2025-06-03 23:43:19 +02:00
|
|
|
- name: Setup Kokoro FastApi
|
|
|
|
|
hosts: kokoro
|
|
|
|
|
vars:
|
|
|
|
|
podman_user: "{{ kokoro_user | default('kokoro') }}"
|
|
|
|
|
podman_group: "{{ kokoro_group | default('kokoro') }}"
|
2025-06-08 21:18:53 +02:00
|
|
|
podman_extra_groups: "users"
|
2025-06-03 23:43:19 +02:00
|
|
|
podman_user_home: "{{ kokoro_home | default('/home/kokoro') }}"
|
2025-06-08 21:18:53 +02:00
|
|
|
podman_user_pubkey: "{{ kokoro_user_pubkey | default(omit) }}"
|
2025-06-03 23:43:19 +02:00
|
|
|
podman_user_folders:
|
|
|
|
|
- data
|
|
|
|
|
pods:
|
|
|
|
|
- name: kokoro
|
|
|
|
|
repo:
|
2025-12-25 16:37:05 +01:00
|
|
|
image: ghcr.io/remsky/kokoro-fastapi-cpu
|
|
|
|
|
env:
|
|
|
|
|
PYTHONPATH: /app:/app/api
|
|
|
|
|
# ONNX Optimization Settings for vectorized operations
|
|
|
|
|
ONNX_NUM_THREADS: 6 # Maximize core usage for vectorized ops
|
|
|
|
|
ONNX_INTER_OP_THREADS: 4 # Higher inter-op for parallel matrix operations
|
|
|
|
|
ONNX_EXECUTION_MODE: parallel
|
|
|
|
|
ONNX_OPTIMIZATION_LEVEL: all
|
|
|
|
|
ONNX_MEMORY_PATTERN: true
|
|
|
|
|
ONNX_ARENA_EXTEND_STRATEGY: kNextPowerOfTwo
|
2025-06-03 23:43:19 +02:00
|
|
|
ports:
|
|
|
|
|
- "0.0.0.0:{{ kokoro_port | default(8880) }}:8880/tcp"
|
2025-12-25 16:37:05 +01:00
|
|
|
# device:
|
|
|
|
|
# - "nvidia.com/gpu=all"
|
2025-06-03 23:43:19 +02:00
|
|
|
|
2026-05-30 23:24:44 +02:00
|
|
|
roles:
|
|
|
|
|
- podman
|