rblundon/homelab
0
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore: apply all changes

Browse Source
This commit is contained in:
Hermes Agent service account committed 2026-09-01 12:28:16 -05:00
1 parent e9924a2524
commit 266b6c7be1
7 files changed
+157 -17
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

ansible/roles/llm-inference-multimodel/defaults/main.yml
+1 -1
View File
@@ -232,7 +232,7 @@ llm_swapmode_models:
gguf_path: "{{ llm_models_dir }}/Qwen3.8-27B-Q4_K_M.gguf"
port: 8105
n_gpu_layers: -1 # -1 = auto-detect / all layers to GPU
ctx_size: 65536
ctx_size: 131072 # t_ca3ff615: 65536 -> 131072 (VRAM pre-validated ~20.3GB < 24GB, t_441470b9)
batch_size: 4096
ubatch_size: 4096
parallel: 1