Thanks to visit codestin.com
Credit goes to github.com

Skip to content
View ikaganacar1's full-sized avatar
☄️
☄️
41 followers · 39 following

Achievements

Achievement: YOLOAchievement: Pull Shark

Achievements

Achievement: YOLOAchievement: Pull Shark

Block or report ikaganacar1

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.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ikaganacar1/README.md

👋 Hi! I’m İsmail Kağan

  • AI researcher and builder passionate about deep learning and open-source innovation
  • Software Engineering major at Atılım University
  • Long-time Linux user (Arch + Hyprland, btw )
  • Explore my useless website: ikaganacar.com

🤖 Artificial Intelligence & Machine Learning

AI stack icons Hugging Face logo Kaggle logo
  • Building, fine-tuning, and deploying LLMs and computer vision models
  • Exploring model compression and edge-AI deployment
  • Interested in practical applications of transformer-based architectures

🌐 Connect with Me

Hugging Face Kaggle LinkedIn Medium

✍️ Latest Writings

  • Read TOON Benchmarks: A Critical Analysis of Different Results (Nov 7, 2025)
  • Read TOON: New data format for LLM Applications (Nov 7, 2025)
  • Read Fine-Tuning LLM with Unsloth: A Practical Guide to Training Models like Qwen3 8B on a Consumer GPU (Sep 25, 2025)
  • Read Anomaly Detection with Holt Winters Simple Exponential Smoothing in Python (Jul 21, 2025)

    • Featured Work

    📊 GitHub Stats

    stats graph languages graph

    Creating bugs since 2019 — now they help debug themselves.

    Pinned Loading

    1. Unsloth-Easy-Finetuning-Pipeline Unsloth-Easy-Finetuning-Pipeline Public

      A production-ready pipeline for fine-tuning large language models using Unsloth with YAML-based configuration, advanced training features, and web-based model serving.

      Python

    2. ismail ismail Public

      ismail is a from-scratch Turkish language model implementation designed for low-end hardware, built and trained on a single RTX 5070 (12GB).

      Python

    3. Reid_Inference_Pipeline Reid_Inference_Pipeline Public

      A TensorRT-optimized testing framework for Person Re-Identification (ReID) and person detection models on NVIDIA GPUs, especially Jetson devices. Supports evaluation of trained models on both singl…

      Python 1 1

    4. ikaganacar.com-Main ikaganacar.com-Main Public

      The source code of my personal website.

      HTML 2