Instructions to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF 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 ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Use Docker

docker model run hf.co/ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ndgold/Qwen3-0.6B-EasyLanguage-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ndgold/Qwen3-0.6B-EasyLanguage-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Ollama
How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with Ollama:
```
ollama run hf.co/ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M
```

Unsloth Studio

How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for ndgold/Qwen3-0.6B-EasyLanguage-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for ndgold/Qwen3-0.6B-EasyLanguage-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ndgold/Qwen3-0.6B-EasyLanguage-GGUF to start chatting

How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with Docker Model Runner:
```
docker model run hf.co/ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M
```

Lemonade

How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3-0.6B-EasyLanguage-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

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 ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ndgold/Qwen3-0.6B-EasyLanguage-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M

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 "ndgold/Qwen3-0.6B-EasyLanguage-GGUF:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

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

Qwen3 0.6B — Easy Language (Q4_K_M GGUF)

A LoRA fine-tune of Qwen/Qwen3-0.6B that rewrites spoken sentences into the Easy Language register of the same language — FALC in French, Leichte Sprache in German, Lectura Fácil in Spanish, and the Inclusion Europe "Easy-to-Read" standard elsewhere.

It is a same-language simplification model, not a translation model. Given French in, it produces simpler French out. It never translates.

Built for Live Linguist, an on-device live captioning app. Everything runs locally; this model exists so that simplification never requires a server.

Prompt contract — this is not optional

The model was trained against one exact rendering, and deviating from it degrades output silently. There is no error, just worse rewrites.

<|im_start|>system
{SYSTEM_PROMPT}<|im_end|>
<|im_start|>user
Original: {SENTENCE}
Rewritten:<|im_end|>
<|im_start|>assistant
<think>

</think>

Three things people get wrong:

The assistant turn is pre-filled with an EMPTY think block — literally <think>\n\n</think>\n\n. This is Qwen3's enable_thinking=false form. Omit it and the model may emit reasoning into your caption.
{SYSTEM_PROMPT} carries a trailing /no_think soft switch. It belongs to the prompt, not the template.
No few-shot examples, and no rolling context. The fine-tune internalised the register, so it was trained on the bare Original:/Rewritten: turn. Injecting examples or prior segments is out-of-distribution and bleeds context into the rewrite.

The system prompts are the per-language framework packs in the app repo under core/src/main/resources/framework_packs/.

Sampling — fixed, and evaluated under these values

parameter	value
temperature	0 (greedy)
repeat_penalty	1.3
repeat_last_n	20
n_predict	128

These are not tuning knobs. The gate results below were produced under exactly these settings.

Evaluation

20-prompt validator-clean rate across 12 languages, scored by the same validators the app uses at runtime (sentence length, one-idea-per-sentence, no invented content, still-in-source-language):

95.00% (19/20) against a 94.50% MLX baseline.

Quantisation

Q4_K_M, which this model tolerates: it scores 95.00% against a 94.50% baseline. The larger 1.7B sibling does not tolerate Q4_K_M and ships at Q5_K_M instead — a quantisation level is a per-model measurement, not a global default.

Limitations

Twenty prompts is a screen, not a benchmark. At n=20 the binomial interval is roughly ±10 pp. This catches gross breakage; it does not establish parity of quality.
Grammar is imperfect at these sizes. A real 0.6B output was "La semaine dernière, nous avons partis à la médina" — the register is right (disfluency removed, run-on split) and the auxiliary is wrong (sommes, not avons).
Simplified text is a paraphrase. The app labels it "~ simplified — may not be exact" for this reason. It should not be relied on where exact wording is legally or medically material.
No real-hardware latency measurement exists yet. The 0.6B has not been benchmarked on a phone.

Licence and attribution

Apache-2.0, inherited from the base model.

Qwen3 by Alibaba Cloud — Qwen/Qwen3-0.6B
Easy Language fine-tune by ndgold

Provenance

LoRA: rank 16, alpha 320 (scale 20.0 × rank), dropout 0.05, on q_proj/k_proj/v_proj/o_proj of the top 16 layers only; max sequence 1024; 1 epoch.
Trained locally on an RTX 5060 Ti (sm_120).
SHA-256 of this artifact: 1be8c19cc1153d33535c3ad6a3306be7d84ecc4ae7a171bf6c1b428c6737ffb8

The app pins this hash and verifies it after download; a mismatch aborts the install.

Downloads last month: 106

GGUF

Model size

0.6B params

Architecture

qwen3

Hardware compatibility

4-bit

Model tree for ndgold/Qwen3-0.6B-EasyLanguage-GGUF

Base model

Qwen/Qwen3-0.6B-Base

Finetuned

Qwen/Qwen3-0.6B

Quantized

(397)

this model

ndgold
/

Qwen3-0.6B-EasyLanguage-GGUF

Qwen3 0.6B — Easy Language (Q4_K_M GGUF)

Prompt contract — this is not optional

Sampling — fixed, and evaluated under these values

Evaluation

Quantisation

Limitations

Licence and attribution

Provenance

Model tree for ndgold/Qwen3-0.6B-EasyLanguage-GGUF

Dataset used to train ndgold/Qwen3-0.6B-EasyLanguage-GGUF