🆕 Microsoft has quietly updated the Phi-3 Mini

- significantly improved code understanding in Python, C++, Rust and Typescript
- Improved output, now it is more structured
- improved understanding of complex sentences
- added support for the <|system|> tag.
- Improved reasoning ability and understanding of long contexts

4K and 128K benchmarks are affected by this update

🤗 Hugging Face

http://t.me/codeprogrammer ✅

Thinking about learning Python 🖥

👩‍💻 Coding in 2024
https://www.youtube.com/shorts/hBgTsN9PVtE


☄️ Where to use Python in Real World
https://www.youtube.com/shorts/gD2YXjMMZs8


Why Python Rules the Coding World! 🐍👑
https://www.youtube.com/shorts/oNDmMSO6ZK0

http://t.me/codeprogrammer ✅

🖥 python pattern notes

💬 #python #notes

http://t.me/codeprogrammer ✅

🖥 How to hide secret text inside an image using Python?

Nothing complicated, you just need the Stegano library:

# pip install stegano

from stegano import lsb
secret = lsb.hide('image.png', 'very secret text')
secret.save('secret_image.png')

print(lsb.reveal('secret_image.png'))

A quick question: how is this ability to store text in a picture implemented?
How easy is it to detect such text hiding?

🖥 GitHub

http://t.me/codeprogrammer ⭐️⭐️

Initialization can have a significant impact on convergence in training deep neural networks.

Read our guide to learn how to initialize neural network parameters effectively:
https://www.deeplearning.ai/ai-notes/initialization/index.html

http://t.me/codeprogrammer ❤️

Football Analysis in Python with Yolov8 & OpenCV

Video: https://youtu.be/yJWAtr3kvPU?si=Ix9t3QiXR3jwbog7

Deep learning web application for football analysis with Streamlit.

Used material links:
Project repo: https://github.com/Hmzbo/Football-Analytics-with-Deep-Learning-and-Computer-Vision

Custom dataset tips: https://docs.ultralytics.com/yolov5/tutorials/tips_for_best_training_results/

Annotation tools:
labelme: https://github.com/wkentaro/labelme

cvat: https://github.com/opencv/cvat

http://t.me/codeprogrammer 💬