Skip to content

Issues: pytorch/torchtitan

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

80 Open 202 Closed
80 Open 202 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

Profiling only a select group of ranks
#1035 opened Mar 31, 2025 by githubsgi
3
Context parallel on Turing GPUs? module: context parallel question Further information is requested
#1034 opened Mar 31, 2025 by dingqingy
6
Linear layer weights are in float32 ? question Further information is requested
#1027 opened Mar 28, 2025 by githubsgi
1
@mori360
14
Any plan to add Llama 1B and/or 3B models ?
#1026 opened Mar 28, 2025 by githubsgi
4
Unable to run flex attention and torch.compile module: flex attention
#1005 opened Mar 22, 2025 by lkhphuc
@fegin
1
Optimizer in backward with grad clipping is broken
#990 opened Mar 20, 2025 by apaz-cli
@mori360
7
Is a PP+FSDP+TP config + toml available for pre-training 405B model ?
#986 opened Mar 19, 2025 by githubsgi
3
[DeepSeek] Create Titan's TrainSpec
#975 opened Mar 17, 2025 by kwen2501 DeepSeek landing
@lessw2020
[DeepSeek] Connect to Titan's checkpointing
#974 opened Mar 17, 2025 by kwen2501 DeepSeek landing
[DeepSeek] Connect to Titan's data loader
#973 opened Mar 17, 2025 by kwen2501 DeepSeek landing
End-to-end training of DeepSeek V3
#942 opened Mar 7, 2025 by EugenHotaj
@kwen2501
5
[Feature] Add Multi-Token Prediction module enhancement New feature or request
#933 opened Mar 5, 2025 by lessw2020
@lessw2020
1
[TP] RuntimeError: shape '[1, 8192, -1, 128]' is invalid for input of size 524288 module: dtensor
#932 opened Mar 5, 2025 by aahehehe
4
[Feature] Support skipping bad grad updates
#916 opened Mar 3, 2025 by yzhangcs
4
[Feature] add preflight NCCL and GEMM check to multinode slurm script
#915 opened Mar 3, 2025 by lessw2020
@lessw2020
[Feature] expose Torch Nan checker as configurable option in toml for those training at scale
#914 opened Mar 3, 2025 by lessw2020
@lessw2020
1
[Checkpointing] fails out if checkpoint folder does not exist when using keep_latest_k bug Something isn't working module: checkpoint
#911 opened Mar 2, 2025 by lessw2020
@fegin
1
[Checkpointing] Using keep_latest_k setting results in failure when using external mounted drive module: checkpoint
#909 opened Mar 1, 2025 by lessw2020
@fegin
1
[Possible PR discuss] Will a PR of training HF model be welcomed? community help wanted huggingface integration
#903 opened Feb 28, 2025 by junjzhang
1
7
Question about triton in deepseek implementtion question Further information is requested
#902 opened Feb 28, 2025 by zqwenn
@kwen2501
2
better handling of tensorwise float8 recipe in configuration
#901 opened Feb 27, 2025 by vkuzo
@vkuzo
1
[Feature request] Integrate DeepGEMM
#889 opened Feb 26, 2025 by lessw2020
@lessw2020
2
dcp.load fails on checkpoints prior to AdamW refactor module: checkpoint
#886 opened Feb 25, 2025 by eminorhan
@fegin
1
Possible to integrate DeepEP?
#885 opened Feb 25, 2025 by ericxsun
4
[Evaluation] Minimal support for downstream tasks community help wanted enhancement New feature or request
#883 opened Feb 24, 2025 by K-H-Ismail
4
ProTip! Updated in the last three days: updated:>2025-03-30.