root
/
tabby
mirror of https://gitee.com/zhang_1334717033/tabby.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

feat: update llama.cpp to fetch latest starcoder support (#452) 

* feat: bump llama.cpp to HEAD

* fix: turn off add_bos by default
release-0.2
Meng Zhang 2023-09-16 11:41:49 +08:00 committed by GitHub
parent 86e48afbe0
commit 97eeb6b926
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
crates/llama-cpp-bindings/llama.cpp

@ -1 +1 @@
Subproject commit 06fc4020de0b92ee13407fdabca7870f53c75de5
Subproject commit 31e7f3c20e1bacb522021e46788e24c045c108f6

2
crates/llama-cpp-bindings/src/engine.cc
View File

@ -47,7 +47,7 @@ class TextInferenceEngineImpl : public TextInferenceEngine {
uint32_t start(const rust::Str prompt, size_t max_input_length) const override {
auto* ctx = ctx_.get();
llama_reset_timings(ctx);
std::vector<llama_token> tokens_list = tokenize(ctx, std::string(prompt), max_input_length, /* add_bos = */ true);
std::vector<llama_token> tokens_list = tokenize(ctx, std::string(prompt), max_input_length, /* add_bos = */ false);
for (size_t i = 0; i < tokens_list.size(); i += N_BATCH) {
const size_t size = std::min(N_BATCH, tokens_list.size() - i);