MCPcopy Create free account

hub / github.com/DevSinghSachan/syntax-augmented-bert / functions

Functions302 in github.com/DevSinghSachan/syntax-augmented-bert

⨍ Functions 302◇ Types & classes 67
Method_read
(self, file_path: str)
dataset_reader/ontonotes_ner_reader.py:55
Method_resize_token_embeddings
(self, new_num_tokens)
model/syntax_bert.py:324
Method_set_loss_scale
(self, value)
fp16/fp16_opt.py:605
Method_set_param_groups
(self, value)
fp16/fp16_opt.py:624
Method_set_state
(self, value)
fp16/fp16_opt.py:614
Function_truncate_seq_pair
Truncates a sequence pair in place to the maximum length.
utils/utils.py:359
Methodbackward
:attr:`backward` performs the following conceptual steps: 1. fp32_loss = loss.float() (see first Note below) 2. scaled_loss
fp16/fp16_opt.py:492
Methodbackward
(self, loss, retain_graph=False)
fp16/loss_scaler.py:167
Functionbackwards_debug_hook
(grad)
fp16/fp16util.py:100
Functioncheck_no_decay
(var_name)
main.py:66
Methodclip_master_grads
Clips fp32 master gradients via ``torch.nn.utils.clip_grad_norm``. Args: max_norm (float or int): max norm of the gradie
fp16/fp16_opt.py:310
Functioncompute_metrics
(task_name, preds, labels)
utils/utils.py:400
Methoddataset_iterator
An iterator over the entire dataset, yielding all sentences processed.
dataset_reader/conll2005_utils.py:100
Methoddataset_iterator
An iterator over the entire dataset, yielding all sentences processed.
dataset_reader/ontonotes_utils.py:184
Methoddepth
(self)
model/tree.py:34
Functionfloat_conversion
(val)
fp16/fp16_opt.py:50
Methodforward
output (FloatTensor): batch_size x n_classes target (LongTensor): batch_size
model/loss.py:20
Methodforward
(self, top_layer_output, labels, sent_len)
model/loss.py:36
Methodforward
(self, input)
model/graph_encoder.py:77
Methodforward
(self, input_ids, token_type_ids=None, attention_mask=None, labels=None, dep_head=None, dep_re
model/graph_encoder.py:92
Methodforward
(self, input_ids_or_bert_hidden, adj=None, dep_rel_matrix=None, wp_seq_lengths=None)
model/graph_encoder.py:204
Methodforward
(self, e, attention_mask, dep_rel_matrix=None)
model/graph_encoder.py:262
Methodforward
(self, e, attention_mask, dep_rel_matrix=None)
model/graph_encoder.py:292
Methodforward
(self, e)
model/graph_encoder.py:326
Methodforward
(self, hidden_states, attention_mask, dep_rel_matrix=None)
model/graph_encoder.py:360
Methodforward
(self, x, y)
model/gate.py:10
Methodforward
(self, x, y)
model/gate.py:20
Methodforward
(self, x, y)
model/gate.py:31
Methodforward
(self, x, y)
model/gate.py:47
Methodforward
(self, x, y)
model/gate.py:62
Methodforward
(self, feats)
model/crf.py:206
Methodforward
(self, input_ids, token_type_ids=None, position_ids=None)
model/syntax_roberta.py:109
Methodforward
(self, input_ids, token_type_ids=None, position_ids=None)
model/syntax_roberta.py:138
Methodforward
(self, input_ids, token_type_ids=None, wp_token_mask=None, labels=None, dep_head=None, dep_rel
model/syntax_roberta.py:213
Methodforward
(self, h, z, attention_mask, attention_probs_dropout_prob=0.1)
model/syntax_bert.py:100
Methodforward
(self, h, z, attention_mask, attention_probs_dropout_prob=0.1)
model/syntax_bert.py:154
Methodforward
(self, input_tensor, z, attention_mask)
model/syntax_bert.py:176
Methodforward
(self, hidden_states, z, attention_mask)
model/syntax_bert.py:194
Methodforward
(self, hidden_states, z, attention_mask)
model/syntax_bert.py:214
Methodforward
(self, hidden_states, token_mask=None, subj_pos=None, obj_pos=None)
model/syntax_bert.py:287
Methodforward
(self, input_ids, token_type_ids=None, wp_token_mask=None, dep_head=None, dep_
model/syntax_bert.py:391
Methodforward
(self, input_ids, token_type_ids=None, wp_token_mask=None, labels=None, dep_head=None, dep_rel
model/syntax_bert.py:461
Methodforward
(self, input_ids, token_type_ids=None, wp_token_mask=None, labels=None, dep_head=None, dep_rel
model/syntax_bert.py:512
Methodforward
(self, *inputs, **kwargs)
fp16/fp16_opt.py:64
Methodforward
(self, input)
fp16/fp16util.py:33
Methodforward
(self, *inputs)
fp16/fp16util.py:95
Methodget_aligned_idxs
(ent_pos)
utils/loader.py:188
Methodget_dev_examples
See base class.
utils/utils.py:113
Methodget_dev_examples
See base class.
utils/utils.py:152
Methodget_dev_examples
See base class.
utils/utils.py:206
Methodget_dev_examples
See base class.
utils/utils.py:267
Methodget_labels
See base class.
utils/utils.py:123
Methodget_labels
See base class.
utils/utils.py:160
Methodget_labels
See base class.
utils/utils.py:216
Methodget_labels
See base class.
utils/utils.py:277
Methodget_labels
See base class.
utils/utils.py:321
Methodget_overall_list
(self)
utils/utils.py:701
Methodget_stats
(self, **kwargs)
utils/utils.py:441
Methodget_stats
Returns ------- A Dict per label containing following the span based metrics: precision : float recall : floa
utils/utils.py:704
Methodget_test_examples
See base class.
utils/utils.py:156
Methodget_test_examples
See base class.
utils/utils.py:211
Methodget_test_examples
See base class.
utils/utils.py:272
Methodget_test_examples
See base class.
utils/utils.py:317
Methodget_test_examples
See base class.
utils/utils.py:354
Methodget_train_examples
See base class.
utils/utils.py:108
Methodget_train_examples
See base class.
utils/utils.py:148
Methodget_train_examples
See base class.
utils/utils.py:201
Methodget_train_examples
See base class.
utils/utils.py:262
Functionhalf_conversion
(val)
fp16/fp16_opt.py:39
Methodhas_overflow
(self, params)
fp16/loss_scaler.py:106
Methodinspect_master_grad_data
When running with :class:`FP16_Optimizer`, ``.grad`` attributes of a model's fp16 leaves should not be regarded as truthful,
fp16/fp16_opt.py:567
Methodload_state_dict
?
utils/optimizers.py:233
Methodload_state_dict
Loads a state_dict created by an earlier call to state_dict(). If ``fp16_optimizer_instance`` was constructed from some ``init_optim
fp16/fp16_opt.py:355
Functionlog_sum_exp
(vec)
model/crf.py:30
Methodloss_scale
(self)
fp16/loss_scaler.py:52
Methodloss_scale
(self)
fp16/loss_scaler.py:161
Methodlr_lambda
(self, step)
utils/schedulers.py:13
Methodlr_lambda
(self, step)
utils/schedulers.py:33
Methodmake_sparse
(values)
utils/optimizers.py:310
Functionnetwork_to_half
Convert model to half precision in a batchnorm-safe way. Retained for legacy purposes. It is recommended to use FP16Model.
fp16/fp16util.py:50
Methodparam_groups
(self)
utils/optimizers.py:208
Functionpearson_and_spearman
(preds, labels)
utils/utils.py:390
Functionprep_param_lists
Creates a list of FP32 master parameters for a given model, as in `Training Neural Networks with Mixed Precision: Real Examples`_. Args
fp16/fp16util.py:103
Functionprepare_sequence
(seq, to_ix)
model/crf.py:23
Methodreset_parameters
Initialize parameters using Truncated Normal Initializer (default in Tensorflow)
model/graph_encoder.py:61
Methodresize_token_embeddings
(self, new_num_tokens=None)
model/graph_encoder.py:170
Methodscale_gradient
(self, module, grad_in, grad_out)
fp16/loss_scaler.py:55
Methodscale_gradient
(self, module, grad_in, grad_out)
fp16/loss_scaler.py:164
Methodstate
?
utils/optimizers.py:225
Methodstate_dict
?
utils/optimizers.py:229
Methodstate_dict
Returns a dict containing the current state of this :class:`FP16_Optimizer` instance. This dict contains attributes of :class:`FP16_O
fp16/fp16_opt.py:334
Methodstep
(self)
utils/optimizers.py:195
Methodstep
?
utils/optimizers.py:219
Functionto_python_float
(t)
fp16/fp16util.py:189
Functionto_python_float
(t)
fp16/loss_scaler.py:19
Functiontree_to_dist
(sent_len, tree)
model/tree.py:213
Methodupdate
(self, prediction, key, *args, **kwargs)
utils/utils.py:423
Methodupdate
Parameters ---------- verb_indices : ``List[Optional[int]]``, required. The indices of the verbal predicate in th
utils/utils.py:618
Methodupdate_scale
(self, overflow)
fp16/loss_scaler.py:135
Methodwrapped_closure
()
fp16/fp16_opt.py:454
← previousnext →201–300 of 302, ranked by callers