[GitHub] asitstands commented on a change in pull request #10970: [MXNET-424] dtype option for multinomial
asitstands commented on a change in pull request #10970: [MXNET-424] dtype option for multinomial URL: https://github.com/apache/incubator-mxnet/pull/10970#discussion_r191984961 ## File path: src/operator/random/sample_multinomial_op.h ## @@ -67,6 +70,10 @@ inline bool SampleMultinomialOpShape(const nnvm::NodeAttrs& attrs, const TShape& ishape = (*in_attrs)[0]; if (!ishape.ndim()) return false; + MSHADOW_TYPE_SWITCH(param.dtype, DType, { +CHECK_LE(ishape[ishape.ndim() - 1], mxnet::common::MaxIntegerValue()); Review comment: I added an error message. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] asitstands commented on a change in pull request #10970: [MXNET-424] dtype option for multinomial
asitstands commented on a change in pull request #10970: [MXNET-424] dtype option for multinomial URL: https://github.com/apache/incubator-mxnet/pull/10970#discussion_r189154689 ## File path: src/operator/random/sample_multinomial_op.h ## @@ -155,9 +158,11 @@ void SampleMultinomialForward(const nnvm::NodeAttrs& attrs, Tensor uniform = ctx.requested[1].get_space_typed(Shape1(N*M), s); prnd->SampleUniform(&uniform, 0, 1); -Kernel::Launch( - s, N, K, M, inputs[0].dptr(), uniform.dptr_, outputs[0].dptr(), - param.get_prob ? outputs[1].dptr() : nullptr); +MSHADOW_TYPE_SWITCH(outputs[0].type_flag_, IType, { Review comment: Sometimes the multinomial samples need further processing in floating point arithmetic, so the samples need to be copied into a new array of floating point type. The copy slows down the training. For example, in RBM, the samples need to be applied by `linalg.gemm` which supports only floating point arrays. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] asitstands commented on a change in pull request #10970: [MXNET-424] dtype option for multinomial
asitstands commented on a change in pull request #10970: [MXNET-424] dtype option for multinomial URL: https://github.com/apache/incubator-mxnet/pull/10970#discussion_r189154689 ## File path: src/operator/random/sample_multinomial_op.h ## @@ -155,9 +158,11 @@ void SampleMultinomialForward(const nnvm::NodeAttrs& attrs, Tensor uniform = ctx.requested[1].get_space_typed(Shape1(N*M), s); prnd->SampleUniform(&uniform, 0, 1); -Kernel::Launch( - s, N, K, M, inputs[0].dptr(), uniform.dptr_, outputs[0].dptr(), - param.get_prob ? outputs[1].dptr() : nullptr); +MSHADOW_TYPE_SWITCH(outputs[0].type_flag_, IType, { Review comment: Sometimes the multinomial samples need further processing in floating point arithmetic, so the samples need to be copied into a new array of floating point type which slow down the training. For example, in RBM, the samples need to be applied by `linalg.gemm` which supports only floating point arrays. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services