Hi @GuillaumePower
Thanks for using Fabric Community.
It sounds like you’re experiencing inconsistent behavior when running your Microsoft Fabric Notebook cells, particularly when using the Spacy package for Natural Language Processing (NLP). This could be due to a few reasons:
Resource Limitations:
If you’re using a free Power BI account with Fabric, there might be limitations on the amount of CPU and memory you can use. This could cause your scripts to terminate prematurely if they exceed these limits.
Concurrency Issues:
If you’re running multiple cells or notebooks at the same time, they might be competing for resources. This could cause some cells to finish quickly while others take longer.
Spacy Package:
Spacy is a powerful NLP library, but it can be resource-intensive. If your dataset is large, it might take a while to process. Also, ensure you’re using the latest version of Spacy as older versions might have performance issues.
To troubleshoot, you could try the following:
Monitor Resource Usage:
Check your capacity usage while your notebook is running to see if you’re hitting any limits.
Run Cells Individually:Try running your cells one at a time to see if the issue persists.
Update Libraries: Ensure that all your Python libraries, including Spacy, are up-to-date.
Hope this helps.
