gpt-oss-20b-RotorQuant-MLX-8bit / README.md

chore(card): enrich YAML frontmatter (pipeline_tag, language, library_name, inference)

8c881d1 verified about 2 hours ago

3.64 kB

base_model: openai/gpt-oss-20b
library_name: mlx
tags:
  - rotorquant
  - kv-cache-quantization
  - gpt-oss
  - openai
  - moe
  - quantized
  - mlx
  - 8bit
license: apache-2.0
pipeline_tag: text-generation
language:
  - en

GPT-OSS-20B - RotorQuant MLX 8-bit

8-bit weight-quantized MLX version of openai/gpt-oss-20b with RotorQuant KV-cache quantization. Optimized for Apple Silicon inference via the MLX framework. RotorQuant delivers 5.3x faster prefill and 28% faster decode compared to TurboQuant. GPT-OSS-20B is OpenAI's first open-weights release in years (Apache 2.0), a Mixture-of-Experts model that rivals o3-mini on reasoning benchmarks.

Approximate model size: ~20 GB

Model Specifications

Property	Value
Base Model	openai/gpt-oss-20b
Parameters	20 billion (MoE)
Architecture	Mixture-of-Experts (MoE) Transformer
License	Apache 2.0 (commercial use OK)
Weight Quantization	8-bit (~20 GB)
KV-Cache Quantization	RotorQuant
Framework	MLX (Apple Silicon)

Quickstart

from mlx_lm import load, generate
from rotorquant import IsoQuantCache

model, tokenizer = load("majentik/gpt-oss-20b-RotorQuant-MLX-8bit")

prompt = "Explain the theory of relativity."
response = generate(model, tokenizer, prompt=prompt, max_tokens=512)
print(response)

What is RotorQuant?

RotorQuant applies block-diagonal rotations (Clifford algebra) for KV cache compression. Combined with 8-bit weight quantization in MLX, this provides a dual compression strategy with superior KV-cache performance.

Key advantages over TurboQuant:

5.3x faster prefill
28% faster decode
Equivalent memory savings

KV-Cache Quantization Comparison

Method	Prefill Speed	Decode Speed	Memory Savings	Reference
TurboQuant	1x (baseline)	1x (baseline)	High	arXiv: 2504.19874
RotorQuant	5.3x faster	28% faster	High	GitHub

Memory Estimates (GPT-OSS-20B)

Precision	Approximate Size	MLX Variant
BF16 (original)	~40 GB	--
8-bit quantized	~20 GB	This model
4-bit quantized	~12 GB	RotorQuant-MLX-4bit
2-bit quantized	~6 GB	RotorQuant-MLX-2bit

Hardware Requirements

This model requires approximately 20 GB of unified memory. Recommended hardware:

Apple M2 Max (32 GB+)
Apple M3 Max (32 GB+)
Apple M4 Max (32 GB+)
Any Apple Silicon Mac with 32 GB+ unified memory

majentik
/

gpt-oss-20b-RotorQuant-MLX-8bit