---
license: other
license_name: proprietary
license_link: https://huggingface.co/AbteeXAILab/lumynax-embed-e5-mistral-7b/resolve/main/LICENSE.txt
library_name: python
pipeline_tag: feature-extraction
language:
- en
tags:
- lumynax
- lumynax-embed-e5-mistral-7b
- embeddings
- abteex-ai-labs
- local-first
---

# LumynaX Embed E5 Mistral 7B

LumynaX Embed E5 Mistral 7B is a hydrated LumynaX embedding package built from the official `intfloat/e5-mistral-7b-instruct`
checkpoint and wrapped for local-first vector generation.

## Provenance

- upstream base model: `intfloat/e5-mistral-7b-instruct`
- packaging identity: `LumynaX` from `AbteeX AI Labs`
- weight claim: this package hydrates the official upstream checkpoint and does not claim a private weight merge

## Runtime Shape

- supported input modalities: `text`
- output modality: `embedding`
- local runtime: Python embedding stack
- capability focus: This package exposes instruction-aware dense embeddings for retrieval and ranking.

## Quick Start

```bash
pip install -r requirements.txt
python quickstart.py "LumynaX packages local models for retrieval."
```

## Included Files

- `merged_model/`: hydrated upstream model payload
- `quickstart.py`: local embedding runner
- `hf_space/`: bundled browser demo shell retained for release continuity
- `release_export_manifest.json`: standalone runtime metadata for this hydrated embedding package

## Publisher

- organization / lab: `AbteeX AI Labs`
- website: `https://abteex.com`
- Hugging Face owner account: `AbteeXAILab`
- recommended model repo: `AbteeXAILab/lumynax-embed-e5-mistral-7b`
- recommended local model name: `lumynax-embed-e5-mistral-7b`