Instructions to use LoneWolfgang/tinyllama-for-abalone-RAG with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LoneWolfgang/tinyllama-for-abalone-RAG with Transformers:

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

pipe = pipeline("text-generation", model="LoneWolfgang/tinyllama-for-abalone-RAG")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("LoneWolfgang/tinyllama-for-abalone-RAG")
model = AutoModelForCausalLM.from_pretrained("LoneWolfgang/tinyllama-for-abalone-RAG")
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 LoneWolfgang/tinyllama-for-abalone-RAG with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/LoneWolfgang/tinyllama-for-abalone-RAG

SGLang

How to use LoneWolfgang/tinyllama-for-abalone-RAG 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 "LoneWolfgang/tinyllama-for-abalone-RAG" \
    --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": "LoneWolfgang/tinyllama-for-abalone-RAG",
		"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 "LoneWolfgang/tinyllama-for-abalone-RAG" \
        --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": "LoneWolfgang/tinyllama-for-abalone-RAG",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use LoneWolfgang/tinyllama-for-abalone-RAG with Docker Model Runner:
```
docker model run hf.co/LoneWolfgang/tinyllama-for-abalone-RAG
```

TinyLlama for Abalone RAG

This is a toy model fine-tuned specifically for the Abalone RAG App. A dataset was developed to finetune this model:

Questions were suggested by GPT-5.
Passages were retrieved from a small index created for the app.
Responses were generated by prompting GPT-5 with the query, the retrieved context, and instructions to produce concise answers referencing the context.

TinyLlama was finetuned on this dataset, resulting in a dedicated TinyLlama model for Abalone RAG. This model is designed to provide concise, context-aware answers about abalone for use in the RAG application.

These are the hyperparamters that were used, and they were not optimized:

Epochs: 3
Batch Size: 2
Gradient Accumulation Steps: 8
Learning Rate: 2e-4

Downloads last month: 6

Safetensors

Model size

1B params

Tensor type

BF16

LoneWolfgang
/

tinyllama-for-abalone-RAG

TinyLlama for Abalone RAG

Space using LoneWolfgang/tinyllama-for-abalone-RAG 1