license: apache-2.0
library_name: mlx
base_model: treadon/granite-4.1-3b-Abliterated-AND-Disinhibited
tags:

• abliteration
• disinhibition
• granite
• mechanistic-interpretability
• mlx
  pipeline_tag: text-generation

granite-4.1-3b-Abliterated-AND-Disinhibited-mxfp8-mlx

Brainwaves

         arc   arc/e boolq hswag obkqa piqa  wino
mxfp8    0.405,0.598,0.843,0.520,0.442,0.713,0.582

Quant    Perplexity      Peak Memory   Tokens/sec
mxfp8   11.595 ± 0.129   6.58 GB       1594

Original model

granite-4.1-3b
mxfp8    0.406,0.581,0.821,0.484,0.434,0.712,0.559

Quant    Perplexity      Peak Memory   Tokens/sec
mxfp8   11.346 ± 0.127   6.58 GB       1690

This model granite-4.1-3b-Abliterated-AND-Disinhibited-mxfp8-mlx was
converted to MLX format from treadon/granite-4.1-3b-Abliterated-AND-Disinhibited
using mlx-lm version 0.31.3.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("granite-4.1-3b-Abliterated-AND-Disinhibited-mxfp8-mlx")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True, return_dict=False,
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)