diff --git a/backends/gaudi/server/text_generation_server/models/custom_modeling/flash_qwen3_moe_modeling.py b/backends/gaudi/server/text_generation_server/models/custom_modeling/flash_qwen3_moe_modeling.py
index 28a96523..f1e73f46 100644
--- a/backends/gaudi/server/text_generation_server/models/custom_modeling/flash_qwen3_moe_modeling.py
+++ b/backends/gaudi/server/text_generation_server/models/custom_modeling/flash_qwen3_moe_modeling.py
@@ -229,21 +229,8 @@ class Qwen3MoE(nn.Module):
         self.process_group = weights.process_group
 
     def forward(self, x: torch.Tensor) -> torch.Tensor:
-        # router_logits: (num_tokens, n_experts)
         router_logits = self.gate(x)
-        # synchronize(x.device)
-        # real_free_memory = get_free_memory(x.device, 1)
-        # log_master(
-        #     logger.debug,
-        #     f"moe forward 1Free memory real: {real_free_memory / 1e9:.2f}GB"
-        # )
         out = self.moe(x, gating_output=router_logits)
-        # synchronize(x.device)
-        # real_free_memory = get_free_memory(x.device, 1)
-        # log_master(
-        #     logger.debug,
-        #     f"moe forward 2 Free memory real: {real_free_memory / 1e9:.2f}GB"
-        # )
 
         # Reduce sum
         if self.process_group.size() > 1:
diff --git a/backends/gaudi/server/text_generation_server/models/flash_causal_lm.py b/backends/gaudi/server/text_generation_server/models/flash_causal_lm.py
index be24ed03..b6cf271d 100644
--- a/backends/gaudi/server/text_generation_server/models/flash_causal_lm.py
+++ b/backends/gaudi/server/text_generation_server/models/flash_causal_lm.py
@@ -1412,7 +1412,6 @@ class FlashCausalLM(Model):
             aliases=aliases,
             weights_loader=weights_loader,
         )
-        print(f"weights: {weights}")
 
         prefix = None
         model = model_class(prefix, config, weights)