Hi @<1859043976472956928:profile|UpsetWhale84> ! Yes, if you specify the output_uri
in Task.init
to the s3 bucket all artifacts will be stored in s3, including model weights. Also, you can specify upload_uri
in OutputModel.update_weights
to the s3 location if you are uploading the model locally
Answered
Hi Is There Any Way To Store Large Model Weights In S3 But Still Using The Model Directory In Clearml (Just Like How We Store Datasets In S3 In Clearml), As If We Keep Storing Them In The File Server It Might Become Full.
Hi is there any way to store large model weights in s3 but still using the model directory in clearml (just like how we store datasets in s3 in clearml), as if we keep storing them in the file server it might become full.
134 Views
1
Answer
21 days ago
19 days ago
Tags
Similar posts