CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-17 11:37:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

CANN: optimize rope operator (#15335)

Browse Source 
* optimize rope ops

* amendment

* delete trailing whitespace

* change the variable name
This commit is contained in:
SHUAI YANG
2025-08-19 21:28:22 +08:00
committed by GitHub
parent 67f09a3a27
commit a6d3cfe7fa
2 changed files with 117 additions and 62 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
ggml/src/ggml-cann/common.h
View File

@@ -368,6 +368,10 @@ struct ggml_backend_cann_context {
std::string name; /**< Name of the device. */
std::string description; /**< Description of the device. */
aclrtEvent copy_event = nullptr; /**< Event for managing copy operations. */
void* init_ptr = nullptr;
void* sin_ptr = nullptr;
void* cos_ptr = nullptr;
int64_t max_prompt_length = 65536;
#ifdef USE_ACL_GRAPH
/// Cached CANN ACL graph used for executing the current ggml computation graph.
std::unique_ptr<ggml_cann_graph> cann_graph;
@@ -414,6 +418,15 @@ struct ggml_backend_cann_context {
ACL_CHECK(aclrtDestroyStream(streams[i]));
}
}
if(init_ptr != nullptr) {
ACL_CHECK(aclrtFree(init_ptr));
}
if(sin_ptr != nullptr) {
ACL_CHECK(aclrtFree(sin_ptr));
}
if(cos_ptr != nullptr) {
ACL_CHECK(aclrtFree(cos_ptr));
}
}
/**