Skip to content

Conversation

@xys-syx
Copy link
Collaborator

@xys-syx xys-syx commented Jan 4, 2026

llvm.func internal @reactant$_Z18__device_stub__fooPi(%arg0: !llvm.ptr {llvm.nocapture, llvm.noundef, llvm.writeonly}) comdat(@__llvm_global_comdat::@foo) attributes {dso_local, frame_pointer = #llvm.framePointerKind<all>, no_infs_fp_math = true, no_inline, no_nans_fp_math = true, no_signed_zeros_fp_math = true, no_unwind, passthrough = ["mustprogress", "nofree", "norecurse", "nosync", ["approx-func-fp-math", "true"], ["no-trapping-math", "true"], ["stack-protector-buffer-size", "8"], ["target-cpu", "sm_120"], ["uniform-work-group-size", "true"]], sym_visibility = "private", target_cpu = "sm_120", target_features = #llvm.target_features<["+ptx88", "+sm_120"]>, unsafe_fp_math = true, will_return} {

@wsmoses @vimarsh6739
Hello, I have several questions that block the progress:

  1. Could you tell where I can find an example for AMD GPU's corresponding to this?
    I only can find something like this:
    https://github.com/llvm/llvm-project/blob/c97de4387b076176d3dbd428229a03b8909941af/clang/test/Driver/amdgpu-mcpu.cl
    https://github.com/llvm/llvm-project/blob/c97de4387b076176d3dbd428229a03b8909941af/clang/test/CodeGenOpenCL/amdgpu-features.cl
    They are not exactly match and I cannot find an example with abiVersion according to https://github.com/llvm/llvm-project/blob/c97de4387b076176d3dbd428229a03b8909941af/mlir/include/mlir/Dialect/LLVMIR/ROCDLOps.td#L2147

  2. And, according to this: https://github.com/llvm/llvm-project/blob/ccb47d0fb9d01d44764fa4ca5c6dcf239ab76ed2/mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td#L6184
    The default parameter for NVVMTargetAttr is not the same comparing to llvm-project, should I modify or not?

@xys-syx xys-syx marked this pull request as draft January 4, 2026 07:49
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: eebedcf Previous: a6af9b3 Ratio
actmtch / JaXPipe / cpu / Primal 0.000007231620011225459 s 0.000007719879995420343 s 0.94
actmtch / Jax / cpu / Primal 0.000007083519994921516 s 0.0000071134600057121135 s 1.00
actmtch / HLOOpt / cpu / Primal 0.00000897795996934292 s 0.000008057620034378488 s 1.11
actmtch / PartOpt / cpu / Primal 0.000008476840048388112 s 0.000006819159971200861 s 1.24
actmtch / IPartOpt / cpu / Primal 0.000008164119972207118 s 0.0000071537999974680136 s 1.14
actmtch / DefOpt / cpu / Primal 0.000009231239992004704 s 0.000007619359957971028 s 1.21
actmtch / IDefOpt / cpu / Primal 0.000009373800003231736 s 0.000007297920028577209 s 1.28
actmtch / JaXPipe / cpu / Forward 0.000012270520037418463 s 0.000010985560029439512 s 1.12
actmtch / Jax / cpu / Forward 0.000011345339971740032 s 0.000010219640034847544 s 1.11
actmtch / HLOOpt / cpu / Forward 0.00001272056003472244 s 0.000011537560048964224 s 1.10
actmtch / PartOpt / cpu / Forward 0.000012344540009507909 s 0.000011458680010036916 s 1.08
actmtch / IPartOpt / cpu / Forward 0.000013333300003068871 s 0.000011927800023840972 s 1.12
actmtch / DefOpt / cpu / Forward 0.000012724440048259566 s 0.00001134841999373748 s 1.12
actmtch / IDefOpt / cpu / Forward 0.000012354179980320622 s 0.000011242060008953558 s 1.10
actmtch / JaXPipe / cpu / PreRev 0.000012289679962123044 s 0.000010634620048222132 s 1.16
actmtch / JaXPipe / cpu / PostRev 0.000012271799987502164 s 0.000010018360007961748 s 1.22
actmtch / JaXPipe / cpu / BothRev 0.000012912379979752586 s 0.000011966400006713229 s 1.08
actmtch / Jax / cpu / BothRev 0.00001031790001434274 s 0.000009409679923919612 s 1.10
actmtch / HLOOpt / cpu / PreRev 0.000012520180016508677 s 0.000011981659990851768 s 1.04
actmtch / HLOOpt / cpu / PostRev 0.000014861840008961736 s 0.000014166439987093329 s 1.05
actmtch / HLOOpt / cpu / BothRev 0.000012889699992228996 s 0.000010792540015245322 s 1.19
actmtch / PartOpt / cpu / PreRev 0.000011934660005863406 s 0.00001240695998603769 s 0.96
actmtch / PartOpt / cpu / PostRev 0.000011677779993988223 s 0.000010077860024466644 s 1.16
actmtch / PartOpt / cpu / BothRev 0.000013563199991040164 s 0.000011556859981283196 s 1.17
actmtch / IPartOpt / cpu / PreRev 0.000012113479997424292 s 0.000011114280014226096 s 1.09
actmtch / IPartOpt / cpu / PostRev 0.000011107159989478533 s 0.000010391919977337238 s 1.07
actmtch / IPartOpt / cpu / BothRev 0.000012543740003820858 s 0.000011432819974288576 s 1.10
actmtch / DefOpt / cpu / PreRev 0.0000122190800448152 s 0.000010761699995782691 s 1.14
actmtch / DefOpt / cpu / PostRev 0.000013079459986329311 s 0.000010955220031974024 s 1.19
actmtch / DefOpt / cpu / BothRev 0.000013364919968807954 s 0.00001106313996388053 s 1.21
actmtch / IDefOpt / cpu / PreRev 0.000011928880048799329 s 0.000011028760045519448 s 1.08
actmtch / IDefOpt / cpu / PostRev 0.000012913579948872212 s 0.000011024740015272982 s 1.17
actmtch / IDefOpt / cpu / BothRev 0.000013238940027804348 s 0.000011336180004946071 s 1.17
actmtch / JaXPipe / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / Jax / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / HLOOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / PartOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / IPartOpt / cuda / Primal 0.0000024 s 0.000002015 s 1.19
actmtch / DefOpt / cuda / Primal 0.0000024 s 0.000002016 s 1.19
actmtch / IDefOpt / cuda / Primal 0.0000024 s 0.000002015 s 1.19
actmtch / JaXPipe / cuda / Forward 0.000009568 s 0.000009408 s 1.02
actmtch / Jax / cuda / Forward 0.00001056 s 0.000010016 s 1.05
actmtch / HLOOpt / cuda / Forward 0.000012929 s 0.000009792 s 1.32
actmtch / PartOpt / cuda / Forward 0.000010591 s 0.000009568 s 1.11
actmtch / IPartOpt / cuda / Forward 0.000010496 s 0.000009696 s 1.08
actmtch / DefOpt / cuda / Forward 0.000011008 s 0.000010144 s 1.09
actmtch / IDefOpt / cuda / Forward 0.000010656 s 0.000010144 s 1.05
actmtch / JaXPipe / cuda / PreRev 0.000010784 s 0.000009664 s 1.12
actmtch / JaXPipe / cuda / PostRev 0.000010815 s 0.000011455999999999998 s 0.94
actmtch / JaXPipe / cuda / BothRev 0.000010848 s 0.000010112 s 1.07
actmtch / Jax / cuda / BothRev 0.000010976 s 0.000010431 s 1.05
actmtch / HLOOpt / cuda / PreRev 0.000011008 s 0.000010752 s 1.02
actmtch / HLOOpt / cuda / PostRev 0.000011008 s 0.000010048 s 1.10
actmtch / HLOOpt / cuda / BothRev 0.000010944 s 0.000010176 s 1.08
actmtch / PartOpt / cuda / PreRev 0.000011072 s 0.000009568 s 1.16
actmtch / PartOpt / cuda / PostRev 0.000010655 s 0.000010176 s 1.05
actmtch / PartOpt / cuda / BothRev 0.000011071 s 0.000010176 s 1.09
actmtch / IPartOpt / cuda / PreRev 0.000011071 s 0.000010209 s 1.08
actmtch / IPartOpt / cuda / PostRev 0.000011104 s 0.000010111 s 1.10
actmtch / IPartOpt / cuda / BothRev 0.000010688 s 0.000009856 s 1.08
actmtch / DefOpt / cuda / PreRev 0.000011264 s 0.000009696 s 1.16
actmtch / DefOpt / cuda / PostRev 0.000010656 s 0.000009888 s 1.08
actmtch / DefOpt / cuda / BothRev 0.00001088 s 0.000010079 s 1.08
actmtch / IDefOpt / cuda / PreRev 0.000010848 s 0.000010336 s 1.05
actmtch / IDefOpt / cuda / PostRev 0.000010815 s 0.000010368 s 1.04
actmtch / IDefOpt / cuda / BothRev 0.000010369 s 0.000009952 s 1.04
actmtch / JaXPipe / tpu / Primal 5.6315e-7 s 5.630500000000001e-7 s 1.00
actmtch / Jax / tpu / Primal 5.96575e-7 s 5.967999999999999e-7 s 1.00
actmtch / HLOOpt / tpu / Primal 0.00000209565 s 0.00000209635 s 1.00
actmtch / PartOpt / tpu / Primal 5.96875e-7 s 5.965499999999999e-7 s 1.00
actmtch / IPartOpt / tpu / Primal 5.52675e-7 s 5.5255e-7 s 1.00
actmtch / DefOpt / tpu / Primal 0.000002155125 s 0.0000021702 s 0.99
actmtch / IDefOpt / tpu / Primal 0.0000020988 s 0.000002107375 s 1.00
actmtch / JaXPipe / tpu / Forward 0.000003826975 s 0.000003831725 s 1.00
actmtch / Jax / tpu / Forward 0.000001209275 s 0.00000121175 s 1.00
actmtch / HLOOpt / tpu / Forward 0.0000039447 s 0.000003942675 s 1.00
actmtch / PartOpt / tpu / Forward 0.000003929825 s 0.000003912325 s 1.00
actmtch / IPartOpt / tpu / Forward 0.000003931475 s 0.000003925624999999999 s 1.00
actmtch / DefOpt / tpu / Forward 0.000003913975 s 0.000003907 s 1.00
actmtch / IDefOpt / tpu / Forward 0.00000392855 s 0.0000039376 s 1.00
actmtch / JaXPipe / tpu / PreRev 0.000003458525 s 0.000003479125 s 0.99
actmtch / JaXPipe / tpu / PostRev 0.0000016368 s 0.00000164015 s 1.00
actmtch / JaXPipe / tpu / BothRev 0.0000034755000000000004 s 0.00000347485 s 1.00
actmtch / Jax / tpu / BothRev 0.0000016433249999999998 s 0.00000165015 s 1.00
actmtch / HLOOpt / tpu / PreRev 0.0000034859250000000004 s 0.000003480225 s 1.00
actmtch / HLOOpt / tpu / PostRev 0.0000034018500000000003 s 0.00000341245 s 1.00
actmtch / HLOOpt / tpu / BothRev 0.000003488425 s 0.0000034801 s 1.00
actmtch / PartOpt / tpu / PreRev 0.000003409825 s 0.0000034184 s 1.00
actmtch / PartOpt / tpu / PostRev 0.0000015884250000000002 s 0.0000015953999999999998 s 1.00
actmtch / PartOpt / tpu / BothRev 0.00000342355 s 0.000003425575 s 1.00
actmtch / IPartOpt / tpu / PreRev 0.000003493375 s 0.000003479925 s 1.00
actmtch / IPartOpt / tpu / PostRev 0.000001632125 s 0.000001635125 s 1.00
actmtch / IPartOpt / tpu / BothRev 0.00000347795 s 0.0000034748 s 1.00
actmtch / DefOpt / tpu / PreRev 0.000003418925 s 0.0000034205000000000005 s 1.00
actmtch / DefOpt / tpu / PostRev 0.00000341995 s 0.000003401125 s 1.01
actmtch / DefOpt / tpu / BothRev 0.0000034109250000000003 s 0.0000034220250000000005 s 1.00
actmtch / IDefOpt / tpu / PreRev 0.0000034811 s 0.000003478225 s 1.00
actmtch / IDefOpt / tpu / PostRev 0.000003417325 s 0.0000034033 s 1.00
actmtch / IDefOpt / tpu / BothRev 0.0000034729 s 0.000003481375 s 1.00
actmtch / JaXPipe / cpu / Primal 0.000012963 s 0.000007719879995420343 s 1.68
actmtch / Jax / cpu / Primal 0.000013277 s 0.0000071134600057121135 s 1.87
actmtch / HLOOpt / cpu / Primal 0.000014007 s 0.000008057620034378488 s 1.74
actmtch / PartOpt / cpu / Primal 0.000013162 s 0.000006819159971200861 s 1.93
actmtch / IPartOpt / cpu / Primal 0.000013404 s 0.0000071537999974680136 s 1.87
actmtch / DefOpt / cpu / Primal 0.000014029 s 0.000007619359957971028 s 1.84
actmtch / IDefOpt / cpu / Primal 0.000014092 s 0.000007297920028577209 s 1.93
actmtch / JaXPipe / cpu / Forward 0.000018708 s 0.000010985560029439512 s 1.70
actmtch / Jax / cpu / Forward 0.000017901 s 0.000010219640034847544 s 1.75
actmtch / HLOOpt / cpu / Forward 0.000018578 s 0.000011537560048964224 s 1.61
actmtch / PartOpt / cpu / Forward 0.000018736 s 0.000011458680010036916 s 1.64
actmtch / IPartOpt / cpu / Forward 0.000018638 s 0.000011927800023840972 s 1.56
actmtch / DefOpt / cpu / Forward 0.000019142 s 0.00001134841999373748 s 1.69
actmtch / IDefOpt / cpu / Forward 0.000018916 s 0.000011242060008953558 s 1.68
actmtch / JaXPipe / cpu / PreRev 0.000018997 s 0.000010634620048222132 s 1.79
actmtch / JaXPipe / cpu / PostRev 0.000017939 s 0.000010018360007961748 s 1.79
actmtch / JaXPipe / cpu / BothRev 0.000018788 s 0.000011966400006713229 s 1.57
actmtch / Jax / cpu / BothRev 0.000017409999999999998 s 0.000009409679923919612 s 1.85
actmtch / HLOOpt / cpu / PreRev 0.000018745 s 0.000011981659990851768 s 1.56
actmtch / HLOOpt / cpu / PostRev 0.000019318 s 0.000014166439987093329 s 1.36
actmtch / HLOOpt / cpu / BothRev 0.000019082 s 0.000010792540015245322 s 1.77
actmtch / PartOpt / cpu / PreRev 0.000019309 s 0.00001240695998603769 s 1.56
actmtch / PartOpt / cpu / PostRev 0.000017956 s 0.000010077860024466644 s 1.78
actmtch / PartOpt / cpu / BothRev 0.000018762 s 0.000011556859981283196 s 1.62
actmtch / IPartOpt / cpu / PreRev 0.000019232 s 0.000011114280014226096 s 1.73
actmtch / IPartOpt / cpu / PostRev 0.000017444 s 0.000010391919977337238 s 1.68
actmtch / IPartOpt / cpu / BothRev 0.000018917 s 0.000011432819974288576 s 1.65
actmtch / DefOpt / cpu / PreRev 0.000019398 s 0.000010761699995782691 s 1.80
actmtch / DefOpt / cpu / PostRev 0.000019419 s 0.000010955220031974024 s 1.77
actmtch / DefOpt / cpu / BothRev 0.000019147 s 0.00001106313996388053 s 1.73
actmtch / IDefOpt / cpu / PreRev 0.000019312 s 0.000011028760045519448 s 1.75
actmtch / IDefOpt / cpu / PostRev 0.000018995 s 0.000011024740015272982 s 1.72
actmtch / IDefOpt / cpu / BothRev 0.000019446 s 0.000011336180004946071 s 1.72
actmtch / JaXPipe / cpu / Primal 0.00001 s 0.000007719879995420343 s 1.30
actmtch / Jax / cpu / Primal 0.000008999999999999999 s 0.0000071134600057121135 s 1.27
actmtch / HLOOpt / cpu / Primal 0.00001 s 0.000008057620034378488 s 1.24
actmtch / PartOpt / cpu / Primal 0.00001 s 0.000006819159971200861 s 1.47
actmtch / IPartOpt / cpu / Primal 0.00001 s 0.0000071537999974680136 s 1.40
actmtch / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000007619359957971028 s 1.18
actmtch / IDefOpt / cpu / Primal 0.00001 s 0.000007297920028577209 s 1.37
actmtch / JaXPipe / cpu / Forward 0.000013 s 0.000010985560029439512 s 1.18
actmtch / Jax / cpu / Forward 0.000039 s 0.000010219640034847544 s 3.82
actmtch / HLOOpt / cpu / Forward 0.000013 s 0.000011537560048964224 s 1.13
actmtch / PartOpt / cpu / Forward 0.000016 s 0.000011458680010036916 s 1.40
actmtch / IPartOpt / cpu / Forward 0.000013 s 0.000011927800023840972 s 1.09
actmtch / DefOpt / cpu / Forward 0.000014 s 0.00001134841999373748 s 1.23
actmtch / IDefOpt / cpu / Forward 0.000013 s 0.000011242060008953558 s 1.16
actmtch / JaXPipe / cpu / PreRev 0.000014 s 0.000010634620048222132 s 1.32
actmtch / JaXPipe / cpu / PostRev 0.000012 s 0.000010018360007961748 s 1.20
actmtch / JaXPipe / cpu / BothRev 0.000014 s 0.000011966400006713229 s 1.17
actmtch / Jax / cpu / BothRev 0.000013 s 0.000009409679923919612 s 1.38
actmtch / HLOOpt / cpu / PreRev 0.000013 s 0.000011981659990851768 s 1.08
actmtch / HLOOpt / cpu / PostRev 0.000015 s 0.000014166439987093329 s 1.06
actmtch / HLOOpt / cpu / BothRev 0.000013 s 0.000010792540015245322 s 1.20
actmtch / PartOpt / cpu / PreRev 0.000014 s 0.00001240695998603769 s 1.13
actmtch / PartOpt / cpu / PostRev 0.000013 s 0.000010077860024466644 s 1.29
actmtch / PartOpt / cpu / BothRev 0.000013 s 0.000011556859981283196 s 1.12
actmtch / IPartOpt / cpu / PreRev 0.000016 s 0.000011114280014226096 s 1.44
actmtch / IPartOpt / cpu / PostRev 0.000013 s 0.000010391919977337238 s 1.25
actmtch / IPartOpt / cpu / BothRev 0.000015 s 0.000011432819974288576 s 1.31
actmtch / DefOpt / cpu / PreRev 0.000014 s 0.000010761699995782691 s 1.30
actmtch / DefOpt / cpu / PostRev 0.000014 s 0.000010955220031974024 s 1.28
actmtch / DefOpt / cpu / BothRev 0.000015 s 0.00001106313996388053 s 1.36
actmtch / IDefOpt / cpu / PreRev 0.000014 s 0.000011028760045519448 s 1.27
actmtch / IDefOpt / cpu / PostRev 0.000015 s 0.000011024740015272982 s 1.36
actmtch / IDefOpt / cpu / BothRev 0.000014 s 0.000011336180004946071 s 1.23
add_one / JaXPipe / cpu / Primal 0.000008403799993175199 s 0.000006649160031884093 s 1.26
add_one / Jax / cpu / Primal 0.00000906087999283045 s 0.00000667218003400194 s 1.36
add_one / HLOOpt / cpu / Primal 0.000008450039995295811 s 0.000006935620012882282 s 1.22
add_one / PartOpt / cpu / Primal 0.00000824023995846801 s 0.000006502260011984618 s 1.27
add_one / IPartOpt / cpu / Primal 0.00000841367997963971 s 0.000007099779977579601 s 1.19
add_one / DefOpt / cpu / Primal 0.000007933479982966674 s 0.0000067235200094728495 s 1.18
add_one / IDefOpt / cpu / Primal 0.000008585900031903293 s 0.000006646379988524132 s 1.29
add_one / JaXPipe / cpu / Forward 0.000011812540014943806 s 0.0000099950800267834 s 1.18
add_one / Jax / cpu / Forward 0.000011013420016752206 s 0.000010441340036777546 s 1.05
add_one / HLOOpt / cpu / Forward 0.00001188326000374218 s 0.000010399800012237392 s 1.14
add_one / PartOpt / cpu / Forward 0.000011826580030174227 s 0.000010128959966095864 s 1.17
add_one / IPartOpt / cpu / Forward 0.000011621679977906753 s 0.00001030195999192074 s 1.13
add_one / DefOpt / cpu / Forward 0.000011764519967982778 s 0.000010688980009945226 s 1.10
add_one / IDefOpt / cpu / Forward 0.000011333400007060844 s 0.000010097639924424584 s 1.12
add_one / JaXPipe / cpu / PreRev 0.000013223320001998216 s 0.00001188040002489288 s 1.11
add_one / JaXPipe / cpu / PostRev 0.000013148019988875604 s 0.000011429259948272375 s 1.15
add_one / JaXPipe / cpu / BothRev 0.00001422017995537317 s 0.000011384459985492868 s 1.25
add_one / Jax / cpu / BothRev 0.00001336704001914768 s 0.00001157989996499964 s 1.15
add_one / HLOOpt / cpu / PreRev 0.000013345339930310729 s 0.000012574260026667616 s 1.06
add_one / HLOOpt / cpu / PostRev 0.00001707736000753357 s 0.000014400119998754236 s 1.19
add_one / HLOOpt / cpu / BothRev 0.000013472800019371788 s 0.000011879859976033913 s 1.13
add_one / PartOpt / cpu / PreRev 0.000012896759990326246 s 0.000011869480040331835 s 1.09
add_one / PartOpt / cpu / PostRev 0.000013836019998052506 s 0.000011798140021710424 s 1.17
add_one / PartOpt / cpu / BothRev 0.000013308959942150978 s 0.000011920180013476055 s 1.12
add_one / IPartOpt / cpu / PreRev 0.000012882319979325984 s 0.000011350700042385142 s 1.13
add_one / IPartOpt / cpu / PostRev 0.00001311074001932866 s 0.000011941640004806684 s 1.10
add_one / IPartOpt / cpu / BothRev 0.000013197779953770805 s 0.000011431320026531464 s 1.15
add_one / DefOpt / cpu / PreRev 0.000013067360005152297 s 0.000011777200015785638 s 1.11
add_one / DefOpt / cpu / PostRev 0.000013897539993195096 s 0.000011858020025101723 s 1.17
add_one / DefOpt / cpu / BothRev 0.000012976280004295404 s 0.000012032380063828897 s 1.08
add_one / IDefOpt / cpu / PreRev 0.000013440420025290222 s 0.000011383379960534513 s 1.18
add_one / IDefOpt / cpu / PostRev 0.000012934719961776864 s 0.000011826420013676398 s 1.09
add_one / IDefOpt / cpu / BothRev 0.000012776780004060128 s 0.00001131926001107786 s 1.13
add_one / JaXPipe / cuda / Primal 0.000002304 s 0.0000019200000000000003 s 1.20
add_one / Jax / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / HLOOpt / cuda / Primal 0.000002304 s 0.0000019200000000000003 s 1.20
add_one / PartOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / IPartOpt / cuda / Primal 0.000002335 s 0.000001919 s 1.22
add_one / DefOpt / cuda / Primal 0.000002335 s 0.000001919 s 1.22
add_one / IDefOpt / cuda / Primal 0.000002335 s 0.0000019200000000000003 s 1.22
add_one / JaXPipe / cuda / Forward 0.00001072 s 0.000009888 s 1.08
add_one / Jax / cuda / Forward 0.000010752 s 0.000010176 s 1.06
add_one / HLOOpt / cuda / Forward 0.000010624 s 0.00000976 s 1.09
add_one / PartOpt / cuda / Forward 0.000010432 s 0.000010111 s 1.03
add_one / IPartOpt / cuda / Forward 0.0000104 s 0.00001008 s 1.03
add_one / DefOpt / cuda / Forward 0.000010944 s 0.000009536 s 1.15
add_one / IDefOpt / cuda / Forward 0.000010592 s 0.000009887 s 1.07
add_one / JaXPipe / cuda / PreRev 0.000025792 s 0.000024704 s 1.04
add_one / JaXPipe / cuda / PostRev 0.000025312 s 0.000024832 s 1.02
add_one / JaXPipe / cuda / BothRev 0.000025663 s 0.00002496 s 1.03
add_one / Jax / cuda / BothRev 0.000038976 s 0.000024831 s 1.57
add_one / HLOOpt / cuda / PreRev 0.000025856 s 0.000024992 s 1.03
add_one / HLOOpt / cuda / PostRev 0.000025312 s 0.000024448 s 1.04
add_one / HLOOpt / cuda / BothRev 0.000025536 s 0.00002512 s 1.02
add_one / PartOpt / cuda / PreRev 0.00002608 s 0.000025408 s 1.03
add_one / PartOpt / cuda / PostRev 0.000026304 s 0.000025184 s 1.04
add_one / PartOpt / cuda / BothRev 0.000026144 s 0.000025024 s 1.04
add_one / IPartOpt / cuda / PreRev 0.000026176 s 0.00002496 s 1.05
add_one / IPartOpt / cuda / PostRev 0.000025664 s 0.000025728 s 1.00
add_one / IPartOpt / cuda / BothRev 0.000026464000000000003 s 0.000026207 s 1.01
add_one / DefOpt / cuda / PreRev 0.000025728 s 0.00002496 s 1.03
add_one / DefOpt / cuda / PostRev 0.0000264 s 0.000025248 s 1.05
add_one / DefOpt / cuda / BothRev 0.00002624 s 0.000025248 s 1.04
add_one / IDefOpt / cuda / PreRev 0.000026432 s 0.000025472000000000003 s 1.04
add_one / IDefOpt / cuda / PostRev 0.000026079 s 0.000025472000000000003 s 1.02
add_one / IDefOpt / cuda / BothRev 0.000026047 s 0.000025248 s 1.03
add_one / JaXPipe / tpu / Primal 0.00000142565 s 0.0000014248250000000002 s 1.00
add_one / Jax / tpu / Primal 0.00000140415 s 0.0000014076500000000002 s 1.00
add_one / HLOOpt / tpu / Primal 0.0000014227500000000005 s 0.000001430525 s 0.99
add_one / PartOpt / tpu / Primal 0.0000014058 s 0.0000014109 s 1.00
add_one / IPartOpt / tpu / Primal 0.0000014303750000000005 s 0.0000014223250000000002 s 1.01
add_one / DefOpt / tpu / Primal 0.00000140715 s 0.0000014062749999999995 s 1.00
add_one / IDefOpt / tpu / Primal 0.0000014245 s 0.00000142675 s 1.00
add_one / JaXPipe / tpu / Forward 0.0000018467 s 0.00000185735 s 0.99
add_one / Jax / tpu / Forward 0.0000018462 s 0.0000018466 s 1.00
add_one / HLOOpt / tpu / Forward 0.00000184815 s 0.000001853 s 1.00
add_one / PartOpt / tpu / Forward 0.000001840575 s 0.000001835275 s 1.00
add_one / IPartOpt / tpu / Forward 0.000001852 s 0.0000018494 s 1.00
add_one / DefOpt / tpu / Forward 0.000001836225 s 0.000001840175 s 1.00
add_one / IDefOpt / tpu / Forward 0.00000184705 s 0.0000018507 s 1.00
add_one / JaXPipe / tpu / PreRev 0.000002241575 s 0.000002237875 s 1.00
add_one / JaXPipe / tpu / PostRev 0.000002231625 s 0.0000022407250000000005 s 1.00
add_one / JaXPipe / tpu / BothRev 0.00000223045 s 0.0000022356 s 1.00
add_one / Jax / tpu / BothRev 0.000002237025 s 0.00000223985 s 1.00
add_one / HLOOpt / tpu / PreRev 0.000002236875 s 0.0000022408250000000003 s 1.00
add_one / HLOOpt / tpu / PostRev 0.000002238525 s 0.00000224435 s 1.00
add_one / HLOOpt / tpu / BothRev 0.0000022328 s 0.000002243225 s 1.00
add_one / PartOpt / tpu / PreRev 0.000002241575 s 0.0000022381250000000003 s 1.00
add_one / PartOpt / tpu / PostRev 0.0000022365750000000003 s 0.0000022467 s 1.00
add_one / PartOpt / tpu / BothRev 0.00000223555 s 0.00000224655 s 1.00
add_one / IPartOpt / tpu / PreRev 0.0000022298 s 0.000002237025 s 1.00
add_one / IPartOpt / tpu / PostRev 0.000002241925 s 0.000002237775 s 1.00
add_one / IPartOpt / tpu / BothRev 0.00000223295 s 0.000002241125 s 1.00
add_one / DefOpt / tpu / PreRev 0.000002239825 s 0.0000022352 s 1.00
add_one / DefOpt / tpu / PostRev 0.000002236775 s 0.000002239775 s 1.00
add_one / DefOpt / tpu / BothRev 0.0000022346 s 0.000002233425 s 1.00
add_one / IDefOpt / tpu / PreRev 0.0000022409 s 0.000002237725 s 1.00
add_one / IDefOpt / tpu / PostRev 0.0000022399250000000004 s 0.0000022399250000000004 s 1
add_one / IDefOpt / tpu / BothRev 0.000002242 s 0.0000022367 s 1.00
add_one / JaXPipe / cpu / Primal 0.000012947 s 0.000006649160031884093 s 1.95
add_one / Jax / cpu / Primal 0.000013013 s 0.00000667218003400194 s 1.95
add_one / HLOOpt / cpu / Primal 0.000012928 s 0.000006935620012882282 s 1.86
add_one / PartOpt / cpu / Primal 0.000012776 s 0.000006502260011984618 s 1.96
add_one / IPartOpt / cpu / Primal 0.000012775 s 0.000007099779977579601 s 1.80
add_one / DefOpt / cpu / Primal 0.000012808 s 0.0000067235200094728495 s 1.90
add_one / IDefOpt / cpu / Primal 0.000012649 s 0.000006646379988524132 s 1.90
add_one / JaXPipe / cpu / Forward 0.000018013999999999997 s 0.0000099950800267834 s 1.80
add_one / Jax / cpu / Forward 0.000017015 s 0.000010441340036777546 s 1.63
add_one / HLOOpt / cpu / Forward 0.000017295000000000003 s 0.000010399800012237392 s 1.66
add_one / PartOpt / cpu / Forward 0.000017191999999999997 s 0.000010128959966095864 s 1.70
add_one / IPartOpt / cpu / Forward 0.000017233 s 0.00001030195999192074 s 1.67
add_one / DefOpt / cpu / Forward 0.000017247999999999998 s 0.000010688980009945226 s 1.61
add_one / IDefOpt / cpu / Forward 0.000017517999999999997 s 0.000010097639924424584 s 1.73
add_one / JaXPipe / cpu / PreRev 0.000019888 s 0.00001188040002489288 s 1.67
add_one / JaXPipe / cpu / PostRev 0.000019526 s 0.000011429259948272375 s 1.71
add_one / JaXPipe / cpu / BothRev 0.000019299 s 0.000011384459985492868 s 1.70
add_one / Jax / cpu / BothRev 0.000019204 s 0.00001157989996499964 s 1.66
add_one / HLOOpt / cpu / PreRev 0.000019635 s 0.000012574260026667616 s 1.56
add_one / HLOOpt / cpu / PostRev 0.000019806 s 0.000014400119998754236 s 1.38
add_one / HLOOpt / cpu / BothRev 0.000019672 s 0.000011879859976033913 s 1.66
add_one / PartOpt / cpu / PreRev 0.000019314 s 0.000011869480040331835 s 1.63
add_one / PartOpt / cpu / PostRev 0.000019396 s 0.000011798140021710424 s 1.64
add_one / PartOpt / cpu / BothRev 0.000019643 s 0.000011920180013476055 s 1.65
add_one / IPartOpt / cpu / PreRev 0.000019382 s 0.000011350700042385142 s 1.71
add_one / IPartOpt / cpu / PostRev 0.000019777 s 0.000011941640004806684 s 1.66
add_one / IPartOpt / cpu / BothRev 0.000019321 s 0.000011431320026531464 s 1.69
add_one / DefOpt / cpu / PreRev 0.000019369 s 0.000011777200015785638 s 1.64
add_one / DefOpt / cpu / PostRev 0.00001966 s 0.000011858020025101723 s 1.66
add_one / DefOpt / cpu / BothRev 0.00001918 s 0.000012032380063828897 s 1.59
add_one / IDefOpt / cpu / PreRev 0.000019815 s 0.000011383379960534513 s 1.74
add_one / IDefOpt / cpu / PostRev 0.000019284 s 0.000011826420013676398 s 1.63
add_one / IDefOpt / cpu / BothRev 0.000019441 s 0.00001131926001107786 s 1.72
add_one / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000006649160031884093 s 1.35
add_one / Jax / cpu / Primal 0.000008999999999999999 s 0.00000667218003400194 s 1.35
add_one / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000006935620012882282 s 1.30
add_one / PartOpt / cpu / Primal 0.000008 s 0.000006502260011984618 s 1.23
add_one / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000007099779977579601 s 1.27
add_one / DefOpt / cpu / Primal 0.000008 s 0.0000067235200094728495 s 1.19
add_one / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000006646379988524132 s 1.35
add_one / JaXPipe / cpu / Forward 0.000012 s 0.0000099950800267834 s 1.20
add_one / Jax / cpu / Forward 0.000012 s 0.000010441340036777546 s 1.15
add_one / HLOOpt / cpu / Forward 0.000012 s 0.000010399800012237392 s 1.15
add_one / PartOpt / cpu / Forward 0.000038 s 0.000010128959966095864 s 3.75
add_one / IPartOpt / cpu / Forward 0.000012 s 0.00001030195999192074 s 1.16
add_one / DefOpt / cpu / Forward 0.000012 s 0.000010688980009945226 s 1.12
add_one / IDefOpt / cpu / Forward 0.000012 s 0.000010097639924424584 s 1.19
add_one / JaXPipe / cpu / PreRev 0.000014 s 0.00001188040002489288 s 1.18
add_one / JaXPipe / cpu / PostRev 0.000014 s 0.000011429259948272375 s 1.22
add_one / JaXPipe / cpu / BothRev 0.000014 s 0.000011384459985492868 s 1.23
add_one / Jax / cpu / BothRev 0.000015 s 0.00001157989996499964 s 1.30
add_one / HLOOpt / cpu / PreRev 0.000014 s 0.000012574260026667616 s 1.11
add_one / HLOOpt / cpu / PostRev 0.000014 s 0.000014400119998754236 s 0.97
add_one / HLOOpt / cpu / BothRev 0.000015 s 0.000011879859976033913 s 1.26
add_one / PartOpt / cpu / PreRev 0.000014 s 0.000011869480040331835 s 1.18
add_one / PartOpt / cpu / PostRev 0.000014 s 0.000011798140021710424 s 1.19
add_one / PartOpt / cpu / BothRev 0.000014 s 0.000011920180013476055 s 1.17
add_one / IPartOpt / cpu / PreRev 0.000014 s 0.000011350700042385142 s 1.23
add_one / IPartOpt / cpu / PostRev 0.000014 s 0.000011941640004806684 s 1.17
add_one / IPartOpt / cpu / BothRev 0.000014 s 0.000011431320026531464 s 1.22
add_one / DefOpt / cpu / PreRev 0.000014 s 0.000011777200015785638 s 1.19
add_one / DefOpt / cpu / PostRev 0.000014 s 0.000011858020025101723 s 1.18
add_one / DefOpt / cpu / BothRev 0.000014 s 0.000012032380063828897 s 1.16
add_one / IDefOpt / cpu / PreRev 0.000014 s 0.000011383379960534513 s 1.23
add_one / IDefOpt / cpu / PostRev 0.000014 s 0.000011826420013676398 s 1.18
add_one / IDefOpt / cpu / BothRev 0.000014 s 0.00001131926001107786 s 1.24
add_two / JaXPipe / cpu / Primal 0.000007631619992025662 s 0.000007013260010353406 s 1.09
add_two / Jax / cpu / Primal 0.000007712080005148892 s 0.0000067600800048239765 s 1.14
add_two / HLOOpt / cpu / Primal 0.000007978880012160516 s 0.0000069772799906786535 s 1.14
add_two / PartOpt / cpu / Primal 0.000007518399988839519 s 0.0000069943000107741686 s 1.07
add_two / IPartOpt / cpu / Primal 0.000007953059994179056 s 0.000007609880012751091 s 1.05
add_two / DefOpt / cpu / Primal 0.000007385280005109962 s 0.000007053520012050285 s 1.05
add_two / IDefOpt / cpu / Primal 0.000007541720005974639 s 0.000006861800020487863 s 1.10
add_two / JaXPipe / cpu / Forward 0.000011630100007096189 s 0.0000102805200003786 s 1.13
add_two / Jax / cpu / Forward 0.000011845959961647168 s 0.000010426179978821892 s 1.14
add_two / HLOOpt / cpu / Forward 0.000012182960017526056 s 0.000010231839987682178 s 1.19
add_two / PartOpt / cpu / Forward 0.000011165679979967536 s 0.0000102741199589218 s 1.09
add_two / IPartOpt / cpu / Forward 0.000012125000002924935 s 0.000010205079997831487 s 1.19
add_two / DefOpt / cpu / Forward 0.000011148640005558264 s 0.000010678859925974392 s 1.04
add_two / IDefOpt / cpu / Forward 0.000011490799970488297 s 0.000010205359985775431 s 1.13
add_two / JaXPipe / cpu / PreRev 0.00001606709997759026 s 0.000014257080019888237 s 1.13
add_two / JaXPipe / cpu / PostRev 0.00001530267996713519 s 0.000013590599992312492 s 1.13
add_two / JaXPipe / cpu / BothRev 0.000015675220010962222 s 0.000014030000056663991 s 1.12
add_two / Jax / cpu / BothRev 0.000015687579989389632 s 0.000014335080013552217 s 1.09
add_two / HLOOpt / cpu / PreRev 0.000015813419940968744 s 0.000014111140008026268 s 1.12
add_two / HLOOpt / cpu / PostRev 0.000018263219981236034 s 0.00001594786002897308 s 1.15
add_two / HLOOpt / cpu / BothRev 0.00001625558000341698 s 0.000013893400000597466 s 1.17
add_two / PartOpt / cpu / PreRev 0.000016251460001512895 s 0.000013874960013708914 s 1.17
add_two / PartOpt / cpu / PostRev 0.000016124259955176966 s 0.000014527659996019793 s 1.11
add_two / PartOpt / cpu / BothRev 0.000015301220009860117 s 0.000014303039952210384 s 1.07
add_two / IPartOpt / cpu / PreRev 0.000015935680003167364 s 0.00001433147995157924 s 1.11
add_two / IPartOpt / cpu / PostRev 0.00001602164001269557 s 0.0000138057399817626 s 1.16
add_two / IPartOpt / cpu / BothRev 0.000015515299992330256 s 0.0000141949400131125 s 1.09
add_two / DefOpt / cpu / PreRev 0.000015580000008412755 s 0.000014339600002131192 s 1.09
add_two / DefOpt / cpu / PostRev 0.000015888719999566093 s 0.000014293080002971692 s 1.11
add_two / DefOpt / cpu / BothRev 0.00001576480003677716 s 0.000013606799984700047 s 1.16
add_two / IDefOpt / cpu / PreRev 0.00001520690002507763 s 0.000014443540039792425 s 1.05
add_two / IDefOpt / cpu / PostRev 0.000015910120018816088 s 0.000014220200000636396 s 1.12
add_two / IDefOpt / cpu / BothRev 0.00001541524000458594 s 0.000014024060028532404 s 1.10
add_two / JaXPipe / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / Jax / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / HLOOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / PartOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / IPartOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
add_two / DefOpt / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / IDefOpt / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
add_two / JaXPipe / cuda / Forward 0.000010624 s 0.000009792 s 1.08
add_two / Jax / cuda / Forward 0.000010208 s 0.000010016 s 1.02
add_two / HLOOpt / cuda / Forward 0.00001072 s 0.000009888 s 1.08
add_two / PartOpt / cuda / Forward 0.000010623 s 0.0000096 s 1.11
add_two / IPartOpt / cuda / Forward 0.0000112 s 0.000009696 s 1.16
add_two / DefOpt / cuda / Forward 0.000010848 s 0.000009984 s 1.09
add_two / IDefOpt / cuda / Forward 0.000010688 s 0.000009856 s 1.08
add_two / JaXPipe / cuda / PreRev 0.000033984 s 0.000032864 s 1.03
add_two / JaXPipe / cuda / PostRev 0.00003344 s 0.000032511 s 1.03
add_two / JaXPipe / cuda / BothRev 0.000033568 s 0.000032288 s 1.04
add_two / Jax / cuda / BothRev 0.00003424 s 0.000032895 s 1.04
add_two / HLOOpt / cuda / PreRev 0.000033824 s 0.000032672 s 1.04
add_two / HLOOpt / cuda / PostRev 0.000033536000000000006 s 0.000032063 s 1.05
add_two / HLOOpt / cuda / BothRev 0.00003392 s 0.000032255 s 1.05
add_two / PartOpt / cuda / PreRev 0.000033376 s 0.000032736 s 1.02
add_two / PartOpt / cuda / PostRev 0.000033568 s 0.000032864 s 1.02
add_two / PartOpt / cuda / BothRev 0.000033056 s 0.000032672 s 1.01
add_two / IPartOpt / cuda / PreRev 0.00003408 s 0.000032800000000000004 s 1.04
add_two / IPartOpt / cuda / PostRev 0.000033663 s 0.000032288 s 1.04
add_two / IPartOpt / cuda / BothRev 0.000034496 s 0.000031712 s 1.09
add_two / DefOpt / cuda / PreRev 0.000033983 s 0.000031808000000000004 s 1.07
add_two / DefOpt / cuda / PostRev 0.00003344 s 0.000032959 s 1.01
add_two / DefOpt / cuda / BothRev 0.000034975 s 0.000032064 s 1.09
add_two / IDefOpt / cuda / PreRev 0.000033664 s 0.000032224 s 1.04
add_two / IDefOpt / cuda / PostRev 0.000033887 s 0.00003184 s 1.06
add_two / IDefOpt / cuda / BothRev 0.000033632 s 0.000032959 s 1.02
add_two / JaXPipe / tpu / Primal 0.000001432125 s 0.0000014266499999999995 s 1.00
add_two / Jax / tpu / Primal 0.000001472 s 0.0000014741000000000002 s 1.00
add_two / HLOOpt / tpu / Primal 0.000001425325 s 0.000001439375 s 0.99
add_two / PartOpt / tpu / Primal 0.0000014723 s 0.0000014709249999999998 s 1.00
add_two / IPartOpt / tpu / Primal 0.0000014258 s 0.0000014249249999999995 s 1.00
add_two / DefOpt / tpu / Primal 0.0000014771 s 0.0000014799750000000002 s 1.00
add_two / IDefOpt / tpu / Primal 0.0000014295 s 0.0000014266 s 1.00
add_two / JaXPipe / tpu / Forward 0.000001834325 s 0.00000182795 s 1.00
add_two / Jax / tpu / Forward 0.00000181965 s 0.000001822175 s 1.00
add_two / HLOOpt / tpu / Forward 0.00000182425 s 0.0000018268 s 1.00
add_two / PartOpt / tpu / Forward 0.00000182705 s 0.0000018263 s 1.00
add_two / IPartOpt / tpu / Forward 0.000001858625 s 0.000001832025 s 1.01
add_two / DefOpt / tpu / Forward 0.000001826925 s 0.000001826825 s 1.00
add_two / IDefOpt / tpu / Forward 0.000001831175 s 0.000001837675 s 1.00
add_two / JaXPipe / tpu / PreRev 0.0000028444750000000004 s 0.000002837325 s 1.00
add_two / JaXPipe / tpu / PostRev 0.000002751 s 0.0000027462 s 1.00
add_two / JaXPipe / tpu / BothRev 0.0000028278 s 0.00000282855 s 1.00
add_two / Jax / tpu / BothRev 0.00000274045 s 0.000002750125 s 1.00
add_two / HLOOpt / tpu / PreRev 0.0000028447 s 0.000002839525 s 1.00
add_two / HLOOpt / tpu / PostRev 0.00000274785 s 0.000002750675 s 1.00
add_two / HLOOpt / tpu / BothRev 0.000002839625 s 0.000002830775 s 1.00
add_two / PartOpt / tpu / PreRev 0.000002748675 s 0.00000274375 s 1.00
add_two / PartOpt / tpu / PostRev 0.000002841875 s 0.00000284045 s 1.00
add_two / PartOpt / tpu / BothRev 0.00000274715 s 0.00000275895 s 1.00
add_two / IPartOpt / tpu / PreRev 0.000002842725 s 0.0000028283 s 1.01
add_two / IPartOpt / tpu / PostRev 0.0000027547 s 0.0000027574 s 1.00
add_two / IPartOpt / tpu / BothRev 0.0000028353250000000003 s 0.000002827875 s 1.00
add_two / DefOpt / tpu / PreRev 0.0000027504 s 0.000002746825 s 1.00
add_two / DefOpt / tpu / PostRev 0.0000028403 s 0.000002836625 s 1.00
add_two / DefOpt / tpu / BothRev 0.0000027552 s 0.00000275395 s 1.00
add_two / IDefOpt / tpu / PreRev 0.0000028357 s 0.0000028386 s 1.00
add_two / IDefOpt / tpu / PostRev 0.000002745425 s 0.000002749025 s 1.00
add_two / IDefOpt / tpu / BothRev 0.000002829325 s 0.0000028399999999999995 s 1.00
add_two / JaXPipe / cpu / Primal 0.000013156 s 0.000007013260010353406 s 1.88
add_two / Jax / cpu / Primal 0.00001275 s 0.0000067600800048239765 s 1.89
add_two / HLOOpt / cpu / Primal 0.000013023 s 0.0000069772799906786535 s 1.87
add_two / PartOpt / cpu / Primal 0.000013069 s 0.0000069943000107741686 s 1.87
add_two / IPartOpt / cpu / Primal 0.000013101 s 0.000007609880012751091 s 1.72
add_two / DefOpt / cpu / Primal 0.000013143 s 0.000007053520012050285 s 1.86
add_two / IDefOpt / cpu / Primal 0.000013232 s 0.000006861800020487863 s 1.93
add_two / JaXPipe / cpu / Forward 0.000017947999999999998 s 0.0000102805200003786 s 1.75
add_two / Jax / cpu / Forward 0.000018127 s 0.000010426179978821892 s 1.74
add_two / HLOOpt / cpu / Forward 0.000017503999999999997 s 0.000010231839987682178 s 1.71
add_two / PartOpt / cpu / Forward 0.000017999 s 0.0000102741199589218 s 1.75
add_two / IPartOpt / cpu / Forward 0.000018124 s 0.000010205079997831487 s 1.78
add_two / DefOpt / cpu / Forward 0.000017969 s 0.000010678859925974392 s 1.68
add_two / IDefOpt / cpu / Forward 0.000017804 s 0.000010205359985775431 s 1.74
add_two / JaXPipe / cpu / PreRev 0.000023252 s 0.000014257080019888237 s 1.63
add_two / JaXPipe / cpu / PostRev 0.000022889 s 0.000013590599992312492 s 1.68
add_two / JaXPipe / cpu / BothRev 0.000022847 s 0.000014030000056663991 s 1.63
add_two / Jax / cpu / BothRev 0.000022273 s 0.000014335080013552217 s 1.55
add_two / HLOOpt / cpu / PreRev 0.000023478 s 0.000014111140008026268 s 1.66
add_two / HLOOpt / cpu / PostRev 0.000023238 s 0.00001594786002897308 s 1.46
add_two / HLOOpt / cpu / BothRev 0.000023675000000000003 s 0.000013893400000597466 s 1.70
add_two / PartOpt / cpu / PreRev 0.000022725 s 0.000013874960013708914 s 1.64
add_two / PartOpt / cpu / PostRev 0.000022951 s 0.000014527659996019793 s 1.58
add_two / PartOpt / cpu / BothRev 0.000022882 s 0.000014303039952210384 s 1.60
add_two / IPartOpt / cpu / PreRev 0.000023086 s 0.00001433147995157924 s 1.61
add_two / IPartOpt / cpu / PostRev 0.000022936 s 0.0000138057399817626 s 1.66
add_two / IPartOpt / cpu / BothRev 0.000022416 s 0.0000141949400131125 s 1.58
add_two / DefOpt / cpu / PreRev 0.000023195 s 0.000014339600002131192 s 1.62
add_two / DefOpt / cpu / PostRev 0.000023136 s 0.000014293080002971692 s 1.62
add_two / DefOpt / cpu / BothRev 0.000023024 s 0.000013606799984700047 s 1.69
add_two / IDefOpt / cpu / PreRev 0.000022922 s 0.000014443540039792425 s 1.59
add_two / IDefOpt / cpu / PostRev 0.000022915 s 0.000014220200000636396 s 1.61
add_two / IDefOpt / cpu / BothRev 0.000022859 s 0.000014024060028532404 s 1.63
add_two / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000007013260010353406 s 1.28
add_two / Jax / cpu / Primal 0.000008999999999999999 s 0.0000067600800048239765 s 1.33
add_two / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.0000069772799906786535 s 1.29
add_two / PartOpt / cpu / Primal 0.000008999999999999999 s 0.0000069943000107741686 s 1.29
add_two / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000007609880012751091 s 1.18
add_two / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000007053520012050285 s 1.28
add_two / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000006861800020487863 s 1.31
add_two / JaXPipe / cpu / Forward 0.000012 s 0.0000102805200003786 s 1.17
add_two / Jax / cpu / Forward 0.000013 s 0.000010426179978821892 s 1.25
add_two / HLOOpt / cpu / Forward 0.000012 s 0.000010231839987682178 s 1.17
add_two / PartOpt / cpu / Forward 0.000013 s 0.0000102741199589218 s 1.27
add_two / IPartOpt / cpu / Forward 0.000013 s 0.000010205079997831487 s 1.27
add_two / DefOpt / cpu / Forward 0.000041 s 0.000010678859925974392 s 3.84
add_two / IDefOpt / cpu / Forward 0.000012 s 0.000010205359985775431 s 1.18
add_two / JaXPipe / cpu / PreRev 0.000016 s 0.000014257080019888237 s 1.12
add_two / JaXPipe / cpu / PostRev 0.000017 s 0.000013590599992312492 s 1.25
add_two / JaXPipe / cpu / BothRev 0.000017 s 0.000014030000056663991 s 1.21
add_two / Jax / cpu / BothRev 0.000016 s 0.000014335080013552217 s 1.12
add_two / HLOOpt / cpu / PreRev 0.000017 s 0.000014111140008026268 s 1.20
add_two / HLOOpt / cpu / PostRev 0.000017 s 0.00001594786002897308 s 1.07
add_two / HLOOpt / cpu / BothRev 0.000017 s 0.000013893400000597466 s 1.22
add_two / PartOpt / cpu / PreRev 0.000017 s 0.000013874960013708914 s 1.23
add_two / PartOpt / cpu / PostRev 0.000017 s 0.000014527659996019793 s 1.17
add_two / PartOpt / cpu / BothRev 0.00005 s 0.000014303039952210384 s 3.50
add_two / IPartOpt / cpu / PreRev 0.000017999999999999997 s 0.00001433147995157924 s 1.26
add_two / IPartOpt / cpu / PostRev 0.000017 s 0.0000138057399817626 s 1.23
add_two / IPartOpt / cpu / BothRev 0.00005 s 0.0000141949400131125 s 3.52
add_two / DefOpt / cpu / PreRev 0.000017 s 0.000014339600002131192 s 1.19
add_two / DefOpt / cpu / PostRev 0.000016 s 0.000014293080002971692 s 1.12
add_two / DefOpt / cpu / BothRev 0.00005 s 0.000013606799984700047 s 3.67
add_two / IDefOpt / cpu / PreRev 0.000017 s 0.000014443540039792425 s 1.18
add_two / IDefOpt / cpu / PostRev 0.000017 s 0.000014220200000636396 s 1.20
add_two / IDefOpt / cpu / BothRev 0.000017 s 0.000014024060028532404 s 1.21
cache / JaXPipe / cpu / Primal 0.000007095439996191999 s 0.000006809860005887458 s 1.04
cache / Jax / cpu / Primal 0.000008598100030212664 s 0.000006390120006471989 s 1.35
cache / HLOOpt / cpu / Primal 0.000008345740006916458 s 0.00000632239999504236 s 1.32
cache / PartOpt / cpu / Primal 0.000008546159970137523 s 0.000006698860024698661 s 1.28
cache / IPartOpt / cpu / Primal 0.000007524620032199891 s 0.000006153299955258262 s 1.22
cache / DefOpt / cpu / Primal 0.000008025879969864036 s 0.000006390019989339635 s 1.26
cache / IDefOpt / cpu / Primal 0.00000811811999483325 s 0.000006239039976208005 s 1.30
cache / JaXPipe / cpu / Forward 0.00001487894005549606 s 0.000015520319975621534 s 0.96
cache / Jax / cpu / Forward 0.00001544368002214469 s 0.00001545653994071472 s 1.00
cache / HLOOpt / cpu / Forward 0.00001552604000607971 s 0.000015742900031909813 s 0.99
cache / PartOpt / cpu / Forward 0.000016202700016947347 s 0.000015216860001601162 s 1.06
cache / IPartOpt / cpu / Forward 0.000015554039982816903 s 0.00001568589999806136 s 0.99
cache / DefOpt / cpu / Forward 0.00001571639998473984 s 0.000015276899976015557 s 1.03
cache / IDefOpt / cpu / Forward 0.000014477820022875677 s 0.000014973079996707384 s 0.97
cache / JaXPipe / cpu / PreRev 0.000017600900018805988 s 0.00001679751999290602 s 1.05
cache / JaXPipe / cpu / PostRev 0.000021162700031709397 s 0.000022183280007084248 s 0.95
cache / JaXPipe / cpu / BothRev 0.000018108080012098072 s 0.000017591719961274065 s 1.03
cache / Jax / cpu / BothRev 0.000021764260000054493 s 0.000020950920006725937 s 1.04
cache / HLOOpt / cpu / PreRev 0.000017341860020678722 s 0.000016981099997792624 s 1.02
cache / HLOOpt / cpu / PostRev 0.000020691780046036003 s 0.00001919976003591728 s 1.08
cache / HLOOpt / cpu / BothRev 0.00001811120002457756 s 0.000016587700056334142 s 1.09
cache / PartOpt / cpu / PreRev 0.000016649280023557367 s 0.00001695807998657983 s 0.98
cache / PartOpt / cpu / PostRev 0.000022177600030772737 s 0.00002165873998819734 s 1.02
cache / PartOpt / cpu / BothRev 0.000017614460030017654 s 0.000017720599962558482 s 0.99
cache / IPartOpt / cpu / PreRev 0.00001638903992898122 s 0.000017051240010914626 s 0.96
cache / IPartOpt / cpu / PostRev 0.00002075978001812473 s 0.00002263604001200292 s 0.92
cache / IPartOpt / cpu / BothRev 0.000016204380017370568 s 0.000016520119988854276 s 0.98
cache / DefOpt / cpu / PreRev 0.000016383080001105555 s 0.00001660762000028626 s 0.99
cache / DefOpt / cpu / PostRev 0.000017617220000829546 s 0.000017032140003720997 s 1.03
cache / DefOpt / cpu / BothRev 0.000017765560041880235 s 0.00001578975998199894 s 1.13
cache / IDefOpt / cpu / PreRev 0.000017343460021947977 s 0.00001630615994145046 s 1.06
cache / IDefOpt / cpu / PostRev 0.000016342199942300794 s 0.000016397000017605023 s 1.00
cache / IDefOpt / cpu / BothRev 0.00001686584004346514 s 0.000016313999985868577 s 1.03
cache / JaXPipe / cuda / Primal 0.000002335 s 0.000002303 s 1.01
cache / Jax / cuda / Primal 0.000002335 s 0.000002303 s 1.01
cache / HLOOpt / cuda / Primal 0.000002335 s 0.00000224 s 1.04
cache / PartOpt / cuda / Primal 0.000002335 s 0.000002271 s 1.03
cache / IPartOpt / cuda / Primal 0.000002335 s 0.000002303 s 1.01
cache / DefOpt / cuda / Primal 0.000002303 s 0.00000224 s 1.03
cache / IDefOpt / cuda / Primal 0.000002304 s 0.00000224 s 1.03
cache / JaXPipe / cuda / Forward 0.000002336 s 0.000002335 s 1.00
cache / Jax / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / HLOOpt / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / PartOpt / cuda / Forward 0.000002336 s 0.000002336 s 1
cache / IPartOpt / cuda / Forward 0.000002336 s 0.000002335 s 1.00
cache / DefOpt / cuda / Forward 0.000002336 s 0.00000224 s 1.04
cache / IDefOpt / cuda / Forward 0.0000023670000000000004 s 0.000002335 s 1.01
cache / JaXPipe / cuda / PreRev 0.00001088 s 0.000010879 s 1.00
cache / JaXPipe / cuda / PostRev 0.000011135 s 0.000010912 s 1.02
cache / JaXPipe / cuda / BothRev 0.000010976 s 0.000010656 s 1.03
cache / Jax / cuda / BothRev 0.000011552 s 0.000011072 s 1.04
cache / HLOOpt / cuda / PreRev 0.000013631 s 0.000013184 s 1.03
cache / HLOOpt / cuda / PostRev 0.000013568 s 0.000013152 s 1.03
cache / HLOOpt / cuda / BothRev 0.000013632 s 0.000015104 s 0.90
cache / PartOpt / cuda / PreRev 0.00001024 s 0.000010783 s 0.95
cache / PartOpt / cuda / PostRev 0.000010688 s 0.000010816 s 0.99
cache / PartOpt / cuda / BothRev 0.00001104 s 0.000010496 s 1.05
cache / IPartOpt / cuda / PreRev 0.000010752 s 0.000010752 s 1
cache / IPartOpt / cuda / PostRev 0.000010496 s 0.000010752 s 0.98
cache / IPartOpt / cuda / BothRev 0.000010784 s 0.0000104 s 1.04
cache / DefOpt / cuda / PreRev 0.000011168 s 0.000010848 s 1.03
cache / DefOpt / cuda / PostRev 0.000010944 s 0.00001056 s 1.04
cache / DefOpt / cuda / BothRev 0.000011167 s 0.000010624 s 1.05
cache / IDefOpt / cuda / PreRev 0.00001136 s 0.000010753 s 1.06
cache / IDefOpt / cuda / PostRev 0.000011072 s 0.000010529 s 1.05
cache / IDefOpt / cuda / BothRev 0.000011424 s 0.000010719 s 1.07
cache / JaXPipe / tpu / Primal 0.000002467525 s 0.0000024726000000000003 s 1.00
cache / Jax / tpu / Primal 0.00000245585 s 0.000002477175 s 0.99
cache / HLOOpt / tpu / Primal 0.000002464075 s 0.000002455625 s 1.00
cache / PartOpt / tpu / Primal 0.000002477375 s 0.00000246615 s 1.00
cache / IPartOpt / tpu / Primal 0.000002468 s 0.0000024747 s 1.00
cache / DefOpt / tpu / Primal 0.000002465375 s 0.00000246345 s 1.00
cache / IDefOpt / tpu / Primal 0.00000248525 s 0.000002483675 s 1.00
cache / JaXPipe / tpu / Forward 0.00000354695 s 0.0000035745 s 0.99
cache / Jax / tpu / Forward 0.0000035303250000000004 s 0.0000035263250000000005 s 1.00
cache / HLOOpt / tpu / Forward 0.00000356945 s 0.00000355895 s 1.00
cache / PartOpt / tpu / Forward 0.00000354375 s 0.00000353245 s 1.00
cache / IPartOpt / tpu / Forward 0.0000035551 s 0.00000354565 s 1.00
cache / DefOpt / tpu / Forward 0.000003545425 s 0.0000035303250000000004 s 1.00
cache / IDefOpt / tpu / Forward 0.000003561325 s 0.000003537875 s 1.01
cache / JaXPipe / tpu / PreRev 0.00000498515 s 0.0000049689 s 1.00
cache / JaXPipe / tpu / PostRev 0.000004974374999999999 s 0.00000497385 s 1.00
cache / JaXPipe / tpu / BothRev 0.000004997825 s 0.0000049721 s 1.01
cache / Jax / tpu / BothRev 0.0000050031250000000005 s 0.0000049660500000000005 s 1.01
cache / HLOOpt / tpu / PreRev 0.0000039447 s 0.000003946125 s 1.00
cache / HLOOpt / tpu / PostRev 0.000004137649999999999 s 0.000004146599999999999 s 1.00
cache / HLOOpt / tpu / BothRev 0.000003947425 s 0.000003937275 s 1.00
cache / PartOpt / tpu / PreRev 0.000004988375 s 0.000004971 s 1.00
cache / PartOpt / tpu / PostRev 0.0000049827 s 0.000004967849999999999 s 1.00
cache / PartOpt / tpu / BothRev 0.000004970025 s 0.000004968549999999999 s 1.00
cache / IPartOpt / tpu / PreRev 0.0000049888500000000005 s 0.00000495325 s 1.01
cache / IPartOpt / tpu / PostRev 0.00000496745 s 0.000004966425 s 1.00
cache / IPartOpt / tpu / BothRev 0.000004963975 s 0.000004970525 s 1.00
cache / DefOpt / tpu / PreRev 0.0000049679 s 0.00000500465 s 0.99
cache / DefOpt / tpu / PostRev 0.000004964825 s 0.000004964275000000001 s 1.00
cache / DefOpt / tpu / BothRev 0.00000496105 s 0.000004978775 s 1.00
cache / IDefOpt / tpu / PreRev 0.00000496695 s 0.00000497575 s 1.00
cache / IDefOpt / tpu / PostRev 0.000004958125 s 0.000004956625 s 1.00
cache / IDefOpt / tpu / BothRev 0.00000498915 s 0.00000495555 s 1.01
cache / JaXPipe / cpu / Primal 0.000012913000000000002 s 0.000006809860005887458 s 1.90
cache / Jax / cpu / Primal 0.000012292 s 0.000006390120006471989 s 1.92
cache / HLOOpt / cpu / Primal 0.000012626 s 0.00000632239999504236 s 2.00
cache / PartOpt / cpu / Primal 0.000012683 s 0.000006698860024698661 s 1.89
cache / IPartOpt / cpu / Primal 0.000012386 s 0.000006153299955258262 s 2.01
cache / DefOpt / cpu / Primal 0.000012685 s 0.000006390019989339635 s 1.99
cache / IDefOpt / cpu / Primal 0.000012444 s 0.000006239039976208005 s 1.99
cache / JaXPipe / cpu / Forward 0.000016459999999999998 s 0.000015520319975621534 s 1.06
cache / Jax / cpu / Forward 0.00001699 s 0.00001545653994071472 s 1.10
cache / HLOOpt / cpu / Forward 0.000017349 s 0.000015742900031909813 s 1.10
cache / PartOpt / cpu / Forward 0.000016768000000000003 s 0.000015216860001601162 s 1.10
cache / IPartOpt / cpu / Forward 0.00001695 s 0.00001568589999806136 s 1.08
cache / DefOpt / cpu / Forward 0.000016838 s 0.000015276899976015557 s 1.10
cache / IDefOpt / cpu / Forward 0.000016857 s 0.000014973079996707384 s 1.13
cache / JaXPipe / cpu / PreRev 0.000017784 s 0.00001679751999290602 s 1.06
cache / JaXPipe / cpu / PostRev 0.000020236 s 0.000022183280007084248 s 0.91
cache / JaXPipe / cpu / BothRev 0.000018029 s 0.000017591719961274065 s 1.02
cache / Jax / cpu / BothRev 0.00002101 s 0.000020950920006725937 s 1.00
cache / HLOOpt / cpu / PreRev 0.000016768000000000003 s 0.000016981099997792624 s 0.99
cache / HLOOpt / cpu / PostRev 0.000018609 s 0.00001919976003591728 s 0.97
cache / HLOOpt / cpu / BothRev 0.000017735999999999998 s 0.000016587700056334142 s 1.07
cache / PartOpt / cpu / PreRev 0.000017943 s 0.00001695807998657983 s 1.06
cache / PartOpt / cpu / PostRev 0.000020051 s 0.00002165873998819734 s 0.93
cache / PartOpt / cpu / BothRev 0.000017603 s 0.000017720599962558482 s 0.99
cache / IPartOpt / cpu / PreRev 0.000017692 s 0.000017051240010914626 s 1.04
cache / IPartOpt / cpu / PostRev 0.000020682 s 0.00002263604001200292 s 0.91
cache / IPartOpt / cpu / BothRev 0.000017639 s 0.000016520119988854276 s 1.07
cache / DefOpt / cpu / PreRev 0.000017394 s 0.00001660762000028626 s 1.05
cache / DefOpt / cpu / PostRev 0.000018148 s 0.000017032140003720997 s 1.07
cache / DefOpt / cpu / BothRev 0.000016845 s 0.00001578975998199894 s 1.07
cache / IDefOpt / cpu / PreRev 0.00001764 s 0.00001630615994145046 s 1.08
cache / IDefOpt / cpu / PostRev 0.000017756 s 0.000016397000017605023 s 1.08
cache / IDefOpt / cpu / BothRev 0.000017708 s 0.000016313999985868577 s 1.09
cache / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000006809860005887458 s 1.32
cache / Jax / cpu / Primal 0.000008 s 0.000006390120006471989 s 1.25
cache / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.00000632239999504236 s 1.42
cache / PartOpt / cpu / Primal 0.000012 s 0.000006698860024698661 s 1.79
cache / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000006153299955258262 s 1.46
cache / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000006390019989339635 s 1.41
cache / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000006239039976208005 s 1.44
cache / JaXPipe / cpu / Forward 0.000015 s 0.000015520319975621534 s 0.97
cache / Jax / cpu / Forward 0.000033 s 0.00001545653994071472 s 2.14
cache / HLOOpt / cpu / Forward 0.000035999999999999994 s 0.000015742900031909813 s 2.29
cache / PartOpt / cpu / Forward 0.000042 s 0.000015216860001601162 s 2.76
cache / IPartOpt / cpu / Forward 0.000035999999999999994 s 0.00001568589999806136 s 2.30
cache / DefOpt / cpu / Forward 0.000034 s 0.000015276899976015557 s 2.23
cache / IDefOpt / cpu / Forward 0.000034 s 0.000014973079996707384 s 2.27
cache / JaXPipe / cpu / PreRev 0.000011 s 0.00001679751999290602 s 0.65
cache / JaXPipe / cpu / PostRev 0.000013 s 0.000022183280007084248 s 0.59
cache / JaXPipe / cpu / BothRev 0.00001 s 0.000017591719961274065 s 0.57
cache / Jax / cpu / BothRev 0.000011 s 0.000020950920006725937 s 0.53
cache / HLOOpt / cpu / PreRev 0.000011 s 0.000016981099997792624 s 0.65
cache / HLOOpt / cpu / PostRev 0.000035000000000000004 s 0.00001919976003591728 s 1.82
cache / HLOOpt / cpu / BothRev 0.000011 s 0.000016587700056334142 s 0.66
cache / PartOpt / cpu / PreRev 0.000011 s 0.00001695807998657983 s 0.65
cache / PartOpt / cpu / PostRev 0.000035999999999999994 s 0.00002165873998819734 s 1.66
cache / PartOpt / cpu / BothRev 0.000011 s 0.000017720599962558482 s 0.62
cache / IPartOpt / cpu / PreRev 0.000011 s 0.000017051240010914626 s 0.65
cache / IPartOpt / cpu / PostRev 0.000011 s 0.00002263604001200292 s 0.49
cache / IPartOpt / cpu / BothRev 0.000032 s 0.000016520119988854276 s 1.94
cache / DefOpt / cpu / PreRev 0.000011 s 0.00001660762000028626 s 0.66
cache / DefOpt / cpu / PostRev 0.000035000000000000004 s 0.000017032140003720997 s 2.05
cache / DefOpt / cpu / BothRev 0.000011 s 0.00001578975998199894 s 0.70
cache / IDefOpt / cpu / PreRev 0.000035999999999999994 s 0.00001630615994145046 s 2.21
cache / IDefOpt / cpu / PostRev 0.000011 s 0.000016397000017605023 s 0.67
cache / IDefOpt / cpu / BothRev 0.000011 s 0.000016313999985868577 s 0.67
Concat / JaXPipe / cpu / Primal 0.000008874160012055653 s 0.000006851299976915471 s 1.30
Concat / Jax / cpu / Primal 0.000008733879958526814 s 0.000006843399969511665 s 1.28
Concat / HLOOpt / cpu / Primal 0.00000872249999702035 s 0.000006759719999536174 s 1.29
Concat / PartOpt / cpu / Primal 0.000009186299994325964 s 0.000006982800005062017 s 1.32
Concat / IPartOpt / cpu / Primal 0.000008495379997839336 s 0.0000068903399915143386 s 1.23
Concat / DefOpt / cpu / Primal 0.00000866658005179488 s 0.000006584719994862098 s 1.32
Concat / IDefOpt / cpu / Primal 0.000008291359963550348 s 0.000006475580039477791 s 1.28
Concat / JaXPipe / cpu / Forward 0.000011060160004490173 s 0.000010126619999937249 s 1.09
Concat / Jax / cpu / Forward 0.000012268579994270112 s 0.000010056619967144796 s 1.22
Concat / HLOOpt / cpu / Forward 0.000012100100011593895 s 0.000010385059958935016 s 1.17
Concat / PartOpt / cpu / Forward 0.000012425020013324684 s 0.000010396620027677272 s 1.20
Concat / IPartOpt / cpu / Forward 0.000011788019992309274 s 0.000010283319952577585 s 1.15
Concat / DefOpt / cpu / Forward 0.000012295920014366856 s 0.000010074919991893694 s 1.22
Concat / IDefOpt / cpu / Forward 0.000011589220021051004 s 0.000010021980042438372 s 1.16
Concat / JaXPipe / cpu / PreRev 0.000013165480013412889 s 0.000011804619998656562 s 1.12
Concat / JaXPipe / cpu / PostRev 0.000013827519960614156 s 0.000011281219958618749 s 1.23
Concat / JaXPipe / cpu / BothRev 0.000012929140011692652 s 0.000011303440014671653 s 1.14
Concat / Jax / cpu / BothRev 0.000013545180036089732 s 0.00001235251998878084 s 1.10
Concat / HLOOpt / cpu / PreRev 0.00001360665999527555 s 0.000012820660031138686 s 1.06
Concat / HLOOpt / cpu / PostRev 0.00001501086001553631 s 0.000013623140011986834 s 1.10
Concat / HLOOpt / cpu / BothRev 0.0000130709200311685 s 0.000011781780012825038 s 1.11
Concat / PartOpt / cpu / PreRev 0.000012772899981428057 s 0.000011703339987434449 s 1.09
Concat / PartOpt / cpu / PostRev 0.000013680619986189411 s 0.000011818699995274074 s 1.16
Concat / PartOpt / cpu / BothRev 0.000013202659956732533 s 0.000011727220016837236 s 1.13
Concat / IPartOpt / cpu / PreRev 0.00001308595999944373 s 0.000011442100021668013 s 1.14
Concat / IPartOpt / cpu / PostRev 0.00001372446000459604 s 0.000011504599997351762 s 1.19
Concat / IPartOpt / cpu / BothRev 0.000012996299992664718 s 0.000011793180028689676 s 1.10
Concat / DefOpt / cpu / PreRev 0.0000128932600182452 s 0.00001143261996730871 s 1.13
Concat / DefOpt / cpu / PostRev 0.000013128399987181182 s 0.000011749939994842862 s 1.12
Concat / DefOpt / cpu / BothRev 0.000013207680003688438 s 0.000011995499980912428 s 1.10
Concat / IDefOpt / cpu / PreRev 0.000013871960054530064 s 0.000011963659990215092 s 1.16
Concat / IDefOpt / cpu / PostRev 0.000013043799990555271 s 0.000011789560048782732 s 1.11
Concat / IDefOpt / cpu / BothRev 0.000012738919995172182 s 0.000011706300028890835 s 1.09
Concat / JaXPipe / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / Jax / cuda / Primal 0.000002464 s 0.0000019200000000000003 s 1.28
Concat / HLOOpt / cuda / Primal 0.000002432 s 0.0000019200000000000003 s 1.27
Concat / PartOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / IPartOpt / cuda / Primal 0.000002431 s 0.0000019200000000000003 s 1.27
Concat / DefOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / IDefOpt / cuda / Primal 0.000002463 s 0.0000019200000000000003 s 1.28
Concat / JaXPipe / cuda / Forward 0.000010976 s 0.00001008 s 1.09
Concat / Jax / cuda / Forward 0.000010816 s 0.000009728 s 1.11
Concat / HLOOpt / cuda / Forward 0.000010911 s 0.000010015 s 1.09
Concat / PartOpt / cuda / Forward 0.000010848 s 0.000010272 s 1.06
Concat / IPartOpt / cuda / Forward 0.000010816 s 0.000010112 s 1.07
Concat / DefOpt / cuda / Forward 0.000010977 s 0.000009952 s 1.10
Concat / IDefOpt / cuda / Forward 0.00001072 s 0.000009984 s 1.07
Concat / JaXPipe / cuda / PreRev 0.000017024 s 0.00001616 s 1.05
Concat / JaXPipe / cuda / PostRev 0.000016736 s 0.000016896000000000002 s 0.99
Concat / JaXPipe / cuda / BothRev 0.000016672 s 0.000017088 s 0.98
Concat / Jax / cuda / BothRev 0.000017312 s 0.000016545 s 1.05
Concat / HLOOpt / cuda / PreRev 0.000017185 s 0.000016544 s 1.04
Concat / HLOOpt / cuda / PostRev 0.000016993 s 0.000016927999999999998 s 1.00
Concat / HLOOpt / cuda / BothRev 0.000017056 s 0.000016544 s 1.03
Concat / PartOpt / cuda / PreRev 0.000016608 s 0.000016768000000000003 s 0.99
Concat / PartOpt / cuda / PostRev 0.00001728 s 0.000016255999999999998 s 1.06
Concat / PartOpt / cuda / BothRev 0.00001712 s 0.00001664 s 1.03
Concat / IPartOpt / cuda / PreRev 0.00002272 s 0.000016768000000000003 s 1.35
Concat / IPartOpt / cuda / PostRev 0.000017056 s 0.000016607 s 1.03
Concat / IPartOpt / cuda / BothRev 0.000016416 s 0.000016511 s 0.99
Concat / DefOpt / cuda / PreRev 0.000016864 s 0.000017568000000000002 s 0.96
Concat / DefOpt / cuda / PostRev 0.000017088 s 0.000016352 s 1.05
Concat / DefOpt / cuda / BothRev 0.000016929 s 0.000016608 s 1.02
Concat / IDefOpt / cuda / PreRev 0.000017183 s 0.000016736 s 1.03
Concat / IDefOpt / cuda / PostRev 0.000017152 s 0.000016383999999999998 s 1.05
Concat / IDefOpt / cuda / BothRev 0.000017472 s 0.000016736 s 1.04
Concat / JaXPipe / tpu / Primal 0.000001540325 s 0.000001531725 s 1.01
Concat / Jax / tpu / Primal 0.00000152805 s 0.00000152335 s 1.00
Concat / HLOOpt / tpu / Primal 0.000001532625 s 0.0000015340999999999998 s 1.00
Concat / PartOpt / tpu / Primal 0.00000151915 s 0.00000151925 s 1.00
Concat / IPartOpt / tpu / Primal 0.0000015403 s 0.000001532975 s 1.00
Concat / DefOpt / tpu / Primal 0.0000015289249999999995 s 0.000001526225 s 1.00
Concat / IDefOpt / tpu / Primal 0.00000154925 s 0.000001531625 s 1.01
Concat / JaXPipe / tpu / Forward 0.0000015727 s 0.00000158145 s 0.99
Concat / Jax / tpu / Forward 0.0000015586749999999995 s 0.000001553625 s 1.00
Concat / HLOOpt / tpu / Forward 0.0000015763999999999998 s 0.000001575525 s 1.00
Concat / PartOpt / tpu / Forward 0.000001551225 s 0.0000015441250000000002 s 1.00
Concat / IPartOpt / tpu / Forward 0.000001575525 s 0.000001570625 s 1.00
Concat / DefOpt / tpu / Forward 0.0000015600249999999995 s 0.000001545675 s 1.01
Concat / IDefOpt / tpu / Forward 0.000001572925 s 0.0000015783 s 1.00
Concat / JaXPipe / tpu / PreRev 0.000002008475 s 0.00000200565 s 1.00
Concat / JaXPipe / tpu / PostRev 0.0000020913 s 0.0000020843 s 1.00
Concat / JaXPipe / tpu / BothRev 0.0000020069250000000003 s 0.00000200775 s 1.00
Concat / Jax / tpu / BothRev 0.00000206815 s 0.0000020726 s 1.00
Concat / HLOOpt / tpu / PreRev 0.000002007025 s 0.0000020133 s 1.00
Concat / HLOOpt / tpu / PostRev 0.000002067425 s 0.00000206675 s 1.00
Concat / HLOOpt / tpu / BothRev 0.000002006875 s 0.000002020075 s 0.99
Concat / PartOpt / tpu / PreRev 0.000002081525 s 0.00000207395 s 1.00
Concat / PartOpt / tpu / PostRev 0.00000201295 s 0.000002011425 s 1.00
Concat / PartOpt / tpu / BothRev 0.00000208145 s 0.0000020691 s 1.01
Concat / IPartOpt / tpu / PreRev 0.000002012 s 0.0000020123250000000003 s 1.00
Concat / IPartOpt / tpu / PostRev 0.000002068325 s 0.0000020733750000000003 s 1.00
Concat / IPartOpt / tpu / BothRev 0.000002011675 s 0.0000020104 s 1.00
Concat / DefOpt / tpu / PreRev 0.00000208035 s 0.00000207675 s 1.00
Concat / DefOpt / tpu / PostRev 0.000002006325 s 0.000002010025 s 1.00
Concat / DefOpt / tpu / BothRev 0.000002073825 s 0.0000020693500000000004 s 1.00
Concat / IDefOpt / tpu / PreRev 0.00000200355 s 0.0000020050000000000003 s 1.00
Concat / IDefOpt / tpu / PostRev 0.0000020717 s 0.00000207435 s 1.00
Concat / IDefOpt / tpu / BothRev 0.000002007575 s 0.000002012 s 1.00
Concat / JaXPipe / cpu / Primal 0.000012575 s 0.000006851299976915471 s 1.84
Concat / Jax / cpu / Primal 0.000012524 s 0.000006843399969511665 s 1.83
Concat / HLOOpt / cpu / Primal 0.00001241 s 0.000006759719999536174 s 1.84
Concat / PartOpt / cpu / Primal 0.00001268 s 0.000006982800005062017 s 1.82
Concat / IPartOpt / cpu / Primal 0.000012714 s 0.0000068903399915143386 s 1.85
Concat / DefOpt / cpu / Primal 0.000012882 s 0.000006584719994862098 s 1.96
Concat / IDefOpt / cpu / Primal 0.000012932 s 0.000006475580039477791 s 2.00
Concat / JaXPipe / cpu / Forward 0.000017208000000000002 s 0.000010126619999937249 s 1.70
Concat / Jax / cpu / Forward 0.00001746 s 0.000010056619967144796 s 1.74
Concat / HLOOpt / cpu / Forward 0.000017295000000000003 s 0.000010385059958935016 s 1.67
Concat / PartOpt / cpu / Forward 0.00001727 s 0.000010396620027677272 s 1.66
Concat / IPartOpt / cpu / Forward 0.000017385 s 0.000010283319952577585 s 1.69
Concat / DefOpt / cpu / Forward 0.000017144 s 0.000010074919991893694 s 1.70
Concat / IDefOpt / cpu / Forward 0.000017334 s 0.000010021980042438372 s 1.73
Concat / JaXPipe / cpu / PreRev 0.000019947 s 0.000011804619998656562 s 1.69
Concat / JaXPipe / cpu / PostRev 0.000019324 s 0.000011281219958618749 s 1.71
Concat / JaXPipe / cpu / BothRev 0.000019294 s 0.000011303440014671653 s 1.71
Concat / Jax / cpu / BothRev 0.000019749 s 0.00001235251998878084 s 1.60
Concat / HLOOpt / cpu / PreRev 0.000019528 s 0.000012820660031138686 s 1.52
Concat / HLOOpt / cpu / PostRev 0.000019611 s 0.000013623140011986834 s 1.44
Concat / HLOOpt / cpu / BothRev 0.000019216 s 0.000011781780012825038 s 1.63
Concat / PartOpt / cpu / PreRev 0.000019476 s 0.000011703339987434449 s 1.66
Concat / PartOpt / cpu / PostRev 0.000019268 s 0.000011818699995274074 s 1.63
Concat / PartOpt / cpu / BothRev 0.000019089 s 0.000011727220016837236 s 1.63
Concat / IPartOpt / cpu / PreRev 0.000019588000000000003 s 0.000011442100021668013 s 1.71
Concat / IPartOpt / cpu / PostRev 0.000018984 s 0.000011504599997351762 s 1.65
Concat / IPartOpt / cpu / BothRev 0.000018889 s 0.000011793180028689676 s 1.60
Concat / DefOpt / cpu / PreRev 0.000019267 s 0.00001143261996730871 s 1.69
Concat / DefOpt / cpu / PostRev 0.000019643 s 0.000011749939994842862 s 1.67
Concat / DefOpt / cpu / BothRev 0.000019347 s 0.000011995499980912428 s 1.61
Concat / IDefOpt / cpu / PreRev 0.000019618 s 0.000011963659990215092 s 1.64
Concat / IDefOpt / cpu / PostRev 0.000019305 s 0.000011789560048782732 s 1.64
Concat / IDefOpt / cpu / BothRev 0.000019156 s 0.000011706300028890835 s 1.64
Concat / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000006851299976915471 s 1.31
Concat / Jax / cpu / Primal 0.000008 s 0.000006843399969511665 s 1.17
Concat / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000006759719999536174 s 1.33
Concat / PartOpt / cpu / Primal 0.000008999999999999999 s 0.000006982800005062017 s 1.29
Concat / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.0000068903399915143386 s 1.31
Concat / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000006584719994862098 s 1.37
Concat / IDefOpt / cpu / Primal 0.000008 s 0.000006475580039477791 s 1.24
Concat / JaXPipe / cpu / Forward 0.000012 s 0.000010126619999937249 s 1.18
Concat / Jax / cpu / Forward 0.000012 s 0.000010056619967144796 s 1.19
Concat / HLOOpt / cpu / Forward 0.000011 s 0.000010385059958935016 s 1.06
Concat / PartOpt / cpu / Forward 0.000012 s 0.000010396620027677272 s 1.15
Concat / IPartOpt / cpu / Forward 0.000012 s 0.000010283319952577585 s 1.17
Concat / DefOpt / cpu / Forward 0.000012 s 0.000010074919991893694 s 1.19
Concat / IDefOpt / cpu / Forward 0.000012 s 0.000010021980042438372 s 1.20
Concat / JaXPipe / cpu / PreRev 0.000014 s 0.000011804619998656562 s 1.19
Concat / JaXPipe / cpu / PostRev 0.000014 s 0.000011281219958618749 s 1.24
Concat / JaXPipe / cpu / BothRev 0.000014 s 0.000011303440014671653 s 1.24
Concat / Jax / cpu / BothRev 0.000014 s 0.00001235251998878084 s 1.13
Concat / HLOOpt / cpu / PreRev 0.000014 s 0.000012820660031138686 s 1.09
Concat / HLOOpt / cpu / PostRev 0.000014 s 0.000013623140011986834 s 1.03
Concat / HLOOpt / cpu / BothRev 0.000014 s 0.000011781780012825038 s 1.19
Concat / PartOpt / cpu / PreRev 0.000015 s 0.000011703339987434449 s 1.28
Concat / PartOpt / cpu / PostRev 0.000014 s 0.000011818699995274074 s 1.18
Concat / PartOpt / cpu / BothRev 0.000014 s 0.000011727220016837236 s 1.19
Concat / IPartOpt / cpu / PreRev 0.000014 s 0.000011442100021668013 s 1.22
Concat / IPartOpt / cpu / PostRev 0.000014 s 0.000011504599997351762 s 1.22
Concat / IPartOpt / cpu / BothRev 0.000014 s 0.000011793180028689676 s 1.19
Concat / DefOpt / cpu / PreRev 0.000014 s 0.00001143261996730871 s 1.22
Concat / DefOpt / cpu / PostRev 0.000014 s 0.000011749939994842862 s 1.19
Concat / DefOpt / cpu / BothRev 0.000014 s 0.000011995499980912428 s 1.17
Concat / IDefOpt / cpu / PreRev 0.000013 s 0.000011963659990215092 s 1.09
Concat / IDefOpt / cpu / PostRev 0.000013 s 0.000011789560048782732 s 1.10
Concat / IDefOpt / cpu / BothRev 0.000013 s 0.000011706300028890835 s 1.11
const_scatter / JaXPipe / cpu / Primal 0.000008372959991902461 s 0.000006616179980483139 s 1.27
const_scatter / Jax / cpu / Primal 0.000008564680001654779 s 0.000006174779955472332 s 1.39
const_scatter / HLOOpt / cpu / Primal 0.000008354280016646953 s 0.000007332740024139639 s 1.14
const_scatter / PartOpt / cpu / Primal 0.000008646819987916387 s 0.000006533879986818647 s 1.32
const_scatter / IPartOpt / cpu / Primal 0.000007952640007715672 s 0.000006258439980229014 s 1.27
const_scatter / DefOpt / cpu / Primal 0.000008807039985185838 s 0.000006923539986019023 s 1.27
const_scatter / IDefOpt / cpu / Primal 0.00000914931999432156 s 0.000006864939996376051 s 1.33
const_scatter / JaXPipe / cpu / Forward 0.000012510260012277284 s 0.000010637440000209609 s 1.18
const_scatter / Jax / cpu / Forward 0.000011119219998363404 s 0.00000980007997895882 s 1.13
const_scatter / HLOOpt / cpu / Forward 0.000012578919986481195 s 0.000011555359997146295 s 1.09
const_scatter / PartOpt / cpu / Forward 0.000012199400007375516 s 0.00001060187999428308 s 1.15
const_scatter / IPartOpt / cpu / Forward 0.000012448580000636869 s 0.000011051679985030204 s 1.13
const_scatter / DefOpt / cpu / Forward 0.000012918959964736131 s 0.00001096950000828656 s 1.18
const_scatter / IDefOpt / cpu / Forward 0.000012293900017539272 s 0.000010708360005082795 s 1.15
const_scatter / JaXPipe / cpu / PreRev 0.0002895711399742 s 0.0002911451400268 s 0.99
const_scatter / JaXPipe / cpu / PostRev 0.0002857495999978 s 0.0002857538600346 s 1.00
const_scatter / JaXPipe / cpu / BothRev 0.0002850913000656 s 0.0002865084399763 s 1.00
const_scatter / Jax / cpu / BothRev 0.0002835299999605 s 0.0002849694199812 s 0.99
const_scatter / HLOOpt / cpu / PreRev 0.0002914470200084 s 0.0002866276799704 s 1.02
const_scatter / HLOOpt / cpu / PostRev 0.0002880887199899 s 0.0002877500000158 s 1.00
const_scatter / HLOOpt / cpu / BothRev 0.0002853247800248 s 0.0002882349399351 s 0.99
const_scatter / PartOpt / cpu / PreRev 0.0002846488799787 s 0.0002865498400205 s 0.99
const_scatter / PartOpt / cpu / PostRev 0.0002855280800031 s 0.0002845963399522 s 1.00
const_scatter / PartOpt / cpu / BothRev 0.0002952634000121 s 0.0002854337799726 s 1.03
const_scatter / IPartOpt / cpu / PreRev 0.0002883831800227 s 0.0002865885999926 s 1.01
const_scatter / IPartOpt / cpu / PostRev 0.0002839672400205 s 0.0002851015399755 s 1.00
const_scatter / IPartOpt / cpu / BothRev 0.000284118460031 s 0.0002857169600247 s 0.99
const_scatter / DefOpt / cpu / PreRev 0.00028622816003 s 0.0002854945399485 s 1.00
const_scatter / DefOpt / cpu / PostRev 0.000286268380014 s 0.0002849405799679 s 1.00
const_scatter / DefOpt / cpu / BothRev 0.0002848124400406 s 0.0002872083000238 s 0.99
const_scatter / IDefOpt / cpu / PreRev 0.0002851092399578 s 0.0002869064000242 s 0.99
const_scatter / IDefOpt / cpu / PostRev 0.0002864153400514 s 0.0002876335599921 s 1.00
const_scatter / IDefOpt / cpu / BothRev 0.0002887705999819 s 0.0003068942400568 s 0.94
const_scatter / JaXPipe / cuda / Primal 0.000002463 s 0.000001888 s 1.30
const_scatter / Jax / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / HLOOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / PartOpt / cuda / Primal 0.000002463 s 0.000001888 s 1.30
const_scatter / IPartOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / DefOpt / cuda / Primal 0.000002464 s 0.000001887 s 1.31
const_scatter / IDefOpt / cuda / Primal 0.000002463 s 0.000001887 s 1.31
const_scatter / JaXPipe / cuda / Forward 0.000010816 s 0.000010432 s 1.04
const_scatter / Jax / cuda / Forward 0.00001088 s 0.000009632 s 1.13
const_scatter / HLOOpt / cuda / Forward 0.000010527 s 0.000010047 s 1.05
const_scatter / PartOpt / cuda / Forward 0.00001088 s 0.00000992 s 1.10
const_scatter / IPartOpt / cuda / Forward 0.000011136 s 0.000009887 s 1.13
const_scatter / DefOpt / cuda / Forward 0.000010688 s 0.000009633 s 1.11
const_scatter / IDefOpt / cuda / Forward 0.000010944 s 0.000009984 s 1.10
const_scatter / JaXPipe / cuda / PreRev 0.00001696 s 0.000016448000000000002 s 1.03
const_scatter / JaXPipe / cuda / PostRev 0.000017856 s 0.00001664 s 1.07
const_scatter / JaXPipe / cuda / BothRev 0.000017567 s 0.000016735 s 1.05
const_scatter / Jax / cuda / BothRev 0.000017632 s 0.000016608 s 1.06
const_scatter / HLOOpt / cuda / PreRev 0.000017312 s 0.000016096 s 1.08
const_scatter / HLOOpt / cuda / PostRev 0.000016672 s 0.000016352 s 1.02
const_scatter / HLOOpt / cuda / BothRev 0.000017247999999999998 s 0.000016416 s 1.05
const_scatter / PartOpt / cuda / PreRev 0.000017568000000000002 s 0.000016608 s 1.06
const_scatter / PartOpt / cuda / PostRev 0.000016927999999999998 s 0.000015776 s 1.07
const_scatter / PartOpt / cuda / BothRev 0.000016927999999999998 s 0.000016576000000000002 s 1.02
const_scatter / IPartOpt / cuda / PreRev 0.000017536 s 0.000016608 s 1.06
const_scatter / IPartOpt / cuda / PostRev 0.000016639 s 0.000016416 s 1.01
const_scatter / IPartOpt / cuda / BothRev 0.00001728 s 0.000016032 s 1.08
const_scatter / DefOpt / cuda / PreRev 0.000017056 s 0.000016383999999999998 s 1.04
const_scatter / DefOpt / cuda / PostRev 0.000017247999999999998 s 0.000016255999999999998 s 1.06
const_scatter / DefOpt / cuda / BothRev 0.000017216 s 0.00001616 s 1.07
const_scatter / IDefOpt / cuda / PreRev 0.000017375999999999998 s 0.00001648 s 1.05
const_scatter / IDefOpt / cuda / PostRev 0.00001728 s 0.000016511 s 1.05
const_scatter / IDefOpt / cuda / BothRev 0.000017088 s 0.00001696 s 1.01
const_scatter / JaXPipe / tpu / Primal 0.000003799025 s 0.00000379415 s 1.00
const_scatter / Jax / tpu / Primal 0.000003831200000000001 s 0.000003819625 s 1.00
const_scatter / HLOOpt / tpu / Primal 0.0000037877 s 0.0000037898 s 1.00
const_scatter / PartOpt / tpu / Primal 0.00000381035 s 0.0000038214 s 1.00
const_scatter / IPartOpt / tpu / Primal 0.0000038028 s 0.00000381755 s 1.00
const_scatter / DefOpt / tpu / Primal 0.000003797625 s 0.0000037987 s 1.00
const_scatter / IDefOpt / tpu / Primal 0.00000377555 s 0.0000037770749999999993 s 1.00
const_scatter / JaXPipe / tpu / Forward 0.000006443075 s 0.0000064722250000000005 s 1.00
const_scatter / Jax / tpu / Forward 0.0000065065 s 0.000006485825 s 1.00
const_scatter / HLOOpt / tpu / Forward 0.000006452599999999999 s 0.0000064555 s 1.00
const_scatter / PartOpt / tpu / Forward 0.0000065162500000000005 s 0.000006474775 s 1.01
const_scatter / IPartOpt / tpu / Forward 0.000006443325 s 0.000006456 s 1.00
const_scatter / DefOpt / tpu / Forward 0.0000065081 s 0.0000064712 s 1.01
const_scatter / IDefOpt / tpu / Forward 0.000006463600000000001 s 0.000006460025 s 1.00
const_scatter / JaXPipe / tpu / PreRev 0.000006590625 s 0.000006612275 s 1.00
const_scatter / JaXPipe / tpu / PostRev 0.00000663245 s 0.000006616525 s 1.00
const_scatter / JaXPipe / tpu / BothRev 0.000006616975 s 0.000006589225 s 1.00
const_scatter / Jax / tpu / BothRev 0.000006629125000000001 s 0.000006626450000000001 s 1.00
const_scatter / HLOOpt / tpu / PreRev 0.000006599150000000001 s 0.00000661935 s 1.00
const_scatter / HLOOpt / tpu / PostRev 0.000006619125 s 0.000006612125 s 1.00
const_scatter / HLOOpt / tpu / BothRev 0.000006595475 s 0.0000065923750000000006 s 1.00
const_scatter / PartOpt / tpu / PreRev 0.00000661285 s 0.000006619375 s 1.00
const_scatter / PartOpt / tpu / PostRev 0.0000065858000000000005 s 0.00000661975 s 0.99
const_scatter / PartOpt / tpu / BothRev 0.00000662395 s 0.0000066182250000000005 s 1.00
const_scatter / IPartOpt / tpu / PreRev 0.000006588425 s 0.000006615250000000001 s 1.00
const_scatter / IPartOpt / tpu / PostRev 0.00000662685 s 0.000006624875 s 1.00
const_scatter / IPartOpt / tpu / BothRev 0.000006572575 s 0.000006599725 s 1.00
const_scatter / DefOpt / tpu / PreRev 0.00000662385 s 0.000006621525 s 1.00
const_scatter / DefOpt / tpu / PostRev 0.000006614225 s 0.000006596475000000001 s 1.00
const_scatter / DefOpt / tpu / BothRev 0.000006608025 s 0.000006611525 s 1.00
const_scatter / IDefOpt / tpu / PreRev 0.000006586775 s 0.000006617875 s 1.00
const_scatter / IDefOpt / tpu / PostRev 0.00000662155 s 0.0000066075 s 1.00
const_scatter / IDefOpt / tpu / BothRev 0.000006607025 s 0.0000065911 s 1.00
const_scatter / JaXPipe / cpu / Primal 0.000012799 s 0.000006616179980483139 s 1.93
const_scatter / Jax / cpu / Primal 0.00001242 s 0.000006174779955472332 s 2.01
const_scatter / HLOOpt / cpu / Primal 0.000013109 s 0.000007332740024139639 s 1.79
const_scatter / PartOpt / cpu / Primal 0.000012612 s 0.000006533879986818647 s 1.93
const_scatter / IPartOpt / cpu / Primal 0.000012615 s 0.000006258439980229014 s 2.02
const_scatter / DefOpt / cpu / Primal 0.000013325 s 0.000006923539986019023 s 1.92
const_scatter / IDefOpt / cpu / Primal 0.00001332 s 0.000006864939996376051 s 1.94
const_scatter / JaXPipe / cpu / Forward 0.00001781 s 0.000010637440000209609 s 1.67
const_scatter / Jax / cpu / Forward 0.000016847 s 0.00000980007997895882 s 1.72
const_scatter / HLOOpt / cpu / Forward 0.000017682 s 0.000011555359997146295 s 1.53
const_scatter / PartOpt / cpu / Forward 0.000017982 s 0.00001060187999428308 s 1.70
const_scatter / IPartOpt / cpu / Forward 0.000017726999999999998 s 0.000011051679985030204 s 1.60
const_scatter / DefOpt / cpu / Forward 0.000017645 s 0.00001096950000828656 s 1.61
const_scatter / IDefOpt / cpu / Forward 0.000017682 s 0.000010708360005082795 s 1.65
const_scatter / JaXPipe / cpu / PreRev 0.000490624 s 0.0002911451400268 s 1.69
const_scatter / JaXPipe / cpu / PostRev 0.00050117 s 0.0002857538600346 s 1.75
const_scatter / JaXPipe / cpu / BothRev 0.000499786 s 0.0002865084399763 s 1.74
const_scatter / Jax / cpu / BothRev 0.000512405 s 0.0002849694199812 s 1.80
const_scatter / HLOOpt / cpu / PreRev 0.000489356 s 0.0002866276799704 s 1.71
const_scatter / HLOOpt / cpu / PostRev 0.000486178 s 0.0002877500000158 s 1.69
const_scatter / HLOOpt / cpu / BothRev 0.0005072729999999 s 0.0002882349399351 s 1.76
const_scatter / PartOpt / cpu / PreRev 0.00051262 s 0.0002865498400205 s 1.79
const_scatter / PartOpt / cpu / PostRev 0.000489988 s 0.0002845963399522 s 1.72
const_scatter / PartOpt / cpu / BothRev 0.0005076859999999 s 0.0002854337799726 s 1.78
const_scatter / IPartOpt / cpu / PreRev 0.000516899 s 0.0002865885999926 s 1.80
const_scatter / IPartOpt / cpu / PostRev 0.000499993 s 0.0002851015399755 s 1.75
const_scatter / IPartOpt / cpu / BothRev 0.000526271 s 0.0002857169600247 s 1.84
const_scatter / DefOpt / cpu / PreRev 0.00052206 s 0.0002854945399485 s 1.83
const_scatter / DefOpt / cpu / PostRev 0.0005194689999999 s 0.0002849405799679 s 1.82
const_scatter / DefOpt / cpu / BothRev 0.000489779 s 0.0002872083000238 s 1.71
const_scatter / IDefOpt / cpu / PreRev 0.0005099709999999 s 0.0002869064000242 s 1.78
const_scatter / IDefOpt / cpu / PostRev 0.000495707 s 0.0002876335599921 s 1.72
const_scatter / IDefOpt / cpu / BothRev 0.000513468 s 0.0003068942400568 s 1.67
const_scatter / JaXPipe / cpu / Primal 0.000008999999999999999 s 0.000006616179980483139 s 1.36
const_scatter / Jax / cpu / Primal 0.000008 s 0.000006174779955472332 s 1.30
const_scatter / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000007332740024139639 s 1.23
const_scatter / PartOpt / cpu / Primal 0.000008999999999999999 s 0.000006533879986818647 s 1.38
const_scatter / IPartOpt / cpu / Primal 0.000008 s 0.000006258439980229014 s 1.28
const_scatter / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000006923539986019023 s 1.30
const_scatter / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000006864939996376051 s 1.31
const_scatter / JaXPipe / cpu / Forward 0.000013 s 0.000010637440000209609 s 1.22
const_scatter / Jax / cpu / Forward 0.000012 s 0.00000980007997895882 s 1.22
const_scatter / HLOOpt / cpu / Forward 0.000013 s 0.000011555359997146295 s 1.13
const_scatter / PartOpt / cpu / Forward 0.000013 s 0.00001060187999428308 s 1.23
const_scatter / IPartOpt / cpu / Forward 0.000013 s 0.000011051679985030204 s 1.18
const_scatter / DefOpt / cpu / Forward 0.000012 s 0.00001096950000828656 s 1.09
const_scatter / IDefOpt / cpu / Forward 0.000013 s 0.000010708360005082795 s 1.21
const_scatter / JaXPipe / cpu / PreRev 0.00034 s 0.0002911451400268 s 1.17
const_scatter / JaXPipe / cpu / PostRev 0.000341 s 0.0002857538600346 s 1.19
const_scatter / JaXPipe / cpu / BothRev 0.000403 s 0.0002865084399763 s 1.41
const_scatter / Jax / cpu / BothRev 0.000375 s 0.0002849694199812 s 1.32
const_scatter / HLOOpt / cpu / PreRev 0.000332 s 0.0002866276799704 s 1.16
const_scatter / HLOOpt / cpu / PostRev 0.000361 s 0.0002877500000158 s 1.25
const_scatter / HLOOpt / cpu / BothRev 0.0003439999999999 s 0.0002882349399351 s 1.19
const_scatter / PartOpt / cpu / PreRev 0.000666 s 0.0002865498400205 s 2.32
const_scatter / PartOpt / cpu / PostRev 0.0003529999999999 s 0.0002845963399522 s 1.24
const_scatter / PartOpt / cpu / BothRev 0.000331 s 0.0002854337799726 s 1.16
const_scatter / IPartOpt / cpu / PreRev 0.00034 s 0.0002865885999926 s 1.19
const_scatter / IPartOpt / cpu / PostRev 0.000431 s 0.0002851015399755 s 1.51
const_scatter / IPartOpt / cpu / BothRev 0.000336 s 0.0002857169600247 s 1.18
const_scatter / DefOpt / cpu / PreRev 0.000379 s 0.0002854945399485 s 1.33
const_scatter / DefOpt / cpu / PostRev 0.000363 s 0.0002849405799679 s 1.27
const_scatter / DefOpt / cpu / BothRev 0.000401 s 0.0002872083000238 s 1.40
const_scatter / IDefOpt / cpu / PreRev 0.0003489999999999 s 0.0002869064000242 s 1.22
const_scatter / IDefOpt / cpu / PostRev 0.000544 s 0.0002876335599921 s 1.89
const_scatter / IDefOpt / cpu / BothRev 0.00034 s 0.0003068942400568 s 1.11
GenDot / JaXPipe / cpu / Primal 0.000009219240018865092 s 0.000007380500019280589 s 1.25
GenDot / Jax / cpu / Primal 0.000008859459976520156 s 0.000006958400008443277 s 1.27
GenDot / HLOOpt / cpu / Primal 0.000009712519977256308 s 0.0000074693799797387325 s 1.30
GenDot / PartOpt / cpu / Primal 0.000009494499954598724 s 0.000007400719969155034 s 1.28
GenDot / IPartOpt / cpu / Primal 0.000009350339978482224 s 0.000007423959987136186 s 1.26
GenDot / DefOpt / cpu / Primal 0.000008783160028542625 s 0.000007572139984404202 s 1.16
GenDot / IDefOpt / cpu / Primal 0.000008654779994685669 s 0.000007332039976972737 s 1.18
GenDot / JaXPipe / cpu / Forward 0.00001279362001696427 s 0.000010938879950117552 s 1.17
GenDot / Jax / cpu / Forward 0.000011891279991687042 s 0.000010100020008394494 s 1.18
GenDot / HLOOpt / cpu / Forward 0.000012783059992216294 s 0.000011385520001567785 s 1.12
GenDot / PartOpt / cpu / Forward 0.000012149980020694784 s 0.00001056173997312726 s 1.15
GenDot / IPartOpt / cpu / Forward 0.000013159920008547488 s 0.000011640999982773792 s 1.13
GenDot / DefOpt / cpu / Forward 0.000012819660050809032 s 0.000010874920017158729 s 1.18
GenDot / IDefOpt / cpu / Forward 0.000012398180006130131 s 0.0000112198199985869 s 1.11
GenDot / JaXPipe / cpu / PreRev 0.00001253266002095188 s 0.000010976620023939176 s 1.14
GenDot / JaXPipe / cpu / PostRev 0.000011565680024432368 s 0.000010879080018639798 s 1.06
GenDot / JaXPipe / cpu / BothRev 0.000012646920013139606 s 0.000011393520044293836 s 1.11
GenDot / Jax / cpu / BothRev 0.000012303839994274311 s 0.000011286379994999153 s 1.09
GenDot / HLOOpt / cpu / PreRev 0.000012952359957125735 s 0.000011677359971145052 s 1.11
GenDot / HLOOpt / cpu / PostRev 0.000015062220018080553 s 0.000013176040029065915 s 1.14
GenDot / HLOOpt / cpu / BothRev 0.000012525360043582624 s 0.000011039759983759723 s 1.13
GenDot / PartOpt / cpu / PreRev 0.000013434740003503977 s 0.000011172839967912297 s 1.20
GenDot / PartOpt / cpu / PostRev 0.000011857460021929 s 0.000010638620024110423 s 1.11
GenDot / PartOpt / cpu / BothRev 0.000012770600014846424 s 0.000011430280028434937 s 1.12
GenDot / IPartOpt / cpu / PreRev 0.000012979520033695734 s 0.00001093669999136182 s 1.19
GenDot / IPartOpt / cpu / PostRev 0.000011540539990164687 s 0.000010266539984513656 s 1.12
GenDot / IPartOpt / cpu / BothRev 0.000012690780004049884 s 0.00001097419997677207 s 1.16
GenDot / DefOpt / cpu / PreRev 0.000011962740036324248 s 0.000010888840006373356 s 1.10
GenDot / DefOpt / cpu / PostRev 0.000013384819985731156 s 0.000011311660045976168 s 1.18
GenDot / DefOpt / cpu / BothRev 0.000013045339992459048 s 0.000011086939975939458 s 1.18
GenDot / IDefOpt / cpu / PreRev 0.000012231780010552029 s 0.000010983839974869624 s 1.11
GenDot / IDefOpt / cpu / PostRev 0.000013365540016820887 s 0.000011606319985730806 s 1.15
GenDot / IDefOpt / cpu / BothRev 0.000012371079983495291 s 0.000011533079996297602 s 1.07
GenDot / JaXPipe / cuda / Primal 0.000002528 s 0.000002016 s 1.25
GenDot / Jax / cuda / Primal 0.000002528 s 0.000002015 s 1.25
GenDot / HLOOpt / cuda / Primal 0.000002527 s 0.000001984 s 1.27
GenDot / PartOpt / cuda / Primal 0.00000256 s 0.000002015 s 1.27
GenDot / IPartOpt / cuda / Primal 0.00000256 s 0.000002015 s 1.27
GenDot / DefOpt / cuda / Primal 0.000002528 s 0.000001984 s 1.27
GenDot / IDefOpt / cuda / Primal 0.000002527 s 0.000001983 s 1.27
GenDot / JaXPipe / cuda / Forward 0.000011008 s 0.000009824 s 1.12
GenDot / Jax / cuda / Forward 0.000010496 s 0.000009856 s 1.06
GenDot / HLOOpt / cuda / Forward 0.00001104 s 0.00000992 s 1.11
GenDot / PartOpt / cuda / Forward 0.000010624 s 0.00000992 s 1.07
GenDot / IPartOpt / cuda / Forward 0.000010656 s 0.000010143 s 1.05
GenDot / DefOpt / cuda / Forward 0.000010624 s 0.000010144 s 1.05
GenDot / IDefOpt / cuda / Forward 0.0000112 s 0.000010048 s 1.11
GenDot / JaXPipe / cuda / PreRev 0.00001072 s 0.00000944 s 1.14
GenDot / JaXPipe / cuda / PostRev 0.000010624 s 0.00000976 s 1.09
GenDot / JaXPipe / cuda / BothRev 0.000010624 s 0.000009472 s 1.12
GenDot / Jax / cuda / BothRev 0.000010912 s 0.000009823 s 1.11
GenDot / HLOOpt / cuda / PreRev 0.000010816 s 0.000010016 s 1.08
GenDot / HLOOpt / cuda / PostRev 0.00001072 s 0.00000992 s 1.08
GenDot / HLOOpt / cuda / BothRev 0.000010656 s 0.000009439 s 1.13
GenDot / PartOpt / cuda / PreRev 0.00001072 s 0.000009985 s 1.07
GenDot / PartOpt / cuda / PostRev 0.000011168 s 0.00001056 s 1.06
GenDot / PartOpt / cuda / BothRev 0.000011232 s 0.000010271 s 1.09
GenDot / IPartOpt / cuda / PreRev 0.000013536 s 0.000010111 s 1.34
GenDot / IPartOpt / cuda / PostRev 0.000010816 s 0.000010367 s 1.04
GenDot / IPartOpt / cuda / BothRev 0.00001072 s 0.000010176 s 1.05
GenDot / DefOpt / cuda / PreRev 0.000011392 s 0.000009984 s 1.14
GenDot / DefOpt / cuda / PostRev 0.00001072 s 0.00000928 s 1.16
GenDot / DefOpt / cuda / BothRev 0.000010688 s 0.000009952 s 1.07
GenDot / IDefOpt / cuda / PreRev 0.00001072 s 0.000009952 s 1.08
GenDot / IDefOpt / cuda / PostRev 0.000009952 s 0.000009951 s 1.00
GenDot / IDefOpt / cuda / BothRev 0.000011072 s 0.000009504 s 1.16
GenDot / JaXPipe / tpu / Primal 9.301e-7 s 9.305e-7 s 1.00
GenDot / Jax / tpu / Primal 9.26e-7 s 9.2595e-7 s 1.00
GenDot / HLOOpt / tpu / Primal 0.00000156665 s 0.0000015745999999999995 s 0.99
GenDot / PartOpt / tpu / Primal 9.26475e-7 s 9.25675e-7 s 1.00
GenDot / IPartOpt / tpu / Primal 9.30125e-7 s 9.3015e-7 s 1.00
GenDot / DefOpt / tpu / Primal 0.0000014846999999999995 s 0.000001488725 s 1.00
GenDot / IDefOpt / tpu / Primal 0.000001574375 s 0.000001568175 s 1.00
GenDot / JaXPipe / tpu / Forward 0.0000031612249999999995 s 0.000003164275 s 1.00
GenDot / Jax / tpu / Forward 0.000002315025 s 0.0000023144250000000003 s 1.00
GenDot / HLOOpt / tpu / Forward 0.000003106975 s 0.0000031074750000000003 s 1.00
GenDot / PartOpt / tpu / Forward 0.00000322565 s 0.000003216575 s 1.00
GenDot / IPartOpt / tpu / Forward 0.000003107 s 0.0000031049 s 1.00
GenDot / DefOpt / tpu / Forward 0.000003205975 s 0.0000032102750000000004 s 1.00
GenDot / IDefOpt / tpu / Forward 0.0000031149 s 0.0000031392 s 0.99
GenDot / JaXPipe / tpu / PreRev 0.0000029536000000000004 s 0.0000029602 s 1.00
GenDot / JaXPipe / tpu / PostRev 0.0000024008 s 0.000002405125 s 1.00
GenDot / JaXPipe / tpu / BothRev 0.000002951075 s 0.0000029486 s 1.00
GenDot / Jax / tpu / BothRev 0.00000239945 s 0.000002403725 s 1.00
GenDot / HLOOpt / tpu / PreRev 0.000002951875 s 0.0000029516 s 1.00
GenDot / HLOOpt / tpu / PostRev 0.000002924875 s 0.00000293085 s 1.00
GenDot / HLOOpt / tpu / BothRev 0.00000295155 s 0.00000295045 s 1.00
GenDot / PartOpt / tpu / PreRev 0.000002929875 s 0.00000292355 s 1.00
GenDot / PartOpt / tpu / PostRev 0.0000023964 s 0.00000239505 s 1.00
GenDot / PartOpt / tpu / BothRev 0.000002950175 s 0.000002926725 s 1.01
GenDot / IPartOpt / tpu / PreRev 0.000002954475 s 0.0000029505 s 1.00
GenDot / IPartOpt / tpu / PostRev 0.0000024123 s 0.000002405125 s 1.00
GenDot / IPartOpt / tpu / BothRev 0.0000029494749999999994 s 0.00000295415 s 1.00
GenDot / DefOpt / tpu / PreRev 0.0000029251 s 0.0000029244 s 1.00
GenDot / DefOpt / tpu / PostRev 0.000002955975 s 0.0000029586 s 1.00
GenDot / DefOpt / tpu / BothRev 0.0000029208 s 0.000002924575 s 1.00
GenDot / IDefOpt / tpu / PreRev 0.000002950875 s 0.00000295695 s 1.00
GenDot / IDefOpt / tpu / PostRev 0.000002933425 s 0.0000029288 s 1.00
GenDot / IDefOpt / tpu / BothRev 0.00000294945 s 0.0000029566250000000003 s 1.00
GenDot / JaXPipe / cpu / Primal 0.000014799 s 0.000007380500019280589 s 2.01
GenDot / Jax / cpu / Primal 0.000014604 s 0.000006958400008443277 s 2.10
GenDot / HLOOpt / cpu / Primal 0.00001397 s 0.0000074693799797387325 s 1.87
GenDot / PartOpt / cpu / Primal 0.000014464 s 0.000007400719969155034 s 1.95
GenDot / IPartOpt / cpu / Primal 0.000014737 s 0.000007423959987136186 s 1.99
GenDot / DefOpt / cpu / Primal 0.000013879 s 0.000007572139984404202 s 1.83
GenDot / IDefOpt / cpu / Primal 0.000014373 s 0.000007332039976972737 s 1.96
GenDot / JaXPipe / cpu / Forward 0.000019419 s 0.000010938879950117552 s 1.78
GenDot / Jax / cpu / Forward 0.000019497 s 0.000010100020008394494 s 1.93
GenDot / HLOOpt / cpu / Forward 0.000018832 s 0.000011385520001567785 s 1.65
GenDot / PartOpt / cpu / Forward 0.000018758 s 0.00001056173997312726 s 1.78
GenDot / IPartOpt / cpu / Forward 0.000019195 s 0.000011640999982773792 s 1.65
GenDot / DefOpt / cpu / Forward 0.00001897 s 0.000010874920017158729 s 1.74
GenDot / IDefOpt / cpu / Forward 0.000018926 s 0.0000112198199985869 s 1.69
GenDot / JaXPipe / cpu / PreRev 0.000019625 s 0.000010976620023939176 s 1.79
GenDot / JaXPipe / cpu / PostRev 0.000020008 s 0.000010879080018639798 s 1.84
GenDot / JaXPipe / cpu / BothRev 0.000019219 s 0.000011393520044293836 s 1.69
GenDot / Jax / cpu / BothRev 0.000020787 s 0.000011286379994999153 s 1.84
GenDot / HLOOpt / cpu / PreRev 0.000019381 s 0.000011677359971145052 s 1.66
GenDot / HLOOpt / cpu / PostRev 0.000019542 s 0.000013176040029065915 s 1.48
GenDot / HLOOpt / cpu / BothRev 0.0000192 s 0.000011039759983759723 s 1.74
GenDot / PartOpt / cpu / PreRev 0.000019128 s 0.000011172839967912297 s 1.71
GenDot / PartOpt / cpu / PostRev 0.000019637 s 0.000010638620024110423 s 1.85
GenDot / PartOpt / cpu / BothRev 0.000019225 s 0.000011430280028434937 s 1.68
GenDot / IPartOpt / cpu / PreRev 0.000019025 s 0.00001093669999136182 s 1.74
GenDot / IPartOpt / cpu / PostRev 0.000020951 s 0.000010266539984513656 s 2.04
GenDot / IPartOpt / cpu / BothRev 0.000018914 s 0.00001097419997677207 s 1.72
GenDot / DefOpt / cpu / PreRev 0.000019261 s 0.000010888840006373356 s 1.77
GenDot / DefOpt / cpu / PostRev 0.000019324 s 0.000011311660045976168 s 1.71
GenDot / DefOpt / cpu / BothRev 0.000019316 s 0.000011086939975939458 s 1.74
GenDot / IDefOpt / cpu / PreRev 0.000019207 s 0.000010983839974869624 s 1.75
GenDot / IDefOpt / cpu / PostRev 0.000019327 s 0.000011606319985730806 s 1.67
GenDot / IDefOpt / cpu / BothRev 0.000018755 s 0.000011533079996297602 s 1.63
GenDot / JaXPipe / cpu / Primal 0.00001 s 0.000007380500019280589 s 1.35
GenDot / Jax / cpu / Primal 0.00001 s 0.000006958400008443277 s 1.44
GenDot / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.0000074693799797387325 s 1.20
GenDot / PartOpt / cpu / Primal 0.00001 s 0.000007400719969155034 s 1.35
GenDot / IPartOpt / cpu / Primal 0.000033 s 0.000007423959987136186 s 4.45
GenDot / DefOpt / cpu / Primal 0.00001 s 0.000007572139984404202 s 1.32
GenDot / IDefOpt / cpu / Primal 0.00001 s 0.000007332039976972737 s 1.36
GenDot / JaXPipe / cpu / Forward 0.000015 s 0.000010938879950117552 s 1.37
GenDot / Jax / cpu / Forward 0.000014 s 0.000010100020008394494 s 1.39
GenDot / HLOOpt / cpu / Forward 0.000013 s 0.000011385520001567785 s 1.14
GenDot / PartOpt / cpu / Forward 0.000014 s 0.00001056173997312726 s 1.33
GenDot / IPartOpt / cpu / Forward 0.000014 s 0.000011640999982773792 s 1.20
GenDot / DefOpt / cpu / Forward 0.000013 s 0.000010874920017158729 s 1.20
GenDot / IDefOpt / cpu / Forward 0.000014 s 0.0000112198199985869 s 1.25
GenDot / JaXPipe / cpu / PreRev 0.000014 s 0.000010976620023939176 s 1.28
GenDot / JaXPipe / cpu / PostRev 0.000014 s 0.000010879080018639798 s 1.29
GenDot / JaXPipe / cpu / BothRev 0.000016 s 0.000011393520044293836 s 1.40
GenDot / Jax / cpu / BothRev 0.000014 s 0.000011286379994999153 s 1.24
GenDot / HLOOpt / cpu / PreRev 0.000014 s 0.000011677359971145052 s 1.20
GenDot / HLOOpt / cpu / PostRev 0.000014 s 0.000013176040029065915 s 1.06
GenDot / HLOOpt / cpu / BothRev 0.000014 s 0.000011039759983759723 s 1.27
GenDot / PartOpt / cpu / PreRev 0.000014 s 0.000011172839967912297 s 1.25
GenDot / PartOpt / cpu / PostRev 0.000015 s 0.000010638620024110423 s 1.41
GenDot / PartOpt / cpu / BothRev 0.000014 s 0.000011430280028434937 s 1.22
GenDot / IPartOpt / cpu / PreRev 0.000013 s 0.00001093669999136182 s 1.19
GenDot / IPartOpt / cpu / PostRev 0.000014 s 0.000010266539984513656 s 1.36
GenDot / IPartOpt / cpu / BothRev 0.000014 s 0.00001097419997677207 s 1.28
GenDot / DefOpt / cpu / PreRev 0.000014 s 0.000010888840006373356 s 1.29
GenDot / DefOpt / cpu / PostRev 0.000013 s 0.000011311660045976168 s 1.15
GenDot / DefOpt / cpu / BothRev 0.000014 s 0.000011086939975939458 s 1.26
GenDot / IDefOpt / cpu / PreRev 0.000014 s 0.000010983839974869624 s 1.27
GenDot / IDefOpt / cpu / PostRev 0.000014 s 0.000011606319985730806 s 1.21
GenDot / IDefOpt / cpu / BothRev 0.000014 s 0.000011533079996297602 s 1.21
hlo_ffi / JaXPipe / cpu / Primal 0.0000114971199491265 s 0.000010507380011404166 s 1.09
hlo_ffi / Jax / cpu / Primal 0.000011023659999409574 s 0.000010191680021307549 s 1.08
hlo_ffi / HLOOpt / cpu / Primal 0.00001076843998816912 s 0.000010493139971003984 s 1.03
hlo_ffi / PartOpt / cpu / Primal 0.000010758739963421248 s 0.000009898840016830945 s 1.09
hlo_ffi / IPartOpt / cpu / Primal 0.000011074819985879004 s 0.00001044735998220858 s 1.06
hlo_ffi / DefOpt / cpu / Primal 0.000010836119972736924 s 0.00000984727998911694 s 1.10
hlo_ffi / IDefOpt / cpu / Primal 0.000010937820034087054 s 0.0000097175999871979 s 1.13
hlo_ffi / JaXPipe / cpu / Forward 0.000015991480013326508 s 0.000014715980014443633 s 1.09
hlo_ffi / Jax / cpu / Forward 0.00001563499999065243 s 0.000014583759984816424 s 1.07
hlo_ffi / HLOOpt / cpu / Forward 0.000016099299964480453 s 0.000014939820002837223 s 1.08
hlo_ffi / PartOpt / cpu / Forward 0.000016378539949073457 s 0.000015221720050249132 s 1.08
hlo_ffi / IPartOpt / cpu / Forward 0.00001678407998952025 s 0.000014407819971893332 s 1.16
hlo_ffi / DefOpt / cpu / Forward 0.00001556542002617789 s 0.000014747760023965385 s 1.06
hlo_ffi / IDefOpt / cpu / Forward 0.00001599095994606614 s 0.000014539300009346334 s 1.10
hlo_ffi / JaXPipe / cpu / PreRev 0.00001636542000596819 s 0.000015279680028470465 s 1.07
hlo_ffi / JaXPipe / cpu / PostRev 0.00001576246005242865 s 0.000014618960030929885 s 1.08
hlo_ffi / JaXPipe / cpu / BothRev 0.000015899540039754355 s 0.000014288980009951048 s 1.11
hlo_ffi / Jax / cpu / BothRev 0.000015510339972024666 s 0.000014828140001554857 s 1.05
hlo_ffi / HLOOpt / cpu / PreRev 0.000016182419958568062 s 0.00001529568003206805 s 1.06
hlo_ffi / HLOOpt / cpu / PostRev 0.000017434820028938704 s 0.000016495360005137626 s 1.06
hlo_ffi / HLOOpt / cpu / BothRev 0.000015553839984931985 s 0.000014466640031969293 s 1.08
hlo_ffi / PartOpt / cpu / PreRev 0.00001562618001116789 s 0.000015213259957818082 s 1.03
hlo_ffi / PartOpt / cpu / PostRev 0.00001534126000478864 s 0.00001446928001314518 s 1.06
hlo_ffi / PartOpt / cpu / BothRev 0.000015374000013252954 s 0.000014484280009128267 s 1.06
hlo_ffi / IPartOpt / cpu / PreRev 0.000015985499985617934 s 0.00001538815998173959 s 1.04
hlo_ffi / IPartOpt / cpu / PostRev 0.000015361799978563794 s 0.000014437160025408956 s 1.06
hlo_ffi / IPartOpt / cpu / BothRev 0.00001547415999993973 s 0.00001444336003260105 s 1.07
hlo_ffi / DefOpt / cpu / PreRev 0.00001559753999572422 s 0.000014806700019107666 s 1.05
hlo_ffi / DefOpt / cpu / PostRev 0.00001551879999169614 s 0.000014446859968302304 s 1.07
hlo_ffi / DefOpt / cpu / BothRev 0.000014975040012359386 s 0.000014170280010148418 s 1.06
hlo_ffi / IDefOpt / cpu / PreRev 0.00001614406006410718 s 0.000014684280004075844 s 1.10
hlo_ffi / IDefOpt / cpu / PostRev 0.000015237719953802296 s 0.000014284059980127494 s 1.07
hlo_ffi / IDefOpt / cpu / BothRev 0.000014898040017214952 s 0.000013923899996370891 s 1.07
hlo_ffi / JaXPipe / cuda / Primal 0.0000023670000000000004 s 0.000001983 s 1.19
hlo_ffi / Jax / cuda / Primal 0.000002368 s 0.000001983 s 1.19
hlo_ffi / HLOOpt / cuda / Primal 0.0000023670000000000004 s 0.000001983 s 1.19
hlo_ffi / PartOpt / cuda / Primal 0.000002368 s 0.000001983 s 1.19
hlo_ffi / IPartOpt / cuda / Primal 0.0000023670000000000004 s 0.000001983 s 1.19
hlo_ffi / DefOpt / cuda / Primal 0.000002368 s 0.000001983 s 1.19
hlo_ffi / IDefOpt / cuda / Primal 0.000002368 s 0.000001983 s 1.19
hlo_ffi / JaXPipe / cuda / Forward 0.000002463 s 0.000002047 s 1.20
hlo_ffi / Jax / cuda / Forward 0.000002463 s 0.000002047 s 1.20
hlo_ffi / HLOOpt / cuda / Forward 0.000002463 s 0.000002048 s 1.20
hlo_ffi / PartOpt / cuda / Forward 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IPartOpt / cuda / Forward 0.000002463 s 0.000002048 s 1.20
hlo_ffi / DefOpt / cuda / Forward 0.000002463 s 0.000002079 s 1.18
hlo_ffi / IDefOpt / cuda / Forward 0.000002463 s 0.000002048 s 1.20
hlo_ffi / JaXPipe / cuda / PreRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / JaXPipe / cuda / PostRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / JaXPipe / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / Jax / cuda / BothRev 0.000002433 s 0.000002047 s 1.19
hlo_ffi / HLOOpt / cuda / PreRev 0.000002431 s 0.000002047 s 1.19
hlo_ffi / HLOOpt / cuda / PostRev 0.000002463 s 0.000002048 s 1.20
hlo_ffi / HLOOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / PartOpt / cuda / PreRev 0.000002433 s 0.000002047 s 1.19
hlo_ffi / PartOpt / cuda / PostRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / PartOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IPartOpt / cuda / PreRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / IPartOpt / cuda / PostRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / IPartOpt / cuda / BothRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / DefOpt / cuda / PreRev 0.000002432 s 0.000002047 s 1.19
hlo_ffi / DefOpt / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / DefOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IDefOpt / cuda / PreRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IDefOpt / cuda / PostRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / IDefOpt / cuda / BothRev 0.000002463 s 0.000002047 s 1.20
hlo_ffi / JaXPipe / tpu / Primal 9.17e-7 s 9.2075e-7 s 1.00
hlo_ffi / Jax / tpu / Primal 9.50175e-7 s 9.50725e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Primal 8.963e-7 s 8.9765e-7 s 1.00
hlo_ffi / PartOpt / tpu / Primal 9.51025e-7 s 9.5555e-7 s 1.00
hlo_ffi / IPartOpt / tpu / Primal 9.00375e-7 s 9.008e-7 s 1.00
hlo_ffi / DefOpt / tpu / Primal 9.5185e-7 s 9.58e-7 s 0.99
hlo_ffi / IDefOpt / tpu / Primal 8.966999999999999e-7 s 8.996999999999999e-7 s 1.00
hlo_ffi / JaXPipe / tpu / Forward 9.4865e-7 s 9.48525e-7 s 1.00
hlo_ffi / Jax / tpu / Forward 9.81125e-7 s 9.819500000000002e-7 s 1.00
hlo_ffi / HLOOpt / tpu / Forward 9.73575e-7 s 9.7365e-7 s 1.00
hlo_ffi / PartOpt / tpu / Forward 9.334e-7 s 9.33825e-7 s 1.00
hlo_ffi / IPartOpt / tpu / Forward 9.7345e-7 s 9.739499999999998e-7 s 1.00
hlo_ffi / DefOpt / tpu / Forward 9.33775e-7 s 9.33575e-7 s 1.00
hlo_ffi / IDefOpt / tpu / Forward 9.7325e-7 s 9.73425e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PreRev 9.31075e-7 s 9.314e-7 s 1.00
hlo_ffi / JaXPipe / tpu / PostRev 9.65025e-7 s 9.64975e-7 s 1.00
hlo_ffi / JaXPipe / tpu / BothRev 9.616e-7 s 9.619e-7 s 1.00
hlo_ffi / Jax / tpu / BothRev 9.64875e-7 s 9.65125e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PreRev 9.612999999999998e-7 s 9.61875e-7 s 1.00
hlo_ffi / HLOOpt / tpu / PostRev 9.64175e-7 s 9.64775e-7 s 1.00
hlo_ffi / HLOOpt / tpu / BothRev 9.61825e-7 s 9.62e-7 s 1.00
hlo_ffi / PartOpt / tpu / PreRev 9.646e-7 s 9.642e-7 s 1.00
hlo_ffi / PartOpt / tpu / PostRev 9.61875e-7 s 9.61425e-7 s 1.00
hlo_ffi / PartOpt / tpu / BothRev 9.649e-7 s 9.64e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PreRev 9.62025e-7 s 9.61425e-7 s 1.00
hlo_ffi / IPartOpt / tpu / PostRev 9.647e-7 s 9.64525e-7 s 1.00
hlo_ffi / IPartOpt / tpu / BothRev 9.6145e-7 s 9.6245e-7 s 1.00
hlo_ffi / DefOpt / tpu / PreRev 9.642e-7 s 9.6415e-7 s 1.00
hlo_ffi / DefOpt / tpu / PostRev 9.61875e-7 s 9.61775e-7 s 1.00
hlo_ffi / DefOpt / tpu / BothRev 9.647e-7 s 9.6445e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PreRev 9.61625e-7 s 9.62e-7 s 1.00
hlo_ffi / IDefOpt / tpu / PostRev 9.65e-7 s 9.649e-7 s 1.00
hlo_ffi / IDefOpt / tpu / BothRev 9.618e-7 s 9.6205e-7 s 1.00
hlo_ffi / JaXPipe / cpu / Primal 0.000017865 s 0.000010507380011404166 s 1.70
hlo_ffi / Jax / cpu / Primal 0.000017827 s 0.000010191680021307549 s 1.75
hlo_ffi / HLOOpt / cpu / Primal 0.000018075 s 0.000010493139971003984 s 1.72
hlo_ffi / PartOpt / cpu / Primal 0.000017823 s 0.000009898840016830945 s 1.80
hlo_ffi / IPartOpt / cpu / Primal 0.000017508 s 0.00001044735998220858 s 1.68
hlo_ffi / DefOpt / cpu / Primal 0.000017629 s 0.00000984727998911694 s 1.79
hlo_ffi / IDefOpt / cpu / Primal 0.000017798 s 0.0000097175999871979 s 1.83
hlo_ffi / JaXPipe / cpu / Forward 0.00002454 s 0.000014715980014443633 s 1.67
hlo_ffi / Jax / cpu / Forward 0.000024067 s 0.000014583759984816424 s 1.65
hlo_ffi / HLOOpt / cpu / Forward 0.000024372 s 0.000014939820002837223 s 1.63
hlo_ffi / PartOpt / cpu / Forward 0.000024303 s 0.000015221720050249132 s 1.60
hlo_ffi / IPartOpt / cpu / Forward 0.000023842 s 0.000014407819971893332 s 1.65
hlo_ffi / DefOpt / cpu / Forward 0.000024169 s 0.000014747760023965385 s 1.64
hlo_ffi / IDefOpt / cpu / Forward 0.000024261 s 0.000014539300009346334 s 1.67
hlo_ffi / JaXPipe / cpu / PreRev 0.000024547 s 0.000015279680028470465 s 1.61
hlo_ffi / JaXPipe / cpu / PostRev 0.000023923 s 0.000014618960030929885 s 1.64
hlo_ffi / JaXPipe / cpu / BothRev 0.000024392 s 0.000014288980009951048 s 1.71
hlo_ffi / Jax / cpu / BothRev 0.000023979 s 0.000014828140001554857 s 1.62
hlo_ffi / HLOOpt / cpu / PreRev 0.00002422 s 0.00001529568003206805 s 1.58
hlo_ffi / HLOOpt / cpu / PostRev 0.000024144 s 0.000016495360005137626 s 1.46
hlo_ffi / HLOOpt / cpu / BothRev 0.000024135 s 0.000014466640031969293 s 1.67
hlo_ffi / PartOpt / cpu / PreRev 0.000024096 s 0.000015213259957818082 s 1.58
hlo_ffi / PartOpt / cpu / PostRev 0.000023941 s 0.00001446928001314518 s 1.65
hlo_ffi / PartOpt / cpu / BothRev 0.000023882 s 0.000014484280009128267 s 1.65
hlo_ffi / IPartOpt / cpu / PreRev 0.000024551 s 0.00001538815998173959 s 1.60
hlo_ffi / IPartOpt / cpu / PostRev 0.000024032 s 0.000014437160025408956 s 1.66
hlo_ffi / IPartOpt / cpu / BothRev 0.000023407 s 0.00001444336003260105 s 1.62
hlo_ffi / DefOpt / cpu / PreRev 0.000024962000000000003 s 0.000014806700019107666 s 1.69
hlo_ffi / DefOpt / cpu / PostRev 0.000023738 s 0.000014446859968302304 s 1.64
hlo_ffi / DefOpt / cpu / BothRev 0.000023519000000000003 s 0.000014170280010148418 s 1.66
hlo_ffi / IDefOpt / cpu / PreRev 0.000024371 s 0.000014684280004075844 s 1.66
hlo_ffi / IDefOpt / cpu / PostRev 0.000024122 s 0.000014284059980127494 s 1.69
hlo_ffi / IDefOpt / cpu / BothRev 0.00002453 s 0.000013923899996370891 s 1.76
hlo_ffi / JaXPipe / cpu / Primal 0.000013 s 0.000010507380011404166 s 1.24
hlo_ffi / Jax / cpu / Primal 0.000013 s 0.000010191680021307549 s 1.28
hlo_ffi / HLOOpt / cpu / Primal 0.000012 s 0.000010493139971003984 s 1.14
hlo_ffi / PartOpt / cpu / Primal 0.000013 s 0.000009898840016830945 s 1.31
hlo_ffi / IPartOpt / cpu / Primal 0.000013 s 0.00001044735998220858 s 1.24
hlo_ffi / DefOpt / cpu / Primal 0.000013 s 0.00000984727998911694 s 1.32
hlo_ffi / IDefOpt / cpu / Primal 0.000012 s 0.0000097175999871979 s 1.23
hlo_ffi / JaXPipe / cpu / Forward 0.000017999999999999997 s 0.000014715980014443633 s 1.22
hlo_ffi / Jax / cpu / Forward 0.000017 s 0.000014583759984816424 s 1.17
hlo_ffi / HLOOpt / cpu / Forward 0.000017 s 0.000014939820002837223 s 1.14
hlo_ffi / PartOpt / cpu / Forward 0.000017999999999999997 s 0.000015221720050249132 s 1.18
hlo_ffi / IPartOpt / cpu / Forward 0.000016 s 0.000014407819971893332 s 1.11
hlo_ffi / DefOpt / cpu / Forward 0.000017 s 0.000014747760023965385 s 1.15
hlo_ffi / IDefOpt / cpu / Forward 0.000017 s 0.000014539300009346334 s 1.17
hlo_ffi / JaXPipe / cpu / PreRev 0.000017 s 0.000015279680028470465 s 1.11
hlo_ffi / JaXPipe / cpu / PostRev 0.000017 s 0.000014618960030929885 s 1.16
hlo_ffi / JaXPipe / cpu / BothRev 0.000017 s 0.000014288980009951048 s 1.19
hlo_ffi / Jax / cpu / BothRev 0.000017 s 0.000014828140001554857 s 1.15
hlo_ffi / HLOOpt / cpu / PreRev 0.000017999999999999997 s 0.00001529568003206805 s 1.18
hlo_ffi / HLOOpt / cpu / PostRev 0.000017999999999999997 s 0.000016495360005137626 s 1.09
hlo_ffi / HLOOpt / cpu / BothRev 0.000017999999999999997 s 0.000014466640031969293 s 1.24
hlo_ffi / PartOpt / cpu / PreRev 0.000017999999999999997 s 0.000015213259957818082 s 1.18
hlo_ffi / PartOpt / cpu / PostRev 0.000017999999999999997 s 0.00001446928001314518 s 1.24
hlo_ffi / PartOpt / cpu / BothRev 0.000017 s 0.000014484280009128267 s 1.17
hlo_ffi / IPartOpt / cpu / PreRev 0.000017 s 0.00001538815998173959 s 1.10
hlo_ffi / IPartOpt / cpu / PostRev 0.000017 s 0.000014437160025408956 s 1.18
hlo_ffi / IPartOpt / cpu / BothRev 0.000017 s 0.00001444336003260105 s 1.18
hlo_ffi / DefOpt / cpu / PreRev 0.000017999999999999997 s 0.000014806700019107666 s 1.22
hlo_ffi / DefOpt / cpu / PostRev 0.000017999999999999997 s 0.000014446859968302304 s 1.25
hlo_ffi / DefOpt / cpu / BothRev 0.000017 s 0.000014170280010148418 s 1.20
hlo_ffi / IDefOpt / cpu / PreRev 0.000017999999999999997 s 0.000014684280004075844 s 1.23
hlo_ffi / IDefOpt / cpu / PostRev 0.000017 s 0.000014284059980127494 s 1.19
hlo_ffi / IDefOpt / cpu / BothRev 0.000017 s 0.000013923899996370891 s 1.22
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.0009070332001101 s 0.0009460127999773 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 / Primal 0.0008897188001355 s 0.0009934444001373 s 0.90
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Primal 0.0009557084001244 s 0.0010689778000596 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 / PartOpt / cpu / Primal 0.0008904666000489 s 0.0010349504000259 s 0.86
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 0.0008868460001394 s 0.0009978774000046 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 / DefOpt / cpu / Primal 0.0009776600000805 s 0.0010341524000978 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.0009453785999539 s 0.0010480445999746 s 0.90
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Forward 0.0021758099999715 s 0.0024604864000139 s 0.88
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 0.0022678528000142 s 0.0025904444000843 s 0.88
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 0.0021896639999795 s 0.0024391827999352 s 0.90
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.0021712452000429 s 0.0025809069999922 s 0.84
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 0.0022081248000176 s 0.0023809604000234 s 0.93
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.0022074225999858 s 0.002511922999929 s 0.88
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 0.0021169382000152 s 0.0025016743999913 s 0.85
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PreRev 0.0052199451998603 s 0.0060883768001986 s 0.86
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.0049577337999835 s 0.006095195799844 s 0.81
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / BothRev 0.005097540800034 s 0.0059681835999072 s 0.85
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / BothRev 0.0068690245999277 s 0.0053168176000326 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 / HLOOpt / cpu / PreRev 0.0032261718000881 s 0.0059495341999536 s 0.54
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.0053606323999702 s 0.0040034562000073 s 1.34
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.0048095441999976 s 0.0061645880000469 s 0.78
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PreRev 0.0054836986000736 s 0.003468521799914 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 / PostRev 0.005098297600125 s 0.0053736331999971 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.0070335463999072 s 0.0041171870000653 s 1.71
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.0047328847999779 s 0.0054194587999518 s 0.87
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PostRev 0.0051473130000886 s 0.0060942769999201 s 0.84
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 0.0054311668000082 s 0.0036507543999505 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 / DefOpt / cpu / PreRev 0.0054712799999833 s 0.003629080399969 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 / DefOpt / cpu / PostRev 0.0030859715999213 s 0.0034791041999596 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 / DefOpt / cpu / BothRev 0.0057488558000841 s 0.0037215792000097 s 1.54
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / PreRev 0.0048766928000077 s 0.0054757472001256 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 / IDefOpt / cpu / PostRev 0.0032340202000341 s 0.0052939145999516 s 0.61
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.0049255152000114 s 0.0036072935998163 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 / JaXPipe / cuda / Primal 0.000295038 s 0.000279774 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / Primal 0.000295006 s 0.000279231 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Primal 0.000301246 s 0.000284959 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 / PartOpt / cuda / Primal 0.000294974 s 0.000278814 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Primal 0.000294751 s 0.000279134 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Primal 0.000302367 s 0.000285471 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / Primal 0.000301503 s 0.00028707 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / Forward 0.000582461 s 0.000555005 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cuda / Forward 0.0005672289999999 s 0.000537277 s 1.06
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / Forward 0.000583196 s 0.000555069 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / Forward 0.000582333 s 0.000553886 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / Forward 0.000582909 s 0.000555292 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / Forward 0.000583037 s 0.000554525 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / Forward 0.0005828449999999 s 0.000554909 s 1.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / PreRev 0.001054778 s 0.001020762 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / PostRev 0.00100681 s 0.000980795 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cuda / BothRev 0.00104889 s 0.001019547 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 / Jax / cuda / BothRev 0.001005467 s 0.000985979 s 1.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PreRev 0.001035642 s 0.0010087619999999 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / PostRev 0.001060027 s 0.001032858 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cuda / BothRev 0.001034746 s 0.001007546 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PreRev 0.0010494019999999 s 0.001020218 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / PostRev 0.001000602 s 0.000972187 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cuda / BothRev 0.001051834 s 0.001021531 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PreRev 0.00105161 s 0.001019771 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / PostRev 0.000997146 s 0.000970139 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cuda / BothRev 0.001049563 s 0.0010194499999999 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / PreRev 0.0010519309999999 s 0.001016314 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 / DefOpt / cuda / PostRev 0.000984379 s 0.000954459 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cuda / BothRev 0.001050939 s 0.001016475 s 1.03
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PreRev 0.001050971 s 0.001013723 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / PostRev 0.001050234 s 0.001011739 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cuda / BothRev 0.001052026 s 0.00101593 s 1.04
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / Primal 0.00012355425 s 0.0001299015 s 0.95
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / Primal 0.000126795 s 0.000124273 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 0.00015223575 s 0.00015880075 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 / tpu / Primal 0.00013398 s 0.000131278 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 / IPartOpt / tpu / Primal 0.0001309379999999 s 0.00013698725 s 0.96
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / Primal 0.00014811225 s 0.0001449175 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 0.00015113825 s 0.000157165 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 / tpu / Forward 0.00021204375 s 0.00021402325 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 0.0002611332499999 s 0.00026188075 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 0.0002122015 s 0.0002204252499999 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 / tpu / Forward 0.0002181497499999 s 0.000213826 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 / IPartOpt / tpu / Forward 0.00021202675 s 0.0002159719999999 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / Forward 0.00021834975 s 0.0002182787499999 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / Forward 0.00021181175 s 0.0002160159999999 s 0.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PreRev 0.0003560035 s 0.0003572715 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / PostRev 0.000256737 s 0.0002582592499999 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / tpu / BothRev 0.0003566477499999 s 0.000357125 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / tpu / BothRev 0.00025712375 s 0.00025970175 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / PreRev 0.00035665025 s 0.00035714525 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / tpu / PostRev 0.0002908265 s 0.00029214375 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 0.000356326 s 0.00035686275 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 0.0003563137499999 s 0.00035705575 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 0.0002724195 s 0.0002741289999999 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 0.00035565925 s 0.000356627 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 0.00035625725 s 0.000356651 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / PostRev 0.00027259475 s 0.0002747175 s 0.99
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / tpu / BothRev 0.00035660775 s 0.0003570965 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / PreRev 0.0003583305 s 0.00035930825 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 0.00028461375 s 0.000285066 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / tpu / BothRev 0.0003580977499999 s 0.0003592965 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 0.0003585865 s 0.000359561 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 0.00030118075 s 0.000301966 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / tpu / BothRev 0.00035889625 s 0.00035916375 s 1.00
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / Primal 0.001735584 s 0.0009460127999773 s 1.83
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 0.001726269 s 0.0009934444001373 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 / HLOOpt / cpu / Primal 0.001765877 s 0.0010689778000596 s 1.65
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.001681798 s 0.0010349504000259 s 1.63
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Primal 0.001731864 s 0.0009978774000046 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 / DefOpt / cpu / Primal 0.001644259 s 0.0010341524000978 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 / Primal 0.001839229 s 0.0010480445999746 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 / JaXPipe / cpu / Forward 0.004877434 s 0.0024604864000139 s 1.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.00535523 s 0.0025904444000843 s 2.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / Forward 0.004994767 s 0.0024391827999352 s 2.05
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.004951845 s 0.0025809069999922 s 1.92
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 0.0047041209999999 s 0.0023809604000234 s 1.98
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / Forward 0.004724548 s 0.002511922999929 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 / IDefOpt / cpu / Forward 0.005201732 s 0.0025016743999913 s 2.08
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 0.009912591 s 0.0060883768001986 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 / JaXPipe / cpu / PostRev 0.009310295 s 0.006095195799844 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 / JaXPipe / cpu / BothRev 0.008167704 s 0.0059681835999072 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 / Jax / cpu / BothRev 0.007795632 s 0.0053168176000326 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 / HLOOpt / cpu / PreRev 0.007863821 s 0.0059495341999536 s 1.32
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.0071633009999999 s 0.0040034562000073 s 1.79
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.007465761 s 0.0061645880000469 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 / PartOpt / cpu / PreRev 0.0069930669999999 s 0.003468521799914 s 2.02
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / PostRev 0.008278797 s 0.0053736331999971 s 1.54
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.007782511 s 0.0041171870000653 s 1.89
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PreRev 0.008081177 s 0.0054194587999518 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 / PostRev 0.007992896 s 0.0060942769999201 s 1.31
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / BothRev 0.008819492 s 0.0036507543999505 s 2.42
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 0.007516503 s 0.003629080399969 s 2.07
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / DefOpt / cpu / PostRev 0.008106545 s 0.0034791041999596 s 2.33
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 0.0090251339999999 s 0.0037215792000097 s 2.43
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 0.008708384 s 0.0054757472001256 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 0.008200996 s 0.0052939145999516 s 1.55
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / BothRev 0.0072457589999999 s 0.0036072935998163 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 / JaXPipe / cpu / Primal 0.003329 s 0.0009460127999773 s 3.52
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Primal 0.001743 s 0.0009934444001373 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 / HLOOpt / cpu / Primal 0.002633 s 0.0010689778000596 s 2.46
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Primal 0.00385 s 0.0010349504000259 s 3.72
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Primal 0.0035619999999999 s 0.0009978774000046 s 3.57
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 0.001829 s 0.0010341524000978 s 1.77
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IDefOpt / cpu / Primal 0.001658 s 0.0010480445999746 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 / JaXPipe / cpu / Forward 0.004432 s 0.0024604864000139 s 1.80
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / Jax / cpu / Forward 0.004777 s 0.0025904444000843 s 1.84
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 0.009007 s 0.0024391827999352 s 3.69
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / Forward 0.004983 s 0.0025809069999922 s 1.93
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / Forward 0.0042169999999999 s 0.0023809604000234 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 0.007331 s 0.002511922999929 s 2.92
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 0.005327 s 0.0025016743999913 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 / PreRev 0.0132609999999999 s 0.0060883768001986 s 2.18
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / JaXPipe / cpu / PostRev 0.011122 s 0.006095195799844 s 1.82
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 0.007751 s 0.0059681835999072 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 / Jax / cpu / BothRev 0.009282 s 0.0053168176000326 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 / HLOOpt / cpu / PreRev 0.010691 s 0.0059495341999536 s 1.80
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / PostRev 0.007776 s 0.0040034562000073 s 1.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / HLOOpt / cpu / BothRev 0.009609 s 0.0061645880000469 s 1.56
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 0.009075 s 0.003468521799914 s 2.62
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 0.010451 s 0.0053736331999971 s 1.94
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / PartOpt / cpu / BothRev 0.012333 s 0.0041171870000653 s 3.00
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 0.0089199999999999 s 0.0054194587999518 s 1.65
llama_dim_288_hidden_dim_768_n_layers_6_n_heads_6_n_kv_heads_6_vocab_size_32000_seq_len_256 / IPartOpt / cpu / PostRev 0.00965 s 0.0060942769999201 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 / IPartOpt / cpu / BothRev 0.017512 s 0.0036507543999505 s 4.80
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 0.0084079999999999 s 0.003629080399969 s 2.32
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 0.012897 s 0.0034791041999596 s 3.71
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 0.0134 s 0.0037215792000097 s 3.60
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 0.014815 s 0.0054757472001256 s 2.71
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 0.015797 s 0.0052939145999516 s 2.98
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 0.008395 s 0.0036072935998163 s 2.33
scatter_sum / JaXPipe / cpu / Primal 0.000010372719980296095 s 0.000007983779933056211 s 1.30
scatter_sum / Jax / cpu / Primal 0.000009912239966070046 s 0.000007660200008103857 s 1.29
scatter_sum / HLOOpt / cpu / Primal 0.000009924140031216666 s 0.000008045299973673536 s 1.23
scatter_sum / PartOpt / cpu / Primal 0.000009251980054614253 s 0.000007703000028413953 s 1.20
scatter_sum / IPartOpt / cpu / Primal 0.000009658720018705936 s 0.00000794093997683376 s 1.22
scatter_sum / DefOpt / cpu / Primal 0.000009918779942381662 s 0.000007402840001304867 s 1.34
scatter_sum / IDefOpt / cpu / Primal 0.000009339999996882396 s 0.000007818260028216173 s 1.19
scatter_sum / JaXPipe / cpu / Forward 0.000013658019970534953 s 0.000012014319991067168 s 1.14
scatter_sum / Jax / cpu / Forward 0.0000133587399704993 s 0.000012560279974422884 s 1.06
scatter_sum / HLOOpt / cpu / Forward 0.000014088160005485406 s 0.000012528000006568618 s 1.12
scatter_sum / PartOpt / cpu / Forward 0.0000136497199855512 s 0.00001192050000099698 s 1.15
scatter_sum / IPartOpt / cpu / Forward 0.000014050600002519786 s 0.000012312659964663909 s 1.14
scatter_sum / DefOpt / cpu / Forward 0.000013517680026779998 s 0.000012823139995816743 s 1.05
scatter_sum / IDefOpt / cpu / Forward 0.000013609839961645776 s 0.000012143160010964494 s 1.12
scatter_sum / JaXPipe / cpu / PreRev 0.000014266079997469204 s 0.000011788539995905013 s 1.21
scatter_sum / JaXPipe / cpu / PostRev 0.000013312580022102338 s 0.000012099339992346358 s 1.10
scatter_sum / JaXPipe / cpu / BothRev 0.00001334502001554938 s 0.000012275820026843575 s 1.09
scatter_sum / Jax / cpu / BothRev 0.000013628260021505411 s 0.000011990059992967873 s 1.14
scatter_sum / HLOOpt / cpu / PreRev 0.00001374890000988671 s 0.000012726860031762044 s 1.08
scatter_sum / HLOOpt / cpu / PostRev 0.00001559102003739099 s 0.00001469114001338312 s 1.06
scatter_sum / HLOOpt / cpu / BothRev 0.000013620279969472905 s 0.000011913419994016297 s 1.14
scatter_sum / PartOpt / cpu / PreRev 0.000013304680014698534 s 0.000011582080005609897 s 1.15
scatter_sum / PartOpt / cpu / PostRev 0.0000137740599893732 s 0.000012619759972949396 s 1.09
scatter_sum / PartOpt / cpu / BothRev 0.000014710540053783916 s 0.000012854620026701011 s 1.14
scatter_sum / IPartOpt / cpu / PreRev 0.000013702019978154566 s 0.000012315319982008077 s 1.11
scatter_sum / IPartOpt / cpu / PostRev 0.000014117320033619762 s 0.00001224900001034257 s 1.15
scatter_sum / IPartOpt / cpu / BothRev 0.000013243819958006497 s 0.000011994440019407193 s 1.10
scatter_sum / DefOpt / cpu / PreRev 0.000013408099957814555 s 0.000011871319975398364 s 1.13
scatter_sum / DefOpt / cpu / PostRev 0.00001372332003484189 s 0.000012172800015832763 s 1.13
scatter_sum / DefOpt / cpu / BothRev 0.000014000059964018874 s 0.00001243236000846082 s 1.13
scatter_sum / IDefOpt / cpu / PreRev 0.000013256960028229512 s 0.000011855739994643957 s 1.12
scatter_sum / IDefOpt / cpu / PostRev 0.00001347286003692716 s 0.000012374520028970436 s 1.09
scatter_sum / IDefOpt / cpu / BothRev 0.000014031419977982296 s 0.000012115460012864787 s 1.16
scatter_sum / JaXPipe / cuda / Primal 0.00001072 s 0.000009792 s 1.09
scatter_sum / Jax / cuda / Primal 0.00001056 s 0.000009952 s 1.06
scatter_sum / HLOOpt / cuda / Primal 0.000010496 s 0.000010368 s 1.01
scatter_sum / PartOpt / cuda / Primal 0.000010368 s 0.000010272 s 1.01
scatter_sum / IPartOpt / cuda / Primal 0.000010688 s 0.000009952 s 1.07
scatter_sum / DefOpt / cuda / Primal 0.000010752 s 0.000009984 s 1.08
scatter_sum / IDefOpt / cuda / Primal 0.000010368 s 0.00000992 s 1.05
scatter_sum / JaXPipe / cuda / Forward 0.000017472 s 0.000017152 s 1.02
scatter_sum / Jax / cuda / Forward 0.000017536 s 0.0000168 s 1.04
scatter_sum / HLOOpt / cuda / Forward 0.000017408 s 0.0000168 s 1.04
scatter_sum / PartOpt / cuda / Forward 0.000018176 s 0.000016927999999999998 s 1.07
scatter_sum / IPartOpt / cuda / Forward 0.000017887 s 0.000017344 s 1.03
scatter_sum / DefOpt / cuda / Forward 0.000017951 s 0.000016896000000000002 s 1.06
scatter_sum / IDefOpt / cuda / Forward 0.000017632 s 0.000017184 s 1.03
scatter_sum / JaXPipe / cuda / PreRev 0.00001744 s 0.000016832 s 1.04
scatter_sum / JaXPipe / cuda / PostRev 0.000017152 s 0.000016639 s 1.03
scatter_sum / JaXPipe / cuda / BothRev 0.000017951 s 0.00001648 s 1.09
scatter_sum / Jax / cuda / BothRev 0.000017728 s 0.000016672 s 1.06
scatter_sum / HLOOpt / cuda / PreRev 0.000017663 s 0.000016704 s 1.06
scatter_sum / HLOOpt / cuda / PostRev 0.000017088 s 0.00001632 s 1.05
scatter_sum / HLOOpt / cuda / BothRev 0.0000176 s 0.000017152 s 1.03
scatter_sum / PartOpt / cuda / PreRev 0.000017663 s 0.000017375999999999998 s 1.02
scatter_sum / PartOpt / cuda / PostRev 0.000018016 s 0.000016768000000000003 s 1.07
scatter_sum / PartOpt / cuda / BothRev 0.000017792 s 0.000017375999999999998 s 1.02
scatter_sum / IPartOpt / cuda / PreRev 0.000017888000000000002 s 0.000017952 s 1.00
scatter_sum / IPartOpt / cuda / PostRev 0.000017726999999999998 s 0.000017183 s 1.03
scatter_sum / IPartOpt / cuda / BothRev 0.000017247999999999998 s 0.000017184 s 1.00
scatter_sum / DefOpt / cuda / PreRev 0.000017983 s 0.000017088 s 1.05
scatter_sum / DefOpt / cuda / PostRev 0.000017247 s 0.000016512 s 1.04
scatter_sum / DefOpt / cuda / BothRev 0.000017472 s 0.000017056 s 1.02
scatter_sum / IDefOpt / cuda / PreRev 0.000017726999999999998 s 0.000016512 s 1.07
scatter_sum / IDefOpt / cuda / PostRev 0.000017375999999999998 s 0.000016896000000000002 s 1.03
scatter_sum / IDefOpt / cuda / BothRev 0.000017344 s 0.000016864 s 1.03
scatter_sum / JaXPipe / tpu / Primal 0.0000013507749999999998 s 0.0000013509499999999995 s 1.00
scatter_sum / Jax / tpu / Primal 0.0000014050000000000003 s 0.000001404675 s 1.00
scatter_sum / HLOOpt / tpu / Primal 0.0000013508 s 0.000001351025 s 1.00
scatter_sum / PartOpt / tpu / Primal 0.000001404725 s 0.0000014045500000000002 s 1.00
scatter_sum / IPartOpt / tpu / Primal 0.000001351025 s 0.00000135125 s 1.00
scatter_sum / DefOpt / tpu / Primal 0.0000014041000000000002 s 0.000001404525 s 1.00
scatter_sum / IDefOpt / tpu / Primal 0.00000135075 s 0.000001351025 s 1.00
scatter_sum / JaXPipe / tpu / Forward 0.00000270235 s 0.0000027067000000000003 s 1.00
scatter_sum / Jax / tpu / Forward 0.00000272265 s 0.000002724975 s 1.00
scatter_sum / HLOOpt / tpu / Forward 0.0000026996750000000005 s 0.0000027009 s 1.00
scatter_sum / PartOpt / tpu / Forward 0.000002690425 s 0.000002691125 s 1.00
scatter_sum / IPartOpt / tpu / Forward 0.000002700275 s 0.00000270055 s 1.00
scatter_sum / DefOpt / tpu / Forward 0.000002691 s 0.00000269025 s 1.00
scatter_sum / IDefOpt / tpu / Forward 0.000002700125 s 0.0000027003 s 1.00
scatter_sum / JaXPipe / tpu / PreRev 0.000002702675 s 0.00000269595 s 1.00
scatter_sum / JaXPipe / tpu / PostRev 0.000002687775 s 0.00000269915 s 1.00
scatter_sum / JaXPipe / tpu / BothRev 0.000002706225 s 0.0000027057 s 1.00
scatter_sum / Jax / tpu / BothRev 0.000002751 s 0.0000027427000000000004 s 1.00
scatter_sum / HLOOpt / tpu / PreRev 0.000002702425 s 0.0000027092000000000003 s 1.00
scatter_sum / HLOOpt / tpu / PostRev 0.00000274695 s 0.0000027511750000000004 s 1.00
scatter_sum / HLOOpt / tpu / BothRev 0.000002706875 s 0.000002708875 s 1.00
scatter_sum / PartOpt / tpu / PreRev 0.00000274225 s 0.000002741775 s 1.00
scatter_sum / PartOpt / tpu / PostRev 0.0000027092000000000003 s 0.000002728425 s 0.99
scatter_sum / PartOpt / tpu / BothRev 0.0000027439750000000003 s 0.000002746375 s 1.00
scatter_sum / IPartOpt / tpu / PreRev 0.000002707375 s 0.0000027063500000000004 s 1.00
scatter_sum / IPartOpt / tpu / PostRev 0.0000027407 s 0.000002742825 s 1.00
scatter_sum / IPartOpt / tpu / BothRev 0.000002704725 s 0.000002707825 s 1.00
scatter_sum / DefOpt / tpu / PreRev 0.000002744825 s 0.00000274005 s 1.00
scatter_sum / DefOpt / tpu / PostRev 0.0000027016750000000003 s 0.000002701725 s 1.00
scatter_sum / DefOpt / tpu / BothRev 0.00000274485 s 0.0000027459 s 1.00
scatter_sum / IDefOpt / tpu / PreRev 0.0000027086500000000003 s 0.000002704125 s 1.00
scatter_sum / IDefOpt / tpu / PostRev 0.000002742325 s 0.0000027453250000000003 s 1.00
scatter_sum / IDefOpt / tpu / BothRev 0.000002704525 s 0.000002705075 s 1.00
scatter_sum / JaXPipe / cpu / Primal 0.000015669 s 0.000007983779933056211 s 1.96
scatter_sum / Jax / cpu / Primal 0.000015351999999999998 s 0.000007660200008103857 s 2.00
scatter_sum / HLOOpt / cpu / Primal 0.000015765999999999998 s 0.000008045299973673536 s 1.96
scatter_sum / PartOpt / cpu / Primal 0.000015615 s 0.000007703000028413953 s 2.03
scatter_sum / IPartOpt / cpu / Primal 0.000015806 s 0.00000794093997683376 s 1.99
scatter_sum / DefOpt / cpu / Primal 0.000015745 s 0.000007402840001304867 s 2.13
scatter_sum / IDefOpt / cpu / Primal 0.000015661 s 0.000007818260028216173 s 2.00
scatter_sum / JaXPipe / cpu / Forward 0.000022765 s 0.000012014319991067168 s 1.89
scatter_sum / Jax / cpu / Forward 0.000022245 s 0.000012560279974422884 s 1.77
scatter_sum / HLOOpt / cpu / Forward 0.000022176 s 0.000012528000006568618 s 1.77
scatter_sum / PartOpt / cpu / Forward 0.000022551 s 0.00001192050000099698 s 1.89
scatter_sum / IPartOpt / cpu / Forward 0.000022482 s 0.000012312659964663909 s 1.83
scatter_sum / DefOpt / cpu / Forward 0.000022475 s 0.000012823139995816743 s 1.75
scatter_sum / IDefOpt / cpu / Forward 0.000022928 s 0.000012143160010964494 s 1.89
scatter_sum / JaXPipe / cpu / PreRev 0.000023017 s 0.000011788539995905013 s 1.95
scatter_sum / JaXPipe / cpu / PostRev 0.000022821 s 0.000012099339992346358 s 1.89
scatter_sum / JaXPipe / cpu / BothRev 0.000022403 s 0.000012275820026843575 s 1.82
scatter_sum / Jax / cpu / BothRev 0.000022471 s 0.000011990059992967873 s 1.87
scatter_sum / HLOOpt / cpu / PreRev 0.000022182 s 0.000012726860031762044 s 1.74
scatter_sum / HLOOpt / cpu / PostRev 0.000022521 s 0.00001469114001338312 s 1.53
scatter_sum / HLOOpt / cpu / BothRev 0.000022343 s 0.000011913419994016297 s 1.88
scatter_sum / PartOpt / cpu / PreRev 0.000022998 s 0.000011582080005609897 s 1.99
scatter_sum / PartOpt / cpu / PostRev 0.000023282 s 0.000012619759972949396 s 1.84
scatter_sum / PartOpt / cpu / BothRev 0.000022513 s 0.000012854620026701011 s 1.75
scatter_sum / IPartOpt / cpu / PreRev 0.000022336 s 0.000012315319982008077 s 1.81
scatter_sum / IPartOpt / cpu / PostRev 0.000022769 s 0.00001224900001034257 s 1.86
scatter_sum / IPartOpt / cpu / BothRev 0.000022825000000000003 s 0.000011994440019407193 s 1.90
scatter_sum / DefOpt / cpu / PreRev 0.000022661 s 0.000011871319975398364 s 1.91
scatter_sum / DefOpt / cpu / PostRev 0.00002272 s 0.000012172800015832763 s 1.87
scatter_sum / DefOpt / cpu / BothRev 0.000022594 s 0.00001243236000846082 s 1.82
scatter_sum / IDefOpt / cpu / PreRev 0.000022463 s 0.000011855739994643957 s 1.89
scatter_sum / IDefOpt / cpu / PostRev 0.000022514 s 0.000012374520028970436 s 1.82
scatter_sum / IDefOpt / cpu / BothRev 0.000023275 s 0.000012115460012864787 s 1.92
scatter_sum / JaXPipe / cpu / Primal 0.000011 s 0.000007983779933056211 s 1.38
scatter_sum / Jax / cpu / Primal 0.00001 s 0.000007660200008103857 s 1.31
scatter_sum / HLOOpt / cpu / Primal 0.000011 s 0.000008045299973673536 s 1.37
scatter_sum / PartOpt / cpu / Primal 0.00001 s 0.000007703000028413953 s 1.30
scatter_sum / IPartOpt / cpu / Primal 0.00001 s 0.00000794093997683376 s 1.26
scatter_sum / DefOpt / cpu / Primal 0.000011 s 0.000007402840001304867 s 1.49
scatter_sum / IDefOpt / cpu / Primal 0.00001 s 0.000007818260028216173 s 1.28
scatter_sum / JaXPipe / cpu / Forward 0.000016 s 0.000012014319991067168 s 1.33
scatter_sum / Jax / cpu / Forward 0.000015 s 0.000012560279974422884 s 1.19
scatter_sum / HLOOpt / cpu / Forward 0.000016 s 0.000012528000006568618 s 1.28
scatter_sum / PartOpt / cpu / Forward 0.000048 s 0.00001192050000099698 s 4.03
scatter_sum / IPartOpt / cpu / Forward 0.000016 s 0.000012312659964663909 s 1.30
scatter_sum / DefOpt / cpu / Forward 0.000016 s 0.000012823139995816743 s 1.25
scatter_sum / IDefOpt / cpu / Forward 0.000015 s 0.000012143160010964494 s 1.24
scatter_sum / JaXPipe / cpu / PreRev 0.000015 s 0.000011788539995905013 s 1.27
scatter_sum / JaXPipe / cpu / PostRev 0.000016 s 0.000012099339992346358 s 1.32
scatter_sum / JaXPipe / cpu / BothRev 0.000015 s 0.000012275820026843575 s 1.22
scatter_sum / Jax / cpu / BothRev 0.000015 s 0.000011990059992967873 s 1.25
scatter_sum / HLOOpt / cpu / PreRev 0.000016 s 0.000012726860031762044 s 1.26
scatter_sum / HLOOpt / cpu / PostRev 0.000017 s 0.00001469114001338312 s 1.16
scatter_sum / HLOOpt / cpu / BothRev 0.000016 s 0.000011913419994016297 s 1.34
scatter_sum / PartOpt / cpu / PreRev 0.000016 s 0.000011582080005609897 s 1.38
scatter_sum / PartOpt / cpu / PostRev 0.000016 s 0.000012619759972949396 s 1.27
scatter_sum / PartOpt / cpu / BothRev 0.000016 s 0.000012854620026701011 s 1.24
scatter_sum / IPartOpt / cpu / PreRev 0.000016 s 0.000012315319982008077 s 1.30
scatter_sum / IPartOpt / cpu / PostRev 0.000017 s 0.00001224900001034257 s 1.39
scatter_sum / IPartOpt / cpu / BothRev 0.000015 s 0.000011994440019407193 s 1.25
scatter_sum / DefOpt / cpu / PreRev 0.000016 s 0.000011871319975398364 s 1.35
scatter_sum / DefOpt / cpu / PostRev 0.000016 s 0.000012172800015832763 s 1.31
scatter_sum / DefOpt / cpu / BothRev 0.000016 s 0.00001243236000846082 s 1.29
scatter_sum / IDefOpt / cpu / PreRev 0.00005 s 0.000011855739994643957 s 4.22
scatter_sum / IDefOpt / cpu / PostRev 0.000016 s 0.000012374520028970436 s 1.29
scatter_sum / IDefOpt / cpu / BothRev 0.000017 s 0.000012115460012864787 s 1.40
slicing / JaXPipe / cpu / Primal 0.000008246720008173724 s 0.00000682609998875705 s 1.21
slicing / Jax / cpu / Primal 0.000008580660005463869 s 0.000006204580004123273 s 1.38
slicing / HLOOpt / cpu / Primal 0.000008068920024015825 s 0.000006755100002919789 s 1.19
slicing / PartOpt / cpu / Primal 0.00000816364000456815 s 0.000006392579962266609 s 1.28
slicing / IPartOpt / cpu / Primal 0.000008538739966752474 s 0.000006374380027409643 s 1.34
slicing / DefOpt / cpu / Primal 0.000008431380038018688 s 0.000006200300022101146 s 1.36
slicing / IDefOpt / cpu / Primal 0.000007858820008550538 s 0.000006124159972387133 s 1.28
slicing / JaXPipe / cpu / Forward 0.00001106811995668977 s 0.000009675040009824445 s 1.14
slicing / Jax / cpu / Forward 0.000011008380042767384 s 0.000009428260009372025 s 1.17
slicing / HLOOpt / cpu / Forward 0.000011262719990554616 s 0.000010320419996787677 s 1.09
slicing / PartOpt / cpu / Forward 0.00001050923996444908 s 0.000009372380027343752 s 1.12
slicing / IPartOpt / cpu / Forward 0.000010775119999379968 s 0.000010439299985591788 s 1.03
slicing / DefOpt / cpu / Forward 0.00001134693999119918 s 0.000009475980032220832 s 1.20
slicing / IDefOpt / cpu / Forward 0.000010927139992418234 s 0.000009360399963043164 s 1.17
slicing / JaXPipe / cpu / PreRev 0.000011676100011754898 s 0.00000984418000371079 s 1.19
slicing / JaXPipe / cpu / PostRev 0.000011612519983827951 s 0.000010596880019875244 s 1.10
slicing / JaXPipe / cpu / BothRev 0.000011619600008998532 s 0.00001076662000741635 s 1.08
slicing / Jax / cpu / BothRev 0.000011673119997794856 s 0.000010217180006293348 s 1.14
slicing / HLOOpt / cpu / PreRev 0.000012137260018789676 s 0.000010931720034932368 s 1.11
slicing / HLOOpt / cpu / PostRev 0.000013634440019814065 s 0.000016274080007860902 s 0.84
slicing / HLOOpt / cpu / BothRev 0.000011684319979394786 s 0.000010077980032292544 s 1.16
slicing / PartOpt / cpu / PreRev 0.000011580680029510404 s 0.000009662459979153936 s 1.20
slicing / PartOpt / cpu / PostRev 0.000011392360011086569 s 0.000010176299974773429 s 1.12
slicing / PartOpt / cpu / BothRev 0.000011671160009427697 s 0.000010779640024338731 s 1.08
slicing / IPartOpt / cpu / PreRev 0.000011516739959915869 s 0.000009989239997594267 s 1.15
slicing / IPartOpt / cpu / PostRev 0.00001209077998282737 s 0.00001034255996273714 s 1.17
slicing / IPartOpt / cpu / BothRev 0.00001150807997873926 s 0.00001012631999401492 s 1.14
slicing / DefOpt / cpu / PreRev 0.000011732060029316926 s 0.000010393999991720193 s 1.13
slicing / DefOpt / cpu / PostRev 0.000011674460001813711 s 0.000010519240031499068 s 1.11
slicing / DefOpt / cpu / BothRev 0.000011196160030522151 s 0.000010008200033553294 s 1.12
slicing / IDefOpt / cpu / PreRev 0.000011172040012752405 s 0.000009962880012608366 s 1.12
slicing / IDefOpt / cpu / PostRev 0.000011410999995860038 s 0.000011089439976785795 s 1.03
slicing / IDefOpt / cpu / BothRev 0.000011557860025277478 s 0.000010662259983291731 s 1.08
slicing / JaXPipe / cuda / Primal 0.000002303 s 0.000001888 s 1.22
slicing / Jax / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / HLOOpt / cuda / Primal 0.000002303 s 0.000001888 s 1.22
slicing / PartOpt / cuda / Primal 0.000002304 s 0.000001887 s 1.22
slicing / IPartOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / DefOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / IDefOpt / cuda / Primal 0.000002303 s 0.000001887 s 1.22
slicing / JaXPipe / cuda / Forward 0.000010624 s 0.000010016 s 1.06
slicing / Jax / cuda / Forward 0.00001088 s 0.000010368 s 1.05
slicing / HLOOpt / cuda / Forward 0.000010752 s 0.000009952 s 1.08
slicing / PartOpt / cuda / Forward 0.000010464 s 0.000009376 s 1.12
slicing / IPartOpt / cuda / Forward 0.000010399 s 0.000009696 s 1.07
slicing / DefOpt / cuda / Forward 0.000010464 s 0.000010112 s 1.03
slicing / IDefOpt / cuda / Forward 0.000010944 s 0.000009728 s 1.13
slicing / JaXPipe / cuda / PreRev 0.0000104 s 0.00000944 s 1.10
slicing / JaXPipe / cuda / PostRev 0.0000104 s 0.000010016 s 1.04
slicing / JaXPipe / cuda / BothRev 0.000010528 s 0.000009887 s 1.06
slicing / Jax / cuda / BothRev 0.000010784 s 0.00001008 s 1.07
slicing / HLOOpt / cuda / PreRev 0.000010784 s 0.000010016 s 1.08
slicing / HLOOpt / cuda / PostRev 0.000010432 s 0.000009536 s 1.09
slicing / HLOOpt / cuda / BothRev 0.000010208 s 0.0000096 s 1.06
slicing / PartOpt / cuda / PreRev 0.000010783 s 0.000009344 s 1.15
slicing / PartOpt / cuda / PostRev 0.000010848 s 0.000010015 s 1.08
slicing / PartOpt / cuda / BothRev 0.00001056 s 0.000009728 s 1.09
slicing / IPartOpt / cuda / PreRev 0.000010177 s 0.000009664 s 1.05
slicing / IPartOpt / cuda / PostRev 0.000010592 s 0.00000816 s 1.30
slicing / IPartOpt / cuda / BothRev 0.00001072 s 0.00001008 s 1.06
slicing / DefOpt / cuda / PreRev 0.000010495 s 0.000010112 s 1.04
slicing / DefOpt / cuda / PostRev 0.000010304 s 0.00000976 s 1.06
slicing / DefOpt / cuda / BothRev 0.000010752 s 0.000010175 s 1.06
slicing / IDefOpt / cuda / PreRev 0.000010752 s 0.000010112 s 1.06
slicing / IDefOpt / cuda / PostRev 0.0000104 s 0.000009408 s 1.11
slicing / IDefOpt / cuda / BothRev 0.000010656 s 0.000009312000000000002 s 1.14
slicing / JaXPipe / tpu / Primal 0.00000102365 s 0.000001027725 s 1.00
slicing / Jax / tpu / Primal 9.6705e-7 s 9.8245e-7 s 0.98
slicing / HLOOpt / tpu / Primal 0.0000010245250000000002 s 0.00000103545 s 0.99
slicing / PartOpt / tpu / Primal 9.72575e-7 s 9.91825e-7 s 0.98
slicing / IPartOpt / tpu / Primal 0.0000010268 s 0.0000010254000000000002 s 1.00
slicing / DefOpt / tpu / Primal 9.7675e-7 s 9.8095e-7 s 1.00
slicing / IDefOpt / tpu / Primal 0.0000010264 s 0.0000010292 s 1.00
slicing / JaXPipe / tpu / Forward 0.000001414675 s 0.000001413875 s 1.00
slicing / Jax / tpu / Forward 0.0000014803250000000003 s 0.000001478675 s 1.00
slicing / HLOOpt / tpu / Forward 0.0000015219750000000002 s 0.00000151735 s 1.00
slicing / PartOpt / tpu / Forward 0.0000014934249999999998 s 0.00000149485 s 1.00
slicing / IPartOpt / tpu / Forward 0.0000015191750000000002 s 0.0000015182499999999998 s 1.00
slicing / DefOpt / tpu / Forward 0.0000014955 s 0.000001494525 s 1.00
slicing / IDefOpt / tpu / Forward 0.000001528625 s 0.00000151995 s 1.01
slicing / JaXPipe / tpu / PreRev 0.000002567175 s 0.0000025668750000000003 s 1.00
slicing / JaXPipe / tpu / PostRev 0.0000025280750000000004 s 0.000002516825 s 1.00
slicing / JaXPipe / tpu / BothRev 0.0000025815 s 0.000002583825 s 1.00
slicing / Jax / tpu / BothRev 0.0000025321500000000004 s 0.00000253285 s 1.00
slicing / HLOOpt / tpu / PreRev 0.000002584525 s 0.00000260035 s 0.99
slicing / HLOOpt / tpu / PostRev 0.000002542625 s 0.000002543725 s 1.00
slicing / HLOOpt / tpu / BothRev 0.00000258025 s 0.0000025989250000000003 s 0.99
slicing / PartOpt / tpu / PreRev 0.0000025315250000000003 s 0.000002535425 s 1.00
slicing / PartOpt / tpu / PostRev 0.0000025793 s 0.000002583175 s 1.00
slicing / PartOpt / tpu / BothRev 0.0000025309 s 0.0000025365 s 1.00
slicing / IPartOpt / tpu / PreRev 0.000002582375 s 0.0000025852000000000003 s 1.00
slicing / IPartOpt / tpu / PostRev 0.000002532925 s 0.000002545375 s 1.00
slicing / IPartOpt / tpu / BothRev 0.0000025836750000000003 s 0.0000025916 s 1.00
slicing / DefOpt / tpu / PreRev 0.0000025380750000000003 s 0.000002536925 s 1.00
slicing / DefOpt / tpu / PostRev 0.000002583 s 0.0000025939250000000003 s 1.00
slicing / DefOpt / tpu / BothRev 0.000002542725 s 0.00000253795 s 1.00
slicing / IDefOpt / tpu / PreRev 0.000002587175 s 0.00000259 s 1.00
slicing / IDefOpt / tpu / PostRev 0.0000025349 s 0.0000025463 s 1.00
slicing / IDefOpt / tpu / BothRev 0.000002582975 s 0.00000258445 s 1.00
slicing / JaXPipe / cpu / Primal 0.000012895 s 0.00000682609998875705 s 1.89
slicing / Jax / cpu / Primal 0.000012435 s 0.000006204580004123273 s 2.00
slicing / HLOOpt / cpu / Primal 0.000012615 s 0.000006755100002919789 s 1.87
slicing / PartOpt / cpu / Primal 0.0000124 s 0.000006392579962266609 s 1.94
slicing / IPartOpt / cpu / Primal 0.00001242 s 0.000006374380027409643 s 1.95
slicing / DefOpt / cpu / Primal 0.000012461 s 0.000006200300022101146 s 2.01
slicing / IDefOpt / cpu / Primal 0.000012462 s 0.000006124159972387133 s 2.03
slicing / JaXPipe / cpu / Forward 0.000016857 s 0.000009675040009824445 s 1.74
slicing / Jax / cpu / Forward 0.000016476999999999998 s 0.000009428260009372025 s 1.75
slicing / HLOOpt / cpu / Forward 0.000016965 s 0.000010320419996787677 s 1.64
slicing / PartOpt / cpu / Forward 0.000016549 s 0.000009372380027343752 s 1.77
slicing / IPartOpt / cpu / Forward 0.00001643 s 0.000010439299985591788 s 1.57
slicing / DefOpt / cpu / Forward 0.000016427 s 0.000009475980032220832 s 1.73
slicing / IDefOpt / cpu / Forward 0.000016459999999999998 s 0.000009360399963043164 s 1.76
slicing / JaXPipe / cpu / PreRev 0.000017117 s 0.00000984418000371079 s 1.74
slicing / JaXPipe / cpu / PostRev 0.000017257 s 0.000010596880019875244 s 1.63
slicing / JaXPipe / cpu / BothRev 0.000016649 s 0.00001076662000741635 s 1.55
slicing / Jax / cpu / BothRev 0.000016767 s 0.000010217180006293348 s 1.64
slicing / HLOOpt / cpu / PreRev 0.000017168 s 0.000010931720034932368 s 1.57
slicing / HLOOpt / cpu / PostRev 0.000017434000000000003 s 0.000016274080007860902 s 1.07
slicing / HLOOpt / cpu / BothRev 0.000016991 s 0.000010077980032292544 s 1.69
slicing / PartOpt / cpu / PreRev 0.000017239000000000002 s 0.000009662459979153936 s 1.78
slicing / PartOpt / cpu / PostRev 0.000017718999999999998 s 0.000010176299974773429 s 1.74
slicing / PartOpt / cpu / BothRev 0.000017001 s 0.000010779640024338731 s 1.58
slicing / IPartOpt / cpu / PreRev 0.000017252 s 0.000009989239997594267 s 1.73
slicing / IPartOpt / cpu / PostRev 0.000017371 s 0.00001034255996273714 s 1.68
slicing / IPartOpt / cpu / BothRev 0.000017089 s 0.00001012631999401492 s 1.69
slicing / DefOpt / cpu / PreRev 0.000017158 s 0.000010393999991720193 s 1.65
slicing / DefOpt / cpu / PostRev 0.000017187 s 0.000010519240031499068 s 1.63
slicing / DefOpt / cpu / BothRev 0.000016986 s 0.000010008200033553294 s 1.70
slicing / IDefOpt / cpu / PreRev 0.000017729000000000003 s 0.000009962880012608366 s 1.78
slicing / IDefOpt / cpu / PostRev 0.000017432 s 0.000011089439976785795 s 1.57
slicing / IDefOpt / cpu / BothRev 0.000017185 s 0.000010662259983291731 s 1.61
slicing / JaXPipe / cpu / Primal 0.000008 s 0.00000682609998875705 s 1.17
slicing / Jax / cpu / Primal 0.000008 s 0.000006204580004123273 s 1.29
slicing / HLOOpt / cpu / Primal 0.000008999999999999999 s 0.000006755100002919789 s 1.33
slicing / PartOpt / cpu / Primal 0.000008999999999999999 s 0.000006392579962266609 s 1.41
slicing / IPartOpt / cpu / Primal 0.000008999999999999999 s 0.000006374380027409643 s 1.41
slicing / DefOpt / cpu / Primal 0.000008999999999999999 s 0.000006200300022101146 s 1.45
slicing / IDefOpt / cpu / Primal 0.000008999999999999999 s 0.000006124159972387133 s 1.47
slicing / JaXPipe / cpu / Forward 0.000012 s 0.000009675040009824445 s 1.24
slicing / Jax / cpu / Forward 0.000012 s 0.000009428260009372025 s 1.27
slicing / HLOOpt / cpu / Forward 0.000012 s 0.000010320419996787677 s 1.16
slicing / PartOpt / cpu / Forward 0.000012 s 0.000009372380027343752 s 1.28
slicing / IPartOpt / cpu / Forward 0.000011 s 0.000010439299985591788 s 1.05
slicing / DefOpt / cpu / Forward 0.000011 s 0.000009475980032220832 s 1.16
slicing / IDefOpt / cpu / Forward 0.000012 s 0.000009360399963043164 s 1.28
slicing / JaXPipe / cpu / PreRev 0.000012 s 0.00000984418000371079 s 1.22
slicing / JaXPipe / cpu / PostRev 0.000012 s 0.000010596880019875244 s 1.13
slicing / JaXPipe / cpu / BothRev 0.000012 s 0.00001076662000741635 s 1.11
slicing / Jax / cpu / BothRev 0.000012 s 0.000010217180006293348 s 1.17
slicing / HLOOpt / cpu / PreRev 0.000039 s 0.000010931720034932368 s 3.57
slicing / HLOOpt / cpu / PostRev 0.000012 s 0.000016274080007860902 s 0.74
slicing / HLOOpt / cpu / BothRev 0.000012 s 0.000010077980032292544 s 1.19
slicing / PartOpt / cpu / PreRev 0.000012 s 0.000009662459979153936 s 1.24
slicing / PartOpt / cpu / PostRev 0.000012 s 0.000010176299974773429 s 1.18
slicing / PartOpt / cpu / BothRev 0.000012 s 0.000010779640024338731 s 1.11
slicing / IPartOpt / cpu / PreRev 0.000012 s 0.000009989239997594267 s 1.20
slicing / IPartOpt / cpu / PostRev 0.000012 s 0.00001034255996273714 s 1.16
slicing / IPartOpt / cpu / BothRev 0.000013 s 0.00001012631999401492 s 1.28
slicing / DefOpt / cpu / PreRev 0.000012 s 0.000010393999991720193 s 1.15
slicing / DefOpt / cpu / PostRev 0.000013 s 0.000010519240031499068 s 1.24
slicing / DefOpt / cpu / BothRev 0.000012 s 0.000010008200033553294 s 1.20
slicing / IDefOpt / cpu / PreRev 0.000012 s 0.000009962880012608366 s 1.20
slicing / IDefOpt / cpu / PostRev 0.000012 s 0.000011089439976785795 s 1.08
slicing / IDefOpt / cpu / BothRev 0.000014 s 0.000010662259983291731 s 1.31
sum / JaXPipe / cpu / Primal 0.000009787039989532786 s 0.000007817779969627737 s 1.25
sum / Jax / cpu / Primal 0.000009181020013784293 s 0.000008294660037790891 s 1.11
sum / HLOOpt / cpu / Primal 0.000009565380023559556 s 0.000008531580006092554 s 1.12
sum / PartOpt / cpu / Primal 0.000009332280023954807 s 0.000008095400007732678 s 1.15
sum / IPartOpt / cpu / Primal 0.000009462899979553186 s 0.000008019259948923719 s 1.18
sum / DefOpt / cpu / Primal 0.000009628019979572857 s 0.00000787671999205486 s 1.22
sum / IDefOpt / cpu / Primal 0.000008968560014182003 s 0.00000840821997371677 s 1.07
sum / JaXPipe / cpu / Forward 0.000013537200047721851 s 0.00001177179999103828 s 1.15
sum / Jax / cpu / Forward 0.000012942639996254002 s 0.000011495280041344812 s 1.13
sum / HLOOpt / cpu / Forward 0.000014053240020075464 s 0.000012031799979013158 s 1.17
sum / PartOpt / cpu / Forward 0.000012780080014636042 s 0.0000115612999888981 s 1.11
sum / IPartOpt / cpu / Forward 0.000012973879993296578 s 0.0000120436600263929 s 1.08
sum / DefOpt / cpu / Forward 0.000013048120017629117 s 0.000011576019978747354 s 1.13
sum / IDefOpt / cpu / Forward 0.000013178179979149715 s 0.000011552179994396284 s 1.14
sum / JaXPipe / cpu / PreRev 0.00001330345997303084 s 0.00001103183999475732 s 1.21
sum / JaXPipe / cpu / PostRev 0.000012805660016965705 s 0.00001130409999859694 s 1.13
sum / JaXPipe / cpu / BothRev 0.000012738639998133297 s 0.000011148939956910907 s 1.14
sum / Jax / cpu / BothRev 0.000012351539999144734 s 0.000011459779998403972 s 1.08
sum / HLOOpt / cpu / PreRev 0.000012515119960880838 s 0.00001164994000646402 s 1.07
sum / HLOOpt / cpu / PostRev 0.0000141828000323585 s 0.000012972059957974124 s 1.09
sum / HLOOpt / cpu / BothRev 0.00001246169998012192 s 0.000010897119991568616 s 1.14
sum / PartOpt / cpu / PreRev 0.000012263099988558678 s 0.000010952600014206836 s 1.12
sum / PartOpt / cpu / PostRev 0.00001231165997523931 s 0.000011012459972334908 s 1.12
sum / PartOpt / cpu / BothRev 0.00001213622000250325 s 0.000011296679977021995 s 1.07
sum / IPartOpt / cpu / PreRev 0.000012534339975900368 s 0.000010750420005933848 s 1.17
sum / IPartOpt / cpu / PostRev 0.00001272238001547521 s 0.000011626079976849725 s 1.09
sum / IPartOpt / cpu / BothRev 0.00001237890004631481 s 0.000011353899972164071 s 1.09
sum / DefOpt / cpu / PreRev 0.000012335860055827652 s 0.000011091079986726982 s 1.11
sum / DefOpt / cpu / PostRev 0.000012521680018835468 s 0.000011270859959040536 s 1.11
sum / DefOpt / cpu / BothRev 0.000012606420023075773 s 0.000010997040017173276 s 1.15
sum / IDefOpt / cpu / PreRev 0.00001241926001057436 s 0.00001100987998142955 s 1.13
sum / IDefOpt / cpu / PostRev 0.000013184599993110169 s 0.000010701340033847374 s 1.23
sum / IDefOpt / cpu / BothRev 0.000013008919995627368 s 0.000011028840035578468 s 1.18
sum / JaXPipe / cuda / Primal 0.000002463 s 0.00000208 s 1.18
sum / Jax / cuda / Primal 0.000002463 s 0.000002079 s 1.18
sum / HLOOpt / cuda / Primal 0.000002463 s 0.00000208 s 1.18
sum / PartOpt / cuda / Primal 0.000002464 s 0.00000208 s 1.18
sum / IPartOpt / cuda / Primal 0.000002463 s 0.000002079 s 1.18
sum / DefOpt / cuda / Primal 0.000002463 s 0.00000208 s 1.18
sum / IDefOpt / cuda / Primal 0.000002463 s 0.000002079 s 1.18
sum / JaXPipe / cuda / Forward 0.00001104 s 0.000010175 s 1.09
sum / Jax / cuda / Forward 0.000010816 s 0.000010465 s 1.03
sum / HLOOpt / cuda / Forward 0.000010879 s 0.000010144 s 1.07
sum / PartOpt / cuda / Forward 0.00001104 s 0.000010336 s 1.07
sum / IPartOpt / cuda / Forward 0.000010784 s 0.000010336 s 1.04
sum / DefOpt / cuda / Forward 0.000010976 s 0.000010208 s 1.08
sum / IDefOpt / cuda / Forward 0.000011072 s 0.000009567 s 1.16
sum / JaXPipe / cuda / PreRev 0.0000104 s 0.000009408 s 1.11
sum / JaXPipe / cuda / PostRev 0.000010433 s 0.00000992 s 1.05
sum / JaXPipe / cuda / BothRev 0.000010272 s 0.000010017 s 1.03
sum / Jax / cuda / BothRev 0.000010208 s 0.000009792 s 1.04
sum / HLOOpt / cuda / PreRev 0.000010592 s 0.000009824 s 1.08
sum / HLOOpt / cuda / PostRev 0.000010687 s 0.000009856 s 1.08
sum / HLOOpt / cuda / BothRev 0.000010303 s 0.000009855 s 1.05
sum / PartOpt / cuda / PreRev 0.000010591 s 0.000010016 s 1.06
sum / PartOpt / cuda / PostRev 0.000010304 s 0.000009824 s 1.05
sum / PartOpt / cuda / BothRev 0.00001056 s 0.000010175 s 1.04
sum / IPartOpt / cuda / PreRev 0.000010527 s 0.000009952 s 1.06
sum / IPartOpt / cuda / PostRev 0.000010368 s 0.0000096 s 1.08
sum / IPartOpt / cuda / BothRev 0.000010464 s 0.000009727 s 1.08
sum / DefOpt / cuda / PreRev 0.000010464 s 0.000010144 s 1.03
sum / DefOpt / cuda / PostRev 0.000010272 s 0.0000096 s 1.07
sum / DefOpt / cuda / BothRev 0.000010624 s 0.00001008 s 1.05
sum / IDefOpt / cuda / PreRev 0.00001056 s 0.000009952 s 1.06
sum / IDefOpt / cuda / PostRev 0.000010176 s 0.000009632 s 1.06
sum / IDefOpt / cuda / BothRev 0.000010816 s 0.000009889 s 1.09
sum / JaXPipe / tpu / Primal 5.09825e-7 s 5.103250000000001e-7 s 1.00
sum / Jax / tpu / Primal 5.47175e-7 s 5.4685e-7 s 1.00
sum / HLOOpt / tpu / Primal 5.100500000000001e-7 s 5.10175e-7 s 1.00
sum / PartOpt / tpu / Primal 5.468e-7 s 5.469999999999999e-7 s 1.00
sum / IPartOpt / tpu / Primal 5.10075e-7 s 5.100500000000001e-7 s 1.00
sum / DefOpt / tpu / Primal 5.47075e-7 s 5.471e-7 s 1.00
sum / IDefOpt / tpu / Primal 5.10525e-7 s 5.10625e-7 s 1.00
sum / JaXPipe / tpu / Forward 0.000001550625 s 0.00000155425 s 1.00
sum / Jax / tpu / Forward 0.000001496575 s 0.0000015014 s 1.00
sum / HLOOpt / tpu / Forward 0.000001528175 s 0.000001529875 s 1.00
sum / PartOpt / tpu / Forward 0.00000149265 s 0.000001500725 s 0.99
sum / IPartOpt / tpu / Forward 0.0000015278 s 0.0000015351999999999995 s 1.00
sum / DefOpt / tpu / Forward 0.0000014927000000000003 s 0.00000149645 s 1.00
sum / IDefOpt / tpu / Forward 0.000001528475 s 0.0000015358 s 1.00
sum / JaXPipe / tpu / PreRev 0.0000010529 s 0.000001048875 s 1.00
sum / JaXPipe / tpu / PostRev 0.0000010895499999999998 s 0.00000108725 s 1.00
sum / JaXPipe / tpu / BothRev 0.00000105315 s 0.0000010524 s 1.00
sum / Jax / tpu / BothRev 0.00000109385 s 0.0000010923 s 1.00
sum / HLOOpt / tpu / PreRev 0.0000010565 s 0.000001055325 s 1.00
sum / HLOOpt / tpu / PostRev 0.000001096275 s 0.000001088125 s 1.01
sum / HLOOpt / tpu / BothRev 0.000001057025 s 0.000001053375 s 1.00
sum / PartOpt / tpu / PreRev 0.00000110015 s 0.000001087875 s 1.01
sum / PartOpt / tpu / PostRev 0.000001054525 s 0.00000104855 s 1.01
sum / PartOpt / tpu / BothRev 0.000001097075 s 0.000001081975 s 1.01
sum / IPartOpt / tpu / PreRev 0.0000010613250000000002 s 0.000001047825 s 1.01
sum / IPartOpt / tpu / PostRev 0.000001092525 s 0.0000010825 s 1.01
sum / IPartOpt / tpu / BothRev 0.000001054425 s 0.000001049375 s 1.00
sum / DefOpt / tpu / PreRev 0.000001092725 s 0.0000010874250000000002 s 1.00
sum / DefOpt / tpu / PostRev 0.0000010587 s 0.000001048975 s 1.01
sum / DefOpt / tpu / BothRev 0.00000109205 s 0.00000108505 s 1.01
sum / IDefOpt / tpu / PreRev 0.0000010528500000000002 s 0.0000010591000000000002 s 0.99
sum / IDefOpt / tpu / PostRev 0.000001092375 s 0.00000108325 s 1.01
sum / IDefOpt / tpu / BothRev 0.00000106675 s 0.0000010499 s 1.02
sum / JaXPipe / cpu / Primal 0.000014385 s 0.000007817779969627737 s 1.84
sum / Jax / cpu / Primal 0.000014307 s 0.000008294660037790891 s 1.72
sum / HLOOpt / cpu / Primal 0.000014176 s 0.000008531580006092554 s 1.66
sum / PartOpt / cpu / Primal 0.000014287 s 0.000008095400007732678 s 1.76
sum / IPartOpt / cpu / Primal 0.000014674 s 0.000008019259948923719 s 1.83
sum / DefOpt / cpu / Primal 0.000014688 s 0.00000787671999205486 s 1.86
sum / IDefOpt / cpu / Primal 0.000014558 s 0.00000840821997371677 s 1.73
sum / JaXPipe / cpu / Forward 0.000020048 s 0.00001177179999103828 s 1.70
sum / Jax / cpu / Forward 0.000019581 s 0.000011495280041344812 s 1.70
sum / HLOOpt / cpu / Forward 0.000019464 s 0.000012031799979013158 s 1.62
sum / PartOpt / cpu / Forward 0.000019664 s 0.0000115612999888981 s 1.70
sum / IPartOpt / cpu / Forward 0.000019925 s 0.0000120436600263929 s 1.65
sum / DefOpt / cpu / Forward 0.000019991 s 0.000011576019978747354 s 1.73
sum / IDefOpt / cpu / Forward 0.000020109 s 0.000011552179994396284 s 1.74
sum / JaXPipe / cpu / PreRev 0.000018629 s 0.00001103183999475732 s 1.69
sum / JaXPipe / cpu / PostRev 0.000018548 s 0.00001130409999859694 s 1.64
sum / JaXPipe / cpu / BothRev 0.000018648 s 0.000011148939956910907 s 1.67
sum / Jax / cpu / BothRev 0.000018709 s 0.000011459779998403972 s 1.63
sum / HLOOpt / cpu / PreRev 0.000018479 s 0.00001164994000646402 s 1.59
sum / HLOOpt / cpu / PostRev 0.000018651 s 0.000012972059957974124 s 1.44
sum / HLOOpt / cpu / BothRev 0.000018781 s 0.000010897119991568616 s 1.72
sum / PartOpt / cpu / PreRev 0.000018582000000000003 s 0.000010952600014206836 s 1.70
sum / PartOpt / cpu / PostRev 0.00001822 s 0.000011012459972334908 s 1.65
sum / PartOpt / cpu / BothRev 0.000018696 s 0.000011296679977021995 s 1.65
sum / IPartOpt / cpu / PreRev 0.000018935 s 0.000010750420005933848 s 1.76
sum / IPartOpt / cpu / PostRev 0.00001855 s 0.000011626079976849725 s 1.60
sum / IPartOpt / cpu / BothRev 0.00001867 s 0.000011353899972164071 s 1.64
sum / DefOpt / cpu / PreRev 0.000018435 s 0.000011091079986726982 s 1.66
sum / DefOpt / cpu / PostRev 0.000018842 s 0.000011270859959040536 s 1.67
sum / DefOpt / cpu / BothRev 0.00001855 s 0.000010997040017173276 s 1.69
sum / IDefOpt / cpu / PreRev 0.000018652 s 0.00001100987998142955 s 1.69
sum / IDefOpt / cpu / PostRev 0.000018535 s 0.000010701340033847374 s 1.73
sum / IDefOpt / cpu / BothRev 0.000018499 s 0.000011028840035578468 s 1.68
sum / JaXPipe / cpu / Primal 0.000034 s 0.000007817779969627737 s 4.35
sum / Jax / cpu / Primal 0.00001 s 0.000008294660037790891 s 1.21
sum / HLOOpt / cpu / Primal 0.00001 s 0.000008531580006092554 s 1.17
sum / PartOpt / cpu / Primal 0.000034 s 0.000008095400007732678 s 4.20
sum / IPartOpt / cpu / Primal 0.00001 s 0.000008019259948923719 s 1.25
sum / DefOpt / cpu / Primal 0.00001 s 0.00000787671999205486 s 1.27
sum / IDefOpt / cpu / Primal 0.00001 s 0.00000840821997371677 s 1.19
sum / JaXPipe / cpu / Forward 0.000014 s 0.00001177179999103828 s 1.19
sum / Jax / cpu / Forward 0.000014 s 0.000011495280041344812 s 1.22
sum / HLOOpt / cpu / Forward 0.000015 s 0.000012031799979013158 s 1.25
sum / PartOpt / cpu / Forward 0.000014 s 0.0000115612999888981 s 1.21
sum / IPartOpt / cpu / Forward 0.000045 s 0.0000120436600263929 s 3.74
sum / DefOpt / cpu / Forward 0.000014 s 0.000011576019978747354 s 1.21
sum / IDefOpt / cpu / Forward 0.000014 s 0.000011552179994396284 s 1.21
sum / JaXPipe / cpu / PreRev 0.000013 s 0.00001103183999475732 s 1.18
sum / JaXPipe / cpu / PostRev 0.000013 s 0.00001130409999859694 s 1.15
sum / JaXPipe / cpu / BothRev 0.000013 s 0.000011148939956910907 s 1.17
sum / Jax / cpu / BothRev 0.000013 s 0.000011459779998403972 s 1.13
sum / HLOOpt / cpu / PreRev 0.000013 s 0.00001164994000646402 s 1.12
sum / HLOOpt / cpu / PostRev 0.000013 s 0.000012972059957974124 s 1.00
sum / HLOOpt / cpu / BothRev 0.000014 s 0.000010897119991568616 s 1.28
sum / PartOpt / cpu / PreRev 0.000013 s 0.000010952600014206836 s 1.19
sum / PartOpt / cpu / PostRev 0.000013 s 0.000011012459972334908 s 1.18
sum / PartOpt / cpu / BothRev 0.000014 s 0.000011296679977021995 s 1.24
sum / IPartOpt / cpu / PreRev 0.000014 s 0.000010750420005933848 s 1.30
sum / IPartOpt / cpu / PostRev 0.000014 s 0.000011626079976849725 s 1.20
sum / IPartOpt / cpu / BothRev 0.000013 s 0.000011353899972164071 s 1.14
sum / DefOpt / cpu / PreRev 0.000013 s 0.000011091079986726982 s 1.17
sum / DefOpt / cpu / PostRev 0.000013 s 0.000011270859959040536 s 1.15
sum / DefOpt / cpu / BothRev 0.000013 s 0.000010997040017173276 s 1.18
sum / IDefOpt / cpu / PreRev 0.000014 s 0.00001100987998142955 s 1.27
sum / IDefOpt / cpu / PostRev 0.000013 s 0.000010701340033847374 s 1.21
sum / IDefOpt / cpu / BothRev 0.000013 s 0.000011028840035578468 s 1.18
value_and_grad / JaXPipe / cpu / Primal 0.000015996520023691118 s 0.000015389159980259137 s 1.04
value_and_grad / Jax / cpu / Primal 0.000015919000015856 s 0.000014474340032393227 s 1.10
value_and_grad / HLOOpt / cpu / Primal 0.00001563824002005276 s 0.0000136315000236209 s 1.15
value_and_grad / PartOpt / cpu / Primal 0.00001516711998192477 s 0.000013677960005225032 s 1.11
value_and_grad / IPartOpt / cpu / Primal 0.000015181440030573869 s 0.000013812460019835271 s 1.10
value_and_grad / DefOpt / cpu / Primal 0.000016178400028366013 s 0.00001391469996633532 s 1.16
value_and_grad / IDefOpt / cpu / Primal 0.000015472060040337967 s 0.000013700100025744177 s 1.13
value_and_grad / JaXPipe / cuda / Primal 0.00003344 s 0.000034048 s 0.98
value_and_grad / Jax / cuda / Primal 0.000033119999999999995 s 0.000033023 s 1.00
value_and_grad / HLOOpt / cuda / Primal 0.000033536000000000006 s 0.000033216 s 1.01
value_and_grad / PartOpt / cuda / Primal 0.000033535 s 0.000033344 s 1.01
value_and_grad / IPartOpt / cuda / Primal 0.000032288 s 0.000033184 s 0.97
value_and_grad / DefOpt / cuda / Primal 0.000033343 s 0.000032447 s 1.03
value_and_grad / IDefOpt / cuda / Primal 0.000033471 s 0.000032639000000000004 s 1.03
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 0.00002286 s 0.000015389159980259137 s 1.49
value_and_grad / Jax / cpu / Primal 0.000022535 s 0.000014474340032393227 s 1.56
value_and_grad / HLOOpt / cpu / Primal 0.000022615 s 0.0000136315000236209 s 1.66
value_and_grad / PartOpt / cpu / Primal 0.000022374000000000003 s 0.000013677960005225032 s 1.64
value_and_grad / IPartOpt / cpu / Primal 0.000022623 s 0.000013812460019835271 s 1.64
value_and_grad / DefOpt / cpu / Primal 0.000022519 s 0.00001391469996633532 s 1.62
value_and_grad / IDefOpt / cpu / Primal 0.000022779 s 0.000013700100025744177 s 1.66
value_and_grad / JaXPipe / cpu / Primal 0.000017 s 0.000015389159980259137 s 1.10
value_and_grad / Jax / cpu / Primal 0.000016 s 0.000014474340032393227 s 1.11
value_and_grad / HLOOpt / cpu / Primal 0.000016 s 0.0000136315000236209 s 1.17
value_and_grad / PartOpt / cpu / Primal 0.000017 s 0.000013677960005225032 s 1.24
value_and_grad / IPartOpt / cpu / Primal 0.000017 s 0.000013812460019835271 s 1.23
value_and_grad / DefOpt / cpu / Primal 0.000016 s 0.00001391469996633532 s 1.15
value_and_grad / IDefOpt / cpu / Primal 0.000017 s 0.000013700100025744177 s 1.24
jaxmd20 / JaXPipe / cuda / Primal 0.001442488 s 0.0013876729999999 s 1.04
jaxmd20 / Jax / cuda / Primal 0.001475224 s 0.001420247 s 1.04
jaxmd20 / HLOOpt / cuda / Primal 0.001353016 s 0.0013118 s 1.03
jaxmd20 / PartOpt / cuda / Primal 0.0013350649999999 s 0.001366872 s 0.98
jaxmd20 / IPartOpt / cuda / Primal 0.00136764 s 0.001347288 s 1.02
jaxmd20 / DefOpt / cuda / Primal 0.0009454019999999 s 0.000914459 s 1.03
jaxmd20 / IDefOpt / cuda / Primal 0.000960603 s 0.000943515 s 1.02
jaxmd20 / JaXPipe / cuda / Forward 0.001625431 s 0.0015452719999999 s 1.05
jaxmd20 / Jax / cuda / Forward 0.00185743 s 0.001751927 s 1.06
jaxmd20 / HLOOpt / cuda / Forward 0.001703478 s 0.001632215 s 1.04
jaxmd20 / PartOpt / cuda / Forward 0.0017094959999999 s 0.0016309039999999 s 1.05
jaxmd20 / IPartOpt / cuda / Forward 0.001707863 s 0.001628055 s 1.05
jaxmd20 / DefOpt / cuda / Forward 0.001697175 s 0.0016312569999999 s 1.04
jaxmd20 / IDefOpt / cuda / Forward 0.001738358 s 0.001616023 s 1.08
jaxmd20 / JaXPipe / cuda / PreRev 0.002750321 s 0.00265237 s 1.04
jaxmd20 / JaXPipe / cuda / PostRev 0.005446531 s 0.005288228 s 1.03
jaxmd20 / JaXPipe / cuda / BothRev 0.002736242 s 0.0026615859999999 s 1.03
jaxmd20 / Jax / cuda / BothRev 0.005424002 s 0.005290403 s 1.03
jaxmd20 / HLOOpt / cuda / PreRev 0.002837168 s 0.002720944 s 1.04
jaxmd20 / HLOOpt / cuda / PostRev 0.005458242 s 0.0052818579999999 s 1.03
jaxmd20 / HLOOpt / cuda / BothRev 0.002780752 s 0.002700274 s 1.03
jaxmd20 / PartOpt / cuda / PreRev 0.002894478 s 0.00279736 s 1.03
jaxmd20 / PartOpt / cuda / PostRev 0.00557872 s 0.005337155 s 1.05
jaxmd20 / PartOpt / cuda / BothRev 0.002836208 s 0.002747601 s 1.03
jaxmd20 / IPartOpt / cuda / PreRev 0.002889297 s 0.0027941919999999 s 1.03
jaxmd20 / IPartOpt / cuda / PostRev 0.005748833 s 0.005379843 s 1.07
jaxmd20 / IPartOpt / cuda / BothRev 0.0028264809999999 s 0.002746865 s 1.03
jaxmd20 / DefOpt / cuda / PreRev 0.002914511 s 0.002808209 s 1.04
jaxmd20 / DefOpt / cuda / PostRev 0.002827311 s 0.002760144 s 1.02
jaxmd20 / DefOpt / cuda / BothRev 0.002818032 s 0.0028778389999999 s 0.98
jaxmd20 / IDefOpt / cuda / PreRev 0.002916207 s 0.002798704 s 1.04
jaxmd20 / IDefOpt / cuda / PostRev 0.002343603 s 0.002310963 s 1.01
jaxmd20 / IDefOpt / cuda / BothRev 0.002817712 s 0.002781455 s 1.01
jaxmd20 / JaXPipe / tpu / Primal 0.009286348125 s 0.009270928125 s 1.00
jaxmd20 / Jax / tpu / Primal 0.009267943125 s 0.009279299375 s 1.00
jaxmd20 / HLOOpt / tpu / Primal 0.00915478375 s 0.009152014375 s 1.00
jaxmd20 / PartOpt / tpu / Primal 0.009196790625 s 0.009205471875 s 1.00
jaxmd20 / IPartOpt / tpu / Primal 0.009199449375 s 0.009201760625 s 1.00
jaxmd20 / DefOpt / tpu / Primal 0.0087965324999999 s 0.008806850625 s 1.00
jaxmd20 / IDefOpt / tpu / Primal 0.008701159375 s 0.008703381875 s 1.00
jaxmd20 / JaXPipe / tpu / Forward 0.017414551875 s 0.017405436875 s 1.00
jaxmd20 / Jax / tpu / Forward 0.01873648375 s 0.0187342325 s 1.00
jaxmd20 / HLOOpt / tpu / Forward 0.0174016375 s 0.0173937212499999 s 1.00
jaxmd20 / PartOpt / tpu / Forward 0.017413895625 s 0.01741813125 s 1.00
jaxmd20 / IPartOpt / tpu / Forward 0.017411084375 s 0.017407166875 s 1.00
jaxmd20 / DefOpt / tpu / Forward 0.0174142825 s 0.017417351875 s 1.00
jaxmd20 / IDefOpt / tpu / Forward 0.01741183875 s 0.017409986875 s 1.00
jaxmd20 / JaXPipe / tpu / PreRev 0.025443471875 s 0.025470619375 s 1.00
jaxmd20 / JaXPipe / tpu / PostRev 0.021870595625 s 0.0218776125 s 1.00
jaxmd20 / JaXPipe / tpu / BothRev 0.02547214375 s 0.0254671168749999 s 1.00
jaxmd20 / Jax / tpu / BothRev 0.02187530625 s 0.0218587225 s 1.00
jaxmd20 / HLOOpt / tpu / PreRev 0.025579310625 s 0.02558389375 s 1.00
jaxmd20 / HLOOpt / tpu / PostRev 0.020719105625 s 0.0207120575 s 1.00
jaxmd20 / HLOOpt / tpu / BothRev 0.02568798 s 0.025689004375 s 1.00
jaxmd20 / PartOpt / tpu / PreRev 0.0254884156249999 s 0.025454488125 s 1.00
jaxmd20 / PartOpt / tpu / PostRev 0.021536606875 s 0.0215167231249999 s 1.00
jaxmd20 / PartOpt / tpu / BothRev 0.0255701725 s 0.025557858125 s 1.00
jaxmd20 / IPartOpt / tpu / PreRev 0.025473325625 s 0.02547195 s 1.00
jaxmd20 / IPartOpt / tpu / PostRev 0.021518866875 s 0.021247565625 s 1.01
jaxmd20 / IPartOpt / tpu / BothRev 0.02556476625 s 0.02556671625 s 1.00
jaxmd20 / DefOpt / tpu / PreRev 0.025489294375 s 0.0254547112499999 s 1.00
jaxmd20 / DefOpt / tpu / PostRev 0.0188234925 s 0.01882540375 s 1.00
jaxmd20 / DefOpt / tpu / BothRev 0.025571261875 s 0.025554230625 s 1.00
jaxmd20 / IDefOpt / tpu / PreRev 0.025474839375 s 0.02547245625 s 1.00
jaxmd20 / IDefOpt / tpu / PostRev 0.018340658125 s 0.018312810625 s 1.00
jaxmd20 / IDefOpt / tpu / BothRev 0.025570789375 s 0.02556315625 s 1.00
jaxmd40 / JaXPipe / cpu / Primal 0.063251761 s 0.071125787 s 0.89
jaxmd40 / Jax / cpu / Primal 0.070615546 s 0.066909728 s 1.06
jaxmd40 / HLOOpt / cpu / Primal 0.082267495 s 0.094894581 s 0.87
jaxmd40 / PartOpt / cpu / Primal 0.069847896 s 0.067299069 s 1.04
jaxmd40 / IPartOpt / cpu / Primal 0.070168536 s 0.071945101 s 0.98
jaxmd40 / DefOpt / cpu / Primal 0.081188583 s 0.090596363 s 0.90
jaxmd40 / IDefOpt / cpu / Primal 0.08240747 s 0.091488309 s 0.90
jaxmd40 / JaXPipe / cpu / Forward 0.153775891 s 0.165743457 s 0.93
jaxmd40 / Jax / cpu / Forward 0.087816248 s 0.096683967 s 0.91
jaxmd40 / HLOOpt / cpu / Forward 0.161387027 s 0.173893948 s 0.93
jaxmd40 / PartOpt / cpu / Forward 0.172594673 s 0.174310663 s 0.99
jaxmd40 / IPartOpt / cpu / Forward 0.1686806589999999 s 0.166700302 s 1.01
jaxmd40 / DefOpt / cpu / Forward 0.163009557 s 0.171433534 s 0.95
jaxmd40 / IDefOpt / cpu / Forward 0.157015731 s 0.181448687 s 0.87
jaxmd40 / JaXPipe / cpu / PreRev 0.247020994 s 0.23507325 s 1.05
jaxmd40 / JaXPipe / cpu / PostRev 0.136783274 s 0.141795398 s 0.96
jaxmd40 / JaXPipe / cpu / BothRev 0.225226648 s 0.222506143 s 1.01
jaxmd40 / Jax / cpu / BothRev 0.1371742879999999 s 0.137150959 s 1.00
jaxmd40 / HLOOpt / cpu / PreRev 0.235933793 s 0.227382734 s 1.04
jaxmd40 / HLOOpt / cpu / PostRev 0.177217543 s 0.179306236 s 0.99
jaxmd40 / HLOOpt / cpu / BothRev 0.2449341709999999 s 0.236692304 s 1.03
jaxmd40 / PartOpt / cpu / PreRev 0.2258393299999999 s 0.22981467 s 0.98
jaxmd40 / PartOpt / cpu / PostRev 0.127516268 s 0.1451506699999999 s 0.88
jaxmd40 / PartOpt / cpu / BothRev 0.238457842 s 0.255750397 s 0.93
jaxmd40 / IPartOpt / cpu / PreRev 0.222299711 s 0.246203902 s 0.90
jaxmd40 / IPartOpt / cpu / PostRev 0.125327601 s 0.139543965 s 0.90
jaxmd40 / IPartOpt / cpu / BothRev 0.242150328 s 0.2444348029999999 s 0.99
jaxmd40 / DefOpt / cpu / PreRev 0.204574316 s 0.229796742 s 0.89
jaxmd40 / DefOpt / cpu / PostRev 0.174820242 s 0.169326369 s 1.03
jaxmd40 / DefOpt / cpu / BothRev 0.249341117 s 0.261869205 s 0.95
jaxmd40 / IDefOpt / cpu / PreRev 0.245146983 s 0.237854778 s 1.03
jaxmd40 / IDefOpt / cpu / PostRev 0.170011579 s 0.1757676749999999 s 0.97
jaxmd40 / IDefOpt / cpu / BothRev 0.251418005 s 0.249661935 s 1.01
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / cuda / Primal 1.701021599 s 1.702607561 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / cuda / Primal 1.703867612 s 1.705338616 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / cuda / Primal 1.71404248 s 1.715820057 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / cuda / Primal 1.6957518299999998 s 1.697166519 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / cuda / Primal 1.692950155 s 1.695060618 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / cuda / Primal 1.6626941 s 1.666226228 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / cuda / Primal 1.910935532 s 1.914439734 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / JaXPipe / tpu / Primal 3.039092575625 s 3.038042935625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / Jax / tpu / Primal 3.039721480625 s 3.03847303625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / HLOOpt / tpu / Primal 3.12200026875 s 3.120727449375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / PartOpt / tpu / Primal 3.06042417 s 3.0592445937500004 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IPartOpt / tpu / Primal 3.060673035625 s 3.059483785625 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / DefOpt / tpu / Primal 2.102638580625 s 2.102149329375 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_24_outer_steps_4 / IDefOpt / tpu / Primal 2.948626418125 s 2.947106015 s 1.00
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / JaXPipe / cpu / Primal 6.335815229 s 6.159330476 s 1.03
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / Jax / cpu / Primal 6.275140779 s 6.086927948 s 1.03
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / HLOOpt / cpu / Primal 6.276289467 s 6.210631889 s 1.01
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / PartOpt / cpu / Primal 6.339062614 s 6.293127436 s 1.01
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IPartOpt / cpu / Primal 6.383657645 s 6.2816959390000005 s 1.02
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / DefOpt / cpu / Primal 2.432598753 s 2.482096093 s 0.98
neuralgcm_v1/deterministic_2_8_deg_inner_steps_2_outer_steps_2 / IDefOpt / cpu / Primal 6.776850834 s 6.689065608 s 1.01

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

@xys-syx xys-syx requested review from vimarsh6739 and wsmoses January 4, 2026 20:18
@xys-syx xys-syx marked this pull request as ready for review January 5, 2026 17:05
@xys-syx xys-syx marked this pull request as draft January 6, 2026 19:18
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.

2 participants