I pretrained a 0.5B model from scratch - Dexter
Modern Llama-style architecture: RoPE, RMSNorm, SwiGLU. Trains nicely on consumer GPUs and already produces coherent, sensible outputs.
Right now it barely competes with low-parameter Qwen models .
I'm deleting this soon because it's a legit cash-printing formula.
๐ฃ๐ฎ๐ถ๐ฑ ๐๐ผ๐๐ฟ๐๐ฒ ๐๐ฅ๐๐ (PART - 3)
1. Artificial Intelligence + Data Analyst
2. Machine Learning + Data Science
3. Cloud Computing + Web Development
4. Ethical Hacking + Hacking
5. Data Analytics + DSA
6. AWS Certified + IBM COURSE
7. Data Science + Deep Learning
8. BIG DATA + SQL COMPLETE COURSE
9. Python + OTHERS
10 MBA + HANDWRITTEN NOTES
(72 Hours only ) Cost About - $500
To get: -
1. Follow (So I can DM you )
2. Like & retweet
3. Reply " Send "
Hey all , sharing on behalf of a friend who runs a talented dev team specializing in fast, reliable software and web solutions.
If you or your network need support building or scaling products, feel free to reach out. I work with them at times and can help connect you directly.