Skip to content

Add heuristic algorithm for speculative #3006

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

Merged
merged 6 commits into from
Sep 14, 2023
Merged

Add heuristic algorithm for speculative #3006

Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
98230ef
Add heuristic algo for speculative
leng-yue Sep 4, 2023
9248528
Constrain minimum n_draft to 2
leng-yue Sep 4, 2023
dddd784
speculative : improve heuristic impl
ggerganov Sep 5, 2023
4e6e951
Merge branch 'master' into HEAD
ggerganov Sep 5, 2023
d9559b7
speculative : be more rewarding upon guessing max drafted tokens
ggerganov Sep 5, 2023
b5efa62
speculative : fix typos
ggerganov Sep 5, 2023
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
Prev Previous commit
Next Next commit
speculative : be more rewarding upon guessing max drafted tokens
  • Loading branch information
@ggerganov
ggerganov committed Sep 5, 2023
commit d9559b78f3bea1a95b927eca685641c4c0082fc5
4 changes: 2 additions & 2 deletions examples/speculative/speculative.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -189,8 +189,8 @@ int main(int argc, char ** argv) {
LOG("all_accepted = %d\n", all_accepted);

if (all_accepted && n_draft == n_dradt_cur) {
LOG(" - max drafted tokens accepted - n_draft += 2\n");
n_draft += 2;
LOG(" - max drafted tokens accepted - n_draft += 8\n");
n_draft = std::min(30, n_draft + 8);
} else if (all_accepted) {
LOG(" - partially drafted tokens accepted - no change\n");
} else {
Expand Down