nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

speculative : fix n_outputs_max and remove draft-simple auto-enable (#23988)

Browse Source 
* speculative : add common_speculative_n_max helper function

Extract the speculative max-draft-size logic from server_n_outputs_max
into a reusable common_speculative_n_max() function in common/speculative.

Assisted-by: llama.cpp:local pi

* cont : draft context always has n_parallel outputs

* llama : log n_outputs_max

* speculative : remove draft-simple auto-enable

* ci : enable server tests on PRs
This commit is contained in:
Georgi Gerganov
2026-06-01 22:26:58 +03:00
committed by GitHub
parent 5aa3a64596
commit 5dcb711666
6 changed files with 40 additions and 51 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-context.cpp
+1
View File
@@ -229,6 +229,7 @@ llama_context::llama_context(
LLAMA_LOG_INFO("%s: freq_base = %.1f\n", __func__, cparams.rope_freq_base);
LLAMA_LOG_INFO("%s: freq_scale = %g\n", __func__, cparams.rope_freq_scale);
LLAMA_LOG_INFO("%s: n_rs_seq = %u\n", __func__, cparams.n_rs_seq);
LLAMA_LOG_INFO("%s: n_outputs_max = %u\n", __func__, cparams.n_outputs_max);
if (cparams.n_ctx_seq < hparams.n_ctx_train) {
LLAMA_LOG_WARN("%s: n_ctx_seq (%u) < n_ctx_train (%u) -- the full capacity of the model will not be utilized\n",