Failure in mlflow.spark.load_model : Random Forrest pretrained model

ashrafkhan94
Databricks Partner

‎08-26-2022 12:17 AM 

model = mlflow.spark.load_model(model_uri=f"models:/{model_name}/{model_version}")

Log:

An error occurred while calling o2861.load.

: org.apache.spark.SparkException: Job aborted due to stage failure: Task 4 in stage 4599.0 failed 4 times, most recent failure: Lost task 4.3 in stage 4599.0 (TID 3270) (10.139.64.5 executor 1): java.lang.AssertionError: assertion failed: Decision Tree load failed. Expected largest node ID to be 53, but found 26

Labels:

sean_owen
Databricks Employee
Databricks Employee

‎09-19-2022 03:37 PM

Hm. Are you training with one version of Spark but loading in another version? though that should be pretty compatible across versions, just trying to rule that in/out.

View solution in original post

Noopur_Nigam
Databricks Employee
Databricks Employee

‎09-30-2022 04:27 AM

Hi @Ashraf Khan​ Did you get a chance to look into Sean's response. Please let us know if you need more help on this.

0 Kudos