Decoders are search strategies for traversing the search space which is spanned by the predictors. Decoders are specified using the --decoder arguments.

Available decoders

  • greedy: Greedy decoding (similar to beam=1)
  • beam: Beam search like in Bahdanau et al, 2015 .
  • sepbeam: Associates predictors with hypos in beam search and applies only one predictor instead of all for hypo expansion.
  • syncbeam: Beam search which compares after consuming a special synchronization symbol instead of after each iteration.
  • syntaxbeam: Beam search which ensures diversity amongst terminal symbol histories.
  • mbrbeam: Diversity encouraging beam search which maximizes the expected BLEU.
  • combibeam: Beam search which applies --combination_scheme at each time step.
  • multisegbeam: Beam search with multiple segmentations.
  • dfs: Depth-first search. This should be used for exact decoding or the complete enumeration of the search space. but it cannot be used if the search space is too large (like for unrestricted NMT) as it performs exhaustive search. If you have not only negative predictor scores, set --early_stopping to false.
  • restarting: Like DFS but with better admissible pruning behavior.
  • astar: A* search. The heuristic function is configured using the --heuristics options.
  • bucket: Works best for bag problems. Maintains buckets for each hypo length and extends a hypo in a bucket by one before selecting the next bucket.
  • flip: This decoder works only for bag problems. It traverses the search space by switching two words in the hypothesis. Do not use bow predictor.
  • bow: Restarting decoder optimized for bag-of-words problems.
  • bigramgreedy: Works best for bag problems. Collects bigram statistics and constructs hypos to score by greedily selecting high scoring bigrams. Do not use bow predictor with this search strategy.
  • vanilla: Original Blocks beam decoder. This bypasses the predictor framework and directly performs pure NMT beam decoding on the GPU. Use this when you do pure NMT decoding as this is usually faster then using a single nmt predictor as the search can be parallelized on the GPU.

Detailed descriptions are available below in the modules.

Relevant modules

cam.sgnmt.decoding.astar module

cam.sgnmt.decoding.beam module

cam.sgnmt.decoding.bigramgreedy module

cam.sgnmt.decoding.bow module

cam.sgnmt.decoding.bucket module

cam.sgnmt.decoding.combibeam module

cam.sgnmt.decoding.core module

cam.sgnmt.decoding.decoder module

cam.sgnmt.decoding.dfs module

cam.sgnmt.decoding.flip module

cam.sgnmt.decoding.greedy module

cam.sgnmt.decoding.heuristics module

cam.sgnmt.decoding.mbrbeam module

cam.sgnmt.decoding.multisegbeam module

cam.sgnmt.decoding.restarting module

cam.sgnmt.decoding.sepbeam module

cam.sgnmt.decoding.syncbeam module

cam.sgnmt.decoding.syntaxbeam module

Module contents

This package contains the central interfaces for the decoder (in the core module ), and the implementations of search strategies (Decoder).