✨8+ Advanced and detailed Pytorch Projects.

✔️ the good and advanced projects , using #PyTorch

Computer Vision + sentiment Analysis + Recommender system and many more

⬇️
https://youtube.com/playlist?list=PLlff-0SljnifT40_TAlibwmyzh1xVhKTR&si=P3wIYBXK9TbGlA5O



#پایتون #Python

📱 @Python4all_pro

❌نحوه نوشتن بدترین کد پایتون

❌بدترین کدهای پایتون چه ویژگی هایی دارند ؟

⬅️ به این موارد دقت کنید اگر اینجوری کدنویسی می کنید سعی کنید روشتون رو اصلاح کنید و کدهای زیبا و سریع تری بنویسید

1. Use incomprehensible names of variables
Call variables X, Y, A, Thing. Abstraction is the key to confusion

def f(x, y, z=None):
    a = x * 2
    b = y + a if z else y - a
    c = [i for i in range(a) if i % 2]
    return sum(c) + b

2. Sake maximum logic in one line
Complex thornar expressions and nested List CompreHance - all in one line.

result = [x if x > 0 else (y if y < 0 else z) for x in data if x or y and not z]

3. Use Eval () and Exec ()
It is slow, unsafe and stupid - but spectacular.

eval("d['" + key + "']")

4. Reprint variables with different types
Let one variable be a line, and a number, and a list - a dynamic typification

value = "42"
value = int(value)
value = [value] * value

5. Use global variables
Change the condition of the application from anywhere. Especially from the inside of the functions.

counter = 0

def increment():
    global counter
    counter += 1

6. Use magic numbers and lines
Without explanation. Let colleagues guess why exactly 42 or "XYZ"

if user.role == "xyz" and user.level > 42:
    access_granted()

7. Ignore style and indentation
No PEP8, no rules. Write as you want

def foo():print("start")
 if True:
  print("yes")
   else:
    print("no")

8. Copy the code from Stack Overflow without delving
Ctrl+C is also a development

def complex_logic(x):
    return (lambda y: (lambda z: z**2)(y + 1))(x)

9. Invent abstraction unnecessary
Instead of a simple function - classes, factories and strategies

class HandlerFactory:
    def get_handler(self):
        class Handler:
            def handle(self, x): return x
        return Handler()

10. Add dead code
Never remove - suddenly comes in handy. And let it be loaded into every launch

def legacy_feature():
    print("This feature is deprecated")
    return
# нигде не вызывается

11. Do not write the documentation
Comments only interfere. Whoever wants to figure it out

def a(x): return x+1

#پایتون #Python

📱 @Python4all_pro

▶️Gremllm - Now your bugs have consciousness 

Gremllm is an unusual Python class in which all methods and properties are created dynamically using LLM. You describe, *what kind of object do you want *, and then Gremllm decides what should happen when the methods are called or the fields are addressed.

🎁 Installation

pip install gremllm

🔧 Example:

from gremllm import Gremllm

counter = Gremllm('counter')
counter.value = 5
counter.increment()
print(counter.value)               # → 6?
print(counter.to_roman_numerals())  # → VI?

🔸
✔️ Opportunities:
- dynamic behavior: everything is determined "on the fly" using LLM
- Support Openai, Claude, Gemini, and local models
- Wet Mode: You can build challenges of calls (methods return objects)
- Verbose Mode: Bodes which code was generated
- smart processing of errors and setting through inheritance

🖥 Github: https://github.com/ur-whitelab/gremllm

🔸

#پایتون #Python

📱 @Python4all_pro