cfn/heterogeneous-distributed-training-framework
Code Issues Proposed changes
5 Commits 4 Branches 0 Tags
6f920623fcdeae94cd698af01445d81c8a8d7889
Go to file
Clone
Open with VS Code Open with VSCodium Open with Intellij IDEA
tianyutong 6f920623fc 实现pp相关逻辑 
Change-Id: If203d8b02ebe3b4669a4333398661f476cc3abd4
2025-05-23 10:09:01 +08:00
.github
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
docs
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
examples
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
images
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
megatron
实现pp相关逻辑
2025-05-23 10:09:01 +08:00
tasks
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
tests
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
tools
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
.coveragerc
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
.gitignore
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
.gitlab-ci.yml
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
.gitreview
Added .gitreview and .zuul.yaml
2025-05-14 13:34:45 +00:00
.zuul.yaml
Added .gitreview and .zuul.yaml
2025-05-14 13:34:45 +00:00
CODEOWNERS
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
CONTRIBUTING.md
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
Dockerfile.ci
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
Dockerfile.linting
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
jet-tests.yml
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
LICENSE
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
MANIFEST.in
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_bert.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_gpt.py
实现dp相关逻辑
2025-05-23 10:04:15 +08:00
pretrain_ict.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_mamba.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_retro.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_t5.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_vision_classify.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_vision_dino.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_vision_inpaint.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pretrain_vlm.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
pyproject.toml
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00
README.md
add readme
2025-05-15 10:45:47 +08:00
setup.py
Initial Commit of Megatron-LM-0.8.0
2025-05-23 09:54:48 +08:00

README.md

Heterogeneous Distributed Training Framework

This repositoy is used for CFN sub-group Heterogeneous Distributed Training Framework

Description
Heterogeneous distributed training framework researches the training method and framework over heterogeneous hardware.
Readme 6.6 MiB
Languages
Python 97.7%
Shell 1.3%
C++ 0.8%