You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Training a ML model usually requires copying remote data into the local storage (a persistent volume in kubernetes) where it will be used for training. The amount of storage required for storing this data varies, however FuseML does not provide a way to specify the size of that storage, instead, FuseML hardcodes that storage size (2 GiB, see: https://github.com/fuseml/fuseml-core/blob/main/pkg/core/tekton/constants.go#L8). Consequently, FuseML cannot be used for training models with more than 2 GiB of data.
The text was updated successfully, but these errors were encountered:
Training a ML model usually requires copying remote data into the local storage (a persistent volume in kubernetes) where it will be used for training. The amount of storage required for storing this data varies, however FuseML does not provide a way to specify the size of that storage, instead, FuseML hardcodes that storage size (2 GiB, see: https://github.com/fuseml/fuseml-core/blob/main/pkg/core/tekton/constants.go#L8). Consequently, FuseML cannot be used for training models with more than 2 GiB of data.
The text was updated successfully, but these errors were encountered: