root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-10 18:51:45 +00:00
Code Issues Actions 13 Packages Projects Releases Wiki Activity

Release encode block when re-setting encoding buffer count if needed

Browse Source
This commit is contained in:
Paul Tsochantaris 2024-10-07 12:21:28 +01:00
parent 5e6358398c
commit 4af03de2a6
1 changed files with 4 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-metal.m
View File

@ -3428,6 +3428,10 @@ static void ggml_backend_metal_set_n_cb(ggml_backend_t backend, int n_cb) {
} }
} }
if(ctx->encode_async) {
Block_release(ctx->encode_async);
}
ctx->encode_async = Block_copy(^(size_t iter) { ctx->encode_async = Block_copy(^(size_t iter) {
const int cb_idx = iter; const int cb_idx = iter;
const int n_cb_l = ctx->n_cb; const int n_cb_l = ctx->n_cb;