AWS Sagemaker ClientError: Unable to initialize the algorithm

Question

Cannot run hyper-parameter auto tuning jobs using the image classification algorithm.

Getting this from Sagemaker job info:

Failure reason ClientError: Unable to initialize the algorithm. Failed to validate input data configuration. (caused by ValidationError) Caused by: Additional properties are not allowed (u'val' was unexpected) Failed validating u'additionalProperties' in schema: {u'$schema': u'http://json-schema.org/draft-04/schema#', u'additionalProperties': False, u'anyOf': [{u'required': [u'train']}, {u'required': [u'validation']}, {u'optional': [u'train_lst']}, {u'optional': [u'validation_lst']}, {u'optional': [u'model']}], u'definitions': {u'data_channel': {u'properties': {u'ContentType': {u'type': u'string'}}, u'type': u'object'}}, u'properties': {u'model': {u'$ref': u'#/definitions/data_channel'}, u'train': {u'$ref': u'#/definitions/data_channel'}, u'train_lst': {u'$ref': u'#/definitions/data_channel'}, u'validation': {u'$ref': u'#/definitio

CloudWatch is giving me this reason:

00:42:35 2018-12-09 22:42:35 Customer Error: Unable to initialize the algorithm. Failed to validate input data configuration. (caused by ValidationError)

Caused by: Additional properties are not allowed (u'val' was unexpected)

Any help please thanks.

Answer 1

as showed in your log, one of input channels was named as val. The correct channel name for validation data should be validation. More details on input configuration can be found here: https://docs.aws.amazon.com/sagemaker/latest/dg/image-classification.html