Skip to content

Issues: stanford-crfm/levanter

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

68 Open 140 Closed
68 Open 140 Closed
Author
Filter by author
Label
Filter by label
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Milestones
Filter by milestone
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

Incompatible devices for jitted computation
#609 opened Jun 1, 2024 by versae
2
Tensorboard FileExistsError on SLURM when training multi-node with 1 process per GPU
#570 opened Apr 30, 2024 by Lauler
3
Possible Ray OOM when procesing long documents
#568 opened Apr 30, 2024 by versae
5
wire up llm-eval-harness markweb p1
#564 opened Apr 23, 2024 by dlwh
Clean up the story for overriding hf configs (e.g. for scan_layers)
#555 opened Apr 15, 2024 by dlwh
add bpb metric markweb
#550 opened Apr 10, 2024 by dlwh
Add TPU Metrics to Weights and Biases Logging enhancement help wanted Extra attention is needed
#544 opened Apr 3, 2024 by Helw150
3
Recommended citation for Levanter?
#537 opened Mar 23, 2024 by haileyschoelkopf
4
Add docs for installing TE from source
#518 opened Mar 8, 2024 by dlwh
2
Integrate Pallas TPU Flash Attention good first issue Good for newcomers help wanted Extra attention is needed p2 perf tpu
#514 opened Mar 7, 2024 by dlwh
Don't use gpt2 tokenizer etc in tests good first issue Good for newcomers p3
#500 opened Feb 28, 2024 by dlwh
5
Fix docs for warmup/warmup_ratio documentation Improvements or additions to documentation good first issue Good for newcomers p2
#499 opened Feb 28, 2024 by dlwh
Many errors while running unit tests on A100 and V100 bug Something isn't working p0
#498 opened Feb 28, 2024 by DwarKapex
@vadam5
5
Make a GH action to launch the small-fast tpu job infrastructure p1 perf
#493 opened Feb 27, 2024 by dlwh
Make a better website documentation Improvements or additions to documentation good first issue Good for newcomers help wanted Extra attention is needed p2
#492 opened Feb 27, 2024 by dlwh
4 tasks
1
Work around #475 (large model initialize on GPU) gpu p1
#482 opened Feb 22, 2024 by dlwh GPU Parity
2
RNG on GPU not sharding correctly, leads to oom gpu upstream
#475 opened Feb 16, 2024 by dlwh GPU Parity
1
Implement Mixtral 8x7b enhancement p2
#473 opened Feb 15, 2024 by dlwh
1
@blahBlahhhJ
NVIDIA JAX toolkit crashes running simple nano run gpu upstream
#471 opened Feb 15, 2024 by dlwh GPU Parity
1
Automatic GH action for bitwise repro tests infrastructure p2
#465 opened Feb 12, 2024 by dlwh
Data Mixture training with RedPajama Dataset
#442 opened Feb 1, 2024 by Ivan-Zhou
2
Provide better exception when we get a timeout sharing wandb data error handling good first issue Good for newcomers p3
#432 opened Jan 25, 2024 by dlwh
FSDP with an odd/weird number of GPUs. enhancement p3
#429 opened Jan 23, 2024 by dlwh GPU Parity
2
TensorParallelism is currently using more memory than not-TP. p2 perf
#428 opened Jan 23, 2024 by dlwh
Reinvestigate Fancy Gradient Checkpointing p3 perf
#427 opened Jan 23, 2024 by dlwh
3 tasks
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.