This message was deleted.
# announcementss
Slackbot
12/14/2019, 5:06 AMThis message was deleted.
s
suhas
12/14/2019, 5:10 AMhow to correct it
b
Bo
12/14/2019, 5:12 AMLet me run the notebook very quick with 0.5.3 version, and go from there.
s
suhas
12/14/2019, 5:13 AMokies , or could you let me know the other stable version, I will install and run the same
b
Bo
12/14/2019, 5:19 AMi just ran the notebook with 0.5.3. It works. I think there is something went wrong during the process of saving bento bundle, and the storage location is not correctly recorded.
I think what we could do now is you rerun the cell that save bento bundle. and we can try again with deployment after that finished.
s
suhas
12/14/2019, 5:25 AMpython bentomlsentiment.py
Killed
suhas
12/14/2019, 5:48 AMThis is how my sentiment file looks like
b
Bo
12/14/2019, 5:49 AM Copy code
bento_tag = '{name}:{version}'.format(name=bento_service.name, version=bento_service.version)
print(bento_tag)Bo
12/14/2019, 5:50 AMname=bento_service.name -> name=loaded_bento_service.name
Bo
12/14/2019, 5:50 AMsame for the version
s
suhas
12/14/2019, 5:58 AMloaded_bento_service = bentoml.load(saved_path)
# Call predict on the restored BentoService
loaded_bento_service.predict(pd.Series(["hello", "hi"]))
bento_tag = '{name}:{version}'.format(name=loaded_bento_service.name, version=bento_service.version)
print(bento_tag)
~
suhas
12/14/2019, 5:59 AMafter this, tried again with python bentomlsentiment.py
Killed
b
Bo
12/14/2019, 5:59 AMVersion=loaded_bento_service.version as well
s
suhas
12/14/2019, 5:59 AMsame output Killed
suhas
12/14/2019, 6:01 AMagain same error "killed"
suhas
12/14/2019, 6:01 AMbento_service = SentimentLRModel.pack(model=sentiment_lr)
# 3) save BentoSerivce to file archive
saved_path = bento_service.save()
# Load exported bentoML model archive from path
loaded_bento_service = bentoml.load(saved_path)
# Call predict on the restored BentoService
loaded_bento_service.predict(pd.Series(["hello", "hi"]))
bento_tag = '{name}:{version}'.format(name=loaded_bento_service.name, version=loaded_bento_service.version)
print(bento_tag)
~
suhas
12/14/2019, 6:29 AMshall i try this with different version, could not able to resolve this issue
suhas
12/14/2019, 6:29 AMI am using ec2 amazon linux instance
2 Views