This PR fix devika #603
This PR fix devika #603
Conversation
If no project is selected, return None immediately
remove backtick (`) from generated files
add normpath to resolvle backslach compatibilite beetwen linux and windows
update save code function to resolve backslash issue
When a project is created, the folder is created immediately. When the "Delete Project" button is triggered, the project is deleted immediately.
Setting the temperature to 0 is not good, because the model will return the prompt repeatedly.
Setting the temperature to 0 is not good, because the model will return the prompt repeatedly.
Setting the temperature to 0 is not good, because the model will return the prompt repeatedly. so setting temperature to 1 will be better option.
Setting the temperature to 0 is not good, because the model will return the prompt repeatedly.
Setting the temperature to 0 is not good, because the model will return the prompt repeatedly.
Change the origin of frontend URL
handle more use cases of validate function of coder
this change fix this error File "C:\Users\pc\Desktop\stream_main\devika\src\agents\coder\coder.py", line 131, in execute
valid_response = self.validate_response(response)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\pc\Desktop\stream_main\devika\src\agents\coder\coder.py", line 37, in validate_response
self.logger.debug(f"Response from the model: {response}")
File "C:\Users\pc\Desktop\stream_main\devika\src\logger.py", line 32, in debug
self.logger.debug(message)
File "C:\Users\pc\Desktop\stream_main\devika\.venv\Lib\site-packages\fastlogging\fastlogging.py", line 248, in debug
self.__log(DEBUG, msg, args, kwargs)
File "C:\Users\pc\Desktop\stream_main\devika\.venv\Lib\site-packages\fastlogging\fastlogging.py", line 238, in __log
self._logMessage(None, (log_time, domain, level, msg, kwargs), 0)
File "C:\Users\pc\Desktop\stream_main\devika\.venv\Lib\site-packages\fastlogging\fastlogging.py", line 423, in _logMessage
print(message, file=self.stdout if level < ERROR else self.stderr)
File "C:\Users\pc\Desktop\stream_main\devika\.venv\Lib\site-packages\colorama\ansitowin32.py", line 47, in write
self.__convertor.write(text)
|
i tested it and its working i would happy if someone develop some sand box fonctionalite to run code and internal server for each project and visualize the internal dev server in Devika browser |
|
Is it possible to upload the entire repaired files? |
|
Do |
|
Do and you will get the updategh pr checkout 603 how? |
|
(devika) C:\Users\m\Desktop\s\devika>python devika.py A module that was compiled using NumPy 1.x cannot be run in If you are a user of the module, the easiest solution will be to Traceback (most recent call last): File "C:\Users\m\Desktop\s\devika\devika.py", line 8, in |
|
(devika) C:\Users\m\Desktop\s\devika>gh pr checkout 603
|
|
24.06.16 13:14:44: root: INFO : Loading sentence-transformer BERT models... |
|
(devika) C:\Users\m\Desktop\s\devika\ui>pip install numpy (devika) C:\Users\m\Desktop\s\devika\ui> |
|
Not working after update |
|
I tried through Pinocchio and the same problem. Does it require gpu |
Do I fresh install, your problem related to Sentencetransformer and the error say , there is just missing dependency ,you should run in Devika root directory not UI directory, and run
|
It's working just you should install all requirements, |
|
(devika) C:\Users\m\Desktop\t\devika>uv pip install -r requirements.txt (devika) C:\Users\m\Desktop\t\devika> |
|
(devika) C:\Users\m\Desktop\t\devika>uv pip install -r requirements.txt (devika) C:\Users\m\Desktop\t\devika>gh pr checkout 603 (devika) C:\Users\m\Desktop\t\devika> |
|
anaconda prompt (base) C:\WINDOWS\system32>cd C:\Users\m\Desktop\p (base) C:\Users\m\Desktop\p>git clone https://github.com/stitionai/devika.git (base) C:\Users\m\Desktop\p>cd devika (base) C:\Users\m\Desktop\p\devika>gh pr checkout 603
(base) C:\Users\m\Desktop\p\devika>uv venv (base) C:\Users\m\Desktop\p\devika>.venv\Scripts\activate (devika) (base) C:\Users\m\Desktop\p\devika>uv pip install -r requirements.txt
(devika) (base) C:\Users\m\Desktop\p\devika>playwright install --with-deps (devika) (base) C:\Users\m\Desktop\p\devika>python devika.py A module that was compiled using NumPy 1.x cannot be run in If you are a user of the module, the easiest solution will be to Traceback (most recent call last): File "C:\Users\m\Desktop\p\devika\devika.py", line 8, in (devika) (base) C:\Users\m\Desktop\p\devika> |
|
(base) C:\Users\m>cd C:\Users\m\Desktop\p (base) C:\Users\m\Desktop\p>git clone https://github.com/stitionai/devika.git (base) C:\Users\m\Desktop\p>cd devika (base) C:\Users\m\Desktop\p\devika>uv venv (base) C:\Users\m\Desktop\p\devika>.venv\Scripts\activate (devika) (base) C:\Users\m\Desktop\p\devika>uv pip install -r requirements.txt
(devika) (base) C:\Users\m\Desktop\p\devika>playwright install --with-deps (devika) (base) C:\Users\m\Desktop\p\devika>python devika.py A module that was compiled using NumPy 1.x cannot be run in If you are a user of the module, the easiest solution will be to Traceback (most recent call last): File "C:\Users\m\Desktop\p\devika\devika.py", line 8, in (devika) (base) C:\Users\m\Desktop\p\devika>python devika.py A module that was compiled using NumPy 1.x cannot be run in If you are a user of the module, the easiest solution will be to Traceback (most recent call last): File "C:\Users\m\Desktop\p\devika\devika.py", line 8, in |
|
After you active venv
|
|
Make sure this requirements is satisfied |
|
why set the temperature to 1? |
When the temperature is 0, the model will always choose the most probable or likely next token based on its training, without any randomness or exploration. This can result in the model getting "stuck" in a loop, generating the same or highly similar text over and over again. When setting temperature to 1. The validate function have low rate of fail , |
|
Now , the temperature is set back to 0 , and I resolved the conflict |
Hello Younes @darrassi1 still even after updating all changes and temp set to 0 , still getting same error RuntimeError: Numpy is not available i think issue here with versions as numpy version i have is site-packages (2.0.0) A module that was compiled using NumPy 1.x cannot be run in If you are a user of the module, the easiest solution will be to another suggestion might be multiple python versions installed in my case i found that i have 2 versions (venv) (base) PS C:\Users\Mahmoud\Desktop\Devika\devika> pip --version i might also need to uninstall one of them and when i do install and run the environments has to be all in one place if someone had same issue and fixed by any of above, please confirm |
I'm not familliar with Devika, but it seems you're running it with embedded python. In that case, you need to run cmd from folder 'C:\Users\Mahmoud\Desktop\Devika\devika' and input command 'pip install numpy==1.26.4', which will replace your 2.0.0 version. |
Thanks a lot AV :) it actually worked after installing older version and indeed it uninstalled the newer |
|
Requirement already satisfied: keybert in c:\users\m\appdata\local\programs\python\python311\lib\site-packages (0.8.5) (devika) C:\Users\m\Desktop\111\devika>python C:\Users\m\Desktop\111\devika\devika.py A module that was compiled using NumPy 1.x cannot be run in If you are a user of the module, the easiest solution will be to Traceback (most recent call last): File "C:\Users\m\Desktop\111\devika\devika.py", line 8, in (devika) C:\Users\m\Desktop\111\devika> |
|
all_embeddings = np.asarray([emb.numpy() for emb in all_embeddings]) (devika) C:\Users\m\Desktop\111\devika>pip install numpy==1.26.4 (devika) C:\Users\m\Desktop\111\devika> |
You have numpy installed globally in your 'c:\users\m\appdata\local\programs\python\python311\lib\site-packages' folder, while Devika uses virtual environment. |
| @@ -26,7 +26,11 @@ | |||
|
|
|||
|
|
|||
| app = Flask(__name__) | |||
| CORS(app, resources={r"/*": {"origins": ["https://localhost:3000"]}}) # Change the origin to your frontend URL | |||
| CORS(app, resources={r"/*": {"origins": # Change the origin to your frontend URL | |||
| [ | |||
No description provided.