A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 10:45 UTC
[Misc] log when using default MoE config (#8971)
Divakar Verma committed
Oct 2, 2024 at 23:31 UTC
01843c89b8ddae00d4a0f0f56b8aa7fbaa3efc42
1 file changed
+3
vllm/model_executor/layers/fused_moe/fused_moe.py
+3
@@ -320,6 +320,9 @@ def get_moe_configs(E: int, N: int,
320
321
# If no optimized configuration is available, we will use the default
322
# configuration
323
+ logger.warning(
324
+ ("Using default MoE config. Performance might be sub-optimal! "
325
+ "Config file not found at %s"), config_file_path)
326
return None
327
328