The scale of the dataset used in the fine-tuning stage

[Davidwhw]

Could you please talk about exactly how much data was used in the ChartVLM’s fine-tuning stage?
In the paper, the source of the data set was explained but the specific amount of data was not mentioned.