Giters

OS-Copilot / OS-Copilot

An self-improving embodied conversational agent seamlessly integrated into the operating system to automate our daily tasks.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

[BUG]: The task list obtained after planning is:[]

RufengSama opened this issue · comments

commented

Describe the bug
The task list obtained after planning is:[], the list is blank and then the system exit.

the log about the issue is:
[QznaSOEze7wIgYaK] 2024-05-16 20:27:06,645 - INFO - Your task is: Move the text files containing the word 'agent' from the folder named 'document' to the path 'working_dir/agent'
[QznaSOEze7wIgYaK] 2024-05-16 20:27:07,352 - INFO - Anonymized telemetry enabled. See https://docs.trychroma.com/telemetry for more information.
[QznaSOEze7wIgYaK] 2024-05-16 20:27:07,472 - ERROR - Error in planning on attempt 1: 'FridayPlanner' object has no attribute 'llm'
[QznaSOEze7wIgYaK] 2024-05-16 20:27:07,473 - ERROR - Error in planning on attempt 2: 'FridayPlanner' object has no attribute 'llm'
[QznaSOEze7wIgYaK] 2024-05-16 20:27:07,473 - ERROR - Error in planning on attempt 3: 'FridayPlanner' object has no attribute 'llm'
[QznaSOEze7wIgYaK] 2024-05-16 20:27:07,473 - ERROR - Max retries reached, planning failed.

thanks for your helping~

commented

Could you please give me more information in your log/temp0325.log?

commented

That is the complete log information everytime I run it.
image

MODEL_NAME="gpt-3.5"
maybe the problem is that I have to use GPT-4? I guess.

commented

Can you shown me you .env file? I guess something went wrong on it. You just need to enable the following three options in the .env file.

image

By the way, you can try to use "gpt-4o-2024-05-13", which is faster and cheaper.

commented

i have a .env file and a .env_template file.
image
image

commented

i have a .env file and a .env_template file. image image

Thank you for providing the information. Have you tried the method I mentioned? If you need a proxy, please uncomment the API_BASE_URL. And your model maybe inappropriate, please check https://openai.com/api/pricing/ for more message.

commented

Hi, I have tried your method these days, but there are still some problems.
Here is the log info and my .env file.
image
image
the API_BASE_URL i tried https://api.cpdd666.cn/v1 and that in picture.
I tried gpt-4 and gpt-3.5-turbo-0125, they both work well in the test web but wrong in OS-Copilot.

commented

Hi, I have tried your method these days, but there are still some problems. Here is the log info and my .env file. image image the API_BASE_URL i tried https://api.cpdd666.cn/v1 and that in picture. I tried gpt-4 and gpt-3.5-turbo-0125, they both work well in the test web but wrong in OS-Copilot.

image

The current issue indicates that the model selection is correct, but there is a problem with the server proxy. If you can access OpenAI without deploying a VPN, please set it up as I have. If a VPN is required, simply uncomment the API_BASE_URL.

commented

Hi, I'm regret to tell you that my OS-Copilot is still error.
I bought a new api-key and test by

curl https://api.xty.app/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer sk-xxx" -d '{ "model": "gpt-4o-2024-05-13", "messages": [{"role": "user", "content": "你好aaaaaa"}], "stream":true, "temperature": 0.7 }'

I can receive data correctly.
image
image
image
image
I tried both using WSL locally and Alibaba Cloud servers, and the situation was the same. help plz[cry cry]

commented

I think maybe something wrong about the key!
I used an openAI official api key and connect correctly but only once.
There are still connection errors after my subsequent attempts, but I can open https://openai.com/ correctly.
How can I solve this problem. Thanks for your listening and helping!

commented

Have you ever tried to set API_BASE_URL="http://127.0.0.1:8079"? I've seen many similar connection issues like yours being resolved in this way.