Instructions to use nraesalmi/tinyllama-kalevala-chat with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nraesalmi/tinyllama-kalevala-chat with Transformers:

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

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

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("nraesalmi/tinyllama-kalevala-chat")
model = AutoModelForCausalLM.from_pretrained("nraesalmi/tinyllama-kalevala-chat", 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 nraesalmi/tinyllama-kalevala-chat with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/nraesalmi/tinyllama-kalevala-chat

SGLang

How to use nraesalmi/tinyllama-kalevala-chat 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 "nraesalmi/tinyllama-kalevala-chat" \
    --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": "nraesalmi/tinyllama-kalevala-chat",
		"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 "nraesalmi/tinyllama-kalevala-chat" \
        --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": "nraesalmi/tinyllama-kalevala-chat",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use nraesalmi/tinyllama-kalevala-chat with Docker Model Runner:
```
docker model run hf.co/nraesalmi/tinyllama-kalevala-chat
```

TinyLlama Kalevala Chat Model

This is a fine-tuned version of TinyLlama/TinyLlama-1.1B-Chat, adapted to answer questions about the Finnish national epic, the Kalevala. The model was fine-tuned using LoRA (Low-Rank Adaptation) on a curated dataset of prompt-response pairs derived from the Kalevala's characters, events, and symbolism.

🧠 Model Description

Base model: TinyLlama-1.1B-Chat
Parameter count: 1.1B
Fine-tuning method: LoRA using Hugging Face Transformers and PEFT
Dataset: nraesalmi/kalevala-dataset, consisting of 300 structured prompt-response pairs
Language: English (with content related to Finnish culture and mythology)

🗃️ Example Usage

from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

model = AutoModelForCausalLM.from_pretrained("nraesalmi/tinyllama-kalevala-chat")
tokenizer = AutoTokenizer.from_pretrained("nraesalmi/tinyllama-kalevala-chat")

chat = pipeline("text-generation", model=model, tokenizer=tokenizer)
chat("Tell me about the Sampo in the Kalevala:")

🧾 Example Output

"The Sampo is a magical artifact in the Kalevala, forged by the smith Ilmarinen. It brings prosperity by grinding out salt, grain, and gold. The struggle for control of the Sampo is a central theme in the epic."

📚 Intended Use

This model is intended for educational, cultural, and research purposes related to Finnish literature and mythology. It may also be useful for generating creative narratives or teaching materials about the Kalevala.

🛑 Limitations

The model may hallucinate or provide incorrect historical details outside its fine-tuned scope.

While aligned to be informative, it does not always follow factual sources beyond its training data.

📜 License

The model follows the same license as the original TinyLlama base model (Apache 2.0). The fine-tuning dataset is public domain or derived from freely accessible summaries and interpretations of the Kalevala.

✍️ Author

Fine-tuned and published by nraesalmi

Downloads last month: 4

Safetensors

Model size

1B params

Tensor type

F16

Model tree for nraesalmi/tinyllama-kalevala-chat

Base model

TinyLlama/TinyLlama-1.1B-Chat-v1.0

Finetuned

(580)

this model

Dataset used to train nraesalmi/tinyllama-kalevala-chat

Paper for nraesalmi/tinyllama-kalevala-chat

LoRA: Low-Rank Adaptation of Large Language Models

Paper • 2106.09685 • Published Jun 17, 2021 • 66