paddlespeech.s2t.decoders.ctcdecoder.swig_wrapper module

Wrapper for various CTC decoders in SWIG.

class paddlespeech.s2t.decoders.ctcdecoder.swig_wrapper.CTCBeamSearchDecoder(vocab_list, batch_size, beam_size, num_processes, cutoff_prob, cutoff_top_n, _ext_scorer, blank_id)

Bases: CtcBeamSearchDecoderBatch

Wrapper for CtcBeamSearchDecoderBatch. Args:

vocab_list (list): Vocabulary list. beam_size (int): Width for beam search. num_processes (int): Number of parallel processes. param cutoff_prob (float): Cutoff probability in vocabulary pruning,

default 1.0, no pruning.

cutoff_top_n (int): Cutoff number in pruning, only top cutoff_top_n

characters with highest probs in vocabulary will be used in beam search, default 40.

param ext_scorer (Scorer): External scorer for partially decoded sentence, e.g. word count

or language model.

Attributes

thisown

The membership flag

Methods

decode
next
reset_state

class paddlespeech.s2t.decoders.ctcdecoder.swig_wrapper.Scorer(alpha, beta, model_path, vocabulary)

Bases: Scorer

Wrapper for Scorer.

Parameters

• alpha (float) -- Parameter associated with language model. Don't use language model when alpha = 0.

• beta (float) -- Parameter associated with word count. Don't use word count when beta = 0.

• vocabulary (list) -- Vocabulary list.

Model_path

Path to load language model.

Attributes

alpha

beta

dictionary

thisown

The membership flag

Methods

get_dict_size
get_log_cond_prob
get_max_order
get_sent_log_prob
is_character_based
make_ngram
reset_params
split_labels

paddlespeech.s2t.decoders.ctcdecoder.swig_wrapper.ctc_beam_search_decoding(probs_seq, vocabulary, beam_size, cutoff_prob=1.0, cutoff_top_n=40, ext_scoring_func=None, blank_id=0)

Wrapper for the CTC Beam Search Decoding function.

Parameters

• probs_seq (2-D list) -- 2-D list of probability distributions over each time step, with each element being a list of normalized probabilities over vocabulary and blank.

• vocabulary (list) -- Vocabulary list.

• beam_size (int) -- Width for beam search.

• cutoff_prob (float) -- Cutoff probability in pruning, default 1.0, no pruning.

• cutoff_top_n (int) -- Cutoff number in pruning, only top cutoff_top_n characters with highest probs in vocabulary will be used in beam search, default 40.

• ext_scoring_func -- External scoring function for partially decoded sentence, e.g. word count or language model.

Returns

List of tuples of log probability and sentence as decoding results, in descending order of the probability.

Return type

list

paddlespeech.s2t.decoders.ctcdecoder.swig_wrapper.ctc_beam_search_decoding_batch(probs_split, vocabulary, beam_size, num_processes, cutoff_prob=1.0, cutoff_top_n=40, ext_scoring_func=None, blank_id=0)

Wrapper for the batched CTC beam search decodeing batch function.

Parameters

• probs_seq (3-D list) -- 3-D list with each element as an instance of 2-D list of probabilities used by ctc_beam_search_decoder().

• vocabulary (list) -- Vocabulary list.

• beam_size (int) -- Width for beam search.

• num_processes (int) -- Number of parallel processes.

• cutoff_prob (float) -- Cutoff probability in vocabulary pruning, default 1.0, no pruning.

• cutoff_top_n (int) -- Cutoff number in pruning, only top cutoff_top_n characters with highest probs in vocabulary will be used in beam search, default 40.

• num_processes -- Number of parallel processes.

• ext_scoring_func -- External scoring function for partially decoded sentence, e.g. word count or language model.

Returns

List of tuples of log probability and sentence as decoding results, in descending order of the probability.

Return type

list

paddlespeech.s2t.decoders.ctcdecoder.swig_wrapper.ctc_greedy_decoding(probs_seq, vocabulary, blank_id)

Wrapper for ctc best path decodeing function in swig.

Parameters

• probs_seq (2-D list) -- 2-D list of probability distributions over each time step, with each element being a list of normalized probabilities over vocabulary and blank.

• vocabulary (list) -- Vocabulary list.

Returns

Decoding result string.

Return type

str