skip rpc tests

Signed-off-by: Yan Chunwei <[email protected]>

Author: Superjomn

tests/unittest/llmapi/test_llm_multi_gpu_pytorch.py

@@ -62,6 +62,7 @@ def test_llama_7b_multi_lora_tp2():
         cuda_graph_config=None)
 
 
+@pytest.mark.skip(reason="https://nvbugs/5560921")
 @skip_ray
 @pytest.mark.gpu2
 def test_llm_rpc_tp2():
@@ -80,6 +81,7 @@ def test_llm_rpc_tp2():
         assert len(res.outputs[0].token_ids) == 10
 
 
+@pytest.mark.skip(reason="https://nvbugs/5560921")
 @skip_ray
 @pytest.mark.gpu2
 @pytest.mark.asyncio

tests/unittest/llmapi/test_llm_pytorch.py

@@ -956,6 +956,7 @@ def test_max_num_token_check(self):
             llm.generate([ids])
 
 
+@pytest.mark.skip(reason="https://nvbugs/5560921")
 @skip_ray
 def test_llm_rpc():
     # TODO: remove the with-statement when shutdown hang issue is fixed
@@ -973,6 +974,7 @@ def test_llm_rpc():
         assert len(res.outputs[0].token_ids) == 10
 
 
+@pytest.mark.skip(reason="https://nvbugs/5560921")
 @skip_ray
 @pytest.mark.asyncio
 async def test_llm_rpc_streaming():