Update test_llm_api_pytorch.py

Signed-off-by: Yi Zhang <187001205+yizhang-nv@users.noreply.github.com>

Author: yizhang-nv

tests/integration/defs/accuracy/test_llm_api_pytorch.py

@@ -1330,7 +1330,7 @@ class TestDeepSeekR1(LlmapiAccuracyTestHarness):
     def test_nvfp4_multi_gpus(self, tp_size, pp_size, ep_size, mtp_nextn, fp8kv,
                               attention_dp, cuda_graph, overlap_scheduler,
                               max_batch_size, moe_backend):
-        kv_cache_config = KvCacheConfig(free_gpu_memory_fraction=0.80)
+        kv_cache_config = KvCacheConfig(free_gpu_memory_fraction=0.70)
         pytorch_config = dict(
             disable_overlap_scheduler=not overlap_scheduler,
             cuda_graph_config=CudaGraphConfig() if cuda_graph else None,