Open zkMyst opened 1 month ago
From this you posted, it is an older version. (python main.py). Try updating and check again.. (run_ui.py) is the new way. Start Docker Desktop before running run_ui.py. (conda create --name agentzero python=3.10) (conda activate agentzero) now do install again from the github page. Follow the instructions and it should work fine then.
Updated, ran the insall commands and:
Connection to Docker failed. Is docker or Docker Desktop running?
Traceback (most recent call last):
File "/opt/miniconda3/envs/agentzero/lib/python3.10/site-packages/docker/api/client.py", line 207, in __init__
self._version = self._retrieve_server_version()
File "/opt/miniconda3/envs/agentzero/lib/python3.10/site-packages/docker/api/client.py", line 230, in _retrieve_server_version
raise DockerException(
docker.errors.DockerException: Error while fetching server API version: ('Connection aborted.', FileNotFoundError(2, 'No such file or directory'))
did you set the correct settings? and change your .env file? put the correct api keys in?
chat_llm = models.get_openai_chat(model_name="gpt-4o-mini", temperature=0)
# chat_llm = models.get_ollama_chat(model_name="gemma2:latest", temperature=0)
# chat_llm = models.get_lmstudio_chat(model_name="lmstudio-community/Meta-Llama-3.1-8B-Instruct-GGUF", temperature=0)
# chat_llm = models.get_openrouter_chat(model_name="mattshumer/reflection-70b:free")
# chat_llm = models.get_azure_openai_chat(deployment_name="gpt-4o-mini", temperature=0)
# chat_llm = models.get_anthropic_chat(model_name="claude-3-5-sonnet-20240620", temperature=0)
# chat_llm = models.get_google_chat(model_name="gemini-1.5-flash", temperature=0)
# chat_llm = models.get_groq_chat(model_name="llama-3.1-70b-versatile", temperature=0)
# utility model used for helper functions (cheaper, faster)
utility_llm = chat_llm # change if you want to use a different utility model
# embedding model used for memory
embedding_llm = models.get_openai_embedding(model_name="text-embedding-3-small")
# embedding_llm = models.get_ollama_embedding(model_name="nomic-embed-text")
# embedding_llm = models.get_huggingface_embedding(model_name="sentence-transformers/all-MiniLM-L6-v2")
# embedding_llm = models.get_lmstudio_embedding(model_name="nomic-ai/nomic-embed-text-v1.5-GGUF")
Yes didn't change anything other than input my api keys for openAI and Perplexity...
@zkMyst Please use Python 3.12, though I don't think it's the only issue there. Are you sure that your user is in the Docker user group?
$ sudo usermod -aG docker $USER
Sorry im on MacOS, dont think there is user management with the desktop app there.
@zkMyst I'm sorry, I've been inactive. Try this, and let me know please: Pull up the terminal and do:
id -nG
that is going to show you all the groups in the system, so you can verify if your user is in the 'docker' group.
Then you should be able to add your user to docker group with this command:
sudo dseditgroup -o edit -a $USER -t user docker
In normal operations it's better to let Docker Desktop manage user permissions, but still. Let me know.
FIXED!
I had this issue as well. I spent nearly 3 hours on it. My fnal fix was to completely uninstall docker (and all its components) and reinstall it. verify it via the terminal (you must enable starting when system starts , and enable cli) otherwise it wont work
P.S. Im also running a mac
FIXED!
I had this issue as well. I spent nearly 3 hours on it. My fnal fix was to completely uninstall docker (and all its components) and reinstall it. verify it via the terminal (you must enable starting when system starts , and enable cli) otherwise it wont work
P.S. Im also running a mac
Thank you, this will prove useful for some users. I had similar issues with Linux VMs that doesn't support nested virtualization and require a docker-ce setup to run A0 with. I'll write docs on the matter.
Ok I found the problem, in macOS you have to check the "Allow the default Docker socket to be used (requires password)" checkbox to allow other tools to interact with docker socket.
Now it works perfectly...
@frdel maybe you could add it to the macOS installation instructions...
Ok I found the problem, in macOS you have to check the "Allow the default Docker socket to be used (requires password)" checkbox to allow other tools to interact with docker socket.
Now it works perfectly...
@frdel maybe you could add it to the macOS installation instructions...
Thank you kindly, I'm adding this exception to the docs!
Hi,
I've got Docker desktop running, but after starting a new execution the agent errors with:
I'm running
python main.py
in a miniconda environement (I can see (base) in my terminal). Does anyone know what this could be due to ?For context this is the most recent Traceback: