Skip to content
This repository has been archived by the owner on Feb 16, 2023. It is now read-only.

Deploying a model to a grid worker #205

Closed
robert-wagner opened this issue May 9, 2019 · 0 comments
Closed

Deploying a model to a grid worker #205

robert-wagner opened this issue May 9, 2019 · 0 comments
Assignees
Labels
Status: Blocked ✖️ Cannot work on this because of some other incomplete work

Comments

@robert-wagner
Copy link
Contributor

robert-wagner commented May 9, 2019

Part of Epic: #203

Blocked by #204, Models not being serializable in Pysyft.

In order to accomplish #203 we need to be able to easily deploy a model to a remote grid worker so that it can be queried.

From an api standpoint we want this to exist as either a function or method on torch models which are also passed a location and details surrounding how to serve them.
either

model.deploy(location, endpoint="name")

This will require a local hook on top of syft's hook.

or

deploy(model, location, endpoint="name")

or

location.deploy_model(model, endpoint="name")

After this call is made models should behave as if they were hosted as in #204

@hereismari hereismari added the Status: Blocked ✖️ Cannot work on this because of some other incomplete work label May 10, 2019
@hereismari hereismari self-assigned this Jun 13, 2019
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
Status: Blocked ✖️ Cannot work on this because of some other incomplete work
Projects
None yet
Development

No branches or pull requests

2 participants