lzhengning
/
JittorMirror
mirror of https://github.com/Jittor/Jittor
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #632 from CHEN-Xinsheng/reduce-memory 

reduce SFRL large block size to 5242880
This commit is contained in:
CHEN Xinsheng 2025-03-18 20:09:33 +08:00 committed by GitHub
parent 4c75b24cc9 95a17684fa
commit 9b53d2b5a7
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
python/jittor/src/mem/allocator/sfrl_allocator.cc
View File

@ -15,8 +15,8 @@
namespace jittor {
DEFINE_FLAG(int, use_sfrl_allocator, 1, "Enable sfrl allocator");
DEFINE_FLAG(int64, sfrl_large_block_size_device, 20971520, "sfrl_large_block_size, larger will reduce memory shard, only affect device");
constexpr int64 sfrl_large_block_size_cpu=20971520;
DEFINE_FLAG(int64, sfrl_large_block_size_device, 5242880, "sfrl_large_block_size, larger will reduce memory shard, only affect device");
constexpr int64 sfrl_large_block_size_cpu=5242880;
std::vector<size_t> CachingBlockPool::block_ids;
//start from 1