---
license: apache-2.0
base_model: google/gemma-4-E4B-it
pipeline_tag: text-generation
library_name: mlx
language:
  - en
tags:
  - medusa
  - gemma
  - mlx
  - 4bit
  - apple-silicon
  - local-inference
  - pocket-polymath
  - structured-reasoning
  - code-assistance
  - arithmetic
  - refusal
  - quantized
---

# MedusaGemma-E4B-MLX-4Bit

![MedusaGemma-E4B-MLX-4Bit](medusafinal.png)

**Tagline:** One spine. Many heads.

> Started as a MacBook Pro experiment and grew into a full composite release.

## Model Summary

MedusaGemma-E4B-MLX-4Bit is the compact Apple Silicon build of MedusaGemma-E4B: a fused local model for offline inference, reasoning, and practical work.

## What Went Into It

This is the 4-bit MLX release for Apple Silicon local use. The donor lineage and adaptation work are baked in so this repo is a single downloadable model artifact.

Primary lineage:

- `google/gemma-4-E4B-it`
- `DavidAU/gemma-4-E4B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking`
- `DavidAU/gemma-4-E4B-it-The-DECKARD-V2-Strong-HERETIC-UNCENSORED-Thinking`
- `Jiunsong/supergemma4-e4b-abliterated`

## Stamsam custom dojo / training

The adaptation stack was built on:

- `pocket-polymath-seam-adapter`
- `pocket-polymath-legacy-adapter`
- `pocket-polymath-legacy-4096-adapter`
- `pocket-polymath-broad-growth-adapter`
- `pocket-polymath-ultimate-dojo-adapter`

Synthetic dojo data came from `glm-5.1:cloud` via Ollama, plus Gemini-generated material from the same workflow.

## Quantization Notes

- Format: MLX
- Quantization: 4-bit
- Target hardware: Apple Silicon

## Best For

- fast Apple Silicon local inference
- offline assistant work
- structured reasoning and business analysis
- general writing and summarization
- practical code generation
- integration with the Pocket Polymath stack

## Keep In Mind

- The bare MLX build can still make arithmetic mistakes without the stack.
- The bare MLX build can still miss exact refusal behavior without the stack.
- The bare MLX build can still miss strict formatting constraints without validators.

## Evaluation Snapshot

- Fresh full holdout: 40/40
- External-lite cleaned: 256/265, 96.6%
- MBPP repair mode: 20/20 final on the cleaned slice and fresh holdout
- Reasoning leaks: 0
- Bad claims: 0

## Safety and Reliability

The Pocket Polymath stack adds:

- reasoning leak guard
- missing-data refusal router
- arithmetic router
- structured-output validators
- completion checks
- optional sandboxed code verifier/repair mode

## Release Names

- Raw model: `MedusaGemma-E4B`
- MLX 4-bit: `MedusaGemma-E4B-MLX-Q4`
- Default stack: `Pocket Polymath: MedusaGemma-E4B RC2`
- Optional code mode: `Pocket Polymath: MedusaGemma-E4B RC2.1 Code Repair`