Functions — github.com/najoungkim/COGS

hub / github.com/najoungkim/COGS / functions

Functions775 in github.com/najoungkim/COGS

⨍ Functions 775◇ Types & classes 138↳ Endpoints 7

↓ 2 callersMethod_build_target_tokens

(self, src, src_vocab, src_raw, pred, attn)

src/OpenNMT-py/onmt/translate/translation.py:36

↓ 2 callersMethod_caller

(self, stats)

src/OpenNMT-py/onmt/utils/earlystopping.py:29

↓ 2 callersMethod_compute_coverage_loss

(self, std_attn, coverage_attn)

src/OpenNMT-py/onmt/utils/loss.py:303

↓ 2 callersMethod_compute_loss

Compute the loss. Subclass must define this method. Args: batch: the current batch. output: the predict out

src/OpenNMT-py/onmt/utils/loss.py:109

↓ 2 callersMethod_decode_and_generate

( self, decoder_in, memory_bank, batch, src_vocabs

src/OpenNMT-py/onmt/translate/translator.py:566

↓ 2 callersMethod_decreasing_or_stopped_status_update

(self, tolerance)

src/OpenNMT-py/onmt/utils/earlystopping.py:203

↓ 2 callersMethod_forward

A naive forward pass for transformer decoder. # T: could be 1 in the case of stepwise decoding or tgt_len Args: inputs

src/OpenNMT-py/onmt/decoders/transformer.py:110

↓ 2 callersMethod_iter_dataset

(self, path)

src/OpenNMT-py/onmt/inputters/inputter.py:796

↓ 2 callersMethod_log_best_step

(self)

src/OpenNMT-py/onmt/utils/earlystopping.py:196

↓ 2 callersFunction_old_style_field_list

Detect old-style text fields. Not old style vocab, old nesting, and text-type fields not using ``TextMultiField``. Args: vocab:

src/OpenNMT-py/onmt/inputters/inputter.py:276

↓ 2 callersFunction_pad_vocab_to_multiple

(vocab, multiple)

src/OpenNMT-py/onmt/inputters/inputter.py:329

↓ 2 callersFunction_pool

(data, batch_size, batch_size_fn, batch_size_multiple, sort_key, random_shuffler, pool_factor)

src/OpenNMT-py/onmt/inputters/inputter.py:619

↓ 2 callersMethod_report_score

(self, name, score_total, words_total)

src/OpenNMT-py/onmt/translate/translator.py:731

↓ 2 callersMethod_rms

(self, x)

src/OpenNMT-py/onmt/utils/optimizers.py:417

↓ 2 callersMethod_run_encoder

(self, batch)

src/OpenNMT-py/onmt/translate/translator.py:551

↓ 2 callersMethod_stats

Args: loss (:obj:`FloatTensor`): the loss computed by the loss criterion. scores (:obj:`FloatTensor`): a score for ea

src/OpenNMT-py/onmt/utils/loss.py:171

↓ 2 callersFunction_threshold_and_support

Sparsemax building block: compute the threshold Args: input: any dimension dim: dimension along which to apply the sparsemax

src/OpenNMT-py/onmt/modules/sparse_activations.py:21

↓ 2 callersMethodadd_child

error handler

src/OpenNMT-py/onmt/bin/train.py:175

↓ 2 callersMethodadd_insertion_noise

(self, tokens, p)

src/OpenNMT-py/onmt/modules/source_noise.py:295

↓ 2 callersMethodall_gather_stats

Gather a `Statistics` object accross multiple process/nodes Args: stat(:obj:Statistics): the statistics object to gather

src/OpenNMT-py/onmt/utils/statistics.py:28

↓ 2 callersFunctionall_reduce_buffer

()

src/OpenNMT-py/onmt/utils/distributed.py:49

↓ 2 callersMethodbackward

Wrapper for backward pass. Some optimizer requires ownership of the backward pass.

src/OpenNMT-py/onmt/utils/optimizers.py:327

↓ 2 callersFunctionbatch_size_fn

(new, count, sofar)

src/OpenNMT-py/onmt/inputters/inputter.py:588

↓ 2 callersMethodblock_ngram_repeats

We prevent the beam from going in any direction that would repeat any ngram of size <block_ngram_repeat> more thant once. Th

src/OpenNMT-py/onmt/translate/decode_strategy.py:115

↓ 2 callersFunctionbuild_align_pharaoh

Convert valid alignment matrix to i-j (from 0) Pharaoh format pairs, or empty list if it's None.

src/OpenNMT-py/onmt/utils/alignment.py:71

↓ 2 callersFunctionbuild_base_model

Build a model from opts. Args: model_opt: the option loaded from checkpoint. It's important that the opts have been updated a

src/OpenNMT-py/onmt/model_builder.py:117

↓ 2 callersFunctionbuild_dataset_iter_multiple

(train_shards, fields, opt)

src/OpenNMT-py/onmt/inputters/inputter.py:912

↓ 2 callersFunctionbuild_save_dataset

(corpus_type, fields, src_reader, tgt_reader, align_reader, opt)

src/OpenNMT-py/onmt/bin/preprocess.py:134

↓ 2 callersFunctionbuild_translator

(opt, report_score=True, logger=None, out_file=None)

src/OpenNMT-py/onmt/translate/translator.py:22

↓ 2 callersFunctioncalc_vocab_load_stats

(vocab, loaded_embed_dict)

src/OpenNMT-py/tools/embeddings_to_torch.py:64

↓ 2 callersMethodckpt_model_opts

(cls, ckpt_opt)

src/OpenNMT-py/onmt/utils/parse.py:84

↓ 2 callersFunctioncollapse_copy_scores

Given scores from an expanded dictionary corresponeding to a batch, sums together copies, with a dictionary word when it is ambiguous.

src/OpenNMT-py/onmt/modules/copy_generator.py:8

↓ 2 callersFunctioncollate_tensor

(batch)

src/OpenNMT-py/tools/vid_feature_extractor.py:67

↓ 2 callersMethodconfig

(fields)

src/OpenNMT-py/onmt/inputters/dataset_base.py:161

↓ 2 callersFunctionconvert_to_torch_tensor

(word_to_float_list_dict, vocab)

src/OpenNMT-py/tools/embeddings_to_torch.py:56

↓ 2 callersMethoddetokenize

Detokenize a single sequence Same args/returns as :func:`tokenize()`

src/OpenNMT-py/onmt/translate/translation_server.py:745

↓ 2 callersMethoddummy_inputs

(cls, params, init_case)

src/OpenNMT-py/onmt/tests/test_embeddings.py:62

↓ 2 callersMethodelapsed_time

compute elapsed time

src/OpenNMT-py/onmt/utils/statistics.py:100

↓ 2 callersMethodensure_max_length

(self)

src/OpenNMT-py/onmt/translate/decode_strategy.py:109

↓ 2 callersMethodensure_min_length

(self, log_probs)

src/OpenNMT-py/onmt/translate/decode_strategy.py:105

↓ 2 callersFunctionfilter_shard_state

(state, shard_size=None)

src/OpenNMT-py/onmt/utils/loss.py:319

↓ 2 callersMethodfirst_step

(self, beam, expected_beam_scores, expected_len_pen)

src/OpenNMT-py/onmt/tests/test_beam_search.py:431

↓ 2 callersMethodget_dropout_mask_

(self, size, p)

src/OpenNMT-py/onmt/models/sru.py:562

↓ 2 callersFunctionget_function_by_path

(path, args=[], kwargs={})

src/OpenNMT-py/onmt/translate/translation_server.py:801

↓ 2 callersFunctionget_pairs

Return set of symbol pairs in a word. word is represented as tuple of symbols (symbols being variable-length strings)

src/OpenNMT-py/tools/apply_bpe.py:126

↓ 2 callersMethodinit_step

(self, beam, expected_len_pen)

src/OpenNMT-py/onmt/tests/test_beam_search.py:414

↓ 2 callersMethodload_model

Load a model given a set of options

src/OpenNMT-py/onmt/translate/translation_server.py:174

↓ 2 callersMethodload_pretrained_vectors

Load in pretrained embeddings. Args: emb_file (str) : path to torch serialized embeddings

src/OpenNMT-py/onmt/modules/embeddings.py:242

↓ 2 callersFunctionmain

()

src/OpenNMT-py/onmt/bin/train.py:205

↓ 2 callersFunctionmain

()

src/OpenNMT-py/onmt/bin/release_model.py:25

↓ 2 callersFunctionmain

()

src/OpenNMT-py/onmt/bin/translate.py:44

↓ 2 callersFunctionmain

()

src/OpenNMT-py/onmt/bin/server.py:149

↓ 2 callersFunctionmain

()

src/OpenNMT-py/onmt/bin/preprocess.py:314

↓ 2 callersFunctionmain

()

src/OpenNMT-py/onmt/bin/average_models.py:39

↓ 2 callersMethodmap_state

(self, fn)

src/OpenNMT-py/onmt/decoders/decoder.py:183

↓ 2 callersMethodname

(self, path)

src/OpenNMT-py/tools/vid_feature_extractor.py:128

↓ 2 callersMethodname_

(path, out_path)

src/OpenNMT-py/tools/vid_feature_extractor.py:119

↓ 2 callersMethodnext

(self)

src/OpenNMT-py/tools/vid_feature_extractor.py:56

↓ 2 callersFunctionnext_batch

(device_id)

src/OpenNMT-py/onmt/bin/train.py:113

↓ 2 callersMethodnoise_source

(self, source, **kwargs)

src/OpenNMT-py/onmt/modules/source_noise.py:70

↓ 2 callersFunctionpatch_fields

(opt, fields)

src/OpenNMT-py/onmt/inputters/inputter.py:192

↓ 2 callersMethodpreload_model

Preloading the model: updating internal datastructure It will effectively load the model if `load` is set

src/OpenNMT-py/onmt/translate/translation_server.py:182

↓ 2 callersMethodprocess

Turn a batch of align-idx to a sparse align idx Tensor

src/OpenNMT-py/onmt/inputters/inputter.py:71

↓ 2 callersMethodread

Read data into dicts. Args: data (str or Iterable[str]): Sequence of audio paths or path to file containing audio

src/OpenNMT-py/onmt/inputters/audio_dataset.py:96

↓ 2 callersMethodread

Read text data from disk. Args: sequences (str or Iterable[str]): path to text file or iterable of the actual tex

src/OpenNMT-py/onmt/inputters/text_dataset.py:12

↓ 2 callersMethodread

Read data into dicts. Args: images (str or Iterable[str]): Sequence of image paths or path to file containing aud

src/OpenNMT-py/onmt/inputters/image_dataset.py:47

↓ 2 callersFunctionrecursive_split

Recursively split segment into smaller units (by reversing BPE merges) until all units are either in-vocabulary, or cannot be split futher.

src/OpenNMT-py/tools/apply_bpe.py:204

↓ 2 callersFunctionrelative_matmul

Helper function for relative positions attention.

src/OpenNMT-py/onmt/utils/misc.py:120

↓ 2 callersFunctionreport_matrix

(row_label, column_label, matrix)

src/OpenNMT-py/onmt/utils/misc.py:142

↓ 2 callersMethodreset_unload_timer

(self)

src/OpenNMT-py/onmt/translate/translation_server.py:601

↓ 2 callersMethodsecond_step

(self, beam, expected_beam_scores, expected_len_pen)

src/OpenNMT-py/onmt/tests/test_beam_search.py:464

↓ 2 callersFunctionshape_transform

Tranform the size of the tensors to fit for conv input.

src/OpenNMT-py/onmt/utils/cnn_factory.py:13

↓ 2 callersFunctionsubword_map_by_joiner

Return word id for each subword token (annotate by joiner).

src/OpenNMT-py/onmt/utils/alignment.py:118

↓ 2 callersFunctionsubword_map_by_spacer

Return word id for each subword token (annotate by spacer).

src/OpenNMT-py/onmt/utils/alignment.py:134

↓ 2 callersMethodthird_step

(self, beam, expected_beam_scores, expected_len_pen)

src/OpenNMT-py/onmt/tests/test_beam_search.py:503

↓ 2 callersMethodto_device

(self, t)

src/OpenNMT-py/onmt/modules/source_noise.py:22

↓ 2 callersMethodtranslate

(self, texts_to_translate, batch_size=8)

src/OpenNMT-py/onmt/translate/translation_server.py:104

↓ 2 callersMethodtranslate_batch

Translate a batch of sentences.

src/OpenNMT-py/onmt/translate/translator.py:516

↓ 2 callersMethodupdate_dropout

(self, dropout, attention_dropout)

src/OpenNMT-py/onmt/encoders/transformer.py:132

↓ 2 callersMethodvalidate_translate_opts

(cls, opt)

src/OpenNMT-py/onmt/utils/parse.py:124

↓ 2 callersFunctionwrite_embeddings

(filename, dict, embeddings)

src/OpenNMT-py/tools/extract_embeddings.py:23

↓ 1 callersMethod__init__

(self, model, root_dir, filenames)

src/OpenNMT-py/tools/vid_feature_extractor.py:37

↓ 1 callersMethod__init__

(self, option_strings, dest, help=None, **kwargs)

src/OpenNMT-py/onmt/opts.py:819

↓ 1 callersMethod__init__

(self, dim=0)

src/OpenNMT-py/onmt/modules/sparse_activations.py:82

↓ 1 callersMethod__init__

(self, dtype)

src/OpenNMT-py/onmt/modules/util_class.py:43

↓ 1 callersMethod__init__

(self, input_size, width=3, dropout=0.2, nopad=False)

src/OpenNMT-py/onmt/utils/cnn_factory.py:21

↓ 1 callersMethod__init__

(self, num_layers, d_model, heads, d_ff, dropout, attention_dropout, embeddings, max_relative

src/OpenNMT-py/onmt/encoders/transformer.py:91

↓ 1 callersMethod__init__

(self, rnn_type, state_dim, bidir_edges, n_edge_types, n_node, bridge_extra_node, n_steps, sr

src/OpenNMT-py/onmt/encoders/ggnn_encoder.py:85

↓ 1 callersMethod__init__

(self, **kwargs)

src/OpenNMT-py/onmt/inputters/inputter.py:66

↓ 1 callersMethod__init__

(self, attentional=True)

src/OpenNMT-py/onmt/decoders/decoder.py:18

↓ 1 callersMethod__init__

(self, num_layers, d_model, heads, d_ff, copy_attn, self_attn_type, dropout, attention_dropou

src/OpenNMT-py/onmt/decoders/transformer.py:219

↓ 1 callersMethod__init__

(self, num_layers, input_size, rnn_size, dropout)

src/OpenNMT-py/onmt/models/stacked_rnn.py:45

↓ 1 callersMethod_accum_batches

(self, iterator)

src/OpenNMT-py/onmt/trainer.py:180

↓ 1 callersMethod_align_forward

For a batch of input and its prediction, return a list of batch predict alignment src indice Tensor in size ``(batch, n_best,)``.

src/OpenNMT-py/onmt/translate/translator.py:473

↓ 1 callersMethod_align_pad_prediction

Padding predictions in batch and add BOS. Args: predictions (List[List[Tensor]]): `(batch, n_best,)`, for each src

src/OpenNMT-py/onmt/translate/translator.py:446

↓ 1 callersMethod_bridge

Forward hidden state through bridge.

src/OpenNMT-py/onmt/encoders/rnn_encoder.py:100

↓ 1 callersMethod_bridge

Forward hidden state through bridge.

src/OpenNMT-py/onmt/encoders/ggnn_encoder.py:264

↓ 1 callersFunction_build_field_vocab

(field, counter, size_multiple=1, **kwargs)

src/OpenNMT-py/onmt/inputters/inputter.py:340

↓ 1 callersMethod_build_rnn

(self, rnn_type, **kwargs)

src/OpenNMT-py/onmt/decoders/decoder.py:322

↓ 1 callersMethod_check_deps

(cls)

src/OpenNMT-py/onmt/inputters/audio_dataset.py:54

← previous next →101–200 of 775, ranked by callers