torch.load()希望能够增加map_location

[PolarisRisingWar]

• [x] I checked to make sure that this is not a duplicate issue
• [x] I'm submitting the request to the correct repository (for model requests, see here)

Is your feature request related to a problem? Please describe.

我遇到的问题是在textgen/gpt/gpt_model.py第359行，因为我保存的设备（双卡）和加载的设备（单卡）不同，我改成adapters_weights = torch.load(checkpoint_name,map_location="cpu")后能够正常运行

Describe the solution you'd like

我的建议是开放一个可以输入设备的接口，或者直接就放CPU上算了

Describe alternatives you've considered

A clear and concise description of any alternative solutions or features you've considered.

Additional Information

Other things you want the developers to know.

[shibing624]

为啥不用 CUDA_VISIBLE_DEVICES=0， cpu就写CUDA_VISIBLE_DEVICES=-1

[PolarisRisingWar]

不是，比如说我之前保存到卡1上了，我现在设置CUDA_VISIBLE_DEVICES=0了，如果不设置map_location的话，它还是会默认调用到卡1上，但是我现在就没有卡1了，就会报错

[shibing624]

get, fixed.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.(由于长期不活动，机器人自动关闭此问题，如果需要欢迎提问)