Whenever I create the pipeline, it suddenly failed. I don’t know why.
Hi @krishnapriya_Anu ,
Could you check the ML logs and let us know the details present in it ?
Also, Let us know what Model you are using and what are the steps followed, so we can confirm on the methods.
May i know where i check the ml log
1.create empty dataset
2.upload a scheme and import the pdf and labeling and export
3.create the ML package
4.Create Pipeline
Looks like a DU Model is being used. However, the Create Pipeline Detailed Steps are missing. Could you also provide them ? Mainly we would need to know what is being configured for Creating the Pipeline.
The Logs for the Failed Pipeline could be found by clicking on that Failed Pipeline and navigate to Logs.
Yes, I am trying to create the DU model. I followed the maintenance steps used for the previous model training those two were successful.
The are the step’s I followed
Could you also check on the Pipeline logs or send the logs here for analysis ?
Can you help me to resolve this
Could you check by disabling the GPU and Start the pipeline and check if it is able to complete successful ?
Also, Noticing there are around 10 days of gap with the last pipeline run.
Could you maybe also check the AI Units assigned to your tenant ? It might all be consumed, if you also had a ML Skill(s) active in place.
If you are using Community cloud, then you would receive the next 5K units in the next month Or when the next half of the Trial Period starts.
Also, do make sure the ML Skills deployed are disabled if not in use, as the deployed ML Skills does consume the AI Units.
Also, I assume the consumption of AI Units is high, it would mean the Skills deployed might have also used GPU, hence the AI units were consumed faster.
This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.