.69
cache / DefOpt / cpu / PostRev 0.000011 s 0.00001623538001695124 s 0.68
cache / DefOpt / cpu / BothRev 0.000011 s 0.00001580041998749948 s 0.70
cache / IDefOpt / cpu / PreRev 0.000011 s 0.00001638905999243434 s 0.67
cache / IDefOpt / cpu / PostRev 0.000011 s 0.000017023899972627986 s 0.65
cache / IDefOpt / cpu / BothRev 0.000035999999999999994 s 0.000015551980022792123 s 2.31
Concat / JaXPipe / cpu / Primal 0.000007922360009615658 s 0.000007405199985441868 s 1.07
Concat / Jax / cpu / Primal 0.000007531280025432352 s 0.000006934959974387311 s 1.09
Concat / HLOOpt / cpu / Primal 0.000007579520024592057 s 0.000009673719987404185 s 0.78
Concat / PartOpt / cpu / Primal 0.000007660500041311024 s 0.000007012420037426637 s 1.09
Concat / IPartOpt / cpu / Primal 0.000007573780021630228 s 0.000007011279994912911 s 1.08
Concat / DefOpt / cpu / Primal 0.00000975070002823486 s 0.000010873839983105428 s 0.90
Concat / IDefOpt / cpu / Primal 0.000007531779974669917 s 0.000006547859966303804 s 1.15
Concat / JaXPipe / cpu / Forward 0.000010926539971478633 s 0.000010484799977348304 s 1.04
Concat / Jax / cpu / Forward 0.000011008120027327096 s 0.000010528139964662842 s 1.05
Concat / HLOOpt / cpu / Forward 0.000015274899942596677 s 0.000014631059975727112 s 1.04
Concat / PartOpt / cpu / Forward 0.000015417600006912835 s 0.000014466300008280088 s 1.07
Concat / IPartOpt / cpu / Forward 0.00001114689996938978 s 0.000010190519969910384 s 1.09
Concat / DefOpt / cpu / Forward 0.000015710439956819755 s 0.000015150820008784649 s 1.04
Concat / IDefOpt / cpu / Forward 0.000011621379999269264 s 0.000009853080027824035 s 1.18
Concat / JaXPipe / cpu / PreRev 0.000012790140035576769 s 0.000012491000043155508 s 1.02
Concat / JaXPipe / cpu / PostRev 0.000012781800014636248 s 0.000011773679998441369 s 1.09
Concat / JaXPipe / cpu / BothRev 0.00001258547999896109 s 0.000014773040029467665 s 0.85
Concat / Jax / cpu / BothRev 0.000012720940012513894 s 0.000011373940033081454 s 1.12
Concat / HLOOpt / cpu / PreRev 0.000012468919985622053 s 0.000011325240011501592 s 1.10
Concat / HLOOpt / cpu / PostRev 0.000016497419983352303 s 0.000015252020057232583 s 1.08
Concat / HLOOpt / cpu / BothRev 0.000014755060010429589 s 0.00001311264002652024 s 1.13
Concat / PartOpt / cpu / PreRev 0.000012810680018446874 s 0.000011417259984227712 s 1.12
Concat / PartOpt / cpu / PostRev 0.000012634620006792829 s 0.000011336720026520196 s 1.11
Concat / PartOpt / cpu / BothRev 0.00001255560002391576 s 0.00001167289998193155 s 1.08
Concat / IPartOpt / cpu / PreRev 0.00001723695994769514 s 0.000011321219981255126 s 1.52
Concat / IPartOpt / cpu / PostRev 0.000012568639976962005 s 0.00001134046001425304 s 1.11
Concat / IPartOpt / cpu / BothRev 0.000012109819999750473 s 0.000011186279998582904 s 1.08
Concat / DefOpt / cpu / PreRev 0.00001293247998546576 s 0.000011393260001568704 s 1.14
Concat / DefOpt / cpu / PostRev 0.00001219276003212144 s 0.000012159280013293028 s 1.00
Concat / DefOpt / cpu / BothRev 0.00001216766001562064 s 0.000011149420024594291 s 1.09
Concat / IDefOpt / cpu / PreRev 0.000012618579967238477 s 0.0000115311199806456 s 1.09
Concat / IDefOpt / cpu / PostRev 0.000012708319973171456 s 0.000012273759984964272 s 1.04
Concat / IDefOpt / cpu / BothRev 0.000012652280001930194 s 0.000011928760068258273 s 1.06
Concat / JaXPipe / tpu / Primal 0.0000015394 s 0.000001517425 s 1.01
Concat / Jax / tpu / Primal 0.000001524125 s 0.0000015343 s 0.99
Concat / HLOOpt / tpu / Primal 0.000001537875 s 0.000001517125 s 1.01
Concat / PartOpt / tpu / Primal 0.0000015332000000000002 s 0.00000152735 s 1.00
Concat / IPartOpt / tpu / Primal 0.00000153685 s 0.00000152425 s 1.01
Concat / DefOpt / tpu / Primal 0.0000015324499999999998 s 0.000001530525 s 1.00
Concat / IDefOpt / tpu / Primal 0.0000015353 s 0.0000015259500000000002 s 1.01
Concat / JaXPipe / tpu / Forward 0.0000015744250000000002 s 0.000001573225 s 1.00
Concat / Jax / tpu / Forward 0.000001556775 s 0.000001535675 s 1.01
Concat / HLOOpt / tpu / Forward 0.000001570875 s 0.000001570275 s 1.00
Concat / PartOpt / tpu / Forward 0.000001554625 s 0.000001538 s 1.01
Concat / IPartOpt / tpu / Forward 0.0000015683749999999998 s 0.000001569825 s 1.00
Concat / DefOpt / tpu / Forward 0.0000015493499999999998 s 0.000001548975 s 1.00
Concat / IDefOpt / tpu / Forward 0.00000157525 s 0.0000015839 s 0.99
Concat / JaXPipe / tpu / PreRev 0.0000020177 s 0.000001999975 s 1.01
Concat / JaXPipe / tpu / PostRev 0.000002080775 s 0.00000208545 s 1.00
Concat / JaXPipe / tpu / BothRev 0.000002008775 s 0.0000020049 s 1.00
Concat / Jax / tpu / BothRev 0.0000020797 s 0.000002071225 s 1.00
Concat / HLOOpt / tpu / PreRev 0.000002012075 s 0.000001991875 s 1.01
Concat / HLOOpt / tpu / PostRev 0.000002072075 s 0.0000020801500000000003 s 1.00
Concat / HLOOpt / tpu / BothRev 0.000002015075 s 0.0000019980750000000003 s 1.01
Concat / PartOpt / tpu / PreRev 0.000002068625 s 0.00000206985 s 1.00
Concat / PartOpt / tpu / PostRev 0.000002015225 s 0.0000019981 s 1.01
Concat / PartOpt / tpu / BothRev 0.00000207515 s 0.000002085225 s 1.00
Concat / IPartOpt / tpu / PreRev 0.00000200555 s 0.000001994425 s 1.01
Concat / IPartOpt / tpu / PostRev 0.00000207525 s 0.00000206995 s 1.00
Concat / IPartOpt / tpu / BothRev 0.00000201485 s 0.0000019993 s 1.01
Concat / DefOpt / tpu / PreRev 0.000002072225 s 0.0000020732249999999995 s 1.00
Concat / DefOpt / tpu / PostRev 0.0000020147 s 0.000001999575 s 1.01
Concat / DefOpt / tpu / BothRev 0.0000020865 s 0.00000207855 s 1.00
Concat / IDefOpt / tpu / PreRev 0.000002009425 s 0.0000020087 s 1.00
Concat / IDefOpt / tpu / PostRev 0.00000207655 s 0.0000020645250000000004 s 1.01
Concat / IDefOpt / tpu / BothRev 0.000002004375 s 0.000001999175 s 1.00
Concat / JaXPipe / cpu / Primal 0.000012947 s 0.000007405199985441868 s 1.75
Concat / Jax / cpu / Primal 0.000012666 s 0.000006934959974387311 s 1.83
Concat / HLOOpt / cpu / Primal 0.000012676 s 0.000009673719987404185 s 1.31
Concat / PartOpt / cpu / Primal 0.000012872 s 0.000007012420037426637 s 1.84
Concat / IPartOpt / cpu / Primal 0.000012546 s 0.000007011279994912911 s 1.79
Concat / DefOpt / cpu / Primal 0.000012749 s 0.000010873839983105428 s 1.17
Concat / IDefOpt / cpu / Primal 0.000012715 s 0.000006547859966303804 s 1.94
Concat / JaXPipe / cpu / Forward 0.000017473 s 0.000010484799977348304 s 1.67
Concat / Jax / cpu / Forward 0.00001743 s 0.000010528139964662842 s 1.66
Concat / HLOOpt / cpu / Forward 0.00001703 s 0.000014631059975727112 s 1.16
Concat / PartOpt / cpu / Forward 0.000017122 s 0.000014466300008280088 s 1.18
Concat / IPartOpt / cpu / Forward 0.000016907999999999998 s 0.000010190519969910384 s 1.66
Concat / DefOpt / cpu / Forward 0.000016726000000000002 s 0.000015150820008784649 s 1.10
Concat / IDefOpt / cpu / Forward 0.000017157 s 0.000009853080027824035 s 1.74
Concat / JaXPipe / cpu / PreRev 0.000019528 s 0.000012491000043155508 s 1.56
Concat / JaXPipe / cpu / PostRev 0.000019527 s 0.000011773679998441369 s 1.66
Concat / JaXPipe / cpu / BothRev 0.000019387000000000003 s 0.000014773040029467665 s 1.31
Concat / Jax / cpu / BothRev 0.000019356 s 0.000011373940033081454 s 1.70
Concat / HLOOpt / cpu / PreRev 0.000019237 s 0.000011325240011501592 s 1.70
Concat / HLOOpt / cpu / PostRev 0.000019366 s 0.000015252020057232583 s 1.27
Concat / HLOOpt / cpu / BothRev 0.000019507 s 0.00001311264002652024 s 1.49
Concat / PartOpt / cpu / PreRev 0.000018996 s 0.000011417259984227712 s 1.66
Concat / PartOpt / cpu / PostRev 0.000019437 s 0.000011336720026520196 s 1.71
Concat / PartOpt / cpu / BothRev 0.000018765 s 0.00001167289998193155 s 1.61
Concat / IPartOpt / cpu / PreRev 0.000019237 s 0.000011321219981255126 s 1.70
Concat / IPartOpt / cpu / PostRev 0.000019126 s 0.00001134046001425304 s 1.69
Concat / IPartOpt / cpu / BothRev 0.000019257 s 0.000011186279998582904 s 1.72
Concat / DefOpt / cpu / PreRev 0.000019098 s 0.000011393260001568704 s 1.68
Concat / DefOpt / cpu / PostRev 0.00001885 s 0.000012159280013293028 s 1.55
Concat / DefOpt / cpu / BothRev 0.00001902 s 0.000011149420024594291 s 1.71
Concat / IDefOpt / cpu / PreRev 0.000019097 s 0.0000115311199806456 s 1.66
Concat / IDefOpt / cpu / PostRev 0.000019212 s 0.000012273759984964272 s 1.57
Concat / IDefOpt / cpu / BothRev 0.00001912 s 0.000011928760068258273 s 1.60
Concat / JaXPipe / cpu / Primal 0.000008 s 0.000007405199985441868 s 1.08
Concat / Jax / cpu / Primal 0.000008 s 0.000006934959974387311 s 1.15
Concat / HLOOpt / cpu / Primal 0.000008 s 0.000009673719987404185 s 0.83
Concat / PartOpt / cpu / Primal 0.000008999999999999999 s 0.000007012420037426637 s 1.28
Concat / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000007011279994912911 s 1.28
Concat / DefOpt / cpu / Primal 0.000008 s 0.000010873839983105428 s 0.74
Concat / IDefOpt / cpu / Primal 0.000008 s 0.000006547859966303804 s 1.22
Concat / JaXPipe / cpu / Forward 0.000012 s 0.000010484799977348304 s 1.14
Concat / Jax / cpu / Forward 0.000012 s 0.000010528139964662842 s 1.14
Concat / HLOOpt / cpu / Forward 0.000012 s 0.000014631059975727112 s 0.82
Concat / PartOpt / cpu / Forward 0.000013 s 0.000014466300008280088 s 0.90
Concat / IPartOpt / cpu / Forward 0.000011 s 0.000010190519969910384 s 1.08
Concat / DefOpt / cpu / Forward 0.000011 s 0.000015150820008784649 s 0.73
Concat / IDefOpt / cpu / Forward 0.000012 s 0.000009853080027824035 s 1.22
Concat / JaXPipe / cpu / PreRev 0.000014 s 0.000012491000043155508 s 1.12
Concat / JaXPipe / cpu / PostRev 0.000013 s 0.000011773679998441369 s 1.10
Concat / JaXPipe / cpu / BothRev 0.000014 s 0.000014773040029467665 s 0.95
Concat / Jax / cpu / BothRev 0.000013 s 0.000011373940033081454 s 1.14
Concat / HLOOpt / cpu / PreRev 0.000013 s 0.000011325240011501592 s 1.15
Concat / HLOOpt / cpu / PostRev 0.000014 s 0.000015252020057232583 s 0.92
Concat / HLOOpt / cpu / BothRev 0.000014 s 0.00001311264002652024 s 1.07
Concat / PartOpt / cpu / PreRev 0.000014 s 0.000011417259984227712 s 1.23
Concat / PartOpt / cpu / PostRev 0.000013 s 0.000011336720026520196 s 1.15
Concat / PartOpt / cpu / BothRev 0.000014 s 0.00001167289998193155 s 1.20
Concat / IPartOpt / cpu / PreRev 0.000013 s 0.000011321219981255126 s 1.15
Concat / IPartOpt / cpu / PostRev 0.000013 s 0.00001134046001425304 s 1.15
Concat / IPartOpt / cpu / BothRev 0.000014 s 0.000011186279998582904 s 1.25
Concat / DefOpt / cpu / PreRev 0.000013 s 0.000011393260001568704 s 1.14
Concat / DefOpt / cpu / PostRev 0.000014 s 0.000012159280013293028 s 1.15
Concat / DefOpt / cpu / BothRev 0.000013 s 0.000011149420024594291 s 1.17
Concat / IDefOpt / cpu / PreRev 0.000013 s 0.0000115311199806456 s 1.13
Concat / IDefOpt / cpu / PostRev 0.000014 s 0.000012273759984964272 s 1.14
Concat / IDefOpt / cpu / BothRev 0.000014 s 0.000011928760068258273 s 1.17
const_scatter / JaXPipe / cpu / Primal 0.000007122739989426919 s 0.000007382560042969999 s 0.96
const_scatter / Jax / cpu / Primal 0.000007276380001712823 s 0.000007396020000669523 s 0.98
const_scatter / HLOOpt / cpu / Primal 0.000007035779990474112 s 0.000008004739993339171 s 0.88
const_scatter / PartOpt / cpu / Primal 0.000007879360036895378 s 0.000006350180010485928 s 1.24
const_scatter / IPartOpt / cpu / Primal 0.000007192259972725878 s 0.000006601120003324468 s 1.09
const_scatter / DefOpt / cpu / Primal 0.000006929580031282967 s 0.000011223479978070827 s 0.62
const_scatter / IDefOpt / cpu / Primal 0.0000066157400124211566 s 0.000007006859996181447 s 0.94
const_scatter / JaXPipe / cpu / Forward 0.000011430920012571731 s 0.000009879959980025888 s 1.16
const_scatter / Jax / cpu / Forward 0.000011180160045114465 s 0.000010793700021167753 s 1.04
const_scatter / HLOOpt / cpu / Forward 0.000015281920004781567 s 0.000009937820004779497 s 1.54
const_scatter / PartOpt / cpu / Forward 0.000015334220033764723 s 0.000014244259991755826 s 1.08
const_scatter / IPartOpt / cpu / Forward 0.00001025938001475879 s 0.000010233359962512622 s 1.00
const_scatter / DefOpt / cpu / Forward 0.000014614200026699108 s 0.000014041699960216648 s 1.04
const_scatter / IDefOpt / cpu / Forward 0.000009836059989538626 s 0.00000954062000346312 s 1.03
const_scatter / JaXPipe / cpu / PreRev 0.0003028447199721 s 0.0003037084600055 s 1.00
const_scatter / JaXPipe / cpu / PostRev 0.0002912651600308 s 0.0002919454400034 s 1.00
const_scatter / JaXPipe / cpu / BothRev 0.000286187040001 s 0.0002827461599645 s 1.01
const_scatter / Jax / cpu / BothRev 0.0002863338400129 s 0.0002841475600234 s 1.01
const_scatter / HLOOpt / cpu / PreRev 0.0002870911600439 s 0.0002834839600018 s 1.01
const_scatter / HLOOpt / cpu / PostRev 0.0002934347400423 s 0.0002885376600443 s 1.02
const_scatter / HLOOpt / cpu / BothRev 0.0002892988799794 s 0.0002866564400028 s 1.01
const_scatter / PartOpt / cpu / PreRev 0.0002894662800099 s 0.0002885445599804 s 1.00
const_scatter / PartOpt / cpu / PostRev 0.0002945420399464 s 0.0002914730799966 s 1.01
const_scatter / PartOpt / cpu / BothRev 0.0002853228200001 s 0.000286857999954 s 0.99
const_scatter / IPartOpt / cpu / PreRev 0.0002926573399872 s 0.0002903088000402 s 1.01
const_scatter / IPartOpt / cpu / PostRev 0.0002935012399848 s 0.0002913657600038 s 1.01
const_scatter / IPartOpt / cpu / BothRev 0.0002850476799994 s 0.0002854645200295 s 1.00
const_scatter / DefOpt / cpu / PreRev 0.0002887255399946 s 0.0002907582999796 s 0.99
const_scatter / DefOpt / cpu / PostRev 0.0002912958800243 s 0.0002929033599957 s 0.99
const_scatter / DefOpt / cpu / BothRev 0.0002866558599998 s 0.0002860883799894 s 1.00
const_scatter / IDefOpt / cpu / PreRev 0.0002925628999946 s 0.0002909318000274 s 1.01
const_scatter / IDefOpt / cpu / PostRev 0.0002904200799821 s 0.0002944353399379 s 0.99
const_scatter / IDefOpt / cpu / BothRev 0.0002862665400334 s 0.0002834542200616 s 1.01
const_scatter / JaXPipe / tpu / Primal 0.00000380315 s 0.000003797475 s 1.00
const_scatter / Jax / tpu / Primal 0.00000382235 s 0.000003825625 s 1.00
const_scatter / HLOOpt / tpu / Primal 9.48825e-7 s 9.42575e-7 s 1.01
const_scatter / PartOpt / tpu / Primal 0.000003808 s 0.000003817975 s 1.00
const_scatter / IPartOpt / tpu / Primal 0.00000379015 s 0.00000380185 s 1.00
const_scatter / DefOpt / tpu / Primal 9.5915e-7 s 9.674e-7 s 0.99
const_scatter / IDefOpt / tpu / Primal 9.37725e-7 s 9.49775e-7 s 0.99
const_scatter / JaXPipe / tpu / Forward 0.0000019227 s 0.0000019167 s 1.00
const_scatter / Jax / tpu / Forward 0.000006503775000000001 s 0.000006496374999999999 s 1.00
const_scatter / HLOOpt / tpu / Forward 0.000001920275 s 0.000001917425 s 1.00
const_scatter / PartOpt / tpu / Forward 0.00000194005 s 0.000001947225 s 1.00
const_scatter / IPartOpt / tpu / Forward 0.000001922625 s 0.000001909375 s 1.01
const_scatter / DefOpt / tpu / Forward 0.00000193745 s 0.000001939725 s 1.00
const_scatter / IDefOpt / tpu / Forward 0.000001915925 s 0.0000019236 s 1.00
const_scatter / JaXPipe / tpu / PreRev 0.000004299325 s 0.000004310600000000001 s 1.00
const_scatter / JaXPipe / tpu / PostRev 0.000006606999999999999 s 0.000006686775000000001 s 0.99
const_scatter / JaXPipe / tpu / BothRev 0.00000431605 s 0.0000042936 s 1.01
const_scatter / Jax / tpu / BothRev 0.000006620049999999999 s 0.0000066556 s 0.99
const_scatter / HLOOpt / tpu / PreRev 0.000004311675 s 0.0000042991 s 1.00
const_scatter / HLOOpt / tpu / PostRev 0.000004301675 s 0.000004303425 s 1.00
const_scatter / HLOOpt / tpu / BothRev 0.000004299350000000001 s 0.00000429245 s 1.00
const_scatter / PartOpt / tpu / PreRev 0.000004296525 s 0.00000430805 s 1.00
const_scatter / PartOpt / tpu / PostRev 0.0000066035 s 0.000006638150000000001 s 0.99
const_scatter / PartOpt / tpu / BothRev 0.000004284825 s 0.000004300725 s 1.00
const_scatter / IPartOpt / tpu / PreRev 0.0000043197 s 0.000004294425 s 1.01
const_scatter / IPartOpt / tpu / PostRev 0.000006614075000000001 s 0.000006673175 s 0.99
const_scatter / IPartOpt / tpu / BothRev 0.000004297150000000001 s 0.0000042975 s 1.00
const_scatter / DefOpt / tpu / PreRev 0.0000042944 s 0.0000043108 s 1.00
const_scatter / DefOpt / tpu / PostRev 0.000004299525 s 0.0000043050750000000005 s 1.00
const_scatter / DefOpt / tpu / BothRev 0.00000429405 s 0.000004284775 s 1.00
const_scatter / IDefOpt / tpu / PreRev 0.000004301575 s 0.0000042975 s 1.00
const_scatter / IDefOpt / tpu / PostRev 0.000004295275 s 0.0000043103 s 1.00
const_scatter / IDefOpt / tpu / BothRev 0.0000042969 s 0.000004288525000000001 s 1.00
const_scatter / JaXPipe / cpu / Primal 0.000012643 s 0.000007382560042969999 s 1.71
const_scatter / Jax / cpu / Primal 0.000012779 s 0.000007396020000669523 s 1.73
const_scatter / HLOOpt / cpu / Primal 0.000012651 s 0.000008004739993339171 s 1.58
const_scatter / PartOpt / cpu / Primal 0.000012706 s 0.000006350180010485928 s 2.00
const_scatter / IPartOpt / cpu / Primal 0.0000124 s 0.000006601120003324468 s 1.88
const_scatter / DefOpt / cpu / Primal 0.000012262 s 0.000011223479978070827 s 1.09
const_scatter / IDefOpt / cpu / Primal 0.000012582 s 0.000007006859996181447 s 1.80
const_scatter / JaXPipe / cpu / Forward 0.000016873 s 0.000009879959980025888 s 1.71
const_scatter / Jax / cpu / Forward 0.000016628 s 0.000010793700021167753 s 1.54
const_scatter / HLOOpt / cpu / Forward 0.000016686 s 0.000009937820004779497 s 1.68
const_scatter / PartOpt / cpu / Forward 0.000016608 s 0.000014244259991755826 s 1.17
const_scatter / IPartOpt / cpu / Forward 0.000016907999999999998 s 0.000010233359962512622 s 1.65
const_scatter / DefOpt / cpu / Forward 0.0000166 s 0.000014041699960216648 s 1.18
const_scatter / IDefOpt / cpu / Forward 0.000016606 s 0.00000954062000346312 s 1.74
const_scatter / JaXPipe / cpu / PreRev 0.000518132 s 0.0003037084600055 s 1.71
const_scatter / JaXPipe / cpu / PostRev 0.000540663 s 0.0002919454400034 s 1.85
const_scatter / JaXPipe / cpu / BothRev 0.000518236 s 0.0002827461599645 s 1.83
const_scatter / Jax / cpu / BothRev 0.0005303419999999 s 0.0002841475600234 s 1.87
const_scatter / HLOOpt / cpu / PreRev 0.0005480429999999 s 0.0002834839600018 s 1.93
const_scatter / HLOOpt / cpu / PostRev 0.000530234 s 0.0002885376600443 s 1.84
const_scatter / HLOOpt / cpu / BothRev 0.000520466 s 0.0002866564400028 s 1.82
const_scatter / PartOpt / cpu / PreRev 0.00051482 s 0.0002885445599804 s 1.78
const_scatter / PartOpt / cpu / PostRev 0.000523116 s 0.0002914730799966 s 1.79
const_scatter / PartOpt / cpu / BothRev 0.000502455 s 0.000286857999954 s 1.75
const_scatter / IPartOpt / cpu / PreRev 0.000514877 s 0.0002903088000402 s 1.77
const_scatter / IPartOpt / cpu / PostRev 0.000523524 s 0.0002913657600038 s 1.80
const_scatter / IPartOpt / cpu / BothRev 0.000521426 s 0.0002854645200295 s 1.83
const_scatter / DefOpt / cpu / PreRev 0.000517227 s 0.0002907582999796 s 1.78
const_scatter / DefOpt / cpu / PostRev 0.00052065 s 0.0002929033599957 s 1.78
const_scatter / DefOpt / cpu / BothRev 0.000487182 s 0.0002860883799894 s 1.70
const_scatter / IDefOpt / cpu / PreRev 0.000511615 s 0.0002909318000274 s 1.76
const_scatter / IDefOpt / cpu / PostRev 0.0005126929999999 s 0.0002944353399379 s 1.74
const_scatter / IDefOpt / cpu / BothRev 0.000523771 s 0.0002834542200616 s 1.85
const_scatter / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000007382560042969999 s 1.22
const_scatter / Jax / cpu / Primal 0.000008 s 0.000007396020000669523 s 1.08
const_scatter / HLOOpt / cpu / Primal 0.000008 s 0.000008004739993339171 s 1.00
const_scatter / PartOpt / cpu / Primal 0.000008 s 0.000006350180010485928 s 1.26
const_scatter / IPartOpt / cpu / Primal 0.000008 s 0.000006601120003324468 s 1.21
const_scatter / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000011223479978070827 s 0.80
const_scatter / IDefOpt / cpu / Primal 0.000008 s 0.000007006859996181447 s 1.14
const_scatter / JaXPipe / cpu / Forward 0.000011 s 0.000009879959980025888 s 1.11
const_scatter / Jax / cpu / Forward 0.000011 s 0.000010793700021167753 s 1.02
const_scatter / HLOOpt / cpu / Forward 0.000012 s 0.000009937820004779497 s 1.21
const_scatter / PartOpt / cpu / Forward 0.000012 s 0.000014244259991755826 s 0.84
const_scatter / IPartOpt / cpu / Forward 0.000011 s 0.000010233359962512622 s 1.07
const_scatter / DefOpt / cpu / Forward 0.000011 s 0.000014041699960216648 s 0.78
const_scatter / IDefOpt / cpu / Forward 0.000011 s 0.00000954062000346312 s 1.15
const_scatter / JaXPipe / cpu / PreRev 0.000331 s 0.0003037084600055 s 1.09
const_scatter / JaXPipe / cpu / PostRev 0.000354 s 0.0002919454400034 s 1.21
const_scatter / JaXPipe / cpu / BothRev 0.000331 s 0.0002827461599645 s 1.17
const_scatter / Jax / cpu / BothRev 0.000356 s 0.0002841475600234 s 1.25
const_scatter / HLOOpt / cpu / PreRev 0.000333 s 0.0002834839600018 s 1.17
const_scatter / HLOOpt / cpu / PostRev 0.000347 s 0.0002885376600443 s 1.20
const_scatter / HLOOpt / cpu / BothRev 0.000359 s 0.0002866564400028 s 1.25
const_scatter / PartOpt / cpu / PreRev 0.000358 s 0.0002885445599804 s 1.24
const_scatter / PartOpt / cpu / PostRev 0.000317 s 0.0002914730799966 s 1.09
const_scatter / PartOpt / cpu / BothRev 0.000334 s 0.000286857999954 s 1.16
const_scatter / IPartOpt / cpu / PreRev 0.000322 s 0.0002903088000402 s 1.11
const_scatter / IPartOpt / cpu / PostRev 0.000333 s 0.0002913657600038 s 1.14
const_scatter / IPartOpt / cpu / BothRev 0.000396 s 0.0002854645200295 s 1.39
const_scatter / DefOpt / cpu / PreRev 0.000335 s 0.0002907582999796 s 1.15
const_scatter / DefOpt / cpu / PostRev 0.000337 s 0.0002929033599957 s 1.15
const_scatter / DefOpt / cpu / BothRev 0.000333 s 0.0002860883799894 s 1.16
const_scatter / IDefOpt / cpu / PreRev 0.000354 s 0.0002909318000274 s 1.22
const_scatter / IDefOpt / cpu / PostRev 0.000365 s 0.0002944353399379 s 1.24
const_scatter / IDefOpt / cpu / BothRev 0.00035 s 0.0002834542200616 s 1.23
GenDot / JaXPipe / cpu / Primal 0.000008412419965679874 s 0.000008016640040295897 s 1.05
GenDot / Jax / cpu / Primal 0.000006871620016681846 s 0.000006963079977140297 s 0.99
GenDot / HLOOpt / cpu / Primal 0.000011450800020611495 s 0.000012155039985373151 s 0.94
GenDot / PartOpt / cpu / Primal 0.000008901319997676182 s 0.000007092280011420371 s 1.26
GenDot / IPartOpt / cpu / Primal 0.000008378759976039873 s 0.000006826300041211652 s 1.23
GenDot / DefOpt / cpu / Primal 0.000012072980034645296 s 0.00001228300000548188 s 0.98
GenDot / IDefOpt / cpu / Primal 0.000007816060015102266 s 0.000007347699993260903 s 1.06
GenDot / JaXPipe / cpu / Forward 0.000011608380018515164 s 0.000011124159982500714 s 1.04
GenDot / Jax / cpu / Forward 0.00001120416002777347 s 0.000010309219987902908 s 1.09
GenDot / HLOOpt / cpu / Forward 0.000016821640037960606 s 0.000016959299964582898 s 0.99
GenDot / PartOpt / cpu / Forward 0.000017084340015571797 s 0.000017031440011123776 s 1.00
GenDot / IPartOpt / cpu / Forward 0.000011800619968198588 s 0.000010438639992571551 s 1.13
GenDot / DefOpt / cpu / Forward 0.000016512340016561212 s 0.000016368179976780085 s 1.01
GenDot / IDefOpt / cpu / Forward 0.000010930039952654624 s 0.000011066980005125514 s 0.99
GenDot / JaXPipe / cpu / PreRev 0.000011901439957000547 s 0.000011087060047429986 s 1.07
GenDot / JaXPipe / cpu / PostRev 0.00001043515999299416 s 0.000015391520009870873 s 0.68
GenDot / JaXPipe / cpu / BothRev 0.000015487840018977296 s 0.000010789599973577425 s 1.44
GenDot / Jax / cpu / BothRev 0.000011384119979993556 s 0.000010168200005864492 s 1.12
GenDot / HLOOpt / cpu / PreRev 0.000011576940023587667 s 0.00001046222003424191 s 1.11
GenDot / HLOOpt / cpu / PostRev 0.000012132000001656708 s 0.000015490820014747442 s 0.78
GenDot / HLOOpt / cpu / BothRev 0.000013101120039209493 s 0.000012611860020115272 s 1.04
GenDot / PartOpt / cpu / PreRev 0.000011155239981235354 s 0.00001058218000252964 s 1.05
GenDot / PartOpt / cpu / PostRev 0.000010662399963621283 s 0.000009967080022761366 s 1.07
GenDot / PartOpt / cpu / BothRev 0.000011535480016391375 s 0.00001042555998537864 s 1.11
GenDot / IPartOpt / cpu / PreRev 0.000013714479991904228 s 0.000015295540033548606 s 0.90
GenDot / IPartOpt / cpu / PostRev 0.000010782680046759196 s 0.000010318080030629064 s 1.05
GenDot / IPartOpt / cpu / BothRev 0.000011581080025280244 s 0.000010373020040788106 s 1.12
GenDot / DefOpt / cpu / PreRev 0.000011405760005800404 s 0.00001111214001866756 s 1.03
GenDot / DefOpt / cpu / PostRev 0.00001164126003459387 s 0.00001102464002542547 s 1.06
GenDot / DefOpt / cpu / BothRev 0.000011582439992707804 s 0.00001104814000427723 s 1.05
GenDot / IDefOpt / cpu / PreRev 0.0000113537399283814 s 0.000010481300005267258 s 1.08
GenDot / IDefOpt / cpu / PostRev 0.000011538840080902446 s 0.000011272080037088016 s 1.02
GenDot / IDefOpt / cpu / BothRev 0.000011781480025092603 s 0.000010840579998330211 s 1.09
GenDot / JaXPipe / tpu / Primal 9.29125e-7 s 9.29875e-7 s 1.00
GenDot / Jax / tpu / Primal 9.352e-7 s 9.359e-7 s 1.00
GenDot / HLOOpt / tpu / Primal 0.00000156705 s 0.00000157835 s 0.99
GenDot / PartOpt / tpu / Primal 9.35525e-7 s 9.35925e-7 s 1.00
GenDot / IPartOpt / tpu / Primal 9.3935e-7 s 9.4065e-7 s 1.00
GenDot / DefOpt / tpu / Primal 0.0000014919 s 0.0000014947 s 1.00
GenDot / IDefOpt / tpu / Primal 0.000001566425 s 0.00000158215 s 0.99
GenDot / JaXPipe / tpu / Forward 0.00000316 s 0.00000316175 s 1.00
GenDot / Jax / tpu / Forward 0.0000023280000000000004 s 0.000002329725 s 1.00
GenDot / HLOOpt / tpu / Forward 0.000003112275 s 0.00000311885 s 1.00
GenDot / PartOpt / tpu / Forward 0.000003217325 s 0.000003216375 s 1.00
GenDot / IPartOpt / tpu / Forward 0.0000031131500000000004 s 0.000003126925 s 1.00
GenDot / DefOpt / tpu / Forward 0.0000032092 s 0.00000322485 s 1.00
GenDot / IDefOpt / tpu / Forward 0.00000310815 s 0.00000312535 s 0.99
GenDot / JaXPipe / tpu / PreRev 0.000002955975 s 0.000002971825 s 0.99
GenDot / JaXPipe / tpu / PostRev 0.000002408225 s 0.00000241625 s 1.00
GenDot / JaXPipe / tpu / BothRev 0.0000029515 s 0.000002956175 s 1.00
GenDot / Jax / tpu / BothRev 0.00000239955 s 0.0000024034500000000003 s 1.00
GenDot / HLOOpt / tpu / PreRev 0.0000029503000000000004 s 0.000002969 s 0.99
GenDot / HLOOpt / tpu / PostRev 0.000002935875 s 0.000002934125 s 1.00
GenDot / HLOOpt / tpu / BothRev 0.0000029545500000000004 s 0.0000029563000000000004 s 1.00
GenDot / PartOpt / tpu / PreRev 0.0000029265750000000004 s 0.00000293125 s 1.00
GenDot / PartOpt / tpu / PostRev 0.0000023921 s 0.0000024151 s 0.99
GenDot / PartOpt / tpu / BothRev 0.0000029324 s 0.000002932425 s 1.00
GenDot / IPartOpt / tpu / PreRev 0.000002956875 s 0.000002953425 s 1.00
GenDot / IPartOpt / tpu / PostRev 0.0000024025000000000003 s 0.000002413975 s 1.00
GenDot / IPartOpt / tpu / BothRev 0.00000296675 s 0.0000029558 s 1.00
GenDot / DefOpt / tpu / PreRev 0.0000029367500000000003 s 0.00000294615 s 1.00
GenDot / DefOpt / tpu / PostRev 0.000002949375 s 0.00000296425 s 0.99
GenDot / DefOpt / tpu / BothRev 0.000002929125 s 0.00000293525 s 1.00
GenDot / IDefOpt / tpu / PreRev 0.000002955925 s 0.000002961 s 1.00
GenDot / IDefOpt / tpu / PostRev 0.000002925625 s 0.0000029478 s 0.99
GenDot / IDefOpt / tpu / BothRev 0.000002949825 s 0.0000029599 s 1.00
GenDot / JaXPipe / cpu / Primal 0.000014613 s 0.000008016640040295897 s 1.82
GenDot / Jax / cpu / Primal 0.000014993 s 0.000006963079977140297 s 2.15
GenDot / HLOOpt / cpu / Primal 0.00001359 s 0.000012155039985373151 s 1.12
GenDot / PartOpt / cpu / Primal 0.000014696 s 0.000007092280011420371 s 2.07
GenDot / IPartOpt / cpu / Primal 0.000014075 s 0.000006826300041211652 s 2.06
GenDot / DefOpt / cpu / Primal 0.000013839 s 0.00001228300000548188 s 1.13
GenDot / IDefOpt / cpu / Primal 0.000013732 s 0.000007347699993260903 s 1.87
GenDot / JaXPipe / cpu / Forward 0.000019341 s 0.000011124159982500714 s 1.74
GenDot / Jax / cpu / Forward 0.000020997 s 0.000010309219987902908 s 2.04
GenDot / HLOOpt / cpu / Forward 0.000019098 s 0.000016959299964582898 s 1.13
GenDot / PartOpt / cpu / Forward 0.000019282 s 0.000017031440011123776 s 1.13
GenDot / IPartOpt / cpu / Forward 0.000018793 s 0.000010438639992571551 s 1.80
GenDot / DefOpt / cpu / Forward 0.000018948 s 0.000016368179976780085 s 1.16
GenDot / IDefOpt / cpu / Forward 0.000019486 s 0.000011066980005125514 s 1.76
GenDot / JaXPipe / cpu / PreRev 0.000019675 s 0.000011087060047429986 s 1.77
GenDot / JaXPipe / cpu / PostRev 0.000021699 s 0.000015391520009870873 s 1.41
GenDot / JaXPipe / cpu / BothRev 0.000019924 s 0.000010789599973577425 s 1.85
GenDot / Jax / cpu / BothRev 0.000020248 s 0.000010168200005864492 s 1.99
GenDot / HLOOpt / cpu / PreRev 0.000019285 s 0.00001046222003424191 s 1.84
GenDot / HLOOpt / cpu / PostRev 0.000019511 s 0.000015490820014747442 s 1.26
GenDot / HLOOpt / cpu / BothRev 0.000019862 s 0.000012611860020115272 s 1.57
GenDot / PartOpt / cpu / PreRev 0.000018717 s 0.00001058218000252964 s 1.77
GenDot / PartOpt / cpu / PostRev 0.000020983 s 0.000009967080022761366 s 2.11
GenDot / PartOpt / cpu / BothRev 0.000020839 s 0.00001042555998537864 s 2.00
GenDot / IPartOpt / cpu / PreRev 0.00001941 s 0.000015295540033548606 s 1.27
GenDot / IPartOpt / cpu / PostRev 0.000021212000000000003 s 0.000010318080030629064 s 2.06
GenDot / IPartOpt / cpu / BothRev 0.00001947 s 0.000010373020040788106 s 1.88
GenDot / DefOpt / cpu / PreRev 0.000019888 s 0.00001111214001866756 s 1.79
GenDot / DefOpt / cpu / PostRev 0.000019151 s 0.00001102464002542547 s 1.74
GenDot / DefOpt / cpu / BothRev 0.000019083 s 0.00001104814000427723 s 1.73
GenDot / IDefOpt / cpu / PreRev 0.000019288 s 0.000010481300005267258 s 1.84
GenDot / IDefOpt / cpu / PostRev 0.000018659 s 0.000011272080037088016 s 1.66
GenDot / IDefOpt / cpu / BothRev 0.000018226 s 0.000010840579998330211 s 1.68
GenDot / JaXPipe / cpu / Primal 0.00001 s 0.000008016640040295897 s 1.25
GenDot / Jax / cpu / Primal 0.00001 s 0.000006963079977140297 s 1.44
GenDot / HLOOpt / cpu / Primal 0.00001 s 0.000012155039985373151 s 0.82
GenDot / PartOpt / cpu / Primal 0.00001 s 0.000007092280011420371 s 1.41
GenDot / IPartOpt / cpu / Primal 0.00001 s 0.000006826300041211652 s 1.46
GenDot / DefOpt / cpu / Primal 0.000008999999999999999 s 0.00001228300000548188 s 0.73
GenDot / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000007347699993260903 s 1.22
GenDot / JaXPipe / cpu / Forward 0.000012 s 0.000011124159982500714 s 1.08
GenDot / Jax / cpu / Forward 0.000013 s 0.000010309219987902908 s 1.26
GenDot / HLOOpt / cpu / Forward 0.000012 s 0.000016959299964582898 s 0.71
GenDot / PartOpt / cpu / Forward 0.000012 s 0.000017031440011123776 s 0.70
GenDot / IPartOpt / cpu / Forward 0.000013 s 0.000010438639992571551 s 1.25
GenDot / DefOpt / cpu / Forward 0.000012 s 0.000016368179976780085 s 0.73
GenDot / IDefOpt / cpu / Forward 0.000013 s 0.000011066980005125514 s 1.17
GenDot / JaXPipe / cpu / PreRev 0.000013 s 0.000011087060047429986 s 1.17
GenDot / JaXPipe / cpu / PostRev 0.000013 s 0.000015391520009870873 s 0.84
GenDot / JaXPipe / cpu / BothRev 0.000013 s 0.000010789599973577425 s 1.20
GenDot / Jax / cpu / BothRev 0.000013 s 0.000010168200005864492 s 1.28
GenDot / HLOOpt / cpu / PreRev 0.000012 s 0.00001046222003424191 s 1.15
GenDot / HLOOpt / cpu / PostRev 0.000013 s 0.000015490820014747442 s 0.84
GenDot / HLOOpt / cpu / BothRev 0.000014 s 0.000012611860020115272 s 1.11
GenDot / PartOpt / cpu / PreRev 0.000013 s 0.00001058218000252964 s 1.23
GenDot / PartOpt / cpu / PostRev 0.000014 s 0.000009967080022761366 s 1.40
GenDot / PartOpt / cpu / BothRev 0.000013 s 0.00001042555998537864 s 1.25
GenDot / IPartOpt / cpu / PreRev 0.000013 s 0.000015295540033548606 s 0.85
GenDot / IPartOpt / cpu / PostRev 0.000013 s 0.000010318080030629064 s 1.26
GenDot / IPartOpt / cpu / BothRev 0.000013 s 0.000010373020040788106 s 1.25
GenDot / DefOpt / cpu / PreRev 0.000013 s 0.00001111214001866756 s 1.17
GenDot / DefOpt / cpu / PostRev 0.000013 s 0.00001102464002542547 s 1.18
GenDot / DefOpt / cpu / BothRev 0.000013 s 0.00001104814000427723 s 1.18
GenDot / IDefOpt / cpu / PreRev 0.000013 s 0.000010481300005267258 s 1.24
GenDot / IDefOpt / cpu / PostRev 0.000013 s 0.000011272080037088016 s 1.15
GenDot / IDefOpt / cpu / BothRev 0.000013 s 0.000010840579998330211 s 1.20
hlo_ffi / JaXPipe / cpu / Primal 0.000010921260018221802 s 0.000011137280016555452 s 0.98
hlo_ffi / Jax / cpu / Primal 0.000010030259982158896 s 0.000011213040006623486 s 0.89
hlo_ffi / HLOOpt / cpu / Primal 0.000013814680023642724 s 0.00001432780003597145 s 0.96
hlo_ffi / PartOpt / cpu / Primal 0.000010144640018552307 s 0.00001080817998627026 s 0.94
hlo_ffi / IPartOpt / cpu / Primal 0.0000099412600229698 s 0.000010937800016108668 s 0.91
hlo_ffi / DefOpt / cpu / Primal 0.000014687580060126492 s 0.000013061580020803376 s 1.12
hlo_ffi / IDefOpt / cpu / Primal 0.00001007190003292635 s 0.000010378080060036154 s 0.97
hlo_ffi / JaXPipe / cpu / Forward 0.000015209000011964237 s 0.00001658268000028329 s 0.92
hlo_ffi / Jax / cpu / Forward 0.000015098399999260436 s 0.00001630907997423492 s 0.93
hlo_ffi / HLOOpt / cpu / Forward 0.000014817760011283098 s 0.000016320479971909663 s 0.91
hlo_ffi / PartOpt / cpu / Forward 0.000015382379961010883 s 0.00001644468002268695 s 0.94
hlo_ffi / IPartOpt / cpu / Forward 0.000014994999974078382 s 0.0000165567999920313 s 0.91
hlo_ffi / DefOpt / cpu / Forward 0.00001514856001449516 s 0.000016407100010837892 s 0.92
hlo_ffi / IDefOpt / cpu / Forward 0.00001503394005339942 s 0.00001647552001486474 s 0.91
hlo_ffi / JaXPipe / cpu / PreRev 0.000014997840025898768 s 0.000016623439951217733 s 0.90
hlo_ffi / JaXPipe / cpu / PostRev 0.000015381139992314274 s 0.000016336880007656874 s 0.94
hlo_ffi / JaXPipe / cpu / BothRev 0.000014938020012777997 s 0.000016133080034705927 s 0.93
hlo_ffi / Jax / cpu / BothRev 0.000014934640003048116 s 0.00001577893999638036 s 0.95
hlo_ffi / HLOOpt / cpu / PreRev 0.000014579299977413027 s 0.000016193839965126244 s 0.90
hlo_ffi / HLOOpt / cpu / PostRev 0.000015399359999719308 s 0.000015816299992366112 s 0.97
hlo_ffi / HLOOpt / cpu / BothRev 0.000016786760006652912 s 0.000016931239943005495 s 0.99
hlo_ffi / PartOpt / cpu / PreRev 0.00001494302003266057 s 0.00001552104004076682 s 0.96
hlo_ffi / PartOpt / cpu / PostRev 0.000014932199974282412 s 0.000016032879957492697 s 0.93
hlo_ffi / PartOpt / cpu / BothRev 0.000014829159972578056 s 0.000015746119988762075 s 0.94
hlo_ffi / IPartOpt / cpu / PreRev 0.000014609800036851083 s 0.0000162041800103907 s 0.90
hlo_ffi / IPartOpt / cpu / PostRev 0.000014884000020174426 s 0.000016090159997474986 s 0.93
hlo_ffi / IPartOpt / cpu / BothRev 0.0000150412000130018 s 0.00001623691998247523 s 0.93
hlo_ffi / DefOpt / cpu / PreRev 0.000015122399963729547 s 0.000016106260018204923 s 0.94
hlo_ffi / DefOpt / cpu / PostRev 0.000015530079990639934 s 0.000015478560026167543 s 1.00
hlo_ffi / DefOpt / cpu / BothRev 0.00001526793999801157 s 0.000016545159987799706 s 0.92
hlo_ffi / IDefOpt / cpu / PreRev 0.000014902159973644302 s 0.000015835739959584315 s 0.94
hlo_ffi / IDefOpt / cpu / PostRev 0.000015500899999096872 s 0.00001573485997141688 s 0.99
hlo_ffi / IDefOpt / cpu / BothRev 0.000015035419983178144 s 0.0000162436400205479 s 0.93
hlo_ffi / JaXPipe / tpu / Primal 9.00025e-7 s 9.27675e-7 s 0.97
hlo_ffi / Jax / tpu / Primal 9.70825e-7 s 9.5865e-7 s 1.01
hlo_ffi / HLOOpt / tpu / Primal 9.36125e-7 s 9.041e-7 s 1.04
hlo_ffi / PartOpt / tpu / Primal 9.73125e-7 s 9.50125e-7 s 1.02
hlo_ffi / IPartOpt / tpu / Primal 9.3335e-7 s 9.07575e-7 s 1.03
hlo_ffi / DefOpt / tpu / Primal 9.733e-7 s 9.51775e-7 s 1.02
hlo_ffi / IDefOpt / tpu / Primal 9.3115e-7 s 9.1045e-7 s 1.02
hlo_ffi / JaXPipe / tpu / Forward 9.49425e-7 s 9.4885e-7 s 1.00
hlo_ffi / Jax / tpu / Forward 9.81775e-7 s 9.816e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Forward 9.73525e-7 s 9.73625e-7 s 1.00
hlo_ffi / PartOpt / tpu / Forward 9.58825e-7 s 9.339e-7 s 1.03
hlo_ffi / IPartOpt / tpu / Forward 9.7355e-7 s 9.73625e-7 s 1.00
hlo_ffi / DefOpt / tpu / Forward 9.58775e-7 s 9.3375e-7 s 1.03
hlo_ffi / IDefOpt / tpu / Forward 9.7425e-7 s 9.736749999999998e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PreRev 9.476e-7 s 9.38325e-7 s 1.01
hlo_ffi / JaXPipe / tpu / PostRev 9.64725e-7 s 9.6435e-7 s 1.00
hlo_ffi / JaXPipe / tpu / BothRev 9.94925e-7 s 9.6035e-7 s 1.04
hlo_ffi / Jax / tpu / BothRev 9.6435e-7 s 9.64725e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PreRev 9.9485e-7 s 9.597e-7 s 1.04
hlo_ffi / HLOOpt / tpu / PostRev 9.651e-7 s 9.64325e-7 s 1.00
hlo_ffi / HLOOpt / tpu / BothRev 9.9505e-7 s 9.60075e-7 s 1.04
hlo_ffi / PartOpt / tpu / PreRev 9.647e-7 s 9.6475e-7 s 1.00
hlo_ffi / PartOpt / tpu / PostRev 9.949e-7 s 9.603e-7 s 1.04
hlo_ffi / PartOpt / tpu / BothRev 9.65225e-7 s 9.64525e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PreRev 9.9445e-7 s 9.5985e-7 s 1.04
hlo_ffi / IPartOpt / tpu / PostRev 9.649e-7 s 9.6465e-7 s 1.00
hlo_ffi / IPartOpt / tpu / BothRev 9.9455e-7 s 9.59875e-7 s 1.04
hlo_ffi / DefOpt / tpu / PreRev 9.64625e-7 s 9.6455e-7 s 1.00
hlo_ffi / DefOpt / tpu / PostRev 9.9495e-7 s 9.60125e-7 s 1.04
hlo_ffi / DefOpt / tpu / BothRev 9.65225e-7 s 9.64375e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PreRev 9.94575e-7 s 9.602e-7 s 1.04
hlo_ffi / IDefOpt / tpu / PostRev 9.650749999999998e-7 s 9.643e-7 s 1.00
hlo_ffi / IDefOpt / tpu / BothRev 9.95075e-7 s 9.59975e-7 s 1.04
hlo_ffi / JaXPipe / cpu / Primal 0.000017629 s 0.000011137280016555452 s 1.58
hlo_ffi / Jax / cpu / Primal 0.000017749000000000002 s 0.000011213040006623486 s 1.58
hlo_ffi / HLOOpt / cpu / Primal 0.000017617 s 0.00001432780003597145 s 1.23
hlo_ffi / PartOpt / cpu / Primal 0.000017471 s 0.00001080817998627026 s 1.62
hlo_ffi / IPartOpt / cpu / Primal 0.000017442 s 0.000010937800016108668 s 1.59
hlo_ffi / DefOpt / cpu / Primal 0.000017485 s 0.000013061580020803376 s 1.34
hlo_ffi / IDefOpt / cpu / Primal 0.000017169 s 0.000010378080060036154 s 1.65
hlo_ffi / JaXPipe / cpu / Forward 0.000024982 s 0.00001658268000028329 s 1.51
hlo_ffi / Jax / cpu / Forward 0.000025126 s 0.00001630907997423492 s 1.54
hlo_ffi / HLOOpt / cpu / Forward 0.000024656 s 0.000016320479971909663 s 1.51
hlo_ffi / PartOpt / cpu / Forward 0.000025159 s 0.00001644468002268695 s 1.53
hlo_ffi / IPartOpt / cpu / Forward 0.000024481 s 0.0000165567999920313 s 1.48
hlo_ffi / DefOpt / cpu / Forward 0.000024868 s 0.000016407100010837892 s 1.52
hlo_ffi / IDefOpt / cpu / Forward 0.000024392 s 0.00001647552001486474 s 1.48
hlo_ffi / JaXPipe / cpu / PreRev 0.000025144 s 0.000016623439951217733 s 1.51
hlo_ffi / JaXPipe / cpu / PostRev 0.000024774 s 0.000016336880007656874 s 1.52
hlo_ffi / JaXPipe / cpu / BothRev 0.000024518 s 0.000016133080034705927 s 1.52
hlo_ffi / Jax / cpu / BothRev 0.000024158 s 0.00001577893999638036 s 1.53
hlo_ffi / HLOOpt / cpu / PreRev 0.000024013 s 0.000016193839965126244 s 1.48
hlo_ffi / HLOOpt / cpu / PostRev 0.00002457 s 0.000015816299992366112 s 1.55
hlo_ffi / HLOOpt / cpu / BothRev 0.000025631 s 0.000016931239943005495 s 1.51
hlo_ffi / PartOpt / cpu / PreRev 0.000025625 s 0.00001552104004076682 s 1.65
hlo_ffi / PartOpt / cpu / PostRev 0.000025415 s 0.000016032879957492697 s 1.59
hlo_ffi / PartOpt / cpu / BothRev 0.000026616 s 0.000015746119988762075 s 1.69
hlo_ffi / IPartOpt / cpu / PreRev 0.000024315 s 0.0000162041800103907 s 1.50
hlo_ffi / IPartOpt / cpu / PostRev 0.000026943000000000003 s 0.000016090159997474986 s 1.67
hlo_ffi / IPartOpt / cpu / BothRev 0.000026261 s 0.00001623691998247523 s 1.62
hlo_ffi / DefOpt / cpu / PreRev 0.00002464 s 0.000016106260018204923 s 1.53
hlo_ffi / DefOpt / cpu / PostRev 0.000025089 s 0.000015478560026167543 s 1.62
hlo_ffi / DefOpt / cpu / BothRev 0.000025705 s 0.000016545159987799706 s 1.55
hlo_ffi / IDefOpt / cpu / PreRev 0.000025406 s 0.000015835739959584315 s 1.60
hlo_ffi / IDefOpt / cpu / PostRev 0.000026274 s 0.00001573485997141688 s 1.67
hlo_ffi / IDefOpt / cpu / BothRev 0.000026267 s 0.0000162436400205479 s 1.62
hlo_ffi / JaXPipe / cpu / Primal 0.000012 s 0.000011137280016555452 s 1.08
hlo_ffi / Jax / cpu / Primal 0.000012 s 0.000011213040006623486 s 1.07
hlo_ffi / HLOOpt / cpu / Primal 0.000011 s 0.00001432780003597145 s 0.77
hlo_ffi / PartOpt / cpu / Primal 0.000012 s 0.00001080817998627026 s 1.11
hlo_ffi / IPartOpt / cpu / Primal 0.000012 s 0.000010937800016108668 s 1.10
hlo_ffi / DefOpt / cpu / Primal 0.000011 s 0.000013061580020803376 s 0.84
hlo_ffi / IDefOpt / cpu / Primal 0.000012 s 0.000010378080060036154 s 1.16
hlo_ffi / JaXPipe / cpu / Forward 0.000016 s 0.00001658268000028329 s 0.96
hlo_ffi / Jax / cpu / Forward 0.000017999999999999997 s 0.00001630907997423492 s 1.10
hlo_ffi / HLOOpt / cpu / Forward 0.000017 s 0.000016320479971909663 s 1.04
hlo_ffi / PartOpt / cpu / Forward 0.000016 s 0.00001644468002268695 s 0.97
hlo_ffi / IPartOpt / cpu / Forward 0.000016 s 0.0000165567999920313 s 0.97
hlo_ffi / DefOpt / cpu / Forward 0.000017999999999999997 s 0.000016407100010837892 s 1.10
hlo_ffi / IDefOpt / cpu / Forward 0.000016 s 0.00001647552001486474 s 0.97
hlo_ffi / JaXPipe / cpu / PreRev 0.000017 s 0.000016623439951217733 s 1.02
hlo_ffi / JaXPipe / cpu / PostRev 0.000016 s 0.000016336880007656874 s 0.98
hlo_ffi / JaXPipe / cpu / BothRev 0.000017 s 0.000016133080034705927 s 1.05
hlo_ffi / Jax / cpu / BothRev 0.000016 s 0.00001577893999638036 s 1.01
hlo_ffi / HLOOpt / cpu / PreRev 0.000016 s 0.000016193839965126244 s 0.99
hlo_ffi / HLOOpt / cpu / PostRev 0.000016 s 0.000015816299992366112 s 1.01
hlo_ffi / HLOOpt / cpu / BothRev 0.000017999999999999997 s 0.000016931239943005495 s 1.06
hlo_ffi / PartOpt / cpu / PreRev 0.000017 s 0.00001552104004076682 s 1.10
hlo_ffi / PartOpt / cpu / PostRev 0.000017 s 0.000016032879957492697 s 1.06
hlo_ffi / PartOpt / cpu / BothRev 0.000017999999999999997 s 0.000015746119988762075 s 1.14
hlo_ffi / IPartOpt / cpu / PreRev 0.000017 s 0.0000162041800103907 s 1.05
hlo_ffi / IPartOpt / cpu / PostRev 0.000017 s 0.000016090159997474986 s 1.06
hlo_ffi / IPartOpt / cpu / BothRev 0.000016 s 0.00001623691998247523 s 0.99
hlo_ffi / DefOpt / cpu / PreRev 0.000017 s 0.000016106260018204923 s 1.06
hlo_ffi / DefOpt / cpu / PostRev 0.000016 s 0.000015478560026167543 s 1.03
hlo_ffi / DefOpt / cpu / BothRev 0.000017 s 0.000016545159987799706 s 1.03
hlo_ffi / IDefOpt / cpu / PreRev 0.000017 s 0.000015835739959584315 s 1.07
hlo_ffi / IDefOpt / cpu / PostRev 0.000016 s 0.00001573485997141688 s 1.02
hlo_ffi / IDefOpt / cpu / BothRev 0.000016 s 0.0000162436400205479 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.0010963728000206 s 0.0011988790001851 s 0.91
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 0.0009653824001361 s 0.0010886000001846 s 0.89
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 0.0009966474000066 s 0.00101588360003 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 / cpu / Primal 0.0009421720000318 s 0.0009838469998612 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 / cpu / Primal 0.00093815980008 s 0.0009344521998173 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 0.0010076167999613 s 0.0010030405999714 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 / Primal 0.0009922549999828 s 0.0010368926001319 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 / Forward 0.0027936741998928 s 0.0028974420001759 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 / cpu / Forward 0.0024394838001171 s 0.0024516601998584 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 / cpu / Forward 0.0023494353999922 s 0.0023232616001223 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 / cpu / Forward 0.002545311599988 s 0.0021805914000651 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 / IPartOpt / cpu / Forward 0.0025244205999115 s 0.0023878779999904 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.0023531380001259 s 0.0024981592000585 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 / Forward 0.0024414490000708 s 0.0022006737999618 s 1.11
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PreRev 0.0070384843999818 s 0.0060858452000502 s 1.16
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 0.0063119567999819 s 0.0055182287998832 s 1.14
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 0.0061667392000344 s 0.0063419092000913 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 / cpu / BothRev 0.0061082141999577 s 0.0062407923999671 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 / cpu / PreRev 0.0057602647999374 s 0.0066440607999538 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 / HLOOpt / cpu / PostRev 0.0062448281998513 s 0.0058188556001368 s 1.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.0067904720001024 s 0.0066429661999791 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 / PartOpt / cpu / PreRev 0.0062055780000264 s 0.0064629129999957 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 / cpu / PostRev 0.0065240349999839 s 0.0054757243999119 s 1.19
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.0059126411999386 s 0.0051480127999639 s 1.15
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.0058074470000974 s 0.0061625360001016 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 / cpu / PostRev 0.006161704400165 s 0.0062819011999636 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 / cpu / BothRev 0.0060245829999075 s 0.0059942748001049 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 / DefOpt / cpu / PreRev 0.0061345305998656 s 0.0052506886001538 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 / DefOpt / cpu / PostRev 0.0057852060000186 s 0.0049551932000213 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 / DefOpt / cpu / BothRev 0.0058817707999878 s 0.004979441999967 s 1.18
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 0.0060444032000305 s 0.0051975696000226 s 1.16
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 0.0061373841999738 s 0.0051738789999944 s 1.19
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.0063915990001078 s 0.007013339399964 s 0.91
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 0.000123864 s 0.00012241775 s 1.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / Primal 0.00012681925 s 0.0001240335 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / Primal 0.0001528655 s 0.000150573 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 / PartOpt / tpu / Primal 0.0001343535 s 0.0001309444999999 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 / tpu / Primal 0.00013114025 s 0.000129236 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 / DefOpt / tpu / Primal 0.0001482035 s 0.00014522725 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / Primal 0.00015073375 s 0.00014919175 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 / Forward 0.0002116855 s 0.000213208 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 / Jax / tpu / Forward 0.000261081 s 0.000260157 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 0.0002125145 s 0.0002128869999999 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 / Forward 0.000218851 s 0.0002109655 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 / Forward 0.000211959 s 0.000213325 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 / DefOpt / tpu / Forward 0.0002182144999999 s 0.0002109575 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 / Forward 0.00021203525 s 0.0002131702499999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PreRev 0.0003541364999999 s 0.0003561732499999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PostRev 0.00025614325 s 0.0002559779999999 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 0.0003550355 s 0.00035705975 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 / Jax / tpu / BothRev 0.000257321 s 0.0002571279999999 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 0.0003546515 s 0.000356954 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / PostRev 0.000291314 s 0.0002910835 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 0.00035490725 s 0.0003566545 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 0.0003555835 s 0.000355736 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 0.00027120725 s 0.00027441325 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 / BothRev 0.000356129 s 0.000355364 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 0.0003550325 s 0.00035684425 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / PostRev 0.0002722179999999 s 0.00027258625 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 0.000354885 s 0.00035727025 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 / DefOpt / tpu / PreRev 0.0003581045 s 0.0003580505 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 0.00028260425 s 0.00028488025 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 / DefOpt / tpu / BothRev 0.00035810725 s 0.0003578225 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 0.00035796725 s 0.0003591957499999 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 0.0002983919999999 s 0.00030137075 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 / BothRev 0.00035694075 s 0.0003590244999999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.002067336 s 0.0011988790001851 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 / Jax / cpu / Primal 0.002509577 s 0.0010886000001846 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 / HLOOpt / cpu / Primal 0.002482158 s 0.00101588360003 s 2.44
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 0.00230179 s 0.0009838469998612 s 2.34
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 0.002484511 s 0.0009344521998173 s 2.66
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 0.002402482 s 0.0010030405999714 s 2.40
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 0.002210942 s 0.0010368926001319 s 2.13
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Forward 0.005818694 s 0.0028974420001759 s 2.01
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.00601174 s 0.0024516601998584 s 2.45
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 0.005919234 s 0.0023232616001223 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 / PartOpt / cpu / Forward 0.006284873 s 0.0021805914000651 s 2.88
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 0.005257612 s 0.0023878779999904 s 2.20
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.005869314 s 0.0024981592000585 s 2.35
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 0.005298674 s 0.0022006737999618 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 / JaXPipe / cpu / PreRev 0.011391945 s 0.0060858452000502 s 1.87
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.009221971 s 0.0055182287998832 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 / JaXPipe / cpu / BothRev 0.009001517 s 0.0063419092000913 s 1.42
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 0.0087771039999999 s 0.0062407923999671 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 / PreRev 0.00805 s 0.0066440607999538 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 / PostRev 0.008167827 s 0.0058188556001368 s 1.40
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.007786473 s 0.0066429661999791 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 / PartOpt / cpu / PreRev 0.008356608 s 0.0064629129999957 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 / PartOpt / cpu / PostRev 0.008819666 s 0.0054757243999119 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 / BothRev 0.008562794 s 0.0051480127999639 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 / IPartOpt / cpu / PreRev 0.0080420789999999 s 0.0061625360001016 s 1.30
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PostRev 0.009378725 s 0.0062819011999636 s 1.49
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 0.008834147 s 0.0059942748001049 s 1.47
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 0.009181879 s 0.0052506886001538 s 1.75
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 0.0082065549999999 s 0.0049551932000213 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 / DefOpt / cpu / BothRev 0.008333038 s 0.004979441999967 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 / IDefOpt / cpu / PreRev 0.008289276 s 0.0051975696000226 s 1.59
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PostRev 0.008605238 s 0.0051738789999944 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 / IDefOpt / cpu / BothRev 0.0083507099999999 s 0.007013339399964 s 1.19
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 0.001552 s 0.0011988790001851 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 / Jax / cpu / Primal 0.001714 s 0.0010886000001846 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 / HLOOpt / cpu / Primal 0.00153 s 0.00101588360003 s 1.51
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.001585 s 0.0009838469998612 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 / Primal 0.001501 s 0.0009344521998173 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 / DefOpt / cpu / Primal 0.001632 s 0.0010030405999714 s 1.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.001502 s 0.0010368926001319 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 / JaXPipe / cpu / Forward 0.004713 s 0.0028974420001759 s 1.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.004606 s 0.0024516601998584 s 1.88
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 0.004585 s 0.0023232616001223 s 1.97
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.003745 s 0.0021805914000651 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 / IPartOpt / cpu / Forward 0.004227 s 0.0023878779999904 s 1.77
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.004023 s 0.0024981592000585 s 1.61
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Forward 0.003826 s 0.0022006737999618 s 1.74
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PreRev 0.007222 s 0.0060858452000502 s 1.19
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 0.008964 s 0.0055182287998832 s 1.62
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / BothRev 0.007261 s 0.0063419092000913 s 1.14
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 0.009207 s 0.0062407923999671 s 1.48
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PreRev 0.007355 s 0.0066440607999538 s 1.11
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.006548 s 0.0058188556001368 s 1.13
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 0.007133 s 0.0066429661999791 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 / PreRev 0.006894 s 0.0064629129999957 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 / PostRev 0.0086679999999999 s 0.0054757243999119 s 1.58
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 0.007215 s 0.0051480127999639 s 1.40
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.0071189999999999 s 0.0061625360001016 s 1.16
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 0.008594 s 0.0062819011999636 s 1.37
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 0.009688 s 0.0059942748001049 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 0.008034 s 0.0052506886001538 s 1.53
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 0.006853 s 0.0049551932000213 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 / DefOpt / cpu / BothRev 0.010272 s 0.004979441999967 s 2.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PreRev 0.007375 s 0.0051975696000226 s 1.42
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 0.008218 s 0.0051738789999944 s 1.59
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.0071639999999999 s 0.007013339399964 s 1.02
scatter_sum / JaXPipe / cpu / Primal 0.00000853531998473045 s 0.000008040239963520435 s 1.06
scatter_sum / Jax / cpu / Primal 0.000008393480011363862 s 0.000007798040014677099 s 1.08
scatter_sum / HLOOpt / cpu / Primal 0.000011315860010654433 s 0.000012431279983502464 s 0.91
scatter_sum / PartOpt / cpu / Primal 0.000008829100015645964 s 0.000008115780010484741 s 1.09
scatter_sum / IPartOpt / cpu / Primal 0.000008460840017505689 s 0.000007551599965154309 s 1.12
scatter_sum / DefOpt / cpu / Primal 0.00000823136001599778 s 0.000007723220005573238 s 1.07
scatter_sum / IDefOpt / cpu / Primal 0.0000087177800196514 s 0.000007809219960108749 s 1.12
scatter_sum / JaXPipe / cpu / Forward 0.000012984540007892064 s 0.00001243014002284326 s 1.04
scatter_sum / Jax / cpu / Forward 0.000012284639960853383 s 0.00001213674003338383 s 1.01
scatter_sum / HLOOpt / cpu / Forward 0.000012204539998492692 s 0.000017723779992593337 s 0.69
scatter_sum / PartOpt / cpu / Forward 0.000018030820010608296 s 0.00001770212002156768 s 1.02
scatter_sum / IPartOpt / cpu / Forward 0.000012395959984132788 s 0.000012014159965474391 s 1.03
scatter_sum / DefOpt / cpu / Forward 0.00001832497999203042 s 0.000017420219955965875 s 1.05
scatter_sum / IDefOpt / cpu / Forward 0.000012411179995979185 s 0.000012483639984566251 s 0.99
scatter_sum / JaXPipe / cpu / PreRev 0.000013455019998218633 s 0.000012508079980761978 s 1.08
scatter_sum / JaXPipe / cpu / PostRev 0.00001361167996947188 s 0.000011829720024252313 s 1.15
scatter_sum / JaXPipe / cpu / BothRev 0.000017379840010107727 s 0.000011907399966730737 s 1.46
scatter_sum / Jax / cpu / BothRev 0.000013375840017033624 s 0.00001214067996443191 s 1.10
scatter_sum / HLOOpt / cpu / PreRev 0.000012885319965789676 s 0.0000119447199813294 s 1.08
scatter_sum / HLOOpt / cpu / PostRev 0.000013063200012766176 s 0.0000160856400270859 s 0.81
scatter_sum / HLOOpt / cpu / BothRev 0.000019102720007140305 s 0.000013727740024478409 s 1.39
scatter_sum / PartOpt / cpu / PreRev 0.000013509640011761804 s 0.000011590940002861317 s 1.17
scatter_sum / PartOpt / cpu / PostRev 0.000012862900011896271 s 0.000011606900006881916 s 1.11
scatter_sum / PartOpt / cpu / BothRev 0.000012754279960063284 s 0.000011817259983217807 s 1.08
scatter_sum / IPartOpt / cpu / PreRev 0.000013059600014457829 s 0.000017593460024727393 s 0.74
scatter_sum / IPartOpt / cpu / PostRev 0.00001364263996038062 s 0.00001198926001052314 s 1.14
scatter_sum / IPartOpt / cpu / BothRev 0.00001357508000182861 s 0.00001175103998320992 s 1.16
scatter_sum / DefOpt / cpu / PreRev 0.000012966519989277003 s 0.000011479659988253844 s 1.13
scatter_sum / DefOpt / cpu / PostRev 0.000013160700027583515 s 0.0000114498199945956 s 1.15
scatter_sum / DefOpt / cpu / BothRev 0.000013031039979978232 s 0.000011325099985697306 s 1.15
scatter_sum / IDefOpt / cpu / PreRev 0.000012806339991584535 s 0.000011430740005380358 s 1.12
scatter_sum / IDefOpt / cpu / PostRev 0.000012870479968114524 s 0.000011938400039070984 s 1.08
scatter_sum / IDefOpt / cpu / BothRev 0.000013368100017032704 s 0.00001204550001602911 s 1.11
scatter_sum / JaXPipe / tpu / Primal 0.0000013509499999999995 s 0.0000013429499999999998 s 1.01
scatter_sum / Jax / tpu / Primal 0.000001414375 s 0.0000014133 s 1.00
scatter_sum / HLOOpt / tpu / Primal 0.0000013607250000000002 s 0.0000013525000000000002 s 1.01
scatter_sum / PartOpt / tpu / Primal 0.00000141395 s 0.00000141395 s 1
scatter_sum / IPartOpt / tpu / Primal 0.0000013602249999999998 s 0.000001352175 s 1.01
scatter_sum / DefOpt / tpu / Primal 0.000001414 s 0.000001413675 s 1.00
scatter_sum / IDefOpt / tpu / Primal 0.0000013602 s 0.000001352625 s 1.01
scatter_sum / JaXPipe / tpu / Forward 0.000002718875 s 0.000002713625 s 1.00
scatter_sum / Jax / tpu / Forward 0.0000027389999999999995 s 0.00000272605 s 1.00
scatter_sum / HLOOpt / tpu / Forward 0.0000027103 s 0.000002716925 s 1.00
scatter_sum / PartOpt / tpu / Forward 0.000002705625 s 0.0000026988 s 1.00
scatter_sum / IPartOpt / tpu / Forward 0.0000027166 s 0.00000271445 s 1.00
scatter_sum / DefOpt / tpu / Forward 0.000002701125 s 0.00000269645 s 1.00
scatter_sum / IDefOpt / tpu / Forward 0.000002709275 s 0.000002718325 s 1.00
scatter_sum / JaXPipe / tpu / PreRev 0.000002707575 s 0.00000269345 s 1.01
scatter_sum / JaXPipe / tpu / PostRev 0.0000026996 s 0.000002698125 s 1.00
scatter_sum / JaXPipe / tpu / BothRev 0.0000027195 s 0.00000271185 s 1.00
scatter_sum / Jax / tpu / BothRev 0.00000275425 s 0.0000027554 s 1.00
scatter_sum / HLOOpt / tpu / PreRev 0.000002718125 s 0.0000027071250000000004 s 1.00
scatter_sum / HLOOpt / tpu / PostRev 0.0000027502 s 0.00000275135 s 1.00
scatter_sum / HLOOpt / tpu / BothRev 0.00000271575 s 0.000002706 s 1.00
scatter_sum / PartOpt / tpu / PreRev 0.000002754475 s 0.000002750575 s 1.00
scatter_sum / PartOpt / tpu / PostRev 0.0000027157 s 0.0000027086500000000003 s 1.00
scatter_sum / PartOpt / tpu / BothRev 0.000002749975 s 0.000002762175 s 1.00
scatter_sum / IPartOpt / tpu / PreRev 0.0000027118 s 0.0000027067249999999995 s 1.00
scatter_sum / IPartOpt / tpu / PostRev 0.0000027482 s 0.0000027543 s 1.00
scatter_sum / IPartOpt / tpu / BothRev 0.0000027123750000000004 s 0.0000027100000000000003 s 1.00
scatter_sum / DefOpt / tpu / PreRev 0.000002748925 s 0.000002751125 s 1.00
scatter_sum / DefOpt / tpu / PostRev 0.0000027128 s 0.0000027078750000000003 s 1.00
scatter_sum / DefOpt / tpu / BothRev 0.0000027535 s 0.000002752525 s 1.00
scatter_sum / IDefOpt / tpu / PreRev 0.0000027142250000000003 s 0.00000270615 s 1.00
scatter_sum / IDefOpt / tpu / PostRev 0.000002755325 s 0.000002755825 s 1.00
scatter_sum / IDefOpt / tpu / BothRev 0.0000027175 s 0.00000271035 s 1.00
scatter_sum / JaXPipe / cpu / Primal 0.000015914 s 0.000008040239963520435 s 1.98
scatter_sum / Jax / cpu / Primal 0.000015371 s 0.000007798040014677099 s 1.97
scatter_sum / HLOOpt / cpu / Primal 0.000016060000000000002 s 0.000012431279983502464 s 1.29
scatter_sum / PartOpt / cpu / Primal 0.000015665 s 0.000008115780010484741 s 1.93
scatter_sum / IPartOpt / cpu / Primal 0.000015917 s 0.000007551599965154309 s 2.11
scatter_sum / DefOpt / cpu / Primal 0.000015628 s 0.000007723220005573238 s 2.02
scatter_sum / IDefOpt / cpu / Primal 0.000016199000000000002 s 0.000007809219960108749 s 2.07
scatter_sum / JaXPipe / cpu / Forward 0.00002292 s 0.00001243014002284326 s 1.84
scatter_sum / Jax / cpu / Forward 0.000022551 s 0.00001213674003338383 s 1.86
scatter_sum / HLOOpt / cpu / Forward 0.000021172 s 0.000017723779992593337 s 1.19
scatter_sum / PartOpt / cpu / Forward 0.000022372 s 0.00001770212002156768 s 1.26
scatter_sum / IPartOpt / cpu / Forward 0.000022327 s 0.000012014159965474391 s 1.86
scatter_sum / DefOpt / cpu / Forward 0.000021873 s 0.000017420219955965875 s 1.26
scatter_sum / IDefOpt / cpu / Forward 0.00002249 s 0.000012483639984566251 s 1.80
scatter_sum / JaXPipe / cpu / PreRev 0.000021756 s 0.000012508079980761978 s 1.74
scatter_sum / JaXPipe / cpu / PostRev 0.00002248 s 0.000011829720024252313 s 1.90
scatter_sum / JaXPipe / cpu / BothRev 0.000024637 s 0.000011907399966730737 s 2.07
scatter_sum / Jax / cpu / BothRev 0.000022955 s 0.00001214067996443191 s 1.89
scatter_sum / HLOOpt / cpu / PreRev 0.000021628 s 0.0000119447199813294 s 1.81
scatter_sum / HLOOpt / cpu / PostRev 0.000023445 s 0.0000160856400270859 s 1.46
scatter_sum / HLOOpt / cpu / BothRev 0.000022493 s 0.000013727740024478409 s 1.64
scatter_sum / PartOpt / cpu / PreRev 0.000021804 s 0.000011590940002861317 s 1.88
scatter_sum / PartOpt / cpu / PostRev 0.000023491000000000003 s 0.000011606900006881916 s 2.02
scatter_sum / PartOpt / cpu / BothRev 0.000023712 s 0.000011817259983217807 s 2.01
scatter_sum / IPartOpt / cpu / PreRev 0.000022096 s 0.000017593460024727393 s 1.26
scatter_sum / IPartOpt / cpu / PostRev 0.000022359 s 0.00001198926001052314 s 1.86
scatter_sum / IPartOpt / cpu / BothRev 0.000022549 s 0.00001175103998320992 s 1.92
scatter_sum / DefOpt / cpu / PreRev 0.000021707 s 0.000011479659988253844 s 1.89
scatter_sum / DefOpt / cpu / PostRev 0.000021986 s 0.0000114498199945956 s 1.92
scatter_sum / DefOpt / cpu / BothRev 0.000022637 s 0.000011325099985697306 s 2.00
scatter_sum / IDefOpt / cpu / PreRev 0.000021313 s 0.000011430740005380358 s 1.86
scatter_sum / IDefOpt / cpu / PostRev 0.000022626 s 0.000011938400039070984 s 1.90
scatter_sum / IDefOpt / cpu / BothRev 0.000022736 s 0.00001204550001602911 s 1.89
scatter_sum / JaXPipe / cpu / Primal 0.00001 s 0.000008040239963520435 s 1.24
scatter_sum / Jax / cpu / Primal 0.00001 s 0.000007798040014677099 s 1.28
scatter_sum / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000012431279983502464 s 0.72
scatter_sum / PartOpt / cpu / Primal 0.00001 s 0.000008115780010484741 s 1.23
scatter_sum / IPartOpt / cpu / Primal 0.00001 s 0.000007551599965154309 s 1.32
scatter_sum / DefOpt / cpu / Primal 0.00001 s 0.000007723220005573238 s 1.29
scatter_sum / IDefOpt / cpu / Primal 0.00001 s 0.000007809219960108749 s 1.28
scatter_sum / JaXPipe / cpu / Forward 0.000015 s 0.00001243014002284326 s 1.21
scatter_sum / Jax / cpu / Forward 0.000015 s 0.00001213674003338383 s 1.24
scatter_sum / HLOOpt / cpu / Forward 0.000015 s 0.000017723779992593337 s 0.85
scatter_sum / PartOpt / cpu / Forward 0.000017 s 0.00001770212002156768 s 0.96
scatter_sum / IPartOpt / cpu / Forward 0.000015 s 0.000012014159965474391 s 1.25
scatter_sum / DefOpt / cpu / Forward 0.000015 s 0.000017420219955965875 s 0.86
scatter_sum / IDefOpt / cpu / Forward 0.000015 s 0.000012483639984566251 s 1.20
scatter_sum / JaXPipe / cpu / PreRev 0.000015 s 0.000012508079980761978 s 1.20
scatter_sum / JaXPipe / cpu / PostRev 0.000015 s 0.000011829720024252313 s 1.27
scatter_sum / JaXPipe / cpu / BothRev 0.000015 s 0.000011907399966730737 s 1.26
scatter_sum / Jax / cpu / BothRev 0.000016 s 0.00001214067996443191 s 1.32
scatter_sum / HLOOpt / cpu / PreRev 0.000015 s 0.0000119447199813294 s 1.26
scatter_sum / HLOOpt / cpu / PostRev 0.000015 s 0.0000160856400270859 s 0.93
scatter_sum / HLOOpt / cpu / BothRev 0.000016 s 0.000013727740024478409 s 1.17
scatter_sum / PartOpt / cpu / PreRev 0.000015 s 0.000011590940002861317 s 1.29
scatter_sum / PartOpt / cpu / PostRev 0.000016 s 0.000011606900006881916 s 1.38
scatter_sum / PartOpt / cpu / BothRev 0.000015 s 0.000011817259983217807 s 1.27
scatter_sum / IPartOpt / cpu / PreRev 0.000016 s 0.000017593460024727393 s 0.91
scatter_sum / IPartOpt / cpu / PostRev 0.000015 s 0.00001198926001052314 s 1.25
scatter_sum / IPartOpt / cpu / BothRev 0.000015 s 0.00001175103998320992 s 1.28
scatter_sum / DefOpt / cpu / PreRev 0.000015 s 0.000011479659988253844 s 1.31
scatter_sum / DefOpt / cpu / PostRev 0.000015 s 0.0000114498199945956 s 1.31
scatter_sum / DefOpt / cpu / BothRev 0.000015 s 0.000011325099985697306 s 1.32
scatter_sum / IDefOpt / cpu / PreRev 0.000016 s 0.000011430740005380358 s 1.40
scatter_sum / IDefOpt / cpu / PostRev 0.000015 s 0.000011938400039070984 s 1.26
scatter_sum / IDefOpt / cpu / BothRev 0.000015 s 0.00001204550001602911 s 1.25
slicing / JaXPipe / cpu / Primal 0.000007544839991169283 s 0.000007253260009747464 s 1.04
slicing / Jax / cpu / Primal 0.000006505359988295823 s 0.000006546520035044523 s 0.99
slicing / HLOOpt / cpu / Primal 0.00001085028004126798 s 0.000010940299989670166 s 0.99
slicing / PartOpt / cpu / Primal 0.000006591539986402495 s 0.000006298379948930233 s 1.05
slicing / IPartOpt / cpu / Primal 0.000006457120016420959 s 0.00000655244000881794 s 0.99
slicing / DefOpt / cpu / Primal 0.000011377099999663187 s 0.000010885439987760035 s 1.05
slicing / IDefOpt / cpu / Primal 0.000006635940026171738 s 0.00000650486000267847 s 1.02
slicing / JaXPipe / cpu / Forward 0.000010717320010371623 s 0.000009839760014074271 s 1.09
slicing / Jax / cpu / Forward 0.000010704340020311064 s 0.000010449979999975768 s 1.02
slicing / HLOOpt / cpu / Forward 0.000014441279981838306 s 0.000014228619993446046 s 1.01
slicing / PartOpt / cpu / Forward 0.00001450065998142236 s 0.00001480601998991915 s 0.98
slicing / IPartOpt / cpu / Forward 0.00001027055998747528 s 0.000009797279990380047 s 1.05
slicing / DefOpt / cpu / Forward 0.000014757559983991087 s 0.00001460727998164657 s 1.01
slicing / IDefOpt / cpu / Forward 0.00001005844000246725 s 0.000009516859954601386 s 1.06
slicing / JaXPipe / cpu / PreRev 0.000010852360010176198 s 0.000010368759976699948 s 1.05
slicing / JaXPipe / cpu / PostRev 0.000011674120032694191 s 0.00001086786002815643 s 1.07
slicing / JaXPipe / cpu / BothRev 0.000010937240022030892 s 0.000014082499974392704 s 0.78
slicing / Jax / cpu / BothRev 0.000011121180014015408 s 0.000010233800012429128 s 1.09
slicing / HLOOpt / cpu / PreRev 0.000010635219996402156 s 0.000010548579948590488 s 1.01
slicing / HLOOpt / cpu / PostRev 0.000011366980006641824 s 0.000010883419954552664 s 1.04
slicing / HLOOpt / cpu / BothRev 0.000012282539973966775 s 0.000011958500008404372 s 1.03
slicing / PartOpt / cpu / PreRev 0.000010513079996599117 s 0.000009984839980461402 s 1.05
slicing / PartOpt / cpu / PostRev 0.000010682120000637952 s 0.000010430199981783517 s 1.02
slicing / PartOpt / cpu / BothRev 0.000010702839972509535 s 0.000010264440015816943 s 1.04
slicing / IPartOpt / cpu / PreRev 0.000010369620003984891 s 0.000010120179967998411 s 1.02
slicing / IPartOpt / cpu / PostRev 0.000011384079980416571 s 0.000010441599988553209 s 1.09
slicing / IPartOpt / cpu / BothRev 0.000010748600006991182 s 0.000010460540033818689 s 1.03
slicing / DefOpt / cpu / PreRev 0.000010736260019257316 s 0.000010200260030615029 s 1.05
slicing / DefOpt / cpu / PostRev 0.000010555260005276068 s 0.000011166700032845255 s 0.95
slicing / DefOpt / cpu / BothRev 0.000011367719989721082 s 0.00001002842001071258 s 1.13
slicing / IDefOpt / cpu / PreRev 0.000011054640008296702 s 0.000010335939996366506 s 1.07
slicing / IDefOpt / cpu / PostRev 0.000011629860009634286 s 0.000011627660060185007 s 1.00
slicing / IDefOpt / cpu / BothRev 0.0000107345800006442 s 0.000010460059975230252 s 1.03
slicing / JaXPipe / tpu / Primal 9.6395e-7 s 0.0000010129 s 0.95
slicing / Jax / tpu / Primal 9.69525e-7 s 9.55525e-7 s 1.01
slicing / HLOOpt / tpu / Primal 9.67825e-7 s 0.000001009725 s 0.96
slicing / PartOpt / tpu / Primal 9.7025e-7 s 9.6235e-7 s 1.01
slicing / IPartOpt / tpu / Primal 9.7415e-7 s 0.00000101265 s 0.96
slicing / DefOpt / tpu / Primal 9.675e-7 s 9.594e-7 s 1.01
slicing / IDefOpt / tpu / Primal 9.643e-7 s 0.000001011825 s 0.95
slicing / JaXPipe / tpu / Forward 0.000001411475 s 0.000001399925 s 1.01
slicing / Jax / tpu / Forward 0.0000014168249999999998 s 0.0000014658750000000002 s 0.97
slicing / HLOOpt / tpu / Forward 0.0000015152 s 0.000001507225 s 1.01
slicing / PartOpt / tpu / Forward 0.00000143625 s 0.000001480825 s 0.97
slicing / IPartOpt / tpu / Forward 0.000001517 s 0.000001509875 s 1.00
slicing / DefOpt / tpu / Forward 0.0000014359749999999998 s 0.0000014848000000000002 s 0.97
slicing / IDefOpt / tpu / Forward 0.00000152695 s 0.0000015096 s 1.01
slicing / JaXPipe / tpu / PreRev 0.0000023814 s 0.0000025669750000000005 s 0.93
slicing / JaXPipe / tpu / PostRev 0.000002512425 s 0.0000025137500000000003 s 1.00
slicing / JaXPipe / tpu / BothRev 0.0000023967 s 0.0000025836750000000003 s 0.93
slicing / Jax / tpu / BothRev 0.000002540525 s 0.000002534225 s 1.00
slicing / HLOOpt / tpu / PreRev 0.00000240895 s 0.000002581275 s 0.93
slicing / HLOOpt / tpu / PostRev 0.000002543375 s 0.0000025394250000000003 s 1.00
slicing / HLOOpt / tpu / BothRev 0.000002406025 s 0.0000025819250000000003 s 0.93
slicing / PartOpt / tpu / PreRev 0.0000025438 s 0.0000025284 s 1.01
slicing / PartOpt / tpu / PostRev 0.000002397475 s 0.000002579375 s 0.93
slicing / PartOpt / tpu / BothRev 0.0000025354500000000004 s 0.000002540475 s 1.00
slicing / IPartOpt / tpu / PreRev 0.000002398425 s 0.00000257935 s 0.93
slicing / IPartOpt / tpu / PostRev 0.00000253405 s 0.000002533825 s 1.00
slicing / IPartOpt / tpu / BothRev 0.0000023909500000000004 s 0.000002585175 s 0.92
slicing / DefOpt / tpu / PreRev 0.00000254265 s 0.0000025453 s 1.00
slicing / DefOpt / tpu / PostRev 0.00000240235 s 0.0000025848 s 0.93
slicing / DefOpt / tpu / BothRev 0.00000254065 s 0.0000025439 s 1.00
slicing / IDefOpt / tpu / PreRev 0.00000240235 s 0.000002598525 s 0.92
slicing / IDefOpt / tpu / PostRev 0.000002541325 s 0.0000025577 s 0.99
slicing / IDefOpt / tpu / BothRev 0.0000023939 s 0.0000025835249999999995 s 0.93
slicing / JaXPipe / cpu / Primal 0.00001291 s 0.000007253260009747464 s 1.78
slicing / Jax / cpu / Primal 0.000012466 s 0.000006546520035044523 s 1.90
slicing / HLOOpt / cpu / Primal 0.000012318 s 0.000010940299989670166 s 1.13
slicing / PartOpt / cpu / Primal 0.000012435 s 0.000006298379948930233 s 1.97
slicing / IPartOpt / cpu / Primal 0.000012233 s 0.00000655244000881794 s 1.87
slicing / DefOpt / cpu / Primal 0.000012337 s 0.000010885439987760035 s 1.13
slicing / IDefOpt / cpu / Primal 0.00001221 s 0.00000650486000267847 s 1.88
slicing / JaXPipe / cpu / Forward 0.000016474 s 0.000009839760014074271 s 1.67
slicing / Jax / cpu / Forward 0.000016255 s 0.000010449979999975768 s 1.56
slicing / HLOOpt / cpu / Forward 0.000016117 s 0.000014228619993446046 s 1.13
slicing / PartOpt / cpu / Forward 0.00001647 s 0.00001480601998991915 s 1.11
slicing / IPartOpt / cpu / Forward 0.000016414999999999998 s 0.000009797279990380047 s 1.68
slicing / DefOpt / cpu / Forward 0.000016388 s 0.00001460727998164657 s 1.12
slicing / IDefOpt / cpu / Forward 0.000016539000000000002 s 0.000009516859954601386 s 1.74
slicing / JaXPipe / cpu / PreRev 0.000017539 s 0.000010368759976699948 s 1.69
slicing / JaXPipe / cpu / PostRev 0.000016833 s 0.00001086786002815643 s 1.55
slicing / JaXPipe / cpu / BothRev 0.000017548999999999997 s 0.000014082499974392704 s 1.25
slicing / Jax / cpu / BothRev 0.000017273 s 0.000010233800012429128 s 1.69
slicing / HLOOpt / cpu / PreRev 0.000017358 s 0.000010548579948590488 s 1.65
slicing / HLOOpt / cpu / PostRev 0.000017860000000000002 s 0.000010883419954552664 s 1.64
slicing / HLOOpt / cpu / BothRev 0.000017163 s 0.000011958500008404372 s 1.44
slicing / PartOpt / cpu / PreRev 0.00001717 s 0.000009984839980461402 s 1.72
slicing / PartOpt / cpu / PostRev 0.000016887 s 0.000010430199981783517 s 1.62
slicing / PartOpt / cpu / BothRev 0.000016846 s 0.000010264440015816943 s 1.64
slicing / IPartOpt / cpu / PreRev 0.000017824 s 0.000010120179967998411 s 1.76
slicing / IPartOpt / cpu / PostRev 0.00001833 s 0.000010441599988553209 s 1.76
slicing / IPartOpt / cpu / BothRev 0.000018072 s 0.000010460540033818689 s 1.73
slicing / DefOpt / cpu / PreRev 0.000017556 s 0.000010200260030615029 s 1.72
slicing / DefOpt / cpu / PostRev 0.000017766 s 0.000011166700032845255 s 1.59
slicing / DefOpt / cpu / BothRev 0.000017825 s 0.00001002842001071258 s 1.78
slicing / IDefOpt / cpu / PreRev 0.000017072000000000002 s 0.000010335939996366506 s 1.65
slicing / IDefOpt / cpu / PostRev 0.000017214 s 0.000011627660060185007 s 1.48
slicing / IDefOpt / cpu / BothRev 0.00001751 s 0.000010460059975230252 s 1.67
slicing / JaXPipe / cpu / Primal 0.000008 s 0.000007253260009747464 s 1.10
slicing / Jax / cpu / Primal 0.000008 s 0.000006546520035044523 s 1.22
slicing / HLOOpt / cpu / Primal 0.000008 s 0.000010940299989670166 s 0.73
slicing / PartOpt / cpu / Primal 0.000008 s 0.000006298379948930233 s 1.27
slicing / IPartOpt / cpu / Primal 0.000008 s 0.00000655244000881794 s 1.22
slicing / DefOpt / cpu / Primal 0.000008 s 0.000010885439987760035 s 0.73
slicing / IDefOpt / cpu / Primal 0.000008 s 0.00000650486000267847 s 1.23
slicing / JaXPipe / cpu / Forward 0.000011 s 0.000009839760014074271 s 1.12
slicing / Jax / cpu / Forward 0.000011 s 0.000010449979999975768 s 1.05
slicing / HLOOpt / cpu / Forward 0.000011 s 0.000014228619993446046 s 0.77
slicing / PartOpt / cpu / Forward 0.000011 s 0.00001480601998991915 s 0.74
slicing / IPartOpt / cpu / Forward 0.000012 s 0.000009797279990380047 s 1.22
slicing / DefOpt / cpu / Forward 0.000011 s 0.00001460727998164657 s 0.75
slicing / IDefOpt / cpu / Forward 0.000011 s 0.000009516859954601386 s 1.16
slicing / JaXPipe / cpu / PreRev 0.000011 s 0.000010368759976699948 s 1.06
slicing / JaXPipe / cpu / PostRev 0.000011 s 0.00001086786002815643 s 1.01
slicing / JaXPipe / cpu / BothRev 0.000011 s 0.000014082499974392704 s 0.78
slicing / Jax / cpu / BothRev 0.000011 s 0.000010233800012429128 s 1.07
slicing / HLOOpt / cpu / PreRev 0.000012 s 0.000010548579948590488 s 1.14
slicing / HLOOpt / cpu / PostRev 0.000011 s 0.000010883419954552664 s 1.01
slicing / HLOOpt / cpu / BothRev 0.000012 s 0.000011958500008404372 s 1.00
slicing / PartOpt / cpu / PreRev 0.000011 s 0.000009984839980461402 s 1.10
slicing / PartOpt / cpu / PostRev 0.000011 s 0.000010430199981783517 s 1.05
slicing / PartOpt / cpu / BothRev 0.000011 s 0.000010264440015816943 s 1.07
slicing / IPartOpt / cpu / PreRev 0.000011 s 0.000010120179967998411 s 1.09
slicing / IPartOpt / cpu / PostRev 0.000012 s 0.000010441599988553209 s 1.15
slicing / IPartOpt / cpu / BothRev 0.000012 s 0.000010460540033818689 s 1.15
slicing / DefOpt / cpu / PreRev 0.000011 s 0.000010200260030615029 s 1.08
slicing / DefOpt / cpu / PostRev 0.000011 s 0.000011166700032845255 s 0.99
slicing / DefOpt / cpu / BothRev 0.000012 s 0.00001002842001071258 s 1.20
slicing / IDefOpt / cpu / PreRev 0.000012 s 0.000010335939996366506 s 1.16
slicing / IDefOpt / cpu / PostRev 0.000012 s 0.000011627660060185007 s 1.03
slicing / IDefOpt / cpu / BothRev 0.000011 s 0.000010460059975230252 s 1.05
sum / JaXPipe / cpu / Primal 0.00000903263997315662 s 0.000008496379950884148 s 1.06
sum / Jax / cpu / Primal 0.000008197939987439895 s 0.000008115659993563895 s 1.01
sum / HLOOpt / cpu / Primal 0.000013719100033995346 s 0.000012578460009535774 s 1.09
sum / PartOpt / cpu / Primal 0.00000927783999031817 s 0.000007630900017829845 s 1.22
sum / IPartOpt / cpu / Primal 0.000008913939955164097 s 0.000007762439990983693 s 1.15
sum / DefOpt / cpu / Primal 0.000009336920029454632 s 0.000007727000047452748 s 1.21
sum / IDefOpt / cpu / Primal 0.000008922800070649828 s 0.000008056840015342458 s 1.11
sum / JaXPipe / cpu / Forward 0.000012861840004916303 s 0.000011537459977262188 s 1.11
sum / Jax / cpu / Forward 0.000012804759990103776 s 0.000011404559991206044 s 1.12
sum / HLOOpt / cpu / Forward 0.000017699319987514172 s 0.000017036620038197723 s 1.04
sum / PartOpt / cpu / Forward 0.00001959044000614085 s 0.000016377980009565364 s 1.20
sum / IPartOpt / cpu / Forward 0.00001295150002988521 s 0.000011404640008549903 s 1.14
sum / DefOpt / cpu / Forward 0.000017629800022405108 s 0.000017290699915974982 s 1.02
sum / IDefOpt / cpu / Forward 0.000012806600016119774 s 0.000011437059974923612 s 1.12
sum / JaXPipe / cpu / PreRev 0.000011964360001002203 s 0.000010980160031976994 s 1.09
sum / JaXPipe / cpu / PostRev 0.000012416879999364027 s 0.000011516420008774731 s 1.08
sum / JaXPipe / cpu / BothRev 0.000011637319967121584 s 0.000012983800006622914 s 0.90
sum / Jax / cpu / BothRev 0.000012390899964884738 s 0.000011333820002619178 s 1.09
sum / HLOOpt / cpu / PreRev 0.000011462580014267589 s 0.000010923279987764544 s 1.05
sum / HLOOpt / cpu / PostRev 0.000015654220023861855 s 0.000010646139990058146 s 1.47
sum / HLOOpt / cpu / BothRev 0.000013502800011337969 s 0.000012727360017379397 s 1.06
sum / PartOpt / cpu / PreRev 0.000011753039998438908 s 0.000011015020063496195 s 1.07
sum / PartOpt / cpu / PostRev 0.000012304299989409628 s 0.000011112040010630152 s 1.11
sum / PartOpt / cpu / BothRev 0.000011991940018560854 s 0.00001066772003468941 s 1.12
sum / IPartOpt / cpu / PreRev 0.00001672601997597667 s 0.000016297819993269513 s 1.03
sum / IPartOpt / cpu / PostRev 0.00001180083999315684 s 0.000011135899985674768 s 1.06
sum / IPartOpt / cpu / BothRev 0.000011649500002022251 s 0.00001092540001081943 s 1.07
sum / DefOpt / cpu / PreRev 0.00001160278000497783 s 0.00001054432000273664 s 1.10
sum / DefOpt / cpu / PostRev 0.000011957380011153873 s 0.0000110990799839783 s 1.08
sum / DefOpt / cpu / BothRev 0.00001130361997638829 s 0.000011000960048477282 s 1.03
sum / IDefOpt / cpu / PreRev 0.000011803960023826222 s 0.00001122497998039762 s 1.05
sum / IDefOpt / cpu / PostRev 0.00001178631999209756 s 0.00001084379997337237 s 1.09
sum / IDefOpt / cpu / BothRev 0.0000116520199480874 s 0.000011360959961166373 s 1.03
sum / JaXPipe / tpu / Primal 5.102750000000001e-7 s 5.102e-7 s 1.00
sum / Jax / tpu / Primal 5.570250000000001e-7 s 5.5785e-7 s 1.00
sum / HLOOpt / tpu / Primal 5.211e-7 s 5.213e-7 s 1.00
sum / PartOpt / tpu / Primal 5.57e-7 s 5.57875e-7 s 1.00
sum / IPartOpt / tpu / Primal 5.20175e-7 s 5.24375e-7 s 0.99
sum / DefOpt / tpu / Primal 5.569e-7 s 5.581e-7 s 1.00
sum / IDefOpt / tpu / Primal 5.204250000000001e-7 s 5.213e-7 s 1.00
sum / JaXPipe / tpu / Forward 0.00000155555 s 0.000001549625 s 1.00
sum / Jax / tpu / Forward 0.0000015010499999999998 s 0.0000015034 s 1.00
sum / HLOOpt / tpu / Forward 0.00000154225 s 0.00000153625 s 1.00
sum / PartOpt / tpu / Forward 0.0000015081 s 0.0000015024750000000002 s 1.00
sum / IPartOpt / tpu / Forward 0.000001536825 s 0.000001533175 s 1.00
sum / DefOpt / tpu / Forward 0.000001495125 s 0.000001499675 s 1.00
sum / IDefOpt / tpu / Forward 0.000001551575 s 0.000001529 s 1.01
sum / JaXPipe / tpu / PreRev 0.000001001825 s 0.0000010481 s 0.96
sum / JaXPipe / tpu / PostRev 0.000001042125 s 0.00000108655 s 0.96
sum / JaXPipe / tpu / BothRev 0.000001003775 s 0.00000104865 s 0.96
sum / Jax / tpu / BothRev 0.0000010456999999999998 s 0.000001084975 s 0.96
sum / HLOOpt / tpu / PreRev 0.0000010112000000000002 s 0.00000104805 s 0.96
sum / HLOOpt / tpu / PostRev 0.000001049775 s 0.00000108465 s 0.97
sum / HLOOpt / tpu / BothRev 0.00000100585 s 0.00000104875 s 0.96
sum / PartOpt / tpu / PreRev 0.00000105495 s 0.0000010845 s 0.97
sum / PartOpt / tpu / PostRev 0.000001007975 s 0.000001047175 s 0.96
sum / PartOpt / tpu / BothRev 0.000001038 s 0.0000010845499999999998 s 0.96
sum / IPartOpt / tpu / PreRev 0.0000010069 s 0.00000104845 s 0.96
sum / IPartOpt / tpu / PostRev 0.0000010447499999999998 s 0.0000010851750000000002 s 0.96
sum / IPartOpt / tpu / BothRev 0.000001005675 s 0.0000010505 s 0.96
sum / DefOpt / tpu / PreRev 0.000001038075 s 0.0000010831 s 0.96
sum / DefOpt / tpu / PostRev 9.9755e-7 s 0.0000010478 s 0.95
sum / DefOpt / tpu / BothRev 0.000001039775 s 0.000001085675 s 0.96
sum / IDefOpt / tpu / PreRev 0.000001005825 s 0.0000010497 s 0.96
sum / IDefOpt / tpu / PostRev 0.00000104235 s 0.0000010868 s 0.96
sum / IDefOpt / tpu / BothRev 0.000001000725 s 0.00000105115 s 0.95
sum / JaXPipe / cpu / Primal 0.000014459 s 0.000008496379950884148 s 1.70
sum / Jax / cpu / Primal 0.000014129 s 0.000008115659993563895 s 1.74
sum / HLOOpt / cpu / Primal 0.000014636 s 0.000012578460009535774 s 1.16
sum / PartOpt / cpu / Primal 0.000014495 s 0.000007630900017829845 s 1.90
sum / IPartOpt / cpu / Primal 0.000014369 s 0.000007762439990983693 s 1.85
sum / DefOpt / cpu / Primal 0.00001481 s 0.000007727000047452748 s 1.92
sum / IDefOpt / cpu / Primal 0.000014459 s 0.000008056840015342458 s 1.79
sum / JaXPipe / cpu / Forward 0.000019969 s 0.000011537459977262188 s 1.73
sum / Jax / cpu / Forward 0.000019274 s 0.000011404559991206044 s 1.69
sum / HLOOpt / cpu / Forward 0.000019225 s 0.000017036620038197723 s 1.13
sum / PartOpt / cpu / Forward 0.000019549 s 0.000016377980009565364 s 1.19
sum / IPartOpt / cpu / Forward 0.000019639 s 0.000011404640008549903 s 1.72
sum / DefOpt / cpu / Forward 0.000019309 s 0.000017290699915974982 s 1.12
sum / IDefOpt / cpu / Forward 0.000019538 s 0.000011437059974923612 s 1.71
sum / JaXPipe / cpu / PreRev 0.000019187 s 0.000010980160031976994 s 1.75
sum / JaXPipe / cpu / PostRev 0.000018513 s 0.000011516420008774731 s 1.61
sum / JaXPipe / cpu / BothRev 0.000019144 s 0.000012983800006622914 s 1.47
sum / Jax / cpu / BothRev 0.000018746 s 0.000011333820002619178 s 1.65
sum / HLOOpt / cpu / PreRev 0.000018558 s 0.000010923279987764544 s 1.70
sum / HLOOpt / cpu / PostRev 0.000019778 s 0.000010646139990058146 s 1.86
sum / HLOOpt / cpu / BothRev 0.000019181 s 0.000012727360017379397 s 1.51
sum / PartOpt / cpu / PreRev 0.000018781 s 0.000011015020063496195 s 1.71
sum / PartOpt / cpu / PostRev 0.000019754 s 0.000011112040010630152 s 1.78
sum / PartOpt / cpu / BothRev 0.000019036 s 0.00001066772003468941 s 1.78
sum / IPartOpt / cpu / PreRev 0.000019025 s 0.000016297819993269513 s 1.17
sum / IPartOpt / cpu / PostRev 0.000018643 s 0.000011135899985674768 s 1.67
sum / IPartOpt / cpu / BothRev 0.000019427 s 0.00001092540001081943 s 1.78
sum / DefOpt / cpu / PreRev 0.000019092000000000003 s 0.00001054432000273664 s 1.81
sum / DefOpt / cpu / PostRev 0.000019186 s 0.0000110990799839783 s 1.73
sum / DefOpt / cpu / BothRev 0.000019659 s 0.000011000960048477282 s 1.79
sum / IDefOpt / cpu / PreRev 0.000018561 s 0.00001122497998039762 s 1.65
sum / IDefOpt / cpu / PostRev 0.000019383 s 0.00001084379997337237 s 1.79
sum / IDefOpt / cpu / BothRev 0.000019345 s 0.000011360959961166373 s 1.70
sum / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000008496379950884148 s 1.06
sum / Jax / cpu / Primal 0.000008999999999999999 s 0.000008115659993563895 s 1.11
sum / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000012578460009535774 s 0.72
sum / PartOpt / cpu / Primal 0.000008999999999999999 s 0.000007630900017829845 s 1.18
sum / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000007762439990983693 s 1.16
sum / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000007727000047452748 s 1.16
sum / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000008056840015342458 s 1.12
sum / JaXPipe / cpu / Forward 0.000013 s 0.000011537459977262188 s 1.13
sum / Jax / cpu / Forward 0.000013 s 0.000011404559991206044 s 1.14
sum / HLOOpt / cpu / Forward 0.000013 s 0.000017036620038197723 s 0.76
sum / PartOpt / cpu / Forward 0.000014 s 0.000016377980009565364 s 0.85
sum / IPartOpt / cpu / Forward 0.000013 s 0.000011404640008549903 s 1.14
sum / DefOpt / cpu / Forward 0.000014 s 0.000017290699915974982 s 0.81
sum / IDefOpt / cpu / Forward 0.000013 s 0.000011437059974923612 s 1.14
sum / JaXPipe / cpu / PreRev 0.000012 s 0.000010980160031976994 s 1.09
sum / JaXPipe / cpu / PostRev 0.000014 s 0.000011516420008774731 s 1.22
sum / JaXPipe / cpu / BothRev 0.000013 s 0.000012983800006622914 s 1.00
sum / Jax / cpu / BothRev 0.000013 s 0.000011333820002619178 s 1.15
sum / HLOOpt / cpu / PreRev 0.000012 s 0.000010923279987764544 s 1.10
sum / HLOOpt / cpu / PostRev 0.000012 s 0.000010646139990058146 s 1.13
sum / HLOOpt / cpu / BothRev 0.000012 s 0.000012727360017379397 s 0.94
sum / PartOpt / cpu / PreRev 0.000012 s 0.000011015020063496195 s 1.09
sum / PartOpt / cpu / PostRev 0.000012 s 0.000011112040010630152 s 1.08
sum / PartOpt / cpu / BothRev 0.000013 s 0.00001066772003468941 s 1.22
sum / IPartOpt / cpu / PreRev 0.000013 s 0.000016297819993269513 s 0.80
sum / IPartOpt / cpu / PostRev 0.000012 s 0.000011135899985674768 s 1.08
sum / IPartOpt / cpu / BothRev 0.000013 s 0.00001092540001081943 s 1.19
sum / DefOpt / cpu / PreRev 0.000014 s 0.00001054432000273664 s 1.33
sum / DefOpt / cpu / PostRev 0.000013 s 0.0000110990799839783 s 1.17
sum / DefOpt / cpu / BothRev 0.000013 s 0.000011000960048477282 s 1.18
sum / IDefOpt / cpu / PreRev 0.000013 s 0.00001122497998039762 s 1.16
sum / IDefOpt / cpu / PostRev 0.000013 s 0.00001084379997337237 s 1.20
sum / IDefOpt / cpu / BothRev 0.000013 s 0.000011360959961166373 s 1.14
value_and_grad / JaXPipe / cpu / Primal 0.000016221320011027273 s 0.0000151070399897435 s 1.07
value_and_grad / Jax / cpu / Primal 0.00001636039994082239 s 0.000014468839963228674 s 1.13
value_and_grad / HLOOpt / cpu / Primal 0.0000153297399810981 s 0.000013816899991070384 s 1.11
value_and_grad / PartOpt / cpu / Primal 0.000014555499992638945 s 0.0000145160000101896 s 1.00
value_and_grad / IPartOpt / cpu / Primal 0.000015121959995667566 s 0.00001387369999974908 s 1.09
value_and_grad / DefOpt / cpu / Primal 0.00001572594002936967 s 0.000014400200034287992 s 1.09
value_and_grad / IDefOpt / cpu / Primal 0.000015286619991456973 s 0.000013790359998893107 s 1.11
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 0.000023481 s 0.0000151070399897435 s 1.55
value_and_grad / Jax / cpu / Primal 0.0000221 s 0.000014468839963228674 s 1.53
value_and_grad / HLOOpt / cpu / Primal 0.00002285 s 0.000013816899991070384 s 1.65
value_and_grad / PartOpt / cpu / Primal 0.000023015 s 0.0000145160000101896 s 1.59
value_and_grad / IPartOpt / cpu / Primal 0.000023238 s 0.00001387369999974908 s 1.67
value_and_grad / DefOpt / cpu / Primal 0.000022145 s 0.000014400200034287992 s 1.54
value_and_grad / IDefOpt / cpu / Primal 0.000023153 s 0.000013790359998893107 s 1.68
value_and_grad / JaXPipe / cpu / Primal 0.000016 s 0.0000151070399897435 s 1.06
value_and_grad / Jax / cpu / Primal 0.000016 s 0.000014468839963228674 s 1.11
value_and_grad / HLOOpt / cpu / Primal 0.000015 s 0.000013816899991070384 s 1.09
value_and_grad / PartOpt / cpu / Primal 0.000016 s 0.0000145160000101896 s 1.10
value_and_grad / IPartOpt / cpu / Primal 0.000016 s 0.00001387369999974908 s 1.15
value_and_grad / DefOpt / cpu / Primal 0.000016 s 0.000014400200034287992 s 1.11
value_and_grad / IDefOpt / cpu / Primal 0.000016 s 0.000013790359998893107 s 1.16
jaxmd20 / JaXPipe / tpu / Primal 0.009278645625 s 0.0092721981249999 s 1.00
jaxmd20 / Jax / tpu / Primal 0.00927935125 s 0.009264746875 s 1.00
jaxmd20 / HLOOpt / tpu / Primal 0.009169834375 s 0.009164909375 s 1.00
jaxmd20 / PartOpt / tpu / Primal 0.0091992575 s 0.00919699375 s 1.00
jaxmd20 / IPartOpt / tpu / Primal 0.009197651875 s 0.009201133125 s 1.00
jaxmd20 / DefOpt / tpu / Primal 0.008750145625 s 0.0087451 s 1.00
jaxmd20 / IDefOpt / tpu / Primal 0.008632813125 s 0.008632471875 s 1.00
jaxmd20 / JaXPipe / tpu / Forward 0.0172678325 s 0.017263674375 s 1.00
jaxmd20 / Jax / tpu / Forward 0.018726079375 s 0.018731593125 s 1.00
jaxmd20 / HLOOpt / tpu / Forward 0.017240084375 s 0.017237675625 s 1.00
jaxmd20 / PartOpt / tpu / Forward 0.01725991125 s 0.01726607375 s 1.00
jaxmd20 / IPartOpt / tpu / Forward 0.017263623125 s 0.01726384375 s 1.00
jaxmd20 / DefOpt / tpu / Forward 0.0172629924999999 s 0.0172619875 s 1.00
jaxmd20 / IDefOpt / tpu / Forward 0.0172669537499999 s 0.017263194375 s 1.00
jaxmd20 / JaXPipe / tpu / PreRev 0.025363226875 s 0.0253415125 s 1.00
jaxmd20 / JaXPipe / tpu / PostRev 0.0218589056249999 s 0.0218886456249999 s 1.00
jaxmd20 / JaXPipe / tpu / BothRev 0.025365874375 s 0.025359194375 s 1.00
jaxmd20 / Jax / tpu / BothRev 0.021858911875 s 0.021893316875 s 1.00
jaxmd20 / HLOOpt / tpu / PreRev 0.0253526475 s 0.025354973125 s 1.00
jaxmd20 / HLOOpt / tpu / PostRev 0.020971758125 s 0.020989665625 s 1.00
jaxmd20 / HLOOpt / tpu / BothRev 0.0252693125 s 0.025265778125 s 1.00
jaxmd20 / PartOpt / tpu / PreRev 0.02536424875 s 0.025360244375 s 1.00
jaxmd20 / PartOpt / tpu / PostRev 0.021506115625 s 0.02151421625 s 1.00
jaxmd20 / PartOpt / tpu / BothRev 0.02528507875 s 0.02529064375 s 1.00
jaxmd20 / IPartOpt / tpu / PreRev 0.025352085625 s 0.025349443125 s 1.00
jaxmd20 / IPartOpt / tpu / PostRev 0.02152209375 s 0.021537013125 s 1.00
jaxmd20 / IPartOpt / tpu / BothRev 0.0252635131249999 s 0.0252666025 s 1.00
jaxmd20 / DefOpt / tpu / PreRev 0.02536238375 s 0.0253634375 s 1.00
jaxmd20 / DefOpt / tpu / PostRev 0.018881056875 s 0.01877251625 s 1.01
jaxmd20 / DefOpt / tpu / BothRev 0.025280159375 s 0.025284674375 s 1.00
jaxmd20 / IDefOpt / tpu / PreRev 0.025352213125 s 0.025350095625 s 1.00
jaxmd20 / IDefOpt / tpu / PostRev 0.0183985068749999 s 0.0181213362499999 s 1.02
jaxmd20 / IDefOpt / tpu / BothRev 0.0252666925 s 0.02526203625 s 1.00
jaxmd40 / JaXPipe / cpu / Primal 0.066482437 s 0.064166112 s 1.04
jaxmd40 / Jax / cpu / Primal 0.062555655 s 0.05890284 s 1.06
jaxmd40 / HLOOpt / cpu / Primal 0.084296924 s 0.09079052 s 0.93
jaxmd40 / PartOpt / cpu / Primal 0.0740173819999999 s 0.068332119 s 1.08
jaxmd40 / IPartOpt / cpu / Primal 0.063397315 s 0.0697432759999999 s 0.91
jaxmd40 / DefOpt / cpu / Primal 0.088728013 s 0.088063969 s 1.01
jaxmd40 / IDefOpt / cpu / Primal 0.091856621 s 0.08888997 s 1.03
jaxmd40 / JaXPipe / cpu / Forward 0.163530489 s 0.175983801 s 0.93
jaxmd40 / Jax / cpu / Forward 0.09028351 s 0.085674646 s 1.05
jaxmd40 / HLOOpt / cpu / Forward 0.171011659 s 0.17782387 s 0.96
jaxmd40 / PartOpt / cpu / Forward 0.17906061 s 0.155398343 s 1.15
jaxmd40 / IPartOpt / cpu / Forward 0.164828605 s 0.156777011 s 1.05
jaxmd40 / DefOpt / cpu / Forward 0.179044703 s 0.16223865 s 1.10
jaxmd40 / IDefOpt / cpu / Forward 0.175765772 s 0.153983223 s 1.14
jaxmd40 / JaXPipe / cpu / PreRev 0.237992751 s 0.215310808 s 1.11
jaxmd40 / JaXPipe / cpu / PostRev 0.148335897 s 0.13967266 s 1.06
jaxmd40 / JaXPipe / cpu / BothRev 0.244074958 s 0.212643407 s 1.15
jaxmd40 / Jax / cpu / BothRev 0.145243478 s 0.152692364 s 0.95
jaxmd40 / HLOOpt / cpu / PreRev 0.2368794629999999 s 0.222208975 s 1.07
jaxmd40 / HLOOpt / cpu / PostRev 0.1854616389999999 s 0.172703218 s 1.07
jaxmd40 / HLOOpt / cpu / BothRev 0.251177356 s 0.2494497849999999 s 1.01
jaxmd40 / PartOpt / cpu / PreRev 0.236831294 s 0.231103632 s 1.02
jaxmd40 / PartOpt / cpu / PostRev 0.142542244 s 0.1311307099999999 s 1.09
jaxmd40 / PartOpt / cpu / BothRev 0.269544394 s 0.247571008 s 1.09
jaxmd40 / IPartOpt / cpu / PreRev 0.2294691479999999 s 0.227675682 s 1.01
jaxmd40 / IPartOpt / cpu / PostRev 0.141298808 s 0.130463223 s 1.08
jaxmd40 / IPartOpt / cpu / BothRev 0.2614049 s 0.250275015 s 1.04
jaxmd40 / DefOpt / cpu / PreRev 0.22405485 s 0.232259637 s 0.96
jaxmd40 / DefOpt / cpu / PostRev 0.182276702 s 0.174756321 s 1.04
jaxmd40 / DefOpt / cpu / BothRev 0.251102428 s 0.251723461 s 1.00
jaxmd40 / IDefOpt / cpu / PreRev 0.23051359 s 0.214849338 s 1.07
jaxmd40 / IDefOpt / cpu / PostRev 0.1764732919999999 s 0.175905601 s 1.00
jaxmd40 / IDefOpt / cpu / BothRev 0.253487939 s 0.240092206 s 1.06
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / tpu / Primal 4.01084235 s 4.027093965 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / tpu / Primal 3.03899991 s 3.038966779375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / tpu / Primal 3.121484056875 s 3.12132557 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / tpu / Primal 3.059339619375 s 3.059262556875 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / tpu / Primal 3.059281688125 s 3.059269564375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / tpu / Primal 2.263579895625 s 2.102676941875 s 1.08
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / tpu / Primal 4.74326887 s 4.356334549375 s 1.09
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / JaXPipe / cpu / Primal 6.292014701 s 6.100846673 s 1.03
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / Jax / cpu / Primal 6.223152087000001 s 6.003130034 s 1.04
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / HLOOpt / cpu / Primal 6.260842503 s 5.945582822 s 1.05
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / PartOpt / cpu / Primal 6.276197390999999 s 5.98866533 s 1.05
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IPartOpt / cpu / Primal 6.307711518 s 5.954800858 s 1.06
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / DefOpt / cpu / Primal 2.519413333 s 2.298179307 s 1.10
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IDefOpt / cpu / Primal 6.862494329 s 6.523067271 s 1.05

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

@ftynse ftynse marked this pull request as ready for review January 29, 2026 22:45
Copy link
Member

@wsmoses wsmoses left a comment

Choose a reason for hiding this comment

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

can you add the requisite test, and add to transform ops?

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