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 2.2.017Mar16151312116543228Feb27262524232019181716151413121110742131Jan30292827252423222120191817161514131098743231Dec3029272618765228Nov26252120192019WIP fix for hpvm runtime with FPGAFixing polling and binary reinvocation fall-back in devtunerAdd new fileFixed necessary changes in other files for FPGAAdding binary reinvocation capability to devtuner.pyMinor fixFix bug that causes crash during parsing of LLVM assemblyAdding VGG16_imagenet_promie and VGG16_imagenet_piped rules to Cmake fileAdding VGG16-imagenet to benchmarks.pyAdding VGG16 imagenet data files - tensor, layer, knob files - need some fixingFixing integer overflows in approximation techniques implementationAdding ImageNet VGG16 - first successHPVM-CAVA-FPGA: Removing unnecessary files; changing visc -> hpvmFixing remaining overflow bugFixing int overflow issuesMergingAdding fixed function dumpCalibrationData2Potential fix for compilation failure with buffer/priv attributesMerge branch 'hpvm-fpga-9' of https://gitlab.engr.illinois.edu/llvm/hpvm into hpvm-fpga-9separate operators and ir, create new abstraction for element wise opsUpdate vgg16_imagenet.pynode without attrInsert hpvm_hint_fpga as neededadding fpga benchmarks to main repo, fpga branchir finalizedcodegen update & hpvm-tensor-rt path updateformat ir and codegen for src.ccAdding Autotuner results directoryadding fpga benchmarks to main repo, fpga branchcodegen part added with emitSource, emitHeader and emitRootHandling double spawn issue - removing multiple instantationsUsing stop_run signal to stop processesAdding piped sources for all DNNs and updating CMake + benchmarks.pyFixing devtuner bugreintroducing basline computationTo devtuner driver script having option for piped executionAdding piping support to devtuner.pyAdding tested AlexNet piped versionAdding batch size as param to Autotuner DNN sourcesAdding confidence dumps to all layer-level Autotuner sources