AgentTuning icon indicating copy to clipboard operation
AgentTuning copied to clipboard

AgentTuning: Enabling Generalized Agent Abilities for LLMs

Results 17 AgentTuning issues
Sort by recently updated
recently updated
newest added

查看huggingface dataset上ALFWorld和Mind2Web的训练数据,发现根据提供的指令,模型不可能产生预期的行为,比如下面两条数据,这个是符合预期的吗?

你好,我下了魔塔上的 AgentInstruct 数据集,但 conversation 都是空值,请问是数据不开源了嘛?

用fastchat部署AgentLM-13B,推理的时候格式是乱的,尤其是streaming的模式,每行只有几个字符就切换到下一行了,一个单词被切成了好几个字母或字母组合。如果手动用transformer加载并用gradio展示的话就没有这个问题,用fastchat的debug模式看了一下,用的是LlamaForCausalLM加载的模型,应该没错

如题,我按照规定要求设置的环境,但是跑不起来,有更清晰的环境设置么?

想问一下,通用数据ShareGPT_Vicuna_unfiltered有9w条,你们是如何筛选到5w条的?能提供一下脚本吗

Thanks for open-sourced agentTuning code , I am quite interested in training the model, i see the training framework is not open-sourced https://github.com/THUDM/AgentTuning/issues/1, The discussion mentioned that it could support...

Hey, thank you for your great work. I just wanted to know how can I run evaluation on the open source AgentInstruct data on the AgentBench repo. I will be...

https://huggingface.co/THUDM/agentlm-7b , I try it,but far below 84% in alfworld-std. Is it the wrong model?