Skip to content

Issues: NVIDIA/bionemo-framework

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

31 Open 7 Closed
31 Open 7 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

Make model weights searchable on NGC enhancement New feature or request
#613 opened Jan 16, 2025 by xinyu-dev
Add descriptions for the params in the config YAML file of MolMIM enhancement New feature or request
#612 opened Jan 16, 2025 by xinyu-dev
Change default overlap param gather of geneformer to match esm2.
#585 opened Jan 9, 2025 by jstjohn
fine-tuning/retraining esm2
#565 opened Jan 2, 2025 by MarjanHJ
Installing SCDL raises a missing dependency
#472 opened Nov 23, 2024 by edawson
Compatibility of model weights trained in newer containers to run in older BioNeMo containers
#454 opened Nov 19, 2024 by xinyu-dev
Enable extraction of gene embeddings from geneformer (averaging of gene embeddings across all cells)
#452 opened Nov 19, 2024 by jstjohn
9
MolMIM checkpoint files ckpt to nemo conversion bug Something isn't working
#427 opened Nov 13, 2024 by xinyu-dev
Geneformer + ESM2 train script and pydantic config equivalence check
#403 opened Nov 6, 2024 by jstjohn
@skothenhill-nv
1
Refactor geneformer_mlm_loss_eval script to use a better mechanism for getting the tokenizer.
#402 opened Nov 5, 2024 by jstjohn
Write test for geneformer_mlm_loss_eval script.
#401 opened Nov 5, 2024 by jstjohn
Convert geneformer_mlm_loss_eval script to use pooch for downloads.
#400 opened Nov 5, 2024 by jstjohn
Add request that contributors from outside NVIDIA add contribution label to their PR documentation Improvements or additions to documentation good first issue Good for newcomers
#309 opened Oct 14, 2024 by jstjohn
1
Enable sequence_parallel and tp_comm_overlap in ESM2
#305 opened Oct 11, 2024 by jstjohn
Support virtual pipeline parallel in global batch size calculation
#286 opened Oct 8, 2024 by jstjohn
1
Automatically determine/override config settings when loading from checkpoint
#275 opened Oct 8, 2024 by jstjohn
@skothenhill-nv
1
Add ability for the from_checkpoint style functions to accept a tag or path object.
#242 opened Oct 3, 2024 by jstjohn
Fix issue where nemo1 to nemo2 conversion tests (esm2 and geneformer) only work when run by themselves.
#226 opened Oct 1, 2024 by jstjohn
1
fix failing unit tests in bionemo1 github related to hardware changes
#195 opened Sep 25, 2024 by dorotat-nv
@dorotat-nv
Add documentation describing what modifications you need (head/loss/inputs) to support different parallelisms of transformer block
#170 opened Sep 17, 2024 by jstjohn
[docs] Emphasize that notebook examples should have saved outputs
#162 opened Sep 16, 2024 by pstjohn
Fix bug checkpoint IO bug once https://github.com/NVIDIA/NeMo/pull/1024 lands bug Something isn't working
#134 opened Sep 4, 2024 by jstjohn
1
Refactor checkpoint loading for fine-tuning for new NeMo2 workflow
#132 opened Sep 4, 2024 by jstjohn
Refactor dataset so that when we upsample we get different masks for different queries of the same data.
#126 opened Aug 28, 2024 by jstjohn
1
Add DCO sign-off enforcement to repo
#83 opened Aug 7, 2024 by jstjohn
ProTip! Follow long discussions with comments:>50.