ver4a/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

CUDA: fix bad asserts for partial offload (#13337)

Browse source
This commit is contained in:
Johannes Gäßler 2025-05-06 13:58:51 +02:00 • committed by GitHub
parent 764b85627b
commit 2356fb1d53
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 21 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

4
ggml/include/ggml.h
View file

@ -673,11 +673,15 @@ extern "C" {
GGML_API bool ggml_is_3d (const struct ggml_tensor * tensor);
GGML_API int ggml_n_dims (const struct ggml_tensor * tensor); // returns 1 for scalars
// returns whether the tensor elements can be iterated over with a flattened index (no gaps, no permutation)
GGML_API bool ggml_is_contiguous (const struct ggml_tensor * tensor);
GGML_API bool ggml_is_contiguous_0(const struct ggml_tensor * tensor); // same as ggml_is_contiguous()
GGML_API bool ggml_is_contiguous_1(const struct ggml_tensor * tensor); // contiguous for dims >= 1
GGML_API bool ggml_is_contiguous_2(const struct ggml_tensor * tensor); // contiguous for dims >= 2
// returns whether the tensor elements are allocated as one contiguous block of memory (no gaps, but permutation ok)
GGML_API bool ggml_is_contiguously_allocated(const struct ggml_tensor * tensor);
// true for tensor that is stored in memory as CxWxHxN and has been permuted to WxHxCxN
GGML_API bool ggml_is_contiguous_channels(const struct ggml_tensor * tensor);