Hello All,
I hope to find you well.
I am using Pyabsa within Databricks through a job cluster to classify sentiments.
When deploying through :**
sent_classifier = APCCheckpointManager.get_sentiment_classifier(checkpoint=r"/checkpoints/fast_lcf_bert_Multilingual_acc_87.28_f1_81.33/", auto_device=False, offline=True, cal_perplexity=False)
It starts deploying but at some certain point , it shows the error :
run failed with error message The size of the run output (31458257 bytes) exceeds the limit (31457280 bytes).
This is because my texts are so long and there is that 30mb of output size in databricks job cluster’s output.
Is there a way to omit the logs provided by absa (like a verbose parameter or so ) so that I can resolve the issue please ?
Thank you