me/env-ansible
1
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

llm, use a more distinct number for the context so that ollama ps better gives away how the model is being used.

Browse Source
This commit is contained in:
me
2026-10-10 19:46:20 -07:00
parent 5541c8251c
commit df0df85d9e
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tasks/workstation/linux/software/llm.yml
+1 -1
View File
@@ -63,7 +63,7 @@
# - qwen3.5:4b would be 3.3GB * 1.3 = 4290, rather than 4000. # - qwen3.5:4b would be 3.3GB * 1.3 = 4290, rather than 4000.
llm_size_coef: 1.3 llm_size_coef: 1.3
# Use the MB of RAM as the context, with a maximum to prevent swapping. # Use the MB of RAM as the context, with a maximum to prevent swapping.
llm_context: "{{ [(ansible_memtotal_mb | round(-3, method='ceil')), 32000] | min | int }}" llm_context: "{{ [(ansible_memtotal_mb | round(-3, method='ceil')), 34543] | min | int }}"
- name: Workstation | Linux | Software | LLMs | Facts 02 - name: Workstation | Linux | Software | LLMs | Facts 02
ansible.builtin.set_fact: ansible.builtin.set_fact: