Instructions to use Henk717/spring-dragon with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Henk717/spring-dragon with Transformers:

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

pipe = pipeline("text-generation", model="Henk717/spring-dragon")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Henk717/spring-dragon")
model = AutoModelForCausalLM.from_pretrained("Henk717/spring-dragon")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Henk717/spring-dragon with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Henk717/spring-dragon"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Henk717/spring-dragon",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Henk717/spring-dragon

SGLang

How to use Henk717/spring-dragon 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 "Henk717/spring-dragon" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Henk717/spring-dragon",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "Henk717/spring-dragon" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Henk717/spring-dragon",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Henk717/spring-dragon with Docker Model Runner:
```
docker model run hf.co/Henk717/spring-dragon
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

This model is a recreation attempt of the AI Dungeon 2 Dragon model, to achieve this text_adventures.txt was used that was bundled with the original AI Dungeon 2 github release prior to the online service. From what we know the same dataset file was used to create the Dragon model, Dragon being a GPT3 175B Davinci model from 2020.

Since LLaMA1 13B has been benchmarking similarly to the original GPT3 175B the hope is that this recreation is faithful to the original Dragon model. But, since it is not known how close it performs without releasing it to former AI Dungeon players we dubbed it "Spring Dragon" instead of "Summer Dragon", consider it Dragon in its growing up phase.

This model is best used with KoboldAI's adventure mode prefixing your actions with You (2020 AI Dungeon did this automatically) and writing in the second person.

Warning: This model is purposefully flawed and should only be used by people Nostalgic for old 2020 era text adventure models. It is not recommended to be used in model merges, and you can very likely get a much better experience from modern instruct models by asking them to "Start a text adventure game about X"

If the recreation was succesfull expect the following recurring themes:

Names: Alison, Annah, Ben, Big Red, Brutus, Camid, Captain Hayes, Captain Roldan, Castus, Catia, Count Grey, Cyrus, Dendrin, Dr. Gaange (also Mr Gaange), Dr. Gossey, Dr. Kessel, Dr. Kovas, Durge, Elder Flynn, Elios, Elizabeth/Eliza, Fay, Father Féval, Fenrir, Great Lich Lord, Grolik, Isabella, *Jacob, *Karth, Kyros, Lilith, Lord Rostov, Magos Cern, Meliodas, Mistress, Mr. Matasan, Mr. Mol, Mr. Reynolds, Naji, Quintus, Ral, Rolomag, Rose, (Sir) Kit, Talia, Tanya, The Emperor, Ulivik, *Vamp/Vampy, Velzix, Yvette, Zalmora/Zal. ( means the AI likes calling the player these)

Locations: Dert, Fort Defiance, Fort Glory, Hessla, Holgard, Klyton, Kyros, Nyttrus, Rask, Teckleville, The Delantium Kingdom, The Empire of Man (also called Imperium of Man), The Felkan Kingdom

Factions: The Black Rats, Chaos Space Marines, The Crimson Talons, The Dark Order, Dornans (worshippers of Dorna), Ebony Claw Syndicate (often called ECS or The Syndicate), The Empire, Eternals, Joachimites (The Church of Joachim), The Nocturnal League, Psykers, The Shadows, Techpriests, Thieves Guild, Vampire Clan.

Deities: Dorna, Joachim, Nyx, Slaanesh, Virgil, Yag.

Species/Races: Eternals, Goliaths, Oalkwardners, The Craxil, ghouls,kobolds, orks, psykers, svelks, vampires, wendigos, werewolves.

Downloads last month: 14

Model tree for Henk717/spring-dragon

Finetunes

1 model

Quantizations

4 models