Instructions to use FredZhang7/danbooru-tag-generator with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FredZhang7/danbooru-tag-generator with Transformers:

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

pipe = pipeline("text-generation", model="FredZhang7/danbooru-tag-generator")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("FredZhang7/danbooru-tag-generator")
model = AutoModelForCausalLM.from_pretrained("FredZhang7/danbooru-tag-generator", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use FredZhang7/danbooru-tag-generator with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FredZhang7/danbooru-tag-generator"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FredZhang7/danbooru-tag-generator",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/FredZhang7/danbooru-tag-generator

SGLang

How to use FredZhang7/danbooru-tag-generator 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 "FredZhang7/danbooru-tag-generator" \
    --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": "FredZhang7/danbooru-tag-generator",
		"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 "FredZhang7/danbooru-tag-generator" \
        --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": "FredZhang7/danbooru-tag-generator",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use FredZhang7/danbooru-tag-generator with Docker Model Runner:
```
docker model run hf.co/FredZhang7/danbooru-tag-generator
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Disclaimer

Danbooru stores millions of tagged anime images, but it doesn't have a way to filter out NSFW content. This model was trained on 100,000 of these tags with up_score ≥ 3 for 3 epochs, so it's possible that some tags might contain NSFW descriptions. So, just be mindful of that. Thank you for your cooperation.

The Safe Version

For details on data preprocessing, prompt engineering, and more, please see Fast Anime PromptGen. I used a very similar approach to train the Danbooru version.

Downloads last month: 411

Model tree for FredZhang7/danbooru-tag-generator

Quantizations

2 models

FredZhang7
/

danbooru-tag-generator

Disclaimer

The Safe Version

Model tree for FredZhang7/danbooru-tag-generator

Dataset used to train FredZhang7/danbooru-tag-generator