Skip to content
Snippets Groups Projects
Select Git revision
  • clockwork default
  • clockwork-v0.6
  • master
  • v0.6.0
  • v0.5
  • v0.4
  • v0.3
  • v0.2
  • v0.1rc
9 results
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.018Nov1615141312111098632130Oct2726252422211918171615141312111098765432129Sep2827262524232221201817131211108765432131Aug30292827262524232221201918171514131110987532131Jul28272625242320191817161514131211109876543130Jun292726232220181716151211105432131May30252423222120181716151412119876543229Apr272625242221201817161514121098231Mar2926251714131211753228Feb27262524222120171613121110987654231Jan302824232220191817161513121110865431Dec301112542129Nov28272625191817642128Oct272623222119181615141312Fixed nnvm issue #239 (#660)[RUNTIME] support limited save without cross compile (#659)Compat for opencl mode between cpu mode and gpu mode (#655)Conv2d scheduler tweaked for super resolution perf (#652)[UNROLL] New unroll option (#647)[TOPI] Add out_dtype argument for conv2d; Add x86 schedules (#646)[APP] improve parameter pack (#645)conv2d perf improved for conv2d_56_64_128, super resolution workloads added (#643)Fix conda packages (#642)[PASS] Fix vthread when extern access touching (#636)[CODEGEN] Enable closure with no argument (#635)[PASS] Update coproc sync (#634)[TUTORIAL] use OpenCL on ARM board (#633)[PASS] Enhance LiftAttrScope (#632)[NNPACK] Add argument nthreads (#631)android gemm for topi/recipe (#628)inline AMD GPU functions (#625)WIP: Add how_to readme to install tvm with nnpack support (#610)Support vector operations for AMD (llvm IR) (#623)conv2d_56_64_128 mark==1 bug fixed (#624)remove minimum 32-bit restriction (#621)add tanh dispatch (#619)[TOPI] fix weight layout in conv2d_transpose (#616)[DLPack] Upgrade dlpack to 0.2 (#609)[TOPI] modify conv2d_transpose schedule (#613)[INTRIN] Enable popcount (#606)Fixed build with metal on MacOS with case-sensitive FS (#601)vgg16 workload error fixed (#598)[TOPI] Support ceil_mode in pooling (#593)add helpful message to topi test (#592)[ROCM] remove fma dispatch (#591)[ROCM] View llvm ir and gcn asm with module.get_source(...) (#590)[BUFFER] Smarter slice to detect compactness (#587)[TOPI] add conv2d_transpose_nchw (#586)[PYTHON] Allow no de-allocation when exit (#583)[CODEGEN] Fix CPU compute attribute (#582)[DOCS] Fix tag_scope example (#581) Update topi/cuda schedules to use target.max_num_threads (#577)[PASS] More robust UnrollLoop configuratin (#576)add friendly tips when not found cl and link (#574)