aws-samples awsome-distributed-training issues

aws-samples / awsome-distributed-training

Collection of best practices, reference architectures, model training examples and utilities to train large models on AWS.

MIT No Attribution

203 stars 86 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Update NCCL tests for both slurm and k8s

#506 KeitaW opened 2 days ago
1
fix: Updating Dockerfile to pin versions and fix the exampel

#505 mnuyens closed 3 days ago
0
pin prometheus to version 2.55.1

#504 nghtm closed 4 days ago
1
Include full Dockerfile for FSDP example

#503 sean-smith closed 3 days ago
2
name SSM doc

#502 sean-smith opened 5 days ago
1
Change aws ofi plugin version 1.13.0

#501 mhuguesaws opened 6 days ago
2
check for ssm document

#500 sean-smith closed 6 days ago
1
Raising the default Kubernetes version in HyperPod EKS template to 1.30

#499 shimomut closed 1 week ago
0
add link to the HP EKS workshop

#498 KeitaW closed 1 week ago
1
create ~/.ssh/config if it doesn't exist

#497 sean-smith closed 1 week ago
1
Issues running easy-ssh.sh script locally

#496 nghtm closed 6 days ago
1
Update 0.nvcr-pytorch-aws.dockerfile

#495 nghtm closed 1 week ago
0
small improvement to login

#494 sean-smith closed 1 week ago
0
Update easy-ssh.sh

#493 sean-smith closed 1 week ago
0
Change slurm exporter to Slinky slurm exporter

#492 mhuguesaws opened 1 week ago
2
FSDP EKS Example failing with: module 'torch.library' has no attribute 'register_fake'

#491 nghtm opened 2 weeks ago
1
easy-ssh.sh switch to default (ubuntu) user

#490 sean-smith closed 1 week ago
0
Synchronizing the CF template with the one we have used in the workshop

#489 shimomut closed 2 weeks ago
0
Improvements/nccl efa update

#488 mhuguesaws closed 2 weeks ago
0
Automate onboarding smhp slurm

#487 amanshanbhag closed 2 weeks ago
3
Generate keypair by default

#486 sean-smith closed 1 week ago
2
add keypair to ~/.ssh/config file

#485 sean-smith closed 2 weeks ago
0
Added onboarding automation script

#484 amanshanbhag closed 2 weeks ago
0
Create fsdp-simple.yaml

#483 KeitaW closed 2 weeks ago
0
Unifying CF template for FSxL for HyperPod Slurm and EKS

#482 shimomut closed 2 weeks ago
3
Add llama3.1 training support in FSDP test case

#481 KeitaW opened 2 weeks ago
0
FutureWarning: `torch.cpu.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cpu', args...)` instead. in FSDP test case

#480 KeitaW opened 2 weeks ago
0
Deprecated dataset "c4"

#479 KeitaW opened 2 weeks ago
0
Sub optimal performance in FSDP test case.

#478 KeitaW opened 2 weeks ago
0
Add github stars to readme file

#477 perifaws closed 2 weeks ago
0
Remove cuda compat

#476 jahaniam closed 3 weeks ago
1
EFA Does not Work on New NVIDA Driver and CUDA Versions `system has unsupported display driver `

#475 kamal-rahimi closed 3 weeks ago
11
Update README.md on NCCL tests example to include MPIOperator setup

#474 KeitaW closed 3 weeks ago
0
don't display login in motd if no login node

#473 sean-smith closed 2 weeks ago
0
nemofw-training is probably deprecated

#472 KeitaW opened 3 weeks ago
0
HyperPod Mountpoint for s3 Lifecycle Script

#471 nghtm closed 3 weeks ago
2
Head node and login node ip in motd

#470 sean-smith closed 4 weeks ago
3
Update requirements.txt

#469 nghtm closed 3 weeks ago
0
Update motd.sh

#468 nghtm closed 4 weeks ago
1
FSDP sample fails with CUDA initialization error on HyperPod EKS

#467 shimomut opened 4 weeks ago
6
Change aws-ofi-plugin for EFA 1.35.0 due to regression

#466 mhuguesaws closed 1 month ago
0
Spelling nits in README.md

#465 jimburtoft closed 1 month ago
0
Update pcluster architecture guidance

#464 KeitaW opened 1 month ago
2
Providing updated pcluster guidance for GENIAC participants.

#463 KeitaW closed 1 month ago
0
add GPU accounting for SMHP

#462 KeitaW opened 1 month ago
1
Bump deepspeed from 0.9.2 to 0.15.1 in /3.test_cases/13.SM-dataparallel-deepspeed/code

#461 dependabot[bot] closed 1 month ago
0
FSDP sample fails model validation

#460 shimomut opened 1 month ago
0
Adding the CF template for FSx Lustre for HyperPod EKS workshop

#459 shimomut closed 1 month ago
5
Install numpy v1 specifically for the FSDP sample app

#458 shimomut closed 1 month ago
0
17.SM-modelparallelv2 uses pytorch binary that depends on deprecated conda packages

#457 junpuf opened 1 month ago
1