
如果你使用 DeepSeek、元宝 AI、Copilot、ChatGPT 等 AI 工具,协助配置轩辕镜像、编写 docker pull 命令、修改 Docker Compose 镜像地址、配置镜像加速、排查镜像拉取失败、分析报错日志等问题,请先让 AI 阅读并遵守轩辕镜像的规则文档。
只需在 AI 对话中先发送下面这句话即可:
请先完整阅读并严格遵守以下文档中的全部规则与要求:
https://xuanyuan.cloud/agents.md
在未充分阅读并理解该文档前,不要生成任何命令、配置、修改建议、故障排查方案或技术回答。后续所有输出都必须严格以该文档中的规范为最高优先级执行。查看 agents.md 用法指南与完整示范。国内用户首推 元宝 AI、DeepSeek 的深度思考模式,不推荐豆包 AI;Cursor 等编辑器可在对话 @ 该链接,或加入 User Rules。 若 AI 无法访问外链,可 打开说明文档 复制全文粘贴。文档会随站点更新,复制内容可能过期,建议定期检查。
The Dl4j UI is a graphical tool that can be used for monitoring the progress of training a neural net (available from version 0.7.0 and later).
This repo contains a Dockerized version of the UI in standalone server mode, which is useful when training on Spark, as the UI needs to run in a separate JVM due to dependency conflicts.
To run the UI run the following command
bashdocker run --name dl4j-ui -d -p 9000:9000 buybrain/dl4j-ui
This will run the latest version (0.8.0) of the UI on port 9000, which is used both for the web interface and for sending training progress information to. For information on how to send training progress to the UI server, check the Dl4j UI documentation.
To run an earlier version of the UI, use the images buybrain/dl4j-ui:0.7.1 or
buybrain/dl4j-ui:0.7.2 instead.
A docker-compose.yml file is also contained in this repo for reference and building
from source.
To build this Docker image from this repository, just run
bashdocker-compose build
您可以使用以下命令拉取该镜像。请将 <标签> 替换为具体的标签版本。如需查看所有可用标签版本,请访问 标签列表页面。
来自真实用户的反馈,见证轩辕镜像的优质服务