vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-10 17:14:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'master' into concedo_experimental

Browse source 
# Conflicts:
#	README.md
#	build.zig
#	flake.nix
#	tests/test-grad0.c
#	tests/test-sampling.cpp
#	tests/test-tokenizer-0.cpp
This commit is contained in:
Concedo 2023-06-25 17:01:15 +08:00
commit d2034ced7b
19 changed files with 346 additions and 149 deletions
Download patch file Download diff file Expand all files Collapse all files

4
gpttype_adapter.cpp
View file

@ -78,6 +78,7 @@ static std::vector<int> smartcontext;
static std::vector<std::string> stop_sequence;
static std::vector<llama_token_data> top_picks;
static int remaining_tokens = 0;
static int stopper_unused_tokens = 0;
static std::string concat_output = "";
inline bool IsNanCheck(float f)
@ -759,6 +760,7 @@ ModelLoadResult gpttype_load_model(const load_model_inputs inputs, FileFormat in
bool gpttype_generate_abort()
{
stopper_unused_tokens = remaining_tokens;
remaining_tokens = 0;
return true;
}
@ -899,7 +901,7 @@ generation_outputs gpttype_generate(const generation_inputs inputs, generation_o
current_context_tokens.resize(n_past);
remaining_tokens = params.n_predict;
int stopper_unused_tokens = 0;
stopper_unused_tokens = 0;
int input_consumed = 0;
std::mt19937 rng(params.seed);
concat_output = "";