Skip to content
View Ammar-Alnagar's full-sized avatar
๐Ÿซ 
Deoding Complexity into Simplicity
๐Ÿซ 
Deoding Complexity into Simplicity

Block or report Ammar-Alnagar

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this userโ€™s behavior. Learn more about reporting abuse.

Report abuse
Ammar-Alnagar/README.md

Greetings, Fellow Tech Wizard! ๐Ÿ‘‹ ๐Ÿง™โ€โ™‚๏ธ

LinkedIn

๐ŸŒŸ Welcome to My Corner of GitHub!

Iโ€™m Ammar Alnagar

Cognitive AI Engineer

Here, innovation meets impact. Iโ€™m passionate about crafting solutions that decode complexity and turn possibilities into reality. From reasoning-driven models to multimodal AI systems, Iโ€™m constantly pushing the boundaries of whatโ€™s possible.

Whether itโ€™s fine-tuning large language models, deploying AI on edge devices, or diving deep into mathematical problem-solving, my mission is simple: solve real-world problems and redefine what AI can do.


๐Ÿš€ Highlights of My Work

๐Ÿ‘พ๐Ÿค–PathfinderAI-S1 (Extreme Reasoning version)

The latest model in the PathfinderAI family , built ontop deepseek-r1 distills, with a mix of Reasoning models traces from o1,o3-mini,deepseek-r1,gemini-thinking. its trained using GRPO , mainly on math and reasoning task and its better than o1-preview and Qwen2.5 qwq in math and hard reasoning tasks.

๐Ÿง Cogito-R1 -Math Wizard

an advanced reasoning and Chain-of-Thought (CoT) model fine-tuned on Qwen2.5-32B using Unsloth, TRL, and GRPO. Designed for complex reasoning, mathematical problem-solving, and structured inference, Cogito-R1 delivers state-of-the-art performance.

๐Ÿ”ฑ CogniLink 70B โ€“ Decoding Complexity

A state-of-the-art reasoning model fine-tuned for advanced Chain-of-Thought (CoT) tasks. Designed to tackle complex decision-making and serve as a cornerstone for solving intricate problems.

๐Ÿง  PathfinderAI 32B โ€“ Redefining Real-Time Thinking

A powerhouse model fine-tuned for reasoning and mathematical problem-solving, delivering unparalleled performance in real-time scenarios.

๐Ÿ“œ Document Sage โ€“ Intelligence Meets Efficiency

An AI-powered semantic search and document processing system that understands context, extracts insights, and accelerates workflows.

๐Ÿ”— RAG Architect โ€“ Building Smarter Agents

Harnessing Retrieval-Augmented Generation (RAG) to create context-aware AI agents that are efficient, reliable, and versatile.


๐ŸŒฑ Continuous Learning & Exploration

Innovation is fueled by curiosity. Hereโ€™s what Iโ€™m diving into:

  • Reinforcement Learning & Multi-Agent Systems: Building adaptive, collaborative AI systems.
  • Mamba2 State Space Models (SSMs): Advancing dynamic modeling to unlock new possibilities.
  • Probabilistic Inference: Enhancing decision-making with Bayesian methods and particle filtering.
  • Few-Shot & Meta-Learning: Maximizing learning efficiency with minimal data.

๐Ÿ’ป How I Can Help

Whether youโ€™re solving a challenge or dreaming big, I bring:
๐Ÿ”น Practical AI Solutions: From document processing to intelligent reasoning systems.
๐Ÿ”น Multimodal AI Expertise: Creating seamless text, visual, and contextual integrations.
๐Ÿ”น LLM Optimization: Deploying powerful AI models on resource-constrained devices like Raspberry Pi.
๐Ÿ”น Real-World Impact: Crafting AI-driven tools that truly make a difference.


๐Ÿค Letโ€™s Build Together

Collaboration fuels innovation! Iโ€™m excited to connect with visionaries who share a passion for:

  • Reasoning-Driven AI Models: Redefining decision-making with powerful tools.
  • Multimodal and Context-Aware Systems: Creating richer, smarter user experiences.
  • Edge-AI Deployments: Unlocking the power of AI in embedded devices.

๐Ÿ“ฌ Letโ€™s connect on LinkedIn, or feel free to explore my work here. Together, we can build the future of AIโ€”one model, one system, one breakthrough at a time!

๐Ÿ† GitHub Stats

ammar-alnagar

Ammar-Alnagar's Stats


GitHub Snake

Pinned Loading

  1. Enlightener Enlightener Public

    Enlightener, the cutting-edge Retrieval-Augmented Generation (RAG) system that revolutionizes query responses. By combining the power of SSM models (Mamba) with the advanced GGUF Mamba models,

    Python 14 2

  2. Cogito-R1- Cogito-R1- Public

    Python

  3. Hybrid-ARAG Hybrid-ARAG Public

    Python

  4. open-r1_Z open-r1_Z Public

    Forked from huggingface/open-r1

    Fully open reproduction of DeepSeek-R1

    Python

  5. PRIME-cogi PRIME-cogi Public

    Forked from PRIME-RL/PRIME

    Scalable RL solution for advanced reasoning of language models

    Python

  6. rStar rStar Public

    Forked from microsoft/rStar

    Python