applying coderabbit suggestion

Signed-off-by: Rakib Hasan <[email protected]>

Author: rakib-hasan

tensorrt_llm/_torch/modules/fused_moe/quantization.py

@@ -6,8 +6,8 @@
 import torch.nn.functional as F
 from torch import nn
 
+import tensorrt_llm.logger as trtllm_logger
 from tensorrt_llm._utils import get_sm_version
-from tensorrt_llm.logger import logger
 from tensorrt_llm.quantization.utils.fp4_utils import (
     float4_sf_dtype, get_reorder_rows_for_gated_act_gemm_row_indices,
     get_shuffle_matrix_a_row_indices, get_shuffle_matrix_sf_a_row_indices)
@@ -743,7 +743,7 @@ def load_weights(self, module: torch.nn.Module, weights: List[Dict],
                     if int(name.split(".")[0]) not in expert_ids:
                         continue
                     weight_name = name.replace("weight_scale_inv", "weight")
-                    logger.debug(f"Resmoothing {weight_name}")
+                    trtllm_logger.logger.debug(f"Resmoothing {weight_name}")
                     weight = weights[weight_name][:]
                     scale = weights[name][:]
                     weights[weight_name], weights[name] = resmooth_to_fp8_e8m0(