Skip to content

Instantly share code, notes, and snippets.

@davidberard98

davidberard98/autogen58-repro.py

Created April 15, 2022 22:41
Show Gist options
  • Save davidberard98/12fdd498e5284a543f2daa55addc8445 to your computer and use it in GitHub Desktop.
Save davidberard98/12fdd498e5284a543f2daa55addc8445 to your computer and use it in GitHub Desktop.
Download ZIP
repro microbenchmark autogen-58.py from https://github.com/pytorch/benchmark/pull/801
Raw
autogen58-repro.py
import torch
import torch.utils.jit.log_extract as log_extract
ir = """graph(%0 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%1 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%2 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%3 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%4 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%5 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%6 : Double(requires_grad=0, device=cuda:0),
%7 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%8 : Double(requires_grad=0, device=cuda:0),
%9 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%10 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%11 : Double(requires_grad=0, device=cuda:0),
%12 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%13 : Double(1, 1, 26, strides=[26, 26, 1], requires_grad=0, device=cuda:0),
%14 : Double(requires_grad=0, device=cuda:0),
%15 : Double(requires_grad=0, device=cuda:0),
%16 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%17 : Double(requires_grad=0, device=cuda:0),
%18 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%19 : Double(requires_grad=0, device=cuda:0),
%20 : Double(requires_grad=0, device=cuda:0),
%21 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%22 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%23 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%24 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%25 : Double(requires_grad=0, device=cuda:0),
%26 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%27 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%28 : Double(requires_grad=0, device=cuda:0),
%29 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%30 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%31 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%32 : Double(requires_grad=0, device=cuda:0),
%33 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%34 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%35 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%36 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%37 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%38 : Double(requires_grad=0, device=cuda:0),
%39 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%40 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%41 : Double(requires_grad=0, device=cuda:0),
%42 : Double(requires_grad=0, device=cuda:0),
%43 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%44 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%45 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%46 : Double(requires_grad=0, device=cuda:0),
%47 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%48 : Double(requires_grad=0, device=cuda:0),
%49 : Double(requires_grad=0, device=cuda:0),
%50 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%51 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%52 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%53 : Double(requires_grad=0, device=cuda:0),
%54 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%55 : Double(requires_grad=0, device=cuda:0),
%56 : Double(requires_grad=0, device=cuda:0),
%57 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%58 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%59 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%60 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%61 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%62 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%63 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%64 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%65 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%66 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%67 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%68 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%69 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0),
%70 : int,
%71 : int,
%72 : int,
%73 : int,
%99 : int,
%100 : int,
%101 : int,
%102 : int,
%103 : int,
%104 : int,
%105 : int,
%106 : int,
%107 : int,
%108 : int,
%109 : int,
%110 : int,
%111 : int,
%112 : int,
%113 : int,
%114 : int,
%115 : int,
%116 : int,
%117 : int,
%118 : int,
%119 : int,
%120 : int,
%121 : int,
%122 : int,
%123 : int,
%124 : int):
%125 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%68, %69, %124)
%126 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%23, %9)
%127 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%67, %22)
%128 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%127, %21)
%129 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%125, %128, %123)
%130 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%129, %7)
%131 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%66, %21)
%132 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%65, %22)
%133 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%132, %21)
%134 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%64, %29)
%135 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%63, %30)
%136 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%61, %7)
%137 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%136, %22)
%138 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%137, %21)
%139 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%61, %62)
%140 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%60, %11)
%141 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%140, %58)
%142 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%59, %11)
%143 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%142, %58)
%144 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%143, %141, %122)
%145 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%144, %139, %121)
%146 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%145, %138, %120)
%147 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%146, %135, %119)
%148 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%147, %134, %118)
%149 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%148, %3)
%150 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%57, %11)
%151 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%150, %56, %117)
%152 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%39, %151)
%153 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%54, %55, %116)
%154 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%153, %152, %115)
%155 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%39, %154)
%156 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%52, %53, %114)
%157 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%156, %155, %113)
%158 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%51, %157)
%159 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%50, %11)
%160 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%159, %49, %112)
%161 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%39, %160)
%162 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%47, %48, %111)
%163 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%162, %161, %110)
%164 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%39, %163)
%165 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%45, %46, %109)
%166 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%165, %164, %108)
%167 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%166, %158, %107)
%168 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%44, %167)
%169 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%43, %11)
%170 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%169, %42, %106)
%171 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%39, %170)
%172 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%40, %41, %105)
%173 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%172, %171, %104)
%174 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%39, %173)
%175 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%37, %38, %103)
%176 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%175, %174, %102)
%177 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%176, %168, %101)
%178 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%177, %149, %100)
%179 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%178, %133, %99)
%180 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%179, %36)
%181 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%180, %131, %98)
%182 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%181, %130, %97)
%183 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%182, %126, %96)
%184 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%183, %20)
%185 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%184, %3)
%186 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%185, %35)
%187 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::reciprocal(%30)
%188 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%187, %28)
%189 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%34, %28)
%190 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%33, %28)
%191 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%190, %32, %95)
%192 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%191, %189, %94)
%193 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%192, %31, %93)
%194 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%13, %193)
%195 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%194, %188)
%196 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%195, %28, %92)
%197 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::pow(%30, %91)
%198 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::div(%194, %197)
%199 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%198, %29)
%200 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%27, %28)
%201 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%26, %11)
%202 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%201, %25, %90)
%203 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%202, %200, %89)
%204 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%203, %24, %88)
%205 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%13, %204)
%206 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%205, %188)
%207 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%206, %199, %87)
%208 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%23, %20)
%209 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%208, %3)
%210 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%209, %207)
%211 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::div(%210, %196)
%212 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::log(%196)
%213 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%208, %22)
%214 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%213, %212)
%215 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%214, %21)
%216 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%129, %20)
%217 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%216, %3)
%218 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%217, %212)
%219 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%10, %19)
%220 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%219, %13)
%221 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%18, %11)
%222 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%221, %3)
%223 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%16, %17, %86)
%224 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%223, %222, %85)
%225 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%224, %220, %84)
%226 : Double(1, 1, 26, strides=[26, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%13, %15)
%227 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%226, %225)
%228 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%227, %12)
%229 : Double(1, 1, 26, strides=[26, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%13, %14)
%230 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%221, %12)
%231 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%10, %11)
%232 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%231, %9)
%233 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%232, %3)
%234 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%7, %8)
%235 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%234, %3)
%236 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%5, %6, %83)
%237 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%236, %4, %82)
%238 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%237, %235, %81)
%239 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%238, %233, %80)
%240 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%239, %230, %79)
%241 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%240, %229, %78)
%242 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%226, %241)
%243 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::mul(%242, %3)
%244 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%243, %228, %77)
%245 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%244, %218, %76)
%246 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%245, %215, %75)
%247 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%246, %211, %74)
%248 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%247, %186, %73)
%249 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%248, %2, %72)
%250 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::sub(%249, %1, %71)
%251 : Double(204, 204, 26, strides=[5304, 26, 1], requires_grad=0, device=cuda:0) = aten::add(%250, %0, %70)
return (%251)
"""
_, inputs = log_extract.load_graph_and_inputs(ir)
log_extract.run_nnc(ir, inputs, dynamic=False)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment