Instructions to use kinmello/qbot-9b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kinmello/qbot-9b 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("kinmello/qbot-9b")

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

llama.cpp

How to use kinmello/qbot-9b with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf kinmello/qbot-9b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf kinmello/qbot-9b:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf kinmello/qbot-9b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf kinmello/qbot-9b:Q4_K_M

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf kinmello/qbot-9b:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf kinmello/qbot-9b:Q4_K_M

Use Docker

docker model run hf.co/kinmello/qbot-9b:Q4_K_M

Llama new
LM Studio
Jan

vLLM

How to use kinmello/qbot-9b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kinmello/qbot-9b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kinmello/qbot-9b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/kinmello/qbot-9b:Q4_K_M

Ollama
How to use kinmello/qbot-9b with Ollama:
```
ollama run hf.co/kinmello/qbot-9b:Q4_K_M
```
Unsloth Desktop

How to use kinmello/qbot-9b with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "kinmello/qbot-9b"

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": "kinmello/qbot-9b"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

MLX LM

How to use kinmello/qbot-9b with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "kinmello/qbot-9b"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "kinmello/qbot-9b"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "kinmello/qbot-9b",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Docker Model Runner
How to use kinmello/qbot-9b with Docker Model Runner:
```
docker model run hf.co/kinmello/qbot-9b:Q4_K_M
```

Lemonade

How to use kinmello/qbot-9b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull kinmello/qbot-9b:Q4_K_M

Run and chat with the model

lemonade run user.qbot-9b-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use kinmello/qbot-9b 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 "kinmello/qbot-9b"

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 kinmello/qbot-9b

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use kinmello/qbot-9b with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "kinmello/qbot-9b"

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 "kinmello/qbot-9b" \
  --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"

QBot 9B (beta)

A 9B assistant for Qlik Sense / Qlik Cloud in English and Brazilian Portuguese, fine-tuned from Qwen3.5-9B by distillation (SFT + rejection-sampling fine-tuning on attempts verified by the Qlik Engine). It writes load scripts and data models, answers function and syntax questions, and works as an agent on Qlik Cloud through the Qlik MCP tools (create apps, fix scripts, reload, build sheets).

This repository hosts the model for the QBot desktop app (beta), which downloads the right build for the computer. manifest.json lists every model version, its builds and the app versions they work with.

Build	Engine	File(s)	Suggested memory
`v1.0.0/mlx-8bit`	MLX (Apple silicon)	safetensors	24 GB
`v1.0.0/mlx-4bit`	MLX (Apple silicon)	safetensors	16 GB
`v1.0.0/gguf/qbot-q8_0.gguf`	llama.cpp, Ollama, LM Studio	GGUF	16 GB VRAM/RAM
`v1.0.0/gguf/qbot-q4_k_m.gguf`	llama.cpp, Ollama, LM Studio	GGUF	10 GB VRAM/RAM

Results (QBot app, Mac M5 Pro 24 GB, MLX 8-bit)

Live tasks on a real Qlik Cloud tenant, graded by the Qlik Engine (0–10):

live48 (48 tasks: star schemas, crosstab, dirty dates, multi-file, dashboards, fixes): 8.93 (38 passed). Teacher Qwen3.5-397B (thinking): 7.76.
App-fixing sample (10 tasks) with the app's automatic thinking mode: 10/10.

4-bit builds score lower on these tasks (live48 5.02 for MLX 4-bit in the app tests); prefer 8-bit when memory allows.

How to use

Sampling used in all evaluations: temperature 0, repetition penalty 1.05 over the last 64 tokens.
No thinking for questions and building apps; thinking (enable_thinking) with the reasoning kept between steps for fixing apps and scripts.
System prompt used in training: see system_prompt.txt.

Limitations

Beta. It can still be wrong about rare functions and syntax details; check important code.
Agent use needs the Qlik MCP tool definitions (97 tools); the app provides them.

QBot is an independent project and is not affiliated with or endorsed by Qlik. Qlik and Qlik Sense are trademarks of QlikTech International AB.

License: Apache 2.0, as the base model.

Downloads last month: 89

MLX

Hardware compatibility

Quantized

GGUF

Model size

9B params

Architecture

qwen35

Hardware compatibility

4-bit

8-bit

Model tree for kinmello/qbot-9b

Base model

Qwen/Qwen3.5-9B-Base

Finetuned

Qwen/Qwen3.5-9B

Adapter

(1,082)

this model