# rinongal/textual_inversion

Repository: https://github.com/rinongal/textual_inversion
Canonical: https://ross.abutalabs.com/products/textual_inversion
Language: Jupyter Notebook
License: MIT
License Family: permissive
Last push: 2023-02-27T08:16:02+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1492, "days_push": 1283, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3055, forks 284 (observed 2026-08-28T04:07:39.746886+00:00)

## What it is
Official implementation of the Textual Inversion paper, which learns new word embeddings in a frozen text-to-image (Latent Diffusion) model from 3-5 example images to capture user-specific concepts. It provides training scripts, configs, and sample inversions for personalized image generation.

## Use cases
- teach a text-to-image model my own concept from a few photos
- learn a new token embedding for a custom object or style
- generate images of a specific subject in new scenes
- invert a small image set into a diffusion model embedding
- personalize latent diffusion text-to-image generation

## When to choose
- you want the original, paper-faithful textual inversion implementation for Latent Diffusion models
- you need to reproduce research results or build on the paper's method
- you work with the LDM text-to-image checkpoint and want embedding-based personalization

## When to avoid
- you mainly want Stable Diffusion personalization in production - newer tools like diffusers or kohya trainers are better maintained
- you need pre-trained embeddings or polished tooling - several TODO items were never completed
- you want a no-training prompt-based approach rather than embedding optimization

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-training, machine-learning, deep-learning, stable-diffusion, image-processing
- domain: deep-learning, machine-learning, artificial-intelligence, image-processing
- platform: python
- tags: textual-inversion, text-to-image, latent-diffusion, personalization, research-code, embeddings, diffusion-models, gpu, linux

## Member repositories
- rinongal/textual_inversion (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:39.746886+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 2026-08-30T07:28:50.996068+00:00, confidence not recorded.
  - readme: https://github.com/rinongal/textual_inversion (fetched 2026-08-28T04:07:39.746886+00:00, sha cc408ef8d676)
- Data as of 2026-08-30T08:39:29.467469+00:00.