Set VP=None for bf16 B300

Signed-off-by: Raghav Hrishikeshan Mukundan <rmukundan@nvidia.com>

Author: rhmukundan

scripts/performance/configs/qwen/qwen3_workload_base_configs.py

@@ -113,7 +113,6 @@
     BASE_QWEN3_235B_A22B_CONFIG,
     num_gpus=64,
     pipeline_model_parallel_size=8,
-    virtual_pipeline_model_parallel_size=4,
     expert_model_parallel_size=8,
     global_batch_size=1024,
     moe_a2a_overlap=False,