Instructions to use Vortex5/Scarlet-Eclipse-12B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Vortex5/Scarlet-Eclipse-12B with Transformers:

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

pipe = pipeline("text-generation", model="Vortex5/Scarlet-Eclipse-12B")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Vortex5/Scarlet-Eclipse-12B")
model = AutoModelForCausalLM.from_pretrained("Vortex5/Scarlet-Eclipse-12B", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Vortex5/Scarlet-Eclipse-12B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Vortex5/Scarlet-Eclipse-12B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Vortex5/Scarlet-Eclipse-12B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Vortex5/Scarlet-Eclipse-12B

SGLang

How to use Vortex5/Scarlet-Eclipse-12B 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 "Vortex5/Scarlet-Eclipse-12B" \
    --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": "Vortex5/Scarlet-Eclipse-12B",
		"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 "Vortex5/Scarlet-Eclipse-12B" \
        --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": "Vortex5/Scarlet-Eclipse-12B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Vortex5/Scarlet-Eclipse-12B with Docker Model Runner:
```
docker model run hf.co/Vortex5/Scarlet-Eclipse-12B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Scarlet Eclipse — 12B

01 // OVERVIEW

Scarlet-Eclipse-12B is a model merged using a custom method — it merges Vortex5/Crimson-Twilight-12B and Vortex5/Harmony-Bird-12B, with Vortex5/LunaMaid-12B as a base.

02 // MERGE CONFIGURATION

Custom Merge Method:

Read full explanation

It is a directional, coherence-aware merge algorithm that moves a base model along the weighted consensus direction defined by its donors rather than averaging them directly. Each donor’s influence is determined by its relative energy (the magnitude of its weight differences from the base), and the method normalizes and scales these offsets to preserve numerical stability.

A small orthogonal adjustment prevents collapse when donors are highly similar, while the strength, trust, and top_k parameters control how far and how selectively the merge travels through parameter space. The result is a controlled shift in model behavior that reflects donor characteristics without discarding the base model’s underlying structure.

Show YAML


merge_method: flowforge
models:
  - model: Vortex5/Crimson-Twilight-12B
  - model: Vortex5/Harmony-Bird-12B
base_model: Vortex5/LunaMaid-12B
parameters:
  strength: 0.9
  trust: 1.0
dtype: bfloat16
tokenizer:
  source: Vortex5/LunaMaid-12B