root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-24 10:24:35 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

server : fix infill prompt format

Browse Source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2024-12-08 22:12:11 +02:00
parent e52522b869
commit b8d1b1a5e1
No known key found for this signature in database
GPG Key ID: 449E073F9DC10735
1 changed files with 14 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
examples/server/server.cpp
View File

@ -3380,7 +3380,20 @@ int main(int argc, char ** argv) {
task.id = ctx_server.queue_tasks.get_new_id();
task.index = i;
task.prompt_tokens = std::move(tokenized_prompts[i]);
if (type == SERVER_TASK_TYPE_INFILL) {
task.prompt_tokens = format_infill(
ctx_server.ctx,
data.at("input_prefix"),
data.at("input_suffix"),
data.at("input_extra"),
ctx_server.params_base.n_batch,
ctx_server.params_base.n_predict,
ctx_server.slots[0].n_ctx, // TODO: there should be a better way
ctx_server.params_base.spm_infill,
tokenized_prompts[i]);
} else {
task.prompt_tokens = std::move(tokenized_prompts[i]);
}
task.params = server_task::params_from_json_cmpl(ctx_server.model, ctx_server.params_base, data);
task.id_selected_slot = json_value(data, "id_slot", -1);