Skip to content

Latest commit

 

History

History

l4t-tlt-r44.7.1

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
.ipynb_checkpoints
.ipynb_checkpoints
 
 
README.md
README.md
 
 

README.md

l4t-tlt

License: MIT

*** Copy Right 2020 Kevin Yu. All rights reserved.

*** Author: Kevin Yu

*** Release date: 2020/07/29

*** Update Time: 2020/07/29

*** Website: www.hikariai.net

The l4t-tlt docker image integrates the Transfer Learning ToolKit developed by NVIDIA with CUDA support. It may help you quickly run inference based on the custom-trained SSD model on the Jetson Devices. This container is compatible with Jetson Nano, TX1/TX2, Xavier NX, and AGX Xavier with the latest JetPack 4.4(L4T R32.4.3) Release.

Package Versions

  • CUDA 10.2
  • TensorRT 7.1.3
  • OpenCV 4.1.1
  • TensorRT-OSS 7.1.3
  • Tensorflow 1.15
  • Pycuda 2019.1.2
  • jupyterlab 2.2
  • ipykernel 5.3.3

Suported Architecture

ARMv8 Nvidia Jetson Platform

Supported Device

NVIDIA Jetson Nano

Notes:

  • This image also works with Jetson Xavier but it will be only using 4 CPU cores.

Application Setup

Use the Pre-Built Image from DockerHub

$ docker pull hikariai/l4t-tlt-r44.7.1:nano

Use the Pre-Built Image from AliCloud (Users from China ONLY)

$ docker pull registry.cn-hangzhou.aliyuncs.com/hikariai/l4t-tlt-r44.7.1:latest

Usage

Run the container

$ sudo docker run --name tlt -it --runtime nvidia -p 3001:8888 hikariai/l4t-tlt-r44.7.1:nano bash
$ jupyter lab --ip 0.0.0.0 --port 8888 --allow-root

For users from China ONLY

$ docker run --name tlt -it --runtime nvidia -p 3001:8888 registry.cn-hangzhou.aliyuncs.com/hikariai/l4t-tlt-r44.7.1:latest bash
$ jupyter lab --ip 0.0.0.0 --port 8888 --allow-root

Notes:

  • Open up a new browser and visit http://localhost:3001, and you should be able to log into JupyterLab with the token displayed in your console.
  • The detailed usage instruction of the tool, including all the setup steps, is available in the notebook

Parameters

FLAG USAGE
-d run the container in detached mode (background mode)
-it run the container in interactive mode
--rm delete the container when it finishes its process
--runtime use a specify runtime (NVIDIA runtime) while running the container
-v add a mounting directory from the host to access and save files inside or outside the container
--name specify the name of the container
--device map a host device to the container
-p map the container port to the host port

Version Tags

NAME VERSION
hikariai/l4t-tlt-r44.7.1 latest (7.1)

License

MIT License