怎们拿到图片对应的tokens而不是embeding?

#11
by duanyu - opened

怎们拿到图片对应的tokens而不是embeding?

OpenGVLab org

outputs = model(pixel_values)

这个输出的outputs是个字典,里面有tokens也有embedding

czczup changed discussion status to closed

Sign up or log in to comment