mirror of
https://github.com/ggerganov/llama.cpp.git
synced 2024-12-24 10:24:35 +00:00
Temporary bump the memory buffer size - hopefully fix issues from 483bab2e
This commit is contained in:
parent
f4f5362edb
commit
31572d9665
@ -632,7 +632,7 @@ static bool llama_eval_internal(
|
|||||||
auto & mem_per_token = lctx.mem_per_token;
|
auto & mem_per_token = lctx.mem_per_token;
|
||||||
|
|
||||||
// TODO: fix this hardcoded size
|
// TODO: fix this hardcoded size
|
||||||
static size_t buf_size = 512u*1024*1024;
|
static size_t buf_size = 2048u*1024*1024; // TMP !!!
|
||||||
static void * buf = malloc(buf_size);
|
static void * buf = malloc(buf_size);
|
||||||
|
|
||||||
if (mem_per_token > 0 && mem_per_token*N > buf_size) {
|
if (mem_per_token > 0 && mem_per_token*N > buf_size) {
|
||||||
|
Loading…
Reference in New Issue
Block a user