Parameter configuration issues of the Nanonets-OCR-s model

[IceBear-huang]

Hello, I'd like to ask. For the model's parameter --gpu-memory-utilization=0.5, setting it to 0.5 or 0.9 doesn't change the model's recognition speed. It seems to only be a memory allocation command and can't improve the model's recognition speed. So what configurations should I make to speed up the model's recognition? Currently, the speed is a bit slow for me—it takes about 27 seconds for content with more information and 9 seconds for content with less. My graphics card is a 3090 Ti.

[mandalsouvik3333]

--gpu-memory-utilization=0.5 Large GPU allocation can be used to do batched processing in parallel. If you are doing 1-2 files at a time, you should reduce this.

speed up the model's recognition

Currently, we resize the image to 2048x2048 by default. You can reduce the image size, which should speed up the entire process. Additionally, you can send multiple requests at once to decrease overall processing time.

docext/docext/app/args.py

Line 72 in 668eee6

"--max_img_size",

[IceBear-huang]

Hello, why is it that some images are not accurately recognized, but when I re-screenshot them and make the screenshot larger, they can be recognized correctly? What's the reason for this? Could it be that image preprocessing is needed to improve accuracy? Do you have any good suggestions for image preprocessing?