BentoML currently does not support that, but I think it’s fairly easy to add this support, assuming the pip dependencies are not changing between versions

Another approach, which I think is more preferable, is to build a separate pipeline that monitors new training data in s3 and triggers training job - if you are using BentoML for model serving deployment, BentoML supports updating an existing deployment to a newer version - that work is still working in progress though. Support for deployment update in AWS lambda and AWS fargate are coming soon.

This message was deleted.

This message was deleted.

Failed to create deployment sentiment-serverless. INTERNAL:BentoML currently not support 0 repository

getting the above error when i triggered

bentoml deployment create sentiment-serverless --bento {bento_tag} --platform aws-lambda --region us-west-2

Hi @suhas. Thanks for the reporting.

I don’t have a good idea what cause that at the moment. Can you retry this with a ‘—verbose’ flag?

bentoml --verbose deployment create sentiment-serverless --bento {bento_tag} --platform aws-lambda --region us-west-2 Usage: bentoml deployment create [OPTIONS] NAME Error: Invalid value for "-b" / "--bento": Bad formatting. Please present in BentoName:Version, for example iris_classifier:v1.2.0

Also what bentoml version you are using?

bentoml --version bentoml, version 0.5.3

ok, thanks.

Could you show me how you defined your Bento Service

This message was deleted.

trying to run the same

is it possible to connect from google colab to aws or we have to work on aws ec2 instance to work efficiently on AWS lambda and bentoml?

Yeah. I think it is possible to use google colab to deploy to aws lambda. You will need to export aws configurations as your envvar

You can find example of it at the QuickStart notebook

check the given url once, not working

Fix the typo

Try again

this is the link, thought you provided me

/usr/local/lib64/python3.6/site-packages/sklearn/feature_extraction/text.py in _count_vocab(self, raw_documents, fixed_vocab) 1140 continue 1141 -> 1142 j_indices.extend(feature_counter.keys()) 1143 values.extend(feature_counter.values()) 1144 indptr.append(len(j_indices)) MemoryError: 12/14/2019 073927 AM INFO: Shutdown kernel 12/14/2019 073927 AM WARNING: Exiting with nonzero exit status

tried to run the code given in the above