https://community.databricks.com/t5/administration-architecture/what-s-the-right-batch-size-in-deep-learning-training/m-p/21414#M83 <P>Using Ganglia you can monitor how busy is the GPU(s). Increasing the batch size would increase that utilization. Bigger batches improve how well each batch updates the model (up to a point) with more accurate gradients. That in turn can allow training to use a higher learning rate, and more quickly reach the point where the model stops improving.</P> <P><A href="https://databricks.com/blog/2019/08/15/how-not-to-scale-deep-learning-in-6-easy-steps.html" target="test_blank">https://databricks.com/blog/2019/08/15/how-not-to-scale-deep-learning-in-6-easy-steps.html</A></P> Thu, 20 Mar 2025 16:38:01 GMT User16789201666 2025-03-20T16:38:01Z https://community.databricks.com/t5/administration-architecture/what-s-the-right-batch-size-in-deep-learning-training/m-p/21414#M83 <P>Using Ganglia you can monitor how busy is the GPU(s). Increasing the batch size would increase that utilization. Bigger batches improve how well each batch updates the model (up to a point) with more accurate gradients. That in turn can allow training to use a higher learning rate, and more quickly reach the point where the model stops improving.</P> <P><A href="https://databricks.com/blog/2019/08/15/how-not-to-scale-deep-learning-in-6-easy-steps.html" target="test_blank">https://databricks.com/blog/2019/08/15/how-not-to-scale-deep-learning-in-6-easy-steps.html</A></P> Thu, 20 Mar 2025 16:38:01 GMT https://community.databricks.com/t5/administration-architecture/what-s-the-right-batch-size-in-deep-learning-training/m-p/21414#M83 User16789201666 2025-03-20T16:38:01Z