CUDAGraphExecutor - add tests for the graph caching mechanism (#670)

thunder/tests/test_networks.py

@@ -83,18 +83,35 @@ def test_nanogpt_complete_cudagraphs(executor, device, dtype):
 
     tom = executor.make_callable(gpt, use_cudagraphs=True, disable_torch_autograd=True)
 
+    # Checking graph cache stats
+    from thunder.executors.cudagraphex import build_cuda_graph
+
+    # Cache stats before test runs
+    build_graph_stats_old = build_cuda_graph.cache_info()
+
     for _ in range(2):
         idx = make((4, 64), dtype=torch.int64, low=0, high=255)
         torch_result = gpt(idx)
 
         thunder_result = tom(idx)
         assert_close(torch_result, thunder_result)
 
+    # Cache stats after test runs
+    build_graph_stats_new = build_cuda_graph.cache_info()
+    # We ran only a single (forward) graph several times.
+    # Test that at most 1 cache miss happened after the runs.
+    assert (build_graph_stats_new.misses - build_graph_stats_old.misses) <= 1
+
     # Check we really run CUDAGraphExecutor {
     assert tom._lc_cd.use_cudagraphs == True
     assert _there_is_cudagraph_sym(thunder.last_traces(tom)[-1])
     # }
 
+    # Let's clear cache if run only in tests
+    # TODO: merge with the cache of the thunder.jit callable
+    if build_graph_stats_old.misses == 0:
+        build_cuda_graph.cache_clear()
+
 
 @instantiate(dtypes=(thunder.float32,), devicetypes=(thunder.devices.DeviceType.CUDA,))
 @requiresCUDA
@@ -107,6 +124,12 @@ def test_nanogpt_complete_cuda_graphs_autograd(executor, device, dtype):
     gpt = nanogpt_model.GPT(config).to(device=device, dtype=tdtype)
     cmodel = executor.make_callable(gpt, use_cudagraphs=True)
 
+    # Checking graph cache stats
+    from thunder.executors.cudagraphex import build_cuda_graph
+
+    # Cache stats before test runs
+    build_graph_stats_old = build_cuda_graph.cache_info()
+
     # Multiple runs to test whether static buffers are properly updated
     for i in range(3):
         x = make_tensor((4, 64), dtype=torch.int64, low=0, high=255, device=device)
@@ -121,12 +144,24 @@ def test_nanogpt_complete_cuda_graphs_autograd(executor, device, dtype):
         assert_close(torch_result, thunder_result)
         assert_close(torch_grads, thunder_grads)
 
+    # Cache stats after test runs
+    build_graph_stats_new = build_cuda_graph.cache_info()
+    # We ran only at most two (forward and backward) graphs several times.
+    # Test that at most 2 cache misses happened after the runs
+    # (at most one per each graph)
+    assert (build_graph_stats_new.misses - build_graph_stats_old.misses) <= 2
+
     # Check we really run CUDAGraphExecutor {
     assert cmodel._lc_cd.use_cudagraphs == True
     assert _there_is_cudagraph_sym(thunder.last_traces(cmodel)[-1])
     assert _there_is_cudagraph_sym(thunder.last_backward_traces(cmodel)[-1])
     # }
 
+    # Let's clear cache if run only in tests
+    # TODO: merge with the cache of the thunder.jit callable
+    if build_graph_stats_old.misses == 0:
+        build_cuda_graph.cache_clear()
+
 
 @instantiate(dtypes=(thunder.float32,))
 def test_nanogpt_csa(executor, device, dtype):