Hitonabi/mission-control-v2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(perf): Optimize KV-Cache to Q4_K and fix Hermes Context bounds

Browse Source
This commit is contained in:
Hitonabi
2026-07-07 17:11:50 +02:00
parent 97be0f1d00
commit edf1614e66
2 changed files with 14 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
deploy/llama-swap.config.yaml
+1 -1
View File
@@ -8,7 +8,7 @@ models:
# nicht mehr); KV Q8_0 macht das speicherneutral zu vorher (65k f16). Bench 2026-07-02:
# Q8_0 kostet 0 t/s, MTP n-max 3 = +26% vs. ohne Spec.
cmd: |
llama-server -m /srv/models/Qwen3.6-35B-A3B-MTP-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf --host 127.0.0.1 --port ${PORT} -c 131072 -ngl 999 -fa on --no-mmap --jinja --parallel 2 -cram 16384 -ctk q8_0 -ctv q8_0 --spec-type draft-mtp --spec-draft-n-max 3
llama-server -m /srv/models/Qwen3.6-35B-A3B-MTP-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf --host 127.0.0.1 --port ${PORT} -c 131072 -ngl 999 -fa on --no-mmap --jinja --parallel 2 -cram 16384 -ctk q4_k -ctv q4_k --spec-type draft-mtp --spec-draft-n-max 3
ttl: 0
aliases:
- hermes