打印了每一步的信息如下: Starting the program... Parsing arguments... Loading model... Model path: [./THUDM/chatglm-6b](https://file+.vscode-resource.vscode-cdn.net/f%3A/demo/01.chatglm/01.ChatGLM-6b/THUDM/chatglm-6b) CUDA is available. Device count: 1 Current device: 0 Device name: NVIDIA GeForce RTX 4060 Memory allocated: 0.00 MB Memory cached: 0.00 MB Creating quantization config... Starting model loading... 到这里,就停止运行了,也不报错,也不开始训练