north-mini-code-mxfp4-mlx

MLX quantization of CohereLabs/North-Mini-Code-1.0 for Apple Silicon.

Variant: Block float MX FP4
Disk size: 16777 MB
Quantized by: sahilchachra

Benchmark results

Evaluated on Apple M4 Pro with MLX. Model loaded once; performance and quality measured in a single pass.

Performance

	This model	FP16 baseline
Decode tok/s (avg, long traces)	59.23	N/A
Peak memory (GB)	18.271	N/A
Disk size (MB)	16777	58175

Quality

Benchmark	This model	FP16 baseline	n
MATH-500 (math reasoning)	96.7% (answered 30/30)	N/A	30
IFEval (instruction following)	38.6%	N/A	44
LiveCodeBench v6 (code, pass@1)	56.7%	N/A	30
HumanEval (code, pass@1)	83.3%	N/A	30

MATH-500 per-level accuracy

Level	This model	FP16 baseline
level 1	100.0%	N/A
level 2	100.0%	N/A
level 3	83.3%	N/A
level 4	100.0%	N/A
level 5	100.0%	N/A

Context scaling (decode tok/s)

Context length	Decode tok/s
~128 tokens	62.8
~256 tokens	62.4
~512 tokens	62.4
~1024 tokens	61.6

Usage

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("sahilchachra/north-mini-code-mxfp4-mlx")
response = generate(model, tokenizer, prompt="Your prompt here", max_tokens=256, verbose=True)

All variants in this collection

Model	Variant
sahilchachra/north-mini-code-mxfp4-mlx	Block float MX FP4 ← this model

Notes

• Requires Apple Silicon (M1 or later) with MLX
• Benchmarks run on Apple M4 Pro, 24 GB unified memory
• License: see CohereLabs/North-Mini-Code-1.0 for the original model's license

Original model

See CohereLabs/North-Mini-Code-1.0 for full model details and intended use.