openfold.data.input_pipeline_multimer¶

Functions

`compose`(x, fs)
`ensembled_transform_fns`(common_cfg, ...)	Input pipeline data transformers that can be ensembled and averaged.
`groundtruth_transforms_fns`()
`map_fn`(fun, x)
`nonensembled_transform_fns`()	Input pipeline data transformers that are not ensembled.
`prepare_ground_truth_features`(tensors)	Prepare ground truth features that are only needed for loss calculation during training
`process_tensors_from_config`(tensors, ...)	Based on the config, apply filters and transformations to the data.

compose(x, fs)¶

ensembled_transform_fns(common_cfg, mode_cfg, ensemble_seed)¶: Input pipeline data transformers that can be ensembled and averaged.

groundtruth_transforms_fns()¶

map_fn(fun, x)¶

nonensembled_transform_fns()¶: Input pipeline data transformers that are not ensembled.

prepare_ground_truth_features(tensors)¶: Prepare ground truth features that are only needed for loss calculation during training

process_tensors_from_config(tensors, common_cfg, mode_cfg)¶: Based on the config, apply filters and transformations to the data.