Support for Models stored in GCS bucket

[kainspraveen]

Is your feature request related to a problem? Please describe.
More often than not in organizations developers don't have access to download models directly from huggingface. It's ususally located in a cloud storage. Would like to add a feature that can directly take gcs link as input inside config.yaml and continue with training the model.

Describe the use case

Describe the solution you'd like
config yaml file will take gcs link as input just like it takes a local storage location of the model. It will load the model into a temporary storage and continue with regular fine-tuning or serving etc..

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

[kainspraveen]

would like to assign myself to this issue. #take #self-assign

[alexsherstinsky]

@kainspraveen Thank you for assigning this issue to yourself. Have you been able to make any progress on it? How can Ludwig maintainers be of help? Please consider joining our Discord and attending our monthly Zoom meetings! Thank you!

[mhabedank]

Hi @kainspraveen is it still relevant to you? Have you done some work on this?