2 Pull requests merged by 2 people

• Fix FSDP venv creation

  #720 merged Jun 8, 2025

• Fsdp regression tests

  #714 merged Jun 6, 2025

2 Issues closed by 1 person

• Fix FSDP CI for venv

  #717 closed Jun 8, 2025

• Add FSDP CI testing for slurm w/ ParallelCluster.

  #713 closed Jun 6, 2025

5 Issues opened by 3 people

• Fix FSDP dataset warning about sequence length being too long.

  #719 opened Jun 7, 2025

• Fix FSDP checkpoint for slurm with container

  #718 opened Jun 7, 2025

• Add HF_TOKEN for FSDP testing.

  #716 opened Jun 7, 2025

• Duplicate content in docker install script after refactoring

  #715 opened Jun 6, 2025

• Implement in CDK

  #712 opened Jun 6, 2025

10 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• new fsdp dataset: nvidia-deeplearningexamples

  #706 commented on Jun 9, 2025 • 3 new comments

• added new tool to scale up-down nodes on an instance group

  #708 commented on Jun 9, 2025 • 1 new comment

• Revisit PR 689 due to version update to NV Container Toolkit

  #711 commented on Jun 6, 2025 • 0 new comments

• Add Nemo2.0 script example

  #579 commented on Jun 7, 2025 • 0 new comments

• Change slurm exporter to Slinky slurm exporter

  #492 commented on Jun 8, 2025 • 0 new comments

• Change `save_state_dict` to `save` checkpoint in FSDP.

  #710 commented on Jun 8, 2025 • 0 new comments

• Update pcluster architecture guidance

  #464 commented on Jun 9, 2025 • 0 new comments

• Enable 1click for SageMaker HyperPod

  #670 commented on Jun 6, 2025 • 0 new comments

• Lustre mount via Ansible for SMHP Slurm LCS

  #682 commented on Jun 6, 2025 • 0 new comments

• HyperPod EKS Helper Script Fixes

  #709 commented on Jun 6, 2025 • 0 new comments