Functions — github.com/bytedance/dplm

CodeHub MCPcopy Create free account

hub / github.com/bytedance/dplm / functions

Functions1,008 in github.com/bytedance/dplm

⨍ Functions 1,008◇ Types & classes 149↳ Endpoints 2

( self, data_dir: str = "data/tape", max_tokens: int = 6000, max_len: int = 10

src/byprot/datamodules/tokenized_protein_datamodule.py:34

( self, data_dir: str = "data/tape", max_tokens: int = 6000, max_len: int = 20

src/byprot/datamodules/uniref50.py:35

( self, data_dir: str = "data/", chain_set_jsonl: str = "chain_set.jsonl", cha

src/byprot/datamodules/cath_datamodule.py:26

( self, sequence_lengths: Iterable, bucket_size: int, num_replicas: int = 1,

src/byprot/datamodules/dataset/uniref.py:36

( self, sampler, max_tokens, max_batch, sample_lengths, max_sq

src/byprot/datamodules/dataset/uniref.py:100

( self, data_dir: str, split: str, max_len=2048, )

src/byprot/datamodules/dataset/uniref.py:193

(self, tokenizer_path=None)

src/byprot/datamodules/dataset/uniref.py:258

( self, dataset, batch_size, max_tokens=6000, drop_last=False,

src/byprot/datamodules/dataset/data_utils.py:384

( self, sequence_lengths: Iterable, bucket_size: int, num_replicas: int = 1,

src/byprot/datamodules/dataset/tokenized_protein.py:102

( self, sampler, max_tokens, max_batch, sample_lengths, max_sq

src/byprot/datamodules/dataset/tokenized_protein.py:167

( self, data_dir: str, split: str, csv_file: str, max_len=2048,

src/byprot/datamodules/dataset/tokenized_protein.py:252

( self, vocab_file, aa_cls_token="<cls_aa>", aa_eos_token="<eos_aa>",

src/byprot/datamodules/dataset/tokenized_protein.py:361

(self, tokenizer)

src/byprot/datamodules/dataset/tokenized_protein.py:562

( self, sequence_lengths: Iterable, bucket_size: int, num_replicas: int = 1,

src/byprot/datamodules/dataset/uniref_hf.py:25

( self, sampler, max_tokens, max_batch, sample_lengths, max_sq

src/byprot/datamodules/dataset/uniref_hf.py:89

( self, data_dir: str, split: str, max_len=2048, )

src/byprot/datamodules/dataset/uniref_hf.py:182

( self, max_len=2048, num_seqs=40, )

src/byprot/datamodules/dataset/uniref_hf.py:214

(self, tokenizer_path=None)

src/byprot/datamodules/dataset/uniref_hf.py:269

( self, alphabet, coord_pad_inf=False, coord_nan_to_zero=True, to_pifo

src/byprot/datamodules/dataset/cath.py:191

(self, alphabet)

src/byprot/datamodules/dataset/cath.py:339

(self, data_cfg)

src/byprot/datamodules/pdb_dataset/pdb_datamodule.py:67

( self, *, sampler_cfg, metadata_csv, seed=123, shuffle=True,

src/byprot/datamodules/pdb_dataset/pdb_datamodule.py:574

src/byprot/datamodules/dataset/uniref.py:61

src/byprot/datamodules/dataset/uniref.py:179

src/byprot/datamodules/dataset/data_utils.py:420

src/byprot/datamodules/dataset/tokenized_protein.py:128

src/byprot/datamodules/dataset/tokenized_protein.py:238

src/byprot/datamodules/dataset/uniref_hf.py:50

src/byprot/datamodules/dataset/uniref_hf.py:168

src/byprot/datamodules/pdb_dataset/pdb_datamodule.py:664

src/byprot/utils/protein/utils.py:70

src/byprot/datamodules/dataset/uniref.py:76

src/byprot/datamodules/dataset/uniref.py:176

src/byprot/datamodules/dataset/uniref.py:211

src/byprot/datamodules/dataset/uniref.py:253

src/byprot/datamodules/dataset/data_utils.py:84

src/byprot/datamodules/dataset/data_utils.py:417

src/byprot/datamodules/dataset/tokenized_protein.py:143

src/byprot/datamodules/dataset/tokenized_protein.py:235

src/byprot/datamodules/dataset/tokenized_protein.py:276

src/byprot/datamodules/dataset/tokenized_protein.py:344

src/byprot/datamodules/dataset/uniref_hf.py:65

src/byprot/datamodules/dataset/uniref_hf.py:165

src/byprot/datamodules/dataset/uniref_hf.py:194

src/byprot/datamodules/dataset/uniref_hf.py:230

src/byprot/datamodules/dataset/uniref_hf.py:261

src/byprot/datamodules/pdb_dataset/pdb_datamodule.py:510

src/byprot/datamodules/pdb_dataset/pdb_datamodule.py:669

Method__post_init__

src/byprot/datamodules/pdb_dataset/protein.py:64

Method__setattr__

(self, name, value)

src/byprot/utils/lr_scheduler.py:66

Method_adaptive_mask

(target_tokens)

src/byprot/tasks/lm/dplm_invfold.py:210

Method_checkpoint_wrapper

(m, o=False, enable=True)

src/byprot/models/structok/modules/folding_utils/trunk.py:140

Method_compute_language_model_representations

Adds bos/eos tokens for the language model, since the structure module doesn't use these.

src/byprot/models/structok/modules/folding_utils/decoder.py:144

Method_esm_idx_to_af2_idx

(self, esmaa, mask)

src/byprot/models/structok/modules/folding_utils/decoder.py:141

Method_featurize

src/byprot/utils/protein/evaluator_dplm2.py:202

Method_format_checkpoint_name

( self, filename, metrics: Dict[str, Tensor], prefix: str = "", auto_i

src/byprot/utils/callbacks.py:206

Method_full_mask

(target_tokens)

src/byprot/tasks/lm/dplm_invfold.py:173

Method_get_resized_embeddings

Build a resized Embedding Module from a provided token Embedding Module. Increasing the size will add newly initialized vectors at the

src/byprot/models/dplm2/modules/dplm2_modeling_esm.py:701

Method_init_progress

(self, trainer)

src/byprot/utils/callbacks.py:124

Method_masked_mean

src/byprot/models/structok/modules/vqvae.py:366

Method_merge_and_log

src/byprot/tasks/lm/dplm_invfold.py:539

Method_mlm_mask

src/byprot/tasks/lm/mlm.py:113

Method_process_csv_row2

(self, processed_file_path)

src/byprot/datamodules/pdb_dataset/pdb_datamodule.py:338

Method_random_mask

(target_tokens)

src/byprot/tasks/lm/dplm_invfold.py:185

Method_reparam_process

This function is used to perform reparameterized decoding. output_tokens: [B, N] output_scores: [B, N] cur_tokens

src/byprot/models/dplm2/dplm2.py:579

Method_save_last_checkpoint

( self, trainer: "pl.Trainer", monitor_candidates: Dict[str, Tensor] )

src/byprot/utils/callbacks.py:290

Method_selected_mask

(target_tokens, sel_mask)

src/byprot/tasks/lm/dplm_invfold.py:204

Method_setup_model

Wraps the model into a :class:`~torch.distributed.fsdp.fully_sharded _data_parallel.FullyShardedDataParallel` module.

src/byprot/utils/strategies.py:30

Method_update_best_and_save

( self, current: Tensor, trainer: "pl.Trainer", monitor_candidates: Dict[str,

src/byprot/utils/callbacks.py:227

Methodaa_cls_token

`str`: End of sentence token. Log an error if used while not having been set.

src/byprot/datamodules/dataset/tokenized_protein.py:420

Methodaa_eos_token

`str`: End of sentence token. Log an error if used while not having been set.

src/byprot/datamodules/dataset/tokenized_protein.py:409

Methodaa_mask_token

`str`: End of sentence token. Log an error if used while not having been set.

src/byprot/datamodules/dataset/tokenized_protein.py:442

Methodaa_unk_token

`str`: End of sentence token. Log an error if used while not having been set.

src/byprot/datamodules/dataset/tokenized_protein.py:431

Functionaccuracy

(pred, target, mask=None, reduction="all")

src/byprot/modules/metrics.py:54

Functionaccuracy_per_sample

(pred, target, mask=None)

src/byprot/modules/metrics.py:62

Functionadd_seqlen

src/byprot/datamodules/dataset/tokenized_protein.py:83

Functionadjust_oxygen_pos

Imputes the position of the oxygen atom on the backbone by using adjacent frame information. Specifically, we say that the oxygen atom is in t

src/byprot/datamodules/pdb_dataset/utils.py:231

Functionbackbone_atom_loss

Computes backbone atom loss. Args: atom37_pred_positions: [*, N_res, 37, 3] predicted positions atom37_gt_positions: [*, N_res, 37, 3

src/byprot/models/structok/modules/loss.py:1662

Methodbasic_ipstream

Default constructor, creates an uninitialised stream.

analysis/TMscore.cpp:569

Methodbasic_opstream

Default constructor, creates an uninitialised stream.

analysis/TMscore.cpp:721

Methodbasic_pstream

Default constructor, creates an uninitialised stream.

analysis/TMscore.cpp:861

Methodbasic_pstreambuf

analysis/TMscore.cpp:1209

Methodbasic_rpstream

Default constructor, creates an uninitialised stream.

analysis/TMscore.cpp:1033

Functionbatch_align_structures

(pos_1, pos_2, mask=None)

src/byprot/datamodules/pdb_dataset/utils.py:199

Methodbits_to_indices

bits: bool tensor of big endian bits, where the last dimension is the bit dimension returns indices, which are long integers from 0

src/byprot/models/structok/modules/lfq.py:241

Functioncalc_aatype_metrics

(generated_aatypes)

src/byprot/utils/protein/utils.py:822

Functioncalc_aligned_rmsd

src/byprot/modules/protein_metrics.py:148

Functioncalc_ca_ca_metrics

(ca_pos, bond_tol=0.1, clash_tol=1.0)

src/byprot/utils/protein/utils.py:881

Functioncalc_distogram

(pos, min_bin, max_bin, num_bins)

src/byprot/utils/protein/utils.py:490

Functioncalc_mdtraj_metrics

src/byprot/utils/protein/utils.py:797

Functioncalc_perplexity

(pred, labels, mask)

src/byprot/modules/protein_metrics.py:71

Functioncalc_rbf

(ca_dists, num_rbf, D_min=1e-3, D_max=22.0)

src/byprot/utils/protein/utils.py:593

Functioncalculate_diversity

( output_dir, metrics_df, top_sample_csv, designable_csv_path )

src/byprot/utils/protein/utils.py:149

Functioncalculate_neighbor_angles

Calculate angles between atoms c <- a -> b. Parameters ---------- R_ac: Tensor, shape = (N,3) Vector from atom a to c.

src/byprot/utils/protein/all_atom.py:205

Functioncalculate_neighbor_angles

Calculate angles between atoms c <- a -> b. Parameters ---------- R_ac: Tensor, shape = (N,3) Vector from atom a to c.

src/byprot/datamodules/pdb_dataset/all_atom.py:205

Functioncalculate_pmpnn_consistency

( output_dir, designable_csv, designable_csv_path )

src/byprot/utils/protein/utils.py:178

Functioncalculate_pmpnn_designability

( output_dir, designable_csv, designable_csv_path, all_mpnn_folds_df_path="pmpnn_results.csv",

src/byprot/utils/protein/utils.py:211

Functioncollate_batch

( batch: List[Dict[str, Any]], batch_converter, transform=None, atoms=("N", "CA", "C", "O"), )

src/byprot/datamodules/dataset/cath.py:143

Functioncommon_pipeline

(config, training=False)

src/byprot/utils/__init__.py:221

(scores, target, label_mask, weights, key="")

src/byprot/modules/cross_entropy.py:265

← previous next →601–700 of 1,008, ranked by callers