mac m3运行后报错，感觉是TensorFlow=2.9.0不支持m3，求指导

[shenxingqi]

The TensorFlow library was compiled to use AVX instructions, but these aren't available on your machine.

[junyangwang0410]

TensorFlow只用于OCR，可以自行替换模型

[kx-kexi]

怎么替换模型呀，是把OC R换成其他的吗

[junyangwang0410]

https://github.com/X-PLUG/MobileAgent/blob/main/Mobile-Agent-v2/MobileAgent/text_localization.py

格式是 text -> List ["text1", "text2", ..., "textn"], coordinate -> List [[x1, y1, x2, y2], [x1, y1, x2, y2], ..., [x1, y1, x2, y2]]

[kx-kexi]

https://github.com/X-PLUG/MobileAgent/blob/main/Mobile-Agent-v2/MobileAgent/text_localization.py

格式是 text -> List ["text1", "text2", ..., "textn"], coordinate -> List [[x1, y1, x2, y2], [x1, y1, x2, y2], ..., [x1, y1, x2, y2]]

啥意思呀，没看懂

[junyangwang0410]

https://github.com/X-PLUG/MobileAgent/blob/main/Mobile-Agent-v2/MobileAgent/text_localization.py 格式是 text -> List ["text1", "text2", ..., "textn"], coordinate -> List [[x1, y1, x2, y2], [x1, y1, x2, y2], ..., [x1, y1, x2, y2]]

啥意思呀，没看懂

修改MobileAgent/text_localization.py文件，保持这一行代码的返回格式就可以，text_data的格式是，List，其中每个元素都是一个文本，["text1", "text2", ..., "textn"], coordinate的格式是List，每个元素都是一个4元组，代表对应index下的文本的bbox，用左上角坐标和右下角坐标代替，[[x1, y1, x2, y2], [x1, y1, x2, y2], ..., [x1, y1, x2, y2]]。

你可以修改这个文件里的ocr模型和调用方式，只需要保持返回值不变，就仍然可以运行项目