-
Notifications
You must be signed in to change notification settings - Fork 0
[serve] Initial version of local_testing_mode (#48477)
#5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: frances/test_promptless3
Are you sure you want to change the base?
Conversation
Adds initial support for a (currently private) `_local_testing_mode` flag to `serve.run`. The intention for this flag is to enable users to write unit tests for their application and model composition logic. User code for each deployment will be run in a background thread using the `UserCallableWrapper` that runs in replica actors. There is a new `Router` and `ReplicaResult` implementation that interacts with the user code to enable the existing `DeploymentHandle` code to work. Before merging: - [x] Figure out why `RAY_SERVE_FORCE_LOCAL_TESTING_MODE=1 pytest -vs tests/test_handle_cancellation.py` hangs locally Follow-ups to this PR: - Fix blocking .result() in resolving args for composition - Make get_replica_context() and other auxiliary APIs work - Support FastAPI TestClient --------- Signed-off-by: Edward Oakes <[email protected]>
|
📝 Documentation updates detected! Promptless has analyzed your changes and created a documentation update PR. You can review the proposed documentation changes here: Please review the suggested updates to ensure they accurately reflect your changes. |
|
📝 Documentation updates detected! Promptless has analyzed your changes and created a documentation update PR. You can review the proposed documentation changes here: Please review the suggested updates to ensure they accurately reflect your changes. |
|
📝 Documentation updates detected! Promptless has analyzed your changes and created a documentation update PR. You can review the proposed documentation changes here: Please review the suggested updates to ensure they accurately reflect your changes. |
|
📝 Documentation updates detected! Promptless has analyzed your changes and created a documentation update PR. You can review the proposed documentation changes here: Please review the suggested updates to ensure they accurately reflect your changes. |
|
📝 Documentation updates detected! Promptless has analyzed your changes and created a documentation update PR. You can review the proposed documentation changes here: Please review the suggested updates to ensure they accurately reflect your changes. |
|
✅ No documentation updates required Promptless has analyzed your changes against existing documentation and determined that no updates are needed at this time. If you believe documentation updates are needed, please update the relevant files manually. |
|
📝 Documentation updates detected! Promptless has analyzed your changes and created a documentation update PR. You can review the proposed documentation changes here: Please review the suggested updates to ensure they accurately reflect your changes. |
Adds initial support for a (currently private)
_local_testing_modeflag toserve.run. The intention for this flag is to enable users to write unit tests for their application and model composition logic.User code for each deployment will be run in a background thread using the
UserCallableWrapperthat runs in replica actors. There is a newRouterandReplicaResultimplementation that interacts with the user code to enable the existingDeploymentHandlecode to work.Before merging:
RAY_SERVE_FORCE_LOCAL_TESTING_MODE=1 pytest -vs tests/test_handle_cancellation.pyhangs locallyFollow-ups to this PR: