Public/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-20 01:31:31 +02:00
Code Issues Packages Projects Releases Wiki Activity

common : enable default speculative config

Browse Source 
Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp
This commit is contained in:
Georgi Gerganov
2026-09-04 10:59:53 +03:00
parent d646c9d155
commit 7c2e0dea43
10 changed files with 196 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/speculative.md
+5 -3
View File
@@ -226,10 +226,12 @@ Use exactly one of these options:
```
--spec-type [none|draft-simple|draft-eagle3|draft-dflash|draft-dspark|draft-mtp|ngram-cache|ngram-simple|ngram-map-k|ngram-map-k4v|ngram-mod]
comma-separated list of types of speculative decoding to use
(default: none)
(default: server/CLI default config, none for other tools)
(env: LLAMA_ARG_SPEC_TYPE)
--spec-default use default speculative decoding config
(enables ngram-mod)
--no-spec-type [none|draft-simple|draft-eagle3|draft-dflash|draft-dspark|draft-mtp|ngram-cache|ngram-simple|ngram-map-k|ngram-map-k4v|ngram-mod]
comma-separated list of types of speculative decoding to remove
(env: LLAMA_ARG_NO_SPEC_TYPE)
--spec-default DEPRECATED: default speculative decoding is enabled by default
```
### Draft Model Parameters