Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • hpvm-release-epochs
  • hpvm-release-epochs0
  • main default
  • v2.0
  • v2.0rc
  • v1.0
  • v0.5
7 results
Created with Raphaël 2.2.012Apr111098765432130Mar292827262423222119181716151312116543228Feb27262524232019181716151413121110742131Jan302928272524232221201918171615141310987working fusion with move-upMergingFix CBE bugs: include address space specifier on pointer casts, account for pointer decay on array bitcasts, and fix switch code genAdding Tuning results for Batch328 - more knobs - 8K iterationsRevert "Added Conv-Bias-ReLU fused operator"Revert "Added fuse test case (LeNet-based)"Added fuse test case (LeNet-based)Added Conv-Bias-ReLU fused operatorWIP: FPGA U&Jadd mobilenetSeparating Knob readiing Utiltiesalexnet & lenet keras/onnxruntimne readyFixing header dependencies for approxhpvm_runtime_utilsalexnet onnxruntime updateUsing increasing knob ordering for AutotunerAdding New Sampling Knobs to tensorConvSampSim -- with half interpolationAdd changes to profiler interfaces to avoid breaking coseadded test for dominance chain and extra check for loop guard branchAdding support for dumping CPU runtime configsmergignAdding baseline configuration to autotuning outputssplit unroll & jam to two passes so that we can run cse, simplifycfg, and loop simplify between themAdding CPU support in profiler and making adding devices more easyAdding CPU tensor operations and approximationsMerge branch 'approx_hpvm' of https://gitlab.engr.illinois.edu/llvm/hpvm into approx_hpvmAdding build support of OpenMP for CPU tensor runtimeUpdating Global Knobs file to include 33% samplingAdding Sampling test for 1*1 filterFix a bug in indexing for skip_offset == 4 for filter samplingModifying tensorConvSampSim to handle 33% samplingPushing latest unit test source - with tests for baseline FP32 FP16Merge branch 'approx_hpvm' of https://gitlab.engr.illinois.edu/llvm/hpvm into approx_hpvmFixed baseline convolution baseline and removed redundant tensorsAdding detailed unit test for 3*3 filter samplingmergingremoving stray printsAdd FP16 support for baseline convolutionAdd regular convolution and bug fixAdding unit tests for sampling and baseline for 1*1 filterAdd fixes to filter sampling