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 15, 21:00 UTC
[Misc] duplicate code in deepseek_v2 (#14106)
wang.yuqi committed
Mar 3, 2025 at 14:10 UTC
e584b85afde5812088b9e7decefc91a66237b226
1 file changed
-1
vllm/model_executor/models/deepseek_v2.py
-1
@@ -105,7 +105,6 @@ class DeepseekV2MoE(nn.Module):
105
self.tp_size = get_tensor_model_parallel_world_size()
106
self.routed_scaling_factor = config.routed_scaling_factor
107
self.n_shared_experts = config.n_shared_experts
108
- self.routed_scaling_factor = config.routed_scaling_factor
108
109
if config.hidden_act != "silu":
110
raise ValueError(f"Unsupported activation: {config.hidden_act}. "