Distribute and run LLMs with a single file.
C++ cross-platformggufllama-cpplocal-ailocal-inferencelocal-llmopen-source-aisingle-file-executablespeech-to-text
Add missing GGML function to CUDA module
Justine Tunney09fdeb3dab24
diff --git a/llama.cpp/ggml-cuda.cu b/llama.cpp/ggml-cuda.cuindex da6c995e..8dac1b98 100644--- a/llama.cpp/ggml-cuda.cu+++ b/llama.cpp/ggml-cuda.cu@@ -223,6 +223,11 @@ static int64_t ggml_nrows(const struct ggml_tensor * tensor) {return tensor->ne[1]*tensor->ne[2]*tensor->ne[3];}+#define ggml_is_permuted ggml_is_permuted_+static bool ggml_is_permuted(const struct ggml_tensor * tensor) {+ return tensor->nb[0] > tensor->nb[1] || tensor->nb[1] > tensor->nb[2] || tensor->nb[2] > tensor->nb[3];+}+#define ggml_is_contiguous ggml_is_contiguous_static bool ggml_is_contiguous(const struct ggml_tensor * tensor) {return