CogAgent icon indicating copy to clipboard operation
CogAgent copied to clipboard

An open-sourced end-to-end VLM-based GUI Agent

Results 30 CogAgent issues
Sort by recently updated
recently updated
newest added

非常nice的一个工作,但是我还有几个问题想咨询一下: 1. SFT 中人工标记的数据集的数据格式 2. Conversion of Agent Datasets 这一步的作用或者目的是什么?

### Feature request / 功能建议 I wonder if there are still up-to-date versions for autoglm phone use and pc use. I can't find the latest version of phone use(android) and...

### Feature request / 功能建议 是否能够进行拖拽或者长按shift+click ### Motivation / 动机 想要制作数据集 ### Your contribution / 您的贡献 https://github.com/THUDM/CogAgent/blob/main/Action_space.md

运行vllm_openai_server.py: ![Image](https://github.com/user-attachments/assets/af4e49ff-0c54-4c04-beec-16bd118404ce) 我的环境: ![Image](https://github.com/user-attachments/assets/49359595-e3c5-4be9-8522-6e9350e4723c)

### System Info / 系統信息 使用vllm代码示例,推理的结果总是Grounded de,是啥原因? ### Who can help? / 谁可以帮助到您? _No response_ ### Information / 问题信息 - [ ] The official example scripts / 官方的示例脚本 - [...

请问是否有使用vllm启动cogagent的详细步骤,我在运行[vllm_openai_server.py]文件时报错:ValueError: ChatGLMForConditionalGeneration has no vLLM implementation and the Transformers implementation is not compatible with vLLM. 我的环境是: 华为npu:910B python3.10.12 vllm0.7.3+empty torch2.5.1 transformers4.50.1

### System Info / 系統信息 cuda==12.4 ### Who can help? / 谁可以帮助到您? @sixsixcoder ### Information / 问题信息 - [x] The official example scripts / 官方的示例脚本 - [x] My own modified...

### System Info / 系統信息 在运行客户端的时候,出现报错: Error code: 404 - {'detail': 'Not Found'},完整报错如下: return self._request( ~~~~~~~~~~~~~^ cast_to=cast_to, ^^^^^^^^^^^^^^^^ ...... retries_taken=retries_taken, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ) ^ File "~/python3.13/site-packages/openai/_base_client.py", line 1064, in _request raise...

我在v100显卡对CogAgent-9B-20241220进行本地部署,单步推理的时间在**8s**左右。请问这个时长是否正常,亦或是我推理方式有问题?

help wanted

Thanks guys for your amazing work! Do you have plan to release evaluation codes on some open-source benchmarks, such as ScreenSpot, OminiAct and OSWorld?