Prerequisites

  • Architecture: aarch64
  • Chip models: Ascend 910B
  • Host driver: 25.5.0
  • Container toolkit (optional) : Ascend-docker-runtime >= 6.0.RC3

Image contents

Built on

harbor.baai.ac.cn/flagos-runtime/flagos-runtime-ascend-cann8.5.0:2.2.0 open_in_new

Python

3.11

Application package

megatron-core[training]==0.17.1

Launch

Published: harbor.baai.ac.cn/flagos-app/megatron_training0.17.1-ascend-cann8.5.0:2.2.0-0.2.3

The image name is long — assign it to a variable first:

  IMG=harbor.baai.ac.cn/flagos-app/megatron_training0.17.1-ascend-cann8.5.0:2.2.0-0.2.3

The two approaches below are alternatives — pick the one that matches how your host runs containers:

With the container toolkit

Start an interactive shell:

  docker run --rm -it \
  --privileged \
  -e ASCEND_VISIBLE_DEVICES=0,1 \
  $IMG bash

Start the app with its default settings:

  docker run --rm -it \
  --privileged \
  -e ASCEND_VISIBLE_DEVICES=0,1 \
  $IMG

Pass arguments to the launcher:

  docker run --rm -it \
  --privileged \
  -e ASCEND_VISIBLE_DEVICES=0,1 \
  $IMG megatron-train --model-type GPT

Without a toolkit — plain docker / podman

Start an interactive shell:

  docker run --rm -it \
  --device /dev/davinci0 \
  --device /dev/davinci1 \
  --device /dev/davinci_manager \
  --device /dev/devmm_svm \
  --device /dev/hisi_hdc \
  -v /usr/local/Ascend/driver:/usr/local/Ascend/driver \
  -v /usr/local/dcmi:/usr/local/dcmi \
  -v /usr/local/sbin/npu-smi:/usr/local/sbin/npu-smi \
  $IMG bash

Start the app with its default settings:

  docker run --rm -it \
  --device /dev/davinci0 \
  --device /dev/davinci1 \
  --device /dev/davinci_manager \
  --device /dev/devmm_svm \
  --device /dev/hisi_hdc \
  -v /usr/local/Ascend/driver:/usr/local/Ascend/driver \
  -v /usr/local/dcmi:/usr/local/dcmi \
  -v /usr/local/sbin/npu-smi:/usr/local/sbin/npu-smi \
  $IMG

Pass arguments to the launcher:

  docker run --rm -it \
  --device /dev/davinci0 \
  --device /dev/davinci1 \
  --device /dev/davinci_manager \
  --device /dev/devmm_svm \
  --device /dev/hisi_hdc \
  -v /usr/local/Ascend/driver:/usr/local/Ascend/driver \
  -v /usr/local/dcmi:/usr/local/dcmi \
  -v /usr/local/sbin/npu-smi:/usr/local/sbin/npu-smi \
  $IMG megatron-train --model-type GPT
Edit this page

Last updated 25 Sep 2026, 10:04 +0800. history