Instructions to use noricha/Qwen2.5-1.5B-Instruct-MLX-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use noricha/Qwen2.5-1.5B-Instruct-MLX-4bit 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("noricha/Qwen2.5-1.5B-Instruct-MLX-4bit")

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 Settings
LM Studio

How to use noricha/Qwen2.5-1.5B-Instruct-MLX-4bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

MLX LM

How to use noricha/Qwen2.5-1.5B-Instruct-MLX-4bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Hermes Agent

How to use noricha/Qwen2.5-1.5B-Instruct-MLX-4bit 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 "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit"

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 noricha/Qwen2.5-1.5B-Instruct-MLX-4bit

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use noricha/Qwen2.5-1.5B-Instruct-MLX-4bit with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "noricha/Qwen2.5-1.5B-Instruct-MLX-4bit" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Qwen2.5-1.5B-Instruct-MLX-4bit

This model is an MLX 4-bit quantized conversion of Qwen/Qwen2.5-1.5B-Instruct, produced with mlx-lm.

Base model

Base: Qwen/Qwen2.5-1.5B-Instruct
License: Apache-2.0
Context length: 32,768 tokens
Generation length: 8,192 tokens
Quantization: 4-bit (group size 64, mixed 4.5 bpw)

Use with MLX-LM

pip install -U mlx-lm

mlx_lm.generate \
  --model noricha/Qwen2.5-1.5B-Instruct-MLX-4bit \
  --system-prompt "あなたは日本語で簡潔に答えるアシスタントです。" \
  --prompt "MLXモデルとは何ですか？" \
  --max-tokens 200 --temp 0.7 --top-p 0.9

Use with LM Studio

LM Studio 0.3.4 以降の MLX Engine で読み込み可能です。アプリ内検索または CLI:

lms get noricha/Qwen2.5-1.5B-Instruct-MLX-4bit

Notes

学習目的で作成した MLX 変換モデルです。用途・制限は元モデルの Qwen2.5-1.5B-Instruct を参照してください。
0.5B 版より日本語の一般知識・自然さが改善しています（M3 Ultra 実測: 生成約 307 tok/s, Peak 0.98 GB）。
反復抑制のため --temp 0.7 --top-p 0.9 などのサンプリングと --system-prompt の併用を推奨します。

Downloads last month: 6

Safetensors

Model size

2B params

Tensor type

U32

BF16

MLX

Hardware compatibility

4-bit

Model tree for noricha/Qwen2.5-1.5B-Instruct-MLX-4bit

Base model

Qwen/Qwen2.5-1.5B

Finetuned

Qwen/Qwen2.5-1.5B-Instruct

Quantized

(371)

this model