Instructions to use albertic-ai/Laika-1-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use albertic-ai/Laika-1-3B with Transformers:

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

pipe = pipeline("text-generation", model="albertic-ai/Laika-1-3B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("albertic-ai/Laika-1-3B")
model = AutoModelForCausalLM.from_pretrained("albertic-ai/Laika-1-3B", 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use albertic-ai/Laika-1-3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "albertic-ai/Laika-1-3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "albertic-ai/Laika-1-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/albertic-ai/Laika-1-3B

SGLang

How to use albertic-ai/Laika-1-3B 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 "albertic-ai/Laika-1-3B" \
    --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": "albertic-ai/Laika-1-3B",
		"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 "albertic-ai/Laika-1-3B" \
        --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": "albertic-ai/Laika-1-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use albertic-ai/Laika-1-3B with Docker Model Runner:
```
docker model run hf.co/albertic-ai/Laika-1-3B
```

Laika-1-3B

Laika-1-3B is a 3 billion parameter language model specialized for space science, astronomy, satellite operations, and aerospace engineering. Named after Laika, the first animal to orbit Earth, this model is designed to assist with space-related tasks and knowledge.

Model Details

Parameter	Value
Architecture	LlamaForCausalLM
Parameters	3B
Hidden Size	3072
Intermediate Size	8192
Attention Heads	24
Hidden Layers	28
Vocab Size	32,000
Context Length	4,096
Precision	bfloat16

Intended Use

Laika-1-3B is designed for:

Space Science Q&A - Answering questions about astronomy, astrophysics, cosmology, and planetary science
Satellite Operations - Assisting with satellite communication, orbit mechanics, and mission planning
Aerospace Engineering - Supporting aerospace design concepts and technical discussions
Space Education - Explaining space-related concepts for educational purposes
Research Assistance - Helping with space science research and literature review

How to Use

Using Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "albertic-ai/Laika-1-3B"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")

messages = [
    {"role": "system", "content": "You are Laika, an AI assistant specialized in space science, astronomy, satellite operations, and aerospace engineering."},
    {"role": "user", "content": "Explain the concept of Lagrange points and their significance for space missions."}
]

input_ids = tokenizer.apply_chat_template(messages, return_tensors="pt", add_generation_prompt=True).to(model.device)

outputs = model.generate(
    input_ids,
    max_new_tokens=512,
    temperature=0.7,
    top_p=0.9,
    do_sample=True,
)

response = tokenizer.decode(outputs[0][input_ids.shape[-1]:], skip_special_tokens=True)
print(response)

Using Pipeline

from transformers import pipeline

pipe = pipeline(
    "text-generation",
    model="albertic-ai/Laika-1-3B",
    torch_dtype="auto",
    device_map="auto",
)

messages = [
    {"role": "user", "content": "What are the main challenges of interplanetary travel?"}
]

output = pipe(messages, max_new_tokens=512)
print(output[0]["generated_text"][-1]["content"])

Chat Template

Laika-1-3B uses the following chat format:

<s>### System:
You are Laika, an AI assistant specialized in space science, astronomy, satellite operations, and aerospace engineering.

### User:
{user_message}

### Assistant:
{assistant_response}</s>

Training

Laika-1-3B was trained on a curated dataset of space science literature, including:

Astronomical research papers and publications
Satellite operations documentation
Aerospace engineering textbooks and materials
Space mission reports and technical documents
Educational astronomy content

Limitations

The model may generate plausible-sounding but incorrect information about specific space missions or technical details
It should not be used as a sole source for mission-critical decisions
Knowledge is limited to training data cutoff
May struggle with highly specialized or niche topics outside its training distribution
Not suitable for real-time satellite operations without human oversight

License

This model is released under the Apache 2.0 License.

Citation

@misc{laika-1-3b,
  title={Laika-1-3B: A Language Model for Space Science},
  author={Albertic AI},
  year={2024},
  url={https://huggingface.co/albertic-ai/Laika-1-3B}
}

albertic-ai
/

Laika-1-3B