CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-12 10:47:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

automatically calculate compute buffer sizes (without graph allocator)

Browse Source
This commit is contained in:
slaren
2023-07-20 02:22:54 +02:00
parent 77ac8deaf1
commit cb205c0d13
5 changed files with 132 additions and 53 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml.c
View File

@@ -4468,7 +4468,7 @@ size_t ggml_get_max_tensor_size(const struct ggml_context * ctx) {
}
struct ggml_backend * ggml_get_ctx_backend(struct ggml_context * ctx) {
return ctx->buffer->backend;
return ctx->buffer->backend_buffer->backend;
}
////////////////////////////////////////////////////////////////////////////////
@@ -4520,7 +4520,7 @@ struct ggml_tensor * ggml_new_tensor_impl(
ggml_assert_aligned(result);
*result = (struct ggml_tensor) {
/*.backend =*/ ctx->buffer->backend,
/*.backend =*/ ggml_get_ctx_backend(ctx),
/*.type =*/ type,
/*.n_dims =*/ n_dims,
/*.ne =*/ { 1, 1, 1, 1 },