将数据推送到远程存储
通过 DVC 远程仓库，将数据集存储在 S3 或 GCS 中
将数据推送到远程存储 是 CoddyKit 上的免费 MLOps Academy 课时。 这是第 3 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 MLOps Academy 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 MLOps Academy 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
The Cache Is Only Local
So far your data lives in DVC's local cache on one machine. Teammates and CI cannot reach it, so you need a shared place to store it.
Meet the DVC Remote
A DVC remote is cloud or shared storage where your data cache is uploaded. Think of it like a Git remote, but for the heavy data instead of code.
Many Backends Supported
DVC speaks to S3, Google Cloud Storage, Azure, SSH, and even a plain shared folder. You pick whatever storage your team already uses. ☁️
Add an S3 Remote
Use dvc remote add to register storage. The -d flag marks it as the default so push and pull use it automatically.
dvc remote add -d storage s3://my-bucket/dvcstoreCommit the Remote Config
The remote setting lives in .dvc/config. Commit it to Git so every teammate inherits the same remote without extra setup.
git add .dvc/config
git commit -m "Add S3 DVC remote"Credentials Stay Out of Git
DVC reads cloud credentials from your environment or AWS profile, never from the committed config. Secrets stay safe and out of version control.
export AWS_ACCESS_KEY_ID=...
export AWS_SECRET_ACCESS_KEY=...Push Data Up
Run dvc push to upload everything in your cache to the remote. Only files not already there are sent, so it stays fast.
dvc pushPush and Git Push Together
A safe habit: push data first, then push code. That way the .dvc pointers in Git always point to data that already exists in the remote.
dvc push
git pushTeammates Pull Data Down
After cloning and git pull, a colleague runs dvc pull to download the exact datasets the .dvc files reference. Code and data reunite.
git clone <repo>
dvc pullFetch vs Pull
dvc fetch downloads data into the cache only, while dvc pull also restores it into your working folder. Pull is fetch plus checkout in one step.
Dedup Saves Bandwidth
Because the remote is content-addressed, unchanged files are never re-uploaded. Push only the bytes that are genuinely new. 🚀
Quick Check
Which command sends your tracked data from the local cache up to shared storage?
Recap
You added a DVC remote, committed its config, and used dvc push to upload data. Teammates run dvc pull to fetch the exact datasets your pointers reference.
常见问题解答
「将数据推送到远程存储」课时是免费的吗？
是的 — 「将数据推送到远程存储」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 MLOps Academy 课程的其余内容，请升级到 CoddyKit PRO。 MLOps Academy 课程共包含 4 节课。
「将数据推送到远程存储」这节课中我会学到什么？
通过 DVC 远程仓库，将数据集存储在 S3 或 GCS 中 你通过在浏览器中直接运行的动手代码来练习 MLOps Academy，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 MLOps Academy 需要有经验吗？
无需任何先前经验。CoddyKit 上的 MLOps Academy 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 3 节课，共 4 节。
「将数据推送到远程存储」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 MLOps Academy 课中编写并运行代码吗？
能。每节 MLOps Academy 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。