Instructions to use p1atdev/dart-v3-pretrain-preview-E with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use p1atdev/dart-v3-pretrain-preview-E with Transformers:

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

pipe = pipeline("text-generation", model="p1atdev/dart-v3-pretrain-preview-E")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("p1atdev/dart-v3-pretrain-preview-E")
model = AutoModelForCausalLM.from_pretrained("p1atdev/dart-v3-pretrain-preview-E")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use p1atdev/dart-v3-pretrain-preview-E with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "p1atdev/dart-v3-pretrain-preview-E"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "p1atdev/dart-v3-pretrain-preview-E",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/p1atdev/dart-v3-pretrain-preview-E

SGLang

How to use p1atdev/dart-v3-pretrain-preview-E 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 "p1atdev/dart-v3-pretrain-preview-E" \
    --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": "p1atdev/dart-v3-pretrain-preview-E",
		"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 "p1atdev/dart-v3-pretrain-preview-E" \
        --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": "p1atdev/dart-v3-pretrain-preview-E",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use p1atdev/dart-v3-pretrain-preview-E with Docker Model Runner:
```
docker model run hf.co/p1atdev/dart-v3-pretrain-preview-E
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

- name: v3 pretrain preview E
  version: v3
  model_name_or_path: p1atdev/dart-v3-pretrain-preview-E
  model_type: eager
  prompt_template_id: V3_PRETRAIN

prompt format:

RATING = Literal["<|rating:general|>", "<|rating:sensitive|>", "<|rating:questionable|>", <|rating:explicit|>]
ASPECT_RATIO = Literal[
  "<|aspect_ratio:too_tall|>",
  "<|aspect_ratio:tall_wallpaper|>",
  "<|aspect_ratio:tall|>",
  "<|aspect_ratio:square|>",
  "<|aspect_ratio:wide|>",
  "<|aspect_ratio:wide_wallpaper|>", 
  "<|aspect_ratio:too_tall|>"
]
LENGTH = Literal["<|length:very_short|>", "<|length:medium|>", "<|length:very_long|>"]

template = (
  "<|bos|>"
  "{rating}{aspect_ratio}{length}"
  "<copyright>{copyright}</copyright>"
  "<character>{character}</character>"
  "<general>{condition}"
)

# for example:
prompt = template.format(
  rating="<|rating:general|>",  # RATING
  aspect_ratio="<|aspect_ratio:tall|>",  # ASPECT_RATIO
  length="<|length:medium|>",  # LENGTH
  copyright="original",
  character="my character",
  condition="1girl", 
)

Downloads last month: -

Safetensors

Model size

0.1B params

Tensor type

BF16

Model tree for p1atdev/dart-v3-pretrain-preview-E

Finetunes

1 model

Collection including p1atdev/dart-v3-pretrain-preview-E

Dart v3 Preview

Collection

6 items • Updated Feb 5, 2025 • 3