Cloud Buckets - S3/ GCS
Import from S3
Once you have made sure the model file structure is in the right format, you can upload the model to Inferless using the"Import Model"
option.
-
Choose the source as
"File"
and choose"Cloud repository"
-
Choose AWS, authenticate, and connect your AWS Account.
-
Enter
Model Name
and paste theS3 Url
link from AWS Sagemaker- Sample S3 link - s3://infer-global-models/sample-module/gpt2-medium-1.5gb.zip
-
Make sure you follow the folder structure as described here while creating the Zip.
-
Make sure to enter the input and output parameters as JSON required for the model.
-
Proceed with the configuration and load the model into Inferless.
Import From Google Cloud buckets
Once you have made sure the model file structure is in the right format, you can upload the model to Inferless using the"Import Model"
option.
-
Choose the source as
"File"
and Choose"Cloud repository"
-
Choose GCP, authenticate and connect your GCP Account.
-
Enter
Model Name
and paste theArtifact Link
link from GCP Cloud Bucket.- A Sample of an artifact link is - gs://pp-samplebucket/gpt2-medium-1.5gb.zip
-
Make sure you follow the folder structure as described here while creating the Zip.
-
Make sure to enter the input and output parameters as JSON required for the model (View here for more Infor).
-
Proceed and load the model into Inferless.