Instructions to use jlm429/birdhouse-in-your-soul-lora-demo with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jlm429/birdhouse-in-your-soul-lora-demo with Transformers:

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

pipe = pipeline("text-generation", model="jlm429/birdhouse-in-your-soul-lora-demo")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("jlm429/birdhouse-in-your-soul-lora-demo", device_map="auto")

PEFT
How to use jlm429/birdhouse-in-your-soul-lora-demo with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use jlm429/birdhouse-in-your-soul-lora-demo with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jlm429/birdhouse-in-your-soul-lora-demo"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jlm429/birdhouse-in-your-soul-lora-demo",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jlm429/birdhouse-in-your-soul-lora-demo

SGLang

How to use jlm429/birdhouse-in-your-soul-lora-demo 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 "jlm429/birdhouse-in-your-soul-lora-demo" \
    --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": "jlm429/birdhouse-in-your-soul-lora-demo",
		"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 "jlm429/birdhouse-in-your-soul-lora-demo" \
        --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": "jlm429/birdhouse-in-your-soul-lora-demo",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jlm429/birdhouse-in-your-soul-lora-demo with Docker Model Runner:
```
docker model run hf.co/jlm429/birdhouse-in-your-soul-lora-demo
```

LoRA Fine-Tuning Demo - Bluebird House (Synthetic Structured Data)

Overview

This repository contains a LoRA adapter fine-tuned on a small synthetic dataset representing structured facts about a fictional “Bluebird House.”

The goal of this project is to demonstrate:

LoRA / PEFT fine-tuning workflow
Structured output learning
Controlled before/after behavior on domain-specific data

This is intentionally a minimal, reproducible example.

Base Model

Qwen/Qwen2.5-3B-Instruct
Architecture: decoder-only causal language model

What This Demonstrates

This project highlights how a small LoRA adapter can override base model priors and inject domain-specific knowledge.

Before fine-tuning, the base model produces generic or incorrect answers.
After fine-tuning, the model reproduces structured, dataset-specific facts.

Example

Prompt

Where is the main water shutoff in the Bluebird House?

Before (base model)

The main water shutoff is typically located under the kitchen sink...

After (LoRA)

Location: basement utility room
Details: north wall, next to the red pressure tank

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

base_model = "Qwen/Qwen2.5-3B-Instruct"
adapter = "jlm429/birdhouse-in-your-soul-lora-demo"

tokenizer = AutoTokenizer.from_pretrained(base_model)

model = AutoModelForCausalLM.from_pretrained(
    base_model,
    device_map="auto",
    trust_remote_code=True
)

model = PeftModel.from_pretrained(model, adapter)

Notes

Dataset is synthetic and intentionally small
Model is expected to memorize structured facts
Demonstrates LoRA workflow rather than generalization

Related Resources

Dataset: https://huggingface.co/datasets/jlm429/bluebird-house-synthetic-dataset
Demo (Space): https://jlm429-bluebird-house-demo.hf.space/

Downloads last month: -; Downloads are not tracked for this model. How to track

Model tree for jlm429/birdhouse-in-your-soul-lora-demo

Base model

Qwen/Qwen2.5-3B

Finetuned

Qwen/Qwen2.5-3B-Instruct

Adapter

(1285)

this model

jlm429
/

birdhouse-in-your-soul-lora-demo