Beyond.
Binary.
Translating complex deep learning theory into efficient, human-centric reality.
The Profile
I am a Computer Science undergraduate at KIIT University. My work exists at the intersection of AI and efficiency—specifically training small language models, fine-tuning LLMs, and building high-performance inference engines.
"name": "Kush Singh",
"role": "ML/Systems Engineer",
"stack": ["C++", "Python", "PyTorch", "Golang"],
"focus": ["LLM Training", "Inference Optimization", "SIMD"],
"currently": "Building lm_forge — a modular LM training engine",
"cgpa": 9.16,
"status": "open_to_opportunities"
}
Selected Work
Inference Engine
High-performance C++ inference engine for LLMs using SIMD/AVX2, with KV-Cache and quantization strategies. Built for raw speed on edge devices—optimized at the silicon level.
View Repository →Mental Health ChatBot
Fine-tuned Llama 3 8B on counseling datasets to enhance empathetic reasoning using PEFT adapters. Compressed via 4-bit Quantization (GGUF) for efficient edge deployment.
Micrograd
A scalar-valued autograd engine built from scratch in Python—understanding the roots of backpropagation and neural networks at the fundamental level.
View Repository →Style Transfer
Neural Image Style Transfer algorithm merging content and artistic style via feature maps.
View Repository →Competencies
// languages
C++ · C · Python · JavaScript · Java · Golang
// deep learning
PyTorch · HF Transformers · NumPy · Pandas · Scikit-Learn
// devops & web
Docker · Kubernetes · Linux · GitHub Actions · HTML5/CSS3
Timeline
B.Tech in Computer Science
Senior Secondary (Class 12)