Instructions to use daslab-testing/Apertus-8B-Instruct-MLX-INT6 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use daslab-testing/Apertus-8B-Instruct-MLX-INT6 with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("daslab-testing/Apertus-8B-Instruct-MLX-INT6")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

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

Notebooks
Google Colab
Kaggle
Local Apps
LM Studio

Pi new

How to use daslab-testing/Apertus-8B-Instruct-MLX-INT6 with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "daslab-testing/Apertus-8B-Instruct-MLX-INT6"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "daslab-testing/Apertus-8B-Instruct-MLX-INT6"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use daslab-testing/Apertus-8B-Instruct-MLX-INT6 with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "daslab-testing/Apertus-8B-Instruct-MLX-INT6"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default daslab-testing/Apertus-8B-Instruct-MLX-INT6

Run Hermes

hermes

MLX LM

How to use daslab-testing/Apertus-8B-Instruct-MLX-INT6 with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "daslab-testing/Apertus-8B-Instruct-MLX-INT6"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "daslab-testing/Apertus-8B-Instruct-MLX-INT6"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "daslab-testing/Apertus-8B-Instruct-MLX-INT6",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Apertus-8B-Instruct-MLX-INT6

File size: 446 Bytes

46db0e1

{
  "add_prefix_space": false,
  "backend": "tokenizers",
  "bos_token": "<s>",
  "clean_up_tokenization_spaces": false,
  "eos_token": "<|assistant_end|>",
  "is_local": true,
  "local_files_only": false,
  "model_input_names": [
    "input_ids",
    "attention_mask"
  ],
  "model_max_length": 1000000000000000019884624838656,
  "pad_token": "<pad>",
  "padding_side": "left",
  "tokenizer_class": "TokenizersBackend",
  "unk_token": "<unk>"
}