Carlo

> > > 很不错，没什么明显问题 欢迎更新后续的评估结果以及详细的训练设定 如有需要，后面做统一的分析回复 > > > > > > 作者大大，dpo_train跑出来的代码只有safetensors格式的文件，你是将其转换成了pth吗 > > 是的，我后面把权重`torch.save`保存成了pth格式，为了和前面模型保持一致 Traceback (most recent call last): File "/data/app/minimind/4-lora_sft.py", line 188, in train_epoch(epoch, wandb) File "/data/app/minimind/4-lora_sft.py",...

> 我看到类似的错误，但训练仍在继续。我不知道这是不是误报。 me too

> Also what happens when you use `--no-content no`? > > `memgpt run --model gpt-4o --no-content no` > > Does that work? What does' no content no 'do

> Is it possible that the OpenAI proxy you're using `https://ai-gateway-test.bondee-inc.com/v1` doesn't support when message content is `None` but OpenAI does? requests data {'model': 'gpt-4o', 'messages': [{'content': 'You are MemGPT,...

> Is it possible that the OpenAI proxy you're using `https://ai-gateway-test.bondee-inc.com/v1` doesn't support when message content is `None` but OpenAI does? messages.[5] { "role": "assistant" }

> Is it possible that the OpenAI proxy you're using `https://ai-gateway-test.bondee-inc.com/v1` doesn't support when message content is `None` but OpenAI does?  

 

@cpacker @sarahwooders

> torch 安装成了 CPU 版本 我也遇到这个问题了

解决了,安装环境的时候使用的是无卡模式,torch 安装成了 CPU 版本了