Actually i wanted to do quantization for my custom trained yolo model. I have already trained my model. Since I am deploying this to nvidia jetson nano therefore, after testing the trained model I felt a huge latency and optimization problems. Now I want to quantize my trained model. If there anyone who can help me guiding how can I quantize my post trained model it would be great. I have tried to convert into onnx but whenever I exported in onnx I lose my class names. Actually I want to know after I quantize or prune the model how could I run inside a video. Thanks in advanced
Most importantly I am using an OBB model.