MCPcopy Create free account
hub / github.com/NVIDIA/TensorRT-LLM / model_files

Function model_files

tests/unittest/test_model_runner_cpp.py:14–21  ·  view source on GitHub ↗
(llm_root: Path, resource_path: Path, results_data_path: Path)

Source from the content-addressed store, hash-verified

12
13@pytest.fixture
14def model_files(llm_root: Path, resource_path: Path, results_data_path: Path):
15 # Model engines and expected outputs need to be generated.
16 print(results_data_path)
17 if not results_data_path.exists():
18 model_cache = llm_models_root()
19 model_cache_arg = ["--model_cache", str(model_cache)
20 ] if model_cache is not None else []
21 prepare_model_tests(llm_root, resource_path, "gpt", model_cache_arg)
22
23
24def test_logits_post_processor(model_files, model_path):

Callers

nothing calls this directly

Calls 2

llm_models_rootFunction · 0.90
prepare_model_testsFunction · 0.50

Tested by

no test coverage detected