Skip to content
View MrudhuhasM's full-sized avatar
📚
📚
7 followers · 3 following

Achievements

Achievement: Pull SharkAchievement: YOLO

Achievements

Achievement: Pull SharkAchievement: YOLO

Block or report MrudhuhasM

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
MrudhuhasM/README.md

Hi, I'm Mrudhuhas 👋

ML Engineer | Training → Inference → Production

8+ years building ML systems end-to-end: from training custom models to optimizing inference at the GPU kernel level.

Current focus

  • GPU performance engineering with CUDA and Triton
  • LLM training, inference optimization, and evaluation
  • Production-oriented RAG and ML systems

Featured work

Project What it demonstrates
gpt2-engine GPT-2 inference with custom Triton kernels, KV caching, quantization, and benchmarks
triton-code Triton kernel experiments and performance analysis
cuda-reduction CUDA reduction strategies benchmarked against CUB
tiny-stories-llm Transformer language-model training from scratch
ssl-foundations Self-supervised representation evaluation with linear probing and k-NN
meditations-rag Agentic RAG with LangGraph, hybrid retrieval, Qdrant, and FastAPI

📫 Open to remote opportunities (contract or full-time) | Flexible on US/EU timezones

LinkedIn • Email

Pinned Loading

  1. meditations-rag meditations-rag Public

    Agentic RAG for Marcus Aurelius' Meditations using LangGraph, hybrid retrieval, Qdrant, FastAPI, and local or hosted LLMs.

    Python

  2. gpt2-engine gpt2-engine Public

    GPT-2 inference engine with custom Triton kernels, KV caching, quantization, and reproducible GPU benchmarks.

    Python 4

  3. cuda-reduction cuda-reduction Public

    CUDA reduction kernels benchmarked against CUB, with optimization experiments across multiple implementation strategies.

    Cuda

  4. ssl-foundations ssl-foundations Public

    Self-supervised learning experiments with encoder evaluation through linear probing and k-nearest-neighbor classification.

    Python

  5. tiny-stories-llm tiny-stories-llm Public

    Transformer language model trained from scratch on TinyStories with configurable experiments and Hugging Face integration.

    Jupyter Notebook

  6. triton-code triton-code Public

    Benchmark-driven Triton kernel experiments covering vector operations, reductions, online softmax, and GPU optimization.

    Python