# kwea123/ngp_pl

Instant-ngp in pytorch+cuda trained with pytorch-lightning (high quality with high speed, with only few lines of legible code)

Repository: https://github.com/kwea123/ngp_pl
Canonical: https://ross.abutalabs.com/products/ngp_pl
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: instant-ngp, cuda, nerf, pytorch-lightning, 3d-reconstruction, pytorch, novel-view-synthesis
Last push: 2023-06-16T01:03:45+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 1525, "days_push": 1175, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1286, forks 150 (observed 2026-08-28T04:04:14.948031+00:00)

## What it is
A PyTorch + CUDA implementation of NVIDIA's Instant-NGP focused on NeRF, trained with PyTorch Lightning for fast, high-quality neural radiance field rendering. It provides a concise, legible codebase intended to facilitate research on NeRF methods.

## Use cases
- train instant-ngp nerf models in pytorch
- novel view synthesis from photos
- 3d scene reconstruction from images
- research codebase for nerf experiments
- fast nerf training on a single gpu
- render synthetic nerf datasets like blender and tanks and temples

## When to choose
- you want a readable pytorch version of instant-ngp for research
- you need fast nerf training with pytorch-lightning on linux with an RTX gpu
- you want to modify or extend NeRF internals with few lines of code

## When to avoid
- you are on Windows or macOS since only Ubuntu 20.04 is supported
- you need a maintained production tool - the author recommends nerfstudio instead
- you lack an NVIDIA GPU with compute capability >= 7.5 and CUDA 11.3

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, graphics, simulation
- domain: computer-vision, graphics, machine-learning, deep-learning
- platform: python
- tags: nerf, instant-ngp, pytorch-lightning, novel-view-synthesis, 3d-reconstruction, neural-rendering, linux, gpu, cuda

## Member repositories
- kwea123/ngp_pl (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:14.948031+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:56:35.263928+00:00, confidence not recorded.
  - readme: https://github.com/kwea123/ngp_pl (fetched 2026-08-28T04:04:14.948031+00:00, sha decf87591b69)
- Data as of 2026-08-30T08:39:29.467469+00:00.