hi i am running rasa with below comment rasa run -m 20191203-124629.tar.gz --enable-api --log-file out.log --cors --endpoints endpoints.yml --remote-storage aws --debug model loaded, but after training the new model is stored in local not in remote storage, i trained the model calling /model/train end point, what i am missing??
Hey @sadque,
Do you get any kind of error or anything in the logs when all this happens? I would be curious to see any logs you might can provide to help on this.
Thanks for your response, not getting any errors, logs same as without remote store, i am using rasa 1.3.6, i think the method persist not calling after model train. i am using /model/train end point to train rasa model
Yes no model is saved in the remote storage by Rasa. You need to do this by yourself.
@sadque Is this giving you any error? I’m assuming you setup the required credentials for aws including the bucket and such?
yes, AWS_SECRET_ACCESS_KEY, AWS_ACCESS_KEY_ID, AWS_DEFAULT_REGION, BUCKET_NAME and AWS_ENDPOINT_URL are added in environmental variable, by running the above run comment i mentioned the model is loaded from aws, but newly trained model is not storing to aws,
hi any new about this please ?