Gemma 4 E4B Instruct Parity GGUF (Q8_0)

This repository contains the canonical GGUF side of the meshllm Gemma 4 parity pair.

  • Source checkpoint: google/gemma-4-E4B-it
  • Conversion path: original checkpoint -> GGUF -> Q8_0
  • Intended use: backend parity testing against the matching MLX artifact

This artifact is not meant to be a general "best available" Gemma 4 release. It exists so that GGUF and MLX can be compared from the same original model lineage with minimal third-party conversion noise.

Canonical pair:

  • GGUF: meshllm/gemma-4-e4b-it-parity-q8_0-gguf
  • MLX: meshllm/gemma-4-e4b-it-parity-8bit-mlx

Latest trusted exact result:

Backend Model Exact
GGUF gemma-4-e4b-it-q8_0.gguf PASS
MLX meshllm/gemma-4-e4b-it-parity-8bit-mlx PASS

Prompt comparison from local same-origin validation:

Prompt GGUF Q8_0 MLX 8bit
primary blue blue
alt-green green green
alt-red red red
capital-france Paris Paris
primary-colors red, green, blue red, green, blue
two-plus-two 4 4
largest-planet Jupiter Jupiter
breathing-gas Oxygen Oxygen
opposite-hot Cold Cold
banana-color Yellow Yellow
after-monday Tuesday Tuesday
Downloads last month
171
GGUF
Model size
8B params
Architecture
gemma4
Hardware compatibility
Log In to add your hardware

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for meshllm/gemma-4-e4b-it-parity-q8_0-gguf

Quantized
(111)
this model