from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

model_path = "supersamdev/Gemma-4-E4B-dailydialog-MLX-8bit"
model, processor = load(model_path)
config = load_config(model_path)

prompt = apply_chat_template(processor, config, "Hello!", num_images=0)
response = generate(model, processor, prompt, max_tokens=200, verbose=True)

Other Quants

Repo	Bits	RAM
Gemma-4-E4B-dailydialog-MLX-2bit	2-bit	~3 GB
Gemma-4-E4B-dailydialog-MLX-4bit	4-bit	~6 GB
Gemma-4-E4B-dailydialog-MLX-6bit	6-bit	~8 GB
Gemma-4-E4B-dailydialog-MLX-8bit	8-bit	~10 GB

Parent Model

supersamdev/Gemma-4-E4B-dailydialog-16bit

Quantization

Converted using mlx-vlm on Apple Silicon.