MCPcopy Create free account

hub / github.com/OFA-Sys/OFA / functions

Functions6,938 in github.com/OFA-Sys/OFA

⨍ Functions 6,938◇ Types & classes 1,161↳ Endpoints 9
↓ 3 callersFunctioncollate
( samples, pad_idx, eos_idx, left_pad_source=False, left_pad_target=False, feature_onl
data/s2t_data/unify_dataset.py:45
↓ 3 callersFunctioncollate
(samples, pad_idx, eos_idx)
data/pretrain_data/unify_dataset.py:55
↓ 3 callersMethodcompute_distances
For every centroid m, computes ||M - m[None, :]||_2 Remarks: - We rely on PyTorch's broadcast
fairseq/fairseq/modules/quantization/pq/em.py:133
↓ 3 callersMethodcompute_loss
(self, feat_out, feat_out_post, eos_out, feat_tgt, eos_tgt, tgt_lens, reduction="mean")
fairseq/fairseq/criterions/tacotron2_loss.py:161
↓ 3 callersMethodcompute_loss_and_acc
(self, model, lprobs, target, reduction='sum')
fairseq/examples/speech_text_joint_to_text/criterions/text_guide_cross_entropy_acc.py:102
↓ 3 callersFunctioncompute_mask_indices
Computes random mask spans for a given shape Args: shape: the the shape for which to compute masks. should be of size 2
data/data_utils.py:437
↓ 3 callersMethodcompute_score_full
(self, L, tau)
fairseq/fairseq/models/nat/insertion_transformer.py:45
↓ 3 callersFunctionconcat_files
(split, src, tgt, extracted_folders, split_urls, path_patterns, to_folder, debug=False)
fairseq/examples/multilingual/data_scripts/download_wmt19_and_before.py:327
↓ 3 callersFunctionconsumer
(tensor)
fairseq/fairseq_cli/preprocess.py:358
↓ 3 callersFunctionconv1x1
1x1 convolution
models/ofa/resnet.py:40
↓ 3 callersFunctionconv3x3
3x3 convolution with padding
models/ofa/resnet.py:34
↓ 3 callersFunctionconvtransformer_espnet
(args)
fairseq/fairseq/models/speech_to_text/convtransformer.py:444
↓ 3 callersFunctioncorpus_bleu
(sys_stream, ref_streams)
fairseq/examples/translation_moe/score.py:112
↓ 3 callersFunctiondataset_dest_prefix
(args, output_prefix, lang)
fairseq/fairseq_cli/preprocess.py:385
↓ 3 callersFunctiondecode_fn
(x)
fairseq/fairseq_cli/generate.py:175
↓ 3 callersMethoddisconnect
(self)
fairseq/fairseq/data/plasma_utils.py:167
↓ 3 callersFunctiondistribution
(dists, window_size)
fairseq/fairseq/modules/kmeans_attention.py:299
↓ 3 callersFunctiondownload_files
(dl_folder, urls, completed_urls={})
fairseq/examples/multilingual/data_scripts/download_wmt19_and_before.py:93
↓ 3 callersFunctionedit_distance2_with_dp
fairseq/fairseq/clib/libnat/edit_dist.cpp:23
↓ 3 callersMethodencode_phone
(self, phone_item)
data/s2t_data/unify_dataset.py:465
↓ 3 callersMethodencode_text
(self, text)
models/clip/model.py:342
↓ 3 callersMethodenergy_from_qk
Compute energy from query and key q_func_value is a tuple looks like (q_proj_func, q_tensor) q_tensor size: bsz, tgt_
fairseq/examples/simultaneous_translation/modules/monotonic_multihead_attention.py:90
↓ 3 callersFunctioneval_f0_error
(samples, distortion_fn)
fairseq/examples/speech_synthesis/evaluation/eval_f0.py:171
↓ 3 callersFunctioneval_metric
(args, hypos, ref)
fairseq/examples/discriminative_reranking_nmt/drnmt_rerank.py:199
↓ 3 callersFunctioneval_str_list
(x, x_type=float)
fairseq/fairseq/dataclass/utils.py:25
↓ 3 callersMethodexists
(prefix_path)
fairseq/fairseq/data/huffman/huffman_mmap_indexed_dataset.py:214
↓ 3 callersMethodexpert_index
(self, i)
fairseq/examples/translation_moe/translation_moe_src/translation_moe.py:139
↓ 3 callersFunctionextend_embedding
( emb: nn.Module, new_vocab_size: int, copy_from_token_id: int )
fairseq/fairseq/tasks/online_backtranslation.py:595
↓ 3 callersMethodextract_features
Similar to *forward* but only return features. Inputs: prev_output_tokens: Tensor(B, T) encoder_out: a dictio
fairseq/fairseq/models/nat/levenshtein_transformer.py:313
↓ 3 callersMethodextract_features
(self, source, get_encoded=False, norm_output=False)
fairseq/examples/textless_nlp/gslm/speech2unit/pretrained/cpc_feature_reader.py:184
↓ 3 callersMethodextract_features_scriptable
Similar to *forward* but only return features. Includes several features from "Jointly Learning to Align and Translate with
fairseq/fairseq/models/transformer/transformer_decoder.py:253
↓ 3 callersFunctionextract_file
(downloaded_file, extract_folder, get_extract_name=get_extract_name, debug=False)
fairseq/examples/multilingual/data_scripts/download_wmt19_and_before.py:131
↓ 3 callersFunctionfasta_file_path
(prefix_path)
fairseq/fairseq/data/fasta_dataset.py:15
↓ 3 callersMethodfilter_indices_by_size
Filter examples that are too large Args: indices (np.array): original array of sample indices dataset (~fair
fairseq/fairseq/tasks/fairseq_task.py:168
↓ 3 callersFunctionfilter_manifest_df
( df, is_train_split=False, extra_filters=None, min_n_frames=5, max_n_frames=3000 )
fairseq/examples/speech_to_text/data_utils.py:244
↓ 3 callersMethodforward
(self)
fairseq/fairseq/sequence_generator.py:744
↓ 3 callersMethodforward
(self, src_tokens, src_lengths=None, **kwargs)
fairseq/examples/speech_text_joint_to_text/models/s2t_dualinputxmtransformer.py:95
↓ 3 callersMethodforward
(self)
models/sequence_generator.py:791
↓ 3 callersMethodforward_attention
Compute attention context vector. Args: value: Transformed value B X n_head X T2 X d_k. scores: Attention score B X n
fairseq/fairseq/modules/espnet_multihead_attention.py:60
↓ 3 callersMethodforward_idx
(self, x)
fairseq/fairseq/modules/gumbel_vector_quantizer.py:137
↓ 3 callersMethodforward_qkv
Transform query, key and value. Args: query: Query tensor B X T1 X C key: Key tensor B X T2 X C value: Va
fairseq/fairseq/modules/espnet_multihead_attention.py:40
↓ 3 callersMethodfrom_file
( cls, filename: str, bos="<s>", pad="<pad>", eos="</s>", unk=
fairseq/fairseq/data/huffman/huffman_coder.py:80
↓ 3 callersMethodgenerate
Generate translations. Match the api of other fairseq generators. Args: models (List[~fairseq.models.FairseqModel]): ensemble of
models/sequence_generator.py:194
↓ 3 callersFunctionget_available_dataset_impl
()
fairseq/fairseq/data/indexed_dataset.py:38
↓ 3 callersMethodget_batch_iterator
( self, dataset, max_tokens=None, max_sentences=None, max_positions=No
fairseq/examples/laser/laser_src/laser_task.py:270
↓ 3 callersMethodget_batch_iterator
( self, dataset, max_tokens=None, max_sentences=None, max_positions=No
fairseq/examples/speech_text_joint_to_text/tasks/speech_text_joint.py:314
↓ 3 callersMethodget_batch_iterator
( self, dataset, max_tokens=None, max_sentences=None, max_positions=No
tasks/ofa_task.py:113
↓ 3 callersMethodget_cache_size
(self)
fairseq/examples/adaptive_span/adaptive_span_model.py:90
↓ 3 callersMethodget_decoder_langtok
(self, tgt_lang, spec=None)
fairseq/fairseq/data/multilingual/multilingual_data_manager.py:483
↓ 3 callersFunctionget_dummy_task_and_parser
to build a fariseq model, we need some dummy parse and task. This function is used to create dummy task and parser to faciliate model/criteri
fairseq/tests/test_sequence_generator.py:52
↓ 3 callersMethodget_encoder_langtok
(self, src_lang, tgt_lang, spec=None)
fairseq/fairseq/data/multilingual/multilingual_data_manager.py:464
↓ 3 callersFunctionget_extract_name
(file_path)
fairseq/examples/multilingual/data_scripts/download_wmt19_and_before.py:127
↓ 3 callersFunctionget_global_group
()
fairseq/fairseq/distributed/utils.py:439
↓ 3 callersMethodget_image_rel_pos_bias
(self, image_position_ids, idx)
models/ofa/unify_transformer.py:732
↓ 3 callersMethodget_input_path
(self, fname=None)
fairseq/examples/wav2vec/wav2vec_featurize.py:168
↓ 3 callersFunctionget_mask_from_lengths
(lengths)
fairseq/examples/textless_nlp/gslm/unit2speech/tacotron2/utils.py:13
↓ 3 callersMethodget_normalized_probs
( self, net_output: Tuple[ torch.Tensor, Optional[Dict[str, List[Optional[torch.Te
fairseq/examples/wav2vec/unsupervised/models/wav2vec_u.py:478
↓ 3 callersMethodget_original_text
(self, i)
fairseq/fairseq/data/indexed_dataset.py:300
↓ 3 callersFunctionget_path_iterator
(tsv, nshard, rank)
fairseq/examples/hubert/simple_kmeans/feature_utils.py:35
↓ 3 callersFunctionget_prefix_no_bpe
(sentence, bpe_symbol, prefix_len)
fairseq/examples/noisychannel/rerank_utils.py:216
↓ 3 callersMethodget_size_ratios
Size ratios for temperature-based sampling (https://arxiv.org/abs/1907.05019)
fairseq/fairseq/data/audio/speech_to_text_dataset.py:422
↓ 3 callersMethodget_smoothed_value
Get a single smoothed value.
fairseq/fairseq/logging/meters.py:291
↓ 3 callersMethodget_speaker_embeddings
(cls, args)
fairseq/fairseq/tasks/text_to_speech.py:98
↓ 3 callersMethodget_symbols
(self, i)
fairseq/fairseq/data/huffman/huffman_mmap_indexed_dataset.py:196
↓ 3 callersFunctionget_symbols_to_strip_from_output
(generator)
fairseq/fairseq_cli/generate.py:52
↓ 3 callersMethodget_target
(self, item)
fairseq/examples/wav2vec/unsupervised/data/random_input_dataset.py:30
↓ 3 callersMethodget_tokens
Normalize tokens by handling CTC blank, ASG replabels, etc.
fairseq/examples/speech_recognition/w2l_decoder.py:90
↓ 3 callersMethodget_total_row_count
(self)
data/file_dataset.py:96
↓ 3 callersMethodget_train_iterator
Return an EpochBatchIterator over the training set for a given epoch.
trainer.py:628
↓ 3 callersMethodget_train_iterator
Return an EpochBatchIterator over the training set for a given epoch.
fairseq/fairseq/trainer.py:612
↓ 3 callersFunctionget_whole_word_mask
(args, dictionary)
fairseq/fairseq/data/encoders/utils.py:10
↓ 3 callersMethodhas_encoder
(self)
fairseq/fairseq/sequence_generator.py:747
↓ 3 callersMethodhas_encoder
(self)
models/sequence_generator.py:794
↓ 3 callersMethodhas_sharded_data
(self, split)
fairseq/fairseq/tasks/fairseq_task.py:130
↓ 3 callersMethodhead_reshape
(self, x)
fairseq/examples/adaptive_span/adaptive_span_model.py:110
↓ 3 callersFunctionimport_pipe
()
fairseq/fairseq/model_parallel/models/pipeline_parallel_transformer/model.py:45
↓ 3 callersMethodincrement
(self, symbol: str, cnt: int)
fairseq/fairseq/data/huffman/huffman_coder.py:183
↓ 3 callersMethodinference
(self, x)
fairseq/examples/textless_nlp/gslm/unit2speech/tacotron2/model.py:193
↓ 3 callersFunctionis_sf_audio_data
(data: bytes)
fairseq/fairseq/data/audio/audio_utils.py:174
↓ 3 callersMethodjoint_forward
(self, x)
fairseq/examples/discriminative_reranking_nmt/models/discriminative_reranking_model.py:263
↓ 3 callersFunctionlabel_smoothed_nll_loss
(lprobs, target, epsilon, ignore_index=None, reduce=True)
fairseq/fairseq/criterions/label_smoothed_cross_entropy.py:33
↓ 3 callersMethodload_model
(self)
fairseq/examples/wav2vec/vq-wav2vec_featurize.py:212
↓ 3 callersFunctionload_tra
(tra_path)
fairseq/examples/wav2vec/unsupervised/kaldi_self_train/st/local/unsup_select.py:32
↓ 3 callersFunctionlowercase
(text)
fairseq/examples/textless_nlp/gslm/unit2speech/tacotron2/cleaners.py:56
↓ 3 callersMethodlr_step
Adjust the learning rate at the end of the epoch.
trainer.py:1110
↓ 3 callersMethodlr_step
Adjust the learning rate at the end of the epoch.
fairseq/fairseq/trainer.py:1088
↓ 3 callersMethodlr_step_update
Update the learning rate after each update.
trainer.py:1116
↓ 3 callersMethodlr_step_update
Update the learning rate after each update.
fairseq/fairseq/trainer.py:1094
↓ 3 callersFunctionmake_binary_alignment_dataset
(input_prefix, output_prefix, num_workers)
fairseq/fairseq_cli/preprocess.py:198
↓ 3 callersMethodmake_dataset
(type, dictionary)
fairseq/fairseq/tasks/sentence_ranking.py:109
↓ 3 callersFunctionmake_lm_config
( data_dir=None, extra_flags=None, task="language_modeling", arch="transformer_lm_gpt2_tiny",
fairseq/tests/test_valid_subset_checks.py:12
↓ 3 callersMethodmask
Helper to get index of mask symbol
fairseq/fairseq/data/legacy/masked_lm_dictionary.py:27
↓ 3 callersMethodmax_decoder_positions
(self)
fairseq/fairseq/sequence_generator.py:753
↓ 3 callersFunctionmerge
(key)
data/pretrain_data/unify_dataset.py:59
↓ 3 callersFunctionmerge
(key)
data/nlg_data/summary_dataset.py:22
↓ 3 callersFunctionmerge
(key)
data/mm_data/caption_dataset.py:37
↓ 3 callersFunctionmerge
(key, left_pad, move_eos_to_beginning=False)
data/mm_data/image_gen_dataset.py:43
↓ 3 callersFunctionmerge
(key)
data/mm_data/refcoco_dataset.py:36
↓ 3 callersMethodmult_rst_grad
(self, rst, ratio)
fairseq/examples/speech_text_joint_to_text/models/s2t_dualinputtransformer.py:276
↓ 3 callersMethodno_sync
A context manager to disable gradient synchronization.
fairseq/fairseq/distributed/legacy_distributed_data_parallel.py:66
← previousnext →601–700 of 6,938, ranked by callers