SSTPredictor stopped when I use it to generate embedding for the structures from pdb file.

[XudongZhang-ny]

As shown in the picture, it stopped when i use:
predictor = SSTPredictor(structure_vocab_size=2048)
result = predictor.predict_from_pdb('../example_data/xxx.pdb')
to generate embedding for the pdb file.
Some pdb files worked, but some pdb files will be stuck, and whenever how long I wait for them.

[zanmaomao]

I encountered the same situation — have you resolved it?

[Lucy3B]

I encountered the same situation and solved this by removing the Nested Parallelism.
I change the following two functions in get_sst_seq.py.

1. def graph_conventer(..)
   for result in tqdm( map(process_subgraph, anchor_nodes), total=len(anchor_nodes), ):

2. def pdb_conventer(..)
   for result in tqdm( map(handle_pdf_file, pdb_files), total=len(pdb_files), ):