From a743d76a01f23038b2c85af1e9048ee836767b44 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Johannes=20G=C3=A4=C3=9Fler?= Date: Thu, 9 May 2024 14:32:02 +0200 Subject: CUDA: generalize FP16 fattn vec kernel (#7061) * CUDA: generalize FP16 fattn vec kernel * disable unsupported head sizes for AMD in test * try AMD fix * fix batch size 2-8 * partially revert changes --- tests/test-backend-ops.cpp | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'tests/test-backend-ops.cpp') diff --git a/tests/test-backend-ops.cpp b/tests/test-backend-ops.cpp index 41718e00..0d66de5d 100644 --- a/tests/test-backend-ops.cpp +++ b/tests/test-backend-ops.cpp @@ -2175,7 +2175,11 @@ static bool test_backend(ggml_backend_t backend, test_mode mode, const char * op test_cases.emplace_back(new test_timestep_embedding()); test_cases.emplace_back(new test_leaky_relu()); +#if defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__) + for (int hs : { 64, 128, }) { // other head sizes not implemented +#else for (int hs : { 64, 80, 128, 256, }) { +#endif // defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__) for (int nh : { 32, }) { for (int kv : { 512, 1024, }) { for (int nb : { 1, 2, 4, 8, }) { -- cgit v1.2.3