Skip to content
View IMvision12's full-sized avatar
😀
AI/ML
😀
AI/ML
40 followers · 2 following

Achievements

Achievement: YOLOAchievement: Pull Sharkx3Achievement: Pair Extraordinairex2Achievement: Quickdraw

Achievements

Achievement: YOLOAchievement: Pull Sharkx3Achievement: Pair Extraordinairex2Achievement: Quickdraw

Highlights

  • Pro

Block or report IMvision12

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 is supported. This note will only be visible to you.
Report abuse

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

Report abuse
IMvision12/README.md

Hi there, I'm Gitesh 👋

Typing SVG

Portfolio LinkedIn Kaggle X

🎯 About Me

Coding

🤖 AI/ML Engineer, building & shipping models 🧩 Author of KerasFormers & Trinix 🏅 2× Kaggle Expert 🤗 Open-source contributor, Hugging Face & Keras 🌍 Based in: Fullerton, USA

I'm an AI/ML engineer who builds open-source libraries that make modern models easy to use, from pretrained Keras 3 model collections to Triton-accelerated PyTorch layers. I work across computer vision, LLMs, RAG, and agentic AI, and I love turning research into clean, reusable code.

✨ Highlights

  • 📦 Author of KerasFormers, Keras 3 pretrained models across Vision, LLM, VLM, Depth & Speech
  • ⚡ Built Trinix, Triton-accelerated PyTorch layers with automatic fallback optimization
  • 🏅 2× Kaggle Expert
  • 🤗 Contributor to Hugging Face (Transformers, Diffusers) and the Keras ecosystem (keras-cv, keras-hub)
  • 🧠 Implemented SOTA vision models in Keras/TF: SegFormer, ConvNeXt, NextViT, YOLO

📊 GitHub Analytics

GitHub Stats GitHub Streak
Contribution Graph

🛠️ Tools & Technology

Python C++ JavaScript PyTorch TensorFlow Keras scikit-learn Hugging Face LangChain vLLM TensorRT bitsandbytes Groq NumPy Pandas PIL OpenCV MySQL MongoDB Oracle Flask React GraphQL AWS Lambda Amazon S3 API Gateway Cognito AppSync CloudWatch Secrets Manager GitHub GitHub Actions CI/CD MLOps LLMOps VS Code Jupyter Google Colab

🚀 Featured Projects

🧠 ML Libraries & Performance

📦 KerasFormers

Pretrained Models for Keras 3

An open-source Keras 3 collection of pretrained models spanning Vision, LLM, VLM, Depth, and Speech, backend-agnostic across TensorFlow, JAX, and PyTorch.

🔥 Highlights:

  • One API, multiple modalities
  • Runs on TF / JAX / PyTorch backends
  • Clean, reusable model implementations

⚡ Tech Stack: Keras 3 JAX TensorFlow PyTorch

⚡ Trinix

Triton-Accelerated PyTorch Layers

Drop-in PyTorch layers accelerated with custom Triton kernels, featuring automatic fallback optimization when a fast path isn't available.

🔥 Highlights:

  • Custom fused Triton kernels
  • Automatic fallback to native ops
  • Plug-and-play with torch.nn

⚡ Tech Stack: Triton PyTorch CUDA

🧮 AdEMAMix Optimizer

Keras Implementation of AdEMAMix

A faithful Keras / TensorFlow implementation of the AdEMAMix optimizer, a mixture of two EMAs for faster, more stable training.

🔥 Highlights:

  • Paper-accurate optimizer port
  • Drop-in keras.optimizers compatible
  • Reproducible training recipes

⚡ Tech Stack: Keras TensorFlow

💻 localml-app

Run Hugging Face Models Locally

Browse the Hugging Face Hub, download with a click, and run detection, segmentation, depth, OCR, SAM, VLMs, ASR, TTS & diffusion fully locally. Your data never leaves your GPU.

🔥 Highlights:

  • No cloud, no upload, 100% local
  • Dozens of task types out of the box
  • One-click model download & run

⚡ Tech Stack: Hugging Face PyTorch Local Inference

🎨 Vision Models & Apps

🖼️ SegFormer-tf

Semantic Segmentation in TensorFlow

A clean TensorFlow / Keras implementation of SegFormer, efficient transformer-based semantic segmentation with hierarchical features.

🔥 Highlights:

  • Transformer encoder + lightweight MLP decoder
  • Reproducible training & inference
  • Readable, well-structured code

⚡ Tech Stack: TensorFlow Keras Transformers

📱 txtcode

Drive Coding CLIs from Your Phone

Control any coding CLI from your phone via WhatsApp, Telegram, Discord, Slack, Teams, or Signal, code from anywhere, no laptop required.

🔥 Highlights:

  • Works with 6+ chat platforms
  • Bridges chat ↔ your terminal
  • Code on the go

⚡ Tech Stack: TypeScript Node.js Chat APIs

github contribution snake



Pinned Loading

  1. KerasFormers KerasFormers Public

    KerasFormers: Open-source Keras 3 collection of pretrained models across Vision, LLM, VLM, Depth, Speech, and more

    Python 11 1

  2. txtcode txtcode Public

    </> Code from your phone control any coding CLI via WhatsApp, Telegram, Discord, Slack, Teams, or Signal

    TypeScript 7

  3. localml-app localml-app Public

    Browse the HuggingFace Hub. Download with a click. Run locally: detection, segmentation, classification, depth, OCR, SAM, VLMs, ASR, TTS, diffusion. No cloud, no upload, your data never leaves your…

    1

  4. trinix trinix Public

    Triton-accelerated PyTorch layers with automatic fallback optimization.

    Python 5 1

  5. yolo-keras yolo-keras Public

    Keras 3 implementation of YoloV5 and YoloV8

    Python

  6. AdEMAMix-Optimizer-Keras AdEMAMix-Optimizer-Keras Public

    A Keras 3 Implementation of AdEMAMix Optimizer

    Python 10 2