Functions — github.com/ServiceNow/HypE

CodeHub MCPcopy Create free account

hub / github.com/ServiceNow/HypE / functions

Functions60 in github.com/ServiceNow/HypE

⨍ Functions 60 ◇ Types & classes 10

↓ 19 callersMethodnum_ent

↓ 12 callersMethodnum_rel

↓ 6 callersMethodconvolve

(self, r_idx, e_idx, pos)

↓ 6 callersMethodpnr

(self, e_idx, r_idx)

↓ 5 callersMethod__init__

(self, dataset, emb_dim, **kwargs)

↓ 5 callersMethodnormalize

(self, normalizer)

↓ 5 callersMethodshift

↓ 3 callersMethodforward

(self, r_idx, e1_idx, e2_idx, e3_idx, e4_idx, e5_idx, e6_idx)

↓ 3 callersMethodinit

↓ 3 callersMethodsave_model

Save the model state to the output folder. If is_best_model is True, then save the model also as best_model.chkpnt

↓ 3 callersMethodtest

Evaluate the given dataset and print results, either by arity or all at once

↓ 2 callersMethodeval_dataset

Evaluate the dataset with the given model.

↓ 2 callersMethodget_model_from_name

Instantiate a model object given the model name

↓ 2 callersMethodpadd

(self, a, max_length)

↓ 2 callersMethodread

(self, file_path)

↓ 2 callersMethodread_test

(self, file_path)

↓ 2 callersMethodtuple2ids

↓ 1 callersMethod__str__

↓ 1 callersMethodadd_fact_and_shred

(self, fact, queries, raw_or_fil)

↓ 1 callersMethodallFactsAsTuples

↓ 1 callersMethodcreate_output_dir

If an output dir is given, then make sure it exists. Otherwise, create one based on time stamp.

↓ 1 callersMethodcreate_queries

(self, fact, position)

↓ 1 callersMethoddecompose_predictions

(self, targets, predictions, max_length)

↓ 1 callersMethodgenerate_neg

(self, pos_batch, neg_ratio)

↓ 1 callersMethodget_ent_id

↓ 1 callersMethodget_rank

(self, sim_scores)

↓ 1 callersMethodget_rel_id

↓ 1 callersMethodload_last_saved_model

Find the last saved model in the output_dir and load it. Load also the best_model If no model found in the output_dir or if t

↓ 1 callersMethodload_model

If a pretrained model is provided, then it will be loaded.

↓ 1 callersMethodneg_each

(self, arr, arity, nr)

↓ 1 callersMethodnext_batch

(self, batch_size, neg_ratio, device)

↓ 1 callersMethodnext_pos_batch

(self, batch_size)

↓ 1 callersMethodpadd_and_decompose

(self, targets, predictions, max_length)

↓ 1 callersMethodsave_hparams

Save the hyperparameters of the model as a json file in the output_dir if train time. If the output_dir contains an hparams.json file

↓ 1 callersMethodshred_facts

↓ 1 callersMethodtest_and_eval

↓ 1 callersMethodtrain_and_eval

↓ 1 callersMethodupdate

(self, rank, raw_or_fil)

↓ 1 callersMethodwas_last_batch

Add two measure objects

(self, dataset, model, valid_or_test, model_name)

(self, dataset, emb_dim, **kwargs)

(self, dataset, emb_dim, **kwargs)

(self, d, emb_dim, **kwargs)

(self, dataset, emb_dim, **kwargs)

(self, ds_name, max_arity=6)

(self, r_idx, e1_idx, e2_idx, e3_idx, e4_idx, e5_idx, e6_idx)

(self, r_idx, e1_idx, e2_idx, e3_idx, e4_idx, e5_idx, e6_idx)

(self, r_idx, e1_idx, e2_idx, e3_idx, e4_idx, e5_idx, e6_idx, ms, bs)

(self, r_idx, e1_idx, e2_idx, e3_idx, e4_idx, e5_idx, e6_idx, ms)

Methodnum_batch

(self, batch_size)

Methodrand_ent_except