https://community.databricks.com/t5/get-started-discussions/speaker-diarization-on-databricks-with-nemo-throwing-error/m-p/108085#M9719 <P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="Nishat_0-1738331572163.png" style="width: 400px;"><img src="https://community.databricks.com/t5/image/serverpage/image-id/14490i961EAB108383FAE2/image-size/medium?v=v2&amp;px=400" role="button" title="Nishat_0-1738331572163.png" alt="Nishat_0-1738331572163.png" /></span>&nbsp;</P><P>The configuration of my compute is&nbsp;<BR /><SPAN>15.4 LTS ML (includes Apache Spark 3.5.0, GPU, Scala 2.12)<BR /></SPAN>Standard_NC8as_T4_v3 on Azure Databricks</P><P>&nbsp;</P> Fri, 31 Jan 2025 14:03:28 GMT Nishat 2025-01-31T14:03:28Z https://community.databricks.com/t5/get-started-discussions/speaker-diarization-on-databricks-with-nemo-throwing-error/m-p/108085#M9719 <P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="Nishat_0-1738331572163.png" style="width: 400px;"><img src="https://community.databricks.com/t5/image/serverpage/image-id/14490i961EAB108383FAE2/image-size/medium?v=v2&amp;px=400" role="button" title="Nishat_0-1738331572163.png" alt="Nishat_0-1738331572163.png" /></span>&nbsp;</P><P>The configuration of my compute is&nbsp;<BR /><SPAN>15.4 LTS ML (includes Apache Spark 3.5.0, GPU, Scala 2.12)<BR /></SPAN>Standard_NC8as_T4_v3 on Azure Databricks</P><P>&nbsp;</P> Fri, 31 Jan 2025 14:03:28 GMT https://community.databricks.com/t5/get-started-discussions/speaker-diarization-on-databricks-with-nemo-throwing-error/m-p/108085#M9719 Nishat 2025-01-31T14:03:28Z https://community.databricks.com/t5/get-started-discussions/speaker-diarization-on-databricks-with-nemo-throwing-error/m-p/108359#M9720 <P>Hi&nbsp;<a href="https://community.databricks.com/t5/user/viewprofilepage/user-id/103118">@Nishat</a>&nbsp;,</P><DIV class=""><DIV class=""><DIV class=""><P class="">It looks like there's a problem with GPU compability. As mentioned in the error message, FlashAttention only supports Ampere GPUs or newer.<BR />According to following thread, GPU architecture you've chosen is not supported<BR /><A href="https://github.com/huggingface/transformers/issues/28188" target="_blank" rel="noopener">RuntimeError: FlashAttention only supports Ampere GPUs or newer. · Issue #28188 · huggingface/transformers</A></P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="szymon_dybczak_0-1738497458699.png" style="width: 400px;"><img src="https://community.databricks.com/t5/image/serverpage/image-id/14524i24AB45DEF47C5B6D/image-size/medium?v=v2&amp;px=400" role="button" title="szymon_dybczak_0-1738497458699.png" alt="szymon_dybczak_0-1738497458699.png" /></span><P>You can try different compute, according to above github thread&nbsp;<SPAN>GPU Type: <STRONG>NVIDIA H100 NVL GPU&nbsp;</STRONG>should work</SPAN></P></DIV></DIV><DIV class=""><DIV class=""><DIV class="">&nbsp;</DIV><DIV class=""><A href="https://learn.microsoft.com/en-us/azure/databricks/compute/gpu#ncads_h100_v5" target="_blank" rel="noopener">GPU-enabled compute - Azure Databricks | Microsoft Learn</A></DIV></DIV></DIV></DIV><DIV class=""><DIV class=""><DIV class=""><DIV class=""><DIV class="">&nbsp;</DIV></DIV></DIV></DIV></DIV> Sun, 02 Feb 2025 12:00:25 GMT https://community.databricks.com/t5/get-started-discussions/speaker-diarization-on-databricks-with-nemo-throwing-error/m-p/108359#M9720 szymon_dybczak 2025-02-02T12:00:25Z