LnL-AI
/

dbrx-base-converted-v2

Text Generation

Inference Endpoints

text-generation-inference

Model card Files Files and versions Community

reduce logging verbosity

#3

by winglian - opened Mar 30

base: refs/heads/main

←

from: refs/pr/3

Discussion Files changed

Files changed (1) hide show

modeling_dbrx.py +1 -1

modeling_dbrx.py CHANGED Viewed

@@ -412,7 +412,7 @@ class DbrxFlashAttention2(DbrxAttention):
         **kwargs: Any,
     ) -> Tuple[torch.Tensor, Optional[torch.Tensor],
                Optional[Tuple[torch.Tensor]]]:
-        logger.info(
             'Implicitly setting `output_attentions` to False as it is not supported in Flash Attention.'
         )
         output_attentions = False

         **kwargs: Any,
     ) -> Tuple[torch.Tensor, Optional[torch.Tensor],
                Optional[Tuple[torch.Tensor]]]:
+        logger.debug(
             'Implicitly setting `output_attentions` to False as it is not supported in Flash Attention.'
         )
         output_attentions = False