Functions — github.com/NUSTM/VLP-MABSA

Method__init__

( self, input_dim, inner_dim, num_classes, pooler_

src/model/modeling_bart.py:842

Method__init__

(self, num_embeddings: int, embedding_dim: int, padding_idx: int, offset)

src/model/modeling_bart.py:870

Method__init__

(self, num_embeddings: int, embedding_dim: int, padding_idx

src/model/modeling_bart.py:903

Method__init__

(self, num_embeddings: int, embedding_dim: int, padding_idx

src/model/modeling_bart.py:985

Method__init__

(self, config: BartConfig)

src/model/modeling_bart.py:1095

Method__init__

(self, config: BartConfig)

src/model/modeling_bart.py:1224

Method__init__

(self, config: BartConfig, **kwargs)

src/model/modeling_bart.py:1410

Method__init__

(self, config)

src/model/modeling_bart.py:1500

Method__init__

(self, num_positions, embedding_dim, padding_idx=None)

src/model/modeling_bart.py:1605

Method__init__

(self, eos_token_id, num_labels, conflict_id,

src/model/metrics.py:7

Method__init__

(self, image_dim, final_dim)

src/model/modules.py:23

Method__init__

(self, config: MultiModalBartConfig, encoder, img_feat_id, cls_token_id)

src/model/modules.py:55

Method__init__

(self, config: MultiModalBartConfig, tokenizer, decoder,

src/model/modules.py:173

Method__init__

(self, config: MultiModalBartConfig, decoder)

src/model/modules.py:339

Method__init__

(self, config: MultiModalBartConfig, decoder)

src/model/modules.py:381

Method__init__

(self, config: MultiModalBartConfig, decoder, senti_ids,

src/model/modules.py:425

Method__init__

(self, config: MultiModalBartConfig, decoder, causal_mask, args)

src/model/modules.py:468

Method__init__

:param Seq2SeqModel seq2seq_model: 序列到序列模型. 会使用seq2seq_model的decoder进行生成 :param int,None bos_token_id: 句子开头的token id :param

src/model/generater.py:17

Method__init__

:param Seq2SeqDecoder decoder: Decoder对象 :param int max_length: 生成句子的最大长度, 每句话的decode长度为max_length + max_len_a*src_len :para

src/model/generater.py:123

Method__init__

(self, config: MultiModalBartConfig, args, bart_model, tokenizer, label_ids)

src/model/MAESC_model.py:72

Method__init__

(self, config: MultiModalBartConfig, bart_model, tokenizer, label_ids, senti_ids, args)

src/model/model.py:80

Method__init__

(self, args, pretrained_model_name='facebook/bart-base', cl

src/data/tokenization_new.py:20

Method__init__

(self, infos)

src/data/dataset.py:13

Method__init__

(self, infos, split)

src/data/dataset.py:125

Method__init__

:param tokenizer: ConditionTokenizer :param mlm_enabled: bool, if use mlm for language modeling. False for autoregressive modeling

src/data/collation.py:14

Method__len__

Number of hypotheses in the list.

src/model/generater.py:759

Method__len__

(self)

src/data/tokenization_new.py:676

Method__len__

(self)

src/data/dataset.py:34

Method__len__

(self)

src/data/dataset.py:143

Function_check_shapes

(shape_1, shape2)

src/model/modeling_bart.py:206

Method_clip_text

(self, text, length)

src/data/collation.py:69

Function_get_shape

(t)

src/model/modeling_bart.py:1086

Method_init_weights

(self, module)

src/model/modules.py:439

Method_reorder_cache

(past, beam_idx)

src/model/modeling_bart.py:1387

Method_reorder_cache

(past, beam_idx)

src/model/mixins.py:420

Methodadjust_logits_during_generation

(self, logits, cur_len, max_length)

src/model/modeling_bart.py:1374

Methodadjust_logits_during_generation

(self, logits, cur_len, max_length)

src/model/mixins.py:400

Functionclip_gradient

(optimizer, grad_clip)

src/model/utils.py:16

Functioncmp

(v1, v2)

src/data/tokenization_new.py:9

Methodconvert_ids_to_tokens

(self, ids)

src/data/tokenization_new.py:670

Methoddummy_inputs

(self)

src/model/modeling_bart.py:187

Methodencode_anp_dis

(self, batch_size)

src/data/tokenization_new.py:336

Methodencode_mlm_sentence

(self, labels)

src/data/tokenization_new.py:173

Functioneval

(args, model, loader, metric, device)

src/eval_utils.py:5

Methodevaluate

(self, aesc_target_span, pred, tgt_tokens)

src/model/metrics.py:34

Methodforward

Args: x (Tensor): input to the layer of shape `(seq_len, batch, embed_dim)` encoder_padding_mask (ByteTensor): binary

src/model/modeling_bart.py:248

Methodforward

Args: input_ids (LongTensor): tokens in the source language of shape `(batch, src_len)` attention_mas

src/model/modeling_bart.py:327

Methodforward

( self, x, encoder_hidden_states, encoder_attn_mask=None,

src/model/modeling_bart.py:426

Methodforward

Includes several features from "Jointly Learning to Align and Translate with Transformer Models" (Garg et al., EMNLP 2019).

src/model/modeling_bart.py:541

Methodforward

(self, x)

src/model/modeling_bart.py:854

Methodforward

Input is expected to be of size [bsz x seqlen].

src/model/modeling_bart.py:881

Methodforward

Input is expected to be of size [bsz x seqlen].

src/model/modeling_bart.py:919

Methodforward

Input is expected to be of size [bsz x seqlen].

src/model/modeling_bart.py:1001

Methodforward

( self, input_ids, attention_mask=None, decoder_input_ids=None

src/model/modeling_bart.py:1113

Methodforward

r""" labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size, sequence_length)`, `optional`): Labels for computing the masked l

src/model/modeling_bart.py:1253

Methodforward

r""" labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size,)`, `optional`): Labels for computing the sequence classification/

src/model/modeling_bart.py:1429

Methodforward

r""" start_positions (:obj:`torch.LongTensor` of shape :obj:`(batch_size,)`, `optional`): Labels for position (index) of the start

src/model/modeling_bart.py:1518

Methodforward

Input is expected to be of size [bsz x seqlen].

src/model/modeling_bart.py:1629

Methodforward

(self, image_features)

src/model/modules.py:27

Methodforward

:param input_ids: LongTensor, tokens in the source language of shape (batch, src_len) :param image_features: list[FloatTensor], imag

src/model/modules.py:98

Methodforward

(self, tokens, state, only_sc=False)

src/model/modules.py:208

Methodforward

(self, labels, input_ids, encoder_outputs, attention_mask, decoder_input_ids, decoder_attentio

src/model/modules.py:347

Methodforward

(self, labels, input_ids, encoder_outputs, attention_mask, decoder_input_ids, decoder_attentio

src/model/modules.py:389

Methodforward

(self, senti_labels, encoder_outputs, attention_mask, senti_decoder_input_ids)

src/model/modules.py:444

Methodforward

(self, mrm_labels, mrm_masks, encoder_outputs, attention_mask, mrm_decoder_input_ids, mrm_deco

src/model/modules.py:489

Methodforward

透传调用seq2seq_model的forward :param torch.LongTensor src_tokens: bsz x max_len :param torch.LongTensor tgt_tokens: bsz x max_le

src/model/generater.py:64

Methodforward

( self, input_ids, image_features, attention_mask=None,

src/model/MAESC_model.py:124

Methodforward

( self, task_type, input_ids, image_features, atte

src/model/model.py:125

Methodgenerate

Generates sequences for models with a LM head. The method currently supports greedy decoding, beam-search decoding, sampling with tem

src/model/mixins.py:33

Methodget_encoder

(self)

src/model/modeling_bart.py:1398

Methodget_input_embeddings

(self)

src/model/modeling_bart.py:1203

Functionget_lr

(optimizer)

src/model/utils.py:11

Methodget_metric

(self, reset=True)

src/model/metrics.py:140

Methodget_output_embeddings

(self)

src/model/modeling_bart.py:1211

Methodget_output_embeddings

(self)

src/model/modeling_bart.py:1401

Methodget_output_embeddings

(self)

src/model/mixins.py:439

Functiongreedy_generate

贪婪地搜索句子 :param Decoder decoder: Decoder对象 :param torch.LongTensor tokens: batch_size x len, decode的输入值，如果为None，则自动从bos_token_id开始生成

src/model/generater.py:223

Methodline

(self)

src/utils.py:55

Functionliner_warmup

(cur_step, t_step, warmup)

src/model/utils.py:25

Functionload_training_data

(path, optimizer=None, scaler=None, map_location=None)

src/utils.py:30

Methodprepare_inputs_for_generation

(self, decoder_input_ids, past, attention_mask, use_cache,

src/model/modeling_bart.py:1360

Methodprepare_inputs_for_generation

(self, decoder_input_ids, past, attention_mask, use_cache, **kwargs)

src/model/mixins.py:386

Methodpri

(self)

src/model/metrics.py:136

Methodreorder_state

(self, indices: torch.LongTensor)

src/model/MAESC_model.py:158

Methodresize_token_embeddings

(self, new_num_tokens: int)

src/model/modeling_bart.py:1232

Functionsample_sentence

(model, input_ids, image_features, attention_mask,

src/model/utils.py:32

Methodset_input_embeddings

(self, value)

src/model/modeling_bart.py:1206

Functionset_lr

(optimizer, lr)

src/model/utils.py:6

Methodset_new_generator

(self, max_length=-1, max_len_a=-1,

src/model/generater.py:176

Methodset_position_embedding

(self, special_tag_start_id, tag_first=True)

src/model/modeling_bart.py:526

Functionsetup_process

(rank, world_size, master_port='12355')

src/utils.py:9

Functions191 in github.com/NUSTM/VLP-MABSA