add test for contextPP4_genTP4

Signed-off-by: Chuang Zhu <[email protected]>

Author: chuangz0

tests/integration/defs/disaggregated/test_configs/disagg_config_ctxpp4_gentp4.yaml

@@ -0,0 +1,36 @@
+model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
+hostname: localhost
+port: 8000
+backend: "pytorch"
+cuda_graph_config: null
+free_gpu_memory_fraction: 0.2
+context_servers:
+  num_instances: 1
+  max_batch_size: 1
+  max_num_tokens: 3000
+  max_seq_len: 4096
+  tensor_parallel_size: 1
+  pipeline_parallel_size: 4
+  kv_cache_config:
+    free_gpu_memory_fraction: 0.2
+    enable_partial_reuse: False
+  disable_overlap_scheduler: True
+  cache_transceiver_config:
+    backend: DEFAULT
+  urls:
+      - "localhost:8001"
+generation_servers:
+  num_instances: 1
+  tensor_parallel_size: 4
+  pipeline_parallel_size: 1
+  max_batch_size: 256
+  max_num_tokens: 4096
+  max_seq_len: 4096
+  kv_cache_config:
+    free_gpu_memory_fraction: 0.2
+    enable_partial_reuse: False
+  disable_overlap_scheduler: True
+  cache_transceiver_config:
+    backend: DEFAULT
+  urls:
+      - "localhost:8002"

tests/integration/defs/disaggregated/test_disaggregated.py

@@ -76,6 +76,8 @@ def get_test_config(test_desc, example_dir, test_root):
         (8, f"{test_configs_root}/disagg_config_ctxtp2pp2_gentp2pp2.yaml"),
         "ctxpp4_genpp4":
         (8, f"{test_configs_root}/disagg_config_ctxpp4_genpp4.yaml"),
+        "ctxpp4_gentp4":
+        (8, f"{test_configs_root}/disagg_config_ctxpp4_gentp4.yaml"),
         "deepseek_v3_lite_fp8_mpi":
         (4,
          f"{test_configs_root}/disagg_config_ctxtp2_gentp2_deepseek_v3_lite_mpi.yaml"
@@ -709,6 +711,26 @@ def test_disaggregated_ctxpp4_genpp4(disaggregated_test_root, llm_venv,
                            cwd=llm_venv.get_working_directory())
 
 
+@pytest.mark.skip_less_device(4)
+@pytest.mark.parametrize("llama_model_root", ['TinyLlama-1.1B-Chat-v1.0'],
+                         indirect=True)
+def test_disaggregated_ctxpp4_gentp4(disaggregated_test_root, llm_venv,
+                                     disaggregated_example_root,
+                                     llama_model_root):
+    src_dst_dict = {
+        llama_model_root:
+        f"{llm_venv.get_working_directory()}/TinyLlama/TinyLlama-1.1B-Chat-v1.0",
+    }
+    for src, dst in src_dst_dict.items():
+        if not os.path.islink(dst):
+            os.makedirs(os.path.dirname(dst), exist_ok=True)
+            os.symlink(src, dst, target_is_directory=True)
+    run_disaggregated_test(disaggregated_example_root,
+                           "ctxpp4_gentp4",
+                           env=llm_venv._new_env,
+                           cwd=llm_venv.get_working_directory())
+
+
 @skip_no_hopper
 @pytest.mark.skip_less_device(4)
 @pytest.mark.parametrize("deepseek_v3_model_root", ['DeepSeek-V3-Lite-fp8'],

tests/integration/test_lists/test-db/l0_dgx_h100.yml

@@ -33,6 +33,7 @@ l0_dgx_h100:
   - disaggregated/test_disaggregated.py::test_disaggregated_ctxpp2_genpp2[TinyLlama-1.1B-Chat-v1.0]
   - disaggregated/test_disaggregated.py::test_disaggregated_ctxtp2_genpp2[TinyLlama-1.1B-Chat-v1.0]
   - disaggregated/test_disaggregated.py::test_disaggregated_ctxpp2_gentp2[TinyLlama-1.1B-Chat-v1.0]
+  - disaggregated/test_disaggregated.py::test_disaggregated_ctxpp4_gentp4[TinyLlama-1.1B-Chat-v1.0]
   - disaggregated/test_disaggregated.py::test_disaggregated_genbs1[TinyLlama-1.1B-Chat-v1.0]
   - accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_auto_dtype[False]
   - accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_auto_dtype[True]