KDD2018
KDD2018
I am interested in Kosmos-2 and appreciate it. I want finetune kosmos-2 on my dataset which included images and Chinese texts, but I couldn't do it successfully. So I tried...
请教下,有大佬尝试过使用两张RTX3090微调CogVLM嘛?有可能性嘛?
deepspeed启用zero3时,在多个迭代训练之后出现RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:1 and cpu! zero3参数配置如下: ` "zero_optimization": { "stage": 3, "offload_optimizer": { "device": "cpu", "pin_memory":...