qinfendekaizhou
qinfendekaizhou
上面不是有现成的环境吗？直接写上路径编译啊
dic = {i: [] for i in range(150)}
没啥错啊，v7和v8我都部署上了
> 希望尽快有大佬的好消息。 > > | | > 伐木累 > | > | > ***@***.*** > | > ---- Replied Message ---- > | From | Xiaoyan ***@***.***> | > |...
> I have the same qustion import torch # 加载原始权重 sam_weight = torch.load('weights/sam_vit_h_4b8939.pth') key_word = 'prompt_encoder' # 或 'mask_decoder' new_weight = {} prefix = f"{key_word}." for key in sam_weight.keys(): #...