when I see someone who promote and really happy about crewAI I immediately understand this guy don’t really understand what’s going on in real AI playground. IMHO. wish you to find out 🙌❤️
It is interesting running the application with crewai run while in the .venv environment vs using my conda environment. When I run it from the .venv environment the line in main.py from ai_research.crew import AiResearch. Which is what I called the project instead of docker, I did not need to remove the ai_research from in front of the crew. Nor, in the main.py file, did I have to add: if __name__ == "__main__": run() It would be very interesting to explore the differences between the two methods of running the code. crewai uses uv, but is using uv in a container production ready? So many questions. I appreciate your videos.
@@TylerReedAI That is interesting. I added it in order to make it work. But then the command runs uv which creates another virtual environment. I thought I was in the conda environment in which I installed Crewai, but I was looking around, so along with VScode not always putting me in the correct virtual environment, I may not have been where I thought I was. Thanks for letting me know.
@@TylerReedAI Error message: hint: The package `crewai-tools` depends on the package `docker` but the name is shadowed by your project. Consider changing the name of the project. To get this message, 1. I deactivate my conda environment. 2. I run the command: source .venv/bin/activate which activates an environment called docker which uv created. 3. I run the command: crewai run If you change the name of the project and follow the same steps the crew runs.
@@TylerReedAI Wow I wrote a whole response to this question and I come back and it is gone. The error was: hint: The package `crewai-tools` depends on the package `docker` but the name is shadowed by your project. Consider changing the name of the project. The error happens when I use the .venv environment created by crewai. In the pyproject.toml file there are scripts. When you activate the .venv environment and run one of the scripts in the pyproject file you get the error above.
@@TylerReedAI Please call your example crews something more unique. "Docker" might be a keyword or class name somewhere and cause conflicts, even if you use lowercase. Calling the crew anything else would help with clarity in the tutorial too. Just "dockerCrew" would make it clear when you're talking about the crew vs the tool, but using an interesting but simple use-case would go pretty far in this area too.
I just saw this. I am learning Crewai with the expectation of putting it in docker, and here it is. Thanks!
Dude, you explain these things so well! Building a crew into an image is a super idea💡
Thank you appreciate it!
This is super handy especially as I bounce from my windows to mac machines
when I see someone who promote and really happy about crewAI I immediately understand this guy don’t really understand what’s going on in real AI playground. IMHO. wish you to find out 🙌❤️
that is pretty cool !! 🤘 and it worked in my machine, yeah! 🤣
I created the docker crew. There was an .env file created with the MODEL var in it.
It is interesting running the application with crewai run while in the .venv environment vs using my conda environment. When I run it from the .venv environment the line in main.py from ai_research.crew import AiResearch. Which is what I called the project instead of docker, I did not need to remove the ai_research from in front of the crew.
Nor, in the main.py file, did I have to add:
if __name__ == "__main__":
run()
It would be very interesting to explore the differences between the two methods of running the code. crewai uses uv, but is using uv in a container production ready? So many questions. I appreciate your videos.
Did the program run continuously or one time only? I mean when we activate the docker.
Just once but we can set it to run continuously with cron!
At about 6:50 in the video, you say you need to add the load_dotenv() command. Why don't you need the python-dotenv in your requirements file?
Hey so because when you install crewai it is installed with it
@@TylerReedAI That is interesting. I added it in order to make it work. But then the command runs uv which creates another virtual environment. I thought I was in the conda environment in which I installed Crewai, but I was looking around, so along with VScode not always putting me in the correct virtual environment, I may not have been where I thought I was. Thanks for letting me know.
Expose as API?
👍❤️.
I've had crewai complain about the name docker. has anyone else had this issue?
What’s it complaining about?
@@TylerReedAI Error message: hint: The package `crewai-tools` depends on the package `docker` but
the name is shadowed by your project. Consider changing the name of the
project.
To get this message,
1. I deactivate my conda environment.
2. I run the command: source .venv/bin/activate which activates an environment called docker which uv created.
3. I run the command: crewai run
If you change the name of the project and follow the same steps the crew runs.
@@TylerReedAI Wow I wrote a whole response to this question and I come back and it is gone.
The error was:
hint: The package `crewai-tools` depends on the package `docker` but
the name is shadowed by your project. Consider changing the name of the
project.
The error happens when I use the .venv environment created by crewai.
In the pyproject.toml file there are scripts. When you activate the .venv environment and run one of the scripts in the pyproject file you get the error above.
@@TylerReedAI Please call your example crews something more unique. "Docker" might be a keyword or class name somewhere and cause conflicts, even if you use lowercase. Calling the crew anything else would help with clarity in the tutorial too. Just "dockerCrew" would make it clear when you're talking about the crew vs the tool, but using an interesting but simple use-case would go pretty far in this area too.