Instructions to use GloathLtd/Seddon1.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GloathLtd/Seddon1.1 with Transformers:

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

pipe = pipeline("text-generation", model="GloathLtd/Seddon1.1")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("GloathLtd/Seddon1.1", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use GloathLtd/Seddon1.1 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 GloathLtd/Seddon1.1:Q8_0
# Run inference directly in the terminal:
llama cli -hf GloathLtd/Seddon1.1:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf GloathLtd/Seddon1.1:Q8_0
# Run inference directly in the terminal:
llama cli -hf GloathLtd/Seddon1.1:Q8_0

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 GloathLtd/Seddon1.1:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf GloathLtd/Seddon1.1:Q8_0

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 GloathLtd/Seddon1.1:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf GloathLtd/Seddon1.1:Q8_0

Use Docker

docker model run hf.co/GloathLtd/Seddon1.1:Q8_0

LM Studio
Jan

vLLM

How to use GloathLtd/Seddon1.1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "GloathLtd/Seddon1.1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GloathLtd/Seddon1.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/GloathLtd/Seddon1.1:Q8_0

SGLang

How to use GloathLtd/Seddon1.1 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 "GloathLtd/Seddon1.1" \
    --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": "GloathLtd/Seddon1.1",
		"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 "GloathLtd/Seddon1.1" \
        --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": "GloathLtd/Seddon1.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use GloathLtd/Seddon1.1 with Ollama:
```
ollama run hf.co/GloathLtd/Seddon1.1:Q8_0
```
Unsloth Desktop
Docker Model Runner
How to use GloathLtd/Seddon1.1 with Docker Model Runner:
```
docker model run hf.co/GloathLtd/Seddon1.1:Q8_0
```

Lemonade

How to use GloathLtd/Seddon1.1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull GloathLtd/Seddon1.1:Q8_0

Run and chat with the model

lemonade run user.Seddon1.1-Q8_0

List all available models

lemonade list

Atomic Chat

Seddon1.1

Seddon1.1 is a small chat and coding assistant, trained in New Zealand by Gloath Limited.

Try it for free at https://ai.gloath.com

This is a research preview. At 128M parameters it is a small model: it can hold a simple conversation and write short pieces of code, but it often gets facts wrong, loses the thread in longer answers, and shouldn't be relied on for anything that matters.

Seddon1.1 uses the standard Qwen3 architecture, so it runs anywhere Qwen3 does. Its weights were trained from scratch by Gloath Limited; they are not based on Qwen or any other model.

Model details


Developer	Gloath Limited, New Zealand
Model type	Causal language model tuned for chat
Architecture	Transformer decoder in the standard Qwen3 layout: 16 layers, hidden size 768, 12 attention heads (4 key-value heads), SwiGLU, RMSNorm, rotary positions, tied embeddings
Parameters	128M
Files	0.26 GB (bfloat16 safetensors)
Context length	2,048 tokens
Languages	English, and code in common programming languages
Released	October 2026
Licence	Apache License 2.0 (see LICENSE)

Quick start

With transformers 4.56.0 or newer:

from transformers import AutoModelForCausalLM, AutoTokenizer

name = "GloathLtd/Seddon1.1"
tokenizer = AutoTokenizer.from_pretrained(name)
model = AutoModelForCausalLM.from_pretrained(name, dtype="auto")

messages = [{"role": "user", "content": "Kia ora! Who are you?"}]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt",
                                       return_dict=True)
output = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(output[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))

The recommended settings (temperature 0.7, top-p 0.9, top-k 50, repetition penalty 1.1) are in generation_config.json, so generate uses them by default.

For LM Studio, Ollama and llama.cpp, use the GGUF files in GloathAI1-GGUF.

Chat format

Seddon1.1 uses its own chat format, which is built into the tokenizer's chat template and the GGUF files, so chat apps apply it for you:

<|endoftext|><|system|>You are a helpful assistant.<|end|><|user|>Kia ora!<|end|><|assistant|>

Each message is its role marker, its text and <|end|>, with no spaces or newlines in between. The model ends its reply with <|end|>.

Intended use and limitations

Good for: trying out a small assistant, learning, research, fine-tuning experiments, and running entirely on your own computer.
Not for: medical, legal, financial or safety advice, or anything where a mistake matters.
It can produce wrong, biased or offensive text. Check its answers, and test any code before using it.
English only: other languages will be poor.

Evaluation

Validation loss at the end of training (cross-entropy per token; lower is better):


Web text (C4)	3.230
Code	1.230
Conversations	2.150

Licence

Seddon1.1 is released under the Apache License 2.0: you may use, modify and share it, including commercially, as long as you include the licence and the NOTICE file. It gives no rights to the names and trade marks of Gloath Limited, such as Gloath and GloathAI and Seddon.

Contact

ai.gloath.com m1@ai.gloath.com

Downloads last month: -

Safetensors

Model size

0.1B params

Tensor type

BF16