7. Deprecated

The following features are deprecated and may be removed in future.

page deprecated

Member gcl::allGather  (poplar::Graph &graph, const poplar::Tensor &data, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allGatherCrossReplica() instead.

Member gcl::allGather  (poplar::Graph &graph, const poplar::Tensor &data, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allGatherCrossReplica() instead.

Member gcl::allReduce  (poplar::Graph &graph, const poplar::Tensor &data, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceCrossReplica() instead.

Member gcl::allReduce  (poplar::Graph &graph, const std::vector< poplar::Tensor > &datas, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceCrossReplica() instead.

Member gcl::allReduce  (poplar::Graph &graph, const std::vector< poplar::Tensor > &datas, popops::CollectiveOperator op, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceCrossReplica() instead.

Member gcl::allReduce  (poplar::Graph &graph, const poplar::Tensor &data, popops::CollectiveOperator op, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceCrossReplica() instead.

Member gcl::allReduceInPlace  (poplar::Graph &graph, poplar::Tensor &data, popops::CollectiveOperator op, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceInPlaceCrossReplica() instead.

Member gcl::allReduceInPlace  (poplar::Graph &graph, std::vector< poplar::Tensor > &datas, popops::CollectiveOperator op, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceInPlaceCrossReplica() instead.

Member gcl::allReduceInPlace  (poplar::Graph &graph, poplar::Tensor &data, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceInPlaceCrossReplica() instead.

Member gcl::allReduceInPlace  (poplar::Graph &graph, std::vector< poplar::Tensor > &datas, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceInPlaceCrossReplica() instead.

Member gcl::allReduceToDestination  (poplar::Graph &graph, const poplar::Tensor &data, poplar::Tensor &destination, popops::CollectiveOperator op, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceToDestinationCrossReplica() instead.

Member gcl::allReduceToDestination  (poplar::Graph &graph, const std::vector< poplar::Tensor > &datas, std::vector< poplar::Tensor > &destinations, popops::CollectiveOperator op, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceToDestinationCrossReplica() instead.

Member gcl::allReduceToDestination  (poplar::Graph &graph, const poplar::Tensor &data, poplar::Tensor &destination, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allReduceToDestinationCrossReplica() instead.

Member gcl::allToAll  (poplar::Graph &graph, const poplar::Tensor &data, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allToAllCrossReplica() instead.

Member gcl::allToAll  (poplar::Graph &graph, const poplar::Tensor &data, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use allToAllCrossReplica() instead.

Member gcl::reduceScatter  (poplar::Graph &graph, const poplar::Tensor &data, popops::CollectiveOperator op, poplar::program::Sequence &prog, const CommGroup &group, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use reduceScatterCrossReplica() instead.

Member gcl::reduceScatter  (poplar::Graph &graph, const poplar::Tensor &data, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use reduceScatterCrossReplica() instead.

Member poplar::Engine::resetExecutionProfile  ()

Use the PVA library instead.

Member poplar::Graph::TraceFn

Tracing via Poplar is deprecated and will be removed.

Member poplar::program::Sequence::Sequence  (T &&...args)

Use Sequence(std::initializer_list<Program>, const DebugContext&) instead.

Class poplin::matmul::PlanningCache

Use poplin::PlanningCache instead.

Member poplin::preplanConvolutions  (poplar::Graph &graph, const std::set< ConvPlanParams > &convs, PlanningCache &cache)

Use preplan() instead.

Member poplin::preplanConvolutions  (const std::set< ConvPlanParams > &convs, PlanningCache &cache)

Use preplan() instead.

Member poplin::preplanMatMuls  (const std::set< MatMulPlanParams > &matmuls, matmul::PlanningCache &cache)

Use preplan() instead.

Member popnn::gru::auGruBwd  (poplar::Graph &graph, const GruParams &params, poplar::program::Sequence &prog, const poplar::Tensor &fwdOutputInit, const poplar::Tensor &fwdIntermediatesSeq, const GruWeights &weights, const poplar::Tensor &fwdInputSeq, const poplar::Tensor &realTimeSteps, const poplar::Tensor &fwdOutput, const poplar::Tensor &gradLayerNext, poplar::Tensor *inputGrad, poplar::Tensor *bwdIntermediates, const poplar::Tensor &attentions, poplar::Tensor *attentionsGrad, const poplar::DebugContext &debugContext, const poplar::OptionFlags &options_, poplin::matmul::PlanningCache *planningCache)

Use previously defined auGruBwd() instead.

Member popnn::gru::auGruBwdWithWU  (poplar::Graph &graph, const GruParams &params, poplar::program::Sequence &prog, const poplar::Tensor &fwdOutputInit, const poplar::Tensor &fwdIntermediates, const GruWeights &weights, const poplar::Tensor &input, const poplar::Tensor &realTimeSteps, const poplar::Tensor &output, const poplar::Tensor &outputGrad, poplar::Tensor *inputGrad, GruWeights &weightsGrad, const poplar::Tensor &attentions, poplar::Tensor *attentionsGrad, const poplar::DebugContext &debugContext, const poplar::OptionFlags &options_, poplin::matmul::PlanningCache *planningCache)

Use previously defined auGruBwdWithWU() instead.

Member popnn::gru::auGruFwd  (poplar::Graph &graph, const GruParams &params, const poplar::Tensor &stateInit, const poplar::Tensor &in, const poplar::Tensor &realTimeSteps, const GruWeights &weights, poplar::Tensor *intermediates, const poplar::Tensor &attScores, poplar::program::Sequence &fwdProg, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={}, poplin::matmul::PlanningCache *planningCache=nullptr)

Use previously defined auGruFwd() instead.

Member popnn::gru::gruBwd  (poplar::Graph &graph, const GruParams &params, poplar::program::Sequence &prog, const poplar::Tensor &fwdOutputInit, const poplar::Tensor &fwdIntermediatesSeq, const GruWeights &weights, const poplar::Tensor &fwdInputSeq, const poplar::Tensor &realTimeSteps, const poplar::Tensor &fwdOutput, const poplar::Tensor &gradLayerNext, poplar::Tensor *inputGrad, poplar::Tensor *bwdIntermediates, const poplar::DebugContext &debugContext, const poplar::OptionFlags &options_, poplin::matmul::PlanningCache *planningCache)

Use previously defined popnn::gruBwd() instead.

Member popnn::gru::gruBwdWithWU  (poplar::Graph &graph, const GruParams &params, poplar::program::Sequence &prog, const poplar::Tensor &fwdOutputInit, const poplar::Tensor &fwdIntermediates, const GruWeights &weights, const poplar::Tensor &input, const poplar::Tensor &realTimeSteps, const poplar::Tensor &output, const poplar::Tensor &outputGrad, poplar::Tensor *inputGrad, GruWeights &weightsGrad, const poplar::DebugContext &debugContext, const poplar::OptionFlags &options_, poplin::matmul::PlanningCache *planningCache)

Use previously defined gruBwdWithWU() instead.

Member popnn::gru::gruFwd  (poplar::Graph &graph, const GruParams &params, const poplar::Tensor &stateInit, const poplar::Tensor &in, const poplar::Tensor &realTimeSteps, const GruWeights &weights, poplar::Tensor *intermediates, poplar::program::Sequence &fwdProg, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={}, poplin::matmul::PlanningCache *planningCache=nullptr)

Use previously defined gruFwd() instead.

Member popnn::gru::GruParams::batchSize

Use rnn::RnnParams.batchSize instead.

Member popnn::gru::GruParams::dataType

Use rnn::RnnParams.dataType instead.

Member popnn::gru::GruParams::layerSizes

Use rnn::RnnParams.layerSizes instead.

Member popnn::gru::GruParams::timeSteps

Use rnn::RnnParams.maxTimeSteps instead.

Member popnn::lstm::lstmBwd  (poplar::Graph &graph, const LstmParams &params, poplar::program::Sequence &prog, const LstmState &fwdStateInit, const poplar::Tensor &fwdIntermediates, const LstmWeights &weights, const poplar::Tensor &input, const poplar::Tensor &output, const poplar::Tensor &outputGrad, const poplar::Tensor *lastCellStateGrad, poplar::Tensor *inputGrad, poplar::Tensor *bwdIntermediates, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={}, poplin::matmul::PlanningCache *planningCache=nullptr)

deprecated Use previously defined popnn::lstmBwd() instead.

Member popnn::lstm::lstmBwdWithWU  (poplar::Graph &graph, const LstmParams &params, poplar::program::Sequence &prog, const LstmState &fwdStateInit, const poplar::Tensor &fwdIntermediates, const LstmWeights &weights, const poplar::Tensor &input, const poplar::Tensor &output, const poplar::Tensor &outputGrad, const poplar::Tensor *lastCellStateGrad, poplar::Tensor *inputGrad, LstmWeights &weightsGrad, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={}, poplin::matmul::PlanningCache *planningCache=nullptr)

deprecated Use previously defined popnn::lstmBwdWithWU() instead.

Member popnn::lstm::LstmParams::batchSize

Use rnn::RnnParams.batchSize instead.

Member popnn::lstm::LstmParams::dataType

Use rnn::RnnParams.dataType instead.

Member popnn::lstm::LstmParams::layerSizes

Use rnn::RnnParams.layerSizes instead.

Member popnn::lstm::LstmParams::timeSteps

Use rnn::RnnParams.maxTimeSteps instead.

Member popnn::rnn::RnnParams::timeSteps

Use RnnParams.maxTimeSteps instead.

Member popops::allGather  (poplar::Graph &graph, const Chunks &toGather, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use gcl::allGatherWithinReplica() instead

Member popops::allReduce  (poplar::Graph &graph, const poplar::Tensor &toReduce, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use gcl::allReduceWithinReplica() instead

Member popops::reduce  (poplar::Graph &graph, const poplar::Tensor &in, const poplar::Type &outType, const std::vector< std::size_t > &dims, ReduceParams params, std::vector< poplar::ComputeSet > &css, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

The reduce overloads that expect a vector of compute sets are deprecated. Please use the reduceMany() function instead.

Member popops::reduceScatter  (poplar::Graph &graph, const poplar::Tensor &toReduce, popops::CollectiveOperator op, poplar::program::Sequence &prog, const poplar::DebugContext &debugContext={}, const poplar::OptionFlags &options={})

Use gcl::reduceScatterWithinReplica() instead