Protocol esm1b: AttributeError: 'dict' object has no attribute 'startswith'
dvonkan opened this issue · 0 comments
dvonkan commented
Metadata
| key | value |
|---|---|
| version | 0.2.3 |
| cuda | True |
Parameter
| key | value |
|---|---|
| type | embed |
| protocol | esm1b |
| reduce | False |
| discard_per_amino_acid_embeddings | False |
Traceback
Traceback (most recent call last):
File "**(directory)**.cache/pypoetry/virtualenvs/biotrainer-8s_lNhFR-py3.10/lib/python3.10/site-packages/bio_embeddings/utilities/pipeline.py", line 322, in execute_pipeline_from_config
stage_output_parameters = stage_runnable(**stage_parameters)
File "**(directory)**.cache/pypoetry/virtualenvs/biotrainer-8s_lNhFR-py3.10/lib/python3.10/site-packages/bio_embeddings/embed/pipeline.py", line 377, in run
embedder: EmbedderInterface = embedder_class(**result_kwargs)
File "**(directory)**.cache/pypoetry/virtualenvs/biotrainer-8s_lNhFR-py3.10/lib/python3.10/site-packages/bio_embeddings/embed/esm_embedder.py", line 39, in __init__
model, alphabet = load_model_and_alphabet_local(self._options["model_file"])
File "**(directory)**.cache/pypoetry/virtualenvs/biotrainer-8s_lNhFR-py3.10/lib/python3.10/site-packages/bio_embeddings/embed/esm_embedder.py", line 23, in load_model_and_alphabet_local
return load_model_and_alphabet_core(model_data, None)
File "**(directory)**.cache/pypoetry/virtualenvs/biotrainer-8s_lNhFR-py3.10/lib/python3.10/site-packages/esm/pretrained.py", line 190, in load_model_and_alphabet_core
if model_name.startswith("esm2"):
AttributeError: 'dict' object has no attribute 'startswith'