test:[nvbug 5415268] add kv_cache_free_gpu_mem_fraction param and llama4 rcca cases (NVIDIA#6430)

Signed-off-by: ruodil <[email protected]>
Signed-off-by: Lanyu Liao <[email protected]>

Author: ruodil

tests/integration/defs/perf/test_perf.py

@@ -375,6 +375,7 @@ def __init__(
         tp_size: int = 1,
         pp_size: int = 1,
         num_gpus: int = 1,
+        kv_cache_free_gpu_mem_fraction: float = 0.9,
     ):
         # The model name.
         self.model_name = model_name
@@ -428,6 +429,8 @@ def __init__(
         self.num_gpus = num_gpus
         # Just build engines
         self.build_only = False
+        # kv cache free gpu mem fraction
+        self.kv_cache_free_gpu_mem_fraction = kv_cache_free_gpu_mem_fraction
 
     def to_string(self,
                   custom_bs: int = None,
@@ -541,6 +544,10 @@ def to_string(self,
         if self.num_gpus > 1:
             entries.append(f"gpus:{self.num_gpus}")
 
+        # Add kv cache free gpu mem fraction.
+        if self.kv_cache_free_gpu_mem_fraction != 0.9:
+            entries.append(f"kv_frac:{self.kv_cache_free_gpu_mem_fraction}")
+
         # Concatenate labels with "-".
         return "-".join(entries)
 
@@ -648,6 +655,11 @@ def load_from_str(self, test_param_labels) -> None:
             self.num_gpus = 1 if not labels[0].startswith("gpus:") else int(
                 labels.pop(0).replace("gpus:", ""))
 
+        if len(labels) > 0:
+            self.kv_cache_free_gpu_mem_fraction = 0.9 if not labels[
+                0].startswith("kv_frac:") else float(
+                    labels.pop(0).replace("kv_frac:", ""))
+
         assert len(
             labels
         ) == 0, f"Invalid test name! Some labels cannot be parsed: {labels}"
@@ -998,7 +1010,8 @@ def get_trtllm_bench_build_command(self, engine_dir) -> list:
             f"--workspace={engine_dir}", f"--model={hf_model_name}",
             f"--model_path={model_dir}", "build", f"--dataset={dataset_path}",
             f"--tp_size={self._config.tp_size}",
-            f"--pp_size={self._config.pp_size}"
+            f"--pp_size={self._config.pp_size}",
+            f"--kv_cache_free_gpu_mem_fraction={self._config.kv_cache_free_gpu_mem_fraction}"
         ]
         max_seq_len = max(self._config.input_lens) + max(
             self._config.output_lens)

tests/integration/test_lists/qa/trt_llm_release_perf_test.yml

@@ -473,19 +473,21 @@ trt_llm_release_perf_test:
 
   #llama_v4_maverick_17b_128e_instruct_fp8
   #pytorch backend
-  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:2000,500-reqs:3000-tp:8-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:500,2000-reqs:3000-tp:8-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:1000,1000-reqs:3000-tp:8-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-input_output_len:128,128-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-input_output_len:512,32-gpus:8]
+  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:2000,500-reqs:3000-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:500,2000-reqs:3000-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:1000,1000-reqs:3000-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-input_output_len:128,128-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-input_output_len:512,32-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  #rcca case
+  - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-input_output_len:20000,2000-reqs:1000-ep:8-tp:8-gpus:8-kv_frac:0.6]
 
   #llama_v4_scout_17b_16e_instruct_fp8
   #pytorch backend
-  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:2000,500-reqs:3000-tp:8-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:500,2000-reqs:3000-tp:8-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:1000,1000-reqs:3000-tp:8-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-input_output_len:128,128-tp:8-gpus:8]
-  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-input_output_len:512,32-tp:8-gpus:8]
+  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:2000,500-reqs:3000-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:500,2000-reqs:3000-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:4096-input_output_len:1000,1000-reqs:3000-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-input_output_len:128,128-ep:8-tp:8-gpus:8-kv_frac:0.6]
+  - perf/test_perf.py::test_perf[llama_v4_scout_17b_16e_instruct_fp8-bench-pytorch-float8-input_output_len:512,32-ep:8-tp:8-gpus:8-kv_frac:0.6]
 
   #deepseek_r1_fp8
   #pytorch backend