Functions — github.com/McGill-NLP/length-generalization

hub / github.com/McGill-NLP/length-generalization / functions

Functions942 in github.com/McGill-NLP/length-generalization

⨍ Functions 942◇ Types & classes 118

↓ 2 callersMethod_create_answer

(self, example: Dict[str, Any])

src/data/data_instance_processor/clutrr_instance_processor.py:57

↓ 2 callersMethod_create_answer

(self, example: Dict[str, Any])

src/data/data_instance_processor/poly_instance_processor.py:60

↓ 2 callersMethod_create_answer

(self, example: Dict[str, Any])

src/data/data_instance_processor/parity_instance_processor.py:58

↓ 2 callersMethod_create_prompt

(self, example: Dict[str, Any])

src/data/data_instance_processor/sort_instance_processor.py:53

↓ 2 callersMethod_create_targets

(self, answer: List[str])

src/data/data_instance_processor/sort_instance_processor.py:63

↓ 2 callersFunction_environment_variables

Wraps `os.environ` to filter out non-encodable values.

src/common/params.py:86

↓ 2 callersFunction_flatten

For dict, list and set objects_ calls itself on the elements and for other types assigns the object_ to the corresponding key

src/common/nest.py:82

↓ 2 callersMethod_get_scratchpad_regex

(self)

src/data/data_instance_processor/data_instance_processor.py:258

↓ 2 callersMethod_get_tokenize_function

( self, add_special_tokens: bool = True, is_training: bool = True )

src/data/s2s_dl_factory.py:270

↓ 2 callersMethod_get_tokenize_function_for_all

( self, add_special_tokens: bool = True, is_training: bool = True )

src/data/token_classification_dl_factory.py:91

↓ 2 callersMethod_patch_config_obj_for_sweep

(self, config: JsonDict)

src/main.py:121

↓ 2 callersMethod_reshape_tensor

(input_tensor)

src/modules/span_extractors/self_attentive_span_extractor.py:330

↓ 2 callersFunction_unflatten

(dic, keys, value)

src/common/nest.py:377

↓ 2 callersFunction_unflatten_asserts

(flat_dict, separator)

src/common/nest.py:356

↓ 2 callersMethodall

(cls)

src/tokenization_utils/__init__.py:13

↓ 2 callersFunctionapply_rotary_pos_emb_new

(x, sincos, offset=0)

src/models/custom_t5_decoder_only.py:82

↓ 2 callersMethodassert_empty

(self, class_name: str)

src/common/params.py:387

↓ 2 callersMethodby_name

Returns a callable function that constructs an argument of the registered class. Because you can register particular functions as co

src/common/registrable.py:148

↓ 2 callersMethodcreate_input_output

Generates an input-output pair.

scripts/dataset_builders/make_dyck_pred_dataset.py:154

↓ 2 callersMethodcreate_model

(self)

src/runtime/seq2seq_runtime.py:354

↓ 2 callersMethodcreate_scratchpad_steps_from_data_instance

( self, data_instance: Dict[str, Any] )

src/data/data_instance_processor/data_instance_processor.py:710

↓ 2 callersMethodcreate_scratchpad_steps_from_data_instance

( self, data_instance: Dict[str, Any] )

src/data/data_instance_processor/sort_instance_processor.py:72

↓ 2 callersMethodextract_answer_from_final_output

(self, final_output: str)

src/data/data_instance_processor/sort_instance_processor.py:122

↓ 2 callersMethodextract_answer_from_prediction

(self, prediction: str)

src/data/data_instance_processor/data_instance_processor.py:32

↓ 2 callersFunctionfixed_pos_embedding

(x, seq_dim=1, seq_len=None)

src/models/custom_t5_decoder_only.py:48

↓ 2 callersMethodflush_local_log

(self)

src/analyzers/analyzer.py:47

↓ 2 callersMethodfrom_file

Load a `Params` object from a configuration file. # Parameters params_file: `str` The path to the configuratio

src/common/params.py:427

↓ 2 callersFunctiongenerate_copy_task_data

( seq_length: int, num_examples: int, args: argparse.Namespace )

scripts/dataset_builders/make_copy_half_dataset.py:98

↓ 2 callersFunctiongenerate_copy_task_data

( seq_length: int, num_examples: int, args: argparse.Namespace )

scripts/dataset_builders/make_copy_dataset.py:181

↓ 2 callersFunctiongenerate_data

( num_variables: int, num_examples: int, permute_clauses: bool = False )

scripts/dataset_builders/make_lego_dataset.py:64

↓ 2 callersFunctiongenerate_data

( seq_length: int, num_examples: int, permute_clauses: bool = False )

scripts/dataset_builders/make_parity_dataset.py:32

↓ 2 callersFunctiongenerate_data

( num_terms: int, num_examples: int, modular: bool = False )

scripts/dataset_builders/make_poly_dataset.py:32

↓ 2 callersFunctiongenerate_data

( seq_length: int, num_examples: int, modular: bool = False, )

scripts/dataset_builders/make_sum_dataset.py:32

↓ 2 callersFunctiongenerate_data

( seq_lengths: Tuple[int, int], num_examples: int, args: argparse.Namespace )

scripts/dataset_builders/make_addition_dataset.py:67

↓ 2 callersFunctiongenerate_data

( seq_length: int, num_examples: int, args: argparse.Namespace )

scripts/dataset_builders/make_sort_dataset.py:118

↓ 2 callersMethodgenerate_list

(self, num, min_size, max_size)

scripts/dataset_builders/make_dyck_pred_dataset2.py:57

↓ 2 callersFunctiongenerate_reverse_task_data

( seq_length: int, num_examples: int, args: argparse.Namespace )

scripts/dataset_builders/make_reverse_dataset.py:141

↓ 2 callersFunctiongenerate_task_data

( seq_length: int, num_examples: int, args: argparse.Namespace )

scripts/dataset_builders/make_count_dataset.py:118

↓ 2 callersMethodget_collate_fn

(self, state: ExperimentStage)

src/data/sequence_classification_dl_factory.py:230

↓ 2 callersMethodget_collate_fn

(self, state: ExperimentStage)

src/data/token_classification_dl_factory.py:227

↓ 2 callersMethodget_collate_fn

(self, state: ExperimentStage)

src/data/s2s_dl_factory.py:557

↓ 2 callersFunctionget_device_of

Returns the device of the tensor.

src/modules/span_extractors/self_attentive_span_extractor.py:12

↓ 2 callersFunctionget_file_name

(config: Dict[str, bool])

scripts/experiment_uploaders/attention_analysis_aggregate.py:130

↓ 2 callersFunctionget_generalization_ranks

( acc_list: List[Tuple[int, int, int]] )

notebooks/plotting/training_plots.py:23

↓ 2 callersFunctionget_param_hash

(params: Dict[str, Any])

scripts/fail_if_sweep_not_complete.py:97

↓ 2 callersFunctionget_project_name

()

scripts/fail_if_sweep_not_complete.py:18

↓ 2 callersFunctionget_queued_jobs

()

scripts/launch_experiment.py:645

↓ 2 callersFunctionget_queued_jobs

()

scripts/launcher.py:812

↓ 2 callersFunctionget_range_vector

Returns a range vector with the desired size, starting at 0. The CUDA implementation is meant to avoid copy data from CPU to GPU.

src/modules/span_extractors/self_attentive_span_extractor.py:22

↓ 2 callersFunctionget_rank

()

src/common/torch_utils.py:22

↓ 2 callersFunctionget_repo_dir

()

scripts/wandb_download_all_bundles.py:14

↓ 2 callersMethodget_search_space_fn

(self)

src/hp_search_space/base_search_space.py:37

↓ 2 callersMethodget_slopes

(self, n)

src/models/custom_t5_decoder_only.py:753

↓ 2 callersFunctionget_true_remote_path

(cluster: str, remote_path: str)

scripts/download_nope_runs_to_local.py:88

↓ 2 callersFunctionget_unique_tokens

()

scripts/dataset_builders/make_count_dataset.py:39

↓ 2 callersMethodhandle_tokenizer

(self, tokenizer: Optional[Tokenizer] = None)

src/models/custom_t5_decoder_only.py:1190

↓ 2 callersFunctioninstall_requirements

(args: argparse.Namespace)

scripts/launch_experiment.py:138

↓ 2 callersMethodis_answer_correct

( self, final_answer: List[str], data_instance: Dict[str, Any] )

src/data/data_instance_processor/sort_instance_processor.py:143

↓ 2 callersMethodlist_available

List default first if it exists

src/common/registrable.py:221

↓ 2 callersFunctionlist_available_models_groups

List a directory recursively in a remote location using SSH with password-less login and SSH configuration. Args: hostname (str): Hos

scripts/experiment_uploaders/list_checkpoint_on_each_clusters.py:7

↓ 2 callersFunctionload_dataframe_from_jsonlines

(path: Path)

scripts/experiment_uploaders/attention_kl_analysis.py:165

↓ 2 callersFunctionload_download_dir

(download_dir: Path)

notebooks/plotting/attention_analysis.py:43

↓ 2 callersFunctionload_download_dir

(download_dir: Path)

notebooks/plotting/attention_analysis_utils.py:45

↓ 2 callersMethodpop_choice

Gets the value of `key` in the `params` dictionary, ensuring that the value is one of the given choices. Note that this `pops` the ke

src/common/params.py:271

↓ 2 callersFunctionread_split

(split)

scripts/dataset_builders/make_pcfg_dataset.py:40

↓ 2 callersFunctionrotate_every_two

Example: [a, b, c, d] -> [-b, a, -d, c]

src/models/custom_t5_decoder_only.py:61

↓ 2 callersFunctionsave_all_len_gen_experiments

( wandb_api: Api, save_path: Optional[Path] = None, group_prefix: str = "SW-", hp_metric_name:

notebooks/plotting/training_plots.py:176

↓ 2 callersMethodset_input_embeddings

(self, new_embeddings)

src/models/custom_t5_decoder_only.py:1231

↓ 2 callersFunctiontest_ds

( ds: Dataset, dl_factory: DataLoaderFactory, stage: ExperimentStage, tokenizer: Tokenizer, )

src/tests/test_all_traditional_len_gen_datasets.py:66

↓ 2 callersFunctiontest_ds

( ds: Dataset, dl_factory: DataLoaderFactory, stage: ExperimentStage, tokenizer: Tokenizer, )

src/tests/test_all_scratchpad_datasets.py:168

↓ 2 callersMethodto_params

Returns a `Params` object that can be used with `.from_params()` to recreate an object just like it. This relies on `_to_par

src/common/from_params.py:815

↓ 2 callersFunctiontorch_compute_symmetric_kl

(p, q)

src/analyzers/attention_kl_analyzer.py:17

↓ 2 callersFunctionunique_experiment_name

(config)

src/common/py_utils.py:37

↓ 2 callersFunctionunique_experiment_name_from_filenames

(config_filenames)

scripts/upload_experiment.py:442

↓ 1 callersFunctionConvertSMVer2Cores

(major, minor)

src/common/gpu_utils.py:31

↓ 1 callersMethod__call__

(self, example: Dict[str, Any])

src/data/data_instance_processor/scan_instance_processor.py:23

↓ 1 callersMethod__init__

(self, module)

src/modules/span_extractors/self_attentive_span_extractor.py:280

↓ 1 callersMethod__init__

(self, *args, **kwargs)

src/analyzers/runtime_analyzer.py:24

↓ 1 callersMethod__init__

(self, **kwargs)

src/data/data_instance_processor/sort_instance_processor.py:15

↓ 1 callersMethod__init__

(self, buffer: io.BytesIO)

src/common/from_params.py:95

↓ 1 callersMethod__init__

(self)

src/runtime/model_inference_shell.py:13

↓ 1 callersMethod__init__

( self, config: Optional[HfModelConfig] = None, tokenizer: Optional[Tokenizer] = None,

src/models/gpt_neo.py:23

↓ 1 callersMethod__init__

( self, config: Optional[HfModelConfig] = None, tokenizer: Optional[Tokenizer] = None,

src/models/gpt2.py:19

↓ 1 callersMethod_analyze_attentions

(self)

src/analyzers/attention_analyzer.py:24

↓ 1 callersMethod_build_base_dataset

(self, path: Path)

src/data/s2s_dl_factory.py:205

↓ 1 callersMethod_build_tokenized_dataset

( self, stage: ExperimentStage, base_ds: Dataset, tokenize: bool = True, )

src/data/s2s_dl_factory.py:227

↓ 1 callersMethod_collect_generation_metrics

(self)

src/analyzers/runtime_analyzer.py:219

↓ 1 callersMethod_collect_inference_metrics

(self)

src/analyzers/runtime_analyzer.py:186

↓ 1 callersMethod_collect_training_metrics

(self)

src/analyzers/runtime_analyzer.py:149

↓ 1 callersMethod_compute_aggregate_measures

(self)

src/analyzers/attention_aggregate_analyzer.py:64

↓ 1 callersMethod_compute_kl

(self)

src/analyzers/attention_kl_analyzer.py:124

↓ 1 callersMethod_compute_max_attention_index_measures

( attention_map_names, attn_analysis_dir, cat, df_data, ds, tokenization_infos )

src/analyzers/attention_aggregate_analyzer.py:184

↓ 1 callersMethod_compute_score_before_after_measures

(attn_analysis_dir, cat, df_data, ds)

src/analyzers/attention_aggregate_analyzer.py:146

↓ 1 callersFunction_convert_dict_to_list

(object_, parent_object, parent_object_key)

src/common/nest.py:420

↓ 1 callersMethod_convert_targets

(self, targets: str)

src/data/data_instance_processor/data_instance_processor.py:64

↓ 1 callersMethod_create_answer

(self, example: Dict[str, Any])

src/data/data_instance_processor/data_instance_processor.py:698

↓ 1 callersMethod_create_answer

(self, example: Dict[str, Any])

src/data/data_instance_processor/sort_instance_processor.py:199

↓ 1 callersMethod_create_notify_script

( self, job_body: Dict[str, Any], persistent_key: str )

scripts/launcher.py:331

↓ 1 callersMethod_create_post_sbatch_launch_script

( self, tmp_exp_dir: Path, persistent_key: str )

scripts/launcher.py:323

↓ 1 callersMethod_create_pre_sbatch_launch_script

( self, tmp_exp_dir: Path, persistent_key: str )

scripts/launcher.py:276

← previous next →101–200 of 942, ranked by callers