Open Devin: Create any Application with Open Source AI Engineer
HTML-код
- Опубликовано: 28 апр 2024
- 🎉 Excited to share something truly groundbreaking with you all today! We're exploring Open Devin, a phenomenal, free, and open-source digital AI software engineer that’s set to transform how we interact with coding projects. Unlike its counterpart, DevOn by Cognition AI, Open Devin offers a cost-free alternative packed with powerful features. Imagine assigning tasks to an AI and watching the magic happen in real-time. 🤖✨
🔍 In this video, I'll take you through the complete setup process of Open Devin, integrating it with OpenAI GPT and Olama models for enhanced performance. Whether you're coding a simple Python script or a full-fledged snake game, Open Devin is your go-to companion. The setup includes a detailed backend and frontend configuration, with everything from cloning the GitHub repository to starting the service on your local machine. 🖥️💡
💡 What You'll Learn:
How to clone and setup Open Devin on your computer.
Step-by-step instructions to integrate OpenAI GPT and Olama models.
Running a digital AI software engineer to perform coding tasks in real-time.
Tips for troubleshooting common issues during setup.
🔗 Don’t forget, I’ll be posting all commands and steps in the video description for easy access. And if you enjoy content on Artificial Intelligence, consider subscribing and hitting the bell icon to stay updated on my latest uploads!
👍 If you find this video helpful, please hit the like button and share it with others who might benefit from it as well.
🔗 Resources:
Patreon: / mervinpraison
Ko-fi: ko-fi.com/mervinpraison
Discord: / discord
Twitter / X : / mervinpraison
Code: github.com/OpenDevin/OpenDevin
Timestamps:
0:00 - Introduction to Open Devin
1:00 - Setting Up the Backend
2:00 - Configuring the Frontend
3:00 - Integrating OpenAI GPT
4:00 - Running Your First Task with Open Devin
5:00 - Integrating Olama Model
6:00 - Advanced Task: Creating a Snake Game
7:00 - Conclusion and What's Next
#OpenDevin #Install #Tutorial
#Devin #DevinAI #AI #CognitionAI #OpenDevin #AISoftwareEngineer #AIAgent #AISoftwareEngineerDevin #Devika #OpenDevinAI #CongitionAI #CognitionAIDevin #DevinAIDemo #DevinAITutorial #DevinAISoftwareEngineer #OpenDevinTutorial #OpenDevinGuide #InstallOpenDevin #HowToInstallOpenDevin #Guide #Tutorial 
Хобби
10/10 for including the instructions on running OpenDevin with Grok and Ollama, Mervin! =)
there is no requirements.txt file in the repository
How's this compare to devika? Would be nice if you'd give a quick comparison when there are similar products you've previously reviewed.
I’ve been using Devika since day one and I have tremendous respect for the community that are adding features and working tirelessly to improve it. That being said, there’s so many parts of it that are broken that it needs time to show its true potential.
Iv been having difficulty getting devils up and running
@@thenextweek2416 Yes, ATM to me it all loads but nothing moves...
@@thenextweek2416 Were you able to run devika successfully, I'm having an issue where I am seeing a white screen on local host 3000, I'll wait for the reply
Thank you for the great video. I have three questions:
1. How can I change Open Devin so that Open Devin accesses local models that are on my PC? I want that on the frontend (website) in the dropdown list "Open Devin Workspace" only lists local models that I have on my PC. How does it work?
2. Can Open Devin work with multiple models at the same time or does it only work with one model at runtime?
3. What exactly does an agent do?
Thank you in advance for your answers
Thank you for creating video on open Devin
Love from developers from Open Devin
Thanks Mervin for another great video, I managed to get the backend and frontend but it is on "Initializing agent (may take up to 10 seconds)..." forever so I'm stuck. The only thing seems to be wrong in the backend is "Error creating controller." line. I'm running it using vscode on windows, any suggestions?
Lord that helicopter scene looks great. WHAT BUDGET ?!
I LOVE YOUR VIDEOS - ADDICTED!!! MUCH LOVE FROM NEW ORLEANS
I missed only to understand what is the difference between Open Devin and Devika... Or how do they compare?
Hello Mervin, can we run this model in remote workspace like github codespace or remote redhat linux server. if possible could you please make one video on this?
At 1:45 I keep getting this error msg: -bash: conda: command not found.
I'm using Terminal on MacOS
Your videos are always very informative. I loved this one specifically since you covered both paid and open source LLM options.
Amazing work!! How does this compare to Devin, which was recently announced by Cognitive AI?
Fantastic! keep up the good work.
Thank you
I have some issues with i18n on the frontend folder :'(
About half the time I run OpenDevin, it disconnects halfway through the task. Says: "Error sending data to client.. Client websocket disconnected (, None).. INFO: connection closed" Is anyone else encountering this when running longer tasks? Has anyone discovered a fix. Could have something to do with a timeout set somewhere in the program, but I couldn't find that. Only seems to happen when the program stalls out for a while which makes me think it's a timeout issue. Thanks all.
This is amazing. I am looking forward to see this project develop. It has a nicer UI than Devika.
Sure we will see soon some use cases
Thanks Mervin! Do you think this has any benefit yet over just using something like simple GPT4 or Claude3 chat to help code? I guess the benefit is that we can use open source without the cost?
I showed basic example. It will be useful when we are a able to perform complex task. We can ask the Ai engineer to redo the code, modify the code., extend the code.. etc.
This is alpha version. once it gets to normal version then we can do all complex tasks.
Thank you! How would you rate this one llm development solution compared to something like autogen?
It’s all similar . It’s like sports shoe and formal shoe.
Devin being formal and AutoGen is like sports shoe
i can't find the requirments.txt file in the repo, do you have a solution please?! 🙏
aiohttp==3.9.3
aiosignal==1.3.1
annotated-types==0.6.0
anyio==4.3.0
asgiref==3.8.1
async-timeout==4.0.3
attrs==23.2.0
azure-core==1.30.1
azure-identity==1.15.0
backoff==2.2.1
bcrypt==4.1.2
beautifulsoup4==4.12.3
bs4==0.0.2
build==1.2.1
cachetools==5.3.3
certifi==2024.2.2
cffi==1.16.0
charset-normalizer==3.3.2
chroma-hnswlib==0.7.3
chromadb==0.4.24
click==8.1.7
coloredlogs==15.0.1
contourpy==1.2.0
cryptography==42.0.5
cycler==0.12.1
dataclasses-json==0.6.4
datasets==2.18.0
deprecated==1.2.14
dill==0.3.8
dirtyjson==1.0.8
distro==1.9.0
docker==7.0.0
exceptiongroup==1.2.0
fastapi==0.110.0
filelock==3.13.3
flatbuffers==24.3.25
fonttools==4.50.0
frozenlist==1.4.1
fsspec==2024.2.0
google-ai-generativelanguage==0.4.0
google-api-core==2.18.0
google-auth==2.29.0
google-generativeai==0.4.1
googleapis-common-protos==1.63.0
greenlet==3.0.3
grpcio==1.62.1
grpcio-status==1.62.1
h11==0.14.0
httpcore==1.0.5
httptools==0.6.1
httpx==0.27.0
huggingface-hub==0.22.2
humanfriendly==10.0
idna==3.6
importlib-metadata==7.0.0
importlib-resources==6.4.0
jinja2==3.1.3
joblib==1.3.2
jsonpatch==1.33
jsonpointer==2.4
kiwisolver==1.4.5
kubernetes==29.0.0
langchain==0.1.13
langchain-community==0.0.29
langchain-core==0.1.36
langchain-text-splitters==0.0.1
langsmith==0.1.38
litellm==1.34.12
llama-index==0.10.25
llama-index-agent-openai==0.2.1
llama-index-cli==0.1.11
llama-index-core==0.10.25.post3
llama-index-embeddings-azure-openai==0.1.6
llama-index-embeddings-huggingface==0.2.0
llama-index-embeddings-ollama==0.1.2
llama-index-embeddings-openai==0.1.7
llama-index-indices-managed-llama-cloud==0.1.5
llama-index-legacy==0.9.48
llama-index-llms-azure-openai==0.1.5
llama-index-llms-openai==0.1.14
llama-index-multi-modal-llms-openai==0.1.4
llama-index-program-openai==0.1.5
llama-index-question-gen-openai==0.1.3
llama-index-readers-file==0.1.12
llama-index-readers-llama-parse==0.1.4
llama-index-vector-stores-chroma==0.1.6
llama-parse==0.4.0
llamaindex-py-client==0.1.15
markupsafe==2.1.5
marshmallow==3.21.1
matplotlib==3.8.3
minijinja==1.0.16
mmh3==4.1.0
monotonic==1.6
mpmath==1.3.0
msal==1.28.0
msal-extensions==1.1.0
multidict==6.0.5
multiprocess==0.70.16
mypy==1.9.0
mypy-extensions==1.0.0
nest-asyncio==1.6.0
networkx==3.2.1
nltk==3.8.1
numpy==1.26.4
oauthlib==3.2.2
onnxruntime==1.17.1
openai==1.14.3
opentelemetry-api==1.24.0
opentelemetry-exporter-otlp-proto-common==1.24.0
opentelemetry-exporter-otlp-proto-grpc==1.24.0
opentelemetry-instrumentation==0.45b0
opentelemetry-instrumentation-asgi==0.45b0
opentelemetry-instrumentation-fastapi==0.45b0
opentelemetry-proto==1.24.0
opentelemetry-sdk==1.24.0
opentelemetry-semantic-conventions==0.45b0
opentelemetry-util-http==0.45b0
orjson==3.10.0
overrides==7.7.0
packaging==23.2
pandas==2.2.1
pillow==10.2.0
portalocker==2.8.2
posthog==3.5.0
proto-plus==1.23.0
protobuf==4.25.3
pulsar-client==3.4.0
pyarrow==15.0.2
pyarrow-hotfix==0.6
pyasn1==0.6.0
pyasn1-modules==0.4.0
pycparser==2.21
pydantic==2.6.4
pydantic-core==2.16.3
pyjwt==2.8.0
pymupdf==1.24.0
pymupdfb==1.24.0
pyparsing==3.1.2
pypdf==4.1.0
pypika==0.48.9
pyproject-hooks==1.0.0
python-dateutil==2.9.0.post0
python-dotenv==1.0.1
pytz==2024.1
pyyaml==6.0.1
regex==2023.12.25
requests==2.31.0
requests-oauthlib==2.0.0
rsa==4.9
ruff==0.3.4
safetensors==0.4.2
scikit-learn==1.4.1.post1
scipy==1.12.0
seaborn==0.13.2
sentence-transformers==2.6.1
setuptools==69.2.0
six==1.16.0
sniffio==1.3.1
soupsieve==2.5
sqlalchemy==2.0.29
starlette==0.36.3
striprtf==0.0.26
sympy==1.12
tenacity==8.2.3
termcolor==2.4.0
threadpoolctl==3.4.0
tiktoken==0.6.0
tokenizers==0.15.2
toml==0.10.2
tomli==2.0.1
torch==2.2.2
tqdm==4.66.2
transformers==4.39.2
typer==0.11.1
typing-extensions==4.10.0
typing-inspect==0.9.0
tzdata==2024.1
urllib3==2.2.1
uvicorn==0.29.0
uvloop==0.19.0
watchfiles==0.21.0
websocket-client==1.7.0
websockets==12.0
wrapt==1.16.0
xxhash==3.4.1
yarl==1.9.4
zipp==3.18.1
Same, this video got dated minutes after uploading. :-(
If you still want to use conda, create a requirements.txt with this following in the contents
datasets
pandas
litellm
termcolor
seaborn
docker
fastapi
uvicorn[standard]
ruff
mypy
langchain
langchain-core
langchain-community
llama-index
llama-index-vector-stores-chroma
chromadb
llama-index-embeddings-huggingface
llama-index-embeddings-azure-openai
llama-index-embeddings-ollama
google-generativeai
toml
..and continue the install. This is gleaned from the Pipfile in the repo, someone decided to change to the Pipenv package installer
python -m pipenv requirements > requirements.txt && python -m pip install -r requirements.txt
What's your favorite open sourced model and favorite app.. I'm a totally newbie and even had trouble running your package which looked pretty awesome too
Local model , I like mistral . But it is not as capable as GPT 4. Mistral is the best local model for now.
I will try to create a video for beginners to do all these things in the near future
Have you looked at hermes 2 pro? Its a finetuned function calling mistral 7b @MervinPraison
@MervinPraison I appreciate it using gpt to sort thru all my error messages helps...alot! Keep up the great work!
do you need to buy credits to use the api?
If you using windows 11 don’t even try it… doesn’t support windows 11 I seen when I was installing in the code.
I would love it if you made a video where you tried to do something with open Devin or similar that is more substantial including web browsing and that kind of thing not just trivial things that can be done one shot in GPT4 anyway.
Could you pretty please do a comparison video comparing the abilities and features of Aider vs Open Devin vs Devika : )
is this from WSL? or are you using a ubuntu operating system?
Thanks!
🙏
Hello, is there a way to use docker toolbox instead?, I have windows home edition which doesn't have and doesn't let install hyper-v (requisite for docker).
on Windows you can install the docker desktop to work with docker as well as the vscode extension and Bob's your uncle.
@@truehighs7845I had installed docker desktop and it didn't work because it required hyper-v (thus the question). I managed to get it working with docker toolbox and Bob's your uncle.
Thanks, frontend and Server are running , but keeping getthing this error no matter i what i select from the model selector on the UI. OpenAi api key are set properly.
Oops. Something went wrong: Error condensing thoughts: OpenAIException - Error code: 404 - {'error': {'message': 'The model `gpt-4-0125-preview` does not exist or you do not have access to it.', 'type': 'invalid_request_error', 'param': None, 'code': 'model_not_found'}}
Make sure you have subscribed or have access to OpenAI API
How to install it in windows 10??
Fronted shoud be desktop based electronjs or tauri for better DX
Why does it show gpt 3.5 turbo on the top right side?
That is the default model to run open devin . You can chance it if you want to
And it doesn’t change if we set ollama or grok.
May be an UI improvement is required in the near future to show the actual model
Ah, ok. Yes, should be. Open Devin seems to be interesting. I have some troubles to get GPT-4-ChatGPT to solve some things. Maybe an agent is able to perform better. Another video on my list to test. Thx!
What's about Windows User.... How to install Open Devin on windows pc ... Please make step by step Video.....
seems like WSL only option im on my way to install !
Can you use OpenDevin with LM Studio?
Yes you can
It uses liteLLM , so you can use their docs to configure LM Studio litellm.vercel.app/docs/providers…
Note. Don’t use /v1 towards the end of the URL . Also try various combination if it doesn’t
@@MervinPraisonThanks for the advice
Can i use it in colab notebook please reply or create a video if possible
Unfortunately you can’t use it in colab for now
opendevin has new update.can you make new installation step?
How can you use this on social media sites for growth
In my upcoming videos I’ll try to cover various use cases
Have you tried Jan?
No. Can you please sent me the link ?
Open Devin: Create any Application with Open Source AI Engineer. Thanks
Hey I was currently thinking developing an Android App using AI , no idea how to start with , if you could guide how to use it efficiently for android app developing that would really be helpful.
Literally ask chatgpt
can i use it with google api key
Yes you can
Just follow this docs. It uses liteLLM at the backend
litellm.vercel.app/docs/providers/gemini
Hi can you make tutorial for android app creating wit no codeing ai
Um. We need an installer AI. Devops and module management are always the hard part of computer programming anyway.
❤
🙏
I don't see any requirements.txt to run in this project.
python -m pipenv requirements > requirements.txt && python -m pip install -r requirements.txt
🎉❤
🙏
Suggesting to use more complicated examples for testing such software. Snake, for loop is too simple. Example should have backend and frontend at least. Also I didn't get how to create software which have multiple files with code.
the installation procedure changed I guess, no requirements file anymore.
Ran into the same challenge
Bro how to use Gemini
litellm.vercel.app/docs/providers/gemini
@@MervinPraison bro in this project
Yo update on this
This Python package management is always the biggest pain, It makes me don't want to use it anymore. In literally EVERY project I am getting these annoying issues. Really I hate it!!!!
why is he ignoring the messages to help those that are stuck
All these AI Engineer tools can only do small functions while RUclipsrs claim they can build anything. Oneday maybe theyll become advanced enough to actually build full apps but for now its more important to lea4n how to install and run them.
dude you go a million miles an hour
SLOW DOWN
Lol it’s not open sourced if you have to use api key from another provider like openai
You can use open source models via ollama , lm studio or others
@@MervinPraison yea but in this video you’re using api keys
All these are boring now. I want something which can directly make money for me.
lazy video with lazy examples, what's the point?
Main aim is to just show how to setup , so everyone can try the tool. It should be capable of doing lot more than what is demonstrated
@@MervinPraison I think you're missing the mark. Many fewer people want to know how to set it up than those who just want to see if it's even worth their time. Most people following this know how to set up a simple GitHub repo at this point. It shouldn't be your focus, you're limiting your views, certainly. Next time I see one of your videos, I'll be much less likely to click and you're catering to a minority of the sub niche.
I too was curious as to how it stacked up against the real Devin demos. I'll give it a shot myself.
The frontend and backend are up, but I keep getting this error, when I type into the UI OF OpenDevin:
Oops. Something went wrong: Error condensing thoughts: OpenAIException - Traceback (most recent call last): File "C:\Users\P S\miniconda3\envs\opendevin\Lib\site-packages\litellm\llms\openai.py", line 376, in completion raise e File "C:\Users\P S\miniconda3\envs\opendevin\Lib\site-packages\litellm\llms\openai.py", line 312, in completion openai_client = openai( ^^^^^^^ File "C:\Users\P S\miniconda3\envs\opendevin\Lib\site-packages\openai\_client.py", line 98, in __init__ raise openaiError( openai.openaiError: The api_key client option must be set either by passing api_key to the client or by setting the OPENAI_API_KEY environment variable
You missed setting your open ai api key step,
Export OPENAI_API_KEY
@@MervinPraison Thanks for replying, actually I followed your directions, I entered the key exactly like this: - OPENAI_API_KEY=sk-XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX (What directory do I enter to fix it)
@@GlobalAiWorkers # On Windows
setx OPENAI_API_KEY ""
# On Linux/Mac
export OPENAI_API_KEY=""
@@MervinPraison It said SUCCESS, but it did not work:
(opendevin) PS C:\Users\Phonetec Services\OpenDevin> setx OPENAI_API_KEY “”
SUCCESS: Specified value was saved.
(opendevin) PS C:\Users\Phonetec Services\OpenDevin>