15 lines
262 B
Plaintext
15 lines
262 B
Plaintext
# Inference server configuration
|
|
HOST=0.0.0.0
|
|
PORT=8000
|
|
|
|
# VRAM management
|
|
VRAM_BUDGET_MB=10240
|
|
STRATEGY=sequential # sequential | concurrent | auto
|
|
|
|
# Model defaults
|
|
YOLO_MODEL=yolov8n.pt
|
|
YOLO_CONFIDENCE=0.3
|
|
|
|
# Device
|
|
DEVICE=auto # auto | cpu | cuda | cuda:0
|