upstream-archive/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-08-05 09:30:49 -05:00
Code Issues Packages Projects Releases Wiki Activity

server: introduce self-speculative decoding

Browse Source
This commit is contained in:
Sascha Rogmann
2025-12-29 20:46:32 +01:00
parent 8f91ca54ec
commit 1fb2658b0d
6 changed files with 105 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
tools/server/server-task.cpp
View File

@@ -234,9 +234,10 @@ task_params server_task::params_from_json_cmpl(
params.sampling.backend_sampling = json_value(data, "backend_sampling", defaults.sampling.backend_sampling);
params.post_sampling_probs = json_value(data, "post_sampling_probs", defaults.post_sampling_probs);
params.speculative.n_min = json_value(data, "speculative.n_min", defaults.speculative.n_min);
params.speculative.n_max = json_value(data, "speculative.n_max", defaults.speculative.n_max);
params.speculative.p_min = json_value(data, "speculative.p_min", defaults.speculative.p_min);
params.speculative.n_min = json_value(data, "speculative.n_min", defaults.speculative.n_min);
params.speculative.n_max = json_value(data, "speculative.n_max", defaults.speculative.n_max);
params.speculative.p_min = json_value(data, "speculative.p_min", defaults.speculative.p_min);
params.speculative.use_self = json_value(data, "speculative.use_self", defaults.speculative.use_self);
params.speculative.n_min = std::min(params.speculative.n_max, params.speculative.n_min);
params.speculative.n_min = std::max(params.speculative.n_min, 0);