ansible/playbook_podman_qwen3.6.yml

- name: Setup Qwen3.6 - Coder
  hosts: qwen3_6_host
  vars:
    podman_user: "{{ qwen3_Z6_user | default('qwen36') }}"
    podman_group: "{{ qwen3_6_group | default('qwen36') }}"
    podman_extra_groups: "users,video"
    podman_user_home: "{{ qwen3_6_home | default('/home/qwen36') }}"
    llama_model: "{{ qwen3_6_model | default('bartowski/Qwen_Qwen3.6-35B-A3B-GGUF:IQ4_XS') }}"
    __qwen3_6_context_length__: "{{ qwen3_6_context_length | default(262144) }}"
    podman_user_folders:
      - models
      - .cache
      - .cache/llama.cpp
    pods:
      - name: qwen3.6
        repo:
          image: ghcr.io/ggml-org/llama.cpp:server-vulkan
        ports:
          - "0.0.0.0:{{ qwen3_6_port | default(8012) }}:8080/tcp"
        command:
          - "-t"
          - "12"
          - "-hf"
          - "{{ llama_model }}"
          - "-ngl"
          - "999"
          - "-c"
          - "{{ __qwen3_6_context_length__ }}"
          - "-np"
          - "1"
          - "-b"
          - "1024"
          - "-ub"
          - "512"
          - "-fa"
          - "on"
          - "-ctk"
          - "q4_0"
          - "-ctv"
          - "q8_0"
          - "-cram"
          - "{{ __qwen3_6_context_length__ }}"
          - "-sm"
          - "layer"
          - "-dev"
          - "Vulkan1,Vulkan2"
          - "-ts"
          - "8,12"
          - "-fit"
          - "off"
          - "-mg"
          - 1
          - "--mmap"
        volumes:
          - "{{ podman_user_home }}/models:/models:Z"
          - "{{ podman_user_home }}/.cache:/app/.cache:Z"
        device:
          - "nvidia.com/gpu=all"
          - "/dev/kfd:/dev/kfd:rw"
          - "/dev/dri:/dev/dri:rw"
  tasks:
    - name: Setup Pods
      ansible.builtin.include_tasks:
        file: tasks/podman.yml
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- name: Setup Qwen3.6 - Coder`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`hosts: qwen3_6_host`
			`vars:`
			`podman_user: "{{ qwen3_Z6_user \| default('qwen36') }}"`
			`podman_group: "{{ qwen3_6_group \| default('qwen36') }}"`
			`podman_extra_groups: "users,video"`
			`podman_user_home: "{{ qwen3_6_home \| default('/home/qwen36') }}"`
			`llama_model: "{{ qwen3_6_model \| default('bartowski/Qwen_Qwen3.6-35B-A3B-GGUF:IQ4_XS') }}"`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`__qwen3_6_context_length__: "{{ qwen3_6_context_length \| default(262144) }}"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`podman_user_folders:`
			`- models`
			`- .cache`
			`- .cache/llama.cpp`
			`pods:`
			`- name: qwen3.6`
			`repo:`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`image: ghcr.io/ggml-org/llama.cpp:server-vulkan`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`ports:`
			`- "0.0.0.0:{{ qwen3_6_port \| default(8012) }}:8080/tcp"`
			`command:`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- "-t"`
			`- "12"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`- "-hf"`
			`- "{{ llama_model }}"`
			`- "-ngl"`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- "999"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`- "-c"`
			`- "{{ __qwen3_6_context_length__ }}"`
			`- "-np"`
			`- "1"`
			`- "-b"`
			`- "1024"`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- "-ub"`
			`- "512"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`- "-fa"`
			`- "on"`
			`- "-ctk"`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- "q4_0"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`- "-ctv"`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- "q8_0"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`- "-cram"`
			`- "{{ __qwen3_6_context_length__ }}"`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- "-sm"`
			`- "layer"`
			`- "-dev"`
			`- "Vulkan1,Vulkan2"`
			`- "-ts"`
			`- "8,12"`
			`- "-fit"`
			`- "off"`
			`- "-mg"`
			`- 1`
			`- "--mmap"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`volumes:`
			`- "{{ podman_user_home }}/models:/models:Z"`
			`- "{{ podman_user_home }}/.cache:/app/.cache:Z"`
			`device:`
			`- "nvidia.com/gpu=all"`
Add GPU monitoring services and configurations 2026-05-29 23:42:14 +02:00			`- "/dev/kfd:/dev/kfd:rw"`
			`- "/dev/dri:/dev/dri:rw"`
Add new model files and update Ansible playbooks for Qwen3.5 and Qwen3.6 2026-05-27 23:53:00 +02:00			`tasks:`
			`- name: Setup Pods`
			`ansible.builtin.include_tasks:`
			`file: tasks/podman.yml`