fix(config): use 1.0e-2 for moe_aux_loss_coeff to ensure correct float parsing (#234)

Author: Xiaoming-AMD

primus/configs/models/megatron/grok_base.yaml

@@ -12,6 +12,6 @@ num_query_groups: 8
 num_experts: 8
 moe_router_topk: 2
 moe_router_load_balancing_type: none
-moe_aux_loss_coeff: 1e-2
+moe_aux_loss_coeff: 1.0e-2
 moe_grouped_gemm: true
 moe_token_dispatcher_type: alltoall

primus/configs/models/megatron/mixtral_base.yaml

@@ -12,6 +12,6 @@ num_query_groups: 8
 num_experts: 8
 moe_router_topk: 2
 moe_router_load_balancing_type: aux_loss
-moe_aux_loss_coeff: 1e-2
+moe_aux_loss_coeff: 1.0e-2
 moe_grouped_gemm: true
 moe_token_dispatcher_type: alltoall