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

llm, add config for Continue (a Codium plugin), fix size_mb not actually getting calculated, change log files to be per-run by making the timestamp very specific, further adjust the ollama environment variables.

Browse Source
This commit is contained in:
me
2026-10-09 14:52:27 -07:00
parent 76702e6433
commit 731a367e6f
1 changed files with 76 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tasks/workstation/linux/software/llm.yml
+76 -37
View File
@@ -32,10 +32,10 @@
zeroclaw: zeroclaw zeroclaw: zeroclaw
zerocode: zerocode zerocode: zerocode
qwen_code: qwen-code qwen_code: qwen-code
# Good for preventing swap usage. # Good to keep low to prevent swap usage.
llm_concurrency: 1 llm_concurrency: "{{ ansible_memtotal_mb / 64000 | round(0, method='ceil') | int }}"
# Allow having multiple chats open without them just waiting each other. # Allow having multiple chats open without them just waiting each other.
llm_parallel: "{{ ansible_processor_vcpus / 4 | round(0, method='ceil') | int }}" llm_parallel: "{{ ansible_processor_vcpus / 8 | round(0, method='ceil') | int }}"
# Most LLMs take a coefficient of about 1.2 to 1.5 of their file size in # Most LLMs take a coefficient of about 1.2 to 1.5 of their file size in
# RAM, which is usually less than guessing a usage of weight*1000. # RAM, which is usually less than guessing a usage of weight*1000.
# Example: # Example:
@@ -71,31 +71,31 @@
ansible.builtin.set_fact: ansible.builtin.set_fact:
llm_installs: [] llm_installs: []
llm_removals: llm_removals:
- { "model": "qwen3:4b", "provider": "qwen_30_0040" } - { "model": "qwen3:4b", "provider": "qwen_30_0040", "size_mb": 4000 }
- { "model": "qwen3.6:27b", "provider": "qwen_36_0270" } - { "model": "qwen3.6:27b", "provider": "qwen_36_0270", "size_mb": 27000 }
- { "model": "qwen3.5:27b", "provider": "qwen_35_0270", "size_mb": llm_size_coef * 17000 } - { "model": "qwen3.5:27b", "provider": "qwen_35_0270", "size_mb": "{{ llm_size_coef * 17000 }}" }
llm_choices: llm_choices:
### QWEN >3.5 - Need to try! ### QWEN >3.5 - Need to try!
- { "model": "qwen3.8:27b", "provider": "qwen_38_0270", "size_mb": 27000 } - { "model": "qwen3.8:27b", "provider": "qwen_38_0270", "size_mb": 27000 }
- { "model": "qwen3.6:35b", "provider": "qwen_36_0350", "size_mb": 35000 } - { "model": "qwen3.6:35b", "provider": "qwen_36_0350", "size_mb": 35000 }
### QWEN 3.5 - Good for generic usage and a little coding. ### QWEN 3.5 - Good for generic usage and a little coding.
- { "model": "qwen3.5:0.8b", "provider": "qwen_35_0008", "size_mb": llm_size_coef * 1200 } - { "model": "qwen3.5:0.8b", "provider": "qwen_35_0008", "size_mb": "{{ llm_size_coef * 1200}}" }
- { "model": "qwen3.5:2b", "provider": "qwen_35_0020", "size_mb": llm_size_coef * 2700 } - { "model": "qwen3.5:2b", "provider": "qwen_35_0020", "size_mb": "{{ llm_size_coef * 2700}}" }
- { "model": "qwen3.5:4b", "provider": "qwen_35_0040", "size_mb": llm_size_coef * 3300 } - { "model": "qwen3.5:4b", "provider": "qwen_35_0040", "size_mb": "{{ llm_size_coef * 3300}}" }
- { "model": "qwen3.5:9b", "provider": "qwen_35_0090", "size_mb": llm_size_coef * 6600 } - { "model": "qwen3.5:9b", "provider": "qwen_35_0090", "size_mb": "{{ llm_size_coef * 6600}}" }
# skip 27b and use 35b # skip 27b and use 35b
- { "model": "qwen3.5:35b", "provider": "qwen_35_0350", "size_mb": llm_size_coef * 22000 } - { "model": "qwen3.5:35b", "provider": "qwen_35_0350", "size_mb": "{{ llm_size_coef * 22000 }}" }
- { "model": "qwen3.5:122b", "provider": "qwen_35_1220", "size_mb": llm_size_coef * 81000 } - { "model": "qwen3.5:122b", "provider": "qwen_35_1220", "size_mb": "{{ llm_size_coef * 81000 }}" }
### QWEN 3.0 Coder - Supposed to be better than 2.5 for machines that can handle it. ### QWEN 3.0 Coder - Supposed to be better than 2.5 for machines that can handle it.
- { "model": "qwen3-coder:30b", "provider": "qwen_30_coder_0300", "size_mb": llm_size_coef * 19000 } - { "model": "qwen3-coder:30b", "provider": "qwen_30_coder_0300", "size_mb": "{{ llm_size_coef * 19000 }}" }
- { "model": "qwen3-coder:420b", "provider": "qwen_30_coder_4200", "size_mb": llm_size_coef * 290000 } - { "model": "qwen3-coder:420b", "provider": "qwen_30_coder_4200", "size_mb": "{{ llm_size_coef * 290000 }}" }
### QWEN 2.5 Coder - Ol'reliable! ### QWEN 2.5 Coder - Ol'reliable!
- { "model": "qwen2.5-coder:0.5b", "provider": "qwen_25_coder_0005", "size_mb": llm_size_coef * 398 } - { "model": "qwen2.5-coder:0.5b", "provider": "qwen_25_coder_0005", "size_mb": "{{ llm_size_coef * 398 }}" }
- { "model": "qwen2.5-coder:1.5b", "provider": "qwen_25_coder_0015", "size_mb": llm_size_coef * 986 } - { "model": "qwen2.5-coder:1.5b", "provider": "qwen_25_coder_0015", "size_mb": "{{ llm_size_coef * 986 }}" }
- { "model": "qwen2.5-coder:3b", "provider": "qwen_25_coder_0030", "size_mb": llm_size_coef * 1900 } - { "model": "qwen2.5-coder:3b", "provider": "qwen_25_coder_0030", "size_mb": "{{ llm_size_coef * 1900 }}" }
- { "model": "qwen2.5-coder:7b", "provider": "qwen_25_coder_0070", "size_mb": llm_size_coef * 4700 } - { "model": "qwen2.5-coder:7b", "provider": "qwen_25_coder_0070", "size_mb": "{{ llm_size_coef * 4700 }}" }
- { "model": "qwen2.5-coder:14b", "provider": "qwen_25_coder_0140", "size_mb": llm_size_coef * 9000 } - { "model": "qwen2.5-coder:14b", "provider": "qwen_25_coder_0140", "size_mb": "{{ llm_size_coef * 9000 }}" }
- { "model": "qwen2.5-coder:32b", "provider": "qwen_25_coder_0320", "size_mb": llm_size_coef * 20000 } - { "model": "qwen2.5-coder:32b", "provider": "qwen_25_coder_0320", "size_mb": "{{ llm_size_coef * 20000 }}" }
### Mistral, Ministral, and Devstral - By a French company? ### Mistral, Ministral, and Devstral - By a French company?
- { "model": "devstral-small-2:24b", "provider": "devstral_20_0240_small", "size_mb": 24000 } - { "model": "devstral-small-2:24b", "provider": "devstral_20_0240_small", "size_mb": 24000 }
#- mistral:7b #- mistral:7b
@@ -118,6 +118,18 @@
#- deepseek-r1:1.5b #- deepseek-r1:1.5b
#- deepseek-r1:7b #- deepseek-r1:7b
- name: Workstation | Linux | Software | LLMs | Facts | Merge | Print
ansible.builtin.debug:
var: "{{ item }}"
loop:
- ansible_processor_vcpus
- ansible_memtotal_mb
- llm_concurrency
- llm_parallel
- llm_installs
- llm_removals
- llm_choices
- name: Workstation | Linux | Software | LLMs | Facts | Choices | Sort - name: Workstation | Linux | Software | LLMs | Facts | Choices | Sort
ansible.builtin.set_fact: ansible.builtin.set_fact:
llm_installs: >- llm_installs: >-
@@ -129,14 +141,10 @@
ansible.builtin.debug: ansible.builtin.debug:
var: "{{ item }}" var: "{{ item }}"
loop: loop:
- ansible_processor_vcpus
- ansible_memtotal_mb
- llm_concurrency
- llm_parallel
- llm_installs - llm_installs
- llm_removals - llm_removals
- name: Workstation | Linux | Software | LLMs | Install - name: Workstation | Linux | Software | LLMs | Packages | Install
ansible.builtin.package: ansible.builtin.package:
name: name:
- "{{ zeroclaw }}" - "{{ zeroclaw }}"
@@ -171,6 +179,7 @@
- name: Workstation | Linux | Software | LLMs | runit | ollama | Script | Run - name: Workstation | Linux | Software | LLMs | runit | ollama | Script | Run
ansible.builtin.copy: ansible.builtin.copy:
dest: "{{ ollama_runit_script }}"
content: | content: |
#!/bin/sh -e #!/bin/sh -e
# Allow pulling LLMs for use with ZeroClaw. # Allow pulling LLMs for use with ZeroClaw.
@@ -186,23 +195,23 @@
export OLLAMA_CONTEXT_LENGTH={{ llm_context }} # default is 4096, https://docs.ollama.com/faq#how-can-i-specify-the-context-window-size export OLLAMA_CONTEXT_LENGTH={{ llm_context }} # default is 4096, https://docs.ollama.com/faq#how-can-i-specify-the-context-window-size
export OLLAMA_MAX_QUEUE=999 # default is 512 (unconfirmed) export OLLAMA_MAX_QUEUE=999 # default is 512 (unconfirmed)
exec chpst -u {{ user }} ollama serve exec chpst -u {{ user }} ollama serve
dest: "{{ ollama_runit_script }}"
mode: "0755" mode: "0755"
- name: Workstation | Linux | Software | LLMs | runit | ollama | Script | Log - name: Workstation | Linux | Software | LLMs | runit | ollama | Script | Log
ansible.builtin.copy: ansible.builtin.copy:
dest: "{{ ollama_runit_log_script }}"
content: | content: |
#!/bin/sh -e #!/bin/sh -e
exec svlogd -tt {{ ollama_runit_log_dest }} exec svlogd -tt {{ ollama_runit_log_dest }}
dest: "{{ ollama_runit_log_script }}"
mode: "0755" mode: "0755"
- name: Workstation | Linux | Software | LLMs | runit | ollama | Script | Log Config - name: Workstation | Linux | Software | LLMs | runit | ollama | Script | Log Config
ansible.builtin.copy: ansible.builtin.copy:
content: |
s10000000 # rotate at 10 MB
n5 # keep 5 old files
dest: "{{ ollama_runit_log_config }}" dest: "{{ ollama_runit_log_config }}"
content: |
s10000000 # rotate at 10 MB
n5 # keep 5 old files
dateformat %Y-%m-%d_%H-%M-%S # each run gets its own file.
owner: "{{ user }}" owner: "{{ user }}"
mode: "0755" mode: "0755"
@@ -229,6 +238,7 @@
- name: Workstation | Linux | Software | LLMs | runit | zeroclaw | Script | File - name: Workstation | Linux | Software | LLMs | runit | zeroclaw | Script | File
ansible.builtin.copy: ansible.builtin.copy:
dest: "{{ zeroclaw_runit_script }}"
content: | content: |
#!/bin/sh -e #!/bin/sh -e
# Allow using the ZeroClaw frontend. # Allow using the ZeroClaw frontend.
@@ -236,28 +246,28 @@
export HOME={{ user_user.home }} export HOME={{ user_user.home }}
export SHELL={{ bash_exec.stdout }} export SHELL={{ bash_exec.stdout }}
exec chpst -u {{ user }} zeroclaw daemon -v --log-level info exec chpst -u {{ user }} zeroclaw daemon -v --log-level info
dest: "{{ zeroclaw_runit_script }}"
mode: "0755" mode: "0755"
- name: Workstation | Linux | Software | LLMs | runit | zeroclaw | Script | Log - name: Workstation | Linux | Software | LLMs | runit | zeroclaw | Script | Log
ansible.builtin.copy: ansible.builtin.copy:
dest: "{{ zeroclaw_runit_log_script }}"
content: | content: |
#!/bin/sh -e #!/bin/sh -e
exec svlogd -tt {{ zeroclaw_runit_log_dest }} exec svlogd -tt {{ zeroclaw_runit_log_dest }}
dest: "{{ zeroclaw_runit_log_script }}"
mode: "0755" mode: "0755"
- name: Workstation | Linux | Software | LLMs | runit | zeroclaw | Script | Log Config - name: Workstation | Linux | Software | LLMs | runit | zeroclaw | Script | Log Config
ansible.builtin.copy: ansible.builtin.copy:
content: |
s1000000 # rotate at 1 MB
n5 # keep 5 old files
dest: "{{ zeroclaw_runit_log_config }}" dest: "{{ zeroclaw_runit_log_config }}"
content: |
s1000000 # rotate at 1 MB
n5 # keep 5 old files
dateformat %Y-%m-%d_%H-%M-%S # each run gets its own file.
mode: "0755" mode: "0755"
when: ansible_service_mgr == "runit" when: ansible_service_mgr == "runit"
- name: Workstation | Linux | Software | LLMs | Enable - name: Workstation | Linux | Software | LLMs | Services | Enable
ansible.builtin.service: ansible.builtin.service:
name: "{{ item }}" name: "{{ item }}"
state: started state: started
@@ -266,7 +276,7 @@
- "{{ ollama }}" - "{{ ollama }}"
- "{{ zeroclaw }}" - "{{ zeroclaw }}"
- name: Workstation | Linux | Software | LLMs | Verify - name: Workstation | Linux | Software | LLMs | Services | Verify
ansible.builtin.uri: ansible.builtin.uri:
url: "{{ item }}" url: "{{ item }}"
status_code: 200 status_code: 200
@@ -372,4 +382,33 @@
create: false create: false
backup: true backup: true
- name: Workstation | Linux | Software | LLMs | Models | Continue | Config
ansible.builtin.copy:
dest: "{{ user_user.home }}/.continue/config.yml"
content: |
# Maintained by `env-ansible/tasks/workstation/linux/software/llm.yml`
name: Main Config
version: 1.0.0
schema: v1
models:
mode: "0644"
- name: Workstation | Linux | Software | LLMs | Models | Continue | Models
ansible.builtin.blockinfile:
path: "{{ user_user.home }}/.continue/config.yml"
marker: "# {mark} ANSIBLE MANAGED BLOCK | llm.yml | Continue"
insertafter: EOF
block: |
{% for llm in llm_installs %}
- name: llm.provider
provider: ollama
model: llm.model
defaultCompletionOptions:
contextLength: {{ llm_context }}
maxTokens: {{ llm_context }}
{% endfor %}
state: present
create: false
backup: true
when: ansible_pkg_mgr == "pacman" when: ansible_pkg_mgr == "pacman"