Skip to content
Snippets Groups Projects
Select Git revision
  • 5-arithmetic-operator-optimization
  • VersionManagement
  • add_clipping_node
  • chore_move_broadcasting_to_core
  • convert
  • dev protected
  • enhance_droptout
  • enhance_maxpool
  • feat/3d-unet-instance-norm
  • feat/DataTypes
  • feat/ImprovedBuild
  • feat/RefGetImpl
  • feat/SplitConv
  • feat/TensorCopy
  • feat_benchmark
  • feat_formatting
  • fix
  • fix_globalavgpool
  • fix_maxpool
  • learning
  • v0.7.0
  • v0.6.1
  • v0.6.0
  • v0.5.1
  • v0.5.0
  • v0.4.1
  • v0.4.0
  • v0.3.2
  • v0.3.1
  • v0.3.0
  • v0.2.3
  • v0.2.2
  • v0.2.1
  • v0.2.0
  • v0.1.0
35 results
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.010Apr983231Mar282726242117161074326Feb25242120171031Jan3029282725222115149Dec65229Nov2726252219181586425Oct23222120191130Sep272625242320181615121110943230Aug29272622139131Jul2618168543130Jun282726242120191817141210654330May28242321171614763221Apr191611109542130Mar292827262524222018151413118654227Feb262320191816151413129876Fix update gradient tensor dimesionsFixed compilation of other modulesReduced OpenMP thresholdsFix backward not being run in unit testFixed Windows buildAdded OpenMP to MatMulFixed wrong offsetsOptimized Fc ImplFixed Windows buildAdded OpenMPupd: remove 'Aidge::' scope from ConvImpl::forward functionsFixed backward for Sqrt and LeakyReLUadd: function to get inference time and output of some GraphView[Issue #250]: Add backward CPU implementation and unit testing for MaxPooling2D CPURefactoring of MaxPoolingImpl_kernels.hpp[Feat] add: unit test for conv2d operator int32_tFix FC backwardfix: move back 'castFromFloat' function in GloabalAvgPoolingImpl_kernel as 'static_cast' truncates and does not round as intendedfix: use aidge 'approxEq' function instead of manual comparison and change CHECKs for REQUIREs in interpolation testfix: error related to 'fma' call with different input types. Also simplify GlobalAvgPooling logicFix handling of negative axisMerge branch 'dev' into topkAdded missing MaxPoolingImpl registrationFixed indices to be conformant to ONNXFIxed TopK registrationFixed missing clampingAdded handling of second outputRemove fmt header from ConvImpl_kernels.hppUpdate Pybind 2.10.4 -> 2.13.6TopK initial implementationFixed duplicate header guardFix clip implementation + add ReLU int8.remove transA and transB attr from FC to be handled in the metaop TransposeFCadd Sum operatoradd hints to support batched biad in FCupdate FC factory callsupport only bias of size outChannels for FCadd transA and transB for FCsupport 2D Bias for FC operatoradd alpha and beta attr to FC impl
Loading