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
↓ 4 callersFunctionbase_architecture
(args)
fairseq/fairseq/models/speech_to_text/s2t_transformer.py:407
↓ 4 callersFunctionbase_architecture
(args)
fairseq/fairseq/models/speech_to_text/xm_transformer.py:502
↓ 4 callersMethodbatch_by_size
Given an ordered set of indices, return batches according to *max_tokens*, *max_sentences* and *required_batch_size_multiple*.
fairseq/fairseq/data/fairseq_dataset.py:104
↓ 4 callersFunctionbatched_index_select
(values, indices)
fairseq/fairseq/modules/kmeans_attention.py:69
↓ 4 callersMethodbinarize_with_mask
(self, txt, prefix, suffix, leading_space, trailing_space)
fairseq/examples/roberta/wsc/wsc_task.py:99
↓ 4 callersFunctionbleu_addngram
fairseq/fairseq/clib/libbleu/libbleu.cpp:73
↓ 4 callersMethodbmuf_process
(self, cfg, args, iterations)
fairseq/tests/distributed/test_bmuf.py:144
↓ 4 callersFunctionbuffered_arange
(max)
fairseq/fairseq/utils.py:271
↓ 4 callersFunctionbuild_dataset
(prefix, data, coder)
fairseq/tests/test_huffman.py:109
↓ 4 callersMethodbuild_model
Build a new model instance.
fairseq/examples/speech_recognition/models/vggtransformer.py:144
↓ 4 callersMethodbuild_model
(self, cfg: FairseqDataclass)
tasks/ofa_task.py:159
↓ 4 callersFunctioncalc_length_from_frac
(bpe_sentence, prefix_frac, bpe_symbol)
fairseq/examples/noisychannel/rerank_utils.py:196
↓ 4 callersFunctioncall
(cmd)
fairseq/examples/multilingual/data_scripts/binarize.py:16
↓ 4 callersFunctioncoordinate_to_offset
(row, col, ncols)
fairseq/examples/speech_recognition/utils/wer_utils.py:47
↓ 4 callersMethoddecode
(self, quant)
models/taming/models/vqgan.py:60
↓ 4 callersFunctiondefault_architecture
(args)
fairseq/fairseq/models/huggingface/hf_gpt2.py:135
↓ 4 callersFunctiondictolist
(d)
fairseq/examples/translation_moe/score.py:51
↓ 4 callersFunctiondump_feature
(reader, generator, num, split, nshard, rank, feat_dir)
fairseq/examples/hubert/simple_kmeans/feature_utils.py:48
↓ 4 callersMethoddump_images
(self, images, text, path, topk=None)
tasks/mm_tasks/image_gen.py:323
↓ 4 callersFunctionemulate_int
(w, bits, method, scale=None, zero_point=None)
fairseq/fairseq/modules/quantization/scalar/ops.py:9
↓ 4 callersMethodend_of_epoch
Returns whether the most recent epoch iterator has been exhausted
fairseq/fairseq/data/iterators.py:378
↓ 4 callersFunctionenhance_level_to_args
(MAX_LEVEL)
utils/vision_helper.py:212
↓ 4 callersFunctioneval_str_dict
(x, type=dict)
fairseq/fairseq/utils.py:785
↓ 4 callersFunctionextract_fbank_features
( waveform: torch.FloatTensor, sample_rate: int, output_path: Optional[Path] = None, n_mel_bin
fairseq/examples/speech_to_text/data_utils.py:73
↓ 4 callersMethodextract_features
Similar to *forward* but only return features. Includes several features from "Jointly Learning to Align and Translate with
fairseq/examples/laser/laser_src/laser_transformer.py:179
↓ 4 callersFunctionfast_conv
Faster convolution evaluation if either kernel size is 1 or length of sequence is 1.
fairseq/examples/speech_synthesis/preprocessing/denoiser/demucs.py:197
↓ 4 callersMethodfrom_tsv
( cls, root: str, cfg: S2TDataConfig, splits: str, tgt_dict, p
fairseq/fairseq/data/audio/speech_to_text_dataset.py:493
↓ 4 callersFunctionget_args
()
fairseq/examples/textless_nlp/gslm/metrics/asr_metrics/ppx.py:40
↓ 4 callersMethodget_decay
(self)
fairseq/fairseq/models/ema/ema.py:131
↓ 4 callersFunctionget_downloaded_file
(dl_folder, url)
fairseq/examples/multilingual/data_scripts/download_wmt19_and_before.py:52
↓ 4 callersFunctionget_encoder
(encoder_json_path, vocab_bpe_path)
fairseq/fairseq/data/encoders/gpt2_bpe_utils.py:131
↓ 4 callersFunctionget_features
( feature_type, checkpoint_path, layer, manifest_path, sample_pct, flatten )
fairseq/examples/textless_nlp/gslm/speech2unit/pretrained/utils.py:70
↓ 4 callersFunctionget_features_or_waveform
Get speech features from .npy file or waveform from .wav/.flac file. The file may be inside an uncompressed ZIP file and is accessed via byte
fairseq/fairseq/data/audio/speech_to_text_dataset.py:65
↓ 4 callersFunctionget_lang_tok
( lang: str, lang_tok_style: str, spec: str = LangTokSpec.main.value )
fairseq/fairseq/data/multilingual/multilingual_utils.py:29
↓ 4 callersMethodget_logits
(self, net_output, is_masked=True)
fairseq/fairseq/models/hubert/hubert.py:536
↓ 4 callersMethodget_logits
( self, net_output: Optional[Dict[str, List[Optional[torch.Tensor]]]], normalize: bool
fairseq/examples/wav2vec/unsupervised/models/wav2vec_u.py:453
↓ 4 callersMethodget_lprobs
(self, model, tokens, mask)
fairseq/examples/roberta/wsc/wsc_criterion.py:49
↓ 4 callersFunctionget_mock_cfg
(finetune_from_model)
fairseq/tests/test_train.py:60
↓ 4 callersMethodget_normalized_probs
Get normalized probabilities (or log probs) from a net's output.
fairseq/fairseq/models/roberta/model.py:230
↓ 4 callersMethodget_normalized_probs
(self, net_output, log_probs, sample=None)
fairseq/tests/speech_recognition/asr_test_base.py:488
↓ 4 callersMethodget_num_updates
Get the number of parameters updates.
fairseq/fairseq/optim/bmuf.py:138
↓ 4 callersMethodget_pos_info
( self, tokens, tgt_pos_embed, src_pos_embed=None,
models/ofa/unify_transformer.py:1465
↓ 4 callersFunctionget_res_file
(file_prefix)
fairseq/examples/speech_recognition/infer.py:159
↓ 4 callersMethodget_res_file
(self, fname: str)
fairseq/examples/speech_recognition/new/infer.py:158
↓ 4 callersFunctionget_self_bleu
(utterances, averaging_mode, weights)
fairseq/examples/textless_nlp/gslm/metrics/asr_metrics/self_auto_bleu.py:61
↓ 4 callersMethodget_source_dictionary
(self, lang)
fairseq/fairseq/data/multilingual/multilingual_data_manager.py:435
↓ 4 callersFunctionget_time_gap
(s, e)
fairseq/fairseq/tasks/translation_multi_simple_epoch.py:27
↓ 4 callersMethodload_dataset
Load a given dataset split. Args: split (str): name of the split (e.g., train, valid, test)
fairseq/examples/speech_recognition/tasks/speech_recognition.py:117
↓ 4 callersMethodload_dataset
(self, split, epoch=1, combine=False, **kwargs)
tasks/mm_tasks/ocr.py:43
↓ 4 callersFunctionload_tsv_to_dicts
(path: Union[str, Path])
fairseq/examples/speech_to_text/data_utils.py:230
↓ 4 callersMethodmax_positions
(self)
fairseq/examples/discriminative_reranking_nmt/tasks/discriminative_reranking_task.py:332
↓ 4 callersMethodmaybe_layer_norm
(self, layer_norm, x, before=False, after=False)
fairseq/fairseq/model_parallel/models/pipeline_parallel_transformer/layers.py:386
↓ 4 callersMethodmaybe_layer_norm
(self, i, x, before=False, after=False)
fairseq/fairseq/models/lightconv.py:710
↓ 4 callersFunctionmerge
(key, left_pad, move_eos_to_beginning=False)
data/s2t_data/unify_dataset.py:58
↓ 4 callersFunctionmerge
(key)
data/cv_data/image_classify_dataset.py:38
↓ 4 callersFunctionmerge
(key)
data/nlu_data/sst2_dataset.py:22
↓ 4 callersFunctionmerge
(key)
data/nlu_data/qqp_dataset.py:22
↓ 4 callersFunctionmerge
(key)
data/nlu_data/rte_dataset.py:22
↓ 4 callersFunctionmerge
(key)
data/nlu_data/cola_dataset.py:22
↓ 4 callersFunctionmerge
(key)
data/nlu_data/mnli_dataset.py:22
↓ 4 callersFunctionmerge
(key)
data/nlu_data/mrpc_dataset.py:22
↓ 4 callersFunctionmerge
(key)
data/nlu_data/qnli_dataset.py:22
↓ 4 callersFunctionmerge
(key)
data/mm_data/snli_ve_dataset.py:36
↓ 4 callersFunctionmerge
(key, left_pad, move_eos_to_beginning=False, pad_to_length=None)
fairseq/fairseq/data/language_pair_dataset.py:29
↓ 4 callersFunctionmerge_result
(worker_result)
fairseq/fairseq_cli/preprocess.py:130
↓ 4 callersFunctionmerge_with_parent
(dc: FairseqDataclass, cfg: DictConfig, remove_missing=True)
fairseq/fairseq/dataclass/utils.py:477
↓ 4 callersFunctionmk_dataset
(num_samples: int, max_len: int, output: Path)
fairseq/tests/test_online_backtranslation.py:31
↓ 4 callersMethodnoising
(self, x, lengths, noising_prob=0.0)
fairseq/fairseq/data/noising.py:36
↓ 4 callersMethodnoising
(self, x, lengths, dropout_prob=None, blank_idx=None)
fairseq/fairseq/data/noising.py:85
↓ 4 callersMethodobt_task
( cls, languages: Sequence[str], data: Path = None, language_mapping: str = None )
fairseq/tests/test_online_backtranslation.py:49
↓ 4 callersFunctionofa_large_architecture
(args)
models/ofa/ofa.py:321
↓ 4 callersMethodordered_indices
(self)
fairseq/fairseq/data/sort_dataset.py:20
↓ 4 callersMethodoutput_layer
(self, features, masked_tokens=None, **unused)
fairseq/fairseq/models/roberta/model.py:515
↓ 4 callersFunctionpack_constraints
Takes a list of list of constraints in tensor form (a list of tensor constraints for each sentence) and transforms it into a packed Tensor. Fo
fairseq/fairseq/token_generation_constraints.py:41
↓ 4 callersMethodpath_requires_pathmanager
Do we require PathManager to access given path?
fairseq/fairseq/file_io.py:135
↓ 4 callersFunctionplot_tts_output
( data_2d, title_2d, data_1d, title_1d, figsize=(24, 4), v_min=DEFAULT_V_MIN, v_max=3, ret_np=
fairseq/fairseq/tasks/text_to_speech.py:263
↓ 4 callersMethodprefetch
(self, indices)
fairseq/fairseq/data/noising.py:333
↓ 4 callersFunctionreadlines
(fd)
fairseq/fairseq_cli/score.py:49
↓ 4 callersMethodregister_classification_head
Register a classification head.
fairseq/fairseq/models/bart/model.py:140
↓ 4 callersMethodreorder_encoder_out
Reorder encoder output according to *new_order*. Args: encoder_out: output from the ``forward()`` method new
fairseq/fairseq/sequence_generator.py:830
↓ 4 callersMethodrerank
(self, reranker, finalized, encoder_input, beam_size)
fairseq/fairseq/iterative_refinement_generator.py:313
↓ 4 callersMethodreset
(self)
fairseq/fairseq/logging/meters.py:47
↓ 4 callersMethodresult_string
(self, order=4)
fairseq/fairseq/scoring/bleu.py:152
↓ 4 callersMethodsample_negatives
(self, y, num, padding_count=None)
fairseq/fairseq/models/wav2vec/wav2vec2.py:484
↓ 4 callersMethodscale
(self, outputs)
fairseq/fairseq/optim/dynamic_loss_scaler.py:28
↓ 4 callersMethodsetUp
(self)
fairseq/tests/speech_recognition/test_vggtransformer.py:126
↓ 4 callersMethodset_specaugment
( self, time_wrap_w: int, freq_mask_n: int, freq_mask_f: int, time_mas
fairseq/examples/speech_to_text/data_utils.py:300
↓ 4 callersFunctionsetup_args
()
fairseq/tests/distributed/test_bmuf.py:89
↓ 4 callersMethodsetup_task
Setup the task (e.g., load dictionaries). Args: cfg (AudioPretrainingConfig): configuration of this task
fairseq/examples/wav2vec/unsupervised/tasks/unpaired_audio_text.py:147
↓ 4 callersMethodskip
Fast-forward the iterator by skipping n elements.
fairseq/fairseq/data/iterators.py:67
↓ 4 callersMethodstop_timeout
(self)
fairseq/fairseq/distributed/distributed_timeout_wrapper.py:65
↓ 4 callersFunctionstr2toks
(str)
fairseq/examples/speech_recognition/utils/wer_utils.py:63
↓ 4 callersFunctionstrip_pronoun
(x)
fairseq/examples/roberta/wsc/wsc_utils.py:117
↓ 4 callersMethodtokenize
(self, sent: str)
fairseq/fairseq/scoring/tokenizer.py:53
↓ 4 callersFunctiontransformer_vaswani_wmt_en_de_big
(args)
fairseq/fairseq/models/transformer/transformer_legacy.py:244
↓ 4 callersMethodupdate
(self)
fairseq/fairseq/optim/dynamic_loss_scaler.py:31
↓ 4 callersMethodzero_grad
(self)
trainer.py:1098
↓ 3 callersFunctionEmbedding
(num_embeddings, embedding_dim, padding_idx=None)
fairseq/fairseq/models/text_to_speech/fastspeech2.py:28
↓ 3 callersFunctionEmbedding
(num_embeddings, embedding_dim, padding_idx)
fairseq/examples/laser/laser_src/laser_lstm.py:531
↓ 3 callersFunctionGatedLinear
Weight-normalized Linear layer (input: B x T x C) with interspersed GLU units
fairseq/fairseq/modules/downsampled_multihead_attention.py:308
← previousnext →401–500 of 6,938, ranked by callers