sdgoij/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-05-10 19:14:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

spec : update common_speculative_init()

Browse Source 
[no ci]
This commit is contained in:
Georgi Gerganov
2026-05-07 10:06:32 +03:00
parent 2466149c25
commit 4eec5542ce
5 changed files with 12 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
tools/server/server-context.cpp
View File

@@ -806,7 +806,8 @@ private:
auto cparams = common_context_params_to_llama(params_dft);
ctx_dft.reset(llama_init_from_model(model_dft.get(), cparams));
params_base.speculative.draft.ctx = ctx_dft.get();
params_base.speculative.draft.ctx_tgt = ctx;
params_base.speculative.draft.ctx_dft = ctx_dft.get();
params_base.speculative.draft.use_ckpt = common_context_can_seq_rm(ctx_dft.get()) == COMMON_CONTEXT_SEQ_RM_TYPE_FULL;
}
@@ -911,7 +912,7 @@ private:
// try speculative decoding
if (ctx_seq_rm_type != COMMON_CONTEXT_SEQ_RM_TYPE_NO) {
try {
slot.spec.reset(common_speculative_init(params_base.speculative, slot.ctx));
slot.spec.reset(common_speculative_init(params_base.speculative));
} catch (const std::exception & e) {
SRV_ERR("failed to initialize speculative decoding context: %s\n", e.what());
}