nnet-compute-from-egs.cc File Reference

#include "base/kaldi-common.h"
#include "util/common-utils.h"
#include "hmm/transition-model.h"
#include "nnet2/train-nnet.h"
#include "nnet2/am-nnet.h"

Include dependency graph for nnet-compute-from-egs.cc:

Go to the source code of this file.

Functions
int	main (int argc, char *argv[])

Function Documentation

main()

int main	(	int	argc,
		char *	argv[]
	)

Definition at line 27 of file nnet-compute-from-egs.cc.

References SequentialTableReader< Holder >::Done(), ParseOptions::GetArg(), NnetExample::input_frames, KALDI_LOG, NnetExample::left_context, Nnet::LeftContext(), SequentialTableReader< Holder >::Next(), kaldi::nnet2::NnetComputation(), ParseOptions::NumArgs(), CompressedMatrix::NumCols(), Nnet::OutputDim(), ParseOptions::PrintUsage(), MatrixBase< Real >::Range(), ParseOptions::Read(), kaldi::ReadKaldiObject(), Nnet::RightContext(), NnetExample::spk_info, CuMatrix< Real >::Swap(), SequentialTableReader< Holder >::Value(), and TableWriter< Holder >::Write().

                                 {
  try {
    using namespace kaldi;
    using namespace kaldi::nnet2;
    typedef kaldi::int32 int32;
    typedef kaldi::int64 int64;

    const char *usage =
        "Does the neural net computation, taking as input the nnet-training examples\n"
        "(typically an archive with the extension .egs), ignoring the labels; it\n"
        "outputs as a matrix the result.  Used mostly for debugging.\n"
        "\n"
        "Usage:  nnet-compute-from-egs [options] <raw-nnet-in> <egs-rspecifier> "
        "<feature-wspecifier>\n"
        "e.g.:  nnet-compute-from-egs 'nnet-to-raw-nnet final.mdl -|' egs.10.1.ark ark:-\n";
    
    ParseOptions po(usage);

    po.Read(argc, argv);
    
    if (po.NumArgs() != 3) {
      po.PrintUsage();
      exit(1);
    }
    
    std::string raw_nnet_rxfilename = po.GetArg(1),
        examples_rspecifier = po.GetArg(2),
        features_or_loglikes_wspecifier = po.GetArg(3);

    Nnet nnet;
    ReadKaldiObject(raw_nnet_rxfilename, &nnet);
    
    int64 num_egs = 0;

    SequentialNnetExampleReader example_reader(examples_rspecifier);
    BaseFloatMatrixWriter writer(features_or_loglikes_wspecifier);
    
    int32 left_context = nnet.LeftContext(),
        context = nnet.LeftContext() + 1 + nnet.RightContext();

    for (; !example_reader.Done(); example_reader.Next()) {
      const NnetExample &eg = example_reader.Value();
      int32 start_offset = eg.left_context - left_context;
      int32 basic_dim = eg.input_frames.NumCols(),
          spk_dim = eg.spk_info.Dim(), dim = basic_dim + spk_dim;
      Matrix<BaseFloat> input_frames(eg.input_frames),
          input_block(context, dim);
      input_block.Range(0, context, 0, basic_dim).CopyFromMat(
          input_frames.Range(start_offset, context, 0, basic_dim));
      if (spk_dim != 0) {
        input_block.Range(0, context, basic_dim, spk_dim).CopyRowsFromVec(
            eg.spk_info);
      }
      CuMatrix<BaseFloat> gpu_input_block;
      gpu_input_block.Swap(&input_block);
      CuMatrix<BaseFloat> gpu_output_block(1, nnet.OutputDim());
      
      bool pad_input = false;
      NnetComputation(nnet, gpu_input_block, pad_input, &gpu_output_block);
      writer.Write("global", Matrix<BaseFloat>(gpu_output_block));
      num_egs++;
    }
    
    KALDI_LOG << "Processed " << num_egs << " examples.";
    
    return (num_egs == 0 ? 1 : 0);
  } catch(const std::exception &e) {
    std::cerr << e.what() << '\n';
    return -1;
  }
}