MiniCPM-V icon indicating copy to clipboard operation
MiniCPM-V copied to clipboard

Can u opensoure the Chinese ocr part data?

Open lucasjinreal opened this issue 1 year ago • 3 comments

Can u opensoure the Chinese ocr part data?

lucasjinreal avatar May 20 '24 14:05 lucasjinreal

Our Chinese OCR data is mainly collected from open-source data, later this week, the technical report will be released~

Cuiunbo avatar May 23 '24 11:05 Cuiunbo

Hi, I might still wondering, the whole training process, does both pretrain and sft have opened vit?

How does the Chinese ocr data used for each period of training?

lucasjinreal avatar May 23 '24 13:05 lucasjinreal

The details can be found in the technical report, which will be released this week~ If you still have any questions at that point, just let us know.

Cuiunbo avatar May 23 '24 13:05 Cuiunbo

where the report

xxlxms avatar Jul 27 '24 14:07 xxlxms