vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

sampling : hide prev behind API and apply #3661

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2023-10-20 18:26:20 +03:00
parent 7e2b5fb1dd
commit 56ba00b923
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
9 changed files with 119 additions and 105 deletions
Download patch file Download diff file Expand all files Collapse all files

9
common/sampling.h
View file

@ -70,6 +70,12 @@ void llama_sampling_reset(llama_sampling_context * ctx);
// Copy the sampler context
void llama_sampling_cp(llama_sampling_context * src, llama_sampling_context * dst);
// Get the last sampled token
llama_token llama_sampling_last(llama_sampling_context * ctx);
// Get a string representation of the last sampled tokens
std::string llama_sampling_prev_str(llama_sampling_context * ctx_sampling, llama_context * ctx_main, int n);
// Print sampling parameters into a string
std::string llama_sampling_print(const llama_sampling_params & params);
@ -99,4 +105,5 @@ llama_token llama_sampling_sample(
void llama_sampling_accept(
struct llama_sampling_context * ctx_sampling,
struct llama_context * ctx_main,
llama_token id);
llama_token id,
bool apply_grammar);