THUDM/AgentTuning Issues
训练数据中指令与模型行为不匹配
Updated本地模型
Closed请问哪里可以找到工作里对于数据库方面的训练数据
Closed 1weight decay确定是0.1吗？
Closed 1基于fastchat部署，推理异常
Updated 3貌似hotpotqa测试脚本跑不起来？
Updated 1训练数据是如何采样的？
Closed 3通用数据如何筛选
Updated 7可以给个简单点的工具调用示例吗
Updated 1期待用 Qwen72B 训练的模型。
Closed 1除了用docker运行，还有其他方式可以运行AgentLM吗？
Closed 6关于TRAJECTORY FILTERING问题
Closed 3请问下agentlm-7b最少需要多少显存可以推理
Closed 5论文中关于损失函数E的问题
Closed 7Finetuning with Mistral or Yi?
Closed 1关于数据集
ClosedNumber of training steps
Closed 1Dataset details 中找不到reward的计算方式
Closed 5关于dataset statics 和 download
Closed 3关于reward
Closed 2agent tuning和toolbench的区别
Closed 1微调显存
Closed 1Start TGI worker
Closed 1请教reward分数的各种情况
Closed 1Inference with `vllm`
Closed 1论文中Table 2中的数字的含义和计算方式
Closed 2Add license
ClosedGrammer mistake in readme
Closed 1Auto comment
Closed 1什么时候上魔塔社区
Closed 1论文中的问题
Closed 1Model Output Length
Closed 2是否可以不在docker里运行
Closed 2底座模型基于llama2，是否支持中文呢
Closed 1fine-tune code
Closed 1微调
Closed 2交互轨迹的Reward如何得到
Closed 1AgentLM能支持openai.api类的接口本地部署吗？
Closed 1train code
Closed 3