Instructions to use aiplanet/LuxLlama with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use aiplanet/LuxLlama with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="aiplanet/LuxLlama")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("aiplanet/LuxLlama")
model = AutoModelForCausalLM.from_pretrained("aiplanet/LuxLlama")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use aiplanet/LuxLlama with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/aiplanet/LuxLlama

SGLang

How to use aiplanet/LuxLlama with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "aiplanet/LuxLlama" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aiplanet/LuxLlama",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "aiplanet/LuxLlama" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aiplanet/LuxLlama",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use aiplanet/LuxLlama 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 aiplanet/LuxLlama 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 aiplanet/LuxLlama to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for aiplanet/LuxLlama to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="aiplanet/LuxLlama",
    max_seq_length=2048,
)

Docker Model Runner
How to use aiplanet/LuxLlama with Docker Model Runner:
```
docker model run hf.co/aiplanet/LuxLlama
```

Request for LoRA Adapter Files (adapter_model.safetensors) for Cloudflare Deployment

by sourabhmnnit - opened Mar 7

Discussion

sourabhmnnit

Mar 7

Hello! First, thank you for the amazing work on LuxLlama.

I am trying to build an app to translate different languages into natural Luxembourgish. I am trying to deploy your model as the final grammar-refinement step using Cloudflare Workers AI (specifically their "Bring Your Own LoRA" feature).

Cloudflare's serverless architecture doesn't allow uploading the full merged model weights (the ~6GB .safetensors files currently in the repo). It only accepts the separated, lightweight LoRA adapter files (adapter_model.safetensors and adapter_config.json).

Would it be possible for you to upload the unmerged LoRA adapter weights to this repository? This would allow developers like me to easily plug your Luxembourgish model into Cloudflare's edge network.

Thank you so much for your time and for contributing to Luxembourgish AI!

sourabhmnnit

Mar 7

I tried to extract the LoRA myself using MergeKit, but it failed with a Tensor Size Mismatch error because the vocabulary sizes don't match. Could you please provide the original adapter files from your training run?

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment