Sagemaker uploading model artifacts without compressing and in a different directory #2392
Replies: 4 comments
-
Hi @inderpartap , it looks like you are saving the model under In your training script, you need to save your model in the |
Beta Was this translation helpful? Give feedback.
-
The Also, I read somewhere that you need to have your models saved as |
Beta Was this translation helpful? Give feedback.
-
The format of the model output is configured by the your training script's model save function and SageMaker will upload contents in Yes, tensorflow-model-server will look for a version number under model-name (model name hashing) directory |
Beta Was this translation helpful? Give feedback.
-
I am facing the same issue. Anyone resolved it? |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Describe the bug
I have a Keras model getting trained using an entry_point script and I am using the following pieces of code to store the model artifacts (in the entry_point script).
Ideally, the model_dir should be
opt/ml/model
and Sagemaker should automatically move the contents of this folder to S3 ass3://<default_bucket>/<training_name>/output/model.tar.gz
When I run the
estimator.fit({'training': training_input_path})
, the training is successful, but the Cloudwatch logs show the following:Even then, Sagemaker does store my model artifacts, with the only difference being that instead of storing them in
s3://<default_bucket>/<training_name>/output/model.tar.gz
, they are now stored unzipped ass3://<default_bucket>/<training_name>/model/model/1/saved_model.pb
along with thevariables
andassets
folder. Because of this,estimator.deploy()
call fails as it is unable to find the artifacts in theoutput/
directory in S3.To reproduce
Estimator code:
System information
A description of your system. Please provide:
Additional context
Add any other context about the problem here.
Beta Was this translation helpful? Give feedback.
All reactions