v10 YOLO26n re-deploy: weights.onnx

use onnxruntime-gpu + cuda128 torch

fix cls_remap: pump/hose swap for v9 (correctly labeled data)

add max_hourly_price_per_gpu=0.50 cap

v9 pseudo-gt: update weights.onnx (mAP50=0.900, fine-tuned from hot_1)

v9 pseudo-gt: update chute_config.yml (mAP50=0.900, fine-tuned from hot_1)

cap hourly price at $0.50, exclude expensive GPUs

Upload weights.onnx with huggingface_hub

Upload miner.py with huggingface_hub

Upload chute_config.yml with huggingface_hub

Upload class_names.txt with huggingface_hub

initial commit