Issues encountered while running the project on Windows 10 with Python 3.11

[SunilWarrier]

I encountered multiple issues while trying to run your project on Windows 10 with Python 3.11 using Windows CMD.

Missing modules after running the requirements: I was unable to find a compatible version of pywin32, so I installed it manually using the following command:

python -m pip install pywin32
I also had to install the following modules manually:

pip install google
pip install readability-lxml
pip install playsound
pip install python-dotenv
pip install docker

Strange characters in the output: After running the program, I noticed some strange characters like ←[32m and ←[94m in the output. This might be an issue with Windows CMD.

AttributeError after answering the AI goals: After answering the question "Enter up to 5 goals for your AI," I encountered the following error:

File "\Auto-GPT\scripts\main.py", line 204, in
assistant_reply = chat.chat_with_ai(
^^^^^^^^^^^^^^^^^^
File "\Auto-GPT\scripts\chat.py", line 68, in chat_with_ai
response = openai.ChatCompletion.create(
^^^^^^^^^^^^^^^^^^^^^
AttributeError: module 'openai' has no attribute 'ChatCompletion'. Did you mean: 'Completion'?

Please help me understand if I missed anything during the setup process and how I can resolve these issues.

I appreciate your support.

Best regards,
Sunil

[SunilWarrier]

Upgrade to openAI solved these issues
pip install --upgrade openai

Now got a new issue
File \AppData\Local\Programs\Python\Python311\Lib\site-packages\openai\util.py", line 186, in default_api_key
raise openai.error.AuthenticationError(
openai.error.AuthenticationError: No API key provided. You can set your API key in code using 'openai.api_key = ', or you can set the environment variable OPENAI_API_KEY=).

[SunilWarrier]

Again dumb me should have read the readme.md correctly renamed the .env file
Now getting error
openai.error.InvalidRequestError: The model: gpt-4 does not exist
I am a paid user in OpenAI and chatGPT

[SunilWarrier]

Maybe I am not allowed, so I revisited and applied for it. Then I changed it to gpt-3.5turbo
it stated working but stuck with
Error: Invalid JSON
What task would you like me to perform?
NEXT ACTION: COMMAND = Error: ARGUMENTS = Invalid JSON
Enter 'y' to authorise command or 'n' to exit program...
Input:y
-=-=-=-=-=-=-= COMMAND AUTHORISED BY USER -=-=-=-=-=-=-=
SYSTEM: Command Error: returned: Unknown command Error:
Error: Invalid JSON
I am ready for your next command. What would you like me to do?
NEXT ACTION: COMMAND = Error: ARGUMENTS = Invalid JSON
Enter 'y' to authorise the command or 'n' to exit the program...
Input:
I am trying to figure out how to fix this.
any suggestions?

[Torantulino]

This last erorr looks like it's due to a bad prompt, try the default one and see if it's any different.

Also please make sure you're on the latest version.

[SunilWarrier]

I tried the defaults and got the same error.
Let me try to debug the JSON at that time. I am not a Python programer, but I can understand.

[32mWelcome to Auto-GPT! ←[0mEnter the name of your AI and its role below. Entering nothing will load defaults.
←[32mName your AI: ←[0mFor example, 'Entrepreneur-GPT'
AI Name: Entrepreneur-GPT
←[94mEntrepreneur-GPT here! ←[0mI am at your service.
←[32mDescribe your AI's role: ←[0mFor example, 'an AI designed to autonomously develop and run businesses with the sole goal of increasing your net worth.'
Entrepreneur-GPT is: an AI designed to autonomously develop and run businesses with the sole goal of increasing your net worth.
←[32mEnter up to 5 goals for your AI: ←[0mFor example: Increase net worth Grow Twitter Account Develop and manage multiple businesses autonomously'
Enter nothing to load defaults, enter nothing when finished.
←[94mGoal←[0m 1:
Error: Prompt file not found
←[31mError: Invalid JSON
←[0mPlease provide a command for me to execute.
←[36mNEXT ACTION: ←[0mCOMMAND = ←[36mError:←[0m ARGUMENTS = ←[36mInvalid JSON←[0m
Enter 'y' to authorise command or 'n' to exit program...
←[35mInput:←[0m

[SunilWarrier]

When I run it in MS Code, I get the colours, and it will load the prompt file.
When I print that I get
Assistant reply JSON: What would you like the next command to be?
Not sure from where this is coming.

PS C:\development\PythonDevelopment\Auto-GPT\scripts> & C:/Users/sunil/AppData/Local/Programs/Python/Python311/python.exe c:/development/PythonDevelopment/Auto-GPT/scripts/main.py
Welcome to Auto-GPT! Enter the name of your AI and its role below. Entering nothing will load defaults.
Name your AI: For example, 'Entrepreneur-GPT'
AI Name: Entrepreneur-GPT
Entrepreneur-GPT here! I am at your service.
Describe your AI's role: For example, 'an AI designed to autonomously develop and run businesses with the sole goal of increasing your net worth.'
Entrepreneur-GPT is: an AI designed to autonomously develop and run businesses with the sole goal of increasing your net worth.
Enter up to 5 goals for your AI: For example: Increase net worth Grow Twitter Account Develop and manage multiple businesses autonomously'
Enter nothing to load defaults, enter nothing when finished.
Goal 1:
Assistant reply JSON: What would you like the next command to be?
Error: Invalid JSON
NEXT ACTION: COMMAND = Error: ARGUMENTS = Invalid JSON
Enter 'y' to authorise command or 'n' to exit program...
Input:y
-=-=-=-=-=-=-= COMMAND AUTHORISED BY USER -=-=-=-=-=-=-=
SYSTEM: Command Error: returned: Unknown command Error:
Assistant reply JSON: What would you like the next command to be?
Error: Invalid JSON
What would you like the next command to be?
NEXT ACTION: COMMAND = Error: ARGUMENTS = Invalid JSON
Enter 'y' to authorise command or 'n' to exit program...
Input:

[SunilWarrier]

I can see we are getting assistant_reply from the OpenAI response.

assistant_reply = response.choices[0].message["content"] in the response is a JSON but the content variable is just a string.
How can we try to load assistant_reply_json = json.loads(assistant_reply) as json in print_assistant_thoughts function?

Here the response is from gpt-3.5-turbo

I have tried the option also - python scripts/main.py --gpt3only but the same results.

Any clue to fix this? Thanks.

[richbeales]

closing as outdated.