You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
defdecode_base64_to_image(base64_string) ->Image:
"""Convert raw data into Pillow image."""image_data=base64.b64decode(base64_string)
image=Image.open(io.BytesIO(image_data))
returnimageimage=decode_base64_to_image(image_bytes)
image.save("tmp/%d.jpg"%index)
看了下目前的接口,图片输入都是基于本地或者http请求的,但是实际使用中拿到的图片输入需要crop之后再输入,无法提供crop之后图片的地址,Qwen-VL是否可以直接输入image bytes来进行训练和推理呢
The text was updated successfully, but these errors were encountered: