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

server : reuse llama_sample_token common util (#3494)

Browse source 
* server : reuse llama_sample_token common function

* common : use n_probs for temperature sampling
This commit is contained in:
Jhen-Jie Hong 2023-10-06 07:44:24 -05:00 • committed by GitHub
parent 16820a5a0d
commit 97af49fa39
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 16 additions and 93 deletions
Download patch file Download diff file Expand all files Collapse all files

9
common/common.cpp
View file

@ -1020,10 +1020,11 @@ llama_token llama_sample_token(
id = llama_sample_token_mirostat_v2(ctx, &cur_p, mirostat_tau, mirostat_eta, &mirostat_mu);
} else {
// Temperature sampling
llama_sample_top_k (ctx, &cur_p, top_k, 1);
llama_sample_tail_free (ctx, &cur_p, tfs_z, 1);
llama_sample_typical (ctx, &cur_p, typical_p, 1);
llama_sample_top_p (ctx, &cur_p, top_p, 1);
size_t min_keep = std::max(1, params.n_probs);
llama_sample_top_k (ctx, &cur_p, top_k, min_keep);
llama_sample_tail_free (ctx, &cur_p, tfs_z, min_keep);
llama_sample_typical (ctx, &cur_p, typical_p, min_keep);
llama_sample_top_p (ctx, &cur_p, top_p, min_keep);
llama_sample_temp(ctx, &cur_p, temp);
{