Skip to content

server : don't overfill the batch during infill #10018

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 1 commit into from
Oct 28, 2024
Merged

server : don't overfill the batch during infill #10018

Show file tree
Hide file tree
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
1 change: 1 addition & 0 deletions examples/server/server.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1880,6 +1880,7 @@ struct server_context {
if (slot.state == SLOT_STATE_STARTED) {
slot.t_start_process_prompt = ggml_time_us();
slot.t_start_generation = 0;

slot.n_past = 0;
slot.n_prompt_tokens = prompt_tokens.size();
slot.state = SLOT_STATE_PROCESSING_PROMPT;
Expand Down
6 changes: 4 additions & 2 deletions examples/server/utils.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -266,8 +266,10 @@ static llama_tokens format_infill(
}

// for now pick FIM context to fit in a batch (ratio prefix:suffix = 3:1, TODO: configurable?)
const int n_suffix_take = std::min<int>(tokens_suffix.size(), (n_batch/4));
const int n_prefix_take = std::min<int>(tokens_prefix.size(), 3*(n_batch/4) - 3);
const int n_prefix_take = std::min<int>(tokens_prefix.size(), 3*(n_batch/4));
const int n_suffix_take = std::min<int>(tokens_suffix.size(), std::max<int>(0, (n_batch/4) - (2 + tokens_prompt.size())));

SRV_DBG("n_prefix_take = %d, n_suffix_take = %d, total = %d\n", n_prefix_take, n_suffix_take, (n_prefix_take + n_suffix_take));

// fill the rest of the context with extra chunks
const int n_extra_take = std::min<int>(std::max<int>(0, n_ctx - (n_batch) - 2*n_predict), extra_tokens.size());
Expand Down
Loading