the input_size of the exported model is fixed?

[wsy-yjys]

问题确认 Search before asking

• [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

when I export inference model with following code， and just use Paddle Inference without TensorRT to test speed, the input_size is fixed? can I use letterbox resize and pad the img which means the input_size is unfixed , in ordder to consistent with other algorithms. Thankyou~

# export inference model
python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams exclude_nms=True

# speed testing with run_benchmark=True
CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_file=demo/000000014439_640x640.jpg --run_mode=paddle --device=gpu --run_benchmark=True

[wangxinxin08]

the input size of PP-YOLO series is fixed and do not use letterbox resize and pad the img which is not matched with our training process.

[wsy-yjys]

If I want to get the fps in the letterbox method， what should I do？Thankyou~