Instructions to use tabularisai/Faust-1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tabularisai/Faust-1 with Transformers:

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

pipe = pipeline("text-generation", model="tabularisai/Faust-1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("tabularisai/Faust-1")
model = AutoModelForCausalLM.from_pretrained("tabularisai/Faust-1", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use tabularisai/Faust-1 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 tabularisai/Faust-1:Q8_0
# Run inference directly in the terminal:
llama cli -hf tabularisai/Faust-1:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf tabularisai/Faust-1:Q8_0
# Run inference directly in the terminal:
llama cli -hf tabularisai/Faust-1:Q8_0

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 tabularisai/Faust-1:Q8_0
# Run inference directly in the terminal:
./build/bin/llama cli -hf tabularisai/Faust-1:Q8_0

Use Docker

docker model run hf.co/tabularisai/Faust-1:Q8_0

Llama new
LM Studio
Jan

vLLM

How to use tabularisai/Faust-1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tabularisai/Faust-1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tabularisai/Faust-1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tabularisai/Faust-1:Q8_0

SGLang

How to use tabularisai/Faust-1 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 "tabularisai/Faust-1" \
    --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": "tabularisai/Faust-1",
		"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 "tabularisai/Faust-1" \
        --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": "tabularisai/Faust-1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use tabularisai/Faust-1 with Ollama:
```
ollama run hf.co/tabularisai/Faust-1:Q8_0
```
Unsloth Desktop
Docker Model Runner
How to use tabularisai/Faust-1 with Docker Model Runner:
```
docker model run hf.co/tabularisai/Faust-1:Q8_0
```

Lemonade

How to use tabularisai/Faust-1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull tabularisai/Faust-1:Q8_0

Run and chat with the model

lemonade run user.Faust-1-Q8_0

List all available models

lemonade list

Atomic Chat

vdmbrsv commited on Jan 22

Commit

8c93af4

verified ·

1 Parent(s): b2ff951

Update README.md

Browse files

Files changed (1) hide show

README.md +4 -30

README.md CHANGED Viewed

@@ -86,24 +86,6 @@ The project is motivated by language sovereignty and practical deployment constr
 ---
-## What this repository provides
-This repository contains a complete conversational checkpoint and tokenizer, ready to be used with standard Hugging Face tooling and local inference runtimes.
-Included artifacts:
-- model.safetensors – main FP16 model weights
-- tokenizer.json, tokenizer_config.json, special_tokens_map.json – German-optimized tokenizer
-- chat_template.jinja – canonical chat template for conversational use
-- generation_config.json – default generation settings
-- faust-1-dpo-golden-v1-1601-q8_0.gguf – quantized GGUF build for llama.cpp-style runtimes
-- faust_bench.png, tokenizer_bench.png – evaluation figures
-- logo-faust.webp – project logo
-For conversational behavior, users should rely on the provided chat template rather than ad-hoc role formatting.
----
 ## Quickstart
 ### Conversational usage (recommended)
@@ -219,19 +201,10 @@ The repository includes a prebuilt Q8_0 GGUF file for efficient local inference.
 ---
-## Limitations
-- Optimized primarily for German; other languages are secondary.
-- Not tuned for structured tool calling or agent frameworks.
-- Small-model limitations apply for very long-horizon reasoning.
-- Non-commercial license restricts usage.
----
 ## Roadmap
-- Reasoning-focused variant
-- Agent-oriented variant
 ---
@@ -241,4 +214,5 @@ A technical paper describing training methodology, tokenizer design, and evaluat
 If you use Faust-1 in academic work, please cite this repository and reference the forthcoming paper when available.
-Developed by tabularis.ai.

 ---
 ## Quickstart
 ### Conversational usage (recommended)
 ---
 ## Roadmap
+- Reasoning-focused variant  (comming soon)
+- Agent-oriented variant  (comming soon)
 ---
 If you use Faust-1 in academic work, please cite this repository and reference the forthcoming paper when available.
+Developed by [tabularis.ai](https://tabularis.ai) in Tübingen.