Xgboost algorithm with entry point file issue

Question

I am trying to create aws glue spark job to train one of the data set . I am using xgboost algorithm in 1.3-1 version . When i try to run the estimator , i am having issue

infrastructure : aws glue 4.00 spark shell

all file folders are s3 path

code snippet.

xgb_script_mode_estimator = XGBoost(
    entry_point="training.py",
    hyperparameters=hyperparameters,
    role=role,
    instance_count=1,
    instance_type=instance_type,
    framework_version="1.3-1",
    output_path="s3://{}/{}/{}/output".format(hyperparameters['bucket_nm'], '/output/', job_name),
   

error :

FileNotFoundError: [Errno 2] No such file or directory: 'training.py'

I placed the "glue script" and training.py in the same job bucket in same folder with init.py file .

The XGBoost function is not recognizing the training.py in the same folder (no name mismatch for the training file including case)

Answer 1

This issue resolved by adding "Source-dir" parameter in the XGBboost function by pointing to trianing.py file location