tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-27 03:55:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

refact : fix convert script + zero out KV cache to avoid nans (#3523)

Browse Source 
* refact : fix convert script + zero out KV cache to avoid nans

* ggml : silu(-inf) should never happen

* metal : assert various kernel requirements
This commit is contained in:
Georgi Gerganov
2023-10-09 14:32:17 +03:00
committed by GitHub
parent dcc09d2596
commit fcca0a7004
6 changed files with 51 additions and 91 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/parallel/parallel.cpp
View File

@ -167,7 +167,7 @@ int main(int argc, char ** argv) {
// the max batch size is as large as the context to handle cases where we get very long input prompt from multiple
// users. regardless of the size, the main loop will chunk the batch into a maximum of params.n_batch tokens at a time
llama_batch batch = llama_batch_init(params.n_ctx, 0);
llama_batch batch = llama_batch_init(n_ctx, 0);
int32_t n_total_prompt = 0;
int32_t n_total_gen = 0;