Ishaan Mathur

The two extra embeddings do indeed come from the `[cls]` and `[eos]` tokens (source: https://github.com/evolutionaryscale/esm/blob/e103c9c1c4047c38e8c7f1215c91f8481268e366/esm/tokenization/sequence_tokenizer.py#L48) To get per-residue embeddings for the original sequence, you can remove the first and last...

Similar to this issue https://github.com/evolutionaryscale/esm/issues/247, I believe this PR https://github.com/evolutionaryscale/esm/pull/276/files#diff-253d59232c58fceb8e86ac22dbcd860107afb9c3d5c4518d9b08ce997a9aab7e addresses the problem of all 1s being written out to the PDB file in the b factor column. I ran...

I was unfortunately unable to reproduce this issue. What version of `esm` are you using? I am using version `3.2.4.a0` and ran `open.py` in the default pixi environment (by running...

Try running `pip install --upgrade esm` to upgrade the version to the latest version released on pypi, which is `3.2.3`

The error message shows that it finds `3.2.3`, but it is ignoring it because your python version isn't compatible. try upgrading to python 3.12 https://github.com/evolutionaryscale/esm/blob/main/pyproject.toml#L6?

@KPHippe I have also noticed a diff between running Colabfold locally vs through the API. Have you been able to identify the reason for this?