mlcommons / training

Reference implementations of MLPerf™ training benchmarks
https://mlcommons.org/en/groups/training
Apache License 2.0
1.61k stars 555 forks source link

can't git clone the submodule of training #446

Closed helinxiao closed 1 year ago

helinxiao commented 3 years ago

I use below command git clone the training package in Ubuntu 16.04 OS. git_ clone --progress --recurse-submodules https://github.com/mlperf/training.git

But will meet below error fatal: repository 'https://github.com/mlperf/community.git/' not found fatal: clone of 'https://github.com/mlperf/community.git' into submodule path 'community' failed

psyhtest commented 3 years ago

Yep, hit the same issue when trying to build the Docker CPU image for DLRM in Inference

root@bded468bf2fb:~/mlperf# git clone --progress --recurse-submodules https://github.com/mlperf/training.git
Cloning into 'training'...
remote: Enumerating objects: 3850, done.
remote: Total 3850 (delta 0), reused 0 (delta 0), pack-reused 3850
Receiving objects: 100% (3850/3850), 78.90 MiB | 7.81 MiB/s, done.
Resolving deltas: 100% (1509/1509), done.
Checking connectivity... done.
Submodule 'community' (https://github.com/mlperf/community.git) registered for path 'community'
Submodule 'compliance/verify_submission/mlperf_submission_helper/mlp_compliance' (https://github.com/bitfort/mlp_compliance) registered for path 'compliance/verify_submission/mlperf_submission_helper/mlp_compliance'
Submodule 'recommendation/dlrm' (https://github.com/facebookresearch/dlrm.git) registered for path 'recommendation/dlrm'
Cloning into 'community'...
Username for 'https://github.com': 

Clicking on community @ eebf719 from the training repo root results in 404 NOT FOUND.

EtoDemerzel0427 commented 3 years ago

I cannot imagine in such a huge project this problem will remain unsolved for this long. Hi @psyhtest, did you find a way to bypass this?

johntran-nv commented 1 year ago

The DLRM engineers are focused on updating the DLRM benchmark for v3.0 (Spring 2023), at which point we'll deprecate the current version. Please reopen if you really need this for the current version and we can consider it.