Closed laurencejackson closed 1 year ago
Since switching mlflow to using proxied artefact storage (scenario 5: https://mlflow.org/docs/latest/tracking.html#scenario-5-mlflow-tracking-server-enabled-with-proxied-artifact-storage-access) there have been issues relating to fetching large artefacts - especially an issue when serving large models.
We should remove the proxied artefact storage and revert to direct s3 artefact storage.
Since switching mlflow to using proxied artefact storage (scenario 5: https://mlflow.org/docs/latest/tracking.html#scenario-5-mlflow-tracking-server-enabled-with-proxied-artifact-storage-access) there have been issues relating to fetching large artefacts - especially an issue when serving large models.
We should remove the proxied artefact storage and revert to direct s3 artefact storage.