vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-05 19:49:03 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Modified batch backend_buffer size to actual size

Browse source
This commit is contained in:
leeetao 2025-07-04 08:52:15 +00:00
parent 82787be7eb
commit da31acbe6a
1 changed files with 1 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
src/llama.cpp
View file

@ -3052,8 +3052,7 @@ struct llama_sbatch {
ubatch_token.resize(!has_embd ? n_ubatch : 0); ubatch_token.resize(!has_embd ? n_ubatch : 0);
ubatch_embd.resize(has_embd ? n_embd * n_ubatch : 0); ubatch_embd.resize(has_embd ? n_embd * n_ubatch : 0);
// TODO: just a guess and test, need to be removed(from tao) ubatch_backend_embd.resize(n_embd * n_tokens + n_tokens);
ubatch_backend_embd.resize(n_embd * n_tokens * 3);
ubatch_out_embd.resize(n_embd * n_tokens); ubatch_out_embd.resize(n_embd * n_tokens);
ubatch_pos.resize(n_ubatch); ubatch_pos.resize(n_ubatch);