- Newest
- Most votes
- Most comments
If you have granted all permissions need as you said, or check my answer on how to do that here, https://repost.aws/questions/QUQzcE1sD0R3SGY_nwiO57IA/sagemaker-canvas-not-able-to-run-predict-either-batch-prediction-single-prediction-on-my-model. If all that is/are checked, you might wanna check further steps below
-
SageMaker Canvas datasets have a limit of 5 GB. If you're working with data larger than 5 GB, it's recommended to use a sampling transform to reduce the size of your dataset before importing it, or you can consider exporting data to Amazon S3 for further processing.
-
You also want to check the schema of the data meets what canvas expected.
-
Depending on what type of ML model Model you want to build or Quick/Standard Canvas model type, I have seen cases where canvas wouldn't allow me to use a dataset that has over 10,000 tabular row. You might wanna pay attention to this too.
Relevant content
- asked 2 years ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated a year ago
No I didn't
I believe I have found the issue that caused my error at least. If any of the columns have a None/Null value as first value it crashes with the error you describe. But I'm reading a parquet file, not csv. If I convert it first to a csv file, then it works even if the first value is None/Null. So I'm afraid that doesn't help you.