Benchmark for testing SVTR -LCNET

[amandhapola]

Hi everyone,

I'm looking into the PP-OCRv3 paper link and have a question about SVTR-LCNET's benchmarking.

The paper details the training and validation dataset, but I'm curious if SVTR-LCNET was also evaluated on standard OCR datasets like IIIT5K, IC15, or similar benchmarks.

Knowing its performance on these common datasets would be really helpful for comparison with other models. Also, if any other benchmarking datasets were used, I'd appreciate that info.

Thanks