Skip to content

Conversation

gmiodice
Copy link
Contributor

  • Add patch to apply on top of llama.cpp to enable the KleidiAI Int4 matmul micro-kernels
  • llama.cpp base hash: 6fcd1331efbfbb89c8c96eba2321bb7b4d0c40e4

- Add patch to apply on top of llama.cpp to enable the KleidiAI Int4 matmul micro-kernels
- llama.cpp base hash: 6fcd1331efbfbb89c8c96eba2321bb7b4d0c40e4

Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
if (t->data == NULL && t->view_src == NULL) {
this_size = GGML_PAD(ggml_backend_buft_get_alloc_size(buft, t), alignment);
+#if defined(GGML_USE_KLEIDIAI)
+ // Temporary solution to allocate more memore if needed for packing the weights.
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Trivial: typo: memory

Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
@phorsman-arm phorsman-arm merged commit 739fdf1 into Arm-Examples:main Jun 17, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants