Skip to content
View gouthamk16's full-sized avatar
8 followers · 2 following

Highlights

  • Pro

Block or report gouthamk16

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.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. 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
gouthamk16/README.md

I build robot learning, perception and ML systems.

Project What it is Result
lc-act Vision-language-action policy for a Franka Panda arm (LIBERO, MuJoCo) 95% task success on LIBERO-Object with 7.6M parameters
Slam Stereo visual SLAM for aerial drones from scratch, with my own bundle adjuster 2 to 17 cm error on 10 of 11 EuRoC flights; 1.74 m on KITTI 00
tiny-diffusion Text diffusion language model, SEDD vs MDLM on one backbone GPT-2 perplexity from 350.5 to 24.75; TensorRT FP8 4.3x faster
miniXLA ML compiler in C that fuses ops and emits PTX FP32 GEMM at 84.3% of cuBLAS
vision-fsd Camera and LiDAR perception on nuScenes Temporal occupancy, object velocity and a local planner

My ML fingerprint on Deep-ML

Pinned Loading

  1. lc-act lc-act Public

    Tiny vision language action (VLA) model built on Language-conditioned Action Chunking (LC-ACT) architecture.

    Python

  2. quantum quantum Public archive

    qubit manipulation and quantum algorithms using Q# and Qiskit for large scale compute.

    Q#

  3. vision-fsd vision-fsd Public

    Surround-camera and LiDAR perception on nuScenes: camera BEV, temporal occupancy, object velocity and a local planner

    Python 2

  4. tiny-diffusion tiny-diffusion Public

    Text diffusion language model: SEDD vs MDLM on a shared 77M backbone, with TensorRT FP16/FP8 inference

    Python 3

  5. memboost memboost Public

    CUDA mixed-precision 2/4-bit weight quantizer for LLMs (reimplements arXiv:2311.16442)

    Python 1

  6. miniXLA miniXLA Public

    A lightweight ML compiler that generates PTX

    C 1