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

Merge branch 'ggerganov:master' into master

Browse source
This commit is contained in:
MaggotHATE 2024-10-04 21:35:23 +05:00 • committed by GitHub
commit d9c9203a0b
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 4 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/include/ggml-alloc.h
View file

@ -24,7 +24,7 @@ GGML_API void ggml_tallocr_alloc(struct ggml_tallocr * talloc, st
// Graph allocator // Graph allocator
/* /*
Example usage: Example usage:
ggml_gallocr_t galloc = ggml_gallocr_new(ggml_bacckend_cpu_buffer_type()); ggml_gallocr_t galloc = ggml_gallocr_new(ggml_backend_cpu_buffer_type());
// optional: create a worst-case graph and reserve the buffers to avoid reallocations // optional: create a worst-case graph and reserve the buffers to avoid reallocations
ggml_gallocr_reserve(galloc, build_graph(max_batch)); ggml_gallocr_reserve(galloc, build_graph(max_batch));

2
ggml/src/ggml-cuda.cu
View file

@ -2448,6 +2448,7 @@ static void ggml_backend_cuda_synchronize(ggml_backend_t backend) {
GGML_UNUSED(backend); GGML_UNUSED(backend);
} }
#ifdef USE_CUDA_GRAPH
static void set_ggml_graph_node_properties(ggml_tensor * node, ggml_graph_node_properties * graph_node_properties) { static void set_ggml_graph_node_properties(ggml_tensor * node, ggml_graph_node_properties * graph_node_properties) {
graph_node_properties->node_address = node->data; graph_node_properties->node_address = node->data;
graph_node_properties->node_op = node->op; graph_node_properties->node_op = node->op;
@ -2498,6 +2499,7 @@ static bool ggml_graph_node_has_matching_properties(ggml_tensor * node, ggml_gra
return true; return true;
} }
#endif
static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) { static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) {
ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context; ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;

2
scripts/sync-ggml.last
View file

@ -1 +1 @@
e5c233e5edbfcfa1d808b9293de9065035c40751 b77f48b1efa671e094696b99fbf566aac8c87d74