todaydeath
todaydeath
/root/miniconda3/lib/python3.7/site-packages/gradio/ 路径下没有 frpc_linux_amd64_v0.2
> > 先对模型进行合并，就可以推理了 > > 谢谢 昨天在issues看到了合并的方法 就可以用了 合并后的模型和基础模型对比，效果明显么
> 问的太对了哥，给我愁死了，这readme写的八成是写错了，那个msg应该是model 写成model会报错，就用msg就好了，不用管返回值。但是在运行的时候，显存占多少啊？我这里24G显存不够用，多显卡运行又报错
> 你好，我已经解决啦，谢谢 请问你是怎么解决的呢？最终占用了多少显存？
https://github.com/OpenBMB/MiniCPM-V/tree/main/finetune#lora-finetuning 这个多卡运行不了，报错： Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cuda:6! (when checking argument for argument src in method wrapper_CUDA_scatter__src)
https://github.com/OpenBMB/MiniCPM-V/tree/main/finetune#lora-finetuning 单卡显存又不够
> > https://github.com/OpenBMB/MiniCPM-V/tree/main/finetune#lora-finetuning 这个多卡运行不了，报错： Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cuda:6! (when checking argument for argument src in method...
> > > > https://github.com/OpenBMB/MiniCPM-V/tree/main/finetune#lora-finetuning 这个多卡运行不了，报错： Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cuda:6! (when checking argument for argument src...
> > > > > > https://github.com/OpenBMB/MiniCPM-V/tree/main/finetune#lora-finetuning 这个多卡运行不了，报错： Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cuda:6! (when checking argument for...
> > > > > > > https://github.com/OpenBMB/MiniCPM-V/tree/main/finetune#lora-finetuning 这个多卡运行不了，报错： Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cuda:6! (when checking argument...