Skip to content

Input Temperature & Output Temperature #4267

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
wants to merge 1 commit into
base: master
Choose a base branch
from
+4 −1
Draft

Input Temperature & Output Temperature #4267

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 4 additions & 1 deletion common/sampling.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -187,13 +187,16 @@ llama_token llama_sampling_sample(
} else {
// temperature sampling
size_t min_keep = std::max(1, params.n_probs);
float output_temp = 1.0; // WIP, this should be defined separately as its own parameter.

llama_sample_temp (ctx_main, &cur_p, temp); // "Initial temp", this is how it's traditionally implemented
llama_sample_top_k (ctx_main, &cur_p, top_k, min_keep);
llama_sample_tail_free(ctx_main, &cur_p, tfs_z, min_keep);
llama_sample_typical (ctx_main, &cur_p, typical_p, min_keep);
llama_sample_top_p (ctx_main, &cur_p, top_p, min_keep);
llama_sample_min_p (ctx_main, &cur_p, min_p, min_keep);
llama_sample_temp (ctx_main, &cur_p, temp);
llama_sample_temp (ctx_main, &cur_p, temp_last); // "Output temp", this is how llama.cpp chose to implement it


id = llama_sample_token(ctx_main, &cur_p);

Expand Down