Downloads · 30 days
0
gemma-4-E4B-it-vindex
rewintous/gemma-4-E4B-it-vindex
gemma-4-E4B-it-vindex is a machine learning model from rewintous. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
--- basemodel: google/gemma-4-E4B-it tags: - larql - vindex - gemma - knowledge-graph - no-gpu language: - en license: apache-2.0 ---
Access
Public
Updated Apr 15, 2026
Repo size
18.6 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin21.2 GB · 100%
At a glance
- License
- apache-2.0
- Access
- Public
- Created
- Apr 14, 2026
- Updated
- Apr 15, 2026
- SHA
- 9ec09520
Model card
From the Hugging Face model README
base_model: google/gemma-4-E4B-it tags:
- larql
- vindex
- gemma
- knowledge-graph
- no-gpu language:
- en license: apache-2.0
gemma-4-E4B-it · LARQL Vindex
Weights from google/gemma-4-E4B-it extracted into a vindex using LARQL — query transformer weights like a graph database. No GPU required.
Vindex Contents
| File | Description |
|---|---|
gate_vectors.bin | W_gate rows — KNN index |
embeddings.bin | W_embed matrix — token lookup |
down_meta.bin | Per-feature output metadata |
index.json | Config, layer bands, provenance |
tokenizer.json | Tokenizer |
relation_clusters.json | Discovered relation types |
feature_labels.json | Probe-confirmed labels |
Extraction Levels
| Level | Size (f16) | Capabilities |
|---|---|---|
| Browse (this repo) | ~3 GB | DESCRIBE, WALK, SELECT |
| Inference | ~6 GB | + INFER |
| All | ~10 GB | + COMPILE |
Links
LARQL · LQL Spec · LQL Guide · Base Model
License
LARQL: Apache-2.0 · Model weights: Gemma Terms of Use
- License
- apache-2.0
- Created
- Apr 14, 2026
- Updated
- Apr 15, 2026